statistical analysis plan amendment 5...ucb 07 sep 2018 statistical analysis plan bimekizumab pa0008...
TRANSCRIPT
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 1 of 183
STATISTICAL ANALYSIS PLAN AMENDMENT 5
Study: PA0008Product: Bimekizumab
A MULTICENTER, PHASE 2B, RANDOMIZED, DOUBLE-BLIND, PLACEBO-CONTROLLED,
PARALLEL-GROUP, DOSE-RANGING STUDY TO EVALUATE THE EFFICACY AND SAFETY OF BIMEKIZUMAB IN ACTIVE
PSORIATIC ARTHRITIS
SAP/Amendment Number DateFinal SAP 12 May 2017SAP Amendment 1 11 Aug 2017SAP Amendment 2 7 Nov 2017SAP Amendment 3 02 Mar 2018SAP Amendment 4 09 Aug 2018SAP Amendment 5 07 Sep 2018
Confidentiality Statement
ConfidentialThis document is the property of UCB and may not – in full or in part – be passed on,
reproduced, published, or otherwise used without the express permission of UCB.
REDACTED COPYTHRTHR
May 2May 2
This do
cumen
t can
not b
e use
d to su
pport
any m
arketi
ng au
thoriz
ation
appli
catio
n and
any e
xtens
ions o
r vari
ation
s the
reof.
ZED, ZED, OLLEDLLED
Y TO Y TOIZUMIZUM
ITISTIS
20172017ugug 20120
Novov 20202 Ma02 M0099
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 2 of 183
TABLE OF CONTENTSLIST OF ABBREVIATIONS.................................................................................................... 7 1 INTRODUCTION ........................................................................................................... 13 2 PROTOCOL SUMMARY............................................................................................... 13 2.1 Study objectives ........................................................................................................... 13
2.1.1 Primary objective................................................................................................. 13 2.1.2 Secondary objectives ........................................................................................... 13 2.1.3 Other objectives ................................................................................................... 14
2.2 Study variables............................................................................................................. 14 2.2.1 Efficacy variables ................................................................................................ 14
2.2.1.1 Primary efficacy variable .......................................................................... 14 2.2.1.2 Secondary efficacy variable(s) .................................................................. 14 2.2.1.3 Other efficacy variables............................................................................. 14
2.2.2 Pharmacokinetic/pharmacodynamic variables .................................................... 15 2.2.2.1 Pharmacokinetic variable .......................................................................... 15 2.2.2.2 Pharmacodynamic variables...................................................................... 16
2.2.3 Pharmacogenomic variables ................................................................................ 16 2.2.4 Immunological variables ..................................................................................... 16 2.2.5 Safety variables.................................................................................................... 16
2.3 Study design and conduct ............................................................................................ 16 2.4 Determination of sample size....................................................................................... 25 3 DATA ANALYSIS CONSIDERATIONS ...................................................................... 25 3.1 General presentation of summaries and analyses ........................................................ 25 3.2 General study level definitions .................................................................................... 27
3.2.1 Relative day ......................................................................................................... 27 3.2.2 Study Periods ....................................................................................................... 27
3.3 Definition of Baseline values....................................................................................... 27 3.4 Protocol deviations....................................................................................................... 28 3.5 Mapping of assessments performed at early termination visit..................................... 28 3.6 Analysis sets................................................................................................................. 28
3.6.1 Enrolled Set (ES) ................................................................................................. 29 3.6.2 Randomized Set (RS) .......................................................................................... 29 3.6.3 Safety Set (SS) ..................................................................................................... 29 3.6.4 Full Analysis Set (FAS)....................................................................................... 29 3.6.5 Per-Protocol Set (PPS)......................................................................................... 29 3.6.6 Pharmacokinetics Per-Protocol Set (PK-PPS)..................................................... 29 3.6.7 Pharmacodynamics Per-Protocol Set (PD-PPS).................................................. 29 3.6.8 Dose-Blind Set (DBS) ......................................................................................... 29
REDACTED .................
..........ee..........
COPY ...............
This do
cumen
t3.63.ca
nnot 1
.6.26.2
bep
alysialysus
ed devev
ing ofng ofto B
viatviatsu
pport
........odsds ....
BaselBasel
anynitioniti
.
marketi
ng ..........
ATIONATIOmmariemmaronon
autho
rizati
on .....
....................
..........
appli
catio
n ...................
.................
and .......... an
y ...........ex
tensio
ns
....................
.........
or ....... varia
tions
. 13131
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 3 of 183
3.6.9 Escape Subject Set (ESS) .................................................................................... 29 3.6.10 Dose-Blind Responder Set (DBRS)..................................................................... 30 3.6.11 Corrected Escape Subject Set (cESS).................................................................. 30 3.6.12 Corrected Dose-Blind Responder Set (cDBRS) .................................................. 30
3.7 Treatment assignment and treatment groups ............................................................... 30 3.8 Center pooling strategy................................................................................................ 31 3.9 Coding dictionaries ...................................................................................................... 31 3.10 Changes to protocol-defined analyses ......................................................................... 31 4 STATISTICAL/ANALYTICAL ISSUES ....................................................................... 31 4.1 Adjustments for covariates .......................................................................................... 31 4.2 Handling of dropouts or missing data.......................................................................... 31
4.2.1 Handling of missing data for efficacy analysis.................................................... 31 4.2.2 Handling of missing data for AE ......................................................................... 34 4.2.3 Handling of missing data for prior and concomitant medication ........................ 35
4.3 Interim analysis and data monitoring........................................................................... 36 4.3.1 Interim analysis Week 12 .................................................................................... 36
4.3.1.1 Changes from interim analysis to SAP-defined analyses.......................... 38 4.3.1.2 Search and selection criteria for AE of special monitoring for interim
analysis ...................................................................................................... 38 4.3.2 Interim analysis Week 48 .................................................................................... 40
4.3.2.1 Changes from interim analysis to SAP-defined analyses.......................... 40 4.3.2.2 Changes during interim analysis to SAP-defined analyses ....................... 41
4.4 Multicenter studies....................................................................................................... 41 4.5 Multiple comparisons/multiplicity............................................................................... 41 4.6 Use of an efficacy subset of subjects ........................................................................... 41 4.7 Active-control studies intended to show equivalence.................................................. 41 4.8 Examination of subgroups ........................................................................................... 41 5 STUDY POPULATION CHARACTERISTICS............................................................. 42 5.1 Subject disposition ....................................................................................................... 42 5.2 Protocol deviations....................................................................................................... 42 6 DEMOGRAPHICS AND OTHER BASELINE CHARACTERISTICS ........................ 43 6.1 Demographics .............................................................................................................. 43 6.2 Other Baseline characteristics...................................................................................... 43 6.3 Medical history and concomitant diseases................................................................... 45 6.4 Prior and concomitant medications.............................................................................. 45 6.5 Prohibited medication and rescue medication ............................................................. 45 7 MEASUREMENTS OF TREATMENT COMPLIANCE............................................... 45 8 EFFICACY ANALYSES ................................................................................................ 46
REDACTED fo...........
..........rim anrim a
nternter
COPY o SAPo SA
for Aor A
This do
cumen
t .226.36.3
cann
otMOMODeDe
be ocooco
OGOGus
edt dispodispool dol
toULAULAsupp
ortudies indies iof subgf subgATAT
any set ofset omark
eting
rim anim an
..........ltiplictipli
autho
rizati
onAE of E of ............
............alysialysi
appli
catio
n ed
........................
P--defindefi
and .......
dicaticatan
y ...........
exten
sions
...................
..........
or ...... va
riatio
ns
311313
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 4 of 183
8.1 Statistical analysis of the primary efficacy variable .................................................... 46 8.1.1 Derivations of primary ACR score and response ................................................ 46
8.1.1.1 Tender joint count (TJC) and swollen joint count (SJC)........................... 47 8.1.1.2 Patient’s Global Assessment of Disease Activity (PGADA) .................... 48 8.1.1.3 Physician’s Global Assessment of Disease Activity (PhGADA) ............. 48 8.1.1.4 Patient’s Assessment of Arthritis Pain (PtAAP) ....................................... 48 8.1.1.5 Health Assessment Questionnaire-Disability Index (HAQ-DI) score....... 48 8.1.1.6 Body surface area (BSA)-Psoriaris ........................................................... 49
8.1.2 Primary analysis of the primary efficacy variable ............................................... 49 8.1.3 Secondary analyses of the primary efficacy variable .......................................... 50 8.1.4 Supportive analyses of the primary efficacy variable.......................................... 51 8.1.5 Sensitivity analyses of the primary efficacy variable .......................................... 51
8.2 Statistical analysis of the secondary efficacy variables ............................................... 52 8.2.1 Derivation of Psoriasis Area and Severity Index (PASI), PASI75, PASI90, and
PASI100............................................................................................................... 52 8.3 Analysis of other efficacy variables............................................................................. 54
8.3.1 Minimal disease activity (MDA) ......................................................................... 55 8.3.1.1 Modified MDA.......................................................................................... 56
8.3.2 Disease Activity Score-28 joint count C-reactive protein (DAS28[CRP]) ......... 56 8.3.3 Bath Ankylosing Spondylitis Disease Activity Index (BASDAI)....................... 56 8.3.4 Modified Nail Psoriasis Severity Index (mNAPSI)............................................. 57 8.3.5 Maastricht Ankylosing Spondylitis Enthesitis (MASES) Index ......................... 57 8.3.6 Leeds Dactylitis Index (LDI)............................................................................... 57 8.3.7 Psoriatic Arthritis Impact of Disease-9 (PsAID-9).............................................. 58 8.3.8 Psoriatic Arthritis Quality of Life (PsAQoL) ...................................................... 59 8.3.9 Short Form – 36 Items Health Survey ................................................................. 59 8.3.10 Hospital Anxiety and Depression Scale (HADS) ................................................ 59 8.3.11 C-reactive protein (CRP) and high-sensitivity C-reactive protein (CRP) ........... 60
8.4 Subgroup analysis ........................................................................................................ 60 9 PHARMACOKINETICS AND PHARMACODYNAMICS .......................................... 60 9.1 Pharmacokinetics ......................................................................................................... 60 9.2 Pharmacodynamics and Immunogenicity .................................................................... 61 10 SAFETY ANALYSES..................................................................................................... 62 10.1 Extent of exposure ....................................................................................................... 62 10.2 Adverse events ............................................................................................................. 65
10.2.1 Exposure duration................................................................................................ 67 10.2.2 Exposure adjusted incidence rate (EAIR) and exposure adjusted event rate
(EAER) ................................................................................................................ 67 10.2.3 AE of Special Interest and AE of Special Monitoring ........................................ 68
REDACTEDcountcouns Dises Dise
SeveriSeverSpoSpo
COPY ....................
This do
cumen
t 0010.10
cann
otPharharPhPh
SS
be MAMA
rmrmus
edup anup anACAC
totivetivesupp
ort36 It36 IAnxietAnxiee p
any p
s Quas Quamark
eting
ondylitndyli
(LDI)LDI)pact opact
autho
rizati
on ... ........
t CC--rereease Aease A
y Indy Ind
appli
catio
n I),)
.........................
...........
and .......
, PAPAan
y ...........
exten
sions
...................
..........
or ...... va
riatio
ns
4848494
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 5 of 183
10.3 Clinical laboratory evaluations .................................................................................... 70 10.4 Potential drug-induced liver injury (PDILI) assessment ............................................. 72 10.5 Vital signs, physical findings, and other observations related to safety ...................... 75
10.5.1 Vital signs ............................................................................................................ 75 10.5.2 Electrocardiograms .............................................................................................. 75 10.5.3 Other safety variables .......................................................................................... 76
Physical Examination .................................................................................................... 76 Assessment of Tuberculosis........................................................................................... 76 Electronic Columbia Suicide Severity Rating Scale (eC-SSRS)................................... 76
11 REFERENCES ................................................................................................................ 78 12 APPENDICES ................................................................................................................. 80 12.1 Classification Criteria for Psoriatic Arthritis (CASPAR) Criteria............................... 80 12.2 Calculation rules for duration of adverse events.......................................................... 80 12.3 Treatment group assignment for tables and figures ..................................................... 81 12.4 Hematology and biochemistry variables for the interim analysis................................ 83 12.5 Health Assessment Questionnaire-Disability Index (HAQ-DI) .................................. 84 12.6 Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) ............................... 87 12.7 Maastricht Ankylosing Spondylitis Enthesitis Index (MASES) questionnaire ........... 88 12.8 Psoriatic Arthritis Impact of Disease-9 (PsAID-9)...................................................... 89 12.9 Psoriatic Arthritis Quality of Life (PsAQoL) .............................................................. 91 12.10 Short-Form (SF-36) Questionnaire .............................................................................. 95 12.11 Classification of the SF-36 questionnaire .................................................................. 100 12.12 Tuberculosis Questionnaire ....................................................................................... 102 12.13 Identification of Opportunistic infections.................................................................. 102 12.14 MedDRA algorithmic approach to anaphylaxis ........................................................ 103 13 AMENDMENT(S) TO THE STATISTICAL ANALYSIS PLAN ............................... 106 13.1 Amendment 1............................................................................................................. 106
13.1.1 Rationale for the amendment............................................................................. 106 13.1.2 Modifications and changes ................................................................................ 106
13.2 Amendment 2............................................................................................................. 113 13.2.1 Rationale for the amendment............................................................................. 113 13.2.2 Modifications and changes ................................................................................ 113
13.3 Amendment 3............................................................................................................. 124 13.3.1 Rationale for the amendment............................................................................. 124 13.3.2 Modification and changes.................................................................................. 124
13.4 Amendment 4............................................................................................................. 147 13.4.1 Rationale for the amendment............................................................................. 147 13.4.2 Modification and changes.................................................................................. 147
REDACTED he9 (Ps9 (Ps
e (PsA(PsAnnairennaireuesue
COPY dvity Indvity In
esitis sitis
This do
cumen
t 13.33.3 cann
ot 13.2.2.2.2
beendnd
RR
used
nnModificodific
m
tonale ale su
pport
c TO THTO T
............f
any unin
apprapprmark
eting
..stionnationna
............istic isti
autho
rizati
ons IndexIndesAIDAID
AQoL)QoL)....
appli
catio
n ..
analysianalyHAQHAQ D
dex (Bex (B
and .......
.....an
y ...........
exten
sions
...................
..........
or ...... va
riatio
ns
7676767
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 6 of 183
13.5 Amendment 5............................................................................................................. 169 13.5.1 Rationale for the amendment............................................................................. 169 13.5.2 Modification and changes.................................................................................. 169
STATISTICAL ANALYSIS PLAN SIGNATURE PAGE................................................... 183
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y exte
nsion
s or v
ariati
ons t
hereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 7 of 183
LIST OF ABBREVIATIONS
AbAb anti-bimekizumab antibody
ACR American College of Rheumatology
ACR20,50,70 American College of Rheumatology 20, 50, 70% response criteria
ACP above the cut point
ADR adverse drug reaction
AE adverse event
AJ assessed joints
ALP alkaline phosphatase
ANCOVA analysis of covariance
ALT alanine aminotransferase
AST aspartate aminotransferase
ATC anatomical therapeutic chemical
BASDAI Bath Ankylosing Spondylitis Disease Activity Index
BCP below the cut point
BKZ bimekizumab
BLQ below the level of quantification
BMI body mass index
BSA body surface area
BUN blood urea nitrogen
CASPAR Classification Criteria for Psoriatic Arthritis
CCP cyclic citrullinated peptide
cDBRS corrected Dose-Blind Responder Set
CDF cumulative distribution function
REDACTED aminoamin
tomictomic
athat
COPY ransfeansf
This do
cumen
tCACA cann
ot
ASAS
be us
ed to
supp
ort an
y mark
eting
h Ankyh Anky
belowbelo
autho
rizati
on erasra
notransotran
al theal the
appli
catio
n
asese
and a
ny ex
tensio
ns or
varia
tions
there
of.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 8 of 183
CDISC Clinical Data Interchange Standards Consortium
cESS corrected Escape Subject Set
CI confidence interval
CL/F total body clearance
CP confirmed positive
CPK creatine phosphokinase
CRP C-reactive protein
CV coefficient of variance
DAP data analysis plan
DAS28(CRP) Disease Activity Score-28 joint count C-reactive protein
DBS Dose-Blind Set
DBRS Dose-Blind Responder Set
DEM data evaluation meeting
DMC Data Monitoring Committee
ECG electrocardiogram
eCRF electronic case report form
eC-SSRS electronic Columbia-Suicide Severity Rating Scale
EAER exposure adjusted event rate
AEIR exposure adjusted incidence rate
ES Enrolled Set
ESS Escape Subject Set
ET early termination
FAS Full Analysis Set
GGT gamma glutamyltransferase
ind Rind R
a a evaleva
COPY etet
This do
cumen
tESSESSca
nnot
be us
ed to
supp
ort an
y mark
eting
ata Monata Mo
elecele
autho
rizati
on
Responespon
uatiouatio
appli
catio
n joint coint c
and a
ny ex
tensio
ns or
varia
tions
there
of.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 9 of 183
H high
HADS Hospital Anxiety and Depression Scale
HADS-A Hospital Anxiety and Depression Scale—Anxiety
HADS-D Hospital Anxiety and Depression Scale—Depression
HAQ-DI Health Assessment Questionnaire – Disability Index
HBcAb-IgM Hepatitis B core antibody-IgM
HBsAg Hepatitis B surface antigen
HIV human immunodeficiency virus
HLGT High-Level Group Term
HLT High Level Term
HRQoL Health-Related Quality of Life
hs-CRP high sensitivity C-reactive protein
INR international normalized ratio
ITT Intent-to-treat
IMP investigational medicinal product
L low
LCL lower confidence limit
LD loading dose
LDH lactate dehydrogenase
LDI Leeds Dactylitis Index
LEF leflunomide
IgM immunoglobulin M
LLN lower limit of normal
LLOQ lower limit of quantification
REDACTED nsitivitnsitivit
ernatiernat
COPY dd QuaQua
This do
cumen
tLEFLEFca
nnot
be us
ed to
supp
ort an
y mark
eting
entnt--toto
inveinv
autho
rizati
onality ality
ty Cy C--r
onal nnal n
appli
catio
n
of
and a
ny ex
tensio
ns x or
varia
tions
there
of.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 10 of 183
LOCF last observation carried forward
MAR missing at random
MASES Maastricht Ankylosing Spondylitis Enthesitis Index
MCH mean corpuscular hemoglobin;
MCHC mean corpuscular hemoglobin concentration
MCMC Markov-Chain Monte Carlo
MCS mental component summary
MCV mean corpuscular volume
MDA Minimal Disease Activity
MedDRA Medical Dictionary for Regulatory Activities
MI multiple imputation
MTX methotrexate
mNAPSI modified Nail Psoriasis Severity Index
N number of subjects
n number of observations
N/A not applicable
NCP not confirmed positive
NRI non-responder imputation
NRS numerical rating scale
NSAID nonsteroidal anti-inflammatory drug
OR odds ratio
PASI Psoriasis Area and Severity Index
PASI75, PASI90, PASI100 Psoriasis Area and Severity Index 75%, 90%, 100%
PCS physical component summary
REDACTED tete
fied Nied N
umbeumb
COPY tionion
This do
cumen
t PAP
cann
ot
RR
be us
ed to
supp
ort an
y nmark
etinger of er of
numbeumb
autho
rizati
on
Nail Psail Ps
ff
appli
catio
n gulatoryulator
and a
ny ex
tensio
ns or
varia
tions
there
of.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 11 of 183
PD pharmacodynamic
PDILI potential drug-induced liver injury
PD-PPS Pharmacodynamics Per-Protocol Set
PGADA Patient’s Global Assessment of Disease Activity
PhGADA Physician’s Global Assessment of Disease Activity
PK pharmacokinetic
PK-PPS Pharmacokinetic Per-Protocol Set
PPS Per-Protocol Set
PsA psoriatic arthritis
PsAID-9 Psoriatic Arthritis Impact of Disease-9
PsAQoL Psoriatic Arthritis Quality of Life
PT preferred term
PtAAP Patient’s Assessment of Arthritis Pain
Q4W every 4 weeks (monthly)
QTcB QT interval corrected for heart rate using Bazett’s formula
QTcF QT interval corrected for heart rate using Fridericia’s formula
RBC red blood cell
RCTC Rheumatology Common Toxicity Criteria
RNA ribonucleic acid
RS Randomized Set
SAE serious adverse event
SAP statistical analysis plan
SC subcutaneous(ly)
REDACTED ermerm
t’s Ast’s A
very very
COPYis Quais Qu
This do
cumen
t RS
SS
cann
ot
S
be us
ed to
supp
ort an
y mark
eting4 we4 we
QT intQT informfor
autho
rizati
on
ssessmssessm
appli
catio
n f DiseaDise
ality oflity o
and a
ny ex
tensio
ns or
varia
tions
there
of.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 12 of 183
SD standard deviation
SE standard error
SF-36 Short-Form 36-item Health Survey
SFU Safety Follow-up
SJC swollen joint count
SMQ Standard MedDRA query
SOC system organ class
SS Safety Set
TEAE treatment-emergent adverse event
TJC tender joint count
TFLs tables, figures, and listings
TNF tumor necrosis factor
TNFα tumor necrosis factor-alpha
TB tuberculosis
UCL upper confidence limit
ULN upper limit of normal
V/F volume of distribution
VAS visual analog scale
WBC white blood cell
WHO-DD World Health Organization Drug Dictionary
Z critical value from the standard normal distribution
REDACTED osis osis
necronecro
ubercuberc
COPYnd lisnd lis
s fas fa
This do
cumen
t can
not DDDD be
used
to su
pport
any um
arketi
ng ulosiulos
upper cpper
autho
rizati
on
actorctor
osis faosis fa
ii
appli
catio
n
tingsings
and t
any e
xtens
ions o
r vari
ation
s the
reof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 13 of 183
1 INTRODUCTIONThe purpose of this statistical analysis plan (SAP) is to provide all necessary information to perform the required interim and final statistical analysis for study PA0008. It also defines the summary tables, figures, and listings (TFLs) to be generated in the clinical study report according to the final Protocol (29 Jul 2016), Protocol Amendment 1 (16 Dec 2016), and Protocol Amendment 2 (09 Mar 2018).
The content of this SAP is compatible with the International Conference on Harmonization/ Food and Drug Administration E9 Guidance documents (1998).
2 PROTOCOL SUMMARYThis is a Phase 2b, multicenter, randomized, double-blind, placebo-controlled, parallel-group, dose-ranging study to investigate the efficacy, safety, pharmacokinetics (PK), and pharmacodynamics (PD) of bimekizumab (also known as UCB4940) compared with placebo in adult subjects with active psoriatic arthritis (PsA) in order to guide the selection of doses and clinical indices in the Phase 3 development program.
The study population will consist of adult subjects (≥18 years of age) fulfilling the Classification Criteria for Psoriatic Arthritis (CASPAR, Table 12‒1) criteria and having active disease with tender joint count (TJC) ≥3 out of 78 and swollen joint count (SJC) ≥3 out of 76. Subjects must be rheumatoid factor and anti-cyclic citrullinated peptide (CCP) antibodies negative and have active psoriatic lesion(s) and/or a documented history of psoriasis. Subjects may be tumor necrosis factor (TNF) inhibitor naïve or may have received 1 prior TNF inhibitor.
Subjects who have been on a TNF inhibitor previously must have:
experienced an inadequate response to previous treatment given for at least 3 months
been intolerant to administration (eg, had a side-effect/adverse event (AE) that led to discontinuation)
lost access to TNF inhibitor for other reasons
An estimated 70 sites in Europe and North America will randomize 200 subjects. Enrollment ofTNF inhibitor experienced subjects will be limited to 30% of the total study population.
The study consists of a Screening Period (14 to 28 days), Double-Blind Period (12 weeks), Dose-Blind Period (36 weeks) and Safety Follow-up Period (20 weeks after last dose; only for subjects who do not enter the extension study). Therefore, the maximum duration of the study is 68weeks.
2.1 Study objectives2.1.1 Primary objectiveThe primary objective is to assess the dose-response based on the efficacy of bimekizumab administered subcutaneously (sc) every 4 weeks (Q4W) for 12 weeks in the treatment of subjects with active PsA.
2.1.2 Secondary objectivesThe secondary objectives of the study are as follows:
REDACTEDhis
y havey have
itor prtor p
nse tonse to
COPYnt cnt pepeptidti
storystory
This do
cumen
t .1.1.ThTh
cann
ot
11
be en us
ed6 wewenter tter
toof aof asupp
ortEuropeuropnced suced s
a SS
anyr for r formark
eting
o preo pr
(eg, ha(eg, h
autho
rizati
on de
y of py of pe receirecei
revioueviou
revrev
appli
catio
nf age) f age)a and ha and
unt (SJnt (SJ(CCP(CCP
and
se anyparear
lectiolectiex
tensio
ns
paralleralle, and and
reded
or va
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 14 of 183
To assess the efficacy of the individual dose regimens of bimekizumab compared to placebo
To assess skin and nail psoriasis in the subgroup of affected subjects at Baseline
To assess the safety and tolerability of bimekizumab
To assess the PK of bimekizumab
To assess the PD of bimekizumab
To assess the immunogenicity of bimekizumab
To assess the exposure:response relationship of bimekizumab as it relates to efficacy and safety
2.1.3 Other objectivesOther objectives of the study are as follows:
To assess the impact on patient-reported quality of life
To assess the impact of bimekizumab treatment on axial disease
To assess the impact of bimekizumab on dactylitis and enthesitis
To assess the impact of administration of bimekizumab on biological pathways relating to disease biology, progression, and response to therapy via biomarker analysis and to enable genomic and related approaches for analysis of subject samples and evaluation of the potential for subject stratification approaches
2.2 Study variables2.2.1 Efficacy variables2.2.1.1 Primary efficacy variableThe primary efficacy variable for this study is as follows:
American College of Rheumatology 50% (ACR50) response at Week 12
2.2.1.2 Secondary efficacy variable(s)The secondary efficacy variables for this study are as follows:
American College of Rheumatology 20% (ACR20) response at Week 12
American College of Rheumatology 70% (ACR70) response at Week 12
Psoriasis Area and Severity Index 90% (PASI90) response at Week 12 in the subgroup of subjects with psoriasis involving at least 3% of body surface area (BSA) at Baseline/Day 1
Psoriasis Area and Severity Index 75% (PASI75) response at Week 12 in the subgroup of subjects with psoriasis involving at least 3% of BSA at Baseline/Day 1
2.2.1.3 Other efficacy variablesOther efficacy variables will be assessed as specified in Table 2‒1:
Time to ACR20 and ACR50 response
REDACTED is ois oachesaches
COPYzumzumtherapthera
of sof
This do
cumen
tPPcann
otoriasiriasisubjecubje
becan Can Cus
edCollegolleg
CoCo
to acyacysu
pport
heueu
ary efary ecy vay v
any
umatumatmark
eting
ariablariab
this sthis
autho
rizati
onpy vy v
subjecubjecap
plica
tion
se
hesitishesiti
b on bib on bvia via
and a
ny ex
tensio
ns cyy ory anan
varia
tions
there
of.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 15 of 183
Psoriasis Area and Severity Index 100% (PASI100) response in the subgroup of subjects with psoriasis involving at least 3% of BSA at Baseline/Day 1
ACR20, ACR50, and ACR70 response
Composite endpoint comprised of ACR50 and PASI90 response in the subgroup of subjects with psoriasis involving at least 3% of BSA at Baseline/Day 1
Minimal Disease Activity (MDA) and modified MDA
Change from Baseline in the Disease Activity Score-28 based on high sensitivity C-reactive protein (DAS28 [CRP])
Change from Baseline in all individual American College of Rheumatology (ACR) score components:
SJC
TJC
Health Assessment Questionnaire—Disability Index (HAQ-DI)
Patient’s Assessment of Arthritis Pain (PtAAP)
Physician’s Global Assessment of Disease Activity (PhGADA)
Patient’s Global Assessment of Disease Activity (PGADA)
high sensitivity C-reactive protein (hs-CRP)
Change from Baseline in the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI)
Change from Baseline in the modified Nail Psoriasis Severity Index (mNAPSI)
Change from Baseline in the Maastricht Ankylosing Spondylitis Enthesitis (MASES) Index
Change from Baseline in the Leeds Dactylitis Index (LDI)
Psoriatic Arthritis Impact of Disease (PsAID)-9
Psoriatic Arthritis Quality of Life (PsAQoL)
Change from Baseline in Physical Component Summary (PCS) and Mental Component Summary (MCS) of the SF-36 (Short-Form 36-Item Health Survey)
Change from Baseline in Hospital Anxiety and Depression Scale (HADS) HADS-Anxiety (A) and HADS-Depression (D) scores
Percentage of subjects with scores below 8 in HADS-A and HADS-D (subjects with normal scores)
2.2.2 Pharmacokinetic/pharmacodynamic variables2.2.2.1 Pharmacokinetic variableThe PK variable is plasma concentration of bimekizumab. The assessment schedule for plasma concentration is available in Table 2‒1.
REDACTED hshs--CRCR
AnkyloAnkylo
difiedifie
COPYActiviActiv
ActivActiv
This do
cumen
ts2.22.2
cann
otand nd
Percenercenscosco
befrofro
d HAd Hus
edMCSMCS
omom
toaselase))
supp
ortct oct o
Qualityualit
linlin
any LeL
of Dof D
marketi
nged Naed Na
aastrichastric
LeedLeed
autho
rizati
onvity (Pity (P
RPP))
osing sing
i
appli
catio
n Q-DD
ty (PhGy (Ph
G
and
DI)DI)
any e
xtens
ions
ACRCR))or
varia
tions
activctiv
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 16 of 183
2.2.2.2 Pharmacodynamic variablesThe PD variables are concentrations of cytokines of relevance to IL 17A/F signaling pathway and PsA biology, and include but are not limited to IL-17A, IL-17F, IL-23, IL-6, and tumor necrosis factor-alpha (TNFα).
2.2.3 Pharmacogenomic variablesWhere local regulations permit, additional blood samples will be collected from consenting subjects at specific time points and stored at -80°C for up to 20 years to allow for potential, exploratory analyses of genomic, genetic, proteomic, and metabolite biomarkers relevant to disease biology and progression, response to therapy, and the inflammatory and immune response processes.
2.2.4 Immunological variablesImmunological variables allow evaluation of immunogenicity as well as immunological biomarkers.
Anti-bimekizumab antibody (AbAb) detection prior to and following study treatment
Serum complement concentrations.
Flow cytometry analysis of key immune cell populations, including but not limited to CD3, CD19, CD4, CD8, and CD69 (using fluorescence activated cell sorting).
Cytokines and other exploratory markers
2.2.5 Safety variablesSafety variables to be assessed as specified in Section 10 are as follows:
Incidence of AEs and serious adverse events (SAEs)
Withdrawal due to AEs
Change from Baseline in vital signs (blood pressure and pulse rate) and body weight
Standard 12-lead electrocardiogram (ECG) intervals (RR, PR, QRS, QT, and QT intervals corrected for heart rate using Bazett’s and Fridericia’s formulas [QTcB and QTcF]), including changes from Baseline
Change from Baseline in clinical laboratory variables (hematology, biochemistry, and urinalysis)
2.3 Study design and conductTo be eligible to participate in this study, subjects must meet the following key inclusion criteria:
Subject has a documented diagnosis of adult-onset PsA classified by CASPAR criteria with symptoms for at least 6 months prior to Screening with active PsA and must have at Baseline TJC ≥3 out of 78 and SJC ≥3 out of 76 (dactylitis of a digit counts as 1 joint each).
Subject must be rheumatoid factor and anti-CCP antibodies negative.
Subject must have active psoriatic lesion(s) and/or a documented history of psoriasis.
REDACTED
cifiedcified
verve
COPYulatiulatnce acnce a
This do
cumen
tSScann
ot be elige elig
be s)) us
edm Bam Ba)
to ges es su
pporttrocarrocar
rate usrate us fromfro
anyital sital
d
marketi
ng d i
rse evese eve
i
autho
rizati
on ctiv
in n SeSe
appli
catio
n ow
ons, incns, inivatevate
and
wingwingan
ymmumm exten
sions
une une
munun
or tot va
riatio
ns
,oo
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 17 of 183
Subjects may be TNF inhibitor naïve or may have received 1 prior TNF inhibitor. Subjects who have been on a TNF inhibitor previously must have:
experienced an inadequate response to previous treatment given for at least 3 months
been intolerant to administration (eg, had a side-effect/AE that led to discontinuation)
lost access to TNF inhibitor for other reasons
Two hundred subjects will be randomized to 1 of 5 treatment arms. Subjects in any treatment group who complete the 12-week Double-Blind Period will enter the 36-week Dose-Blind Period. Treatment during the Dose-Blind Period will start at Week 12 and be administered Q4W thereafter through Week 44.
Screening Period/ Baseline
During the Screening Period, the Investigator will obtain laboratory data and verify that the doses of methotrexate (MTX), leflunomide (LEF), nonsteroidal anti-inflammatory drugs (NSAIDs), or corticosteroids, if used, are stable. The Screening Period will also enable washout of medications not permitted for use during the study. The Screening Period will last 14 to 28 days.
Double-Blind Period
During the Double-Blind Period, subjects will be randomized 1:1:1:1:1 (stratified by region andprior TNF inhibitor exposure) to receive the following blinded study treatment regimens:
Placebo
Bimekizumab 16mg administered sc Q4W
Bimekizumab 160mg administered sc Q4W
Bimekizumab 320mg administered sc Q4W
Bimekizumab 320mg loading dose followed by 160mg administered sc starting at Week 4and Q4W thereafter
Enrollment of TNF inhibitor experienced subjects will be limited to approximately 30% of thetotal study population.
Investigational medicinal product (IMP) will be administered sc in the clinic at Baseline, Week 4, and Week 8. Additional non-dosing study visits in the Double-Blind Period will occur at Week 1 and Week 2. At Week 12, subjects will transition from the Double-Blind Period into the 36-week Dose-Blind Period. Note that the other assessments are planned to be performed before the IMP administration and will be analyzed for the Double-Blind Period.
Dose-Blind Period
The first IMP for the Dose-Blind Period will be administered at Week 12. At the Week 12 Visit, subjects will be allocated to bimekizumab treatment regimens as follows:
Subjects in the placebo group will be rerandomized 1:1 to bimekizumab 160mg or bimekizumab 320mg Q4W
REDACTED llowllo
sc Q4sc Q4
redred
COPY randorand
owinwi
This do
cumen
tDoseosentThT
cann
otk DoDoMP adMP a
e
bed Wd Wosos
used
medimedi8. AddAd
to dicidicisu
pport
ibitor ebitor
n.
anyng dng dm
arketi
ng
sc Q4Wsc Q4
ered scred s
dosd
autho
rizati
on omimi
ng bling bli
WW
appli
catio
n ing ng
mizedized
andod wid wi
g PePe
any
d d ammamma
llll
exten
sions
er
verver
oreded Qva
riatio
ns
nt
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 18 of 183
Subjects in the bimekizumab 16mg dose group will be rerandomized 1:1 to bimekizumab 160mg or bimekizumab 320mg Q4W
Subjects in the bimekizumab 160mg dose group will continue to receive bimekizumab 160mg Q4W
Subjects in the bimekizumab 320mg dose group will continue to receive bimekizumab 320mg Q4W
Subjects in the bimekizumab 320mg (loading)/160mg dose group will continue to receive bimekizumab 160mg Q4W
All assessments for the Dose-Blind Period will be done from Week 18 to Week 48. The last IMP administration is on Week 44.
The total duration of treatment is 36 weeks.
During the 36-week Dose-Blind Period, subjects will be evaluated for inadequate response totreatment at defined time points. Subjects who do not show an improvement in tender andswollen joint count will be eligible to receive rescue therapy as defined in Section 6.5.
Safety Follow-up/Extension study
At the completion of the Dose-Blind Period, Investigators should discuss treatment options withthe subject. Subjects will be given the opportunity to enter an extension study at Week 48,provided they do not qualify for rescue therapy.
REDACTED yy.. COPYgatorgatoy to eny to e
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on nterter ap
plica
tion finein
shouldshouer an r an
andvemeveme
ned ed
any adequadequ
en
exten
sionsThe lThe l
or va
riatio
ns
ive ve
thereo
f.
UC
B07
Sep
201
8St
atis
tical
Ana
lysi
s Pla
nB
imek
izum
abPA
0008
Con
fiden
tial
Page
19
of 1
83
Tab
le2‒
1:Sc
hedu
le o
f ass
essm
ents
Vis
ita /W
eek
Prot
ocol
ac
tivity
Screening (V1) (14-28 days)
Baseline (Day 1, V2)(firstdose)
Dou
ble-
Blin
d Pe
riod
(wee
ks a
fter
firs
t do
se)
Dos
e-B
lind
Peri
od(w
eeks
aft
er fi
rst d
ose)
ET
SFU b
3 / 14/ 2
5/ 46/ 8
7/ 128/ 16
9/ 2 0
1 0/ 2 4
1 1/ 2 8
1 2/ 3 2
1 3/ 3 6
1 4/ 4 015
/44
16/
48
Writ
ten
Info
rmed
con
sent
X
Dem
ogra
phic
dat
aX
PsA
his
tory
X
Sign
ifica
nt p
ast m
edic
al
hist
ory
and
conc
omita
nt
dise
ases
XX
c
Incl
usio
n/ex
clus
ion
crite
riaX
X
Prio
r med
icat
ion
X
Con
com
itant
med
icat
ion
XX
XX
XX
XX
XX
XX
XX
XX
XX
Adv
erse
eve
nts
XX
XX
XX
XX
XX
XX
XX
XX
X
eC-S
SRS
XX
XX
XX
XX
XX
XX
XX
XX
XX
HA
DS
XX
XX
XX
XX
XX
XX
XX
XX
X
PsA
QoL
XX
XX
XX
XX
XX
BA
SDA
IX
XX
XX
XX
XX
XX
X
HA
Q-D
IX
XX
XX
XX
XX
XX
X
SF-3
6X
XX
XX
XX
XX
REDACTED
ACD
DAARERRRR
COPY YPYYY YCO
This do
cumen
t ntca
nnot t
nnbe
bb
used
dus
to tttosu
pport
rt
XX
rtrtrt rt
XXup
any y
XX
yyy nymark
eting
intininn
g
arkin
XX
ar ar ar ar
autho
rizati
on
oooon
hotio
au
appli
catio
n ati at at at
ppat
pppppppp
and 22 88 aaaaa
any 2/2/ 33 an
exten
sions
riodiod
first
dir
st d
//11en
ex
or va
riatio
ns
on
va
thereo
f.
00
UC
B07
Sep
201
8St
atis
tical
Ana
lysi
s Pla
nB
imek
izum
abPA
0008
Con
fiden
tial
Page
20
of 1
83
Tab
le2‒
1:Sc
hedu
le o
f ass
essm
ents
Vis
ita /W
eek
Prot
ocol
ac
tivity
Screening (V1) (14-28 days)
Baseline (Day 1, V2)(firstdose)
Dou
ble-
Blin
d Pe
riod
(wee
ks a
fter
firs
t do
se)
Dos
e-B
lind
Peri
od(w
eeks
aft
er fi
rst d
ose)
ET
SFU b
3 / 14/ 2
5/ 46/ 8
7/ 128/ 16
9/ 2 0
1 0/ 2 4
1 1/ 2 8
1 2/ 3 2
1 3/ 3 6
1 4/ 4 015
/44
16/
48
PGA
DA
XX
XX
XX
XX
XX
XX
X
PsA
ID-9
XX
XX
XX
XX
XX
Tube
rcul
osis
que
stio
nnai
reX
XX
XX
XX
X
Hei
ght
X
Vita
l sig
ns (p
ulse
, te
mpe
ratu
re, B
P)dX
XX
XX
XX
XX
XX
XX
XX
XX
X
Wei
ght
XX
XX
XX
XX
Phys
ical
exa
min
atio
neX
XX
XX
X
SJC
and
TJC
XX
XX
XX
XX
XX
XX
X
BSA
aff
ecte
d by
PSO
(B
SA p
alm
met
hod)
XX
XX
XX
PASI
fX
XX
XX
XX
XX
X
mN
APS
IX
XX
XX
X
MA
SES
XX
XX
XX
XX
XX
LDI
XX
XX
XX
XX
XX
PhG
AD
AX
XX
XX
XX
XX
XX
X
REDACTED D
XXTE
EAC
EEDRE
COPY Y
XX
PYYY YCO
This do
cumen
t ntca
nnott t
nnbe
bb
used
dus
to tttosu
pport
rt
XXup
any ny anynyyny
marketi
ng
in etark
XXmaaaa
autho
rizati
on
oooon tio
XXizaho
a
appli
catio
n XX ati at at at
ppat
pppppppp
and 22 88 aaaaa
any 2/2/ 33 an
exten
sions
riodiod
first
dir
st d
//11en
ex
or va
riatio
ns
on
va
thereo
f.
00
UC
B07
Sep
201
8St
atis
tical
Ana
lysi
s Pla
nB
imek
izum
abPA
0008
Con
fiden
tial
Page
21
of 1
83
Tab
le2‒
1:Sc
hedu
le o
f ass
essm
ents
Vis
ita /W
eek
Prot
ocol
ac
tivity
Screening (V1) (14-28 days)
Baseline (Day 1, V2)(firstdose)
Dou
ble-
Blin
d Pe
riod
(wee
ks a
fter
firs
t do
se)
Dos
e-B
lind
Peri
od(w
eeks
aft
er fi
rst d
ose)
ET
SFU b
3 / 14/ 2
5/ 46/ 8
7/ 128/ 16
9/ 2 0
1 0/ 2 4
1 1/ 2 8
1 2/ 3 2
1 3/ 3 6
1 4/ 4 015
/44
16/
48
ECG
XX
XX
X
Hem
atol
ogy/
bioc
hem
istry
/ur
inal
ysis
XX
XX
XX
XX
XX
XX
XX
XX
XX
hs-C
RPg
XX
XX
XX
XX
XX
XX
XX
CR
PgX
XX
X
Preg
nanc
y te
stin
ghX
XX
XX
Hep
atiti
s B a
nd C
test
ing
X
HIV
test
ing
X
RF,
ant
i-CC
P an
tibod
ies
X
Blo
od sa
mpl
e fo
r bi
mek
izum
ab p
lasm
a co
ncen
tratio
nsi
XX
XX
XX
XX
XX
XX
X
Bim
ekiz
umab
ant
ibod
y de
tect
ioni
XX
XX
XX
XX
XX
X
Blo
od sa
mpl
es fo
r cy
toki
nes,
com
plem
ent,
and
biom
arke
r ana
lysi
si
XX
XX
XX
X
REDACTED DEDDD
ACD
ACACACE
ACEEDRE
COPY Y
XXO
This do
cumen
t tcan
not
issi c
t tbe bb
used
du
to otsu
pport
rt
s
any ny XXanynyyny
marketi
ng
in etark maaaa
autho
rizati
on XXn
tio
XX
at
hohhohohoa
appli
catio
n ati
XX
at at at
ppat
a
and 22 88 aaaaa
any 2/2/ 33 an
exten
sions
riodiod
first
dir
st d
//11en
ex
or va
riatio
ns
on
va
thereo
f.
00
UC
B07
Sep
201
8St
atis
tical
Ana
lysi
s Pla
nB
imek
izum
abPA
0008
Con
fiden
tial
Page
22
of 1
83
Tab
le2‒
1:Sc
hedu
le o
f ass
essm
ents
Vis
ita /W
eek
Prot
ocol
ac
tivity
Screening (V1) (14-28 days)
Baseline (Day 1, V2)(firstdose)
Dou
ble-
Blin
d Pe
riod
(wee
ks a
fter
firs
t do
se)
Dos
e-B
lind
Peri
od(w
eeks
aft
er fi
rst d
ose)
ET
SFU b
3 / 14/ 2
5/ 46/ 8
7/ 128/ 16
9/ 2 0
1 0/ 2 4
1 1/ 2 8
1 2/ 3 2
1 3/ 3 6
1 4/ 4 015
/44
16/
48
Blo
od sa
mpl
es fo
r ge
nom
ic,
prot
eom
ic/m
etab
olom
ics
anal
yses
i,j
XX
XX
XX
X
For p
artic
ipat
ing
cent
ers:
B
lood
sam
ples
for f
low
cy
tom
etry
i
XX
XX
XX
X
Blo
od sa
mpl
es fo
r ge
netic
/epi
gene
tic
anal
ysis
i,j
XX
XX
IGR
A T
uber
culo
sis t
estk
XX
XX
Eval
uatio
n fo
r res
cuel
XX
X
Che
st x
-ray
mX
Han
d an
d fo
ot x
-ray
snX
Ente
r IX
RSo
XX
XX
XX
XX
XX
XX
XX
XX
XX
Bim
ekiz
umab
or p
lace
bo
adm
inis
tratio
np,q
XX
XX
XX
XX
XX
XX
BA
SDA
I=B
ath
Ank
ylos
ing
Spon
dylit
is D
isea
se A
ctiv
ity In
dex;
BP=
bloo
d pr
essu
re; B
SA=b
ody
surf
ace
area
; CC
P= c
yclic
citr
ullin
ated
pep
tide;
C
RP=
C-r
eact
ive
prot
ein;
eC
RF=
elec
troni
c C
ase
Rep
ort F
orm
; eC
-SSR
S=el
ectro
nic
Col
umbi
a-Su
icid
e Se
verit
y R
atin
g Sc
ale;
EC
G=e
lect
roca
rdio
gram
;
REDACTED
ACD
EDRE
COPY YCO
This do
cumen
t
otei
n;ei
n;
can
ng S
png
Span
not t
anbe
ebus
ed d
XX
eddddu
to tsu
pport
rtpo
s
any ny annn
marketi
ng
arkin
m
autho
rizati
on
o
XX
ooon
hotio
au
appli
catio
n XX
ppati
and 22 88 aaaaa
any 2/2/ 33 an
exten
sions
riodiod
first
dir
st d
//11en
ex
or va
riatio
ns
on
va
thereo
f.
