challenges for the pathologist in the era of personalized medicine prof keith m kerr department of...

Challenges for the Pathologistin the era of

Personalized Medicine

Prof Keith M KerrDepartment of Pathology

Aberdeen University Medical SchoolAberdeen, UK

Disclosures

Acted as Speaker and/or Consultant for:AstraZeneca, Roche, Boehringer Ingeheim,

Bristol-Myers-Squibb, Clovis, Eli Lilly, MSD, Novartis, Pfizer

Personalised approach to treating cancerUnderstanding the molecular basis of cancer has identified specific

drivers and allowed for sub-classification of the disease, revealing the potential for targeted agents

‘ONE SIZE FITS ALL’

Drug X

PERSONALISED THERAPY

Drug X Y Z

• Personalised treatment consists of three essential components :– Oncogenic target that drives cancer growth

– Predictive biomarker that detects presence of the target

– Well conducted clinical studies that confirm treatment efficacy in the identified patient group

Diagnostic steps in lung cancer

Identification of Actionable genetic alterations

Morphological diagnosis of lung cancerSeparate SCLC from NSCLC

Subtype NSCLC where possible

Predictive ImmunohistochemistryOnly when needed

The Challenge: Subtyping non-small cell

carcinomas?• Squamous (Cis/Gem) vs Non-squamous (Cis/Pem)

cytotoxic chemotherapy• Contraindication of anti-angiogenic therapy for

squamous carcinomas• Triage for molecular testing

• Simple immunohistochemistry has solved this problem (TTF1, p63, p40, CK5/6)

• NSCLC-NOS rate should be <10%

It is worthwhile finding an Actionable genetic alteration in Lung cancer

Driver detected – Targeted Rx

Kris MG et al. JAMA 2014; 311, 1998-2006

Kerr KM. J Clin Pathol 2013;66:832–838 ROS1 fusion genes1.1-2.6% AdenocarcinomasFusion correlates with protein (IHC)Sensitive to crizotinib

NTRK1 fusionMPRIP-NTRK1 and CD74-NTRK13.3% of ‘onco-negative’ adenocarcinomasTrk inhibitors existVaishnavi A et al. Nat Med 2013; 19, 1469-72

CD74-NRG1 fusionSearch in ‘onco-negative’ adenocarcinomasERBB3 and PI3K-AKT pathway activationMucinous adenocarcinomasPotential therapeutic targetFernandez-Cuesta L et al. Can Disc 2014; jan30 epub

BRAF mutations2.5-4% AdenocarcinomasIncl V600E other V600 Smoking vs non-smokingVemurafenib

HER2 mutations1-2% AdenocarcinomasMutually exclusive of EGFR, KRASTraztuzamab?

MET upregulation4% amplification, ~50% overexpressionBiomarker issuesFailed trials

KRAS mutations25-35% AdenocarcinomasMEK inhibition?

RET fusion genes~1% AdenocarcinomasVandetanib & others

Clin Can Res 2012;18, 2443-51

FGFR1 amplificationBiomarker issuesDefinition of amplification20% may be overestimate?Ponatinib – FGFR1 inhibitorWynes MW et al. Clin Cancer Res 2014;20:3299-3309

Discoid Domain Receptor 2 mutationPrevalence 3.8%Good in vitro target – miRNA & DasatinibLimited clinical evidenceHammerman PS et al. Cancer Discov 2011

PI3Kinase~30% amplification~6% mutations – addictive??Inhibitors exist

IGFR1FigitumumabSome effect in squamousToxicity

EGFRTKI vs MoAbMutations – rarity (vIII – 8%)Targeting the receptor

METInhibitors existSo far no success

Panel of 54 genes with potentially druggable alterations

Govindan et al. Cell. 2012 Sep 14;150:1121-34

The Challenge: Lots of targets – Lots of drugs?

• Practical reality– EGFR mutation– ALK gene fusion

• ‘Nearly routine’?– BRAF mutation– ROS1 gene fusion

• Many more potential drugs with biomarkers

Methods of biomarker analysis

Transcription

Change in DNA sequence

Change in Gene copy #

mRNA transcript


Transcription



mRNA transcript

FISH/CISH

DNA mutational analysis

Microarray

RT-PCR


Transcription



mRNA transcript

FISH/CISH


Microarray

RT-PCR

Next-generation sequencing


Transcription



FISH/CISH


Microarray

RT-PCR


Protein

Translation

mRNA transcript


Transcription



FISH/CISH


Microarray

RT-PCR


Protein

Translation

mRNA transcript

Immunohistochemistry


Transcription



FISH/CISH


Microarray

RT-PCR


Translation

Immunohistochemistry

Biological ActivityOncogenesisDrug target

mRNA transcript

Protein

The Challenge:Complex testing landscape

• Biomarkers at different stages between gene and protein

• Complex testing strategy• Multiplex testing offers some solutions

• Can be confusing: EGFR• Sometimes unclear which approach is the best• Multiple biomarkers in the same ‘diagnostic space’

Methods of ALK biomarker analysis

Transcription



Translation

Biological ActivityOncogenesisDrug target

Break-apart FISH

IHC for ALK protein

PCR for ALK fusion gene transcripts

mRNA transcript

Protein

NGS

Prognostic and predictive biomarkers are used to guide treatment decisions in oncology

Provides information on outcome, independent of the administered therapy

Provides information on outcome with regards to

a specific therapy

Prognostic biomarkers may help define patient’s prognosis, risk of recurrence

or the duration of survival

Predictive biomarkers estimate response or survival of a specific patient on a specific therapy and can be a target

for therapy

Biomarkers for NSCLC

ERCC1Ki67/MIB1

p53

BRAFEGFRKRASMET

PD-L1

ALKHER2RET

ROS1

Prognostic Predictive

High levels if PD1 or PDL1 protein expression (IHC) may inhibitImmune response

Block PD1 or PDL1Immune damage to tumour

Chen, et al. Clin Cancer Res 2012

Biomarkers for Immunotherapy?Biomarkers for Immunotherapy?