00
UC
B07
Sep
201
8St
atis
tical
Ana
lysi
s Pla
nB
imek
izum
abPA
0008
Con
fiden
tial
Page
23
of 1
83
Tab
le2‒
1:Sc
hedu
le o
f ass
essm
ents
Vis
ita /W
eek
Prot
ocol
ac
tivity
Screening (V1) (14-28 days)
Baseline (Day 1, V2)(firstdose)
Dou
ble-
Blin
d Pe
riod
(wee
ks a
fter
firs
t do
se)
Dos
e-B
lind
Peri
od(w
eeks
aft
er fi
rst d
ose)
ET
SFU b
3 / 14/ 2
5/ 46/ 8
7/ 128/ 16
9/ 2 0
1 0/ 2 4
1 1/ 2 8
1 2/ 3 2
1 3/ 3 6
1 4/ 4 015
/44
16/
48ET
=Ear
ly T
erm
inat
ion;
HA
DS=
Hos
pita
l Anx
iety
and
Dep
ress
ion
Scal
e; H
AQ
-DI=
Hea
lth A
sses
smen
t Que
stio
nnai
re—
Dis
abili
ty In
dex;
HIV
=hum
an
imm
unod
efic
ienc
y vi
rus;
hs-
CR
P=hi
gh se
nsiti
vity
CR
P; IG
RA
=int
erfe
ron
gam
ma
rele
ase
assa
y; IM
P=in
vest
igat
iona
l med
icin
al p
rodu
ct; I
XR
S=in
tera
ctiv
e vo
ice
or w
eb re
spon
se sy
stem
; mN
APS
I=m
odifi
ed N
ail P
soria
sis S
ever
ity In
dex;
LD
I=Le
eds D
acty
litis
Inde
x; P
ASI
=pso
riasi
s are
a se
verit
y in
dex;
PG
AD
A=P
atie
nt’s
Glo
bal A
sses
smen
t of D
isea
se A
ctiv
ity; P
hGA
DA
=Phy
sici
an’s
Glo
bal A
sses
smen
t of D
isea
se A
ctiv
ity; P
sA=p
soria
tic a
rthrit
is;
PsA
ID=P
soria
tic A
rthrit
is Im
pact
of D
isea
se; P
sAQ
oL=P
soria
tic A
rthrit
is Q
ualit
y of
Life
; PSO
=pso
riasi
s; P
tAA
P=Pa
tient
’s A
sses
smen
t of A
rthrit
is P
ain;
R
F=rh
eum
atoi
d fa
ctor
; SF-
36=s
hort
form
-36;
SFU
=Saf
ety
Follo
w-U
p; S
JC=s
wol
len
join
t cou
nt; T
B=t
uber
culo
sis;
TJC
=ten
der j
oint
cou
nta
Vis
it w
indo
ws o
f +/-3
day
s fro
m th
e fir
st d
ose
at a
ll vi
sits
thro
ugh
Wee
k 12
and
+/-4
day
s for
all
visi
ts a
fter W
eek
12 e
xcep
t SFU
. The
Saf
ety
Follo
w-u
p V
isit
win
dow
is -3
and
+7
days
from
last
dos
e.b
Safe
ty F
ollo
w-U
p V
isit
occu
rs 2
0 w
eeks
afte
r the
last
dos
e fo
r all
subj
ects
who
com
plet
e th
e st
udy
and
do n
ot e
nter
the
exte
nsio
n st
udy
or w
ho d
isco
ntin
ue
early
, inc
ludi
ng th
ose
with
draw
n fr
om st
udy
treat
men
t.c
Ensu
re th
ere
are
no si
gnifi
cant
cha
nges
in m
edic
al h
isto
ry th
at w
ould
exc
lude
the
subj
ect b
ased
on
the
excl
usio
n cr
iteria
.d
At B
asel
ine/
Day
1, c
olle
ct p
ulse
and
BP
prio
r to
drug
adm
inis
tratio
n an
d th
en a
t 30
min
utes
and
1 h
our a
fter d
osin
g. A
t all
othe
r vis
its c
olle
ct p
ulse
and
BP
prio
r to
drug
adm
inis
tratio
n an
d on
ce a
fter d
osin
g. A
ll ot
her p
roce
dure
s are
don
e pr
ior t
o do
sing
.e
Incl
udes
eva
luat
ion
of si
gns a
nd sy
mpt
oms o
f act
ive
TB a
nd ri
sk fo
r exp
osur
e to
TB
.f
PASI
will
be
asse
ssed
in a
ll su
bjec
ts w
ith B
asel
ine/
Day
1 B
SA a
ffect
ed b
y PS
O ≥
3% a
s det
erm
ined
by
the
BSA
pal
m m
etho
d.g
Hig
h se
nsiti
vity
-CR
P is
per
form
ed a
t Scr
eeni
ng, B
asel
ine,
and
all
othe
r stu
dy v
isits
as m
arke
d w
ith a
nX
. CR
P w
ill b
e pe
rfor
med
at B
asel
ine,
Wee
k 4,
W
eek
12, a
nd W
eek
24 o
nly.
Afte
r Scr
eeni
ng, t
he h
s-C
RP
and
CR
P da
ta w
ill n
ot b
e se
nt to
the
Inve
stig
ator
to p
rote
ct th
e bl
inde
d na
ture
of t
he tr
eatm
ent
assi
gnm
ents
. h
A se
rum
pre
gnan
cy te
st w
ill b
e pe
rfor
med
at S
cree
ning
for a
ll w
omen
of c
hild
bear
ing
pote
ntia
l. Th
is re
com
men
datio
n al
so a
pplie
s to
wom
en o
f chi
ldbe
arin
g po
tent
ial w
ith in
freq
uent
or i
rreg
ular
men
stru
al c
ycle
s and
wom
en d
urin
g m
enop
ause
. Nat
ural
men
opau
se is
reco
gniz
ed to
hav
e oc
curr
ed a
fter 1
2 co
nsec
utiv
e m
onth
s of a
men
orrh
ea, f
or w
hich
ther
e is
no
othe
r obv
ious
pat
holo
gica
l or p
hysi
olog
ical
cau
se (I
nter
natio
nal M
enop
a use
Soc
iety
[IM
S], 2
015)
. A
urin
e pr
egna
ncy
test
is a
lso
requ
ired
at th
e B
asel
ine,
Wee
k 48
, ET,
and
at S
FU V
isits
. A u
rine
preg
nanc
y te
st w
ill a
lso
be p
erfo
rmed
at a
ny st
udy
visi
t w
here
ther
e ha
s bee
n a
dela
y in
men
ses.
This
reco
mm
enda
tion
also
app
lies t
o w
omen
of c
hild
bear
ing
pote
ntia
l with
infr
eque
nt o
r irr
egul
ar m
enst
rual
cyc
les.
Preg
nanc
y te
st re
sults
mus
t be
nega
tive
prio
r to
adm
inis
terin
g IM
P.i
At d
osin
g vi
sits
, all
bloo
d sa
mpl
es a
re ta
ken
prio
r to
dosi
ng. B
lood
sam
ples
for b
imek
izum
ab a
nd a
nti-b
imek
izum
ab a
ntib
ody
dete
ctio
n w
ill b
e pr
oces
sed
as
per i
nstru
ctio
ns in
the
labo
rato
ry m
anua
l. j
A se
para
te In
form
ed C
onse
nt F
orm
will
be re
quire
d fo
r sub
ject
s who
pro
vide
sam
ples
for g
enet
ic te
stin
g. T
he In
form
ed C
onse
nt F
orm
mus
t be
sign
ed p
rior
to c
olle
ctin
g sa
mpl
es.
REDACTEDs12
and
12 a
n
ubje
cts
bjec
ts
wou
ldw
ould
tt
COPY
s Glo
bs G
lolit
y of
Lty
of
wol
lew
olle
This do
cumen
t
Co
es.s.
cann
otneam
ple
mpl
ebo
rato
ror
ator
onse
nns
en
be
men
sen
egat
ieg
atius
edwhi
cw
hied
at t
hd
at th
nses
.se
s.
to
trual
trua
ichch
supp
ort
hss--C
RC
R
Scre
eni
Scre
eni
alcy
alc
any
BSA
BS
ne, a
nne
, an
RP
marketi
ng
d ex
cld
excl
tion
and
on a
nro
cedu
reoc
edu
d ris
k fo
risk
fSASA
autho
rizati
on
Life
Lif
en jo
inen
join
+/+/--4
d4 d
s who
cw
ho c
appli
catio
n
smen
t sm
ent
ay; I
MP
y; IM
Pds
Dac
ts D
acal
Ass
esA
sse
fe; P
Se;
P
and 22 88 aaaa
any 2/2/ 33 an
exten
sions
riodiod
first
dir
st d
//11en
ex
or va
riatio
ns
on
va
thereo
f.
00
UC
B07
Sep
201
8St
atis
tical
Ana
lysi
s Pla
nB
imek
izum
abPA
0008
Con
fiden
tial
Page
24
of 1
83
Tab
le2‒
1:Sc
hedu
le o
f ass
essm
ents
Vis
ita /W
eek
Prot
ocol
ac
tivity
Screening (V1) (14-28 days)
Baseline (Day 1, V2)(firstdose)
Dou
ble-
Blin
d Pe
riod
(wee
ks a
fter
firs
t do
se)
Dos
e-B
lind
Peri
od(w
eeks
aft
er fi
rst d
ose)
ET
SFU b
3 / 14/ 2
5/ 46/ 8
7/ 128/ 16
9/ 2 0
1 0/ 2 4
1 1/ 2 8
1 2/ 3 2
1 3/ 3 6
1 4/ 4 015
/44
16/
48k
It is
reco
mm
ende
d th
at Q
uant
ifero
n TB
GO
LD te
st b
e pe
rfor
med
. l
At W
eeks
16,
24,
and
36,
subj
ects
will
be
eval
uate
d fo
r ina
dequ
ate
resp
onse
to tr
eatm
ent.
Subj
ects
who
do
not s
how
an
impr
ovem
ent i
n te
nder
and
swol
len
join
t cou
nts w
ill b
e el
igib
le to
rece
ive
resc
ue th
erap
y.m
If a
subj
ect h
as h
ad a
rece
nt ra
diog
raph
of t
he c
hest
with
in 3
mon
ths p
rior t
o th
e Sc
reen
ing
Vis
it, it
may
be
used
as t
he S
cree
ning
che
st x
-ray
.n
Han
d an
d fo
ot x
-ray
s are
not
requ
ired
if th
e su
bjec
t has
han
d an
d/or
foot
x-ra
ys (o
nly
hand
OR
foot
is re
quire
d) th
at a
re n
o m
ore
than
2 y
ears
from
the
date
of
Scr
eeni
ng, a
nd a
re a
vaila
ble
for s
ubm
issi
on w
ithin
the
Scre
enin
g Pe
riod
to th
e ce
ntra
l rad
iolo
gy re
ader
and
who
se re
adin
g is
con
sist
ent w
ith C
ASP
AR
C
riter
ia #
5 sh
owin
g ill
-def
ined
oss
ifica
tion
near
join
t mar
gin
(but
exc
ludi
ng o
steo
phyt
e fo
rmat
ion)
. For
thos
e su
bjec
ts w
ithou
t prio
r x-r
ays t
hat m
eet t
his
Crit
erio
n an
d th
ose
who
se su
bmitt
ed fi
lms a
re n
ot re
ad w
ith c
hang
es c
onsi
sten
t with
CA
SPA
R, h
and
and
foot
x-r
ays m
ay b
e do
ne d
urin
g th
e Sc
reen
ing
Perio
d.o
IXR
S is
use
d to
regi
ster
the
Subj
ect a
t Scr
eeni
ng, r
ando
miz
e th
e Su
bjec
t at B
asel
ine/
Day
1, r
eran
dom
ize
the
Subj
ect a
t Wee
k 12
(Sub
ject
s in
the
plac
ebo
or
bim
ekiz
umab
16m
g tre
atm
ent g
roup
s), a
nd to
regi
ster
all
visi
ts.
pTh
e m
inim
um ti
me
betw
een
dose
s dur
ing
the
Dou
ble-
Blin
d Pe
riod
shou
ld b
e no
less
than
25
days
and
no
mor
e th
an 3
1da
ys a
nd d
urin
g th
e D
ose-
Blin
d Pe
riod
shou
ld b
e no
less
than
24
days
and
no
mor
e th
an 3
2 da
ys.
q A
t Wee
k 12
, the
Dos
e-B
lind
Perio
d do
sing
will
star
t.
REDACTEDo ng
ost
ng o
son
sist
eon
sist
e
Subj
ecSu
bje
.. ioi
COPY
e Sc
ree
Scr
(onl
y (o
nly
the
cth
e c
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
d sh
oul
shou
lys
..
autho
rizati
on
han
han
entra
l en
tral
eoph
yteo
phyt
ent w
ithnt
with
t at B
aat
Ba
appli
catio
n
ject
s wh
ects
w
enin
g V
iin
g V
nd O
ndO
and 22 88 aaaa
any 2/2/ 33 an
exten
sions
riodiod
first
dir
st d
//11en
ex
or va
riatio
ns
on
va
thereo
f.
00
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 25 of 183
2.4 Determination of sample sizeA total of 200 subjects (40 in each treatment group) are planned to be randomized in this study.
The sample size is calculated based on the ACR50 response data from the Phase 2 bimekizumab study in subjects with PsA (PA0007). The ACR50 responses at Week 12 were reported to be 80.0% (95% CI: 37.6%; 96.4%), 52.6% (95% CI: 31.7%; 72.7%), and 33.3% (95% CI: 9.7%; 70.0%) for the bimekizumab 560/320mg, bimekizumab 240/160mg, and bimekizumab 80/40mg doses, respectively. Placebo ACR50 response at Week 12 is based on the PA0007 study (0%), the FUTURE 1 study (8.0%; Mease et al, 2015) and FUTURE 2 study (6.0%; McInnes et al, 2015).
Since the uncertainty of ACR50 responder rates of PA0007 is high due to the small number of subjects, the midpoint of the point estimate and the lower 95% CI was used for the sample size calculation of all active dose groups as a conservative assumption. For the bimekizumab 16mg dose, the midpoint of the 80/40mg dose was reduced by 6.5%.
Hence, ACR50 responder rates of 58.5%, 42.2%, 15.0%, and 8% at the end of a 12-week treatment period for bimekizumab 320mg, bimekizumab 160mg, bimekizumab 16mg, and placebo have been assumed, respectively. Since those ACR50 responder rates are based on a study population of TNF inhibitor naïve subjects, an adjustment was made for a reduction in ACR50 response of 25% by a maximum of 30% of subjects with TNF inhibitor experience (ie, 12 subjects per group).
The sample size for the primary objective of evaluating the dose response relationship was calculated using a 2-sided test for detecting a linear trend across proportions (Nam, 1987) at a 2-sided significance level of 0.05. With 40 subjects in each treatment group, the test for detecting the overall dose response based on ACR50 response is powered at >99%.
The sample size calculations were performed using the software nQuery Advisor® 7.0.
3 DATA ANALYSIS CONSIDERATIONS3.1 General presentation of summaries and analysesAll computations and generation of outputs will be performed using SAS® Version 9.3 or later. All tables and listing will use Courier New font size 9.
All original and derived variables will be listed and described using summary statistics (number of observations [n], mean, standard deviation [SD], median, minimum and maximum, unless otherwise stated) or frequency counts (number of subjects [N] and percentages). For multiple post-Baseline assessments at a specific visit, the first non-missing measurement will be used for summary statistics or frequency counts.
Descriptive statistics will be displayed to provide an overview of the study results. For categorical variables, the number and percentage of subjects in each category will be presented. The denominator for percentages will be based on the number of subjects appropriate for the purpose of analysis. Unless otherwise noted, all percentages will be displayed to 1 decimal place. No percentage will be displayed for zero counts, and no decimal will be presented when the percentage is 100%. For continuous variables, descriptive statistics will include number of subjects with available measurements (n), mean, SD, median, minimum, and maximum.
REDACTEDevaluevalug a ling a lin
40 s0 subjubjACR50ACR5
COPYsubjesubj
This do
cumen
t DescesccatecateTT
cann
otselinelin
mary stary s
scripcrip
betateate
ineineus
ed derier
s [n], s [n], ed)ed)
toriverivesu
pport
enerationerat
will uswill us
any sentsenmark
eting
0 rere
rformerform
YSIS CSIS
autho
rizati
on
ating ating near trear tr
bjects iects respesp
appli
catio
n bimimrespondespon
ment waent wcts withts wi
andhe enhe en
mekmek
any
ndd
exten
sions
ll numl num
the sathe samekimeki
or va
riatio
ns
mg g %), ), al, al,
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 26 of 183
Decimal places for descriptive statistics will always apply the following rules:
“n” will be an integer.
Mean, SD, and median will use 1 additional decimal place compared to the original data.
Coefficient of variance (CV[%]) will be presented with 1 decimal place.
Minimum and maximum will have the same number of decimal places as the original value.
Statistical tests of efficacy variables will be presented as 2-sided p-values rounded to 3 decimal places. P-values less than 0.001 will be presented as “<0.001” and p-values greater than 0.999 will be presented as “>0.999” Statistical comparisons will be performed by two-sided statistical tests at the 0.050 level of significance.
The SAS output for Cochran-Mantel-Haenszel test, logistics regression, and analysis of covariance (ANCOVA), as well as multiple imputation (MI) will be provided as a separate PDF document in addition to TFLs. The SAS output will be included in the ‘Documentation of Statistical Methods’ section of the clinical study report.
In the Double-Blind Period the order of treatment groups will be presented in tables from left to right will be
Placebo
Bimekizumab 16mg
Bimekizumab 160mg
Bimekizumab 160mg with loading dose 320mg
Bimekizumab 320mg
The general principle is to go from the lowest to highest dose when moving from left to right. Tables may also include columns for all subjects or all subjects on bimekizumab. An overview of the treatment group assignment is available in Section 12.3.
Selected tables which are specified in the TFL shells will only use data from the Dose-Blind Period. For listings and the selected tables, the label and order of treatment groups will be presented as follows:
Placebo to bimekizumab 160mg at Week 12
Placebo to bimekizumab 320mg at Week 12
Bimekizumab 16mg to bimekizumab 160mg at Week 12
Bimekizumab 16mg to bimekizumab 320mg at Week 12
Bimekizumab 160mg with loading dose 320mg to bimekizumab 160mg at Week 12
Bimekizumab 160mg
Bimekizumab 320mg
The abbreviation for bimekizumab is BKZ and will be used in tables and listings headers. In the TFLs subjects on bimekizumab 160mg with loading dose 320mg will be displayed as “BKZ 160mg w/LD”. All subjects on bimekizumab will be labeled as “All BKZ” in the TFLs.
REDACTED
dose 3dose 3
COPY
This do
cumen
t BBca
nBB nnotmekizkiz
imeime
be to
zuzuus
edimekmek
o bimbim
to kikisu
pport
specipecithe selthe se
anyt is ais
ifieif
marketi
ng
he lowhe lowfor allfor alava
autho
rizati
on
320mg20mg
appli
catio
npresepres and a
ny edd
DocumDocumex
tensio
nsd stasta
nalysialysid as d as
or0.90.9atisati
varia
tions
e.
imal mal 999999
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 27 of 183
A complete set of data listings containing all documented data and all derived data (eg, change from Baseline) will be generated.
Unless otherwise stated, listings will be sorted by treatment, subject number within each treatment group (not randomization number), variable (if applicable) and visit (if applicable; including timing relative to dosing if applicable). For listings including nonrandomized subjects, the nonrandomized subjects will be shown first in the listing, ordered by subject number. All listings will include repeat and unscheduled measurements; such measurements will appear in chronological order together with the scheduled visits, ie, a repeated measurement will appear directly after the visit and time relative to dosing for which the repeat measurement was performed. In all the listings dates will be presented in the format ‘YYYY-MM-DD’ and times will be presented in 24h clock format as ‘hh:mm’.
3.2 General study level definitions3.2.1 Relative dayRelative day will be calculated as the current date minus the date of first dose of study drug administration plus 1 for days on or after the day of first dose of study drug dose. Relative day 1 is the date of first bimekizumab administration.
Relative days before first administration of bimekizumab will have the prefix ‘-‘ and will be calculated as date of fist dose of study drug administration minus the current date.
For days after the last bimekizumab administration, relative day will be calculated as the current date minus the date of last dose of study drug administration including the prefix ‘+’.
Calculations of ‘Relative Day’ should not include partial dates, but should be left blank in these instances.
For AEs, relative days for start and stop dates will be calculated as the number of days since the first injection of the medication. For non-treatment emergent AEs, relative day of onset will be negative if the event started and stopped before the first dose. A complete date must be established in order to correctly identify the AEs. Section 4.2.2 describes imputation rules in case of missing data for AEs.
3.2.2 Study PeriodsThe following study periods are defined for the classification by study period:
Pre-treatment period (Screening period): up to 28 days, ends with the visit date of the first dose of study medication (Visit 2)
Double-Blind treatment period: starts with the visit date of the first dose of study medication (Visit 2), ends at Week 12 visit date
Dose-Blind treatment period: starts at the Week 12 visit, ends at Week 48 visit.
Post-treatment period: the post-treatment period (Follow-up period) is the period after the Week 48 visit
3.3 Definition of Baseline valuesThe last valid measurement before first dose of study drug (Visit 2) will be used as the Baseline value. If a scheduled Baseline assessment is taken on the same day and after the first
REDACTED atioatioug admug adm
not incnot inc
COPY mabmastratiostrati
on, ron
This do
cumen
tDDca
nnot ubleuble
(Visit(Visit
bef studf stu used
dy
ment pent pdyd
toy pey pesup
port
PerioPerioi
any
ststly idey ide
marketi
ng
op datep daor nonnon
toppeopp
autho
rizati
on
relativrelatiministrminist
clude plude p
appli
catio
n tudyud
will havill han minn min
andfirst drst d
dy dy d
any
dosd
exten
sionsandand or d td t
varia
tions
pear ear
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 28 of 183
administration of study medication, it will be analyzed as the first post-Baseline assessment. If a scheduled Baseline assessment is taken on the same day as first administration of study medication and no time is collected, it will be assumed to have been taken prior to study medication.
The exception of the above rule are the questionnaires collected from the vendor ERT and measurements for CRP and hs-CRP. For those measurement, the last valid measurement at the same day as Visit 2 will be used as the Baseline value.
However, for some variables, assessments may be scheduled for Screening only and not for Baseline. In this case the Screening value will be utilized as Baseline value. If a Baseline measurement is missing, and a Screening value available, the Screening value will be utilized as Baseline instead.
3.4 Protocol deviationsImportant protocol deviations are defined as those deviations from the protocol likely to have ameaningful impact on the primary efficacy outcomes for an individual subject. Important protocol deviations will be identified and classified by the deviation types defined in the appropriate protocol-specific document. All protocol deviations will be reviewed as part of the ongoing data cleaning process and all important deviations will be identified and documented prior to unblinding to confirm exclusion from Per Protocol Set (PPS), Pharmacokinetics Per-Protocol Set (PK-PPS) and Pharmacodynamics Per-Protocol Set (PD-PPS). The exclusion from PPS is limited to the Double-Blind Period, ie. only subjects with important protocol deviationsprior to re-randomization at Visit 7 (Week 12) will be excluded from PPS. Subjects with important protocol deviations after Visit 7 (Week 12) will not be excluded from PPS.
3.5 Mapping of assessments performed at early termination visitStudy assessments at an early termination visit where visit date matches the visit date of a scheduled visit will be summarized at the scheduled visit with the same visit date. Premature study termination visit assessments that do not have a scheduled visit with a matching date will be assigned to the next scheduled site visit following the last visit where assessments were available regardless whether or not there will be an assessment on this visit. The assessment of AbAb is an exception to this rule: AbAb will be mapped to the next visit where antibody levels are measured. For subjects who discontinue study treatment early and return for the Week 12 visits as per the protocol, the assessments collected at that visit are summarized as Week 12 assessments.
3.6 Analysis setsThe primary efficacy variable will be analyzed for all subjects in the Full Analysis Set (FAS). The supportive analyses for the primary efficacy variables will include the Randomized Set (RS) and PPS. All other efficacy variables will be based on the FAS. Demographics tables will be performed for FAS, Safety Set (SS), and Dose-Blind Set (DBS). Safety variables will be summarized using the SS. In addition, safety analyses during the Dose-Blind Period will be conducted on all subjects in the DBS. PK variables will be analyzed for all subjects in the PK-PPS. PD variables will be analyzed for all subjects in the PD-PPS.
At the time of the Week 48 interim it was discovered that the Dose-Blind Responder Set (DBRS) and Escape Subject Set (ESS) were defined incorrectly from the original intent. The original
REDACTEDonlyonl12) wi12) wi
7 (W7 (W
ssmessme
COPYrotocrotocerer-ProPrly sly s
This do
cumen
tandand PPpepe
cann
ot primarrimare supsup
be us
edprotrot to
ubjjotoctoc
supp
orther orer orto this o thi
bjectsject
anyntsnt
uled sled smark
eting
mententation vation v
d at theat thts ths th
autho
rizati
onotoctoc
subjecubjecwill be ll be
Week 1eek 1
ntsts
appli
catio
n on tn ts will bwill
wiwill be ll beol Set (l Set colcol
andal subl sub
typtyp
any otocootoco
bjebj
exten
sionsn
be utile utior e e
varia
tions
or r
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 29 of 183
intent of the ESS analysis set was to have only those subjects that were eligible to receive rescue therapy, that went on to actually receive rescue therapy, with all other subjects remaining in the DBRS. As such, two extra sets, the Corrected Dose-Blind Responder Set (cDBRS) and Corrected Escape Subject Set (cESS) have been added. These analysis sets will correctly select out those subjects that were eligible and received rescue medication. All outputs created on the DBRS and ESS will also be produced on the cDBRS and cESS to give full transparency of the difference. Further all outputs using the DBRS and ESS will be repeated on the DBS and this analysis will be used as the main analysis of the Dose-Blind as this analysis set is the most unbiased set.
3.6.1 Enrolled Set (ES)The Enrolled Set (ES) will consist of all subjects who have given informed consent.
3.6.2 Randomized Set (RS)The RS consists of all randomized subjects.
3.6.3 Safety Set (SS)The SS consists of all randomized subjects who received at least 1 dose of the IMP.
3.6.4 Full Analysis Set (FAS)The FAS consists of all randomized subjects who received at least 1 dose of the IMP and have valid measurement of the primary efficacy variable at Baseline.
3.6.5 Per-Protocol Set (PPS)The PPS consists of all subjects in the FAS who had no important protocol deviation affecting the primary efficacy variable. The subjects with important protocol deviations will be predefined and evaluated during a data evaluation meeting prior to unblinding of the data.
3.6.6 Pharmacokinetics Per-Protocol Set (PK-PPS)The PK-PPS is a subset of the FAS, consisting of those subjects who had no important protocol deviations affecting the pharmacokinetic variables, as confirmed during ongoing data cleaning meetings prior to database lock.
3.6.7 Pharmacodynamics Per-Protocol Set (PD-PPS)The PD-PPS consists of all randomized subjects who took at least one dose of the IMP and provided at least 1 PD measurement post-dose that is without important protocol deviation affecting that time point.
3.6.8 Dose-Blind Set (DBS)The DBS consists of all subjects who started the Dose-Blind Period and who received at least 1 dose of IMP during the Dose-Blind Period.
3.6.9 Escape Subject Set (ESS)The ESS consists of all subjects who started the Dose-Blind Period and who received at least 1 dose of the IMP during the Dose-Blind Period and who have not achieved at least a 10% reduction from Baseline in SJC and TJC at Week 16, Week 24, or Week 36.
REDACTED S whoS who
ects wects won meon me
COPYeiveeive at Be at B
This do
cumen
t oses
3.63.6ca
nnot DBS cDBS c
e of e of
be us
edt 1 P1 P
time time to
sts ots oPDPD
supp
ort lockock
acodycodyof aof
anyacokacok
k.k.mark
eting
eetinetin
Perer--PrPS, conS, cokinki
autho
rizati
on Bas
o had had with imith imingng
appli
catio
n do
at leasat leaselineseline
and ose oose
any e
xtens
ions o
r vari
ation
s the
reof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 30 of 183
3.6.10 Dose-Blind Responder Set (DBRS)The DBRS consists of all subjects who started the Dose-Blind Period and who received at least 1 dose of the IMP during the Dose-Blind Period and who have achieved at least a 10% reduction from Baseline in SJC and TJC at Week 16, Week 24, and Week 36.
3.6.11 Corrected Escape Subject Set (cESS)The cESS is a corrected version of the ESS, which was incorrectly defined previously. The cESS consists of subjects that have shown less than a 10% improvement in SJC and TJC at either Week 16, Week 24, or Week 36 and received rescue therapy.
3.6.12 Corrected Dose-Blind Responder Set (cDBRS)The cDBRS is a corrected version of the DBRS, which was incorrectly defined previously. The cDBRS consists of subjects that have shown at least a 10% improvement in SJC or TJC at Week 16, Week 24 and Week 36. Subjects that would be in the cESS, including subjects that discontinue, that did not receive rescue therapy will be in the cDBRS.
3.7 Treatment assignment and treatment groupsIt is expected that subjects receive treatment as randomized and hence safety analyses will be based on the SS, as randomized. However, if after unblinding it is determined that subjects randomized to placebo received bimekizumab at any time, then for safety analyses these subjects will be reallocated to the appropriate bimekizumab treatment group. Subjects randomized to bimekizumab will only be reallocated to the placebo treatment group if they did not receive bimekizumab at any time during the study.
Depending on the analysis sets, subjects will be summarized and listed based on the actual received treatment or according to the randomization which will be considered as planned treatment:
All subjects screened/ES: planned treatment
RS: planned treatment
SS: actual treatment
FAS: planned treatment
PPS: planned treatment
PK-PPS: actual treatment
PD-PPS: actual treatment
DBS: planned treatment (demographics, baseline characteristics, and efficacy analyses) or actual treatment
ESS: planned treatment
DBRS: planned treatment
cESS: planned treatment
cDBRS: planned treatment
REDACTEDplaceplace
ts willts wilhe ranhe ran
COPYlil
ny timy timab treab tre
bb
This do
cumen
taacann
otPPSPS
DBS:DBS:act
be a
S:S:us
edd tretre
actuactu
to eateatsu
pport
atmenttmen
any m
arketi
ng ndomdom
nned trened t
autho
rizati
oneatmeeatmbo treao trea
l be sube smi
appli
catio
n ou
d henchenng it is g it is
e, then, theen
and upsups
anysubjsub ex
tensio
ns
previoprevioSJC orJC objecje
or va
riatio
ns
SS SS
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 31 of 183
3.8 Center pooling strategyCenters will be pooled into geographic regions for analysis purposes. Centers will be grouped in the geographic regions North America and Europe.
3.9 Coding dictionariesMedical history and AEs will be coded using the Medical Dictionary for Regulatory Activities (MedDRA) version 19.0. Previous and ongoing medical history will be classified by primary system organ class (SOC) and preferred term (PT).
All AEs will be classified by primary SOC, high level term and PT. Prior and concomitant medications will be coded using version SEP2015 of the World Health Organization Drug Dictionary (WHO-DD) and will be classified by Anatomical Main Group, Pharmacological Subgroup, and PT. Medical procedures will not be coded.
3.10 Changes to protocol-defined analysesThe PK variables total body clearance (CL/F) and volume of distribution (V/F) will be analyzed separately. CL/F and V/F variables were removed from the PK variable list.
Following two subgroup analyses were added: time since first diagnosis of PsA and body weight.
Instead of stopping the pairwise testing of each bimekizumab dose versus placebo once it failed to reach significance at a significance level of α=0.05, the pairwise testing will continue and further pairwise comparisons are seen as non-significant.
No fixed sequence testing will be used for the secondary efficacy variables.
The other efficacy variable modified MDA was added to the analysis. The definition for the modified MDS is available in Section 8.3.1.
Time to a given response is defined as the length in weeks from Baseline until the first date when the response is achieved. In the definition the length was changed from days to weeks due toeasier interpretation.
Two addition analyses were added in the section analysis of other efficacy variables. A frequencies table counting the number of subjects with PSAID-9 ≤ 3 as well as cumulative distribution function (CDF) plots of PSAID-9 change from Baseline to Week 12 and Week 12 to Week 48 by treatment group.
4 STATISTICAL/ANALYTICAL ISSUES4.1 Adjustments for covariatesThe primary efficacy analysis and secondary analysis will be adjusted for treatment, geographic region, and prior TNF inhibitor exposure (yes/no).
4.2 Handling of dropouts or missing data4.2.1 Handling of missing data for efficacy analysisThe analysis for the binary primary, secondary, and other efficacy variables will use non-responder imputation (NRI) for handling missing data. In NRI, each subject with missing data or who has discontinued Double-Blind study treatment prior to Week 12 will be counted as a non-responder.
REDACTED ignign
the sethe se
MDA wMDA wonon 8.38.3
COPYekizuekizu0.05, t0.05,nifinif
This do
cumen
t egiogi
4.24.2ca
nnot primaprima
ion, on,
be us
ed me
STASTto(
ent ent sup
porre ade ad
ng the ng th(CDF)CDF
any
ddedd
marketi
ng .11..
as the ls the efinitiofiniti
autho
rizati
on the he
icant.cant.
econdcond
was awas a
appli
catio
n iaba
diagnodiagn
mab domab de paipa
and tionon
able ble
any
(V(V
exten
sionsita
n Drun Drumacoloacolo
or ant nt va
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 32 of 183
Sensitivity analysis will be performed for the secondary analysis of the primary variable (ACR50 response at Week 12) using MI assuming that missingness is missing at random (MAR). The MI method will be applied as follows:
1. Create a data set, sorted by treatment groups, of subjects with observed values and those needing estimation by multiple imputation. For the imputation step, missing values will be separated into non-monotone missing values (ie, intermittent missing values between completed assessments) and monotone missing values (ie, missing values after the patient dropped out). The procedure will sequentially estimate an imputation model for the ACR components at each post-Baseline visit where ACR components are collected, with geographic region, and prior TNF inhibitor exposure as covariates separated between the treatment groups.
For the imputation of intermediate missing values, the missing ACR components in each data set will be filled in using the Markov-Chain Monte Carlo (MCMC) method with multiple chains and monotone imputing. A total number of imputations will be 100. The seed used for these imputations will be 2017.
Note: All other multiple imputation procedures described in this SAP will use this same seed as well.
Note: The imputation model based on the MCMC method will only allow continuous variables as predictors. Therefore, prior TNF inhibitor exposure and geographic region will be re-coded as indicator variables (with values of 0 or 1 for each level of the variable).
Note: To avoid that imputed values are outside of the pre-defined range of values for the ACR components (eg, HAQ-DI [0-3]) maximum and minimum values for imputed variable values are specified.
If the intermediate missing data are imputed, the monotone missing data will be imputed for all patients with monotone regression including geographic region, and prior TNF inhibitor exposure as covariates. The dataset is the output dataset of the partial imputation. Since this dataset already has 100 imputed values at each visit, only one imputation will be performed.
Note: Maximum and minimum values are specified for imputed variable values to avoid values outside of the pre-defined range of values.
2. The ACR50 response will be calculated using the complete datasets. That is, the values of the ACR components at Week 12 based on the complete datasets will be compared to their corresponding Baseline values to calculate an ACR50 response for each subject (as described in Section 8.1.1). Each complete imputed data set will then be analyzed based on a logisticregression model with factors of treatment group, geographic region, and prior TNF inhibitor exposure as fixed effects.
3. The Week 12 results from the logistic regression analysis of each of the 100 imputed data sets will be combined for overall inference using Rubin’s rules, which account for the uncertainty associated with the imputed values (Rubin, 1987).
REDACTED(wit(wi
lues arlues arDI [0DI [0
COPYCMCCMCNF inNF inithith
This do
cumen
trr cann
ot coco
orresporrespinin SeS
be R5R5
ompompus
edutsidutsid
5050
tomummumd
supp
ortthis dhis d be perbe p
m a
any otoot
covariovarmark
eting
-33])]
data adata tone one
autho
rizati
onnhibhib
valuevalue
re oute out])) mm
appli
catio
nd in thin th
methomethbitorbito
and
hi
any MCM
ationsationex
tensio
nsweeneen
ompoompoMC)C)
or n thtva
riatio
ns
CR R
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 33 of 183
Some key points to consider relative to the calculation of the odds ratios and corresponding confidence intervals (CIs) are noted below:
As the estimates of the odds ratios from the logistic regression model in step 3 follow a lognormal distribution, a log transformation is needed to normalize these estimates since the procedures for combining results from multiple imputed datasets are based on the assumption that the statistics estimated from each imputed dataset are normally distributed. Therefore, the log of the odds ratio estimates from the logistic regression model are used when combining into a single inference (step 3). Additionally, the standard errors (SE) for the odds ratios are transformed as follows:
(1)
where UCL and LCL are the upper and lower confidence limit, respectively, for the CI of the odds ratio from the logistic regression model, and Zα/2 is the relevant critical value from the standard normal distribution (1.96 for a 95% CI). The estimates of the log odds ratio for each bimekizumab dose relative to placebo and the corresponding upper and lower confidence limits will be provided. The odds ratio is estimated by exponentiating the estimate of the log odds ratio. The confidence limits of the odds ratio are then estimated as follows:
(2)
(3)
where OR is the back-transformed estimate of the odds ratio just described, SE is the standard error of the log odds ratio and Zα/2 is the relevant critical value from the standard normal distribution (1.96 for a 95% CI). These calculations will be done such that odds ratios and corresponding CIs are calculated for the odds ratio of each bimekizumab dose versus placebo. Note that the p-values presented in the tables are not impacted by the transformations described above.
The supportive analysis number 3 (analysis of the ACR components), and all continuous secondary efficacy variables will use the same MI method as described above. Instead of using the logistic regression model the ANCOVA with treatment group, geographic region, and TNF inhibitor exposure as fixed effects and the Baseline values as covariate will be used. No log transformation is needed for the ANCOVA estimations.
Following Rubin (1987), multiple imputation estimates of descriptive statistics are computed by simply averaging the estimates from independent repetitions of the imputation algorithm:
(4)
where is the estimate of θ from the completed data set Berglund, 2014).
REDACTED
stimatstimatthth
COPY ateat
This do
cumen
tlgorgorcann
otwing Rng Rply avly avorithrith
beion on us
edssiosio
sure aure an isn is
to var
iononsu
pport
s numbnum
ariablriabl
anyed in ed inmark
eting
te e ooe relevrelev
hese calese cafor thefor th
th
autho
rizati
on
of tof
appli
catio
nper anper anng the eng the
as follo foll
and ritiiti
he loge loan
yly, foly, fical ca
exten
sions
or va
riatio
ns
nto a nto a
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 34 of 183
The imputation model will be applied for each treatment group separately. However, in the event there are computational challenges with the imputation model (eg, due to a standard deviation of 0 for responses of a given imputation), it is acceptable to modify the imputation model to include treatment as a variable in the model rather than running a separate model for each treatment group. It should be noted that doing so assumes that treatment does not interact with any of the other variables in the imputation model.
4.2.2 Handling of missing data for AEFor analyses of AEs, a complete date must be established in order to correctly identify the AE as occurring during treatment or not. For purposes of imputing missing components of partially-reported start and stop dates for AEs, the algorithms listed below will be followed. Start and stop dates of AEs will be displayed as reported in the subject data listings (ie, no imputed values will be displayed in data listings).
Partial AE start dates will be imputed as follows:
Imputation of Partial Start Dates:
If only the month and year are specified and the month and year of first dose is not the same as the month and year of the start date and the subject did not switch treatment during that month and year, then use the 1st of the month.
If only the month and year are specified and the month and year of first dose is the same as the month and year of the start, then use the date of first dose.
If only the month and year are specified and the subject did switch treatment during that month and year, then use the date of first switch treatment.
If only the year is specified, and the year of first dose is not the same as the year of the start date and the subject did not switch treatment during that year, then use the 1st of January of the year of the start date.
If only the year is specified, and the year of first dose is the same as the year of the start date, then use the date of first dose.
If only the year is specified, and the year of first dose is the same as the year the subject did switch treatment, then use the date of first switch treatment.
If only the year is specified, and the year of first dose is the same as the year of the start date and the same as the year the subject did switch treatment, then use the date of first dose.
If the start date is completely unknown and the stop date is unknown or not prior to the date of first dose, then use the date of first dose.
If the date of first study medication or switch treatment is partial, then the above imputation approach will be applied:
If only the day of first study medication or switch treatment administration is unknown, then this will be set to either the first of the month, or the day of first treatment allocation visit (excluding placebo run-in), whichever is later.
If both the day and month of first study medication or switch treatment are unknown, then this date will be imputed to be the date of first treatment allocation visit.
REDACTED e dae d
d and d and f first f first
the ythe y
COPY he mohe m
datedate
This do
cumen
t f thethapprappr
cann
ot e stasta
f first dfirst
he de d
be
tarttartus
edyear iear ime ame a
to , t
isu
pport
dod
pecifiepecifithen uhen
any, and, an
ose.ose
marketi
ngyear oyear h treatmh treat
d td
autho
rizati
ononthnth
e of firof fir
d the suthe suswitcswitc
appli
catio
near oar onot swnot sw
th anh an
and
off
any e
xtens
ions
artStart atart
uted vuted v
orialliallva
riatio
ns
AE aAE
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 35 of 183
If first treatment allocation visit date is missing then the all activities events will be assumedto be treatment emergent.
Imputation of Partial Stop Dates:
If only the month and year are specified, then use the last day of the month,
If only the year is specified, then use December 31st of that year.
If the stop date is completely unknown, do not impute the stop date.
In the event of ambiguity or incomplete data which makes it impossible to determine whether the AE was treatment emergent, the AE will be considered treatment emergent.
If the intensity of an AE is unknown, it is considered as severe.
If the relationship to study drug is missing, it is considered as related.
The imputation rule for missing seriousness differ between the interim and final analysis. For the interim analysis no imputations rule will be applied. For the final analysis the worst case approach will be applied. If the seriousness of an AE is missing for the final analysis, it is considered as serious.
4.2.3 Handling of missing data for prior and concomitant medicationAny medications with incomplete start and end dates/times will be handled according to the following rules for classification as prior and concomitant. Such imputations will only be performed for these classifications and calculations; in the listings all data will be shown as recorded on the electronic case report form (eCRF).
Imputation of partial start dates:
If only the month and year are specified and the month and year of first dose is not the same as the month and year of the start date and the subject did not switch treatment during that month and year, then use the 1st of the month.
If only the month and year are specified and the month and year of first dose is the same as the month and year of the start, then use the date of first dose.
If only the month and year are specified and the subject did switch treatment during that month and year, then use the date of first switch treatment.
If only the year is specified, and the year of first dose is not the same as the year of the start date and the subject did not switch treatment during that year, then use the 1st of January of the year of the start date.
If only the year is specified, and the year of first dose is the same as the year of the start date, then use the date of first dose.
If only the year is specified, and the year of first dose is the same as the year the subject did switch treatment, then use the date of first switch treatment.
If only the year is specified, and the year of first dose is the same as the year of the start date and the same as the year the subject did switch treatment, then use the date of first dose.
REDACTEDon
ulationulationm (eCm (eC
ecec
COPY r rtes/timtes/t
ncomcom
This do
cumen
tththcan
notyear ear
If onlIf onltheh
bed thed thr oo
used
ar
e year yeare
tor, the, th supp
ortearar
f the stthe s
and yand y
any
r are r are mark
eting
cified aified a
art datet date1stst of of
autho
rizati
on me
mitantmitantns; in s; in
CRF).RF).
appli
catio
n r th
d conccones wies wi
andlysisysis
the fhe fan
ynd finnd fins ths th
exten
sions
oretheth va
riatio
ns
herher
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 36 of 183
If only the year is specified and the end date is before date of first dose, then set the start date to the 1st of January of the year of the start date.
If only the year and day are specified and month is missing, then only the year will be considered and the month and day will be imputed with the rules above
If the start date is completely unknown and the stop date is unknown or not prior to the date of first dose, then use the date of first dose.
If the start date is completely unknown and the stop date is prior to the date of first dose, then set the start date to the 1st of January of the year of the end date.
If the date of first study medication or switch treatment is partial, then the above imputation approach will be applied:
If only the day of first study medication or switch treatment administration is unknown, then this will be set to either the first of the month, or the day of first treatment allocation visit (excluding placebo run-in), whichever is later.
If both the day and month of first study medication or switch treatment are unknown, then this date will be imputed to be the date of first treatment allocation visit.
Imputation of Partial Stop Dates:
If only the month and year are specified, then use the last day of the month,
If only the year is specified, then use December 31st of that year.
If the stop date is completely unknown, do not impute the stop date.
There will be no imputation of any other missing data.
4.3 Interim analysis and data monitoringTwo interim analyses are planned for the study after the subjects have completed 12 and 48 weeks.
4.3.1 Interim analysis Week 12After all enrolled subjects have completed the 12 weeks Double-Blind Period, an interim analysis will be performed (i) to analyze the dose:exposure response for ACR and PASI response criteria to determine the optimal therapeutic dose(s) for subsequent studies, and (ii) to perform a comprehensive evaluation of all Double-Blind data of the study.
No separate SAP for the interim analysis will be provided. The interim analysis is a limited version of the final analysis and will focus on specific topics, eg, primary and secondary efficacy analysis and comprise the results of the 12 weeks Double-Blind Period. The TFL shells for the interim and final analysis will be provided in two different documents.
The interim analysis will summarize disposition, demographics, PsA history, Baseline characteristics, PsA concomitant medication, efficacy analysis, AEs and treatment-emergent AEs (derivation Section 10.2), hematology and biochemistry laboratory data, and Electronic Columbia-Suicide Severity Rating Scale (eC-SSRS). Not all collected hematology and biochemistry laboratory variables will be summarized. For the interim analysis only selected
REDACTEDcembecemb
wn, down, do
other mother m
COPY use thuse t
This do
cumen
tinternter
ThT
cann
ot ateat
n of thn of thlysis aysis rimi
be
e SAe SAus
edminemine
ve evae evatorformfor
e thtsu
pportnalysalys
ects hacts hrmedm
any
sis
marketi
ng mismis
nd nd ddfor thfor th
autho
rizati
onhe le l
er 31er 31stst
o not inot i
ssinsin
appli
catio
nreateaocationocatio
lastlast
and
atmtm
anyment ment ex
tensio
nsmputatimputat
on is un is u
or va
riatio
ns
e, the, the
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 37 of 183
hematology and biochemistry variables will be provided as listed in Table 12‒4. Listings of AEs, hematology, biochemistry and C-SSRS data will be provided.
The interim efficacy analysis will focus on the primary and secondary efficacy analyses including supportive and sensitivity analysis. In addition, summary tables for the primary and secondary efficacy variables, the ACR components, and MDA will be provided. Categorical variables will be summarized using frequency tables by each visit. Categorical variables will be summarized using frequency tables by each visit. Continuous variables will be summarized using descriptive statistics by each visit. Time to onset of ACR50 response and ACR20 response will be summarized and plotted as described in Section 8.3.
PK and AbAb data will be analyzed in the interim analysis as described in Section 9.1 and Section 9.2 .Biomarker data will not be analyzed in the interim analysis.
The interim analysis includes selected data up to Week 12 and in addition data up to Week 24. The data up to Week 24 may include partial results up to the time the database for Week 12 was locked for ongoing subjects. Following variables will be analyzed: primary and secondary efficacy variables, ACR components, MDA, PK, AbAb, TEAEs, biochemistry and hematology laboratory results, and Hy’s Law. It depends on the variable how those additional time points will be presented in the interim analysis.
The efficacy analyses and the plasma concentration analyses for the period between Week 12 and Week 24 are only for those treatment groups that are not rerandomized. Two summary tables for each of the primary and secondary efficacy variables, the ACR components, and MDA will be presented. The first table will include data up to Week 12 by treatment group at Baseline. The second table will include data up to Week 24 by the following treatment groups: bimekizumab 160mg and bimekizumab 160mg with loading dose 320mg, and bimekizumab 320mg. The summary table for plasma concentration will present only data up to Week 12 for the Baseline treatment groups Placebo and bimekizumab 16mg. Data up to Week 24 will be presented for bimekizumab 160mg, bimekizumab 160mg with loading dose 320mg, and bimekizumab 320mg. All Listings which include data up to Week 24 will use the same approach as for plasma concentration tables.
The safety analysis (TEAEs, biochemistry and hematology laboratory results, and Hy’s Law) will be handled as follows: Two summary tables for each of the safety analysis will be presented. The first table will include data up to Week 12 by treatment group at Baseline. The second table will include data up to Week 24 and will also present subjects under their Baseline treatment group. For the second table any event/finding that is unique to a treatment group should only be presented in the “All subjects” column and the treatment columns should remain blank. Following rules will be applied for the TEAEs: If the relative TEAE start date is less or equal to 84 (12*7) days, then the TEAE will be presented in the up to Week 12 table. If the relative start date is less or equal to 168 (24*7) days, then the TEAE will be presented in the up to Week 24 table.
The treatment group information will not be displayed for the interim listings.
If an AE occurs on the same day as the treatment switch (Visit 7, Week 12) then the AE will be allocated to the Double-Blind treatment. An exception from this general rule is made for adverse drug reactions (ADRs), and AE of hypersensitivity reactions and anaphylactic reactions. If those
REDACTEDy varvata up tta up t
k 24 bk 24 bh loadih load
ation ation
COPYanalyanalthat athat a
ariaaria
This do
cumen
tdate ate tabltab
cann
otwing ruing r12*7)12*7)
e isis
behe
n then thus
eda up up e secosec
toinclinc supp
ort
AEs, biAEs, blows: Tows: lud
anyup up mark
etingn will pwill
zumabzumabab 160mb 160p to Wp to
autho
rizati
onare ne n
ablbles, es, to Weo We
by by thetheng dong do
appli
catio
nw thosthos
ses fores fonot not
and marma
ochemcheman
ybasebasery anry
exten
sions9.11 anan
ta up a up e
or va
riatio
ns
sing sing e will will
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 38 of 183
AEs occur on the same days as the treatment switch, the AEs will be allocated to the Dose-Blind treatment and will not be presented for the interim analysis.
Exposure of study medication will not be summarized but a listing will be provided including following information: date of first dose, date of last dose, and duration of exposure. The last dose for the Double-Blind Period is planned on Visit 6 (Week 8).
For the interim analysis, the database will be locked and the treatment codes will be made available to UCB personnel with exception of operational staff working on the study. An interim report will be written. The Investigators and subjects will remain blind to the assigned bimekizumab dosing regimen until the subject completes the Dose-Blind Period at Week 48.Further details are available in the PA0008 blinding plan.
Safety data will be provided to an independent Data Monitoring Committee (DMC). The DMC will review those safety data periodically. The composition and operation of the DMC will be defined in the DMC charter. The presentation and analysis of data for the DMC meetings is described within a separate DMC SAP.
4.3.1.1 Changes from interim analysis to SAP-defined analysesFor TEAE two summary tables were presented as described in previous section with one modification: the first table was displaying all TEAE during the Double-Blind Period instead of limiting it TEAE with a relative TEAE start date less equal 84 days.
The definitions of AE of special monitoring were updated with the third SAP Amendment after the interim analysis. The interim analysis used the previous definition which is displayed in Section 4.3.1.2. The interim analysis only displayed the overall number of AE of special monitoring by treatment group.
The baseline definition for CRP and hs-CRP measurements were updated with the fourth SAP Amendment after the interim analysis. The interim analysis used the previous definition: The last valid measurement before first dose of study drug (Visit 2) will be used as the Baseline value.
4.3.1.2 Search and selection criteria for AE of special monitoring for interim analysis
Following AEs are defined as TEAEs of special monitoring:
Fungal infectious disorder
Opportunistic infection (including TB)
Malignant or unspecified tumor
Malignant tumor
Major cardiovascular event
Haematopoietic cytopenias
Neuropsychiatric events
Inflammatory bowel disease
Hypersensitivity and anaphylactic reaction
ere ereed theed the
y disply disp
hshs
COPYdudu
ss equss equ
e upup
This do
cumen
tMMcann
ot gn
Malignalign
MaM
benant onant us
ed ou
stic intic itous dus dsu
pport
fined ained
isi
anyselecselecmark
eting
CRP mCRP
is. Thes. These of sse of
autho
rizati
on
pdateddatede prevprev
layed layed
appli
catio
n nedn previoprevi
g the Dthe Dal 84 al 84
and
ed ad aan
yDMCDM exten
sions
e
MC). MC). the Dthe D
MCMC
ork 48k 4va
riatio
ns
erim rim
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 39 of 183
Hepatic events
Fungal Infectious disorderAll TEAEs which code into the High Level Group Term (HLGT) of “Fungal infectious disorders” are classified as fungal infectious disorder.
Opportunistic infectionAll TEAEs identified using UCB-defined search criteria as described in Section 12.13.
Malignant or unspecified tumorThe search criteria is based on the Standardized MedDRA Query (SMQ) = “Malignant or unspecified tumours (SMQ)”.
Malignant tumorThe search criteria is based on the SMQ=”Malignant tumours (SMQ)”.
Major cardiovascular eventsThe major cardiovascular events are identified using the following UCB-defined search criteria:
All serious TEAEs which code to a PT included in the Scope=Broad and/or Scope=Narrow groups of the following SMQs:
Haemorrhagic central nervous system vascular conditions (SMQ)
Ischaemic central nervous system vascular conditions (SMQ)
All serious TEAEs which code to a PT included in the HLT “Ischaemic coronary artery disorders” except events coding to PT “Chest Pain” or “Chest discomfort”
All serious TEAEs which code to a PT included in any of the following HLTs: “Heart Failures NEC”, “Left Ventricular Failures”, or “Right Ventricular Failures” and which also code to the SOC of “Cardiac Disorders” as Primary SOC.
Haematopoietic cytopenias The search criteria is based on the SMQ = “Haematopoietic cytopenias”. The SMQ search should include all serious TEAEs which code to a PT included in the Scope=Broad and/or Scope=Narrow groups within the SMQ.
Neuropsychiatric eventsThe search criteria is based on the SMQ = “Depression and suicide/self-injury (SMQ)”. The SMQ search will include all TEAEs which code to a PT included in the Scope=Broad and/or Scope=Narrow groups within the SMQ.
Inflammatory bowel diseaseAll TEAEs which code into the HLT of “Colitis (excl infective)” are identified as inflammatory bowel disease.
Hypersensitivity reactions and anaphylactic reactionsHypersensitivity reactions and anaphylactic reactions will be identified as follow:
REDACTED asculaascula
PT inPT into PT to PT
COPY scularcula
This do
cumen
tInflanflaAA
cann
ot ch h
searchearchpe=Npe=N
be
critcritus
ed rouou
hiatriiatrio
riups ups
supp
ort niania
ased onsed oious Tous T
any DD
as asmark
etingCC
a PT ia PT lar Faiar Fa
DisorDiso
autho
rizati
onar conar con
ar condr cond
ncludecludeCheChe
appli
catio
nng UCng UC
cope=Bcope=
d
and
CC
any e
xtens
ionsant ornt o or
varia
tions
there
of.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 40 of 183
a) Hypersensitivity reactions: All TEAEs that either emerged on the same day as when a study medication injection reaction was received, or that emerged one day after a study medication injection was received, which code to a PT which contains the term “hypersensitivity” will be considered to be a hypersensitivity reaction and included in the summary table as such.
b) Anaphylactic reactions: An algorithmic approach will be used to identify TEAEs that are considered to be anaphylactic reactions. PTs are separated into the 4 distinct categories (A, B, C, D) prior to the algorithmic approach being applied.
All TEAEs that either emerged on the same day as when a study medication injection reaction was received, or that emerged one day after a study medication injection was received, and which fulfill any of the 3 criteria described in Section 12.14 will be included in the summary table.
Hepatic eventsThe search criteria is based on all TEAEs in the SMQ “Drug related hepatic disorders -comprehensive search (SMQ)”. Note that the following two sub-SMQs are to be excluded: “Liver neoplasms, benign (incl cysts and polyps)” and “Liver neoplasms, malignant and unspecified (SMQ)”.
The SMQ search should include all TEAEs (regardless of whether they have been judged as related to study medication or not) which code to a PT included in the Scope=Narrow group within each SMQ.
4.3.2 Interim analysis Week 48After all enrolled subjects have completed the Week 48 or Early termination visit, a second interim analysis will be performed to analyze the key efficacy and safety data for the whole treatment period.
No separate SAP for that interim analysis will be provided. The interim analysis is a subset of the final analysis and will focus on the primary and secondary efficacy analysis. The TFL shells for the final analysis will be used.
The snapshot for the PA0008 interim analysis Week 48 will occur before all subjects have completed the PA0008 study. Specifically, subjects who do not enter the extension study will need to complete the SFU Visit, which occurs 20 weeks after the last dose of study treatment. The Week 48 interim database lock will be performed based on the last subject completing the Week 48 visit. It is anticipated that there will be some subjects still awaiting the SFU at that time. The number of subjects in the SFU period is expected to be low, and the minimal data to be collected from their SFU visits is not considered mandatory to evaluation of the key efficacy and safety objectives of the study.
4.3.2.1 Changes from interim analysis to SAP-defined analysesThe baseline definition for CRP and hs-CRP measurements were updated with the fourth SAP Amendment after the interim analysis. The interim analysis used the previous definition: The last valid measurement before first dose of study drug (Visit 2) will be used as the Baseline value.
REDACTED 4848
pleted pleted o ano an
COPYdless odless oo a PTa PT
This do
cumen
t afefe
4.34.3ca
nnothe nue nu
cted fcted fety oty o
beisit. isit.umum
used
e thetheinteriinterItIt
to008008e Se S
supp
ort sed.ed.
A0008A0008 stu8 st
anyn the n themark
eting
thnalyze alyze
analyanaly
autho
rizati
onT incT inc
he Whe W
appli
catio
n MQQ
eoplasoplas
f whetwhel
andhepatepa
Qs aQs an
y ticti
exten
sionsre
ved, aned, anhe sumhe sum
oractiactva
riatio
ns
s
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 41 of 183
4.3.2.2 Changes during interim analysis to SAP-defined analysesAt the time of the Week 48 interim it was discovered that the DBRS and ESS were defined incorrectly from the original intent. The original intent of the ESS analysis set was to have only those subjects that were eligible to receive rescue therapy, that went on to actually receive rescue therapy, with all other subjects remaining in the DBRS. As such, two extra sets, the cDBRS and cESS have been added. These analysis sets will correctly select out those subjects that were eligible and received rescue medication. For the final analysis, all outputs created on the DBRS and ESS will also be produced on the cDBRS and cESS to give full transparency of the difference. Further all outputs using the DBRS and ESS will be repeated on the DBS and this analysis will be used as the main analysis of the Dose-Blind as this analysis set is the most unbiased set.
4.4 Multicenter studiesThe data from all centers will be pooled for the purposes of the analysis. There will be no formal statistical evaluation of the effect of center on the results obtained.
4.5 Multiple comparisons/multiplicityIn order to control for the overall Type I error rate, the pairwise comparisons of bimekizumab will be formally evaluated for statistical significance only if the primary efficacy analysis is statistically significant at the two-sided 5% level. In addition, the pairwise comparisons will follow a sequential testing sequence and the formal evaluation of statistical significance of each comparison is dependent upon the previous comparison achieving statistical significance at the two-sided 5% level. If the sequential testing fails to reach significance at a significance level of α=0.05, then the pairwise testing will continue and the comparison are seen as non-significant. The p-values will be displayed as nominal p-values. More details can be found in Section 8.1.3.
4.6 Use of an efficacy subset of subjectsThe primary dose-response analysis and pairwise comparisons will be repeated for (1) all subjects in the PPS, and (2) for all subjects in the RS as a supportive analysis. The purpose is (1) to evaluate the effect of important protocol deviations on the analysis, and (2) to evaluate the consistency of the FAS with the intent-to-treat (ITT) principle.
4.7 Active-control studies intended to show equivalenceNot applicable.
4.8 Examination of subgroupsThe following variables for subgroup analyses will be used:
Age (<45 years, ≥45 to <65 years, ≥65 years)
Gender (male, female)
Geographic region (North America, Europe)
BASDAI (<4 [mild disease]; ≥4 to ≤7 [moderate disease]; >7 to ≤10 [severe disease])
Prior TNF inhibitor exposure (yes, no)
Treatment-emergent AbAb status (positive, negative)
REDACTEDcompacompng failg fail
ontinuontinuominalominal
COPY onlonn addin add
mal evmal ev
This do
cumen
t Gca
nnotllowinowi
Age Age
be us
ed
ExEx
to vesu
pport
portortwith thwith t
ee--cco
anyr all sall
rtanttan
marketi
ng al pp-
y subsy subsis andis an
autho
rizati
onvaluavaluaarison arison
ls to rls to rue andue andvava
appli
catio
n se come com
f the pthe ption, tion,
i
and a
nyThereTher exten
sionse momo ord thd thostt
varia
tions
RS RS
hishis
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 42 of 183
Extent of psoriasis involvement (<3%, ≥3% to <10%, ≥10%)
hs-CRP level (<= upper limit normal, > upper limit normal)
Time since first diagnosis of PsA (< 2 years, ≥ 2 years)
Body weight (<100 kg, ≥100 kg)
Subjects will be counted to have a positive treatment-emergent AbAb status in case the first AbAb positivity occurred up to Visit 7 (Week 12). The definition of first AbAb occurrence is described in Section 9.2.
The derivation of time since first diagnosis of PsA is described in Section 6.2.
5 STUDY POPULATION CHARACTERISTICS5.1 Subject dispositionSummary tables of all subjects screened will be presented including reason for screen failures and disposition of subjects screened. The disposition of subjects screened will include the number of subjects included in each analysis set (ES, RS, SS, FAS, PPS, PK-PPS, PD-PPS, DBS, ESS, and DBRS) overall and by site.
The number and percentage of subjects who discontinued study medication, who discontinued the study, and who discontinued due to AEs will be summarized for subjects in the RS and DBS. The table using RS will analyze subjects for the complete study as well as the Double-Blind Period. Completed study is defined as completed the 12-week Double-Blind Period and the 36-week Dose-Blind Period including the safety follow-up visit. Subjects will be allocated to the treatment group at Baseline. The table using DBS will only summarize subjects in the Dose-Blind Period.
Study eligibility criteria will be listed and a separate listing of subjects who did not meet the eligibility criteria will be presented.
Subject disposition will be listed for all subjects screened and will include, inter alia, the following information: subject status, date of informed consent, randomization, first and last study medication, and primary reason for premature study termination. Listings of subject analysis sets, study and study medication discontinuation, and visit dates will be presented by subject including the relative study day (calculated as described in Section 3.2.1) for each visit
5.2 Protocol deviationsImportant protocol deviations will be identified and classified by the deviation types. Following different important protocol deviation types will be classified: inclusion criteria deviation, exclusion criteria deviation, withdrawal criteria deviation, prohibited concomitant medication use, incorrect treatment or dose, treatment non-compliance, and procedural non-compliance.
Summary tables of the number and percentage of subjects with an important protocol deviation will be provided including a summary of subjects excluded from the PPS, PK-PPS, and PD-PPS due to important protocol deviations. Summary tables will be presented for FAS and Baseline treatment groups.
A listing of all important protocol deviations identified at the data evaluation meeting will be presented by subject for all subjects in the RS, and will include deviation type and description.
REDACTEDeted teted fety foety fo
using using
COPY nuenue sue sum
compcomp
This do
cumen
t se, e,
SuSu
cann
otnt imt imusion sion incinc
beprotoprotus
ed
ProtProttohe rehe r su
pport
ject ectrimary imary
nd studd studelal
anyed fod fo
t stat st
marketi
ng
d and a and ad.
or
autho
rizati
onplete plete the 12he 12
ollowollowDBS DBS
appli
catio
n S, P, P
study mudy marizemariz
andened ned
PPSPPS
any on foron fo
w
exten
sions
or va
riatio
ns
is
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 43 of 183
6 DEMOGRAPHICS AND OTHER BASELINE CHARACTERISTICS
6.1 DemographicsTables with descriptive statistics (number of subjects, mean, SD, minimum, median, and maximum) and listings will be given for the demographic variables age (at time of informed consent), gender, racial group, ethnicity, weight at Screening, height at Screening, body mass index (BMI), country, and geographic region. Age and BMI will be summarized as a continuous variable and as categorical variables. Age will be combined based on the following categories: ≤18, 19-<65, ≥65 years (clinicaltrials.gov requirement), 18-<65, 65-<85, ≥85 years (EudraCT requirement), and 18-<45, ≥45-<65, ≥65 years. BMI will be summarized based on the following categories: <25, 25-<30, ≥30 kg/m2.
BMI in kg/m2 is calculated based on the height (in m) and the weight (in kg) using the formula:
For adults:
(5)
Even if they are available in the database, these variables are calculated during analysis (if applicable). These calculated values are used in the statistical analysis since they are considered more accurate.
A frequency table for Lifestyle will be presented as well as a corresponding listing. Data collected in the CRF for subject’s childbearing potential and birth control will only be listed.
The summary tables will be performed on the SS and repeated using the FAS and the DBS. If the SS and FAS analysis sets are identical the summaries will not be repeated. The Listing will be provided for all subjects screened, except the Lifestyle listing will use the RS.
6.2 Other Baseline characteristicsPsA history will be summarized for subjects in the FAS and SS including the time since first diagnosis of PsA, age at first diagnosis date, and PsA subtype (polyarticular – symmetric arthritis, oligoarticular – asymmetric arthritis, distal interphalangeal, joint predominant, spondylitis predominant, arthritis mutilans). Time since first diagnosis will be summarized as a continuous variable and as categorical variables based on following categories: <2, ≥2 years. The history will be listed for all subjects in the RS.
Time since first diagnosis of PsA will be calculated as:
(6)
If the date of diagnosis is partial, it should be imputed to the most recent feasible date (ie, last day of the month if only day is missing, or the last day of the year if day and month are missing).
Age at first diagnosis will be calculated as:
REDACTED esentesente
dbearindbeari
med omed o
COPYiablesablehe stahe sta
This do
cumen
t can
notince fnce be
l be l be used
ominminriableiablelili
to larinanina
supp
ortmarizedarizeat first t firs
rr –– a
anyne cne
d
marketi
ngon ton theheal the sl the
excepexcep
autho
rizati
onatistitisti
ed as wed as wing pong po
appli
catio
n
are caare cica
and a
ny )) ex
tensio
nsEudrud
thethe fof
usinusin
or orior
draCdraCva
riatio
ns
nuous uous riesries
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 44 of 183
(7)
These formulas may result in incorrect ages after rounding if the birth day falls on the date of first diagnosis. In that case the age is calculated as the number of years between the year of birth and the year of the randomization visit. The age for individual subjects is presented with 1 decimal. The rounding is done downwards. For subjects enrolled at German sites, only the year of birth may be entered into the eCRF for this study. For these subjects age will be calculated after imputing their date of birth to be 01 Jan XXXX.
Baseline characteristics (including Scores relevant for inclusion and exclusion criteria) will be summarized by treatment group and overall for FAS, SS, and DBS. Following variables will be summarized:
TJC
SJC
hs-CRP
Rheumatoid factor
Anti-CCP antibodies
Prior NSAID therapy, and prior anti-TNF therapy
Psoriasis BSA
Nail psoriasis
Dactylitis
NSAID therapy
Synthetic DMARDs
MTX
Sulfasalazine
Hydroxychloroquine
A rheumatoid factor value < 15 IU/mL is defined as negative and a value ≥ 15 IU/mL is defined as positive. Anti-CCP antibodies negative is a value < 5 U/mL and a value ≥ 5 U/mL is defined as positive.
For hs-CRP following rules will be applied:
If Baseline hs-CRP values are missing, then take the Screening hs-CRP values.
If Screening hs-CRP values are missing, then take the median hs-CRP values.
The corresponding listings will be presented for the RS.
REDACTED herhe COPY
erapyrapy
This do
cumen
t can
not ive.ve.
hss--CRCR
be. An. An us
ed oq
factofactontiti
toquinquinsup
port a
ny m
arketi
ng au
thoriz
ation
yy
appli
catio
n and
any e
xtens
ionsia) wa) w
ariableriable
or i
varia
tions
ar
ed d
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 45 of 183
6.3 Medical history and concomitant diseasesMedical history and ongoing medical conditions will be summarized by MedDRA SOC and PT by treatment and overall including the number and percentage of subjects with each condition. The denominator for the percentages will be the number of subjects in the SS and FAS for each treatment or overall.
Medical history and ongoing medical conditions will be listed by treatment and subject including the reported term, PT, and SOC for the RS. The start date (month and year only) and end date (or ongoing if applicable) will also be included in the listing. A glossary of all medical history conditions will also be presented including the reported term, PT and SOC.
6.4 Prior and concomitant medicationsPrior medications include any medications that started and ended prior to the start date of study medication. Concomitant medications are medications taken at least one day in common with the study medication dosing period. For bimekizumab, the dosing period is typically from the date offirst dose up to (but not including) one dosing interval post last dose. Thus, a concomitant medication is any medication whose start date is on or after the date of first study medication and prior to the date of last study medication administration + 28 days (1 dosing interval), or whose stop date is either missing, or on or after the date of first study medication administration.
In the case of missing data, the classification of medications as prior or concomitant will be performed as described in Section 4.2.3.
The number and percentage of subjects taking prior medications will be summarized by treatment group, overall and by anatomical therapeutic chemical (ATC) class, presenting Anatomical Main Group (ATC Level 1), Pharmacological Subgroup (ATC level 3), and PT.
The number and percentage of subjects taking concomitant medications will be summarized similarly. Tables for prior and concomitant medications will be presented for SS and FAS.
Prior and concomitant medications will be listed by treatment and subject for RS.
6.5 Prohibited medication and rescue medicationProhibited medications are defined in the Protocol (Section 7.8.2).
Rescue medication will be allowed for some subjects evaluated at weeks 16, 24, and 36. Further details are defined in the Protocol (Section 5.1.4). The number and percentage of subjects who used rescue medication will be summarized and listed.
The SS and FAS will be used for both summary tables and the RS will be used for the listings.
7 MEASUREMENTS OF TREATMENT COMPLIANCECompliance will be summarized as the number of doses received relative to the number of dosesscheduled:
(8)
where the total number of expected doses is derived relative to when the subject finishes treatment. If a subject completes treatment 12 doses are expected (Baseline and every 4th week
REDACTED ng pring pri
cal theal the1), Ph1), P
ctscts
COPY edicatiedicat
This do
cumen
tComomschsch
cann
ot ndnd
mp
be
d FAd FAus
eded ind in
medicaedicatowilwil
n tnsu
pport
d mmare deare de
ill bell
any nsn
medmemark
eting
s takingtaking
omitantmita
ns wils w
autho
rizati
on o
ior meor meerarapeupeu
armaarma
appli
catio
nate oe oays (1 dys (1
dy meddy me
ons asns as
andThuThu
of fof f
any y iy
ypicalypicas
exten
sions
start dtart din cin
or yy va
riatio
ns
ng ng te (or e (or
y
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 46 of 183
afterwards until Week 44). If a subject discontinues early, then the number of expected doses isbased on the time of early discontinuation relative to the dosing visits. If a dose is not completely given at a specific visit (eg, subject only received one injection instead of the two planned injections), then the subject will be considered to have no compliance for the visit. In the formula above it will be counted as no dose received at this visit.
A summary of percent treatment compliance categorized as ≤80% and >80% will be provided by treatment group for the overall treatment period as well as for Double-Blind treatment period.For the Double-Blind treatment period compliance will refer to the first 12 weeks and will be presented by treatment received at baseline. For the overall treatment period, the compliance will be calculated for the following three groups: bimekizumab 160mg and 160mg with loading dose, bimekizumab 320mg, and all bimekizumab. Treatment compliance for the bimekizumab 160mg and 160mg with loading dose group will be calculated for the time the subject receives 160mg or 160mg with loading dose, eg for a subject who switches from Placebo or bimekizumab 16mg to bimekizumab 160mg at Week 12, the compliance will only be calculated for the time the subject receives bimekizumab 160mg. The same approach will be done for bimekizumab 320mg. The all bimekizumab group will consist of all the doses of bimekizumab including bimekizumab 16mg, and will only exclude the time subjects receive Placebo.
A by-subject listing of treatment compliance will be provided, presenting. percent compliance and numbers of expected and received doses for the Double-Blind Period and for treatment with any dose of bimekizumab.
8 EFFICACY ANALYSES8.1 Statistical analysis of the primary efficacy variable8.1.1 Derivations of primary ACR score and responseACR20/50/70 response represents at least 20%, 50%, or 70% respective improvement from Baseline for each of the following:
TJC (based on 78 joints)
SJC (based on 76 joints)
At least 3 of the 5 remaining core set measures
Disease activity as assessed by PGADA
Disease activity as assessed by PhGADA
Pain as assessed by PtAAP
Physical function as assessed by the HAQ-DI
Acute phase response as assessed by the hs-CRP
For all non-missed visits, if any of the component scores are missing, then those scores will be considered as not having met the criteria for improvement for the respective component. For component scores with a Baseline value of 0, the percentage improvement is not calculable and is treated as missing. However, if ACR20, ACR50, or ACR70 response is used as the primary endpoint in a study, consideration should be given to requiring a value >0 for each of the component endpoints in order to enter the study. For component scores with a Baseline value >0
REDACTED SSof thof th
marymary
COPYDouDou
This do
cumen
t FF
cann
otPainain
PhPh
be as
n asn aus
edactivtiv
se ace ac
to ivityvitysu
pport
nts)ts)
remairemai
any m
arketi
ngy ACAC
least 2least
autho
rizati
on
he e pprirCRCR
appli
catio
n
edd, pres, preblele--BliBl
andmekmek
cludinludinan
yfor thfor tkizukiz
exten
sionsoadinadi
zumabzumareceivreceiv
mekizekizh
orancancng
varia
tions
y
be e cece
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 47 of 183
and a post-Baseline value of 0, percentage improvement will be calculated as for any other non-missing data.
Missing items for TJC, SJC, and HAQ-DI will be handled as described in Section 8.1.1.1 and Section 8.1.1.5.
8.1.1.1 Tender joint count (TJC) and swollen joint count (SJC)8.1.1.1.1 78/76 joint evaluation for ACR responseThe Principal Investigator, another delegated physician, or an appropriately qualified medical professional (based on local requirements) who has had documented training on how to perform these assessments correctly assesses the following joints for tenderness: the distal interphalangeal, proximal interphalangeal and metacarpophalangeal joints of the hands, and metatarsophalangeal joints of the feet, the carpometacarpal and wrist joints (counted separately), the elbows, shoulders, acromioclavicular, sternoclavicular, hip, knee, talo-tibial, and midtarsaljoints. All of these except for the hips are assessed for swelling.
Artificial and ankylosed joints, as well as missing joints (ie, amputated joints), are excluded from both tenderness and swelling assessments.
One dactylitic digit is to be counted as 1 swollen joint (instead of counting as 3 in the finger or 2 in the toe).
Table 8‒1 summarizes the swelling and tenderness grading criteria.
Table 8‒1: Swelling and tenderness grading
The tender joint counts (TJC) and swollen joint counts (SJC) are weighted joint counts. If there are missing observations in the tender or swollen joint assessments, then the remaining observations will be assessed and weighted by the number of the assessed joints (AJ):
(9)
(10)
where n is the number of total joints. If a joint is missing at Baseline, then that joint is set to missing throughout the study. If more than 50% of the tender joint assessments or 50% of the
CTED s gras gra
REDACTCOPY
ss grass gra
This do
cumen
t can
not b
e use
d rvava
will bell betont
ationatiosu
ppts (TJs (TJppo
rt any
mark
eting
ho
rizati
onadingding
adingding
autho
appli
catio
n ad of coad of c
g c
andted joied jo
any e
xtens
ions
handshandsountedunted
bial, ial,
orperper varia
tions
ical cal f
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 48 of 183
swollen joint assessments are missing at the time of a given assessment, then no imputation will be done and the total TJC or SJC will be set to missing for that visit.
Injected joints will be counted as swollen and tender from the date of injection up to 52 weeks after injection, ie, swelling and tenderness grading is 1.
8.1.1.1.2 28 joint evaluation for determination of DAS28(CRP)The following 28 joints will be used for calculation of the DAS28(CRP).
Upper extremity (26)-bilateral shoulders, elbows, wrists (includes radiocarpal, and carpal and carpometacarpal bones considered as a single unit), MCP I, II, III, IV, and V, thumb interphalangeals (IP), PIP II, III, IV, and V
Lower extremity (2)-knees
Injected joints will be counted as swollen and tender from the date of injection up to 52 weeks after injection, ie, swelling and tenderness grading is 1.
8.1.1.2 Patient’s Global Assessment of Disease Activity (PGADA)Subjects will score their Global Assessment of Disease Activity using a visual analog scale (VAS) where 0 is “very good, no symptoms” and 100 is “very poor, severe symptoms”. The subject should be asked to consider both joint and skin components in their response to this question.
8.1.1.3 Physician’s Global Assessment of Disease Activity (PhGADA)The Investigator will assess the overall status of the subject with respect to their PsA signs and symptoms and functional capacity (considering both joint and skin components) using a numerical rating scale (NRS) where 0 is “very good, asymptomatic and no limitation of normal activities” and 100 is “very poor, very severe symptoms which are intolerable and inability to carry out all normal activities”.
8.1.1.4 Patient’s Assessment of Arthritis Pain (PtAAP)The PtAAP VAS is part of the ACR core set of measures in arthritis (Felson et al, 1993). Subjects will assess their arthritis pain using a VAS where 0 is “no pain” and 100 is “most severe pain”.
8.1.1.5 Health Assessment Questionnaire-Disability Index (HAQ-DI) scoreThe HAQ-DI contains 20 items divided into 8 domains that measure: dressing and grooming (2items), arising (2 items), eating (3 items), walking (2 items), hygiene (3 items), reach (2 items),grip (3 items), and common daily activities (3 items). The questionnaire is available in theAppendix Section 12.5. Subjects are required to indicate the degree of difficulty they have experienced in each domain in the past week on a 4-point scale that ranges from 0 (without difficulty) to 3 (unable to do). The highest score in each category is then summed (0 to 24) and divided by the number of categories scored to give a score that ranges from 0 to 3.
Any individual score of less than 2 is adjusted to 2 if the activity requires assistance from another individual or the use of an assistive device. If the category score is already a 2, and a device related to that category is used, or help from another person is provided for the category, the
REDACTEDsmesmeatus oftus of
nsiderinsideri0 is 0 is ““yy
COPYin coin c
enen
This do
cumen
txpexpediffidiffidd
cann
otitemtemendixndix
erieerie
be singsing us
ed eaa
contacontatoalthlth
supp
ort thhe e
r arthriarthran
ysssse ACe A
marketi
ng“veryver
severesevere
smesm
autho
rizati
on
ntt of Dof f the sf the s
ing bong boryry
appli
catio
nusingusingy poor,y poor
mponenmpone
andvity (vity (
any
P
exten
sions
on upn up
or va
riatio
ns
pal anpal a
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 49 of 183
score for that category remains 2. The following list details how each aid and device is associated with the category scores (Table 8‒2):
Table 8‒2: Aid or Device associated with HAQ-DI domain
If all questions within a given category are unanswered, no score will be provided for that category (this rule applies even if aids and devices are non-missing). The HAQ-DI will be calculated by dividing the sum of the category scores by the number of categories with at least 1 question answered. If fewer than 6 categories have responses, no disability score will be calculated. A lower HAQ disability score indicates an improvement in function.
8.1.1.6 Body surface area (BSA)-PsoriarisThe BSA palm method will be used for the evaluation of BSA affected by psoriasis as follows:
The subject’s hand, including the palm, fingers and thumb, is used as the reference point for measuring how much of their skin is affected by psoriasis, representing roughly 1% of the body’s surface.
Subject’s palm=1%
Head and neck=10% (10 palms)
Upper extremities=20 % (20 palms)
Trunk=30% (30 palms)
Lower extremities=40 % (40 palms)
Total BSA=100%.
8.1.2 Primary analysis of the primary efficacy variableThe primary efficacy variable (ACR50 at Week 12) will be analyzed for all subjects in the FAS.
The dose-response relationship between treatment and ACR50 response will be assessed with an ordered categorical analysis using a nonparametric correlation statistic of Mantel and Haenszel
REDACTEDvicesvicegory scgory sc
gories goriesscore iscore i
((
COPYsweredwere
s a
PY
This do
cumen
t TotaTota
cann
otunk=3nk=3
LowLow
beextreextr us
edeck=eck=
emem
to =1%1%su
pport
eir sir s
%%
anythe pthe p
skinsk
marketi
ng indnd
(BSA)BSA)ed for td for
autho
rizati
oned, noed, noare noare no
cores ores s havehavedicadica
app
o spp
licatio
n and
any e
xtens
ions o
r
ons
varia
tion
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 50 of 183
(Mantel and Haenszel, 1959) and modified ridit scores (Bross, 1958) with the corresponding p-value.
The analysis will include geographic region and prior TNF inhibitor exposure as stratification factors. Prior TNF inhibitor exposure will be used as a stratification factor as it may have an impact on efficacy. The use of geographic region as a stratification factor is intended to combine study centers in similar geographic regions. The geographic regions are defined in Section 3.8.The correlation between dose and ACR50 response will be evaluated at a two-sided significance level of α=0.05. This evaluation of dose-response will constitute the primary efficacy analysis.
Bimekizumab 160mg with loading dose will be excluded from the dose-response analysis. Atable will present the responder rates for placebo, BKZ 16mg, BKZ 160mg, BKZ 320mg, and the correlation statistic of Mantel and Haenszel (ridit score) with the corresponding p-value.
NRI will be used to account for missing data in the primary analysis. That is, subjects with a missing ACR score at Week 12 or who discontinued study treatment prior to the Week 12 visit will be considered non-responders for the primary analysis.
8.1.3 Secondary analyses of the primary efficacy variableAs the secondary analysis for the primary efficacy variable, a logistic regression model will be used to assess the effect of each individual dose versus placebo on ACR50 response. The model will include fixed effects for treatment, geographic region, and prior TNF inhibitor exposure (yes/no). To avoid the problem of the so-called monotone likelihood resulting in infinite large confidence intervals (eg, if one of the cell counts in the 2x2 table is equal to zero), a penalized maximum likelihood approach based on the modified score procedure of Firth (eg, Heinze and Schemper, 2002) will be used in the logistic models. If the logistic regression model is unable to converge, then geographic region may be dropped from the model to facilitate convergence. If the logistic regression model is unable to converge after dropping the geographic region, then prior TNF inhibitor exposure may be dropped as well. Comparisons will be made for each dose versus placebo at a 2-sided significance level of α=0.05. For each dose, the odds ratio versus placebo, the 95% CI, and the corresponding p-value will be calculated. In addition, a comparison will be made for the bimekizumab 320mg loading dose followed by bimekizumab 160mg Q4W versus bimekizumab 160mg Q4W using the same logistic model described above.
The pairwise comparisons of bimekizumab versus placebo will be formally evaluated for statistical significance only if the primary dose response efficacy analysis is statistically significant. Once the dose response has been established, pairwise testing of each bimekizumab dose versus placebo will account for multiplicity by using a fixed sequence testing procedure with each bimekizumab dose being tested sequentially from the highest dose to the lowest dose. If the dose-response relationship fails to reach significance at a significance level of α=0.05, then the further pairwise comparisons are seen as non-significant.
A table will present the responder rates for placebo, BKZ 16mg, BKZ 160mg w/ LD, BKZ 160mg, BKZ 320mg, and for each dose the odds ratios (differences to placebo), 95% CIs, and corresponding p-values from the logistic regression.
NRI will be used to account for missing data in the primary analysis. That is, subjects with a missing ACR score at Week 12 or who discontinued study treatment prior to the Week 12 visit will be considered non-responders for the primary analysis.
REDACTEDunts iunts he mode mod
gistic mgistic may be day be le tle t
COPYs p
regregioiomonotmono
This do
cumen
t A taA ta11
cann
oth bih bi
dosedosefurthfurth
bes plaplabimbim
used
canannce thece thaceace
tori
nce ncesu
pport
izumzum0mg Q4mg Q
sons sons
anyorresorre
umabma
marketi
ng drodro
to convo cone droppdrop
cancecancsp
autho
rizati
ontone tone n the 2the 2
difiedifiedmodelmode
roppopp
appli
catio
n vlogisticogisti
cebo onebo on, and , and
likl
and
variaarian
yto thto texte
nsion
sys
20mg0mgp-valuvalu
, subjesubjethth
oris. is. Ava
riatio
ns
nce nce ysis.sis.
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 51 of 183
8.1.4 Supportive analyses of the primary efficacy variableThe following supportive analyses are planned:
1. The primary dose response analysis and pairwise comparisons will be repeated for all subjects in the PPS to evaluate the effect of important protocol deviations on the analysis.
2. The primary dose response analysis and pairwise comparisons will be repeated for all subjects in the RS to evaluate the consistency of the FAS with the ITT principle. Subjects with no valid measurement of the primary efficacy variable at Baseline will be included as non-responders. NRI will be used to account for missing data.
3. The change from Baseline in all ACR components (ie, TJC, SJC, PGADA, PhGADA, PtAAP, HAQ-Di, and hs-CRP) will be analyzed using the ANCOVA with treatment group, geographic region, and prior TNF inhibitor exposure as fixed effects and the Baseline values as covariate. MI will be used for missing data (Section 4.2.1).
For the supportive analysis number 1 and 2 two tables will be displayed. One table will present the responder rates for placebo, BKZ 16mg, BKZ 160mg w/ LD, BKZ 160mg, BKZ 320mg, and the correlation statistic of Mantel and Haenszel (ridit score) with the corresponding p-value. The other table will present the responder rates, and for each dose the odds ratios (differences to placebo), 95% CIs, and corresponding p-values from the logistic regression.
For the supportive analysis number 3 tables will present the least-square means and standard error for placebo and bimekizumab dose for the ANCOVA models. For the pairwise comparison between placebo and bimekizumab dose the least-square means, standard error, corresponding p-values, and 95% confidence intervals for the contrasts will be provided.
8.1.5 Sensitivity analyses of the primary efficacy variableTwo sensitivity analyses will be performed for the primary dose response analysis and for the supportive analysis number 3:
1. This analysis evaluates the effectiveness of the treatment (de facto hypothesis) and will be based on the ordered categorical analysis to evaluate dose-response as specified in the primary analysis. In an attempt to prevent missing data during the study, efforts will be made to collect data from subjects that withdraw early from the study. An analysis will be performed in which all available data at Week 12 will be considered. In this case, subjectswill be analyzed according to their randomized treatment, even if they discontinued prior to Week 12 and were no longer on the randomized study treatment when the assessment was performed at Week 12. If Week 12 data for the ACR50 response are not collected, the subject will be assumed to be a non-responder. It should be noted that this measures something different from the primary analysis and could be confounded by placebo subjects who withdraw and are subsequently on another active medication at the time of the Week 12 assessment.
2. An additional sensitivity analysis will be based on observed data only for subjects who are still on the initially randomized treatment at Week 12. Subjects with missing data or who have prematurely discontinued study treatment will be excluded from the analysis. The same procedure described for the primary efficacy analysis of dose response will be used.
REDACTED e Ae A least least
the cothe c
es of tes of
COPY theh
presentresenANCAN
This do
cumen
t a2
cann
otbe abe adifferefferewithwith
bemed med asas
used
yzed zedand wnd w
tohichhichd acd ac
supp
ortegeg
an attemn attem subjesubjch alh a
anyeffeceffe
goricagoricamark
eting
thrmed fmed f
ectiect
autho
rizati
on nt thth
COVACOVAsquarsqua
ontrasontra
he e pp
appli
catio
n the che cthe oddhe od
gistic rgistic
the lehe l
and
d. d. KZ 16Z 16
co
any
One One ex
tensio
nsADAADAeatmenatme
he Base Ba
or va
riatio
ns
d as as
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 52 of 183
The following sensitivity analyses will be performed for the supportive analysis for the primary efficacy variable (pairwise comparison analysis of the primary efficacy variable) in order to assess the missing data assumptions:
1. Missing data will be imputed using MI. In MI, the missing value is replaced by a set ofplausible values, where each value is a Bayesian draw from the conditional distribution of the missing data given the observed data. Intermittent missing data will be imputed using the MCMC method, followed by monotone regression for monotone missing data to evaluate the effect of the method for handling missing data on the analysis. The ACR components will be imputed individually and then the ACR response will be calculated using the complete datasets. The multiply imputed data sets will be analyzed using a logistic regression model with fixed effects fixed effects for treatment, geographic region, and prior TNF inhibitor exposure. Finally, the results will be combined into a single inference using Rubin’s rule (Carpenter and Kenward, 2013). This procedure assumes a MAR pattern of missingness and corresponds to an estimand of what has been called the difference in outcome improvement if all subjects tolerated or adhered to treatment (Mallinckrodt et al, 2012). More details are available in Section 4.2.1.
2. The logistic regression model will be applied where all available data at Week 12 will be considered, as described in sensitivity analysis (1) for the primary efficacy variable.
3. The logistic regression model will be applied where only observed data for subjects still on the initially randomized treatment at Week 12 will be considered, as described in sensitivity analysis (2) for the primary efficacy variable.
The same tables as used for the primary and secondary efficacy analysis for the primary efficacy variable and for the supportive analysis number 3 with the different imputation analysis will be presented.
8.2 Statistical analysis of the secondary efficacy variablesThe secondary efficacy variables will be analyzed for all subjects in the FAS.
All categorical variables (ie, ACR20/70 response and PASI75/90 response) will be analyzed for treatment effects using pairwise comparisons using the same logistic model as described in the secondary analysis of the primary variable (Section 8.1.3).
The responder rates for placebo, BKZ 16mg, BKZ 160mg w/ LD, BKZ 160mg, BKZ 320mg, and for each dose the odds ratios (differences to placebo), 95% CIs, and corresponding p-values from the logistic regression will be provided.
The categorical efficacy variables will be analyzed using observed cases as treated and imputed with NRI.
8.2.1 Derivation of Psoriasis Area and Severity Index (PASI), PASI75, PASI90, and PASI100
PASI will be assessed in all subjects with Baseline BSA affected by psoriasis ≥3% determined by the method described in Section 8.1.1.6 (ie, the BSA palm method).
The PASI is the most commonly used and validated assessment for grading the severity of psoriasis in clinical studies (Feldman, 2004). The PASI quantifies the severity and extent of thedisease and weighs these with the percentage of BSA involvement.
REDACTED able.able.
and seand seis numis num
COPY where where
2 wil2 wil
This do
cumen
t8.2.1.2.1can
notegoregor
NRI.NRI.
11
begistiist
ii
used
tes esose thse thtictic
tos fors fosu
pport
e, AAairwiseirwi
f the prthe pr
any s ws w
ACRACR
marketi
ng
sis ofsis owill will
autho
rizati
ono
ll be cll be c
econdcondmber 3ber 3
appli
catio
n ilable dlable e prima prim
only only o
and01212an
yutcoutco2). M2)
exten
sionsion mon m
F inhiF inhg RubiRubi
of mof m
or ete te mo
varia
tions
the he
will beill be
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 53 of 183
Table 8‒3: Body areas for calculation of percent BSA for PASI
Body area Details of area BSA Degree of involvement of body areaa
Head Face, back of head 10% 0 to 6
Upper limbs Left, right, upper lower, flexorsurface, extensor surface
20% 0 to 6
Trunk Front, back, groin 30% 0 to 6
Lower limbs Left, right, upper lower, flexorsurface, extensor surface, includingbuttocks
40% 0 to 6
Total 100%
BSA=body surface area; PASI=Psoriasis Area and Severity Indexa Where 0=none; 1=1% to <10% affected, 2=10% to <30% affected, 3=30% to <50% affected, 4=50% to <70% affected, 5=70% to <90% affected, 6=90% to 100% affected
The body is divided into 4 sections (head (h) (10% of a person's skin); arms (a) (20%); trunk (t) (30%); legs (l) (40%), Table 8‒3). Each of these sections is scored by itself, and then the 4 scores are combined into the final PASI. For each section, the percent of area of skin affected (A), is estimated and then transformed into a grade from 0 to 6:
0% of involved area, grade: 0
< 10% of involved area, grade: 1
10-29% of involved area, grade: 2
30-49% of involved area, grade: 3
50-69% of involved area, grade: 4
70-89% of involved area, grade: 5
90-100% of involved area, grade: 6
Within each body area, the severity is estimated by three clinical signs: redness (R), thickness (T), and scaliness (S) (each on a 5-point scale): 0=none, 1=slight, 2=moderate, 3=marked, and 4=very marked. The PASI is a measure of the average redness, thickness, and scaliness of the psoriatic skin lesions, multiplied by the involved psoriasis area score of the respective section, and weighted by the percentage person’s skin for the respective section:
(11)
where
Rh, Ru, Rt, and Rl is the redness score of plaques on the head, upper limbs, trunk, and lower limbs;
REDACTEDrom rom
22
COPYa a
ctionsctionon,n, thethe
00
This do
cumen
t can
noc skiski
weightweightbe
rkedrkedinin
used
y ararness (Sess (dd
to reareasu
pportarea, grea, g
ved arved a
anygradegrademark
eting
de: 3 e: 3
autho
rizati
one pere per0 to 6:to 6:
appli
catio
n <5
son's skon's sis scois scrce
and
0% a0%
any
aany ex
tensio
ns
ex
or orva
riatio
ns
on
ar
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 54 of 183
Th, Tu, Tt, and Tl is the thickness score of plaques on the head, upper limbs, trunk, and lower limbs;
Sh, Su, St, and Sl is the scaliness score of plaques on the head, upper limbs, trunk, and lower limbs;
Ah, Au, At, and Al is the degree of involvement for the head, upper limbs, trunk, and lower limbs.
The PASI score ranges from 0 to 72 with a higher score indicating increased disease severity.
In case up to 2 severity measurements are missing for a certain region, the average of the remaining will be utilized to substitute the missing(s). If the area of affected skin and/or all severity measurements for up to 2 body areas are missing, missing (R+T+S) x A for a body area will be substituted by the average of the available (R+T+S) x A. Otherwise the PASI will be setto missing.
The PASI50, PASI75, PASI90, and PASI100 responses are based on at least 50%, 75%, 90%, and 100% improvement from Baseline in the PASI score.
8.3 Analysis of other efficacy variablesAll other efficacy variables will be analyzed for all subjects in the FAS, DBRS, ESS, cDBRS, cESS, and DBS.
All binary variables will be summarized using frequency tables by each visit. All continuous variables will be summarized using descriptive statistics by each visit.
The Hospital Anxiety and Depression Scale (HADS) will be analyzed separately for HADS-Aand HADS-D. The summary statistics for HADS-A and HADS-D will be presented in two different tables.
Subgroup of subjects will be used for the analyses of mNAPSI, MASES, and LDI. The MASES will be analyzed for subjects with enthesitis at Baseline (MASES > 0). The LDI will be summarized for subjects with dactylitis at Baseline (LDI > 0) and the mNAPSI will be analyzed on a subgroup of subjects with nail disease at Baseline (mNAPSI > 0).
CDF plots of PSAID-9 change from Baseline to Week 12 and from Week 12 to Week 48 by treatment group will be presented as well as a frequency table with number of subjects with PSAID-9 ≤ 3.
Times series plots will be provided for the response rate for the primary and secondary efficacy variables. The plots will show the weekly response rate over the first 12 weeks.
Time to onset of ACR20/50 response will be estimated and presented using the Kaplan-Meier product-limit method for each treatment. Time to a given response will be defined as the length in weeks from Baseline until the first date when the response is achieved. Following derivation will be used: Days from Baseline until the first date divided by 7. There will be no rounding for the Kaplan Meier estimates. Subjects who discontinue study prior to achieving a response will be censored at the date of study discontinuation. Subjects who reach the Week 12 visit without achieving a response will be censored at the date of the Week 12 visit. Missing response data will be imputed with NRI.
REDACTED freive staive st
cale (Hcale (s for Hs for H
COPY bjbj
equeeque
This do
cumen
tinn wewewiwi
cann
ot to onto onductduct
bes p
The pThe us
ed
plotslots
to ll beb su
pport
with with
9 changchanbe pre p
anydactydacty
h nan
marketi
ng HA
for tor the aheentheenth
autho
rizati
on
ency tncy tatisticatistic
HADHADADSADS
appli
catio
n
in then the
and
t lel any
east 5east 5ex
tensio
nsd/or ad/or for a bfor a
e PASPAS
or e e varia
tions
ty.y.
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 55 of 183
The median time to response including the two-sided 95% CI will be calculated for each treatment. Between groups differences (each bimekizumab dose versus placebo) will be analyzed using a log-rank test stratified by geographic region and prior TNF inhibitor exposure. Missing response data will be imputed with NRI.
Summary tables will be presented for the time to onset of ACR20 response and ACR50 response including following information: number of subjects achieving ACR20/ACR50 response, number of subjects censored, descriptive statistics (minimum, 25th percentile, median, 75th percentile, maximum) for the Kaplan-Meier estimates including the p-value (difference to placebo). Additional Kaplan-Meier Plots will be provided.
All other efficacy variables will be analyzed using observed cases as treated and imputed with NRI for binary variables and multiple imputation for categorical/continuous variables.
Subjects in the ESS who discontinue with bimekizumab treatment during the Dose-Blind Period will be treated as missing and categorical/continuous variables will be imputed.
In order to assess the potential bias of the potentially unblinded subjects based on data, the results of the final analysis will be utilized. Results of the ACR50 and PASI90 response of the rerandomized 320mg and 160mg treatment group at Week 48 will be compared to the potentially unblinded, randomized 320mg and 160mg treatment groups. If the percentage of ACR50 and PASI90 responders is comparable as assessed by overlapping 95% confidence intervals, the potential bias is estimated to be at minimum.