PD-L1 Negative PD-L1 Positive (predictive of response)

Gandhi L, et al. AACR 2014. Abstract CT105.

Less response More response

?1% 5% 10% 50% cell positive

Intensity of staining?Immune cell staining?Several therapeutics

Several companion diagnostics…………

PD-L1 as a predictive immune biomarker: assays,sample collection and analysis in NSCLC studies

PembrolizumabMerck

• Prototype or clinical trial IHC assay (22C3 Ab)1,2

• Surface expression of PD-L1 on tumour specimen1,2

• Ph I: Fresh or archival tissue1,2

IHC Staining:• Strong vs weak expression1,2

• PD-L1 expression required for NSCLC for enrollment1

• Note that one arm of KEYNOTE 001 trial requires PD-L1- tumours1

Tumour PD-L1 expression:1,2 • ≥50% PD-L1+ cut-off:

32% (41/129)• 1–49% PD-L1+ cut-off: 36%

(46/129)

NivolumabBristol-Myers Squibb

• Dako automated IHC assay (28-8 Ab)3,4

• Surface expression of PD-L1 on tumour cells3,4

• Archival or fresh tissue3,4

IHC Staining:• Strong vs weak expression3,4

• Patients not restricted by PD-L1 status in 2nd- & 3rd-line

• Ph III 1st-line trial in PD-L1+5

Tumour PD-L1 expression:• 5% PD-L1+ cut-off: 59% (10/17)3

• 5% PD-L1+ cut-off: 49% (33/68)4

MPDL3280ARoche/Genentech

• Central laboratory IHC assay6

• Surface expression of PD-L1 on TILs or tumour cells6,7

• Archival or fresh tissue6

IHC Staining Intensity (0, 1, 2, 3):•IHC 3 (≥10% PD-L1+)6,7

•IHC 2,3 (≥5% PD-L1+)6,7

•IHC 1,2,3 (≥1% PD-L1+)6,7

•IHC 0,1,2,3 (all patients with evaluable status)6,7

•PD-L1 expression required for NSCLC for enrolment in Ph II trials6

•x

TIL PD-L1 expression:6

•IHC 3 (≥10% PD-L1+): 11% (6/53) •PD-L1 low (IHC 1, 0): 62% (33/53)

MEDI4736AstraZeneca

• Ventana automated IHC (BenchMark ULTRA using Ventana PD-L1 (SP263) clone)8,9

• Surface expression of PD-L1 on tumour cells8,9

• Unknown

IHC Staining Intensity:•Not presented to date8–10

Tumour PD-L1 expression (all doses):8

•PD-L1+: 34% (20/58)•PD-L1-: 50% (29/58)

†Definition of PD-L1 positivity differs between assay methodologies.1. Garon EB, et al. Presented at ESMO 2014 (abstr. LBA43); 2. Rizvi NA, et al. Presented at ASCO 2014 (abstr. 8007); 3. Gettinger S et al. Poster p38 presented at ASCO 2014 (abstr. 8024);

4. Brahmer JR et al. Poster 293 presented at ASCO 2014 (abstr. 8112^); 5. http://www.clinicaltrials.gov/ct2/show/NCT02041533 Accessed January 2015;6 . Rizvi NA et al. Poster presented at ASCO 2014 (abstr. TPS8123); 7. Soria J-C, et al. ESMO 2014 (abstr. 1322P); 8. Brahmer JR, et al. Poster presented at ASCO 2014 (abstr. 8021^);

9. Segal NH, et al. Presented at ASCO 2014 (abstr. 3002^); 10. Segal NH, et al. ESMO 2014 (abstr. 1058PD).

Ab, antibody; IHC, immunohistochemistry

http://www.clinicaltrials.gov/ct2/show/NCT02041533

The Challenge: Delivery• Reliable• Accurate• Timely• Relevant• Cost

The Challenge: Delivery• Reliable – will the test work?

– Pre-analytics– Formalin fixed-Paraffin embedded tissue (FFPE)– DNA for mutations

• DNA quality – fragmentation• Single test – 1.5% complete failure• Multiple tests - ?

– FISH <10%– RNA PCR >10% ?– IHC <5%

The Challenge: Delivery• Reliable – will the test work?• Accurate – is the test outcome correct?

– False positive tests – contamination– False negative test

• DNA quantity• DNA ‘purity’ – how much is from tumour

– Enough for multiplex testing?– FISH: 10-20% ALK tests challenged– RNA PCR: ~20% failed in recent ETOP study (ALK)– IHC: antibody specificity, non-specific staining– Biomarker Heterogeneity

The Challenge: Delivery• Reliable – will the test work?• Accurate – is the test outcome correct?• Timely - How quickly do you really need the results?

– More tests – more time– More complexity - more time– Communication Patient-Physician-Lab-Physician-Patient


• Relevant– Are the biomarker tests needed?– Has the lab performed the correct test?


• Relevant• Cost

– Technology is not cheap– Manpower is not cheap– Companion diagnostics are not cheap– Pathology and Therapy: budgets & reimbursement

Challenges for the Pathologistin the era of

Personalized Medicine

• Morphological assessment• Range of Biomarkers• Molecular diversity• Testing complexity• Sample limitations• Service delivery

challenges for the pathologist in the era of personalized medicine prof keith m kerr department of...

Documents

clin cancer res

molecular basis of cancer

squamous carcinomastriage

cancer discov 2011pi3kinase

potential drugs

nonsmall cell carcinomas

lung cancerthe challenge

lung cancerdriver