8.3.1 Minimal disease activity (MDA)MDA is a state of disease activity deemed a useful target of treatment by both the patient and physician, given current treatment possibilities and limitations. Criteria shown in Table 8‒4covering all of the domains of the disease have been developed to determine whether or not patient has reached MDA based on key outcome measures in PsA. A subject is considered as having MDA if 5 or more of the following 7 criteria in Table 8‒4 are fulfilled:
Table 8‒4: MDA criteria in PsA
A subject is classified as in MDA when they meet 5 of 7 of the following criteria:
TJC ≤1
SJC ≤1
PASI ≤1 or BSA ≤3
PtAAP ≤15
PGADA ≤20
HAQ-DI ≤0.5
Tender enthesial points ≤1
BSA=body surface area; HAQ=Health Assessment Questionnaire; MDA=minimal disease activity; PASI=Psoriasis Area and Severity Index; PGADA=Patient’s Global Assessment of Disease Activity; PtAAP=Patient’s Assessment of Arthritis Pain; SJC=swollen joint count; TJC=tender joint count
REDACTEDMDAMDAd a usd a us
sibilitsibilitiseasesease
COPYerlaperlap
A)
This do
cumen
tHAQAQ
TeTenumu
cann
ot 155
DA ≤DA ≤no
QQca
be S
beus
ed
A ≤3A ≤3ed to o
supp
ort
n MDAn MDrtpp
anyin Psin Pamark
etinge have hav
ey outcey outllowinlowin
autho
rizati
on
A)seful teful t
ies anes an
appli
catio
n anwill be ill be
If the If theping 95ng 9
andcts cts bb
nd Pd Pan
yputedpute
basb
exten
sionsputedputed
ables.ables.
e DoseDosedd
or va
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 56 of 183
Following rule will be applied for subjects with BSA < 3 at Baseline due to missing post-Baseline measurements for BASI and PASI: Subjects with BSA < 3 will always meet the criteria PASI ≤ 1 or BSA ≤ 3.
The MASES items will be used to check whether a subject has less or equal one tender enthesial point.
8.3.1.1 Modified MDASubjects are classified as in modified MDA when they meet the criteria PASI ≤ 1 or BSA ≤ 3 and 4 of the following criteria: TJC ≤1, SJC ≤ 1, PtAAP ≤ 15, PGADA ≤ 20, HAQ-DI ≤ 0.5 or tender enthesial points ≤ 1. The difference between the MDA and the modified MDA is that the PASI/BSA criteria is a mandatory criteria.
8.3.2 Disease Activity Score-28 joint count C-reactive protein (DAS28[CRP])
The components for DAS28(CRP) include the TJC and SJC based on 28 joints, hs-CRP (mg/L),and the PGADA (mm). DAS28(CRP) is calculated as follows:
(12)
The hs-CRP values below the limit of quantification should be set to half the limit of quantification for the calculations. The limit of quantification for hs-CRP is 0.16 mg/L.
8.3.3 Bath Ankylosing Spondylitis Disease Activity Index (BASDAI)The BASDAI is the most commonly used instrument to measure the disease activity of ankylosing spondylitis. The BASDAI is a validated self-reported instrument which consists of 6 NRSs, each with 10 units to measure the severity of the 5 major symptoms: fatigue, spinal pain, peripheral joint pain and swelling, enthesitis, and morning stiffness (both severity and duration) over the last week. The BASDAI questionnaire is available in the Appendix Section 12.6. Togive each symptom equal weighting, the average of the 2 scores relating to morning stiffness is taken. The resulting 0 to 50 sum score is divided by 5 to give a final BASDAI score between 0 and 10, with lower scores indicating lower disease activity.
The BASDAI is calculated as follows:
(13)
where Q1 – Q6 are the six question from the BASDAI questionnaire.
If 1 of the 2 morning stiffness measurements (ie, questions: and
) is missing, the other one will be used for themorning stiffness calculation.
REDACTEDof quof qu
ndylitndylitused iused i
AI isAI is
COPY tion shion sh
This do
cumen
t he
If 1 If 1 m
cann
ot
ere Qre Q
be us
ed a tolcullcul s
uppo
rt eigei
50 sum50 sumores indres in
t
anyAI quAI q
ghtinghtinmark
etingin
a valia vale the sethe s
, entheenthquequ
autho
rizati
onshoulshoulantificntific
tis Disis Dinstrustru
appli
catio
n
dd
and
28 8 any
jointjointex
tensio
ns is ts t
ein n
or 0.50
thathva
riatio
ns
≤ 3 3 .5 o5 o
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 57 of 183
If 1 major symptom of the BASDAI is missing, the sum score of the remaining symptoms will be divided by the number of symptoms assessed. If more than 1 major symptom is missing, the sum score will be set to missing.
8.3.4 Modified Nail Psoriasis Severity Index (mNAPSI)Subjects with psoriatic nail disease will have a target nail selected at the Baseline visit for evaluation using the mNAPSI. The nail selected should be the most affected nail observed at Baseline and should be the only one assessed throughout the study. The target nail will be scored (0 to 3) for onycholysis/oil drop dyschromia, nail plate crumbling, and pitting and will be scored (0 for “no” or 1 for “yes”) for leukonychia, nail bed hyperkeratosis, splinter haemorrhages and red spots in the lunula. The score for an individual nail ranges from 0 to 13 with higher scores indicative of more severe nail disease. Subjects with nail disease at Baseline are defined as those with a mNAPSI score >0 at Baseline.
If 1 or 2 response items scored on the 0 to 1 scale are missing, the missing response(s) will be imputed by the average of the available responses. Otherwise, the total mNAPSI score will be set to missing.
8.3.5 Maastricht Ankylosing Spondylitis Enthesitis (MASES) IndexThe MASES Index comprises 13 items (bilateral costochondral 1, costochondral 7, anterior superior iliac spine, posterior iliac spine, iliac crest and proximal insertion of the achilles tendon sites, and the fifth lumbar vertebral body spinous process) (Heuft-Dorenbosch et al, 2003) each scored as 0 = yes or 1 = no and then summed for a possible score of 0 to 13, with higher scores indicating worse enthesitis. The questionnaire is available in Section 12.7.
If 7 or more items are available, MASES will be imputed by dividing the sum score with the number of assessments and multiplying the result with 13. If less than 7 times are available, MASES will be treated as missing.
Presence of enthesitis at Baseline should be defined as a Baseline MASES score >0.
8.3.6 Leeds Dactylitis Index (LDI)Dactylitis, the swelling of an entire digit related to articular and periarticular inflammation, is a characteristic of inflammatory spondyloarthropathies, including PsA. Presence of dactylitis will be assessed using the LDI basic which evaluates for a ≥10% difference in the circumference of the digit compared to the opposite digit (Healy and Helliwell, 2007; Helliwell et al, 2005). The percent difference between circumferences will be multiplied by a tenderness score (0 for non-tender, 1 for tender). The digits involved and the matching contralateral digit will also be recorded at the same visits. The results from each digit with dactylitis will then be summed to produce a final score.
The following is a table of normative values, which will be used to provide the comparison, if matching digits are thought to be involved (Table 8‒5).
Table 8‒5: Normative values for LDI
Digit Men Women
Hand Thumb 70 58
REDACTEDfor for aaire isaire is
ES wiES wiying tying t
COPYochc
and pand ps procproc
This do
cumenThe he matma
cann
otd at at uce a uce a
fofo
beor tenor tethth
used
ed tod tence bence ndd
to he Lhe Lto tto t
supp
ortf an enan emmatormatoLDILD
anyitis Iitis mark
etingthe rethe re
shouldhou
autho
rizati
onp
cess) cess) possipossi
s availavail
ll be ll be
appli
catio
n sitis (sitis ndral 1,dral 1
roximoxim(H(
and mNm an
yrespresmNAPNA
exten
sionsag
gher scher sdefinedefine
p
ores aes va
riatio
ns
cored ored scorescore
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 58 of 183
Table 8‒5: Normative values for LDI
Digit Men Women
Index 63 54
Middle 63 54
Ring 59 50
Little 52 44
Foot Great toe 82 72
Second 52 46
Middle 50 44
Fourth 50 44
Little 52 45
The following rules will be applied for the LDI calculation in case of unclear data. If both digits of a given pair are recorded as affected, then each digit will be compared to the normative value. If both comparisons result in a difference greater than 10%, then both digits will contribute to the LDI score. If the circumference of the affected digit is smaller than the unaffected digit, then the LDI will be calculated by comparing the smaller digit to the normative value. If a digit is recorded as affected and the circumference of the contralateral digit is missing, the normative value will be used for comparison with the affected digit.
If the LDI CRF page is left empty at Baseline, it will be assumed that no digits were affected. At post-Baseline visits, when LDI cannot be determined, last observation carried forward (LOCF) imputation should be applied.
8.3.7 Psoriatic Arthritis Impact of Disease-9 (PsAID-9)The PsAID-9 is a patient-reported outcome measure for assessing the impact of PsA in 9 physical and psychological domains, including pain, fatigue, skin problems, work and/or leisure activities, functional capacity, discomfort, sleep disturbance, coping, and anxiety/fear/uncertainty. The questionnaire is available in the Appendix Section 12.8.
Each domain is assessed with a single question using a 0 to 10 NRS. Each domain score is multiplied by a weighting factor and the results are then summed to provide the total score.
(14)
The total score ranges from 0 to 10, with higher scores indicating a worse status. A score below 4 out of 10 is considered a patient-acceptable status. A change of 3 or more points is considered relevant absolute change.
REDACTEDf the f thee affecaffec
BaselinBaselnot beot be
COPY 101it is smit is sm
digitdigit
This do
cumen
t can
not b
e use
dasseassea weia we
to esssssu
pportl domdom
pacity,acitynty. Thty. T
any rted orted
ma
marketi
nge detee dete
is Impis Im
autho
rizati
ont to thto thcontracontra
cted ded d
ne, it e, it
appli
catio
nse of use of e compcomp
, then bthenmaller aller
h
and nd
any 444
454any
aex
tensio
ns
nsns
4 ex
or rva
riatio
ns
nnnonia
v
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 59 of 183
If one of the 9 domains is missing, the missing value is imputed using the mean value of the 8other (non-missing) domains. The imputed value will be used to calculate the PsAID with the formula above. If 2 or more of the domains are missing, the PsAID will be treated as missing.
8.3.8 Psoriatic Arthritis Quality of Life (PsAQoL)The use of the PsAQoL is recommended by the health regulatory authorities as one of the disease specific Health-Related Quality of Life (HRQoL) measures in PsA (CHMP/EWP/438/04). The questionnaire is available in Section 12.9. The PsAQoL comprises 20 items (0 for “not true” or 1 for “true”) so that the score ranges from 0 to 20 with higher scores indicating worse HRQoL.
If 6 or less item responses are missing, the missing responses will be imputed with the mean of available responses to calculate a total score. If more than 6 items responses are missing, the total score will be left missing.
8.3.9 Short Form – 36 Items Health SurveyThe SF-36 measures the following 8 health domains as rated by the subjects over the past four weeks: Physical Functioning, role Physical, Bodily Pain, General Health, Vitality, Social Functioning, Role Emotional, and Mental Health. The questionnaire is available in Section 12.10and the classification of the questionnaire items to the health domains is shown in Section 12.11.
The SF-36 PCS and MCS are used to measure the two broad components, or aspects, of health:physical and mental. PCS and MCS are based on the aggregate of 8 health concepts described above and all of the eight health domain scales are used to score both components summary measures.
One additional item asks respondents about health change over the past year.
The SF-36 will be calculated using QualityMetric’s Health Outcomes™ Scoring Software 4.5.The software uses updates 2009 U.S. population norms and applies a Full Missing Score Estimation method as follows:
A health domain score (except the physical functioning domain) will be estimated provided that at least one non-missing response is available within that domain.
For the physical functioning domain item response theory will be used to develop a model for estimates of the missing score.
Regression methods are then applied to estimate the PCS and the MCS on the basis of the available domains.
8.3.10 Hospital Anxiety and Depression Scale (HADS)The HADS was chosen for its well-established psychometric properties and its use in clinical research on biological therapy in subjects with chronic plaque psoriasis (Langley et al, 2010; Dauden et al, 2009). The HADS scores for anxiety and for depression range from 0 to 21 with higher scores indicating worse state. A score below 8 is considered to be normal whereas a score of 15 and above is considered severe (Snaith and Zigmond, 1994).
REDACTED arar
bout hbout h
QualitQualit
COPYwo brwo bthe aghe a
re usre u
This do
cumen
tse
DaudDaudhihi
cann
ot HADSADSearcharch
be us
edmethmeth
domaiomato hodhod
supp
ort sinsi
ctionctioninhe mise mi
anyept tept
ng reng remark
eting
tyMeyMpopulatopula
theth
autho
rizati
on ggregr
sed ted too
health alth
MetMet
appli
catio
n aire isire idomainmain
ad comad coregatega
and
bjejeealth, alth, iss
any
ects octs oex
tensio
ns e mm
issingssingor meamea
varia
tions
or 1 or 1 oL.L.
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 60 of 183
8.3.11 C-reactive protein (CRP) and high-sensitivity C-reactive protein (CRP)
Blood will be collected for measurement of CRP and hs-CRP. CRP and hs-CRP are indicators for inflammation and are measured in the blood. Hs-CRP is a component of composite efficacy variables, and in addition is also summarized and analyzed as individual variables. For hs-CRP the summary statistics should contain n, geometric mean, geometric CV, median, first and third quartile (Q1 and Q3), minimum, and maximum, where the geometric CV (%) is calculated using:
(15)
with SDln – the standard deviation of the ln-transformed hs-CRP values.
For descriptive statistics, the observed values and ratio to Baseline values will be displayed.
hs-CRP and CRP values below the limit of quantification should be set to should be set to half the limit of quantification for the calculations. The limit of quantification for hs-CRP is 0.16 mg/L and 0.4 mg/dL = 4.00 mg/L for CRP.
8.4 Subgroup analysisSubgroup analyses will be performed on the primary and secondary efficacy variables. The variables for subgroup analyses are defined in Section 4.8.
Subjects will be counted to have a positive treatment-emergent AbAb status in case the first AbAb positivity occurred up to Visit 7 (Week 12). The definition of first AbAb occurrence is described in Section 9.2.
The derivation of time since first diagnosis of PsA is described in Section 6.2.
All subgroup analyses are based on imputed data and will be summarized using descriptive statistics only.
9 PHARMACOKINETICS AND PHARMACODYNAMICS 9.1 PharmacokineticsBimekizumab plasma concentrations will be summarized for each treatment at each scheduled visit using the PK-PPS, and using a subset of the DBS which includes subjects that belong to both analysis sets, the DBS and the PK-PPS.
If bimekizumab plasma concentration measurements are deemed to be below the level of quantification (BLQ), then for calculation of the derived statistics this sample result will be set to half the lower limit of quantification (LLOQ). The subjects with at least 1 result that is defined as BLQ will also be listed within the respective analysis table. Descriptive statistics will be calculated if at least 2/3 of the values are above the LLOQ. If this is not the case, only median, minimum, and maximum will be presented.
In addition geometric mean bimekizumab plasma concentration time curves will be plotted by treatment group.
The bimekizumab concentrations will also be listed.
REDACTEDatmatmek 12ek 12
agnosiagnosi
COPY andanionion 44
menmen
This do
cumen
tcalcucalcumimi
cann
ot catioatio
he lowe lowQ wilQ wi
l
beumabmaionion
usedets, ths, th
ab pab p
toPPSPPSh
supp
ortcokinokinconceconce
S a
anyOKIOK
ne
marketi
ngis ofo
mputedmpute
NN
autho
rizati
on .
nt--emeeme2). The. The
of Pf P
appli
catio
n
econdacond8..
andion on an
yshoushon forfo
exten
sions
l be dbe d
or(15(15varia
tions
ing:ing:
5)5)
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 61 of 183
9.2 Pharmacodynamics and ImmunogenicityThe biomarker data will be separated between LGC assays, mulitplex proinflammatory cytokine and chemokine analysis, flow cytometry - TBNK panel, and flow cytometry -Th1/Th2/Th17/Th22 panel. Biomarker variables will be summarized and listed for each treatment at each scheduled visit using the PD-PPS, and using a subset of the DBS which includes subjects that belong to both analysis sets, the DBS and the PK-PPS.
In addition to the PD variables, whole blood will be stored to isolate deoxyribonucleic acid which may be used to examine genetic and epigenetic changes. The blood samples for genetic and epigenetic, and genomic, proteomic/metabolomics will not be analyzed in the interim or final analysis. Variables will be analyzed in a separate analysis as ad-hoc analysis.
The AbAb status will be determined for each visit where samples are taken for drug concentration measures. A cut point will be determined by the bioanalytical laboratory and will be used to determine the AbAb status as “above the cut point” (ACP) or “below the cut point” (BCP). For any AbAb levels that are ACP, a further confirmatory assay will be performed, the results of which will be determined as either “confirmed positive” (CP) or “not confirmed positive” (NCP). For samples that are CP, a further titre assay will be performed and the AbAb titre will be reported.
At each visit:
Samples that are either BCP or ACP and NCP are defined as AbAb-
Samples that are ACP and CP are defined as AbAb+
In addition, overall AbAb status should be determined:
Subject AbAb positivity is defined as having with AbAb+ at any time in the treatment period. This does not include Baseline/pre-treatment
Subject AbAb negativity is defined with AbAb- at any time in the treatment period. This does also include AbAb+ at Baseline/pre-treatment
If there is AbAb+ at Baseline/pre-treatment and there is a 4-fold increase in titre at least at one visit during the treatment period, then the subject has also an overall AbAb positivity status.
The visit of the first occurrence of AbAb positivity is defined as the visit when a subject is AbAb+ for the first time during treatment period excluding Baseline/pre-treatment. If there is AbAb+ at Baseline/pre-treatment and there is a 4-fold increase in titre at least at one visit during the treatment period, then the subject’s first occurrence visit is the Baseline Visit.
The number and percentage of subjects with AbAb levels above the specified cut point will be summarized.
In addition, the time point of the first occurrence of AbAb positivity during the Double-Blind Period, and the entire treatment period (excluding Baseline and pre-treatment) will be summarized for each treatment group.
REDACTED ed as Aed as A
be debe de
ed ased as
COPY P areP are
This do
cumen
tsummumm
InIn
cann
otatmentmen
e numnum
beBaseBaseuse
dfirst irst e firse firs
l
to osu
pportBaselinaseli
treatmtreatm
anyt Bast Ba
ne
marketi
ngs havs haveline/peline/p
fined finedse
autho
rizati
ondefindefi
AbAbAbAb
etermiterm
ii
appli
catio
nill be ll be and y wy w
CP) oCP) oan
ybelowbelowiwill bll
exten
sions
rim
drug drug laboraabora
orm om ova
riatio
ns
netic etic
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 62 of 183
All individual subject-level AbAb results will be listed including the screening assay, confirmatory assay, and titres if applicable. Note, that titre results will only be available, if the confirmatory assay is positive.
Immunogenicity data will be summarized and listed using the PK-PPS analysis set.
10 SAFETY ANALYSESAll safety summaries and listings will be performed using all subjects in the SS.
10.1 Extent of exposureThe duration of exposure and time at risk will be summarized for the Double-Blind Period and the entire treatment period. For the entire treatment period the duration of exposure and time at risk will be calculated for bimekizumab 160mg and 160mg with loading dose, bimekizumab 320mg, and all bimekizumab. The calculation of exposure duration and time at risk for the all bimekizumab group is different from that of the other two groups. For all bimekizumab the duration of exposure and time at risk will include the time a subject received any dose of bimekizumab (including 16mg) while for the other two treatment groups only the time a subject received 160mg, 160mg with loading dose, or 320mg will be included into the calculation.
Duration of exposure Double-Blind Period
The duration of exposure (in days) during the Double-Blind Period will be calculated as:
(16)
28 days refer to one half-life of bimekizumab.
Note: If the date of last injection (Double-Blind Period) + 28 extends to a date beyond the date of first injection (Dose-Blind Period), then this calculation reverts to
(17)
For subjects who die during the Double-Blind Period, then this calculation reverts to:
(18)
Duration of exposure entire treatment period
For subjects who do not switch study treatments, who receive bimekizumab 160mg with loading dose at Baseline, or who receive bimekizumab 16mg and will be summarized under all bimekizumab group:
REDACTED
ekizekiz
COPY bleble--BlB
This do
cumen
t DuraDurameF
cann
ot be
usedho dieo dieto
supp
ort an
y )) mark
eting
umabumab
oubleouble), then, the
autho
rizati
on ap
plica
tioncludedluded
ind Pend P
and ceiveiv
oupoupssd
any bimebim
ved ved
exten
sions
erioriore and e and
bimekbimekat risat ris
or ododva
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 63 of 183
(19)
Note: If the date of last injection +28 extends to a date beyond the final visit date (not including SFU), then this calculation reverts to:
(20)
For subjects who die, then this calculation reverts to:
(21)
For subjects who receive Placebo or bimekizumab 16mg in the Double-Blind Period and will besummarized under their Dose-Blind treatment in the overall table group bimekizumab 160mg or 320mg, or subjects who received Placebo and will be summarized in the overall table under their Dose-Blind treatment:
(22)
Note: If the date of last dose +28 extends to a date beyond the final visit date (not including SFU), then this calculation reverts to:
(23)
For subjects who die during the Dose-Blind Period, then this calculation reverts to:
(24)
The sum over all duration of exposure (in years) will be calculated for the total subject-years of duration of exposure.
The duration of exposure will be summarized by treatment group using descriptive statistics. A by-subject listing of date of first and last dose in Double-Blind Period, the date of first and last dose in Dose-Blind Period and the duration of exposure during Double-Blind Period and under bimekizumab treatment (any dose of bimekizumab) will be performed.
Time at risk Double-Blind Period
For subjects who complete the final visit of the Double-Blind Period and continue to the Dose-Blind Period:
REDACTED to a dto a d
COPY
This do
cumen
tbyy--susudosdos
cann
otn of en of
e duradurabb
bever aver ause
d
ll
to su
pport
g theth any
hehe DD
marketi
ng au
thoriz
ation
date bedate b
appli
catio
n inn andup bup b
n thethean
y lind Plind imi
exten
sions
or va
riatio
ns
0))
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 64 of 183
(25)
For subjects who discontinue on or prior to the final visit of the Double-Blind Period, use the minimum of the following:
(26)
(27)
(28)
where 140 days refers to 5*half-life of bimekizumab.
For subjects who die during the Double-Blind Period, then this calculation reverts to:
(29)
Time at risk entire treatment period
For subjects who do not switch study treatments, who receive bimekizumab 160mg with loading dose at Baseline, or who receive bimekizumab 16mg and will be summarized under all bimekizumab group:
For subjects who complete the Dose-Blind Period and enter the extension study:
(30)
For subjects who die prior to the final visit:
(31)
For all other subjects, use the minimum of the following:
(32)
(33)
For subjects who receive Placebo or bimekizumab 16mg in the Double-Blind Period and will besummarized under their Dose-Blind treatment in the overall table with bimekizumab 160mg or 320mg, or for subjects who receive Placebo and will be summarized in the overall table under their Dose-Blind treatment:
For subjects who complete the Dose-Blind Period and enter the extension study:
REDACTEDents, ents, wumab mab
--BlinBli
COPY
This do
cumen
tor r summsumm3232
anno
t
subjsubj
be us
ed tocts, uts, su
pport
use
any the finhe fimark
etingnd Pend Pe
autho
rizati
on
who rewho re16mg16mg
appli
catio
nalculalculand
l
any e
xtens
ions((o
r ((22
varia
tions
there
of.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 65 of 183
(34)
For subjects who die during the Dose-Blind Period:
(35)
For all other subjects, use the minimum of the following:
(36)
(37)
The sum over all time at risk (in years) will be calculated for the total subject-years of time at risk.
The days on treatment and days on bimekizumab treatment will be calculated as follows:
(38)
(39)
10.2 Adverse eventsAn AE is any untoward medical occurrence in a patient or clinical investigation subject administered a pharmaceutical product that does not necessarily have a causal relationship with this treatment. An AE can therefore be any unfavorable and unintended sign (including an abnormal laboratory finding), symptom, or disease temporally associated with the use of a medicinal (investigational) product, whether or not related to the medicinal (investigational) product.
AEs (including serious AEs) are characterized as either non-treatment or treatment emergent according to the following criteria:
Non-treatment emergent are the events with onset date and time prior to the very first administration of study medication (bimekizumab or placebo) or after a 140-day period after the final drug administration.
Treatment-emergent AEs (TEAE) are those with onset date at or after the very first administration of study medication. The events that emerge within 140 days after the final drug administration, will also be considered as treatment emergent (eg, in the case of premature discontinuation or during the Safety Follow-up [SFU] period).
REDACTED COPY mm
This do
cumen
t Ncann
otcludlud
ding tding t
NoNo
be
dindinus
edstigatiga toy finy fi supp
ortdicadica
eutical uticaE can thcan th
ndd
any tsts
cal oal
marketi
ng
ss
autho
rizati
on ap
plica
tion
ota
will bewill b
nd
al sul suan
y exte
nsion
s or v
ariati
ons t
hereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 66 of 183
All AEs occurring during the study (i.e., after signature of the informed consent document) will be recorded in the eCRF. For each AE the following information will be recorded in the eCRF: AE term (verbatim term), date of onset, whether or not the AE was classified as a SAE, as an AE of special interest, intensity, relationship to study medication, action taken with study medication, other action taken, outcome, date of outcome, and whether the AE led to study drug discontinuation or to study discontinuation.
ADRs are defined as a response to a drug which is noxious and unintended and which occurs at doses normally used in man for prophylaxis, diagnosis, or therapy of disease or for modification of physiological function. Any AE that is considered “Related” to study treatment will be classed as an ADR.
If an AE occurs on the same day as the treatment switch (Visit 7, Week 12) then the AE will be allocated to the Double-Blind treatment. An exception from this general rule is made for ADRs,and AE of hypersensitivity reactions and anaphylactic reactions. If those AEs occur on the same days as the treatment switch, the AEs will be allocated to the Dose-Blind treatment.
The incidence of TEAEs will be summarized by MedDRA SOC, high level term, and PT. Tables with incidences of classified TEAEs by maximum intensity, by relationship, and by subject number will be provided. The incidence of non TEAEs, non-serious TEAEs above the reporting threshold of 5% of subjects and relationship will be summarized. Furthermore the incidence of all TEAEs, serious TEAEs, non-serious TEAEs, and TEAEs leading to study discontinuation and/or permanent withdrawal of study medication, adverse drug reactions, fungal infectious disorder TEAEs, opportunistic infection (including tuberculosis) TEAEs, malignant or unspecified tumor TEAEs, malignant tumor TEAEs, major cardiovascular event TEAEs, haematopoietic cytopenias TEAEs, neuropsychiatric events TEAEs, inflammatory bowel disease TEAEs, hypersensitivity and anaphylactic Reaction TEAEs, hepatic events TEAEs, and TEAEs by timing of onset relative to AbAb Status will be summarized. In addition, an overall summary table will be provided.
The tables will be split into the Double-Blind Period and the complete treatment period (exception non TEAE table and TEAEs by timing of onset relative to AbAb Status). Presentations for the Double-Blind Period will summarize AEs that start prior to or at Visit 7 (see details given above) by treatment group as randomized for the Double-Blind Period.Presentations for the complete treatment period will only summarize AEs that occur under treatment with bimekizumab, irrespective of the treatment period. For summaries of thebimekizumab 160mg, 160mg with loading dose and bimekizumab 320mg treatment groups, patients randomized to Placebo or bimekizumab 16mg at Baseline will only be included with AEs that start in the Dose-Blind Period. The all bimekizumab treatment group will also present AEs occurring under bimekizumab 16mg treatment in the Double-Blind Period. AEs for each subject will be summarized based on the treatment actually received at the onset of each particular AE.
Exposure adjusted incidence rate (EAIR) and exposure adjusted event rate (EAER) will only be calculated for the complete treatment period for following tables in the final analysis: all TEAEs, serious TEAEs, TEAEs leading to study discontinuation and/or permanent withdrawal, adverse drug reactions, fungal infectious disorder TEAEs, opportunistic infection (including tuberculosis) TEAEs, malignant or unspecified tumor TEAEs, malignant tumor TEAEs, major cardiovascular event TEAEs, haematopoietic cytopenias TEAEs, neuropsychiatric events
REDACTEDludinludinor TEor TE
uropsyuropsyylacticylacticStaSt
COPY ummumnd TEnd TE
n, adn, ad
This do
cumen
tbjjpartiparti
EE
cann
ot t stastaoccurrccurr
ject ect
bendomndomartart
used
imekme160m160mmimi
to e ccekizekiz
supp
ortblele--BBve) by t) bycompom
any Doo
nd TEnd TEli
marketi
ngc ReRe
atus watus w
oubloub
autho
rizati
onEA
dversedverseg tubeg tube
EAEs, EAEs, ychiatchiateaceac
appli
catio
nelatioelatioerious rious
arized. rizedAEs leAEs le
and gh levh lev
any Es Es
treatmtreatex
tensio
ns
the Athe Ais mads mad
s ococ
ore cle c varia
tions
at at
cationationclascla
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 67 of 183
TEAEs, inflammatory bowel disease TEAEs, hypersensitivity and anaphylactic Reaction TEAEs, hepatic events TEAEs, and TEAEs by timing of onset relative to AbAb Status.
10.2.1 Exposure durationThe duration of each AE will be calculated as follows:
+ 1 (40)
For the calculation of duration for AEs with missing start dates, the dates will be imputed as described in Section 4.2.2.
The time to first dose for each AE will be calculated as follows for all TEAEs:
(41)
The time to most recent dose for each AE will be calculated as follows for all TEAEs:
(42)
Time to first dose and time to most recent dose will not be calculated for pretreatment AEs.
For the calculation of time to first bimekizumab and time to most recent bimekizumab dose the same formulas as above (41 and 42) will be used but the date of first bimekizumab dose will be used instead of the date of first dose. The time to first dose and time to first bimekizumab dose differs only for those subjects who receive placebo at Baseline.
10.2.2 Exposure adjusted incidence rate (EAIR) and exposure adjusted event rate (EAER)
The EAIR is defined as the number of subjects (nAE) with a specific AE adjusted for the exposure and will be scaled to 100 patient-years:
(43)
where is a subject’s exposure time in years up to the first occurrence of the AE of interest (equation [40] in years) at the level of coding evaluated, nnoAE the number of subjects without the specific AE and the total time at risk scaled to 100 patient-years (Section 10.1).
If a subject has multiple events at the level of coding evaluated, the time of exposure is calculated to the first occurrence of the AE of interest. If a subject has no events, the total time at risk (Section 10.1) in years is used. As indicated above, exposure-adjusted incidence rates will only be calculated for the overall treatment period. These presentations do not include AEs that occur under Placebo treatment. Therefore, a subject’s exposure time will only start at the first dose of bimekizumab in the Dose-Blind Period for subjects randomized to Placebo at Baseline. Also, for subject’s randomized to bimekizumab 16mg at Baseline, exposure time will only be
REDACTED w
umab mab be usbe u
The tiThe teceiecei
COPYwill nwill n
This do
cumen
t If a If ac
cann
ot tion [ion [cific cific
be
isisus
ed to
supp
orted to d to an
y umbermbe1
marketi
ngm
ive plave pl
ed incid incR)R)
autho
rizati
on
not benot be
and tind tiseded bubu
me tome to
appli
catio
n ands for fo an
y r a
exten
sions
or va
riatio
ns
as s
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 68 of 183
considered from the start of Dose-Blind Period, when presenting AEs for the bimekizumab 160mg and bimekizumab160mg with loading dose, or bimekizumab 320mg groups. All exposure time on any bimekizumab dose is considered for the all bimekizumab column.
Exact Poisson 95% CIs for incidence rates are calculated using the relationship between the Poisson and the Chi-square distribution (Ulm, 1990; Fay and Feuer, 1997):
(44)
(45)
(46)
(47)
where nAE is the number of subjects with a specific AE for the incidence rate of interest and is the basis for the number of the degrees of freedom for the chi-square quantile for the upper tail probability χ2, is a subject’s exposure time in years up to the first occurrence of the AE of interest (equation [40] in years) at the level of coding evaluated, nnoAE the number of subjects without the specific AE and the total time at risk scaled to 100 patient-years (Section 10.1).
The EAER will be the number of AEs including repeat occurrences in individual subjects divided by the total time at risk scaled to 100 patient-years and calculated using:
(48)
where NAE is the total number of AEs, is a subject’s total time at risk in years and nAll thenumber of subjects.
No CI will be computed for EAER.
10.2.3 AE of Special Interest and AE of Special MonitoringAE of special interest is any AE which meets the Hy’s Law criteria, defined as ≥3x upper limit of normal (ULN) alanine aminotransferase (ALT) or aspartate aminotransferase (AST) with coexisting ≥2xULN total bilirubin in the absence of ≥2xULN alkaline phosphatase (ALP), with no alternative explanation for the biochemical abnormality.
AE of special monitoring for this study include:
REDACTEDecifieciffreedofreedo
sure tisure te levee levehe the t
COPY
ici
This do
cumen
t10.20.2AA
cann
otof sof s
CI wilI wil
be s ts t
subsubus
ed
thehe
to su
pport
any f Af A
k scalk scamark
eting
el o ototal tiotal t
AEs AE
autho
rizati
on
AE foAE foom form for
ime inme inof coof c
appli
catio
n at
and a
ny ex
tensio
ns or
varia
tions
)
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 69 of 183
Infections (serious, opportunistic, fungal and TB)
Malignancies, including lymphoma
Major cardiovascular events
Neutropenia
Neuropsychiatric events (in particular depression and suicide)
Inflammatory bowel disease
Anaphylactic reaction
Hepatic events
For the definitions of AE of special monitoring the Bimekizumab Safety Topics of Interest (Version date 19Feb2018) will be used.
The incidence of TEAEs of special monitoring will be summarized by MedDRA system organ class, high level term, and PT. EAIR with associated 95% CI and the EAER will be included in the summary tables. Serious infections are also classified as AE of special monitoring but no separate table will be produced.
The output table for the search criteria Standardized MedDRA Query (SMQ) = “Malignant or unspecified tumours (SMQ)” will include two different overall rows:
The first overall incidence row will summarize “Any malignancies (including unspecified)” and this row will summarize the incidence of all AEs flagged for inclusion in the table, regardless of the High Level Term (HLT) it codes to.
The second overall incidence row will summarize “Any malignancy (including unspecified, excluding non-melanomic skin cancers)” and this row will summarize the incidence of AEs flagged for inclusion in the table, excluding those which code to an HLT of “skin neoplasms malignant and unspecified (excl melanoma)”.
The output table for the search criteria SMQ=”Malignant tumours (SMQ)” will include two different overall rows:
The first overall incidence row will summarize “Any malignancies” and this row will summarize the incidence of all AEs flagged for inclusion in the table, regardless of the High Level Term (HLT) it codes to.
The second overall incidence row will summarize “Any malignancy (excluding non-melanomic skin cancers)” and this row will summarize the incidence of AEs flagged for inclusion in the table, excluding those which code to an HLT of “skin neoplasms malignant and unspecified (excl melanoma)”.
The output table for Anaphylactic reaction will include three different overall rows:
The first row within the body of the table will be labeled “Any hypersensitivity/anaphylactic reaction” and will represent the overall incidence of subjects who reported at least one hypersensitivity reaction or at least one anaphylactic reaction.
REDACTEDmarizemarizence once o
(HLT)(HLT)
w wilw wi
COPYMedMefferentfferen
This do
cumen
t aThTh
cann
otlanomanomnclusinclus
andand
becondcondused
e ie m (HL(HL
toin
inciincisu
pport
rchch
ncidenncide
any excex
h crith crmark
etingill sumll sum
ancers)”ncersle, exce, exc
xcl mcl m
autho
rizati
ont ovov
e “Any“Anyof all Af all A
) it coit co
appli
catio
nof speof sp
DRA QRA Qveravera
and
MMe EAEEAE
any
edDedDex
tensio
ns
cs of s of
or va
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 70 of 183
The second row within the body of the table will be labeled “Any hypersensitivity reaction” and will represent the overall incidence of subjects who reported at least one hypersensitivity reaction.
The third row within the body of the table will be labeled “Any anaphylactic reaction” and will represent the overall incidence of subjects who reported at least one anaphylactic reaction.
10.3 Clinical laboratory evaluationsThe routine clinical laboratory evaluations specified in Table 10‒1 and will be summarized. If any additional analytes are also recorded then these will be listed only.
Different summary tables for hematology and biochemistry variables will be provided, based on data from scheduled visits: observed values and change from Baseline, shift from Baseline to maximum post-Baseline value (in Double-Blind Period), shift from Baseline to maximum post-Baseline value (in Dose-Blind Period), shift from Baseline to minimum post-Baseline value (in Double-Blind Period), shift from Baseline to minimum post-Baseline value (in Dose-Blind Period), shift from Baseline to end of treatment (in Double-Blind Period), shift from Baseline to end of treatment (in Dose-Blind Period), and markedly abnormal laboratory data.
End of treatment (in Double-Blind Period) will be defined as the last visit in the Double-Blind Period or the early termination assessment depending if the subject discontinued in the Double-Blind Period.
End of treatment (in Dose-Blind Period) will be defined as the Week 48 visit or the early termination assessment depending if the subject discontinued early or not.
In addition, number of subjects who meet the Hy’s Law criteria (Section 10.2) will be described using the frequencies.
All laboratory data will be listed by treatment, subject and visit including changes from Baseline for numeric variables, flags for measurements outside the normal ranges, the relative study day, a flag for whether the test was not done and a flag for whether the subject was fasting.
Additional listings will be presented for Hepatitis B and C, human immunodeficiency virus (HIV), genomic proteomic/metabolomics, and genetic/epigenetic tests. C-reactive protein (CRP) and hs-CRP will be listed separately in the efficacy section.
Table 10‒1: Laboratory measurements
Hematology Biochemistry Urinalysis
Basophils Calcium Bacteria
Eosinophils Chloride Crystals
Lymphocytes Magnesium Glucose
Monocytes Potassium pH
Neutrophils Sodium RBC
Hematocrit Glucose WBC
REDACTED ll be dll be d
subjecsubjec
meet meet
COPY finefining if ing if
This do
cumen
t EosinEosintmLmm
eum
cann
otologylogy
ophilsophilsnnc
be
yye us
ed
LabLabtoe listlis su
pport
be presee pre
omic/momic/mstedste
any meme
not doot domark
etingthehe
y treatmy treateasureasu
autho
rizati
on th
defineefinect disct dis
he He H
appli
catio
n Perermal laboal lab
as the s thehe subhe su
ande valuvalu
riodriod
any e te
postpost-BBex
tensio
ns
videdvidedrom Bom Bto mto
oreded varia
tions
d Id I
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 71 of 183
Table 10‒1: Laboratory measurements
Hematology Biochemistry Urinalysis
Hemoglobin BUN Urine dipstick for pregnancy testingb
MCH Creatinine
MCHC hs-CRP/CRPa
MCV AST
Platelet count ALT
RBC count GGT
WBC count ALP
Total bilirubin
LDH
Total cholesterol
Uric acid
Albumin
Serum pregnancy testingb
ALP=alkaline phosphatase; ALT=alanine aminotransferase; AST=aspartate aminotransferase; BUN=blood urea nitrogen; CRP=C-reactive protein; GGT=gamma glutamyltransferase; hs-CRP=high sensitivity C-reactive protein; LDH=lactate dehydrogenase; MCH=mean corpuscular hemoglobin; MCHC=mean corpuscular hemoglobin concentration; MCV=mean corpuscular volume; RBC=red blood cell; SFU=Safety Follow-up; WBC=white blood cella Both CRP and hs-CRP will be tested at specified visits per Protocol Table 5.1.b A serum pregnancy test will be performed at Screening for all women of childbearing potential. This recommendation also applies to women of childbearing potential with infrequent or irregular menstrual cycles and women during menopause. Natural menopause is recognized to have occurred after 12 consecutive months of amenorrhea, for which there is no other obvious pathological or physiological cause (International Menopause Society, 2015). A urine pregnancy test is also required at the Baseline, Week 48, ET, and at SFU visits. A urine pregnancy test will also be performed at any study visit where there has been a delay in menses. This recommendation also applies to women of childbearing potential with infrequent or irregular menstrual cycles. Pregnancy test results must be negative prior to administering IMP.
Markedly abnormal values for biochemistry and hematology will be defined as laboratory values graded 3 or 4 according to the Rheumatology Common Toxicity Criteria (RCTC). Definitions of the markedly abnormal values are given in Table 10‒2 and Table 10‒3 and are based on the RCTC units. All units in the tables below will be converted to the standard units based on Clinical Data Interchange Standards Consortium (CDISC) standards. The markedly abnormal laboratory results will be listed separately.
Values that are below the lower limit of the reference range will be flagged as ‘L’ (low) and values that are above the upper limit of the reference range will be flagged as ‘H’ (high) and listed as well.
REDACTtransferansfmma glumma gl
corpuscorpuumum
CTED nancy nancyDCT
COPY OP
C
This do
cumen
tRCTCTClCl
cann
ot y aby ab
d 3 ord 3 ormarkmarkC
be
abnoabnous
edlso apo apesults sults
to p
o beo besu
pport
womomtural mural m
e is no ois noegnanegnan
anyformeorme
men omen
marketi
ng scularcular
me; RBCe; RBC
specifspecid
autho
rierase; Arase; A
utamylamylori
zatio
n testintestinti
o
ori
appli
catio
n iiionatilicap
and n
any
any ex
tensio
ns
nsns
ex
or rva
riatio
ns
nnnonia
v
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 72 of 183
For tables where data are summarized by visit, unscheduled and repeat visits will not be summarized, but these data will be included in listings. In the case where laboratory values are below the LLOQ, then these will be set to the midpoint between 0 and the LLOQ for the purpose of summarizing the data.
Table 10‒2: Definitions of markedly abnormal hematology values
Variable (RCTC Units) Markedly Abnormal Definition
Low High
Hemoglobin (g/dL) <LLNAND>2.0 decrease from Baseline
N/A
Hemoglobin (g/dL) <8.0 N/A
Leukocytes (total x 1000) <2.0 N/A
Lymphocytes (x 1000) <0.5 N/A
Neutrophils (x 1000) <1.0 N/A
Platelets (x 1000) <50 N/A
LLN=lower limit of normal; N/A=not applicable; RCTC= Rheumatology Common Toxicity Criteria.
Table 10‒3: Definitions of Markedly Abnormal Biochemistry Values
Variable (RCTC Units) Markedly Abnormal Definition
Low High
ALP N/A >3 x ULN
ALT N/A >3 x ULN
AST N/A >3 x ULN
Calcium (mg/dL) <7.0 >12.5
Creatinine (mg/dL) N/A >1.8 x ULN
Glucose (mg/dL) <40 >250
Potassium (mmol/L) <3.0 >6.4
Sodium (mmol/L) <125 N/A
Total bilirubin N/A ≥2 x ULN
Uric acid N/A ≥3 x ULN
ALP=alkaline phosphatase; AST=aspartate aminotransferase; N/A=Not applicable; RCTC= Rheumatology Common Toxicity Criteria; ULN=upper limit of normal
10.4 Potential drug-induced liver injury (PDILI) assessment All potential drug-induced liver injury (PDILI) events require immediate action, testing, and monitoring. The measurements and additional information required for the assessment of PDILI
REDACTEDAbnorAbno
edly Aedly ACT
ww EDAR
COPY RheumRheumP
Y
This do
cumenALAen
tUric ric en
caaccann
ot mmmm
bilirubiliruot
ca
be
mol/mol/eeus
ed
mol/Lol/Led to o
supp
ort t
pps
any N/Nyanaaaamark
eting
tin
g
N/AN/Aarkm
autho
rizati
onmatolatol
rmal rmal
Abnormbnormor
au
appli
catio
n N
N/Nonaticaaaa
logylogyapp
and
/AA ndan
y nyex
tensio
ns
xte
or rva
riatio
ns
onononria
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 73 of 183
events when there is a reasonable possibility that they may have been caused by the IMP are included but not limited to those listed in Table 10‒4 (laboratory measurements) and Table 10‒5(additional information).
PDILI laboratory results and additional PDILI information will only be listed by treatment group and subject. If specific PDILI information collected separately is matching to the entries in the standard eCRF pages collected for all subjects, the specific PDILI information will be added to the corresponding listing for the standard eCRF information (eg, lifestyle information is collected for all study subjects, the additional PDILI information for alcohol and illicit drug use will be included in the listings for lifestyle). For information collected on top (eg, family history of PDILI) a new listing will be generated.
Table 10‒4: PDILI laboratory measurements
Virology-related Hepatitis A IgM antibody
HBsAg
Hepatitis E IgM antibody
HBcAb-IgM
Hepatitis C RNA
Cytomegalovirus IgM antibody
Epstein-Barr viral capsid antigen IgM antibody (if unavailable, obtain heterophile antibody or monospot testing)
Immunology Anti-nuclear antibody (qualitative and quantitative)
Anti-smooth muscle antibody (qualitative and quantitative)
Type 1 anti-liver kidney microsomal antibodies (qualitative and quantitative)
Hematology Eosinophil count
Urinalysis Toxicology screena
Chemistry Amylase
ALT, AST
If total bilirubin ≥1.5xULN, obtain fractionated bilirubin to obtain % directbilirubin
Serum CPK and LDH to evaluate possible muscle injury causing transaminase elevation
Additional Prothrombin time/INRb
Serum pregnancy test
PK sample
REDACTEDpsid apsid ady or mdy or m
antibodantibodAC
h muh muRED
COPY tibodyibodyP
YC
This do
cumen
t ddd
umumum
cann
ot
ditioitioanbe
used
to su
pport
oxicxi
AmyAmort
uAuupu
any phiph
colocolny mark
etinguscle auscle a
-liver kliver eti
hil coil cma
autho
rizati
on
ntigenntigenmonosmonos
io
dy y (qu(qutho
a
appli
catio
n onca
app
and nd
any nyex
tensio
ns
xt
ory hisy hi varia
tions
g use use istoist
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 74 of 183
Table 10‒4: PDILI laboratory measurements
Virology-related Hepatitis A IgM antibody
HBsAg
Hepatitis E IgM antibody
HBcAb-IgM
Hepatitis C RNA
Cytomegalovirus IgM antibody
Epstein-Barr viral capsid antigen IgM antibody (if unavailable, obtain heterophile antibody or monospot testing)
ALT=alanine aminotransferase; AST=aspartate aminotransferase; CPK=creatine phosphokinase; HBcAb-IgM=hepatitis B core antibody-IgM; HBsAg=hepatitis B surface antigen; IgM=immunoglobulin M; INR=international normalized ratio; LDH=lactate dehydrogenase; PDILI=potential drug-induced liver injury; PK=pharmacokinetic; RNA=ribonucleic acid; ULN=upper limit of normala For detecting substances (ie, amphetamines, benzodiazepines, opioids, marijuana, cocaine, phencyclidine, and tricyclic antidepressants), additional tests may be performed based on the Investigator’s medical judgment and patient’s history.b Measured only for subjects with ALT >8xULN, elevations in total bilirubin, and symptoms of hepatitis or hypersensitivity. Hepatitis symptoms include fatigue, nausea, vomiting, and right upper quadrant pain or tenderness; hypersensitivity symptoms include eosinophilia (>5%), rash, and fever (without clear alternative cause).
Table 10‒5: Additional PDILI information
New or updated information
Concomitant prescription and over-the-counter medications (eg, acetaminophen, herbal remedies,vitamins); dosages and dates should be included.
Pertinent medical history, including the following:
History of liver disease (eg, autoimmune hepatitis, nonalcoholic steatohepatitis or other “fatty liver disease”)
Adverse reactions to drugs
Allergies
Relevant family history or inheritable disorders (eg, Gilbert’s syndrome, alpha-1 antitrypsin deficiency)
Recent travel
Progression of malignancy involving the liver (Note: Metastatic disease to the liver, by itself, should not be used as an explanation for significant AST and/or ALT elevations.)
The appearance or worsening of clinical symptoms of hepatitis or hypersensitivity (eg, fatigue, nausea,vomiting, right upper quadrant pain or tenderness, decreased appetite, abdominal pain, jaundice, fever, orrash)
REDACTED as
formaformR
COPYtotal totaa, vom, vom
sh, anh, an
This do
cumen
tRRum
cann
otlevantevantdeficiefici
be es us
edctionstionsto
supp
ort cludinludi
ase (egse (eg
annganyld beld b
anmark
eting
ang
hee--councoue i
rke
autho
rizati
on mitintin
nd feved feve
ationtion
appli
catio
n uana, ana,nvestigavestig
bilirubinilirubing, ang, a
and
drugru
c
anyglobuglobu
gg--indin
exten
sions
e, obta, obtans
nase; nase;te
or rva
riatio
ns
nnnonia
v
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 75 of 183
Table 10‒5: Additional PDILI information
New or updated information
Recent clinically significant hypotension or hypoxemia with compromised cardiopulmonary function
Alcohol and illicit drug use
Results of liver imaging or liver biopsy, if done
Results of any specialist or hepatology consult, if done
Any postmortem/pathology reports
ALT=alanine aminotransferase; AST=aspartate aminotransferase; PDILI=potential drug-induced liver injury
10.5 Vital signs, physical findings, and other observations related to safety
10.5.1 Vital signsThe following vital signs measurements will be assessed:
Systolic Blood Pressure (mmHg)
Diastolic Blood Pressure (mmHg)
Pulse Rate (bpm)
Temperature (C)
Vital signs measurements (except temperature) will be summarized and listed by visit and timing relative to dosing including changes from Baseline. The listing will also include details to abnormal values as defined in Table 10‒6. Temperature measurements will only be listed and not summarized in a table.
Table 10‒6: Definitions of markedly abnormal blood pressure values
Variable (Unit) Markedly Abnormal Low Markedly Abnormal High
Systolic blood pressure (mmHg) <90 and a decrease from Baseline of ≥20
>180 and an increase fromBaseline of ≥20
Diastolic blood pressure (mmHg) <50 and a decrease fromBaseline of ≥15
>105 and an increase fromBaseline of ≥15
10.5.2 ElectrocardiogramsThe following ECG variables will be assessed:
PR interval (ms)
RR interval (ms)
QRS duration (ms)
QT interval (ms)
QTcF interval (ms)
REDACTED
peratureraturfrom from 101
COPY
This do
cumen
t he he cann
ot 2e follfoll
be
bus
ed
pressupressudto
supp
ort t
(mmH(mmHpo
anymama
aaaamark
eting
m BasBa0‒‒66.. TeTe
ark
autho
rizati
on
ree)) wiwiasease
appli
catio
n and
any e
xtens
ions
iver injver inns
tionsions
or rva
riatio
ns
nnnonia
v
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 76 of 183
QTcB interval (ms)
The date and time of the ECG will be recorded in the eCRF together with the Investigator interpretation and details of any abnormalities.
A summary of the number and percentage of subjects with normal, abnormal not clinically significant and abnormal clinically significant ECG results at all applicable visits will be presented.
All ECG variables will be summarized (absolute values and change from Baseline) and listed by visit.
10.5.3 Other safety variablesPhysical ExaminationAbnormal results of the physical examination together with details of abnormalities will be listed by treatment group, subject, and visit.
Assessment of TuberculosisListings of history of latent TB, TB test results and ‘Evaluation of signs and symptoms of tuberculosis’ questionnaire data will be provided by treatment and subject.
Electronic Columbia Suicide Severity Rating Scale (eC-SSRS)The eC-SSRS is an assessment tool that evaluates suicidal ideation and behavior. The eC-SSRS contains 9 categories with binary responses (yes/no):
Category 1 –
Category 2 –
Category 3 –
Category 4 –
Category 5 –
Category 6 –
Category 7 –
Category 8 –
Category 9 –
Following composite endpoints based on the above categories are defined as:
Suicidal ideation: A “yes” answer at any time during treatment to any one of the five suicidal ideation questions (Categories 1-5).
Suicidal behavior: A “yes” answer at any time during treatment to any one of the five suicidal behavior questions (Categories 6-9).
Suicidal ideation or behavior: A “yes” answer at any time during treatment to any one of the ten suicidal ideation and behavior questions (Categories 1-9).
TEDyes/nyes/n
REDACTECOPYg Scag Sc
s suis sui
This do
cumen
tidi cann
oting cng c
SuiciSuicid
be ry 9 y 9 us
ed99 ––
ed to
supp
ort an
y mark
eting
izatio
ncidalcidalo):o):
autho
riz
appli
catio
nof of signsigt and st and
ale (eCle (ei
and a
ny ex
tensio
ns
malitimaliti
or va
riatio
ns
ted byed by
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 77 of 183
The incidence of subjects with suicidal ideation, behavior and injuries including the composite endpoints will be summarized by treatment group and visit. A by-subject listing of the eC-SSRS data will be provided.
Self-injurious behavior without suicidal intent is defined as event in the category non-suicidal self-injurious behavior.
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y exte
nsion
s or v
ariati
ons t
hereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 78 of 183
11 REFERENCESBerglund P, Heeringa S. Multiple imputation of missing data using SAS®. Cary, NC: SAS Institute Inc. 2014.
Bross I. D. J. How to use ridit analysis. Biometrics. 1958;14 (1):18-38.
CHMP/EWP/438/04. Guideline on Clinical Investigation of Medicinal Products for the Treatment of Psoriatic Arthritis (EMEA) Dec 2006.
Dauden E, Griffiths CE, Ortonne JP, Kragballe K, Molta CT, Robertson D, et al. Improvements in patient-reported outcomes in moderate-to-severe psoriasis patients receiving continuous or paused etanercept treatment over 54 weeks: the CRYSTEL study. J Eur Acad Dermatol Venereol. 2009;23(12):1374-82.
Fay MP, Feuer EJ. Confidence intervals for directly standardized rates: a method based on the Gamma distribution. Statistics in Medicine. 1997;16, 791-801.
Feldman SR. A quantitative definition of severe psoriasis for use in clinical trials. J Dermatolog Treat. 2004;15(1):27-9.
Felson DT, Anderson JJ, Boers M, Bombardier C, Chernoff M, Fried B, et al. The American College of Rheumatology preliminary core set of disease activity measures for rheumatoid arthritis clinical trials. The Committee on Outcome Measures in Rheumatoid Arthritis Clinical Trials. Arthritis Rheum. 1993;36(6):729-40.
Gossec L, de Wit M, Klitz U, Braun J, Kalyoncu U, Scrivo R, et al. A patient-derived and patient-reported outcome measure for assessing psoriatic arthritis: elaboration and preliminary validation of the Psoriatic Arthritis Impact of Disease (PsAID) questionnaire, a 13-country EULAR initiative. Ann Rheum Dis. 2014;73:1012-19.
Healy PJ, Helliwell PS. Measuring dactylitis in clinical trials: Which is the best instrument to use? J Rheumatol. 2007;34:1302-6.
Heinze, G., Schemper, M. 2002. A solution to the problem of separation in logistic regression. Statistics in Medicine, 21, 2109-2119.
Helliwell PS, Firth J, Ibrahim GH, Melsom RD, Shah I, Turner DE. Development of an assessment tool for dactylitis in patients with psoriatic arthritis. J Rheumatol. 2005;32:1745-50.
Heuft-Dorenbosch L, Spoorenberg A, van Tubergen A, Landewé R, van der Tempel H, Mielants H, et al. Assessment of enthesitis in ankylosing spondylitis. Ann Rheum Dis. 2003;62;127-132.
International Menopause Society. Cornwall, UK: Menopause Terminology; c2015. http://www.imsociety.org/menopause terminology.php. Accessed 25 Jan 2016.
Langley RG, Feldman SR, Han C, Schenkel B, Szapary P, Hsu MC, et al. Ustekinumab significantly improves symptoms of anxiety, depression, and skin-related quality of life in patients with moderate-to-severe psoriasis: Results from a randomized, doubleblind, placebo-controlled phase III trial. J Am Acad Dermatol. 2010;63(3):457-65.
Mallinckrodt CH, Lin Q, Lipkovich I, Molenberghs G. A structured approach to choosing estimands and estimators in longitudinal clinical trials. Pharm Stat. 2012;11:456-61.
REDACTED oncu Uoncu
sessingessinmpact opact o20142014
COPY aseasMeasMea
This do
cumen
tLangangsignsignp
cann
otionaona//wwwwww
glegle
be es
al Mal
sedosch Lsch L
ssmesme
todacdacsu
pport
2109210
Ibrahimbrahiactylcty
any02. A 02. A
99 2
marketi
ng4;73:4;73
dactylitactyli6..
autho
rizati
on
U, ScU, Scg psorg pso
of Disf Dis11
appli
catio
n M, FriedFrie
ctivity tivityures inures
andlinicinic an
y all
exten
sions
uomatol atol
thod bhod b
orus ous va
riatio
ns
mentment
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 79 of 183
Mantel N, Haenszel W. Statistical aspects of the Analysis of Data From Retrospective Studies of Disease. J Natl Cancer Inst. 1959; 22:719-48.
McInnes IB, Mease PJ, Kirkham B, Kavanaugh A, Ritchlin CT, Rahman P, et al. Secukinumab, a human anti-interleukin-17A monoclonal antibody, in patients with psoriatic arthritis (FUTURE 2): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet 2015;386:1137-46.
Mease PJ, McInnes IB, Kirkham B, Kavanaugh, A, Rahman, P, van der Heijde D, et al. Secukinumab inhibition of Interleukin-17A in patients with psoriatic arthritis. N Engl J Med 2015;373:1329-39.
Nam, J.M. A simple approximation for calculating sample sizes for detecting linear trend in proportions. Biometrics 1987;43(3):701-5.
Rubin D.B. 1987. Multiple Imputation for Nonresponse in Surveys. New York: John Wiley and Sons.
Snaith RP, Zigmond AS. The hospital anxiety and depression scale, with the irritability depression-anxiety scale and the Leeds situational anxiety scale manual. 1994.
Ulm K. A simple method to calculate the confidence interval of a standardized mortality ratio. American Journal of Epidemiology. 1990;131, 373-375.
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tiona stana stan
and
th hnual. 1ual. 1
any
the the iiex
tensio
nsrendend
k: John: Joh
or d ind iva
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 80 of 183
12 APPENDICES12.1 Classification Criteria for Psoriatic Arthritis (CASPAR) CriteriaA score will be derived by adding the points from each of the 5 categories. The maximum score will be 6. The CASPAR criteria are fulfilled, if the score is ≥3 and Inflammatory articular disease (joint, spine, or entheseal):
Table 12‒1: CASPAR Criteria
Category Definition Points
1) Evidence of psoriasis:Current psoriasis or
Personal history of psoriasis or
Family history of psoriasis
Psoriatic skin or scalp disease present today as judged by a dermatologist or rheumatologistA history of psoriasis that may be obtained from the subject, family physician, dermatologist, rheumatologist, or other qualified health care providerA history of psoriasis in a first- or second-degree relative according to subject report
2 points
1 point
1 point
2) Psoriatic nail dystrophy Typical psoriatic nail dystrophy, including onycholysis, pitting, and hyperkeratosis, observed on current physical examination
1 point
3) A negative test for rheumatoid factor By any method except latex, but preferably by enzyme-linked immunosorbent assay (ELISA) or nephelometry, according to the local laboratory reference range
1 point
4) Dactylitis:Current dactylitis orHistory of dactylitis
Swelling of an entire digitA history of dactylitis recorded by a rheumatologist
1 point1point
5) Radiologic evidence of juxta-articular new bone formation
Ill-defined ossification near joint margins (but excluding osteophyte formation) on plain radiographs of the hand or foot
1 point
CASPAR=Classification Criteria for Psoriatic Arthritis
12.2 Calculation rules for duration of adverse eventsThe calculation rules for duration of AEs are presented in Table 12‒2. AE duration is computed and reported in day.
Table 12‒2: Calculation rules for duration of adverse events
Data Availability
Onset Date
Outcome Date
Calculation Rules
Complete data D1 D2 Duration = D2 – D1 + 1
REDACTEBy any y anyenzyenzynen
EDrrentrrent
TE
ED
COPY oriatic oriatic
ysis, psis, pp
Y
This do
cumen
tand rnd r
TT
cann
ot e calcucalcu
re
be a us
ed
ssificsificed
to su
pport
of juxtaof juxtp
oan
y mark
eting
ephphrefererefere
kekekeket
ke
autho
rizy methomethymeme--linlihelohel
izatio
npittingittingphysichysiciza
appli
catio
n a firsta firstsubject ubjec
nail dyail dya
pl
andied heed he
anyaineain
atologatologex
tensio
ns
y as as ogistgistnedned
ss orPoPoova
riatio
ns
oiniva
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 81 of 183
Table 12‒2: Calculation rules for duration of adverse events
Data Availability
Onset Date
Outcome Date
Calculation Rules
Start date missing - D2 Duration = < D2 – D0 + 1Where, for a subject in the SS, D0 is the date of the first intake of IMP, and for screen failures, D0 is the date of the Screening visit day
End date missing D1 - Duration = > Final contact date – D1 + 1For resolved and ongoing AE Duration
Start and end date missing
- - Duration = > Final contact date – D0 + 1For resolved and ongoing AE Duration Where, for subjects in the SS, D0 is the date of the first intake of IMP, and for screen failures, D0 is the date of the Screening visit day
12.3 Treatment group assignment for tables and figuresTable 12‒3 displays the treatment group labels for each data type. This overview clarifies what kind of treatment groups will be used for producing the tables and figures separated between the different table types.
Table 12‒3: Treatment group assignment for tables and figures
Placebo BKZ 16mg, BKZ 160mg w/LD /BKZ 160mg/ BKZ 320mg
All BKZ
All subjects
Subject disposition X X X
Important protocol deviation X X X
Demographics/Lifestyle X X X
Ankylosing spondylitis history X X X
Baseline characteristics X X X
Tuberculosis testing X X X
Previous and ongoing medical history
X X X
Prior and concomitant medication
X X X
nmennmen
ceboceboACACACAC
COPY ch dchg the tg the
This do
cumen
tPreviPrevihishis
umu
cann
ot culosiulosit
ca
bearactacb
used
dylitiliti
terterse
to tis htissu
pport
or
s
any
Xyy
anmark
eting
BBBng
Xke
autho
rizati
on ta
nt fornt for
BKBKth
appli
catio
n es ands anta typea typ
ables bles
and eenen
aan
yD0 iD0 failufai
exten
sions
+ 1+ 1rationation
ioor
varia
tions
st e of of on
va
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 82 of 183
Table 12‒3: Treatment group assignment for tables and figures
Placebo BKZ 16mg, BKZ 160mg w/LD /BKZ 160mg/ BKZ 320mg
All BKZ
All subjects
Rescue and prohibited medication
X X X
Bimekizumab compliance X X X
Extent of exposure X X X
Efficacy analysis X X
Plasma/Bimekizumab concentration
X X
Pharmacokinetic variables/Biomarker data/AbAb
X X
AEs X X X
Safety laboratory tests X X X
Vital signs/Body weight/ECG/ Physical examination
X X X
eC-SSRS X X X
BKZ=bimekizumab; eC-SSRS= Electronic Columbia-Suicide Severity Rating Scale; LD=loading dose.
XX D
ACTCCCCTAC
COPY YC
This do
cumen
t can
not b
e use
d to s
uppo
rt any
ctronicctron
marketi
ng
ngeti
m
autho
rizati
on
tio
XX ho
appli
catio
n
pp
and
aan
y y exten
sions
ns
ns
e
or ova
riatio
ns
onria
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 83 of 183
12.4 Hematology and biochemistry variables for the interim analysis
Table 12‒4: Selected hematology and biochemistry variables for the interim analysis
Hematology Biochemistry
Lymphocytes AST
Neutrophils ALT
Hematocrit Total bilirubin
Hemoglobin
Platelet count
ALT=alanine aminotransferase; AST=aspartate aminotransferase.
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd
anan
y ny ex
tensio
nss
en
or va
riatio
ns
onononon
ar
thereo
f.
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y exte
nsion
s or v
ariati
ons t
hereo
f.
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y exte
nsion
s or v
ariati
ons t
hereo
f.
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y exte
nsion
s or v
ariati
ons t
hereo
f.
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y exte
nsion
s or v
ariati
ons t
hereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 88 of 183
12.7 Maastricht Ankylosing Spondylitis Enthesitis Index (MASES) questionnaire
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd d aaaaaaaannnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn
yyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyy
sions
eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeexxxxxteeeeeeeennnnnnnnnnnnnnnnnnnnnnnnnnnn
sior
varia
tions
there
of.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 89 of 183
12.8 Psoriatic Arthritis Impact of Disease-9 (PsAID-9)
RRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRREEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAACCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCTEEEEEDDDDDDDDDDDDDDDDD CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCOOOOOOOOOOOOOOOOOOPPPPPPYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYY
This do
cummmmmmeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeennnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn
ttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttt cccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaannn
nooooooooottttttttttttttttt bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb
eeeeee uuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuussssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssseeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee
dddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddddd ttto ss
uuuuuuuuuuuuuppppppppppppppppppppppppppppppppppooooooooooooooooooo
rrrrrrrrrrrrrrrrrttttttttttttttttttttttttt aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaannnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyy
mmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmaaaarkk
eeeeeeeetttttttttttiiiiiiiiiiinnnngggggggggggggggggggggggggggggggggggggggggggggggggggggg
aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaauuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuutttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttthhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhoooooo
rizaaaaaaaaaaaaaaatttttttttttttttttiiiiiiiiii
ooooooooooooooooooooooooonnnnnnnnnnnnnnnnnnnnnnnnnnnnnn appppppppppp
pppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllliiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiicccccccccccccca
ttiioooooooonnnn aaaaaaaaaaaaaaaaaaaaaaaaaaaa
nnnnnnnnnnnnddddddddddddd an
y exte
nsionnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn
sssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssss or
riatio
ns
varia
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 90 of 183
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y eexxxxxttttteee
nnnnssssiiiiiiiiiioooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooonnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn
ssssssssssssssssssssssssssss or
ions
varia
tiothe
reof.
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y exte
nsion
s or v
ariati
ons t
hereo
f.
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y exte
nsion
s or v
ariati
ons t
hereo
f.
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y exte
nsion
s or v
ariati
ons t
hereo
f.
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y exte
nsion
s or v
ariati
ons t
hereo
f.
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y exte
nsion
s or v
ariati
ons t
hereo
f.
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y exte
nsion
s or v
ariati
ons t
hereo
f.
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y exte
nsion
s or v
ariati
ons t
hereo
f.
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y exte
nsion
s or v
ariati
ons t
hereo
f.
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y exte
nsion
s or v
ariati
ons t
hereo
f.
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y exte
nsion
s or v
ariati
ons t
hereo
f.
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y exte
nsion
s or v
ariati
ons t
hereo
f.
REDACTED COPY
This do
cummen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y exte
nsion
s or v
ariati
ons tt
hhereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 103 of 183
Step 1: Refer to column B of the spreadsheet (Opportunistic infections MedDRA v 19.xlsx) which identifies the PTs to be classified as opportunistic infections using either a single ‘x’ or a double ‘xx’.
TEAEs which code to a PT flagged with a single ‘x’ need to also be serious in order to be considered an opportunistic infection.
All TEAEs which code to a PT flagged with a double ‘xx’ are considered to be an opportunistic infection, regardless of seriousness.
Step 2: Refer to column C of the spreadsheet which identifies the PTs that need to be evaluated on a case-by-case basis by the study physician in order to determine whether it is a true opportunistic infection or not. The process for physician review is as follows:
1. Study programming team creates a spreadsheet which lists all of the subjects with a TEAE present in the database which codes to a PT identified as case-by-case. Information from the AE dataset to be included in the spreadsheet: Subject ID, AE verbatim term, SOC, High Level Term, Lower Level Term, PT, AE start date, AE end date, seriousness, severity, relationship to study medication, action taken. Additionally, a column will be included where the study physician can document their decision on the case.
2. Study physician reviews the cases in the spreadsheet and indicates in the additional column which AEs are confirmed to be opportunistic infections via a single ‘x’.
3. Study programming team incorporates these decisions into the AE dataset by merging the study physician decisions for individual subjects / PTs and flagging the confirmed opportunistic infections as such in the dataset.
All subjects with a case-by-case PT reported that has been confirmed by the study physician to be an opportunistic infection will be summarized as such in the stand-alone table, along with all of the events identified in Step 1 of this process.
The timing and frequency of Step 2 should be outlines and agreed to by the study team at the beginning of the study. It is suggested that this process be executed multiple times throughout the course of the study, more frequently in the weeks leading up to database lock, and one final time immediately prior to database lock.
Following the initial physician review of case-by-case events, subsequent reviews will be based on the cumulative set of case-by-case events present in the database at each time point of spreadsheet creation. Physician decisions from previous runs should be retained in each subsequent run. The final run of the spreadsheet, with all study physician decisions on the full set of case-by-case events, will be archived at the conclusion of the study.
12.14 MedDRA algorithmic approach to anaphylaxisThe SMQ Anaphylactic reaction consists of three parts:
A narrow search: If a subject reports any TEAE which codes to a PT included in Category A, then the event will be flagged as an anaphylactic reaction and summarized as such in the table.
Category A
REDACTED se dse dsubjesubje
e datae data
reportreportsumsum
COPY eeteenfectionfectio
decidec
This do
cumen
t The The ca
nnotbyby--cc
1414
bet runt runusedve sese
reatioeatiotoal phal p
ettsu
pportfrequrequ
atabasetabas
physhy
any p p
uggestggesuen
marketi
ng ted ed
mmarimmarithis prohis p
2 sh2 sh
autho
rizati
on
cisionssionsects / Pcts / P
aset.set.
d thatha
appli
catio
n coloe.
nd indicd indns vians via
andserioerio
olumlum
any nn
m termm terus
exten
sions
ue e
cts wits wnformfor
oraluaalu va
riatio
ns
uateuat
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 104 of 183
A broad search: If a subject reports any TEAE which codes to a PT included in Category B AND reports any TEAE which codes to a PT included in Category C, and both TEAEs have the same start date, then both events will be flagged as anaphylactic reactions and summarized as such in the table.
Category B
Category C
RRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRREEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAACCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDD CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPPYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYY
This do
cumen
t can
not b
e use
d to s
ussuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp
ort nyan
yyyyyyyyyyyyy mmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaarrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkk
eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeetttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiinnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnggggggggggggggggggggggggggggggggggggggggggg
aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaauuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuutttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttthhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhoooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrriiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiizzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaatttttttttttttttttttttttttttttttttttttttttttttttttttttttiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii
oooooooooooooooooon
ation
appli
cati
and a
ny ex
tensio
nsCategCategoth TEth TE
ions anns a
or va
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 105 of 183
Category D
An algorithmic approach: If a subject reports any TEAE which codes to a PT included in Category D AND reports (either a TEAE which codes to a PT included in Category B OR a TEAE which codes to a PT included in Category C), and both TEAEs have the same start date, then both events will be flagged as anaphylactic reactions and summarized as such in the table.
REDACTED ports ports
EAE wEAE wed in Ced in
aggedagged
COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
etingd as ad as aa
uthori
zatio
n any Tany T
which whichCategCateg
appli
catio
n and
any e
xtens
ions o
r vari
ation
s the
reof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 106 of 183
13 AMENDMENT(S) TO THE STATISTICAL ANALYSIS PLAN13.1 Amendment 113.1.1 Rationale for the amendmentThe main objectives of this SAP amendment are to describe and specify the extension of time point up to Week 24 for the interim analysis. The SAP has been amended to:
Add additional analyses for the time points up to Week 24 for ACR, PASI, MDA, TEAEs, Hy’s Law, PK, and AbAb data
Add the definition for rheumatoid factor and anti-CCP antibodies positivity and negativity
Add clarification for handling injected joints for the swollen and tender grading
Add assessment of potential bias of potentially unblinding at the entry of Dose-Blind phase
Refine the categories of the subgroup variable Psoriasis BSA
13.1.2 Modifications and changesChange # 1:
Section 4.3 Interim analysis and data monitoring
After all enrolled subjects have completed the 12 weeks Double-Blind Period, an interim analysis will be performed (i) to analyze the dose:exposure response for ACR and PASI response criteria to determine the optimal therapeutic dose(s) for subsequent studies, and (ii) to perform a comprehensive evaluation of all double-blind data of the study.
No separate SAP for the interim analysis will be provided. The interim analysis is a limited version of the final analysis and will focus on specific topics, eg, primary and secondary efficacy analysis and comprise the results of the 12 weeks Double-Blind Period only. The TFL shells for the interim and final analysis will be provided in two different documents.
The interim analysis will summarize disposition, demographics, PsA history, Baseline characteristics, PsA concomitant medication, efficacy analysis, AEs and treatment-emergent AEs (derivation Section 10.2), hematology and biochemistry laboratory data, and Electronic Columbia-Suicide Severity Rating Scale (eC-SSRS). Corresponding listings will be provided. Not all collected hematology and biochemistry laboratory variables will be summarized and listed. For the interim analysis only selected variables will be provided as listed in Table 12‒4.
If an AE occurs on the same day as the treatment switch (Visit 7, Week 12) then the AE will be allocated to the Double-Blind treatment. An exception from this general rule is made for adverse drug reactions (ADRs), and AE of hypersensitivity reactions and anaphylactic reactions. If those AEs occur on the same days as the treatment switch, the AEs will be allocated to the Dose-Blind treatment and will not be presented for the interim analysis.
Exposure of study medication will not be summarized but a listing will be provided including following information: date of first dose, date of last dose, and duration of exposure. The last dose for the Double-Blind Period is planned on Visit 6 (Week 8). For subject who completed the Double-Blind Period the Week 8 date is the date of the last dose. If a subject discontinued within the Double-Blind Period, the discontinuation date will be presented for the date of last dose.
REDACTED oseosedose(dose(
blind dblind d
ysis wysis wl focl foc
COPY weeweeksk
e:exp:ex
This do
cumen
t EsEstreattreat
E
cann
otd to to reactireact
s ococ
beccurccurthth
used
hemheinterimnterim
toematmasu
pport
mitanmita), hem), hem
verityverity
anymarizmariz
ant
marketi
ng wilw
cus onus onthe 12 he 12
be probe pr
autho
rizati
on s DD
xposurposur(s) fors) for
data odata o
ll bel be
appli
catio
n
DoubDou
and a
ny ex
tensio
nsnegaega
ingng
f DoseDos
or ativtiva
riatio
ns
Es, s,
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 107 of 183
The interim efficacy analysis will focus on the primary and secondary efficacy analyses including supportive and sensitivity analysis. In addition, summary tables for the primary and secondary efficacy variables, and the ACR components will be provided. Categorical variables will be summarized using frequency tables by each visit. Continuous variables will be summarized using descriptive statistics by each visit. Time to onset of ACR50 response and ACR20 response will be summarized and plotted as described in Section 8.3.
PK and PD data will not be analyzed in the interim analysis. A separate PKPD analysis based on a separate Data Analysis Plan (DAP) will be performed after Week 12.
…
Has been changed toAfter all enrolled subjects have completed the 12 weeks Double-Blind Period, an interim analysis will be performed (i) to analyze the dose:exposure response for ACR and PASI response criteria to determine the optimal therapeutic dose(s) for subsequent studies, and (ii) to perform a comprehensive evaluation of all double-blind data of the study.
No separate SAP for the interim analysis will be provided. The interim analysis is a limited version of the final analysis and will focus on specific topics, eg, primary and secondary efficacy analysis and comprise the results of the 12 weeks Double-Blind Period only. The TFL shells for the interim and final analysis will be provided in two different documents.
The interim analysis will summarize disposition, demographics, PsA history, Baseline characteristics, PsA concomitant medication, efficacy analysis, AEs and treatment-emergent AEs (derivation Section 10.2), hematology and biochemistry laboratory data, and Electronic Columbia-Suicide Severity Rating Scale (eC-SSRS). Corresponding listings will be provided.Not all collected hematology and biochemistry laboratory variables will be summarized and listed. For the interim analysis only selected hematology and biochemistry variables will be provided as listed in Table 12‒4. Listings of AEs, hematology, biochemistry and C-SSRS data will be provided.The interim efficacy analysis will focus on the primary and secondary efficacy analyses including supportive and sensitivity analysis. In addition, summary tables for the primary and secondary efficacy variables, and the ACR components, and MDA will be provided. Categorical variables will be summarized using frequency tables by each visit. Continuous variables will be summarized using descriptive statistics by each visit. Time to onset of ACR50 response and ACR20 response will be summarized and plotted as described in Section 8.3.
PK and PD data AbAb data will not be analyzed in the interim analysis as described in Section 9.1 and Section 9.2 . A separate PKPD analysis based on a separate Data Analysis Plan (DAP) will be performed after Week 12. Biomarker data will not be analyzed in the interim analysis.
The interim analysis includes selected data up to Week 12 and in addition data up to Week 24. The data up to Week 24 may include partial results up to the time the database for Week 12 was locked for ongoing subjects. All analyses for the period between Week 12 and Week 24 are only for those treatment groups that are not rerandomized. Following variables will be analyzed: primary and secondary efficacy variables, ACR components, MDA, PK, AbAb, TEAEs and Hy’s Law. It depends on the analysis how those additional
REDACTED n, n, n, efficn, effic
d biocd biocale (eCale (eC
ochemochem
COPY ublbwo diffwo dif
demde
This do
cumen
tanalanalen
TT
cann
ot DDtn 9.19.1
AP) wAP) wca
be
D datD datbus
edng dng dnse wnse w
toummmmes
supp
ort is ws w
d sensitsensriablesiablemarma
any
will wilmark
etingmistrmistr
electedelecteListingsisting
ma
autho
rizati
on
mogramogracacy aacy a
chemichemC-SSRSSR
appli
catio
nnterimterimeg, primg, pri
lind Pind Pferent erent
and
m
anyes, anes, anex
tensio
ns
an inan inR anR an
or va
riatio
ns
on on
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 108 of 183
time points will be presented in the interim analysis. Two summary tables for each of the primary and secondary efficacy variables, the ACR components, and MDA will be presented. The first table will include data up to Week 12 by treatment group at Baseline. The second table will include data up to Week 24 by following treatment groups: bimekizumab 160mg and bimekizumab 160mg with loading dose 320mg, and bimekizumab 320mg. The same approach will be done for all TEAEs tables and the Hy’s Law criteria tables. The summary table for plasma concentration will present only data up to Week 12 for the Baseline treatment groups Placebo and bimekizumab 16mg. Data up to Week 24 will be presented for bimekizumab 160mg, bimekizumab 160mg with loading dose 320mg, and bimekizumab 320mg. All Listings which include data up to Week 24 will use the same approach as for plasma concentration tables.If an AE occurs on the same day as the treatment switch (Visit 7, Week 12) then the AE will be allocated to the Double-Blind treatment. An exception from this general rule is made for adverse drug reactions (ADRs), and AE of hypersensitivity reactions and anaphylactic reactions. If those AEs occur on the same days as the treatment switch, the AEs will be allocated to the Dose-Blind treatment and will not be presented for the interim analysis.
Exposure of study medication will not be summarized but a listing will be provided including following information: date of first dose, date of last dose, and duration of exposure. The last dose for the Double-Blind Period is planned on Visit 6 (Week 8). For subject who completed the Double-Blind Period the Week 8 date is the date of the last dose. If a subject discontinued within the Double-Blind Period, the discontinuation date will be presented for the date of last dose.
…
Change # 2:
Section 8.1.2.1 Tender joint count (TJC) and swollen joint count (SJC)
Section 8.1.2.1.1 78/76 joint evaluation for ACR response The Principal Investigator, another delegated physician, or an appropriately qualified medical professional (based on local requirements) who has had documented training on how to perform these assessments correctly assesses the following joints for tenderness: the distal interphalangeal, proximal interphalangeal and metacarpophalangeal joints of the hands, and metatarsophalangeal joints of the feet, the carpometacarpal and wrist joints (counted separately), the elbows, shoulders, acromioclavicular, sternoclavicular, hip, knee, talo-tibial, and midtarsaljoints. All of these except for the hips are assessed for swelling.
Artificial and ankylosed joints, as well as missing joints (ie, amputated joints), are excluded from both tenderness and swelling assessments.
One dactylitic digit is to be counted as 1 swollen joint (instead of counting as 3 in the finger or 2 in the toe).
Table 8‒1 summarizes the swelling and tenderness grading criteria.
REDACTEDdate date
(T(T
COPYit 6 (Wt 6 (of theof thee we w
This do
cumen
t nenein thin th
T
cann
ot l anan
endernnder
e dacdac
be
nd and aus
edulderlders. Als. Al
toal joal jors
supp
ort l rere
tly asly assmal inmal inointoin
anyther ther
equirequmark
eting
TJC) anJC) an
uation ationded
autho
rizati
one laslas
will beill beap
plica
tion ting wiing w
nd durnd duWeek 8Week st dst d
andalloclocan
y acactic ticcatecat
exten
sions
thsen the n the
is mis m
orrhe she sva
riatio
ns
2 t24 4 ar20mg0mgv
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 109 of 183
Table 13‒1: Swelling and tenderness grading
The tender joint counts (TJC) and swollen joint counts (SJC) are weighted joint counts. If there are missing observations in the tender or swollen joint assessments, then the remaining observations will be assessed and weighted by the number of the assessed joints (AJ):
(9)
(10)
where n is the number of total joints. If a joint is missing at Baseline, then that joint is set to missing throughout the study. If more than 50% of the tender joint assessments or 50% of the swollen joint assessments are missing at the time of a given assessment, then no imputation willbe done and the total TJC or SJC will be set to missing for that visit.
Section 8.1.2.1.2 28 joint evaluation for determination of DAS28(CRP)
The following 28 joints will be used for calculation of the DAS28(CRP).
Upper extremity (26)-bilateral shoulders, elbows, wrists (includes radiocarpal, and carpal and carpometacarpal bones considered as a single unit), MCP I, II, III, IV, and V, thumb interphalangeals (IP), PIP II, III, IV, and V
Lower extremity (2)-knees
Has been changed to
Section 8.1.2.1 Tender joint count (TJC) and swollen joint count (SJC)
Section 8.1.2.1.1 78/76 joint evaluation for ACR response The Principal Investigator, another delegated physician, or an appropriately qualified medical professional (based on local requirements) who has had documented training on how to perform these assessments correctly assesses the following joints for tenderness: the distal interphalangeal, proximal interphalangeal and metacarpophalangeal joints of the hands, and metatarsophalangeal joints of the feet, the carpometacarpal and wrist joints (counted separately), the elbows, shoulders, acromioclavicular, sternoclavicular, hip, knee, talo-tibial, and midtarsaljoints. All of these except for the hips are assessed for swelling.
REDACTEDnt is mnt is mn 50%50%
at the tat the tll be sll be s
COPY
This do
cumen
t he heprofeprofethth
cann
ot n 8n 8.11e PriPri
be1.2.11.2.1used
nged nged
toy (2y (2supp
orts conconP), PIP), PIP
2)2) k
any teral seral
i
marketi
ngset toet t
n for dfor d
ed fored fo
autho
rizati
on
missingmissin% of th% of th
time otime ototo
appli
catio
n and
any e
xtens
ions unts. Iunts. I
maininaininints (Ants (A
s oor
ns
varia
tion
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 110 of 183
Artificial and ankylosed joints, as well as missing joints (ie, amputated joints), are excluded from both tenderness and swelling assessments.
One dactylitic digit is to be counted as 1 swollen joint (instead of counting as 3 in the finger or 2 in the toe).
Table 8‒1 summarizes the swelling and tenderness grading criteria.
Table 13‒1: Swelling and tenderness grading
The tender joint counts (TJC) and swollen joint counts (SJC) are weighted joint counts. If there are missing observations in the tender or swollen joint assessments, then the remaining observations will be assessed and weighted by the number of the assessed joints (AJ):
(9)
(10)
where n is the number of total joints. If a joint is missing at Baseline, then that joint is set to missing throughout the study. If more than 50% of the tender joint assessments or 50% of the swollen joint assessments are missing at the time of a given assessment, then no imputation will be done and the total TJC or SJC will be set to missing for that visit.
Injected joints will be counted as swollen and tender from the date of injection up to 52 weeks after injection, ie, swelling and tenderness grading is 1.
Section 8.1.2.1.2 28 joint evaluation for determination of DAS28(CRP)
The following 28 joints will be used for calculation of the DAS28(CRP).
Upper extremity (26)-bilateral shoulders, elbows, wrists (includes radiocarpal, and carpal and carpometacarpal bones considered as a single unit), MCP I, II, III, IV, and V, thumb interphalangeals (IP), PIP II, III, IV, and V
Lower extremity (2)-knees
Injected joints will be counted as swollen and tender from the date of injection up to 52 weeks after injection, ie, swelling and tenderness grading is 1.
Change # 3
REDACTED COPYmbemb
This do
cumen
tcccann
otlowinowi
UppeUppeca
be 2.2.
ngus
eddectioctioed1.21.2
tol bel betoonsu
pporty
are mare mJC or SC or
e coe c
any inin
If moIf mmi
marketi
ng
nts. Ints.
autho
rizati
on ap
plica
tione weig weig
ments, mentsof the of the
anand a
ny ex
tensio
ns or
ariati
ons
arvar
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 111 of 183
Section 6.2 Other Baseline characteristics
…
Baseline characteristics (including Scores relevant for inclusion and exclusion criteria) will be summarized by treatment group and overall for FAS and SS. Following variables will be summarized:
TJC
SJC
hs-CRP
Rheumatoid factor
Anti-CCP antibodies
Prior NSAID therapy, and prior anti-TNF therapy
Psoriasis BSA
Nail psoriasis
Dactylitis
NSAID therapy
Synthetic DMARDs
MTX
Sulfasalazine
Hydroxychloroquine
The corresponding listings will be presented for the RS.
Has been changed to
…
Baseline characteristics (including Scores relevant for inclusion and exclusion criteria) will be summarized by treatment group and overall for FAS and SS. Following variables will be summarized:
TJC
SJC
hs-CRP
Rheumatoid factor
Anti-CCP antibodies
Prior NSAID therapy, and prior anti-TNF therapy
Psoriasis BSA
Nail psoriasis
REDACTED COPY
This do
cumen
t can
not C
hshs--CC
be us
ed toatmeatm su
pport
cs (inccs (incen
any m
arketi
ng
e presepres
autho
rizati
on ap
plica
tion a
nd an
y exte
nsion
s or v
ariati
ons t
hereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 112 of 183
Dactylitis
NSAID therapy
Synthetic DMARDs
MTX
Sulfasalazine
Hydroxychloroquine
A rheumatoid factor value < 15 IU/mL is defined as negative and a value ≥ 15 IU/mL is defined as positive. Anti-CCP antibodies negative is a value < 5 U/mL and a value ≥ 5 U/mL is defined as positive.
The corresponding listings will be presented for the RS.
Change # 4:
Following text has been added at the end of
Section 8.3 Analysis of other efficacy variables
…In order to assess the potential bias of the potentially unblinded subjects based on data, the results of the final analysis will be utilized. Results of the ACR50 and PASI90 response of the rerandomized 320mg and 160mg treatment group at Week 48 will be compared to the potentially unblinded, randomized 320mg and 160mg treatment groups. If the percentage of ACR50 and PASI90 responders is comparable as assessed by overlapping 95% confidence intervals, the potential bias is estimated to be at minimum.
Change # 5:
Section 8.4 Subgroup analysis
and
Section 4.8 Examination of subgroups
…
Extent of psoriasis involvement >3% (yes, no)
…
Has been changed to
…
Extent of psoriasis involvement (<3%, ≥3% to <10%, ≥10%)
…
REDACTEDu
group groupg treattreat
ssessessesseminimuminimu
COPY lly unbly unb
lts olts o
This do
cumen
tEEca
nnoten chn chb
e haha
usedoriasoriasto
sisi
supp
ort
n of sun of san
y mark
eting
umum.. autho
rizati
onnbl
f the f the Aat Weat We
tment tment ed by od by o
appli
catio
n
lindeinde
and a
ny ex
tensio
nsmsalue ≥ alue ≥
ionormL imL o
varia
tions
there
of.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 113 of 183
13.2 Amendment 213.2.1 Rationale for the amendmentThe SAP has been amended to:
Add ‘End of Treatment’ visit for the final analysis
Add clarification for the different handling of additional time points in the interim analysis
Add rule for missing hs-CRP values at Baseline
Correct inflammatory bowel disease coding
Add rule for handling missing data for prior and concomitant medication
Modify Section 10
13.2.2 Modifications and changesSection 3.5 Mapping of assessments performed at early termination visit
Study assessments at an early termination visit where visit date matches the visit date of a scheduled visit will be summarized at the scheduled visit with the same visit date. Premature study termination visit assessments that do not have a scheduled visit with a matching date will be assigned to the next scheduled site visit following the last visit where assessments were available regardless whether or not there will be an assessment on this visit. The assessment of AbAb is an exception to this rule: AbAb will be mapped to the next visit where antibody levels are measured. For subjects who discontinue study treatment early and return for the Week 12 visits as per the protocol, the assessments collected at that visit are summarized as Week 12 assessments.
Has been changed toStudy assessments at an early termination visit where visit date matches the visit date of a scheduled visit will be summarized at the scheduled visit with the same visit date. Premature study termination visit assessments that do not have a scheduled visit with a matching date will be assigned to the next scheduled site visit following the last visit where assessments were available regardless whether or not there will be an assessment on this visit. The assessment of AbAb is an exception to this rule: AbAb will be mapped to the next visit where antibody levels are measured. For subjects who discontinue study treatment early and return for the Week 12 visits as per the protocol, the assessments collected at that visit are summarized as Week 12 assessments.
A visit called ‘End of Treatment’ will be added for the TFLs in the final analysis. The visit will display the Week 48 assessments or the early termination assessment depending if the subject discontinued early or not.
Section 4.2.3 Handling of missing data for prior and concomitant medicationAny medications with incomplete start and end dates/times will be handled according to the following rules for classification as prior and concomitant and for the calculation of relative study days. Such imputations will only be performed for these classifications and calculations; in the listings all data will be shown as recorded on the electronic case report form (eCRF).
REDACTEDbe mbe me ste studud
nts collts col
COPYscs
ng the ng thean assan assma
This do
cumen
tsubjubjenSeS
cann
ot t callecallennl dispdispcaec
be ts.ts. us
edor sur su
he proe proto
on ton ubub
supp
ort ssmsm
chedulhedulhetherhetheto tto
anyzedzed
mentmentmark
eting
minatiominatid atd a
autho
rizati
onsessmsessmappedpped
dy treay trealectedected
appli
catio
nmatchatchthe samhe sa
duled vuled vlast viast vme
andon vn v
hehe
any
visiis
exten
sions
or va
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 114 of 183
Imputation of partial start dates:
If only the month and year are specified and the month and year of first dose is not the same as the month and year of the start date and the subject did not switch treatment during that month and year, then use the 1st of the month.
If only the month and year are specified and the month and year of first dose is the same as the month and year of the start, then use the date of first dose.
If only the month and year are specified and the subject did switch treatment during that month and year, then use the date of first switch treatment.
If only the year is specified, and the year of first dose is not the same as the year of the start date and the subject did not switch treatment during that year, then use the 1st of January of the year of the start date.
If only the year is specified, and the year of first dose is the same as the year of the start date, then use the date of first dose.
If only the year is specified, and the year of first dose is the same as the year the subject did switch treatment, then use the date of first switch treatment.
If only the year is specified, and the year of first dose is the same as the year of the start date and the same as the year the subject did switch treatment, then use the date of first dose.
If the start date is completely unknown and the stop date is unknown or not prior to the date of first dose, then use the date of first dose.
…
Has been changed to
Any medications with incomplete start and end dates/times will be handled according to the following rules for classification as prior and concomitant and for the calculation of relative study days. Such imputations will only be performed for these classifications and calculations; in the listings all data will be shown as recorded on the electronic case report form (eCRF).
Imputation of partial start dates:
If only the month and year are specified and the month and year of first dose is not the same as the month and year of the start date and the subject did not switch treatment during that month and year, then use the 1st of the month.
If only the month and year are specified and the month and year of first dose is the same as the month and year of the start, then use the date of first dose.
If only the month and year are specified and the subject did switch treatment during that month and year, then use the date of first switch treatment.
If only the year is specified, and the year of first dose is not the same as the year of the start date and the subject did not switch treatment during that year, then use the 1st of January of the year of the start date.
If only the year is specified, and the year of first dose is the same as the year of the start date, then use the date of first dose.
REDACTEDd thed those.ose.
COPY dosedostreatmtreatm
he she s
This do
cumen
t I cann
ot ly ty
e mon mon
If oIf o
be
the theus
edh andandd yearyear
toth ath ad yd
supp
ort showhow
art dateart dat
andan
anywill owill
own wn
marketi
ng
art andrt anas prios pri
on
autho
rizati
on m
stop dtop dap
plica
tionme me
s the sathe sment, tent, t
and
e asas
anye yee yeex
tensio
nsr of thr of thof Jaof Ja
or va
riatio
ns
at
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 115 of 183
If only the year is specified, and the year of first dose is the same as the year the subject did switch treatment, then use the date of first switch treatment.
If only the year is specified, and the year of first dose is the same as the year of the start date and the same as the year the subject did switch treatment, then use the date of first dose.
If the start date is completely unknown and the stop date is unknown or not prior to the date of first dose, then use the date of first dose.
If the start date is completely unknown and the stop date is prior to the date of first dose, then set the start date to the 1st of January of the year of the end date.
…
Section 4.3 Interim analysis and data monitoring
…
The interim analysis includes selected data up to Week 12 and in addition data up to Week 24. The data up to Week 24 may include partial results up to the time the database for Week 12 was locked for ongoing subjects. All analyses for the period between Week 12 and Week 24 are only for those treatment groups that are not rerandomized. Following variables will be analyzed: primary and secondary efficacy variables, ACR components, MDA, modified MDA, PK, AbAb, TEAEs and Hy’s Law. It depends on the variable how those additional time points will be presented in the interim analysis. Two summary tables for each of the primary and secondary efficacy variables, the ACR components, MDA and modified MDA will be presented. The first table will include data up to Week 12 by treatment group at Baseline. The second table will include data up to Week 24 by following treatment groups: bimekizumab 160mg and bimekizumab 160mg with loading dose 320mg, and bimekizumab 320mg. The same approach will be done for all TEAEs tables and the Hy’s Law criteria tables. The summary table for plasma concentration will present only data up to Week 12 for the Baseline treatment groups Placebo and bimekizumab 16mg. Data up to Week 24 will be presented for bimekizumab 160mg, bimekizumab 160mg with loading dose 320mg, and bimekizumab 320mg. All Listings which include data up to Week 24 will use the same approach as for plasma concentration tables. …
For the interim analysis, the database will be locked and the treatment codes will be made available to UCB personnel with exception of operational staff working on the study. An interim report will be written. The Investigators and subjects will remain blind to the assigned bimekizumab dosing regimen until the subject completes the Dose-Blind Period at Week 48.
…
Has been changed to
…
The interim analysis includes selected data up to Week 12 and in addition data up to Week 24. The data up to Week 24 may include partial results up to the time the database for Week 12 was locked for ongoing subjects. All analyses for the period between Week 12 and Week 24 are only for those treatment groups that are not rerandomized. The analysis of following variables will include data up to Week 24Following variables will be analyzed: primary and secondary
yDA anDA a
reatmeeatmeng treang trea
ose 32ose 32d td
COPY ononhow thhow th
tabletabled
This do
cumen
t ……ca
nnot
s beebee
beab db dus
ed B pepe
writtewrittedodo
to ysyersoerso
supp
ortWeekeek
sis, this, th
anywith lwith
k 24k 2
marketi
ng20m0m
the Hyhe Hynly datnly da
Data uDatall
autho
rizati
ones for s fornd modnd mo
ent grnt gratmentmenmgmg
appli
catio
nWeekWeeng varig var
ts, MDs, MDose adse ad
and
ionone datae data
kk
any
n datn datex
tensio
ns or
varia
tions
rsttva
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 116 of 183
efficacy variables, ACR components, MDA, modified MDA, PK, AbAb, TEAEs, biochemistry and hematology laboratory results, and Hy’s Law. It depends on the variable how those additional time points will be presented in the interim analysis.
The efficacy analyses and the plasma concentration analyses for the period between Week 12 and Week 24 are only for those treatment groups that are not rerandomized. Two summary tables for each of the primary and secondary efficacy variables, the ACR components, and MDA and modified MDA will be presented. The first table will include data up to Week 12 by treatment group at Baseline. The second table will include data up to Week 24 by the following treatment groups: bimekizumab 160mg and bimekizumab 160mg with loading dose 320mg, and bimekizumab 320mg. The same approach will be done for all TEAEs tables and the Hy’s Law criteria tables. The summary table for plasma concentration will present only data up to Week 12 for the Baseline treatment groups Placebo and bimekizumab 16mg. Data up to Week 24 will be presented for bimekizumab 160mg, bimekizumab 160mg with loading dose 320mg, and bimekizumab 320mg. All Listings which include data up to Week 24 will use the same approach as for plasma concentration tables.
The safety analysis (TEAEs, biochemistry and hematology laboratory results, and Hy’s Law) will be handled as follows: Two summary tables for each of the safety analysis will be presented. The first table will include data up to Week 12 by treatment group at Baseline. The second table will include data up to Week 24 and will also present subjects under their Baseline treatment group. For the second table any event/finding that is unique to a treatment group should only be presented in the “All subjects” column and the treatment columns should remain blank. Following rules will be applied for the TEAEs: If the relative TEAE start date is less or equal to 84 (12*7) days, then the TEAE will be presented in the up to Week 12 table. If the relative start date is less or equal to 168 (24*7) days, then the TEAE will be presented in the up to Week 24 table.
The treatment group information will not be displayed for the interim listings.
…
For the interim analysis, the database will be locked and the treatment codes will be made available to UCB personnel with exception of operational staff working on the study. An interim report will be written. The Investigators and subjects will remain blind to the assigned bimekizumab dosing regimen until the subject completes the Dose-Blind Period at Week 48. Further details are available in the PA0008 blinding plan.
…
Section 6.2 Other Baseline characteristics
…
Time since first diagnosis of PsA will be calculated as:
(6)
If the date of diagnosis is partial, it should be imputed to the most recent feasible date (ie, last day of the month if only day is missing, or the last day of the year if day and month are missing).
REDACTEDn thn thED
rulesrulesCTE
al to 8al to 8DAC
elativelativED
up tup tR
COPYYandandPY
e ae any nyCO
hehe
This do
cumen
tTimeTimeca
nnot6.2 6.2 b
e OO
usedare areed
tong reng ra
supp
ort e dad
nel witnel wThe InThe I
regireg
any
atabatamark
etingto Wto W
ting
will nwill nark
autho
rizati
oneveeveon“All “All ats will will
ori84 (1284 (12thststarara
appli
catio
nboraoran ach of ach oati2 by tr by t
plic
will will alapentent
and
ratoato
anywill will e
xtens
ionsbles les s
t onlyonly. Data Data
ding ding
org dog doan
varia
tions
12 12
dosedos
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 117 of 183
…
A rheumatoid factor value < 15 IU/mL is defined as negative and a value ≥ 15 IU/mL is defined as positive. Anti-CCP antibodies negative is a value < 5 U/mL and a value ≥ 5 U/mL is defined as positive.
The corresponding listings will be presented for the RS.
Has been changed to
…
Time since first diagnosis of PsA will be calculated as:
(6)
If the date of diagnosis is partial, it should be imputed to the most recent feasible date (ie, last day of the month if only day is missing, or the last day of the year if day and month are missing).
The absolute values of the formula (6) will be used to display positive values.
…
A rheumatoid factor value < 15 IU/mL is defined as negative and a value ≥ 15 IU/mL is defined as positive. Anti-CCP antibodies negative is a value < 5 U/mL and a value ≥ 5 U/mL is defined as positive.
For hs-CRP following rules will be applied:
If Baseline hs-CRP values are missing, then take the Screening hs-CRP values.
If Screening hs-CRP values are missing, then take the median hs-CRP values.
The corresponding listings will be presented for the RS.
Section 8.4 Subgroup analysisSubgroup analyses will be performed on the primary and secondary efficacy variables. The following variables for subgroup analyses will be defined:
Age (<45 years, ≥45 to <65 years, ≥65 years)
Gender (male, female)
Geographic region (North America, Europe)
BASDAI (<4 [mild disease]; ≥4 to ≤7 [moderate disease]; >7 to ≤10 [severe disease])
Prior TNF inhibitor exposure (yes, no)
Treatment-emergent AbAb status (positive, negative)
Extent of psoriasis involvement (<3%, ≥3% to <10%, ≥10%)
hs-CRP level (<= upper limit normal, > upper limit normal)
Time since first diagnosis of PsA (< 2 years, ≥ 2 years)
REDACTED plied: plied:
ACare miare m
RE
COPY s negas neg
lue <lue <
This do
cumen
tPPcann
ot ra
BASDAASDA
PriPr
beaphicaphicus
ed s
male, fale, to, ≥4, ≥4su
pporte perfe perf
r subgrsubg
5
any isismark
eting
ississgare miare mke
e presepres
autho
rizati
on< 5 U< 5 U
singing
appli
catio
npositposiio
ative ative a/
andday aay a
t
any feasibfeasi
nd
exten
sions
or va
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 118 of 183
Body weight (<100 kg, ≥100 kg)
Subjects will be counted to have a positive treatment-emergent AbAb status in case the positivity occurred between the first dose and Visit 7 (Week 12).
The derivation of time since first diagnosis of PsA is described in Section 6.2.
All subgroup analyses are based on imputed data and will be summarized using descriptive statistics only.
Has been changed toSubgroup analyses will be performed on the primary and secondary efficacy variables. The following variables for subgroup analyses are defined in Section 4.8. will be defined:
Age (<45 years, ≥45 to <65 years, ≥65 years)
Gender (male, female)
Geographic region (North America, Europe)
BASDAI (<4 [mild disease]; ≥4 to ≤7 [moderate disease]; >7 to ≤10 [severe disease])
Prior TNF inhibitor exposure (yes, no)
Treatment emergent AbAb status (positive, negative)
Extent of psoriasis involvement (<3%, ≥3% to <10%, ≥10%)
hs CRP level (<= upper limit normal, > upper limit normal)
Time since first diagnosis of PsA (< 2 years, ≥ 2 years)
Body weight (<100 kg, ≥100 kg)
Subjects will be counted to have a positive treatment-emergent AbAb status in case the positivity occurred between the first dose and Visit 7 (Week 12).
The derivation of time since first diagnosis of PsA is described in Section 6.2.
All subgroup analyses are based on imputed data and will be summarized using descriptive statistics only.
Section 8.3 Analysis of other efficacy variables…
Time to onset of ACR20/50 response will be estimated and presented using the Kaplan-Meier product-limit method for each treatment. Time to a given response will be defined as the length in weeks from Baseline until the first date when the response is achieved. Subjects who discontinue study treatment prior to achieving a response will be censored at the date of study treatment discontinuation. Subjects who reach the Week 12 visit without achieving a response will be censored at the date of the Week 12 visit. Missing response data will be imputed with NRI.
…
Has been changed to
REDACTED % to% t
> uppe> uppeTE< 2 ye< 2 yeDA
))
COPYative)ativePYto <1o <C
This do
cumen
tdiscodiscotretre
cann
ot onsons
uctct--limlimweeksweek
be
nsetnsetus
edlysiysi to sis ois o
supp
ort fi
re basee basan
y irst drst d
marketi
ng
positiveositivnd Visd V
autho
rizati
on
10%, 0%, tioer limr limriz
ears, ≥ars, ≥ut
appli
catio
n to ≤to ≤n and
≤1≤1
any e
xtens
ionss. ThT
ed:ed:onor TheThe
varia
tions
there
of.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 119 of 183
…
Time to onset of ACR20/50 response will be estimated and presented using the Kaplan-Meier product-limit method for each treatment. Time to a given response will be defined as the length in weeks from Baseline until the first date when the response is achieved. Following derivation will be used: Days from Baseline until the first date divided by 7. There will be no rounding for the Kaplan Meier estimates. Subjects who discontinue study treatment prior to achieving a response will be censored at the date of study treatment discontinuation. Subjects who reach the Week 12 visit without achieving a response will be censored at the date of the Week 12 visit. Missing response data will be imputed with NRI.
…
Section 10.1 Extent of exposureThe duration of exposure (in days) will be calculated as:
(15)
28 days refer to one half-life of bimekizumab. If a change of treatment occurs prior to completion of the dosing interval, duration of exposure to study medication will be calculated as:
(16)
For subjects who have died the exposure will be as follows:
(17)
The sum over all duration of exposure (in years) will be calculated for the total subject-years of duration of exposure.
The duration of exposure will be summarized by treatment group using descriptive statistics. A by-subject listing of date of first and last dose and the duration of exposure will be performed.
Time at risk (in days) is defined as:
(18)
where 140 days refers to 5*half-life of bimekizumab.
The sum over all time at risk (in years) will be calculated for the total subject-years of time at risk.
Has been changed toThe duration of exposure (in days) will be calculated as:
(15)
REDACTED
re willre will
COPYure tore
This do
cumen
t skk
HaHaca
nnot
40 d0
suum om ok
be
daydayus
ed to
ys) i) supp
orwill bill b
e of fire of f
is deis d
any
bebe
marketi
ng
osure (isure (
autho
rizati
on
l be asbe a
appli
catio
n treatmetreatm
study mtudy
and a
ny ex
tensio
ns or
varia
tions
he he
sit. t.
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 120 of 183
28 days refer to one half-life of bimekizumab. If a change of treatment occurs prior to completion of the dosing interval, duration of exposure to study medication will be calculated as:
(16)
For subjects who have died the exposure will be as follows:
(17)
The sum over all duration of exposure (in years) will be calculated for the total subject-years of duration of exposure.
The duration of exposure will be summarized by treatment group using descriptive statistics. A by-subject listing of date of first and last dose and the duration of exposure will be performed.
Time at risk (in days) is defined as:
(18)
where 140 days refers to 5*half-life of bimekizumab.
The sum over all time at risk (in years) will be calculated for the total subject-years of time at risk.
The days on treatment and days on bimekizumab treatment will be calculated as follows:
(19)
(20)
Section 10.2.1 Exposure duration
The duration of each AE will be calculated as follows:
+ 1 (19)
For the calculation of duration for AEs with missing start dates, the dates will be imputed as described in Section 4.2.2.
The time to first dose for each AE will be calculated as follows for all TEAEs:
(20)
The time to most recent dose for each AE will be calculated as follows for all TEAEs:
REDACTED ekizumekizumCT
R
COPY culateculat
This do
cumen
t TheThe
cann
ot he calce calccribecribe
beus
edeachachto
os
h Ah Asu
pport
rure dure d
any y
amark
eketin
gtin
gke
autho
rizati
on ed
mab tmab tor
appli
catio
n
for tfor t
and a
ny wiw ex
tensio
nsctct--yey
iptive ptivewill bill
or eava
riatio
ns
7)7)
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 121 of 183
(21)
Time to first dose and time to most recent dose will not be calculated for pretreatment AEs.
For the calculation of time to first bimekizumab and time to most recent bimekizumab dose the same formulas as above (20 and 21) will be used but the date of first bimekizumab dose will be used instead of the date of first dose. The time to first dose and time to first bimekizumab dose differs only for those subjects who receive placebo at Baseline.
The days on treatment and days on bimekizumab treatment will be calculated as follows:
(22)
(23)
Has been changed to
The duration of each AE will be calculated as follows:
+ 1 (1921)
For the calculation of duration for AEs with missing start dates, the dates will be imputed as described in Section 4.2.2.
The time to first dose for each AE will be calculated as follows for all TEAEs:
(220)
The time to most recent dose for each AE will be calculated as follows for all TEAEs:
(231)
Time to first dose and time to most recent dose will not be calculated for pretreatment AEs.
For the calculation of time to first bimekizumab and time to most recent bimekizumab dose the same formulas as above (20 and 21) will be used but the date of first bimekizumab dose will be used instead of the date of first dose. The time to first dose and time to first bimekizumab dose differs only for those subjects who receive placebo at Baseline.
The days on treatment and days on bimekizumab treatment will be calculated as follows:
REDACTED follfoll
AEs wAEs w
COPY
This do
cumen
t imm
ForFor
anno
t
me toe to
be us
ed to
ecencensupp
ort
nt dnt
any m
arketi
ngwith mwith m
E will wil
autho
rizati
on
ws:ws:
appli
catio
n and
any e
xtens
ions llowsllowsor
varia
tions
be be
dose ose
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 122 of 183
(22)
(23)
Section 10.2.2 Exposure adjusted incidence rate (EAIR) and exposure adjusted event rate (EAER)The EAIR is defined as the number of subjects (n) with a specific AE adjusted for the exposure and will be scaled to 100 patient-years:
(24)
where is a subject’s exposure time in years up to the first occurrence of the AE of interest.
If a subject has multiple events, the time of exposure is calculated to the first occurrence of the AE of interest. If a subject has no events, the total time at risk is used.
Exact Poisson 95% CIs for incidence rates are calculated using the relationship between the Poisson and the Chi-square distribution (Ulm, 1990; Fay and Feuer, 1997):
(25)
(26)
where n is the number of subjects with a specific AE for the incidence rate of interest and is the basis for the number of the degrees of freedom for the chi-square quantile for the upper tail probability χ2.
The EAER will be the number of AEs including repeat occurrences in individual subjects divided by the total time at risk scaled to 100 patient-years and calculated using:
(27)
where NAE is the total number of AEs and is a subject’s total time at risk in years.
No CI will be computed for EAER.
Has been changed to
The EAIR is defined as the number of subjects (n) with a specific AE adjusted for the exposure and will be scaled to 100 patient-years:
REDACTEDc
m, 199m, 199COPYme ame a
alculalcula
This do
cumen
t whewheca
nnot
be us
edbe thbe the totatota
to hhsu
pportsubjecubje
f the dethe d
any
t
marketi
ng au
thoriz
ationated uated u
90; Fay0; Fay
appli
catio
noccurroccur
ulated latedt risk isrisk i
and a
ny ex
tensio
ns
r the ethe e
or nt rt va
riatio
ns
ratrat
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 123 of 183
(24)
where is a subject’s exposure time in years up to the first occurrence of the AE of interest (equation [21] in years) at the level of coding evaluated.
If a subject has multiple events at the level of coding evaluated, the time of exposure is calculated to the first occurrence of the AE of interest. If a subject has no events, the total time at risk (15) is used.
Exact Poisson 95% CIs for incidence rates are calculated using the relationship between the Poisson and the Chi-square distribution (Ulm, 1990; Fay and Feuer, 1997):
(25)
(26)
where n is the number of subjects with a specific AE for the incidence rate of interest and is the basis for the number of the degrees of freedom for the chi-square quantile for the upper tail probability χ2.
The EAER will be the number of AEs including repeat occurrences in individual subjects divided by the total time at risk scaled to 100 patient-years and calculated using:
(27)
where NAE is the total number of AEs and is a subject’s total time at risk in years.
No CI will be computed for EAER.
Section 10.2.3 AE of Special Monitoring…
Inflammatory bowel disease
These events will be presented in a stand-alone table which includes EAIR and EAER. The table will include all TEAEs which code into the HLGT of “Colitis excl infective”.
…
Has been changed to
…
Inflammatory bowel diseaseThese events will be presented in a stand-alone table which includes EAIR and EAER. The table will include all TEAEs which code into the HLGT of “Colitis excl infective”.
REDACTED uding uding
o 100 o 100
COPY for forr the cr the
This do
cumen
t HasHas
cann
otludeudebents wnts w
used
bowebowe
to su
pport
or EAr EA
SpeciaSpeci
anyf Af A
AEAE
marketi
ng
AEsAE
autho
rizati
on chihi
g repearepeapatienpatien
appli
catio
n
he incie inci--squsq
and a
ny ex
tensio
nseen ten or thth
varia
tions
me at me at
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 124 of 183
13.3 Amendment 313.3.1 Rationale for the amendmentThe SAP has been amended to:
Add Section with changes from interim analysis to SAP-defined analyses
Change definition of AE of special monitoring for final analysis
Add rule for value below limit of quantification for DAS28(CRP)
Add Section for CRP and hs-CRP
Update AE section to apply the interim rules to the final analysis
Update markedly abnormal laboratory units and add the RCTC reference
Update AbAb Section to clarify the overall AbAb status and the first occurrence of AbAb positivity
Change subgroup analysis sections to change the definitions of treatment-emergent AbAb status
Remove rounding information for BMI
Update prohibited medication section to remove the rule that efficacy results should be set to missing after receiving prohibited medication
Update pharmacokinetics section to be aligned with TFLs shells
13.3.2 Modification and changesList of abbreviation
Has been changed to
…
CDF cumulative distribution function
CI confidence interval
…
RBC red blood cell
RNA ribonucleic acid
…
…
CDF cumulative distribution function
CDISC Clinical Data Interchange Standards Consortium
CI confidence interval
REDACTED ionon
e alignalign
hangehange
COPY ve theve th
This do
cumen
t ……
CC
cann
ot been beenbe
used
to su
pport
any cm
arketi
ng
cu
autho
rizati
one ruru
ned wied wi
ess
appli
catio
n f tr
ule tule
and
reatmeatan
yccurccu exten
sions
urr
or va
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 125 of 183
Section 4.3.1 has been added
Section 4.3.1 Changes from interim analysis to SAP-defined analyses
For TEAE two summary tables were presented as described in previous section with one modification: the first table was displaying all TEAE during the Double-Blind Period instead of limiting it TEAE with a relative TEAE start date less equal 84 days.
The definitions of AE of special monitoring were updated with the third SAP Amendment after the interim analysis. The interim analysis used the previous definition which is displayed in Section 4.3.1.2. The interim analysis only displayed the overall number of AE of special monitoring by treatment group.
The definition of time since first diagnosis of PsA was updated with the third SAP Amendment after the interim analysis. In the previous definitions the date of informed consent was subtracted from the date of diagnosis which results in negative values. For the interim analysis another calculations were used:
Section 4.3.2 has been added
Section 4.3.2 Search and selection criteria for AE of special monitoring for interim analysis
Following AEs are defined as TEAEs of special monitoring:
Fungal infectious disorder
Opportunistic infection (including TB)
Malignant or unspecified tumor
Malignant tumor
Major cardiovascular event
Haematopoietic cytopenias
Neuropsychiatric events
Inflammatory bowel disease
Hypersensitivity and anaphylactic reaction
Hepatic events
…
RBC Red blood cell
RCTC Rheumatology Common Toxicity CriteriaRNA Ribonucleic acid
…
REDACTED gngDe usede used
TE
R
COPYwaswaPYreviorevi
COnosisosisC
This do
cumen
t Hca
nnotlignaigna
noMajoMajoca
be antnt
beus
edstic inic iednt ort or
to uso
nsu
pport
tned ased app
os disodisos
any ectionectioamark
eting
ing
autho
rizati
on ous
ns whicwhicati
od:d:
riz
appli
catio
n ed thed ton
updatepdatpps defdea
and
ththds defidefi
anthethe
any y
hird Sird Saex
tensio
ns
ion won wnslind Plind
xte4 daydaye
or va
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 126 of 183
Fungal Infectious disorder
All TEAEs which code into the High Level Group Term (HLGT) of “Fungal infectious disorders” are classified as fungal infectious disorder.
Opportunistic infection
All TEAEs identified using UCB-defined search criteria as described in Section 12.13.
Malignant or unspecified tumor
The search criteria is based on the Standardized MedDRA Query (SMQ) = “Malignant or unspecified tumours (SMQ)”.
Malignant tumor
The search criteria is based on the SMQ=”Malignant tumours (SMQ)”.
Major cardiovascular events
The major cardiovascular events are identified using the following UCB-defined search criteria:
All serious TEAEs which code to a PT included in the Scope=Broad and/or Scope=Narrow groups of the following SMQs:
Haemorrhagic central nervous system vascular conditions (SMQ)
Ischaemic central nervous system vascular conditions (SMQ)
All serious TEAEs which code to a PT included in the HLT “Ischaemic coronary artery disorders” except events coding to PT “Chest Pain” or “Chest discomfort”
All serious TEAEs which code to a PT included in any of the following HLTs: “Heart Failures NEC”, “Left Ventricular Failures”, or “Right Ventricular Failures” and which also code to the SOC of “Cardiac Disorders” as Primary SOC.
Haematopoietic cytopenias
The search criteria is based on the SMQ = “Haematopoietic cytopenias”. The SMQ search should include all serious TEAEs which code to a PT included in the Scope=Broad and/or Scope=Narrow groups within the SMQ.
Neuropsychiatric events
The search criteria is based on the SMQ = “Depression and suicide/self-injury (SMQ)”. The SMQ search will include all TEAEs which code to a PT included in the Scope=Broad and/or Scope=Narrow groups within the SMQ.
Inflammatory bowel disease
All TEAEs which code into the HLT of “Colitis (excl infective)” are identified as inflammatory bowel disease.
Hypersensitivity reactions and anaphylactic reactions
Hypersensitivity reactions and anaphylactic reactions will be identified as follow:
REDACTEDvascuvascuTE
a PT inPT inDAC
to PTto PTRED
toto
COPY vascuvascuCO
This do
cumen
tInflanflaen
AA
cann
otMQ sMQ sno/or Scor Scan
beh crih crbeus
ed
atric etric eus
et
tooupouptosu
pport
tased oased pprious Trious sups w
any m
arketi
ng
a PT a PT etular Faular Farrdiac rdiacm
autho
rizati
onular lar
onlar colar coiza
ncludncludth“Ch“Cha
appli
catio
n Scope=Scopca
coco
andng UCg UC
anan
y CB
exten
sions
or ana
orva
riatio
ns
ant ont o
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 127 of 183
c) Hypersensitivity reactions: All TEAEs that either emerged on the same day as when a study medication injection reaction was received, or that emerged one day after a study medication injection was received, which code to a PT which contains the term “hypersensitivity” will be considered to be a hypersensitivity reaction and included in the summary table as such.
d) Anaphylactic reactions: An algorithmic approach will be used to identify TEAEs that are considered to be anaphylactic reactions. PTs are separated into the 4 distinct categories (A, B, C, D) prior to the algorithmic approach being applied.
All TEAEs that either emerged on the same day as when a study medication injection reaction was received, or that emerged one day after a study medication injection was received, and which fulfill any of the 3 criteria described in Section 12.14 will be included in the summary table.
Hepatic events
The search criteria is based on all TEAEs in the SMQ “Drug related hepatic disorders -comprehensive search (SMQ)”. Note that the following two sub-SMQs are to be excluded: “Liver neoplasms, benign (incl cysts and polyps)” and “Liver neoplasms, malignant and unspecified (SMQ)”.
The SMQ search should include all TEAEs (regardless of whether they have been judged as related to study medication or not) which code to a PT included in the Scope=Narrow group within each SMQ.
Section 4.8 Examination of subgroups
…
Subjects will be counted to have a positive treatment-emergent AbAb status in case the positivity occurred between the first dose and Visit 7 (Week 12).
Has been changed to
…
Subjects will be counted to have a positive treatment-emergent AbAb status in case the firstAbAb positivity occurred up to between the first dose and Visit 7 (Week 12). The definition of first AbAb occurrence is described in Section 9.2.
Section 6.1 Demographics
…
BMI in kg/m2 is calculated based on the height (in m) and the weight (in kg) using the formula:
For adults:
(5)
Even if they are available in the database, these variables are calculated during analysis (if applicable). These calculated values are used in the statistical analysis since they are considered more accurate. BMI is rounded to 1 decimal.
REDACTED
i
COPY ardlesardleOP
ode tode tC
This do
cumen
tFor aor acan
not
I in kgin kg
dd
be D us
edrrereed
emoemo
to curuencenc
supp
ort
ed to hd to rred rred
any m
arketi
ng
sitive trtive td Visit Visit
autho
rizati
onto a Pto a Ptio
appli
catio
n b-Sner neopr neoati
o
s of ws of wapPT
andated hted ndSMSM
any
heh
exten
sions
ctisioion wn w
ionll be il be ins
or on on ova
riatio
ns
d.va
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 128 of 183
…
Has been changed to…
BMI in kg/m2 is calculated based on the height (in m) and the weight (in kg) using the formula:
For adults:
(5)
Even if they are available in the database, these variables are calculated during analysis (if applicable). These calculated values are used in the statistical analysis since they are considered more accurate. BMI is rounded to 1 decimal.
…
Section 6.2 Other Baseline characteristics
….
Time since first diagnosis of PsA will be calculated as:
(6)
If the date of diagnosis is partial, it should be imputed to the most recent feasible date (ie, last day of the month if only day is missing, or the last day of the year if day and month are missing).
The absolute values of the formula (6) will be used to display positive values.
…
Has been changed to
….
Time since first diagnosis of PsA will be calculated as:
(6)
If the date of diagnosis is partial, it should be imputed to the most recent feasible date (ie, last day of the month if only day is missing, or the last day of the year if day and month are missing).
The absolute values of the formula (6) will be used to display positive values.
…
Section 6.5 Prohibited medication and rescue medication
REDACTEDuld be uld be
ng, orng, or
))
COPY s:
This do
cumen
t ay y
TheThemeca
nnot e date date
of of tt
be us
edto
supp
ort
nosis oosis o
any m
arketi
ng r tht
) will bwill bau
thoriz
ation
impumpuhe lahe la
appli
catio
n and
any e
xtens
ions
nalysislysisey arey are
or (5(5 va
riatio
ns
5)5)
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 129 of 183
Prohibited medications are defined in the Protocol (Section 7.8.2). All efficacy data collected at assessments after the use of prohibited medications should be treated as missing. Prohibited medication use and date of first usage should be determined and documented during the data evaluation meeting (DEM) prior to the database lock. The number and percentage of subjects who used prohibited medication will be summarized and also be listed.
…
Has been changed toProhibited medications are defined in the Protocol (Section 7.8.2). All efficacy data collected at assessments after the use of prohibited medications should be treated as missing. Prohibited medication use and date of first usage should be determined and documented during the data evaluation meeting (DEM) prior to the database lock. The number and percentage of subjects who used prohibited medication will be summarized and also be listed.
…
Section 8.3.2 Disease Activity Score-28 joint count C-reactive protein (DAS28[CRP])The components for DAS28(CRP) include the TJC and SJC based on 28 joints, hs-CRP (mg/L), and the PGADA (mm). DAS28(CRP) is calculated as follows:
(12)
Has been changed to
The components for DAS28(CRP) include the TJC and SJC based on 28 joints, hs-CRP (mg/L), and the PGADA (mm). DAS28(CRP) is calculated as follows:
(12)
The hs-CRP values below the limit of quantification should be set to half the limit of quantification for the calculations. The limit of quantification for hs-CRP is 0.16 mg/L.
Section 8.3.11 has been added
Section 8.3.11 C-reactive protein (CRP) and high-sensitivity C-reactive protein (CRP)
Blood will be collected for measurement of CRP and hs-CRP. CRP and hs-CRP are indicators for inflammation and are measured in the blood. Hs-CRP is a component of composite efficacy variables, and in addition is also summarized and analyzed as individual variables. For hs-CRP the summary statistics should contain n, geometric mean, geometric CV, median, first and third quartile (Q1 and Q3), minimum, and maximum, where the geometric CV (%) is calculated using:
(15)
with SDln – the standard deviation of the ln-transformed hs-CRP values.
REDACTED
ncludencludeP) isP) is
COPY PY
This do
cumen
tariaariantCVCVm
cann
otl ttors fors fnompositposit
caabab
bebe cbe bus
ed ss
1 CC--rrustobeebee
supp
ort
ow thew thepoe calcucalcsu
p
any m
arketi
ng e th
s calcucalcu
ark
autho
rizati
on
he The T
appli
catio
ned on ed ons:s:
and otein otein
any e
xtens
ionshibithibis
ng theg theionage of ge ofns
orecteectritediteo
varia
tions
tedtedv
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 130 of 183
For descriptive statistics, the observed values and ratio to Baseline values will be displayed.
hs-CRP and CRP values below the limit of quantification should be set to half the limit of quantification for the calculations. The limit of quantification for hs-CRP is 0.16 mg/L and 0.4 mg/dL = 4.00 mg/L for CRP.
Section 9.2 Pharmacodynamics and Immunogenicity…
In addition, overall AbAb status should be determined:
Subject AbAb positivity is defined as having with AbAb+ at any time in the treatment period. This does not include Baseline/pre-treatment
Subject AbAb negativity is defined with AbAb- at any time in the treatment period. This does also include AbAb+ at Baseline/pre-treatment
Treatment-emergent AbAb positivity is defined when a subject is AbAb+ for the first time during treatment period excluding Baseline/pre-treatment. If there is AbAb+ at Baseline/pre-treatment and there is a pre-defined fold increase in titre at least at one visit during the treatment period, then the subject has also a treatment-emergent AbAb positivity status.
Note: The fold increase from Baseline required will be defined with the development of the assay prior to database lock.
…
Has been changed to…
In addition, overall AbAb status should be determined:
Subject AbAb positivity is defined as having with AbAb+ at any time in the treatment period. This does not include Baseline/pre-treatment
Subject AbAb negativity is defined with AbAb- at any time in the treatment period. This does also include AbAb+ at Baseline/pre-treatment
Treatment emergent AbAb positivity is defined when a subject is AbAb+ for the first time during treatment period excluding Baseline/pre treatment. If there is AbAb+ at Baseline/pre-treatment and there is a pre defined 4-fold increase in titre at least at one visit during the treatment period, then the subject has also an treatment emergent overall AbAb positivity status.
Note: The fold increase from Baseline required will be defined with the development of the assay prior to database lock.
The visit of the first occurrence of AbAb positivity is defined as the visit when a subject is AbAb+ for the first time during treatment period excluding Baseline/pre-treatment. If there is AbAb+ at Baseline/pre-treatment and there is a 4-fold increase in titre at least at one visit during the treatment period, then the subject’s first occurrence visit is the Baseline Visit.
…
REDACTED
uldul
COPYbe dbe d
This do
cumen
tassayssaytThThm
cann
ot s..
e: Thee: Thanay pyc
beent pent pus
edmententdand thnd th
perper
pto rgeng
nt pet ptsu
pport
ty is y isAb+ aAb+
ent Ant As
anye Be B
s des d
marketi
ng
d be debe de
fined asned aBaseBa
autho
rizati
on ap
plica
tionast atst a
ergent Agent
fined wined
and
Abbre is Ae is Aatt
any
Ab+ Ab+ ex
tensio
ns menen
nt perit per
or ntntva
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 131 of 183
Section 8.4 Subgroup analysis
…
Subjects will be counted to have a positive treatment-emergent AbAb status in case the positivity occurred between the first dose and Visit 7 (Week 12).
…
Has been changed to
…
Subjects will be counted to have a positive treatment-emergent AbAb status in case the first AbAb positivity occurred up to between the first dose and Visit 7 (Week 12). The definition of first AbAb occurrence is described in Section 9.2.
…
Section 9.1Pharmacokinetics
…
In addition geometric mean bimekizumab plasma concentration time curves will be plotted by treatment group, and by cumulative antibody status (ie, positive, negative) for subjects randomized to bimekizumab.
…
Has been changed to
…
In addition geometric mean bimekizumab plasma concentration time curves will be plotted by treatment group, and by cumulative antibody status (ie, positive, negative) for subjects randomized to bimekizumab.
…
Section 10.1 Extent of exposure
The duration of exposure (in days) will be calculated as:
(15)
28 days refer to one half-life of bimekizumab. If a change of treatment occurs prior to completion of the dosing interval, duration of exposure to study medication will be calculated as:
(16)
For subjects who have died the exposure will be as follows:
(17)
REDACTED
umaum
COPY e, pe,
This do
cumen
t can
nots referefe
pletiopletio
be us
edto
supp
ort
xposuxposu
sure (iure (
any m
arketi
ng
mab plaab plaantibodntiboke
autho
rizati
on ap
plica
tion
on timen timsitive, ntive,
and a
ny ex
tensio
nsthehe fifnse defidefins
or iva
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 132 of 183
The sum over all duration of exposure (in years) will be calculated for the total subject-years of duration of exposure.
The duration of exposure will be summarized by treatment group using descriptive statistics. A by-subject listing of date of first and last dose and the duration of exposure will be performed.
Time at risk (in days) is defined as:
(18)
where 140 days refers to 5*half-life of bimekizumab.
The sum over all time at risk (in years) will be calculated for the total subject-years of time at risk.
The days on treatment and days on bimekizumab treatment will be calculated as follows:
(19)
(20)
Has been changed toThe duration of exposure (in days) will be calculated as:
(156)
28 days refer to one half-life of bimekizumab. If a change of treatment occurs prior to completion of the dosing interval, duration of exposure to study medication will be calculated as:
(167)
For subjects who have died the exposure will be as follows:
(178)
The sum over all duration of exposure (in years) will be calculated for the total subject-years of duration of exposure.
The duration of exposure will be summarized by treatment group using descriptive statistics. A by-subject listing of date of first and last dose and the duration of exposure will be performed.
Time at risk (in days) is defined as:
(189)
REDACTED
l be cal be c
COPY
This do
cumen
tduratdurat
TT
cann
ote sumsum
be us
ed
who hwho hto
supp
ortntervervan
yf bimf bimrval,val
marketi
ng au
thoriz
ation
alculaalcula
appli
catio
n and a
ny ex
tensio
ns
s of timof tim
d as fas f
or va
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 133 of 183
where 140 days refers to 5*half-life of bimekizumab.
The sum over all time at risk (in years) will be calculated for the total subject-years of time at risk.
The days on treatment and days on bimekizumab treatment will be calculated as follows:
(1920)
(201)
Section 10.2 Adverse events
…
The following code lists will be used for AE recording:
Pattern of event: intermittent or continuous
Intensity of event: mild, moderate or severe
Relationship: related or not related
Action taken with IMP: dose not changed, dose reduced, dose increased, drug temporarily, interrupted, drug permanently withdrawn or not applicable
Outcome: resolving, not resolved, resolved, resolved with sequelae, worsened, fatal or unknown
For all TEAEs the following variables will be calculated:
Duration
Time since first dose
ADRs are defined as a response to a drug which is noxious and unintended and which occurs at doses normally used in man for prophylaxis, diagnosis, or therapy of disease or for modification of physiological function. Any AE that is considered “Related” to study treatment will be classed as an ADR.
…
Has been changed to…
The following code lists will be used for AE recording:
Pattern of event: intermittent or continuous
Intensity of event: mild, moderate or severe
Relationship: related or not related
REDACTED ged, dged, d
drawndrawn
d, resd, res
COPY
This do
cumen
t ……
T
cann
ot s bbeeneen
be us
ed seded
al funl funto
s ad ind in
supp
ort
a respa resp
any m
arketi
ngsolvesolve
ables wbles
autho
rizati
on
dose reose ren oror nono
d
appli
catio
n and
any e
xtens
ions o
r vari
ation
s ))
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 134 of 183
Action taken with IMP: dose not changed, dose reduced, dose increased, drug temporarily, interrupted, drug permanently withdrawn or not applicable
Outcome: resolving, not resolved, resolved, resolved with sequelae, worsened, fatal or unknown
For all TEAEs the following variables will be calculated:
Duration
Time since first dose
ADRs are defined as a response to a drug which is noxious and unintended and which occurs at doses normally used in man for prophylaxis, diagnosis, or therapy of disease or for modification of physiological function. Any AE that is considered “Related” to study treatment will be classed as an ADR.
If an AE occurs on the same day as the treatment switch (Visit 7, Week 12) then the AE will be allocated to the Double-Blind treatment. An exception from this general rule is made for ADRs, and AE of hypersensitivity reactions and anaphylactic reactions. If those AEs occur on the same days as the treatment switch, the AEs will be allocated to the Dose-Blind treatment.
…
Section 10.2.1 Exposure duration
The duration of each AE will be calculated as follows:
+ 1 (19)
For the calculation of duration for AEs with missing start dates, the dates will be imputed as described in Section 4.2.2.
The time to first dose for each AE will be calculated as follows for all TEAEs:
(20)
The time to most recent dose for each AE will be calculated as follows for all TEAEs:
(21)
…
Has been changed to
The duration of each AE will be calculated as follows:
+ 1 (1922)
REDACTED as foas fo
COPY
This do
cumen
t ……
HH
cann
ot be
used
t recrecto
cecesu
pport
anyAE wAE mark
eting
Es withEs wit
autho
rizati
on
llowslows
appli
catio
n reaeanllocateocatati
andhis ges gendactiacti
any ek 12ek 1anen
exten
sions
h ococr modmod
ent wient wi
or ccuccuva
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 135 of 183
For the calculation of duration for AEs with missing start dates, the dates will be imputed as described in Section 4.2.2.
The time to first dose for each AE will be calculated as follows for all TEAEs:
(2023)
The time to most recent dose for each AE will be calculated as follows for all TEAEs:
(2124)
Section 10.2.2 Exposure adjusted incidence rate (EAIR) and exposure adjusted event rate (EAER)The EAIR is defined as the number of subjects (n) with a specific AE adjusted for the exposure and will be scaled to 100 patient-years:
(24)
where is a subject’s exposure time in years up to the first occurrence of the AE of interest (equation [21] in years) at the level of coding evaluated.
If a subject has multiple events at the level of coding evaluated, the time of exposure is calculated to the first occurrence of the AE of interest. If a subject has no events, the total time at risk (15) is used.
Exact Poisson 95% CIs for incidence rates are calculated using the relationship between the Poisson and the Chi-square distribution (Ulm, 1990; Fay and Feuer, 1997):
(25)
(26)
where n is the number of subjects with a specific AE for the incidence rate of interest and is the basis for the number of the degrees of freedom for the chi-square quantile for the upper tail probability χ2.
The EAER will be the number of AEs including repeat occurrences in individual subjects divided by the total time at risk scaled to 100 patient-years and calculated using:
(27)
REDACTEDears uears ing evng ev
level olevel othe Athe A
COPY Y
This do
cumen
tprobrob
ThT
cann
ot n is thn is this for s for
abb
be us
ed to
supp
ort an
ytributribumark
etingAE ofAE of
nce ratnce rautit
autho
rizati
on
up to thp to thvaluataluat
of codf codff
appli
catio
n and
E adjE adj
any adjad ex
tensio
ns
((22
djusju
or va
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 136 of 183
where NAE is the total number of AEs and is a subject’s total time at risk in years.
No CI will be computed for EAER.
Has been changed toThe EAIR is defined as the number of subjects (n) with a specific AE adjusted for the exposure and will be scaled to 100 patient-years:
(245)
where is a subject’s exposure time in years up to the first occurrence of the AE of interest (equation [21] in years) at the level of coding evaluated.
If a subject has multiple events at the level of coding evaluated, the time of exposure is calculated to the first occurrence of the AE of interest. If a subject has no events, the total time at risk (15) is used.
Exact Poisson 95% CIs for incidence rates are calculated using the relationship between the Poisson and the Chi-square distribution (Ulm, 1990; Fay and Feuer, 1997):
(256)
(267)
where n is the number of subjects with a specific AE for the incidence rate of interest and is the basis for the number of the degrees of freedom for the chi-square quantile for the upper tail probability χ2.
The EAER will be the number of AEs including repeat occurrences in individual subjects divided by the total time at risk scaled to 100 patient-years and calculated using:
(278)
where NAE is the total number of AEs and is a subject’s total time at risk in years.
No CI will be computed for EAER.
Section 10.2.3 AE of Special Interest and AE of Special Monitoring
AE of special interest is any AE which meets the Hy’s Law criteria, defined as ≥3x upper limit of normal (ULN) alanine aminotransferase (ALT) or aspartate aminotransferase (AST) with coexisting ≥2xULN total bilirubin in the absence of ≥2xULN alkaline phosphatase (ALP), with no alternative explanation for the biochemical abnormality.
AEs of special monitoring for this study include: serious infections (including opportunistic infections and tuberculosis [TB]), cytopenias, hypersensitivities, suicide ideation or behavior
REDACTED
AC
thth
COPY OCO
This do
cumen
t AE AE o
cann
otwill bwill b
tion tion
beis ths thus
ed
hehe
to su
pport
berat riskat risk
any
r of Ar of Amark
eting
h a speca spec
of freeof fre
autho
rizati
on
th
appli
catio
n he rele reFeuer, 1euer,
and
elala
anyevenevenex
tensio
ns
AE of AE of
xposxpo
or va
riatio
ns
455))vav
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 137 of 183
(assessed using the eC-SSRS), depression and anxiety (assessed using the HADS), major cardiovascular events and liver function test changes/enzyme elevations (ALT, AST, and bilirubin), malignancies, and inflammatory bowel diseases.
The incidence of TEAEs of special monitoring will be summarized by MedDRA system organ class, high level term, and PT. EAIR with associated 95% CI, and the EAER will be included in the summary tables for following TEAEs of special monitoring:
Fungal infectious disorder
Opportunistic infection (including TB)
Malignant or unspecified tumor
Malignant tumor
Major cardiovascular event
Haematopoietic cytopenias
Neuropsychiatric events
Inflammatory bowel disease
Hypersensitivity and anaphylactic reaction
Hepatic events
Fungal infectious disorderFungal infections will be summarized in a stand-alone table which presents EAIR and EAER. The table will include all TEAEs which code into the High Level Group Term (HLGT) of “Fungal infectious disorders”.
Opportunistic infection
Opportunistic infections (including TB) will be summarized in a stand-alone table which presents EAIR and EAER. The table will include all TEAEs identified using UCB-defined search criteria as described in Section 12.13.
Malignant or unspecified tumorThese events will be presented in one stand-alone table which will include EAIR and EAER. The table will be based on the criteria Standardized MedDRA Query (SMQ) = “Malignant or unspecified tumours (SMQ)”.
The output table will include 2 different overall incidence rows:
4. The first overall incidence row will summarize “Any malignancies (including unspecified)” and this row will summarize the incidence of all AEs flagged for inclusion in the table, regardless of the High Level Term (HLT) it codes to.
5. The second overall incidence row will summarize “Any malignancy (including unspecified, excluding non-melanomic skin cancers)” and this row will summarize the incidence of AEs flagged for inclusion in the table, excluding those which code to an HLT of “skin neoplasms malignant and unspecified (excl melanoma)”.
n a stan a staich coich co
COPY
This do
cumen
taacann
otutput ttput t
The The
bed tumd tumus
edll bebe
ased asedmo
to e pe psu
pport
ThThed in d in SS
cified ified
anyding ding
he tabe tabmark
eting
odede
g Tg T
autho
rizati
on
andnd--alale ine in
appli
catio
n and
any e
xtens
ions o
r vari
ation
s the
reof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 138 of 183
Malignant tumorThese events will be presented in one stand-alone table which will include EAIR and EAER. The table will be based on the criteria SMQ=”Malignant tumours (SMQ)”.
Note that the events included in the “Malignant tumours” table will be a subset of the events included in the “Malignant or unspecified tumours” table. The SMQ search should include all TEAEs which code to a PT included in the Scope=Narrow group within each SMQ.
The output table will include 2 different overall incidence rows:
6. The first overall incidence row will summarize “Any malignancies” and this row will summarize the incidence of all AEs flagged for inclusion in the table, regardless of the HLT it codes to.
7. The second overall incidence row will summarize “Any malignancy (excluding non-melanomic skin cancers)” and this row will summarize the incidence of AEs flagged for inclusion in the table, excluding those which code to an HLT of “skin neoplasms malignant and unspecified (excl melanoma)”.
Major cardiovascular events
The major cardiovascular events are identified using the following UCB-defined search criteria:
All serious TEAEs which code to a PT included in the Scope=Broad and/or Scope=Narrow groups of the following SMQs:
Haemorrhagic central nervous system vascular conditions (SMQ)
Ischaemic central nervous system vascular conditions (SMQ)
All serious TEAEs which code to a PT included in the HLT “Ischaemic coronary artery disorders” except events coding to PT “Chest Pain” or “Chest discomfort”
All serious TEAEs which code to a PT of “Cardiac failure congestive”
Haematopoietic cytopeniasThese events will be presented in a stand-alone table that is based on the SMQ = “Haematopoietic cytopenias”. The SMQ search should include all serious TEAEs which code to a PT included in the Scope=Broad and/or Scope=Narrow groups within the SMQ.
Neuropsychiatric eventsThese events will be presented in a stand-alone table including EAIR and EAER. The table will be based on the SMQ = “Depression and suicide/self-injury (SMQ)”. The SMQ search will include all TEAEs which code to a PT included in the Scope=Broad and/or Scope=Narrow groups within the SMQ.
Inflammatory bowel diseaseThese events will be presented in a stand-alone table which includes EAIR and EAER. The table will include all TEAEs which code into the HLT of “Colitis excl infective”.
Hypersensitivity reactions and anaphylactic reactions
Anaphylactic reactions will be summarized together in a stand-alone table.
em vasem va
em vaem va
o a Po a P
COPY he
ed in td in
This do
cumen
tgrouprou
InIn
cann
oted ond onude alde aupsups
bents wnts wus
ed
atric etric etoe Sce Scs
uppo
rt esentedsentedopeniaspenias
cop
any m
arketi
ngPT incPT into PT “o PT
e to ae to
autho
rizati
on the
ascularscular
ascularscula
appli
catio
n
lowingowin
e Scoe Sc
and
kinin anyf AEf A
n neon neoex
tensio
nsof thof th
ludingudinAEAE
or ll l varia
tions
there
of.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 139 of 183
The first row within the body of the table will be labeled “Any hypersensitivity/anaphylactic reaction” and will represent the overall incidence of subjects who reported at least one hypersensitivity reaction or at least one anaphylactic reaction.
The second row within the body of the table will be labeled “Any hypersensitivity reaction” and will represent the overall incidence of subjects who reported at least one hypersensitivity reaction.
The third row within the body of the table will be labeled “Any anaphylactic reaction” and will represent the overall incidence of subjects who reported at least one anaphylactic reaction.
Following these three overall incidence rows, all TEAEs that have been identified as either a hypersensitivity reaction or an anaphylactic reaction will be summarized (together – not broken out by type) by SOC, HLT and PT.
Hypersensitivity reactions and anaphylactic reactions will be identified as follow:
e) Hypersensitivity reactions: All TEAEs that either emerged on the same day as when a study medication injection reaction was received, or that emerged one day after a study medication injection was received, which code to a PT which contains the term “hypersensitivity” will be considered to be a hypersensitivity reaction and included in the summary table as such.
f) Anaphylactic reactions: An algorithmic approach will be used to identify TEAEs that are considered to be anaphylactic reactions. PTs are separated into the 4 distinct categories (A, B, C, D) prior to the algorithmic approach being applied.
All TEAEs that either emerged on the same day as when a study medication injection reaction was received, or that emerged one day after a study medication injection was received, and which fulfill any of the 3 criteria described in Section 12.14 will be included in the summary table.
Hepatic eventsAlthough not officially considered to be AEs of special monitoring but hepatic events are nonetheless considered to be interesting enough to be summarized in stand-alone tables.
Hepatic events will be summarized in a stand-alone table that includes all TEAEs in the SMQ “Drug related hepatic disorders - comprehensive search (SMQ)”. Note that the following two sub-SMQs are to be excluded: “Liver neoplasms, benign (incl cysts and polyps)” and “Liver neoplasms, malignant and unspecified (SMQ)”.
The SMQ search should include all TEAEs (regardless of whether they have been judged as related to study medication or not) which code to a PT included in the Scope=Narrow group within each SMQ.
Has been changed toAE of special interest is any AE which meets the Hy’s Law criteria, defined as ≥3x upper limit of normal (ULN) alanine aminotransferase (ALT) or aspartate aminotransferase (AST) with coexisting ≥2xULN total bilirubin in the absence of ≥2xULN alkaline phosphatase (ALP), with no alternative explanation for the biochemical abnormality.
REDACTED . P. Papproappro
ame dame day afteay aftescribscrib
COPY pprproacoa
PTs PTs
This do
cumen
t ith
HasHaca
nnotQ seQ se
ed to sd to shin ehin e
be ma
eaea
sedto be o be
alignalig
totic dtic dsup
port
sidido be inbe in
e summsumdisdi
any
derederedmark
eting
er abed in bed in
autho
rizati
on ch wh w
are seare soach bach b
day as ay as a stua stu
appli
catio
nhich cich sitivity tivity
will will
and
he heerged rged
c
any
o
samsamex
tensio
nseitheith
–– notnot
llowlow
or herherva
riatio
ns
will ill
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 140 of 183
AEs of special monitoring for this study include: serious infections (including opportunistic infections and tuberculosis [TB]), cytopenias, hypersensitivities, suicide ideation or behavior (assessed using the eC SSRS), depression and anxiety (assessed using the HADS), major cardiovascular events and liver function test changes/enzyme elevations (ALT, AST, and bilirubin), malignancies, and inflammatory bowel diseases.
AE of special monitoring for this study include:
Infections (serious, opportunistic, fungal and TB)
Malignancies, including lymphoma
Major cardiovascular events
Neutropenia
Neuropsychiatric events (in particular depression and suicide)
Inflammatory bowel disease
Anaphylactic reaction
Hepatic events
For the definitions of AE of special monitoring the Bimekizumab Safety Topics of Interest (Version date 19Feb2018) will be used. The incidence of TEAEs of special monitoring will be summarized by MedDRA system organ class, high level term, and PT. EAIR with associated 95% CI, and the EAER will be included in the summary tables. for following TEAEs of special monitoring: Serious infections are also classified as AE of special monitoring but no separate table will be produced.
The output table for the search criteria Standardized MedDRA Query (SMQ) = “Malignant or unspecified tumours (SMQ)” will include two different overall rows:
• The first overall incidence row will summarize “Any malignancies (including unspecified)” and this row will summarize the incidence of all AEs flagged for inclusion in the table, regardless of the High Level Term (HLT) it codes to.
• The second overall incidence row will summarize “Any malignancy (including unspecified, excluding non-melanomic skin cancers)” and this row will summarize the incidence of AEs flagged for inclusion in the table, excluding those which code to an HLT of “skin neoplasms malignant and unspecified (excl melanoma)”.
The output table for the search criteria SMQ=”Malignant tumours (SMQ)” will include two different overall rows:
• The first overall incidence row will summarize “Any malignancies” and this row will summarize the incidence of all AEs flagged for inclusion in the table, regardless of the High Level Term (HLT) it codes to.
• The second overall incidence row will summarize “Any malignancy (excluding non-melanomic skin cancers)” and this row will summarize the incidence of AEs flagged for inclusion in the table, excluding those which code to an HLT of “skin neoplasms malignant and unspecified (excl melanoma)”.
REDACTEDng wing wassociassoci
Es of Es of ACing buing bRE
COPYe Bie BiPY
llll
This do
cumen
tTTntsumsummL
cann
ottputputno
tdifferifferan
ThTh
be oplpe
t tt tus
edududdAEs flaEs flseplasmlasm
todingdintsu
pport
w wwrtthe Hihe Hpo
all inll ins
anye rowronywill swill mark
eting
ut nut geria Steria Skeurs (SMrs (S
maww
autho
rizati
on
l be sube suiated ated
specispecithno no
appli
catio
n
mekizuekizuap
p
and a
ny ex
tensio
ns or
varia
tions
there
of.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 141 of 183
The output table for Anaphylactic reaction will include three different overall rows:
• The first row within the body of the table will be labeled “Any hypersensitivity/anaphylactic reaction” and will represent the overall incidence of subjects who reported at least one hypersensitivity reaction or at least one anaphylactic reaction.
• The second row within the body of the table will be labeled “Any hypersensitivity reaction” and will represent the overall incidence of subjects who reported at least one hypersensitivity reaction.
• The third row within the body of the table will be labeled “Any anaphylactic reaction” and will represent the overall incidence of subjects who reported at least one anaphylactic reaction.
Fungal infectious disorder
Opportunistic infection (including TB)
Malignant or unspecified tumor
Malignant tumor
Major cardiovascular event
Haematopoietic cytopenias
Neuropsychiatric events
Inflammatory bowel disease
Hypersensitivity and anaphylactic reaction
Hepatic events
Fungal infectious disorder
Fungal infections will be summarized in a stand alone table which presents EAIR and EAER. The table will include all TEAEs which code into the High Level Group Term (HLGT) of “Fungal infectious disorders”.
Opportunistic infectionOpportunistic infections (including TB) will be summarized in a stand alone table which presents EAIR and EAER. The table will include all TEAEs identified using UCB defined search criteria as described in Section 12.13.
Malignant or unspecified tumorThese events will be presented in one stand alone table which will include EAIR and EAER. The table will be based on the criteria Standardized MedDRA Query (SMQ) = “Malignant or unspecified tumours (SMQ)”.
The output table will include 2 different overall incidence rows:
8. The first overall incidence row will summarize “Any malignancies (including unspecified)” and this row will summarize the incidence of all AEs flagged for inclusion in the table, regardless of the High Level Term (HLT) it codes to.
REDACTED
reactireactiDA
COPY
This do
cumen
ttabletablentununm
cese ese eanno
tnant ant noese evse evca
beria aia e us
edfectiectdR and andse
as
ioto
tiontiontsu
pport
EAAders”.ers”.or
iononsu
anyarizarizy
AEs wAEs wamark
eting
zedze
autho
rizati
on
ononau
appli
catio
n and
any e
xtens
ionsaphyph
nsor
actiactor hylhyl
varia
tions
tiontion
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 142 of 183
9. The second overall incidence row will summarize “Any malignancy (including unspecified, excluding non melanomic skin cancers)” and this row will summarize the incidence of AEs flagged for inclusion in the table, excluding those which code to an HLT of “skin neoplasms malignant and unspecified (excl melanoma)”.
Malignant tumorThese events will be presented in one stand alone table which will include EAIR and EAER. The table will be based on the criteria SMQ=”Malignant tumours (SMQ)”.
Note that the events included in the “Malignant tumours” table will be a subset of the events included in the “Malignant or unspecified tumours” table. The SMQ search should include all TEAEs which code to a PT included in the Scope=Narrow group within each SMQ.
The output table will include 2 different overall incidence rows:
10. The first overall incidence row will summarize “Any malignancies” and this row will summarize the incidence of all AEs flagged for inclusion in the table, regardless of the HLT it codes to.
11. The second overall incidence row will summarize “Any malignancy (excluding nonmelanomic skin cancers)” and this row will summarize the incidence of AEs flagged for inclusion in the table, excluding those which code to an HLT of “skin neoplasms malignant and unspecified (excl melanoma)”.
Major cardiovascular events
The major cardiovascular events are identified using the following UCB defined search criteria:
All serious TEAEs which code to a PT included in the Scope=Broad and/or Scope=Narrow groups of the following SMQs:
Haemorrhagic central nervous system vascular conditions (SMQ)
Ischaemic central nervous system vascular conditions (SMQ)
All serious TEAEs which code to a PT included in the HLT “Ischaemic coronary artery disorders” except events coding to PT “Chest Pain” or “Chest discomfort”
All serious TEAEs which code to a PT of “Cardiac failure congestive”
Haematopoietic cytopeniasThese events will be presented in a stand-alone table that is based on the SMQ = “Haematopoietic cytopenias”. The SMQ search should include all serious TEAEs which code to a PT included in the Scope=Broad and/or Scope=Narrow groups within the SMQ.
Neuropsychiatric events
These events will be presented in a stand alone table including EAIR and EAER. The table will be based on the SMQ = “Depression and suicide/self-injury (SMQ)”. The SMQ search will include all TEAEs which code to a PT included in the Scope=Broad and/or Scope=Narrow groups within the SMQ.
Inflammatory bowel disease
REDACTED tified tified CT
a PT a PTED
COPYe to e toY
This do
cumen
tNeureurtThThm
canincluncluno
ttopoopotncludcludnn
uroroc
bents wts we
oieoibus
edic cytc cytedwilwil
to AEsAEs wwsu
ppor
ich cch crtvents coents po
s whs ws
anyus sus y
codco
marketi
ng in
ous systus sysrksyssy
autho
rizati
on
usingusinghoncluncluau
appli
catio
nlignancignanionhe incidhe incican HLTn HLp
and
le,e anyd thd thy
, regaregaaex
tensio
nscludcluds
Q.Q. i
hishi
or ntnr
dedeova
riatio
ns
he he o
ntstsv
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 143 of 183
These events will be presented in a stand alone table which includes EAIR and EAER. The table will include all TEAEs which code into the HLT of “Colitis excl infective”.
Hypersensitivity reactions and anaphylactic reactions
Anaphylactic reactions will be summarized together in a stand alone table.
The first row within the body of the table will be labeled “Any hypersensitivity/anaphylactic reaction” and will represent the overall incidence of subjects who reported at least one hypersensitivity reaction or at least one anaphylactic reaction.
The second row within the body of the table will be labeled “Any hypersensitivity reaction” and will represent the overall incidence of subjects who reported at least one hypersensitivity reaction.
The third row within the body of the table will be labeled “Any anaphylactic reaction” and will represent the overall incidence of subjects who reported at least one anaphylactic reaction.
Following these three overall incidence rows, all TEAEs that have been identified as either a hypersensitivity reaction or an anaphylactic reaction will be summarized (together – not broken out by type) by SOC, HLT and PT.
Hypersensitivity reactions and anaphylactic reactions will be identified as follow:
g) Hypersensitivity reactions: All TEAEs that either emerged on the same day as when a study medication injection reaction was received, or that emerged one day after a study medication injection was received, which code to a PT which contains the term “hypersensitivity” will be considered to be a hypersensitivity reaction and included in the summary table as such.
h) Anaphylactic reactions: An algorithmic approach will be used to identify TEAEs that are considered to be anaphylactic reactions. PTs are separated into the 4 distinct categories (A, B, C, D) prior to the algorithmic approach being applied.
All TEAEs that either emerged on the same day as when a study medication injection reaction was received, or that emerged one day after a study medication injection was received, and which fulfill any of the 3 criteria described in Section 12.14 will be included in the summary table.
Hepatic eventsAlthough not officially considered to be AEs of special monitoring but hepatic events are nonetheless considered to be interesting enough to be summarized in stand alone tables.
Hepatic events will be summarized in a stand alone table that includes all TEAEs in the SMQ “Drug related hepatic disorders comprehensive search (SMQ)”. Note that the following two sub SMQs are to be excluded: “Liver neoplasms, benign (incl cysts and polyps)” and “Liver neoplasms, malignant and unspecified (SMQ)”.
The SMQ search should include all TEAEs (regardless of whether they have been judged as related to study medication or not) which code to a PT included in the Scope=Narrow group within each SMQ.
Section 10.3 Clinical laboratory evaluations
REDACTEDre
which cwhich EDred toed toCT
algoralgorR
COPY ww
t eitheeitheOPeeceivceivC
This do
cumen
tsubub SSntnenem
cann
ot tic eveic evnrug reug reca
bet
s cons cobeus
ed
officofficusto
supp
ortged oed rt3 criter critepo
onany d on d on tny
oneon
marketi
ngrithmirithmngc reactireactkelgorithmgorithmah
autho
rizati
on er
ved, oved, oiocode tcode tza
o be a be a hor
appli
catio
n ar
e idente idenlicemeremerap
anden idn idd
rizedizedaan
yylacylaydende
exten
sionstivityivityns
reactireactxtactict
oron”on”orva
riatio
ns
”
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 144 of 183
…
Markedly abnormal values for biochemistry and hematology are defined in Table 10‒2 and Table 10‒3.The markedly abnormal laboratory results will be listed separately.
Values that are below the lower limit of the reference range will be flagged as “L” (low) and values that are above the upper limit of the reference range will be flagged as “H” (high) and listed as well.
For tables where data are summarized by visit, unscheduled and repeat visits will not be summarized, but these data will be included in listings. In the case where laboratory values are below the lower limit of quantification, then these will be set to the midpoint between 0 and the lower limit of quantification for the purpose of summarizing the data.
Table 13‒2: Definitions of Markedly Abnormal Biochemistry Values
Variable (Standard international units)
Markedly abnormal definition
Low High
ALP N/A >3 x ULN
ALT N/A >3 x ULN
AST N/A >3 x ULN
Calcium (mg/dL) <7.0 >12.5
Creatinine (mg/dL) N/A >1.8 x ULN
Glucose (mg/dL) <40 >250
Potassium (mmol/L) <3.0 >6.4
Sodium (mmol/L) <125 N/A
Total bilirubin N/A ≥2 x ULN
Uric acid N/A ≥3 x ULN
ALP=alkaline phosphatase; AST=aspartate aminotransferase; N/A=Not applicable; ULN=upper limit of normal.
Table 13‒3: Definitions of Markedly Abnormal Hematology Values
Variable (Standard international units)
Markedly abnormal definition
Low High
Hemoglobin (g/dL) <LLNAND>2.0 decrease from Baseline
N/A
Hemoglobin (g/dL) <8.0 N/A
Leukocytes (total x 1000) <2.0 N/A
Lymphocytes (x 1000) <0.5 N/A
Neutrophils (x 1000) <1.0 N/A
REDACTED A ED
<40<40AC
R<RRRER
COPY YCO
This do
cumen
t
umumumum
cann
oogloboglobno
t no
be an us
ed efinfin
ndardared
to initinitsu
pport
AST=aAST
or
su
any
any mark
eting
<3.03.0 g<12<ke
tarmaaaa
autho
rizati
on
oniza
tha
appli
catio
n at
pp
and HiHidanaaaa
any y
aex
tensio
nsn 0 a0 a
e
ores esandan
varia
tions
s areare
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 145 of 183
Table 13‒2: Definitions of Markedly Abnormal Biochemistry Values
Variable (Standard international units)
Markedly abnormal definition
Low High
Platelets (x 1000) <50 N/A
LLN=lower limit of normal; N/A = not applicable.
Has been changed to
…
Markedly abnormal values for biochemistry and hematology will be defined as laboratory values graded 3 or 4 according to the Rheumatology Common Toxicity Criteria (RCTC). Definitions of the markedly abnormal values are given in Table 10‒2 and Table 10‒3 and are based on the RCTC units. All units in the tables below will be converted to the standard units based on Clinical Data Interchange Standards Consortium (CDISC) standards. The markedly abnormal laboratory results will be listed separately.
Values that are below the lower limit of the reference range will be flagged as “L” (low) and values that are above the upper limit of the reference range will be flagged as “H” (high) and listed as well.
For tables where data are summarized by visit, unscheduled and repeat visits will not be summarized, but these data will be included in listings. In the case where laboratory values are below the lower limit of quantification, then these will be set to the midpoint between 0 and the lower limit of quantification for the purpose of summarizing the data.
Table 13‒4: Definitions of Markedly Abnormal Biochemistry Values
Variable (Standard international RCTC units)
Markedly abnormal definition
Low High
ALP N/A >3 x ULN
ALT N/A >3 x ULN
AST N/A >3 x ULN
Calcium (mg/dL) <7.0 >12.5
Creatinine (mg/dL) N/A >1.8 x ULN
Glucose (mg/dL) <40 >250
Potassium (mmol/L) <3.0 >6.4
Sodium (mmol/L) <125 N/A
Total bilirubin N/A ≥2 x ULN
Uric acid N/A ≥3 x ULN
ALP=alkaline phosphatase; AST=aspartate aminotransferase; N/A=Not applicable; RCTC= Rheumatology Common Toxicity Criteria; ULN=upper limit of normal.
REDACTEDun
in listin lishen thehen the
urpose urpose
dd
COPY ana
nschnsch
This do
cumen
tPotasotas
SoSoen
umuuu
cassicann
ot e (m(m
ose (mse (mot
ca
be
mg/mg/eus
ed
L)L) edto o
supp
ort t
pps
any
amark
eting
dly Abny Abn
MMrrrke
ar
autho
rizati
on
heduleeduletings. ings.
hese wese wof suof su
appli
catio
n d s
ill be fll be will bewill b
andnsorsor
ndsepaepaan
yyvertvertanrtiuti
exten
sions
as labas labnsriteriaiteria
xted d TaTaee
or va
riatio
ns
nnnthe
reof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 146 of 183
Table 13‒4: Definitions of Markedly Abnormal Biochemistry Values
Variable (Standard international RCTC units)
Markedly abnormal definition
Low High
Table 13‒5: Definitions of Markedly Abnormal Hematology Values
Variable (Standard internationalRCTC units)
Markedly abnormal definition
Low High
Hemoglobin (g/dL) <LLNAND>2.0 decrease from Baseline
N/A
Hemoglobin (g/dL) <8.0 N/A
Leukocytes (total x 1000) <2.0 N/A
Lymphocytes (x 1000) <0.5 N/A
Neutrophils (x 1000) <1.0 N/A
Platelets (x 1000) <50 N/A
LLN=lower limit of normal; N/A = not applicable; RCTC= Rheumatology Common Toxicity Criteria.
REDACTED COPY
== RheRhCOOP
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
onumatumaonap
plica
tion on
lic
tolap
and N/N
nnnndanan
y A
/A/Aany ex
tensio
ns
ex
or orva
riatio
ns
iononoari
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 147 of 183
13.4 Amendment 413.4.1 Rationale for the amendmentThe main objectives of this SAP amendment are to describe an additional interim analysis Week 48 and correct small errors.
13.4.2 Modification and changesSection 1 IntroductionThe purpose of this statistical analysis plan (SAP) is to provide all necessary information to perform the required interim and final statistical analysis for study PA0008. It also defines the summary tables, figures, and listings (TFLs) to be generated in the clinical study report according to the final Protocol (29 Jul 2016) and Protocol Amendment 1 (16 Dec 2016).
…
Has been changed toThe purpose of this statistical analysis plan (SAP) is to provide all necessary information to perform the required interim and final statistical analysis for study PA0008. It also defines the summary tables, figures, and listings (TFLs) to be generated in the clinical study report according to the final Protocol (29 Jul 2016), and Protocol Amendment 1 (16 Dec 2016), and Protocol Amendment 2 (09 Mar 2018).
…
Section 2.3 Study design and conductTable 2-1 updated to align with Table 5-1 from the Protocol Amendment 2.
Section 3.1 General presentation of summaries and analyses
…
All original and derived variables will be listed and described using summary statistics (number of observations [n], mean, standard deviation [SD], median, minimum and maximum, unless otherwise stated) or frequency counts (number of subjects [N] and percentages).
…
Statistical tests of efficacy variables will be presented as 2-sided p-values rounded to 4 decimal places. P-values less than 0.0001 will be presented as “<0.0001” and p-values greater than 0.9999 will be presented as “>0.9999.” Statistical comparisons will be performed by two-sidedstatistical tests at the 0.0500 level of significance.
…
Has been changed to
…
All original and derived variables will be listed and described using summary statistics (number of observations [n], mean, standard deviation [SD], median, minimum and maximum, unlessotherwise stated) or frequency counts (number of subjects [N] and percentages). For multiple
REDACTED
e 55--1 f1 f
f sf
COPYotocootoc
This do
cumen
t Has Has ca
nnotal teal te bel be l be
tt
used
of eff efues leses le
to ffiffisu
pport
stananuency ency
anyles wles w
ndarndamark
eting
summaumma
wil
autho
rizati
on
om tom t
appli
catio
ne
dy PAdy PAin the cin the
AmenAmen
and ecesseces
any e
xtens
ionsfineine
eport port ec 201c 201
or toto
esesva
riatio
ns
oo
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 148 of 183
post-Baseline assessments at a specific visit, the first non-missing measurement will be used for summary statistics or frequency counts.…
Statistical tests of efficacy variables will be presented as 2-sided p-values rounded to 34 decimal places. P-values less than 0.0001 will be presented as “<0.0001” and p-values greater than 0.9999 will be presented as “>0.9999.” Statistical comparisons will be performed by two-sidedstatistical tests at the 0.0500 level of significance.
…
Section 3.2.2 Study Periods
The following study periods are defined for the classification by study period:
Pre-treatment period (Screening period): up to 28 days, ends with the first dose of bimekizumab.
Double-Blind treatment period: starts with the first dose of study medication (Visit 2), ends at Week 12 prior to the treatment re-randomization.
Dose-Blind treatment period: starts at the Week 12 visit after the treatment re-randomization, ends at Week 48.
Post-treatment period: The post-treatment period (Follow-up period) is the period after the last dose of bimekizumab administration.
Has been changed to
The following study periods are defined for the classification by study period:
Pre-treatment period (Screening period): up to 28 days, ends with the visit date of the first dose of study medication (Visit 2) first dose of bimekizumab.
Double-Blind treatment period: starts with the visit date of the first dose of study medication (Visit 2), ends at Week 12 visit date prior to the treatment re randomization.
Dose-Blind treatment period: starts at the Week 12 visit after the treatment re randomization,ends at Week 48 visit.
Post-treatment period: The post-treatment period (Follow-up period) is the period after the Week 48 visitlast dose of bimekizumab administration.
Section 3.3 Definition of Baseline valuesUnless otherwise specified, the last valid measurement before first dose of study drug (Visit 2)will be used as the Baseline value. The same Baseline definition will be used for the Follow-upperiod. However, for some variables, assessments may be scheduled for Screening only and not for Baseline. In this case the Screening value will be utilized as Baseline value. If a Baseline measurement is missing, and a Screening value available, the Screening value will be utilized as Baseline instead.
If a scheduled Baseline assessment is taken on the same day and after the first administration of study medication, it will be analyzed as the first post-Baseline assessment. If a scheduled
REDACTED
ned foned fo
pe
COPY riod (riod (
This do
cumen
tl
perioperiofofo
cann
ot .3
ss othes othl be ube u
be3 De3 Deus
ednt pnt p
visitisitlalusto perper
supp
ort 122
nt periot pervisitisit.su
anyiod:iod
visitvisitamark
eting
or
eriod): riod): it 2)t 2) firfiark
d: std: s
autho
rizati
on(Foll(Foll
the cthe
appli
catio
n after thafter th
ow
andy medimed
any e
xtens
ions
st dost do
or va
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 149 of 183
Baseline assessment is taken on the same day as first administration of study medication and no time is collected, it will be assumed to have been taken after the study medication.
Has been changed to
Unless otherwise specified, t The last valid measurement before first dose of study drug (Visit 2) will be used as the Baseline value. If a scheduled Baseline assessment is taken on the same day and after the first administration of study medication, it will be analyzed as the first post-Baseline assessment. If a scheduled Baseline assessment is taken on the same day as first administration of study medication and no time is collected, it will be assumed to have been taken after the study medication. The same Baseline definition will be used for the Follow up period.
The exception of the above rule are the questionnaires collected from the vendor ERT and measurements for CRP and hs-CRP. For those measurement, the last valid measurement at the same day as Visit 2 will be used as the Baseline value.However, for some variables, assessments may be scheduled for Screening only and not for Baseline. In this case the Screening value will be utilized as Baseline value. If a Baseline measurement is missing, and a Screening value available, the Screening value will be utilized as Baseline instead.
If a scheduled Baseline assessment is taken on the same day and after the first administration of study medication, it will be analyzed as the first post Baseline assessment. If a scheduled Baseline assessment is taken on the same day as first administration of study medication and no time is collected, it will be assumed to have been taken after the study medication.
Section 3.5 Mapping of assessments performed at early termination visit
…
A visit called ‘End of Treatment’ will be added for the TFLs in the final analysis. The visit will display the Week 48 assessments or the early termination assessment depending if the subject discontinued early or not.
Has been changed to
…
A visit called ‘End of Treatment’ will be added for the TFLs in the final analysis. The visit will display the Week 48 assessments or the early termination assessment depending if the subject discontinued early or not.
Section 3.6 Analysis setsThe primary efficacy variable will be analyzed for all subjects in the Full Analysis Set (FAS). The supportive analyses for the primary efficacy variables will include the Randomized Set (RS) and PPS. All other efficacy variables will be based on the FAS. Demographics tables will be performed for FAS as well as Safety Set (SS). Safety variables will be summarized using the SS. In addition, safety analyses during the Dose-Blind Period will be conducted on all subjects in the Dose-Blind Set (DBS). PK variables will be analyzed for all subjects in the PK-PPS. PD variables will be analyzed for all subjects in the PD-PPS.
…
REDACTEDy as fy as fDas fias fED
ve beene beenCT
perfoperfo
COPY same dsame PY
postpost BCOi
This do
cumen
tThe The anan
cann
otn 3.6 3.6 e primprim
beed eaed ee us
ednd od odeek 48ek 48se
arar
to of Tof Ttsu
pport
any m
arketi
ng
will be all beor theor th
autho
rizarst adst atio
nBaseBaseonst admt adaten taketakeori
ormedrmed
appli
catio
n inecreeninreeni
day anay anapli
andeninenin
e vavaan
y ng o
exten
sions
endorendotenalid mlid mex
or rhe he orva
riatio
ns
as s rio hav havv
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 150 of 183
Section 3.6.9 Escape Subject Set (ESS)The Escape Subject Set (ESS) consists of all subjects who started the Dose-Blind Period and who received at least 1 dose of the IMP during the Dose-Blind Period and who have not achieved at least a 10% reduction from Baseline in SJC and TJC at Week 16, Week 24, or Week 36.
Section 3.6.10 Dose-Blind Responder Set (DBRS)
The Dose-Blind Responder Set (DBRS) consists of all subjects who started the Dose-Blind Period and who received at least 1 dose of the IMP during the Dose-Blind Period and who have achieved at least a 10% reduction from Baseline in SJC and TJC at Week 16, Week 24, and Week 36.
Has been changed toThe primary efficacy variable will be analyzed for all subjects in the Full Analysis Set (FAS). The supportive analyses for the primary efficacy variables will include the Randomized Set (RS) and PPS. All other efficacy variables will be based on the FAS. Demographics tables will be performed for FAS, as well as Safety Set (SS), and Dose-Blind Set (DBS). Safety variables will be summarized using the SS. In addition, safety analyses during the Dose-Blind Period will be conducted on all subjects in the Dose Blind Set (DBS). PK variables will be analyzed for all subjects in the PK-PPS. PD variables will be analyzed for all subjects in the PD-PPS.
At the time of the Week 48 interim it was discovered that the Dose-Blind Responder Set (DBRS) and Escape Subject Set (ESS) were defined incorrectly from the original intent. The original intent of the ESS analysis set was to have only those subjects that were eligible to receive rescue therapy, that went on to actually receive rescue therapy, with all other subjects remaining in the DBRS. As such, two extra sets, the Corrected Dose-Blind Responder Set (cDBRS) and Corrected Escape Subject Set (cESS) have been added. These analysis sets will correctly select out those subjects that were eligible and received rescue medication. All outputs created on the DBRS and ESS will also be produced on the cDBRS and cESS to give full transparency of the difference. Further all outputs using the DBRS and ESS will be repeated on the DBS and this analysis will be used as the main analysis of the Dose-Blind as this analysis set is the most unbiased set.…
Section 3.6.9 Escape Subject Set (ESS)The Escape Subject Set (ESS) consists of all subjects who started the Dose-Blind Period and who received at least 1 dose of the IMP during the Dose-Blind Period and who have not achieved at least a 10% reduction from Baseline in SJC and TJC at Week 16, Week 24, or Week 36.
Section 3.6.10 Dose-Blind Responder Set (DBRS)
The Dose Blind Responder Set (DBRS) consists of all subjects who started the Dose-Blind Period and who received at least 1 dose of the IMP during the Dose-Blind Period and who have achieved at least a 10% reduction from Baseline in SJC and TJC at Week 16, Week 24, and Week 36.
3.6.11 Corrected Escape Subject Set (cESS)
REDACTED scocDere defre de
TEset wset wAC
on toon toED
As suAs suR
COPY ). P.zed fozed fo
overove
This do
cumen
t 6..
SecSecca
nnoteiveive
eved aved abee Sue Subeeded
used
Escapscapbjbj
to su
pportptd on ton or analyanaly
up
any d od yarencarenath
marketi
nguch, tch, ingcted Ected E
keout thout thmaon ton
autho
rizati
on
red thed tatiefinedfinedrizwas to as tothactuactuatt
appli
catio
nSet (Det (onng the Dg the variabvariab
r all suall s
and
ththmograpograp
DD
anyAnalAna
he Rae Raex
tensio
ns24, a4, a
alyaly
or o ho
andanva
riatio
ns
havhav
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 151 of 183
The cESS is a corrected version of the ESS, which was incorrectly defined previously. The cESS consists of subjects that have shown less than a 10% improvement in SJC and TJC at either Week 16, Week 24, or Week 36 and received rescue therapy.
3.6.12 Corrected Dose-Blind Responder Set (cDBRS)
The cDBRS is a corrected version of the DBRS, which was incorrectly defined previously. The cDBRS consists of subjects that have shown at least a 10% improvement in SJC or TJC at Week 16, Week 24 and Week 36. Subjects that would be in the cESS, including subjects that discontinue, that did not receive rescue therapy will be in the cDBRS.
Section 3.7 Treatment assignment and treatment groups
…
For AE tables and listings, AEs for each subject will be summarized based on the treatment actually received at the onset of each particular AE. It should be noted that this could result in one subject being summarized in two different treatment groups within one output if they were on a treatment regimen that involved treatment switching and experienced AEs under both treatment allocations.
Efficacy analyses will be performed according to randomization and not actual treatment received.
Has been changed to
…
For AE tables and listings, AEs for each subject will be summarized based on the treatment actually received at the onset of each particular AE. It should be noted that this could result in one subject being summarized in two different treatment groups within one output if they were on a treatment regimen that involved treatment switching and experienced AEs under both treatment allocations.
Efficacy analyses will be performed according to randomization and not actual treatment received.
Depending on the analysis sets, subjects will be summarized and listed based on the actualreceived treatment or according to the randomization which will be considered as planned treatment:
All subjects screened/ES: planned treatment
RS: planned treatment
SS: actual treatment
FAS: planned treatment
PPS: planned treatment
PK-PPS: actual treatment
PD-PPS: actual treatment
REDACTED
subjesubjACparticparticED
o dio diR
COPY
This do
cumen
tFFcann
otplaplaotSS: aS: aca
be ec
eannan
used
cts scts su
to t ororto
supp
ort
alysis slysisupor acr ac
anyormedormenymark
etingcu
ifferenfferenngd treatmtreatmke
d
autho
rizati
on
ect wict withoular Alar Aau
appli
catio
n tion andtion an
and
ooencedenced
anythisth
ne oune ouex
tensio
ns
n the tthe isis
or va
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 152 of 183
DBS: planned treatment (demographics, baseline characteristics, and efficacy analyses)or actual treatment
ESS: planned treatment
DBRS: planned treatment
cESS: planned treatment
cDBRS: planned treatment
Section 4.2.3 Handling of missing data for prior and concomitant medicationAny medications with incomplete start and end dates/times will be handled according to the following rules for classification as prior and concomitant and for the calculation of relative study days. Such imputations will only be performed for these classifications and calculations; in the listings all data will be shown as recorded on the electronic case report form (eCRF).
Imputation of partial start dates:
If only the month and year are specified and the month and year of first dose is not the same as the month and year of the start date and the subject did not switch treatment during that month and year, then use the 1st of the month.
If only the month and year are specified and the month and year of first dose is the same as the month and year of the start, then use the date of first dose.
If only the month and year are specified and the subject did switch treatment during that month and year, then use the date of first switch treatment.
If only the year is specified, and the year of first dose is not the same as the year of the start date and the subject did not switch treatment during that year, then use the 1st of January of the year of the start date.
If only the year is specified, and the year of first dose is the same as the year of the start date, then use the date of first dose.
If only the year is specified, and the year of first dose is the same as the year the subject did switch treatment, then use the date of first switch treatment.
If only the year is specified, and the year of first dose is the same as the year of the start date and the same as the year the subject did switch treatment, then use the date of first dose.
If the start date is completely unknown and the stop date is unknown or not prior to the date of first dose, then use the date of first dose.
If the start date is completely unknown and the stop date is prior to the date of first dose, then set the start date to the 1st of January of the year of the end date.
If the date of first study medication or switch treatment is partial, then the above imputation approach will be applied:
If only the day of first study medication or switch treatment administration is unknown, then this will be set to either the first of the month, or the day of first treatment allocation visit (excluding placebo run-in), whichever is later.
REDACTED and thand t
first swfirst s
the yethe yech tch t
COPY e mone mon
date odate o
This do
cumen
tssIfIf
cann
otfirst drst d
If theIf th
betart dtart us
ed ar iar
me as me as tois sis s
supp
ort
ecifiedcifiethen ushen u
any anan
ose.ose.mark
eting
ear ar treatmreatm
nd thnd t
autho
rizati
onof firsof fir
the subhe subwitchwitch
r ofof
appli
catio
n ar oot switt swi
th andh andt
and
of fof fan
y foro exten
sionsng to tg to
n of rof reeand cand c
orm rm
or va
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 153 of 183
If both the day and month of first study medication or switch treatment are unknown, then this date will be imputed to be the date of first treatment allocation visit.
If first treatment allocation visit date is missing then the all activities events will be assumed to be treatment emergent..
…
Has been changed to
Any medications with incomplete start and end dates/times will be handled according to the following rules for classification as prior and concomitant and for the calculation of relative study days. Such imputations will only be performed for these classifications and calculations; in the listings all data will be shown as recorded on the electronic case report form (eCRF).
Imputation of partial start dates:
If only the month and year are specified and the month and year of first dose is not the same as the month and year of the start date and the subject did not switch treatment during that month and year, then use the 1st of the month.
If only the month and year are specified and the month and year of first dose is the same as the month and year of the start, then use the date of first dose.
If only the month and year are specified and the subject did switch treatment during that month and year, then use the date of first switch treatment.
If only the year is specified, and the year of first dose is not the same as the year of the start date and the subject did not switch treatment during that year, then use the 1st of January of the year of the start date.
If only the year is specified, and the year of first dose is the same as the year of the start date, then use the date of first dose.
If only the year is specified, and the year of first dose is the same as the year the subject did switch treatment, then use the date of first switch treatment.
If only the year is specified, and the year of first dose is the same as the year of the start date and the same as the year the subject did switch treatment, then use the date of first dose.
If only the year is specified and the end date is before date of first dose, then set the start date to the 1st of January of the year of the start date.
If only the year and day are specified and month is missing, then only the year will be considered and the month and day will be imputed with the rules above
If the start date is completely unknown and the stop date is unknown or not prior to the date of first dose, then use the date of first dose.
If the start date is completely unknown and the stop date is prior to the date of first dose, then set the start date to the 1st of January of the year of the end date.
If the date of first study medication or switch treatment is partial, then the below aboveimputation approach will be applied:
REDACTED witcwit
ear of ar of treatmtreatm
COPY of ff
he subhe sutchtch
This do
cumen
t I cann
ot nly tly t
otonsidensidennIf thIf th
be
thetheus
edyear earedto thto thu
to th
iisu
pport
use se
pecifiepecifihe yeae yea
any, and, an
the the
marketi
ng
he yearhe yea
d td
autho
rizati
onbjecjec
treatmtreatm
first dirst dment duent d
appli
catio
n d year oyear
t dose.dose
ect dct d
andtch trch t an
yrst dorst dorea
exten
sionsa
alculalcula(eCRFeCRF
ortivetiveorva
riatio
ns
he e
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 154 of 183
If only the day of first study medication or switch treatment administration is unknown, then this will be set to either the first of the month, or the day of first treatment allocation visit (excluding placebo run-in), whichever is later.
If both the day and month of first study medication or switch treatment are unknown, then this date will be imputed to be the date of first treatment allocation visit.
If first treatment allocation visit date is missing then the all activities events will be assumed to be treatment emergent..
…
Section 4.3 Interim analyses and data monitoring
…
Exposure of study medication will not be summarized but a listing will be provided including following information: date of first dose, date of last dose, and duration of exposure. The last dose for the Double-Blind Period is planned on Visit 6 (Week 8). For subject who completed the Double-Blind Period the Week 8 date is the date of the last dose. If a subject discontinued within the Double-Blind Period, the discontinuation date will be presented for the date of last dose.
…
Section 4.3.1 Changes from interim analysis to SAP-defined analyses
For TEAE two summary tables were presented as described in previous section with one modification: the first table was displaying all TEAE during the Double-Blind Period instead of limiting it TEAE with a relative TEAE start date less equal 84 days.
The definitions of AE of special monitoring were updated with the third SAP Amendment after the interim analysis. The interim analysis used the previous definition which is displayed in Section 4.3.1.2. The interim analysis only displayed the overall number of AE of special monitoring by treatment group.
The definition of time since first diagnosis of PsA was updated with the third SAP Amendment after the interim analysis. In the previous definitions the date of informed consent was subtracted from the date of diagnosis which results in negative values. For the interim analysis another calculations were used:
Section 4.3.2 Search and selection criteria for AE of special monitoring for interim analysis…
Has been changed to
Section 4.3 Interim analyses and data monitoring
Two interim analyses are planned for the study after the subjects have completed 12 and 48 weeks.
REDACTEDd as d as all TEall T
start dstart d
nitorinitori
COPY SAPSAP
dd
This do
cumen
t ……ca
nnot
tion tion
be us
ed
sededto
supp
ortfifi
In theIn thosis whosis w
d:d:
any
rst dirst dimark
eting
ing wng wysis usesis us
is onlys only
autho
rizati
on dede
describescribEAE EAE
date leate le
wew
appli
catio
n f a nted foted fo
efineefin
andr subjsubj
a subsuan
yo
of expf expjecj
exten
sions
ovidovid
or va
riatio
ns
ed d ithe
reof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 155 of 183
Section 4.3.1 Interim analysis Week 12
….
Exposure of study medication will not be summarized but a listing will be provided including following information: date of first dose, date of last dose, and duration of exposure. The last dose for the Double-Blind Period is planned on Visit 6 (Week 8). For subject who completed the Double Blind Period the Week 8 date is the date of the last dose. If a subject discontinued within the Double Blind Period, the discontinuation date will be presented for the date of last dose.
…
Section 4.3.1.1 Changes from interim analysis Week 12 to SAP-defined analyses
For TEAE two summary tables were presented as described in previous section with one modification: the first table was displaying all TEAE during the Double-Blind Period instead of limiting it TEAE with a relative TEAE start date less equal 84 days.
The definitions of AE of special monitoring were updated with the third SAP Amendment after the interim analysis. The interim analysis used the previous definition which is displayed in Section 4.3.1.2. The interim analysis only displayed the overall number of AE of special monitoring by treatment group.
The definition of time since first diagnosis of PsA was updated with the third SAP Amendment after the interim analysis. In the previous definitions the date of informed consent was subtracted from the date of diagnosis which results in negative values. For the interim analysis another calculations were used:
The baseline definition for CRP and hs-CRP measurements were updated with the fourth SAP Amendment after the interim analysis. The interim analysis used the previous definition: The last valid measurement before first dose of study drug (Visit 2) will be used as the Baseline value.
Section 4.3.2 Interim analysis Week 48
After all enrolled subjects have completed the Week 48 or Early termination visit, a second interim analysis will be performed to analyze the key efficacy and safety data for the whole treatment period.
No separate SAP for that interim analysis will be provided. The interim analysis is a subset of the final analysis and will focus on the primary and secondary efficacy analysis. The TFL shells for the final analysis will be used.
The snapshot for the PA0008 interim analysis Week 48 will occur before all subjects have completed the PA0008 study. Specifically, subjects who do not enter the extension study will need to complete the SFU Visit, which occurs 20 weeks after the last dose of study treatment. The Week 48 interim database lock will be performed based on the last subject completing the Week 48 visit. It is anticipated that there will be some subjects still awaiting
REDACTEDgativgatiD
EDR
COPYwas upwas uPYons thons thCOve
This do
cumen
t f ththntTFLTFL
meca
nnot
nt pnt pot
separaeparaanhe fhe f
bealyslybe
perper
used
lled sled sus
eysissis
torim im tosu
pporttd meamea
po
an
any P P y nterimnteriamark
eting
ng
etieti
andan
autho
rizati
onhe dahe dane valuvaluati
a
appli
catio
nio
numbenumb
pdated datedppatet
andird Srd S
on whn wan
y SAPSA
exten
sions
s
with owith d PeriPer
or va
riatio
ns
in in othe
reof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 156 of 183
the SFU at that time. The number of subjects in the SFU period is expected to be low, and the minimal data to be collected from their SFU visits is not considered mandatory to evaluation of the key efficacy and safety objectives of the study.
Section 4.3.2.1 Changes from interim analysis to SAP-defined analyses
The baseline definition for CRP and hs-CRP measurements were updated with the fourth SAP Amendment after the interim analysis. The interim analysis used the previous definition: The last valid measurement before first dose of study drug (Visit 2) will be used as the Baseline vale.
Section 4.3.2.2 Changes during interim analysis to SAP-defined analyses
At the time of the Week 48 interim it was discovered that the DBRS and ESS were defined incorrectly from the original intent. The original intent of the ESS analysis set was to have only those subjects that were eligible to receive rescue therapy, that went on to actually receive rescue therapy, with all other subjects remaining in the DBRS. As such, two extra sets, the cDBRS and cESS have been added. These analysis sets will correctly select out those subjects that were eligible and received rescue medication. For the final analysis, all outputs created on the DBRS and ESS will also be produced on the cDBRS and cESS to give full transparency of the difference. Further all outputs using the DBRS and ESS will be repeated on the DBS and this analysis will be used as the main analysis of the Dose-Blind as this analysis set is the most unbiased set.
Section 6.1 Demographics…
The summary tables will be performed on the SS and repeated using the FAS. If the SS and FAS analysis sets are identical the summaries will not be repeated. The Listing will be provided for all subjects screened, except the Lifestyle listing will use the RS.
Has been changed to
…
The summary tables will be performed on the SS and repeated using the FAS and the DBS. If the SS and FAS analysis sets are identical the summaries will not be repeated. The Listing will be provided for all subjects screened, except the Lifestyle listing will use the RS.
Section 6.2 Other Baseline characteristics
…
Baseline characteristics (including Scores relevant for inclusion and exclusion criteria) will be summarized by treatment group and overall for FAS and SS. Following variables will be summarized:
…
Has been changed to
…
REDACTED
ed on ed on ariearie
COPYused usedPY
set.set.CO
This do
cumen
tsummummca
nnot ine chne ch
mmarimar
be us
ed
ther Bher toy
subjsub supp
ort
ll be pebe pysis setsis se
je
any m
arketi
ng tht
es will s will yle listine list
autho
rizati
on
he Se S
appli
catio
n n. .nd on thon tati
outs usinuts uslicas the s the
app
andwill cill cnd FoFo
any
ntyS. S. AsAsanor
exten
sions
S werwer
nsis set s set xtet onone
or va
riatio
ns
usedsedria
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 157 of 183
Baseline characteristics (including Scores relevant for inclusion and exclusion criteria) will be summarized by treatment group and overall for FAS, and SS and DBS. Following variables will be summarized:
…
Section 6.4 Prior and concomitant medicationsPrior medications include any medications that started prior to the start date of study medication. Past medications are a subset of prior medications, and include prior medications with a stop date before the date of first study medication administration. Concomitant medications are medications taken at least one day in common with the study medication dosing period. For bimekizumab, the dosing period is typically from the date of first dose up to (but not including) one dosing interval post last dose. Thus, a concomitant medication is any medication whose start date is on or after the date of first study medication and prior to the date of last study medication administration + 28 days (1 dosing interval), and whose stop date is either missing, or on or after the date of first study medication administration. Medications may be both prior and concomitant.
In the case of missing data, the classification of medications as prior or concomitant will be performed as described in Section 4.2.3. Imputations of missing data will be performed before calculation of relative study days.
…
Has been changed toPrior medications include any medications that started and ended prior to the start date of study medication. Past medications are a subset of prior medications, and include prior medications with a stop date before the date of first study medication administration. Concomitant medications are medications taken at least one day in common with the study medication dosing period. For bimekizumab, the dosing period is typically from the date of first dose up to (but not including) one dosing interval post last dose. Thus, a concomitant medication is any medication whose start date is on or after the date of first study medication and prior to the date of last study medication administration + 28 days (1 dosing interval), and whose stop date is either missing, or on or after the date of first study medication administration. Medications may be both prior and concomitant.
In the case of missing data, the classification of medications as prior or concomitant will be performed as described in Section 4.2.3. Imputations of missing data will be performed before calculation of relative study days.
…
Section 7 Measurements of treatment complianceCompliance will be summarized as the number of doses received relative to the number of doses scheduled:
(8)
REDACTED
ons thaons thubset oubset oEDrst srst sR
COPY
This do
cumen
tSectiectiCoC
cann
ot on nt
ioi
be as ds d
n of rn of rbus
ed
missinmissindescdesc
to su
pport
er tr on + 28n + 2
first stfirst s
anypost post
the dthemark
eting
of ostudy mtudy ing
t least leasting ping pe
lal
autho
rizati
on
hat start starf priopria
appli
catio
n prior oprior sing daing da
and botot an
ymismisth prth p
exten
sions
od. Fd. Fnot incot inc
cation ationast stust stu
ss
or FoFova
riatio
ns
on. on. op datep date
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 158 of 183
where the total number of expected doses is derived relative to when the subject finishes treatment. If a subject completes treatment 12 doses are expected (Baseline and every 4th week afterwards until Week 48). If a subject discontinues early, then the number of expected doses is based on the time of early discontinuation relative to the dosing visits. If a dose is not completely given at a specific visit (eg, subject only received one injection instead of the two planned injections), then the subject will be considered to have no compliance for the visit. In the formula above it will be counted as no dose received at this visit.
A summary of percent treatment compliance categorized as ≤80% and >80% will be provided by treatment group.
…
Has been changed toCompliance will be summarized as the number of doses received relative to the number of doses scheduled:
(8)
where the total number of expected doses is derived relative to when the subject finishes treatment. If a subject completes treatment 12 doses are expected (Baseline and every 4th week afterwards until Week 4448). If a subject discontinues early, then the number of expected doses is based on the time of early discontinuation relative to the dosing visits. If a dose is not completely given at a specific visit (eg, subject only received one injection instead of the two planned injections), then the subject will be considered to have no compliance for the visit. In the formula above it will be counted as no dose received at this visit.
A summary of percent treatment compliance categorized as ≤80% and >80% will be provided by treatment group for the overall treatment period as well as for double-blind treatment period.
…
Section 8.1.1.3 Physician’s Global Assessment of Disease Activity (PhGADA)
The Investigator will assess the overall status of the subject with respect to their PsA signs and symptoms and functional capacity (considering both joint and skin components) using a numerical rating scale (NRS) where 0 is “very good, asymptomatic and no limitation of normal activities” and 10 is “very poor, very severe symptoms which are intolerable and inability to carry out all normal activities”.
Has been changed to
The Investigator will assess the overall status of the subject with respect to their PsA signs and symptoms and functional capacity (considering both joint and skin components) using a numerical rating scale (NRS) where 0 is “very good, asymptomatic and no limitation of normal activities” and 100 is “very poor, very severe symptoms which are intolerable and inability to carry out all normal activities”.
Section 8.3 Analysis of other efficacy variables
REDACTED ntintin relatin relat
ubjectubjecwill be will be
no dono do
COPYrelatrelaes arees ar
tinueinu
This do
cumen
t as
TheTheca
nnot a
out allut all
s beebe
be atin
and and se
dfunctfunctng sng s
toill all a supp
ort
cian’s ian’s
ss
any m
arketi
ng ose rese r
mpliancmplianreatmeeatmm
autho
rizati
on e exex
es ears earive tove to
t only t onlyconsiconsi
appli
catio
n n
ve to we to expecxpe
and a
ny ex
tensio
ns
the nhe n
or va
riatio
ns
ded byed by
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 159 of 183
All other efficacy variables in the Double-Blind Period will be analyzed for all subjects in the FAS. For the Dose-Blind period, other efficacy variables will be analyzed for all subjects in the DBRS and ESS.
…
Time to onset of ACR20/50 response will be estimated and presented using the Kaplan-Meier product-limit method for each treatment. Time to a given response will be defined as the length in weeks from Baseline until the first date when the response is achieved. Following derivation will be used: Days from Baseline until the first date divided by 7. There will be no rounding for the Kaplan Meier estimates. Subjects who discontinue study treatment prior to achieving a response will be censored at the date of study treatment discontinuation. Subjects who reach the Week 12 visit without achieving a response will be censored at the date of the Week 12 visit. Missing response data will be imputed with NRI.
…
Has been changed toAll other efficacy variables in the Double Blind Period will be analyzed for all subjects in the FAS. For the Dose Blind period, other efficacy variables will be analyzed for all subjects in the DBRS and ESS.
Other efficacy variables will be analyzed for all subjects in the FAS, DBRS, ESS, cDBRS, cESS, and DBS.
…
Time to onset of ACR20/50 response will be estimated and presented using the Kaplan-Meier product-limit method for each treatment. Time to a given response will be defined as the length in weeks from Baseline until the first date when the response is achieved. Following derivation will be used: Days from Baseline until the first date divided by 7. There will be no rounding for the Kaplan Meier estimates. Subjects who discontinue study treatment prior to achieving a response will be censored at the date of study treatment discontinuation. Subjects who reach the Week 12 visit without achieving a response will be censored at the date of the Week 12 visit. Missing response data will be imputed with NRI.
…
Section 10.1 Extent of exposure
The duration of exposure (in days) will be calculated as:
(16)
28 days refer to one half-life of bimekizumab. If a change of treatment occurs prior to completion of the dosing interval, duration of exposure to study medication will be calculated as:
(17)
For subjects who have died the exposure will be as follows:
REDACTED
will bewill bement. ment.
dd
COPY l subjsubCO
This do
cumen
t 28 28
cann
ot be
on ofon o usedxtentxtent
f ef e
to su
pport
ievinevinwill be will be
any
jhe dathe dang
marketi
ng TimTim
date whate wntil thentil th
ects wects
autho
rizati
on jectecn
e estimestimmeme
appli
catio
nnalyzalyznbe anale anaati
cts ints i
and
ze
any e
xtens
ions
nggwho reho re
Week 1eek 1
or g a g a va
riatio
ns
ion on ng forg fo
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 160 of 183
(18)
The sum over all duration of exposure (in years) will be calculated for the total subject-years of duration of exposure.
The duration of exposure will be summarized by treatment group using descriptive statistics. A by-subject listing of date of first and last dose and the duration of exposure will be performed.
Time at risk (in days) is defined as:
(19)
where 140 days refers to 5*half-life of bimekizumab.
The sum over all time at risk (in years) will be calculated for the total subject-years of time at risk.
The days on treatment and days on bimekizumab treatment will be calculated as follows:
(20)
(21)
Has been changed to
The duration of exposure (in days) will be calculated as:
(16)
28 days refer to one half-life of bimekizumab. If a change of treatment occurs prior to completion of the dosing interval, duration of exposure to study medication will be calculated as:
(17)
For subjects who have died the exposure will be as follows:
(1718)
The sum over all duration of exposure (in years) will be calculated for the total subject-years of duration of exposure.
The duration of exposure will be summarized by treatment group using descriptive statistics. A by-subject listing of date of first and last dose and the duration of exposure will be performed.
Time at risk (in days) is defined as:
REDACTEDCOPY
This do
cumen
t TT
otbjectsjects
cann
be us
edededto
os supp
ort
lf-f life olife ing ining insu
any m
arketi
ng
ill be cill be
autho
rizati
on ap
plica
tion a
ndalculaculaany
atet
exten
sions
t-yearyear
or(19(19va
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 161 of 183
(1819)
where 140 days refers to 5*half-life of bimekizumab.
The sum over all time at risk (in years) will be calculated for the total subject-years of time at risk.
The days on treatment and days on bimekizumab treatment will be calculated as follows:
(1920)
(2021)
Section 10.2.1 Exposure duration
The duration of each AE will be calculated as follows:
+ 1 (22)
For the calculation of duration for AEs with missing start dates, the dates will be imputed as described in Section 4.2.2.
The time to first dose for each AE will be calculated as follows for all TEAEs:
(23)
The time to most recent dose for each AE will be calculated as follows for all TEAEs:
(24)
Time to first dose and time to most recent dose will not be calculated for pretreatment AEs.
For the calculation of time to first bimekizumab and time to most recent bimekizumab dose the same formulas as above (20 and 21) will be used but the date of first bimekizumab dose will be used instead of the date of first dose. The time to first dose and time to first bimekizumab dose differs only for those subjects who receive placebo at Baseline.
Has been changed to
The duration of each AE will be calculated as follows:
+ 1 (2122)
missimissi
be cabe ca
COPY
inging
This do
cumen
tdiffeiffe
HH
cann
otl
formuormud insteinsteersers
beculaculaus
ed
dose aose ato
supp
ort an
yr ear eamark
eting
ach
autho
rizati
on
g startstar
lculatculat
appli
catio
n and
any e
xtens
ions ((s or 191922
varia
tions
there
of.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 162 of 183
For the calculation of duration for AEs with missing start dates, the dates will be imputed as described in Section 4.2.2.
The time to first dose for each AE will be calculated as follows for all TEAEs:
(2223)
The time to most recent dose for each AE will be calculated as follows for all TEAEs:
(2324)
Time to first dose and time to most recent dose will not be calculated for pretreatment AEs.
For the calculation of time to first bimekizumab and time to most recent bimekizumab dose the same formulas as above (2220 and 2321) will be used but the date of first bimekizumab dose will be used instead of the date of first dose. The time to first dose and time to first bimekizumab dose differs only for those subjects who receive placebo at Baseline.
Section 10.2.2 Exposure adjusted incidence rate (EAIR) and exposure adjusted event rate (EAER)The EAIR is defined as the number of subjects (n) with a specific AE adjusted for the exposure and will be scaled to 100 patient-years:
(25)
where is a subject’s exposure time in years up to the first occurrence of the AE of interest (equation [21] in years) at the level of coding evaluated.
If a subject has multiple events at the level of coding evaluated, the time of exposure is calculated to the first occurrence of the AE of interest. If a subject has no events, the total time at risk (15) is used.
Exact Poisson 95% CIs for incidence rates are calculated using the relationship between the Poisson and the Chi-square distribution (Ulm, 1990; Fay and Feuer, 1997):
(26)
(27)
where n is the number of subjects with a specific AE for the incidence rate of interest and is the basis for the number of the degrees of freedom for the chi-square quantile for the upper tail probability χ2.
REDACTEDts (n)ts (n COPYEAIEAI
This do
cumen
t can
not b
e use
d5% C% he Che Ch
to CICsu
pportventsents
ccurrencurrean
yevel eve
s a
marketi
ng
re timee timl ol
autho
rizati
on
with with
appli
catio
nd timtimaseline.seline
R) andR) and
and nt bt
f firstf firstan
y etret
bimebimex
tensio
ns
((22
trearea
or va
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 163 of 183
The EAER will be the number of AEs including repeat occurrences in individual subjects divided by the total time at risk scaled to 100 patient-years and calculated using:
(28)
where NAE is the total number of AEs and is a subject’s total time at risk in years.
No CI will be computed for EAER.
Has been changed toThe EAIR is defined as the number of subjects (n) with a specific AE adjusted for the exposure and will be scaled to 100 patient-years:
(2425)
where is a subject’s exposure time in years up to the first occurrence of the AE of interest (equation [21] in years) at the level of coding evaluated, nnoAE the number of subjects without the specific AE and the total time at risk scaled to 100 patient-years (equation [18] in years).If a subject has multiple events at the level of coding evaluated, the time of exposure is calculated to the first occurrence of the AE of interest. If a subject has no events, the total time at risk (1815) in years is used.
Exact Poisson 95% CIs for incidence rates are calculated using the relationship between the Poisson and the Chi-square distribution (Ulm, 1990; Fay and Feuer, 1997):
(2526)
(2627)
(27)
(28)
where n is the number of subjects with a specific AE for the incidence rate of interest and is the basis for the number of the degrees of freedom for the chi-square quantile for the upper tail probability χ2, is a subject’s exposure time in years up to the first occurrence of the
REDACTED of codof cod
AE of AE of
e rae ra
COPY edd,scaledscaleOP
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
ates aretes ar
tion (Uon (U
autho
rizati
on
ding eing eintereintere
appli
catio
n t occuroccu
oAEAE thethplpl to 10to 10ap
and a
ny ex
tensio
ns
thethe eeor
varia
tions
there
of.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 164 of 183
AE of interest (equation [21] in years) at the level of coding evaluated, nnoAE the number of subjects without the specific AE and the total time at risk scaled to 100 patient-years (equation [18] in years).The EAER will be the number of AEs including repeat occurrences in individual subjects divided by the total time at risk scaled to 100 patient-years and calculated using:
(2928)
where NAE is the total number of AEs and is a subject’s total time at risk in years and nAll
the number of subjects.
No CI will be computed for EAER.
Section 10.3 Clinical laboratory evaluations…
Different summary tables for hematology and biochemistry variables will be provided: observed values and change from Baseline, from Baseline to maximum post-Baseline value, from Baseline to minimum post-Baseline value, shift from Baseline to end of treatment, shift from Baseline to end of double-blind treatment, and markedly abnormal laboratory data.
…
Table 13‒1: Laboratory measurements
Hematology Biochemistry Urinalysis
Basophils Calcium Albumin
Eosinophils Chloride Bacteria
Lymphocytes Magnesium Crystals
Atypical lymphocytes Potassium Glucose
Monocytes Sodium pH
Neutrophils Glucose RBC
Hematocrit BUN WBC
Hemoglobin Creatinine Urine dipstick for pregnancy testingb
MCH hs-CRP/CRPa
MCHC AST
MCV ALT
Platelet count GGT
RBC count ALP
REDACTED
mentsmentsRE
COPY axiaxne to ne to
normaorma
This do
cumen
tMCMCntnt
mumumum
cann
ot itt
oglobiglobiot b
e e used
ed
to osu
pport
tpp
s
any ChCyanaaaamark
eting
chemischemi
g
CalciumCalciurk
m
autho
rizati
one
al labal labap
plica
tion
ariableriablum posm po
nd ofnd of
and a
ny ex
tensio
ns
n yearsyearor
varia
tions
288))a
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 165 of 183
Table 13‒1: Laboratory measurements
Hematology Biochemistry Urinalysis
WBC count Total bilirubin
LDH
Total cholesterol
Serum pregnancy testingb
ALP=alkaline phosphatase; ALT=alanine aminotransferase; AST=aspartate aminotransferase; BUN=blood urea nitrogen; CRP=C-reactive protein; GGT=gamma glutamyltransferase; hs-CRP=high sensitivity C-reactive protein; LDH=lactate dehydrogenase; MCH=mean corpuscular hemoglobin; MCHC=mean corpuscular hemoglobin concentration; MCV=mean corpuscular volume; RBC=red blood cell; SFU=Safety Follow-up; WBC=white blood cella Both CRP and hs-CRP will be tested at specified visits per Protocol Table 5.1.b A serum pregnancy test will be performed at Screening for all women of childbearing potential. This recommendation also applies to women of childbearing potential with infrequent or irregular menstrual cycles and women during menopause. Natural menopause is recognized to have occurred after 12 consecutive months of amenorrhea, for which there is no other obvious pathological or physiological cause (International Menopause Society, 2015). A urine pregnancy test is also required at the Baseline, Week 48, ET, and at SFU visits. A urine pregnancy test will also be performed at any study visit where there has been a delay in menses. This recommendation also applies to women of childbearing potential with infrequent or irregular menstrual cycles. Pregnancy test results must be negative prior to administering IMP.
Has been changed to
…
Different summary tables for hematology and biochemistry variables will be provided, based on data from scheduled visits: observed values and change from Baseline, from Baseline to maximum post-Baseline value, from Baseline to minimum post-Baseline value, shift from Baseline to end of treatment, shift from Baseline to end of double blind treatment, and markedly abnormal laboratory data.
End of treatment will be defined as the Week 48 visit or the early termination assessment depending if the subject discontinued early or not.
…
Table 13‒1: Laboratory measurements
Hematology Biochemistry Urinalysis
Basophils Calcium Albumin
Eosinophils Chloride Bacteria
Lymphocytes Magnesium Crystals
Atypical lymphocytes Potassium Glucose
Monocytes Sodium pH
REDACTED terieri
ologolog
COPYeliel
there thereential wential
ing Iing
This do
cumenEoEoe
ntBasoasoen
umuuu
cann
ot atologtologt
ophhca
be 1:
bus
ed
LL
to jo su
pport
be defbe deppject dect ds
anyhift fift y mark
eting
gy andgy and
ed valud valuom Baom Bfrof
autho
rizati
onwithwith
IMP.MP.ap
plica
tion
t od after d after
ical caucal caWeek 4Week
has beenas beh infh inf
andring ping p
or irrr irr
any
o
exten
sionsbloodlood
reactiveactivhemogemog
p; WB; WB
or ddor va
riatio
ns
oniononoari
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 166 of 183
Table 13‒1: Laboratory measurements
Hematology Biochemistry Urinalysis
Neutrophils Glucose RBC
Hematocrit BUN WBC
Hemoglobin Creatinine Urine dipstick for pregnancy testingb
MCH hs-CRP/CRPa
MCHC AST
MCV ALT
Platelet count GGT
RBC count ALP
WBC count Total bilirubin
LDH
Total cholesterol
Albumin
Serum pregnancy testingb
ALP=alkaline phosphatase; ALT=alanine aminotransferase; AST=aspartate aminotransferase; BUN=blood urea nitrogen; CRP=C-reactive protein; GGT=gamma glutamyltransferase; hs-CRP=high sensitivity C-reactive protein; LDH=lactate dehydrogenase; MCH=mean corpuscular hemoglobin; MCHC=mean corpuscular hemoglobin concentration; MCV=mean corpuscular volume; RBC=red blood cell; SFU=Safety Follow-up; WBC=white bloodcella Both CRP and hs-CRP will be tested at specified visits per Protocol Table 5.1.b A serum pregnancy test will be performed at Screening for all women of childbearing potential. This recommendation also applies to women of childbearing potential with infrequent or irregular menstrual cycles and women during menopause. Natural menopause is recognized to have occurred after 12 consecutive months of amenorrhea, for which there is no other obvious pathological or physiological cause (International Menopause Society, 2015). A urine pregnancy test is also required at the Baseline, Week 48, ET, and at SFU visits. A urine pregnancy test will also be performed at any study visit where there has been a delay in menses. This recommendation also applies to women of childbearing potential with infrequent or irregular menstrual cycles. Pregnancy test results must be negative prior to administering IMP.
…
Table 10-4 updated to align with Table 12-4 from the Protocol Amendment 2.
Table 10-5 updated to align with Table 12-5 from the Protocol Amendment 2.
Section 12.2 Calculation rules for duration of adverse events
The calculation rules for duration of AEs are presented in Table 12‒2. AE duration is computed and reported in day.
REDACTtransferansfmma mma glug
corpuscorpuumum
CTED nancy nancyDCT
COPY OP
C
This do
cumen
tTablabl
SeS
cann
ot e 1010 44
be us
edlso apo apesults sults
top
o be o be su
pport
womomtural mural m
e is no ois noegnanegnan
anyformeorme
men omen
marketi
ng scularcular
me; RBCe; RBC
specifspecid
autho
rierase; Arase; A
utamylamylori
zatio
n testintestinti
o
ori
appli
catio
n iiionatilicap
and n
any
any ex
tensio
ns
nsns
ex
or rva
riatio
ns
ony ionono
v
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 167 of 183
Table 13‒2: Calculation rules for duration of adverse events
Data Availability
Onset Date
Outcome Date
Calculation Rules
Complete data D1 D2 Duration = D2 – D1 + 1
Start date missing - D2 Duration = < D2 – D0Where, for a subject in the FAS, D0 is the date of the first intake of IMP, and for screen failures, D0 is the date of the Screening visit day
End date missing D1 - Duration = > Discharge day – D1For resolved and ongoing AE DurationWhere discharge refers to the date of Visit 16 (Week48) or date of discontinuation
Start and end date missing
- - Duration = > Discharge day – D0For resolved and ongoing AE Duration Where, for a subjects in the FAS, D0 is the date of the first intake of IMP, and for screen failures, D0 is the date of the Screening visit day,Discharge refers to the date of Visit 16 (Week 48) ordate of discontinuation.
Has been changed toThe calculation rules for duration of AEs are presented in Table 12‒2. AE duration is computed and reported in day.
Table 13‒3: Calculation rules for duration of adverse events
Data Availability
Onset Date
Outcome Date
Calculation Rules
Complete data D1 D2 Duration = D2 – D1 + 1
Start date missing - D2 Duration = < D2 – D0 + 1Where, for a subject in the SSFAS, D0 is the date of the first intake of IMP, and for screen failures, D0 is the date of the Screening visit day
End date missing D1 - Duration = > Final contact dateDischarge day – D1 +1For resolved and ongoing AE DurationWhere discharge refers to the date of Visit 16 (Week48) or date of discontinuation
REDACTEDdatedate
TTTTE
AEs aAEs a
COPYf the Sf the schargcharg
This do
cumen
t
um
cann
ot ate mite miotbe
used
ed
to -- supp
ort
posu
any OuOu
D
ya
marketi
ng are re
s for s for
autho
rizati
onge refe ref
of discf disciz
e prpr
appli
catio
noingingects in tcts in
MP, anMP, aScreenincreen
ferf
and day ay ––
g Ag A
anDDany
an
exten
sions
onone of V of V
s or ova
riatio
ns
the he s the the
onthe
reof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 168 of 183
Table 13‒3: Calculation rules for duration of adverse events
Data Availability
Onset Date
Outcome Date
Calculation Rules
Start and end date missing
- - Duration = > Final contact date Discharge day – D0 +1For resolved and ongoing AE Duration Where, for a subjects in the SSFAS, D0 is the date of the first intake of IMP, and for screen failures, D0 is the date of the Screening visit day,Discharge refers to the date of Visit 16 (Week 48) ordate of discontinuation.
REDACTED COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y exte
nsi(We(Weon
s WeekWeeio
xt
orD0D va
riatio
nate of e of D0 iD0
ons
onthe
reof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 169 of 183
13.5 Amendment 513.5.1 Rationale for the amendmentThe main objectives are to describe the different presentation of the safety assessments.
13.5.2 Modification and changesText added for Section 4.2.1 Handling of missing data for efficacy analysis
The imputation model will be applied for each treatment group separately. However, in the event there are computational challenges with the imputation model (eg, due to a standard deviation of 0 for responses of a given imputation), it is acceptable to modify the imputation model to include treatment as a variable in the model rather than running a separate model for each treatment group. It should be noted that doing so assumes that treatment does not interact with any of the other variables in the imputation model.
Section 6.4 Prior and concomitant medicationsPrior medications include any medications that started and ended prior to the start date of study medication. Concomitant medications are medications taken at least one day in common with the study medication dosing period. For bimekizumab, the dosing period is typically from the date of first dose up to (but not including) one dosing interval post last dose. Thus, a concomitant medication is any medication whose start date is on or after the date of first study medication and prior to the date of last study medication administration + 28 days (1 dosing interval), and whose stop date is either missing, or on or after the date of first study medication administration.
…
Has been changed toPrior medications include any medications that started and ended prior to the start date of study medication. Concomitant medications are medications taken at least one day in common with the study medication dosing period. For bimekizumab, the dosing period is typically from the date of first dose up to (but not including) one dosing interval post last dose. Thus, a concomitant medication is any medication whose start date is on or after the date of first study medication and prior to the date of last study medication administration + 28 days (1 dosing interval), or whose stop date is either missing, or on or after the date of first study medication administration.
…
Section 7 Measurements of treatment compliance
…
A summary of percent treatment compliance categorized as ≤80% and >80% will be provided by treatment group for the overall treatment period as well as for Double-Blind treatment period.
A by-subject listing of treatment compliance will be provided.
Has been changed to
…
A summary of percent treatment compliance categorized as ≤80% and >80% will be provided by treatment group for the overall treatment period as well as for Double-Blind treatment period.
date odate
atiat
COPYpp
n or afn or astratiotratio
off
This do
cumen
treatmeatm
A bA
cann
ot ummarmmatmetme
be Mea us
ed
asurasu
to su
ppor
ion won wstudtudy my
sing, sing,
anying) oing)
whwh
marketi
ng
tions thons thons are ns are
For bimor b
autho
rizati
onon + on + f first f first
appli
catio
nast oast operiperiodo
last doast doter theer th
282
andior to or to
onon
any
h
exten
sions
impmpns
eparaeparasiothat trhat trenl. e
or andanormpupu
varia
tions
in then thearndanda
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 170 of 183
For the Double-Blind treatment period compliance will refer to the first 12 weeks and will be presented by treatment received at baseline. For the overall treatment period, the compliance will be calculated for the following three groups: bimekizumab 160mg and 160mg with loading dose, bimekizumab 320mg, and all bimekizumab. Treatment compliance for the bimekizumab 160mg and 160mg with loading dose group will be calculated for the time the subject receives 160mg or 160mg with loading dose, eg for a subject who switches from Placebo or bimekizumab 16mg to bimekizumab 160mg at Week 12, the compliance will only be calculated for the time the subject receives bimekizumab 160mg. The same approach will be done for bimekizumab 320mg. The all bimekizumab group will consist of all the doses of bimekizumab including bimekizumab 16mg, and will only exclude the time subjects receive Placebo.A by-subject listing of treatment compliance will be provided, presenting percent compliance and numbers of expected and received doses for the Double-Blind Period and for treatment with any dose of bimekizumab.
Section 9.1 Pharmacokinetics
Bimekizumab plasma concentrations will be summarized for each treatment at each scheduled visit using the PK-PPS analysis set.
…
Has been changed toBimekizumab plasma concentrations will be summarized for each treatment at each scheduled visit using the PK-PPS, analysis set and using a subset of the DBS which includes subjects that belong to both analysis sets, the DBS and the PK-PPS.
…
Section 9.2 Pharmacodynamics and Immunogenicity
The biomarker data will be separated between LGC assays, mulitplex proinflammatory cytokine and chemokine analysis, flow cytometry - TBNK panel, and flow cytometry -Th1/Th2/Th17/Th22 panel. Biomarker variables will be summarized and listed for each treatment at each scheduled visit using the PD-PPS analysis set.
…
In addition the time point of the first occurrence of AbAb positivity during the treatment period (excluding Baseline and pre-treatment) will be summarized for each treatment group.
…
Has been changed to
The biomarker data will be separated between LGC assays, mulitplex proinflammatory cytokine and chemokine analysis, flow cytometry - TBNK panel, and flow cytometry -Th1/Th2/Th17/Th22 panel. Biomarker variables will be summarized and listed for each treatment at each scheduled visit using the PD-PPS, analysis set and using a subset of the DBS which includes subjects that belong to both analysis sets, the DBS and the PK-PPS.…
REDEDACTED e summe summ
usingusingAC
DBSDBSED
COPY
This do
cumen
t The Thea
cann
ot
s beebee
beBasBasus
ed
e timetimeselsel
to su
pportw w
el. Biol. Bioduled vduled
anyaratearat
cytocytomark
eting
nd Imnd Im
tedted
autho
rizati
on
marizmarizg a suba su
thandanda
appli
catio
nh treah tre and a
ny ex
tensio
ns ans
cent ccent cted andd ande
orrnd wndva
riatio
ns
ek k tioab b arimab ab v
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 171 of 183
In addition, the time point of the first occurrence of AbAb positivity during the Double-Blind Period, and the entire treatment period (excluding Baseline and pre-treatment) will be summarized for each treatment group.
…
Section 10.1 Extent of exposureThe duration of exposure (in days) will be calculated as:
(16)
28 days refer to one half-life of bimekizumab.
For subjects who have died the exposure will be as follows:
(17)
The sum over all duration of exposure (in years) will be calculated for the total subject-years of duration of exposure.
The duration of exposure will be summarized by treatment group using descriptive statistics. A by-subject listing of date of first and last dose and the duration of exposure will be performed.
Time at risk (in days) is defined as:
(18)
where 140 days refers to 5*half-life of bimekizumab.
The sum over all time at risk (in years) will be calculated for the total subject-years of time at risk.
The days on treatment and days on bimekizumab treatment will be calculated as follows:
(19)
(20)
Has been changed to
The duration of exposure and time at risk will be summarized for the Double-Blind Periodand the entire treatment period. For the entire treatment period the duration of exposure and time at risk will be calculated for bimekizumab 160mg and 160mg with loading dose, bimekizumab 320mg, and all bimekizumab. The calculation of exposure duration and time at risk for the all bimekizumab group is different from that of the other two groups. For all bimekizumab the duration of exposure and time at risk will include the time a subject
REDACTED
of bof b
COPYatmeatmed the dd the
This do
cumenHas Has
ThThm
cann
ot
bb
be us
ed to
supp
ort an
y ys on ys onmark
etingimekiimek
ars) wilrs) w
autho
rizati
on durauraap
plica
tioned ford fo
t groupgrourationatio
and
r
any e
xtens
ions o
r(16(16varia
tions
6))
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 172 of 183
received any dose of bimekizumab (including 16mg) while for the other two treatment groups only the time a subject received 160mg, 160mg with loading dose, or 320mg will be included into the calculation.
Duration of exposure Double-Blind Period
The duration of exposure (in days) during the Double-Blind Period will be calculated as:
(16)
28 days refer to one half-life of bimekizumab.
Note: If the date of last injection (Double-Blind Period) + 28 extends to a date beyond the date of first injection (Dose-Blind Period), then this calculation reverts to
(17)
For subjects who die during the Double-Blind Period, then this calculation reverts to:
(18)
Duration of exposure entire treatment period
For subjects who do not switch study treatments, who receive bimekizumab 160mg with loading dose at Baseline, or who receive bimekizumab 16mg and will be summarized under all bimekizumab group:
(19)
Note: If the date of last injection +28 extends to a date beyond the final visit date (not including SFU), then this calculation reverts to:
(20)
For subjects who die, then this calculation reverts to:
(21)
For subjects who receive Placebo or bimekizumab 16mg in the Double-Blind Period and will be summarized under their Dose-Blind treatment in the overall table group
REDACTED
ment ment R
COPYriodriodPY
This do
cumen
t ForFom
cann
ot be
used
e of ofdU), thU), thus
to f laf lasu
pport
any m
arketi
ngperiperingudy tredy trrko receivrecem
autho
rizati
on
ioi
appli
catio
n
then then ap
p
and a
ny toto
ny ex
tensio
ns
date date exo
or va
riatio
ns
6)6)
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 173 of 183
bimekizumab 160mg or 320mg, or subjects who received Placebo and will be summarized in the overall table under their Dose-Blind treatment:
(22)
Note: If the date of last dose +28 extends to a date beyond the final visit date (not including SFU), then this calculation reverts to:
(23)
For subjects who die during the Dose-Blind Period, then this calculation reverts to:
(24)
The sum over all duration of exposure (in years) will be calculated for the total subject-years of duration of exposure.
The duration of exposure will be summarized by treatment group using descriptive statistics. A by-subject listing of date of first and last dose in Double-Blind Period, the date of first and last dose in Dose-Blind Period and the duration of exposure during Double-Blind Period and under bimekizumab treatment (any dose of bimekizumab) will be performed.
Time at risk Double-Blind Period
For subjects who complete the final visit of the Double-Blind Period and continue to the Dose-Blind Period:
(25)
For subjects who discontinue on or prior to the final visit of the Double-Blind Period, use the minimum of the following:
(26)
(27)
(28)
where 140 days refers to 5*half-life of bimekizumab.
REDACTED ed by ted by
dose dose ihe durhe du
ny dony doR
COPY ill be ill be
This do
cumen
t can
notmebe us
edwho dho dsem of tof u
to su
pport
any
fify mark
etingose ofose oing
final inalm
autho
rizati
on
treatmreatmin Don Dothation ation
ff
appli
catio
n
calculaalcul
and a
nyion ron an
y exten
sions
or va
riatio
ns
ding ing ria
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 174 of 183
For subjects who die during the Double-Blind Period, then this calculation reverts to:
(29)
Time at risk entire treatment period
For subjects who do not switch study treatments, who receive bimekizumab 160mg with loading dose at Baseline, or who receive bimekizumab 16mg and will be summarized under all bimekizumab group:
For subjects who complete the Dose-Blind Period and enter the extension study:
(30)
For subjects who die prior to the final visit:
(31)
For all other subjects, use the minimum of the following:
(32)
(33)
For subjects who receive Placebo or bimekizumab 16mg in the Double-Blind Period and will be summarized under their Dose-Blind treatment in the overall table withbimekizumab 160mg or 320mg, or for subjects who receive Placebo and will besummarized in the overall table under their Dose-Blind treatment:
For subjects who complete the Dose-Blind Period and enter the extension study:
(34)
For subjects who die during the Dose-Blind Period:
(35)
For all other subjects, use the minimum of the following:
(36)
(37)
The sum over all time at risk (in years) will be calculated for the total subject-years of time at risk.
REDACTED ekizumekizum
TEBlindBlind
ACor or subsuED
ndernderR
COPY
This do
cumen
t
othethot
cann
ot be
her sher s
used
to su
pport
duringdurinup
any m
arketi
ng theithei
ingsee--BlinBlrk
autho
rizati
onmab 1mab 1riztreattreatthbjectsjectsa
appli
catio
n and
any e
xtens
ions o
r vari
ation
s der er
atithe
reof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 175 of 183
The days on treatment and days on bimekizumab treatment will be calculated as follows:
(38)
(39)
Section 10.2 Adverse events
…
The incidence of TEAEs will be summarized by MedDRA SOC, high level term, and PT. Tables with incidences of classified TEAEs by maximum intensity, by relationship, and by subject number will be provided. The incidence of non-serious TEAEs above the reporting threshold of 5% of subjects and relationship will be summarized. Furthermore the incidence of all TEAEs, serious TEAEs, non-serious TEAEs, and TEAEs leading to study discontinuation and/or permanent withdrawal of study medication will be summarized. In addition an overall summary table will be provided.
Exposure adjusted incidence rate (EAIR) and exposure adjusted event rate (EAER) will be calculated for following tables in the final analysis: all TEAEs, serious TEAEs, TEAEs leading to study discontinuation and/or permanent withdrawal, adverse drug reactions, fungal infectious disorder TEAEs, opportunistic infection (including tuberculosis) TEAEs, malignant or unspecified tumor TEAEs, malignant tumor TEAEs, major cardiovascular event TEAEs, haematopoietic cytopenias TEAEs, neuropsychiatric events TEAEs, inflammatory bowel disease TEAEs, hypersensitivity and anaphylactic Reaction TEAEs, hepatic events TEAEs, and TEAEs by timing of onset relative to AbAb Status.
Has been changed to
…
The incidence of TEAEs will be summarized by MedDRA SOC, high level term, and PT. Tables with incidences of classified TEAEs by maximum intensity, by relationship, and by subject number will be provided. The incidence of non TEAEs, non-serious TEAEs above the reporting threshold of 5% of subjects and relationship will be summarized. Furthermore the incidence of all TEAEs, serious TEAEs, non-serious TEAEs, and TEAEs leading to study discontinuation and/or permanent withdrawal of study medication, adverse drug reactions, fungal infectious disorder TEAEs, opportunistic infection (including tuberculosis) TEAEs, malignant or unspecified tumor TEAEs, malignant tumor TEAEs, major cardiovascular event TEAEs, haematopoietic cytopenias TEAEs, neuropsychiatric events TEAEs, inflammatory bowel disease TEAEs, hypersensitivity and anaphylactic Reaction TEAEs, hepatic events TEAEs, and TEAEs by timing of onset relative to AbAb Status will be summarized. In addition, an overall summary table will be provided.
REDACTEDlysislysiwithdrwithdr
(inclu(inclt tumotumo
neurneur
COPY posureposureis:is:
This do
cumen
tinclnclntTEATEm
cann
ottg to s to snoverse derse dcaludlu
beng
eore tore bus
edder wilwilseg thrthru
toencencetolsu
pport
AEs wAEs wsues
any m
arketi
ngor
ropsycropsylactic Ractic
Ab Statb Sta
autho
rizati
one adjadj
all TEall Tdrawalrawal
udingudingTEATEA
appli
catio
ny discy disced. In aed. In
jusjus
andthe he
e ince incan
yip, anip, ae repre
exten
sions
erm, erm,
or va
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 176 of 183
The tables will be split into the Double-Blind Period and the complete treatment period (exception non TEAE table and TEAEs by timing of onset relative to AbAb Status). Presentations for the Double-Blind Period will summarize AEs that start prior to or at Visit 7 (see details given above) by treatment group as randomized for the Double-Blind Period. Presentations for the complete treatment period will only summarize AEs that occur under treatment with bimekizumab, irrespective of the treatment period. For summaries of the bimekizumab 160mg, 160mg with loading dose and bimekizumab 320mg treatment groups, patients randomized to Placebo or bimekizumab 16mg at Baseline will only be included with AEs that start in the Dose-Blind Period. The all bimekizumab treatment group will also present AEs occurring under bimekizumab 16mg treatment in the Double-Blind Period. AEs for each subject will be summarized based on the treatment actually received at the onset of each particular AE.
Exposure adjusted incidence rate (EAIR) and exposure adjusted event rate (EAER) will only be calculated for the complete treatment period for following tables in the final analysis: all TEAEs, serious TEAEs, TEAEs leading to study discontinuation and/or permanent withdrawal, adverse drug reactions, fungal infectious disorder TEAEs, opportunistic infection (including tuberculosis) TEAEs, malignant or unspecified tumor TEAEs, malignant tumor TEAEs, major cardiovascular event TEAEs, haematopoietic cytopenias TEAEs, neuropsychiatric events TEAEs, inflammatory bowel disease TEAEs, hypersensitivity and anaphylactic Reaction TEAEs, hepatic events TEAEs, and TEAEs by timing of onset relative to AbAb Status.
Section 10.2.1 Exposure duration
The duration of each AE will be calculated as follows:
+ 1 (21)
For the calculation of duration for AEs with missing start dates, the dates will be imputed as described in Section 4.2.2.
The time to first dose for each AE will be calculated as follows for all TEAEs:
(22)
The time to most recent dose for each AE will be calculated as follows for all TEAEs:
(23)
Time to first dose and time to most recent dose will not be calculated for pretreatment AEs.
For the calculation of time to first bimekizumab and time to most recent bimekizumab dose the same formulas as above (22 and 23) will be used but the date of first bimekizumab dose will be used instead of the date of first dose. The time to first dose and time to first bimekizumab dose differs only for those subjects who receive placebo at Baseline.
REDACTED
ated aated a
COPYEs, iEs, PY
AEs, hAEsCO
This do
cumen
t TT
cann
ot be
m used
most rost
to su
pport
ach Ach any
AEA
marketi
ng
r AEs wAEs
autho
rizati
on he
n
s follfoll
pappli
catio
nignangnantioevent Tvent
icaflammlammppepatiepatia
andontinti
nd
s disdisaan
ybles iles an
yinuinu
exten
sions
meshe treahe treaio
ate (ate (ei
orent entva
riatio
ns
mg g tiwill ill ari
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 177 of 183
Has been changed to
The duration of each AE will be calculated as follows:
+ 1 (4021)
For the calculation of duration for AEs with missing start dates, the dates will be imputed as described in Section 4.2.2.
The time to first dose for each AE will be calculated as follows for all TEAEs:
(4122)
The time to most recent dose for each AE will be calculated as follows for all TEAEs:
(4223)
Time to first dose and time to most recent dose will not be calculated for pretreatment AEs.
For the calculation of time to first bimekizumab and time to most recent bimekizumab dose the same formulas as above (4122 and 4223) will be used but the date of first bimekizumab dose will be used instead of the date of first dose. The time to first dose and time to first bimekizumab dose differs only for those subjects who receive placebo at Baseline.
Section 10.2.2 Exposure adjusted incidence rate (EAIR) and exposure adjusted event rate (EAER)
The EAIR is defined as the number of subjects (nAE) with a specific AE adjusted for the exposure and will be scaled to 100 patient-years:
(24)
where is a subject’s exposure time in years up to the first occurrence of the AE of interest (equation [21] in years) at the level of coding evaluated, nnoAE the number of subjects without the specific AE and the total time at risk scaled to 100 patient-years (equation [18] in years).
If a subject has multiple events at the level of coding evaluated, the time of exposure is calculated to the first occurrence of the AE of interest. If a subject has no events, the total time at risk (18) in years is used.
Exact Poisson 95% CIs for incidence rates are calculated using the relationship between the Poisson and the Chi-square distribution (Ulm, 1990; Fay and Feuer, 1997):
(25)
be time time
eceiveeceive
ncidennciden
COPY ot t
nd timnd tiusedused
This do
cumen
talccrisk (risk (
E
cann
ot AE
ubjectubjectculatulat
be1]
E andE anus
eda subsub] in yin
to jjsu
pport
any m
arketi
ng nc
of subof sub0 pati0 pat
autho
rizati
on me
d but d but to firsto firs
e place plac
ce rae ra
appli
catio
n
calculaalcula
to mto m
and a
ny ex
tensio
ns
((44
l TEATEA
or va
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 178 of 183
(26)
(27)
(28)
where nAE is the number of subjects with a specific AE for the incidence rate of interest and is the basis for the number of the degrees of freedom for the chi-square quantile for the upper tail probability χ2, is a subject’s exposure time in years up to the first occurrence of the AE of interest (equation [21] in years) at the level of coding evaluated, nnoAE the number of subjects without the specific AE and the total time at risk scaled to 100 patient-years (equation [18] in years).
The EAER will be the number of AEs including repeat occurrences in individual subjects divided by the total time at risk scaled to 100 patient-years and calculated using:
(29)
where NAE is the total number of AEs, is a subject’s total time at risk in years and nAll thenumber of subjects.
No CI will be computed for EAER.
Has been changed to
The EAIR is defined as the number of subjects (nAE) with a specific AE adjusted for the exposure and will be scaled to 100 patient-years:
(4324)
where is a subject’s exposure time in years up to the first occurrence of the AE of interest (equation [4021] in years) at the level of coding evaluated, nnoAE the number of subjects without the specific AE and the total time at risk scaled to 100 patient-years (equation [18] in yearsSection 10.1).
If a subject has multiple events at the level of coding evaluated, the time of exposure is calculated to the first occurrence of the AE of interest. If a subject has no events, the total time at risk (18 Section 10.1) in years is used. As indicated above, exposure-adjusted incidence rates will only be calculated for the overall treatment period. These presentations do not include
REDACTED E
ss
COPY at oat ntnt--yeayea
This do
cumen
t(equa(equathethe
cann
ot ere re
be us
ed to
e s supp
ort
s the nuthe scacalele
anyER.ER mark
eting
RR
autho
rizati
on ar
iza
appli
catio
n 10000
currencurrenrs ands and
andE the the
0 pa0 pa
anyccurrccur
nu
exten
sions
of intef intee forfor
or(28(28varia
tions
8)
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 179 of 183
AEs that occur under Placebo treatment. Therefore, a subject’s exposure time will only start at the first dose of bimekizumab in the Dose-Blind Period for subjects randomized to Placebo at Baseline. Also, for subject’s randomized to bimekizumab 16mg at Baseline, exposure time will only be considered from the start of Dose-Blind Period, when presenting AEs for the bimekizumab 160mg and bimekizumab160mg with loading dose, or bimekizumab 320mg groups. All exposure time on any bimekizumab dose is considered for the all bimekizumab column.Exact Poisson 95% CIs for incidence rates are calculated using the relationship between the Poisson and the Chi-square distribution (Ulm, 1990; Fay and Feuer, 1997):
(4425)
(4526)
(4627)
(4728)
where nAE is the number of subjects with a specific AE for the incidence rate of interest and is the basis for the number of the degrees of freedom for the chi-square quantile for the upper tail probability χ2, is a subject’s exposure time in years up to the first occurrence of the AE of interest (equation [4021] in years) at the level of coding evaluated, nnoAE the number of subjects without the specific AE and the total time at risk scaled to 100 patient-years (equation [18] in years Section 10.1).
The EAER will be the number of AEs including repeat occurrences in individual subjects divided by the total time at risk scaled to 100 patient-years and calculated using:
(4829)
where NAE is the total number of AEs, is a subject’s total time at risk in years and nAll thenumber of subjects.
No CI will be computed for EAER.
Section 10.3 Clinical laboratory evaluations
…
REDACTED ED
withwith
COPY
This do
cumen
twherwhernunu
cann
ot be
used
tal al tohe
l timtimsu
pport
1).).
e numnum
anyars) aars) mark
etingth a sph a sp
es of fres of s exposexpo
at
autho
rizati
on
za
appli
catio
n at
and a
ny ex
tensio
ns or
varia
tions
e
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 180 of 183
Different summary tables for hematology and biochemistry variables will be provided, based on data from scheduled visits: observed values and change from Baseline, from Baseline to maximum post-Baseline value, from Baseline to minimum post-Baseline value, shift from Baseline to end of treatment, and markedly abnormal laboratory data.
End of treatment will be defined as the Week 48 visit or the early termination assessment depending if the subject discontinued early or not.
…
Table 13‒1: Laboratory measurements
Hematology Biochemistry Urinalysis
Basophils Calcium Bacteria
Eosinophils Chloride Crystals
Lymphocytes Magnesium Glucose
Monocytes Potassium pH
Neutrophils Sodium RBC
Hematocrit Glucose WBC
Hemoglobin BUN Urine dipstick for pregnancy testingb
MCH Creatinine
MCHC hs-CRP/CRPa
MCV AST
Platelet count ALT
RBC count GGT
WBC count ALP
Total bilirubin
LDH
Total cholesterol
Albumin
Serum pregnancy testingb
REDACTED
nene CTE
CRP/CCRP/CDA
R
COPY YCO
This do
cumen
t can
notot
anbe
e used
ed
u
to su
pport
tpp
s
anyGGGanaaaamark
eting
CRPRPng
ALTALT rkm
autho
rizati
on n
ori
PPaa ut
appli
catio
n pHpHtitititioati
pla
anducosecosed
aan
y saex
tensio
ns
ns
ene
or va
riatio
ns th
ereof.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 181 of 183
Table 13‒1: Laboratory measurements
Hematology Biochemistry Urinalysis
ALP=alkaline phosphatase; ALT=alanine aminotransferase; AST=aspartate aminotransferase; BUN=blood urea nitrogen; CRP=C-reactive protein; GGT=gamma glutamyltransferase; hs-CRP=high sensitivity C-reactive protein; LDH=lactate dehydrogenase; MCH=mean corpuscular hemoglobin; MCHC=mean corpuscular hemoglobin concentration; MCV=mean corpuscular volume; RBC=red blood cell; SFU=Safety Follow-up; WBC=white blood cella Both CRP and hs-CRP will be tested at specified visits per Protocol Table 5.1.b A serum pregnancy test will be performed at Screening for all women of childbearing potential. This recommendation also applies to women of childbearing potential with infrequent or irregular menstrual cycles and women during menopause. Natural menopause is recognized to have occurred after 12 consecutive months of amenorrhea, for which there is no other obvious pathological or physiological cause (International Menopause Society, 2015). A urine pregnancy test is also required at the Baseline, Week 48, ET, and at SFU visits. A urine pregnancy test will also be performed at any study visit where there has been a delay in menses. This recommendation also applies to women of childbearing potential with infrequent or irregular menstrual cycles. Pregnancy test results must be negative prior to administering IMP.
Has been changed to
…
Different summary tables for hematology and biochemistry variables will be provided, based on data from scheduled visits: observed values and change from Baseline, shift from Baseline to maximum post-Baseline value (in Double-Blind Period), shift from Baseline to maximum post-Baseline value (in Dose-Blind Period), shift from Baseline to minimum post-Baseline value (in Double-Blind Period), shift from Baseline to minimum post-Baseline value (in Dose-Blind Period), shift from Baseline to end of treatment (in Double-Blind Period), shift from Baseline to end of treatment (in Dose-Blind Period), and markedly abnormal laboratory data.
End of treatment (in Double-Blind Period) will be defined as the last visit in the Double-Blind Period or the early termination assessment depending if the subject discontinued in the Double-Blind Period.
End of treatment (in Dose-Blind Period) will be defined as the Week 48 visit or the early termination assessment depending if the subject discontinued early or not.
…
Table 13‒1: Laboratory measurements
Hematology Biochemistry Urinalysis
Basophils Calcium Bacteria
Eosinophils Chloride Crystals
Lymphocytes Magnesium Glucose
Monocytes Potassium pH
Neutrophils Sodium RBC
REDACTEDd c
lind Plind TEodd),), s
ACfromfromEDeline eline
ii
COPY chemihem
chanchan
This do
cumen
tBasBasn
tnt
umumumum
cann
ot 3‒‒11
matomatono
be
11::us
ed sm tomentmen su
pport
ererrtd..p
DoseDose Bst d
anyBlinBli
nyrminminamark
eting
to ento en DoseDoke
ndnd
autho
rizati
on istr
nge froge froPerioPeriorizhift hift ff
thom BaseBaseauen
appli
catio
n
try vary va
and
rrere anymensemens
gulargul
exten
sions
rual cyrual cy
ve mone mononal Mnal M
SFU SFU
or va
riatio
ns
od d
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 182 of 183
Table 13‒1: Laboratory measurements
Hematology Biochemistry Urinalysis
Hematocrit Glucose WBC
Hemoglobin BUN Urine dipstick for pregnancy testingb
MCH Creatinine
MCHC hs-CRP/CRPa
MCV AST
Platelet count ALT
RBC count GGT
WBC count ALP
Total bilirubin
LDH
Total cholesterol
Uric acid
Albumin
Serum pregnancy testingb
ALP=alkaline phosphatase; ALT=alanine aminotransferase; AST=aspartate aminotransferase; BUN=blood urea nitrogen; CRP=C-reactive protein; GGT=gamma glutamyltransferase; hs-CRP=high sensitivity C-reactive protein; LDH=lactate dehydrogenase; MCH=mean corpuscular hemoglobin; MCHC=mean corpuscular hemoglobin concentration; MCV=mean corpuscular volume; RBC=red blood cell; SFU=Safety Follow-up; WBC=white blood cella Both CRP and hs-CRP will be tested at specified visits per Protocol Table 5.1.b A serum pregnancy test will be performed at Screening for all women of childbearing potential. This recommendation also applies to women of childbearing potential with infrequent or irregular menstrual cycles and women during menopause. Natural menopause is recognized to have occurred after 12 consecutive months of amenorrhea, for which there is no other obvious pathological or physiological cause (International Menopause Society, 2015). A urine pregnancy test is also required at the Baseline, Week 48, ET, and at SFU visits. A urine pregnancy test will also be performed at any study visit where there has been a delay in menses. This recommendation also applies to women of childbearing potential with infrequent or irregular menstrual cycles. Pregnancy test results must be negative prior to administering IMP.
minotrminotrammamm
DACTED DpregnaregnCTDA
COPY OP
C
This do
cumen
t can
not tese be
ion aon st resst re
used
rine pne pill alsoll alsalsoals
tothertherprpr
supp
orte perfperfs to womto wo
e. Natur Nature ise
any ed at sed at
f
marketi
ng ansfnsf
ma glutaa glutaorpuscuorpusc
volume; olume
ausferaferuth
oriza
tion
tio
ancy tency teori
ut
appli
catio
n iiionatilicap
and n
any
any ex
tensio
ns
nsns
ex
or rva
riatio
ns
onia
v
thereo
f.
UCB 07 Sep 2018Statistical Analysis Plan Bimekizumab PA0008
Confidential Page 183 of 183
STATISTICAL ANALYSIS PLAN SIGNATURE PAGEThis document has been reviewed and approved per the Review and Approval of Clinical Documents Standard Operating Procedures. Signatures indicate that the final version of the Statistical Analysis Plan (SAP) or amended SAP is released for execution.
COPY
This do
cumen
t can
not b
e use
d to s
uppo
rt any
mark
eting
autho
rizati
on ap
plica
tion a
nd an
y exte
nsion
s or v
ariati
ons t
hereo
f.