diagnostic and prognostic potential of joint imaging in ... · decrease in bone mineral density...

Linköping University Medical Dissertations No. 1658

Diagnostic and prognostic potential of

joint imaging in patients with

anti-citrullinated protein antibodies

Michael Ziegelasch

Department of Clinical and Experimental Medicine

Linköping University, Sweden

Linköping 2018

Tabellexempel m.m.

Michael Ziegelasch, 2018

Cover/picture/Illustration/Design: Ultrasound image of an inflamed joint taken

by the author

Published article has been reprinted with the permission of the copyright holder.

Printed in Sweden by LiU-Tryck, Linköping, Sweden, 2018

ISBN: 978-91-7685-145-6

ISSN: 0345-0082

Isidor’s idea of TIRx /T-Rex

To

Edna,

Isidor & HugoMax

ABSTRACT

The introduction of novel therapeutic strategies set new goals for the patients’

outcome, which aims to achieve remission. This goal requires early diagnosis of

RA and prompt efficient pharmacotherapy. The introduction of anti-citrullinated

protein antibodies (ACPA) two decades ago allowed an earlier RA diagnosis.

However, there are indications that ACPA positivity is still associated with higher

rates of radiographic damage. As the small joints in hands and feet commonly are

the first involved sites of inflammation, the role of different imaging modalities

were studied regarding their diagnostic and prognostic impact for assessment of

arthritis in RA. Further, ultrasound (US) and radiography were used to study the

association between RA-specific antibodies and the occurrence of arthritis and

joint damage in systemic lupus erythematosus (SLE).

The use of US allows assessment of soft tissue like joint capsules, tendons and

bursae. Used for a live scanning, it is easy to detect effusions and edema. Doppler

indicates vasoproliferation were inflammation is present. Also, US seems to be

more sensitive than radiography to detect minimal structural changes located at

bone surfaces. We wanted to investigate whether US findings in a pre-RA stage

can predict development of arthritis.

Digital X-ray radiogrammetry (DXR) is a technique based on computerized

analyses of standard hand radiographs to calculate peripheral bone mineral density

(BMD) of the three middle metacarpal bones (DXR-BMD). In order for early

treatment decisions, we aimed to study whether changes in DXR-BMD loss after 3

months can predict radiographic damage in early RA.

In conclusion, the studies showed that ACPA-positivity is still associated with a

higher risk of radiographic damage regardless of early treatment decisions.

Therefore, close radiographic monitoring and readiness to intensive treatment is

warranted in ACPA-positive patients. This thesis also shows that erosions detected

by US in ACPA-positive patients with arthralgia predict development of clinical

arthritis. Also, the magnitude of DXR-BMD loss helps identify patients at higher

risk for future radiographic damage, and may therefore help to improve early

treatment decisions. Finally, US and radiography confirm a higher rate of arthritis

and erosions also in SLE patients who are positive for RA-specific antibodies.

1

CONTENTS

Populärvetenskaplig sammanfattning

3

Papers in the thesis

4

Abbreviations

5

1. Introduction

6

1.1. Rheumatic diseases and arthritis

6

1.2. The course of RA 6

1.2.1. Preclinical RA 6

1.2.2. Clinical phase

8

1.3. Treat to Target 10

1.3.1. The Treat to Target concept 10

1.3.2. Arthritis in the diagnosis of rheumatic diseases 10

1.3.2.1. Classification criteria for rheumatoid arthritis 10

1.3.2.2. Arthritis in other rheumatic diseases 12

1.3.2.3. Diagnosis by clinical features and the role of imaging 13

1.3.3. Antirheumatic treatment 14

1.3.4. Monitoring 15

1.3.5. Assessment of clinical remission

16

1.4. Alternatives of imaging for diagnosing and predicting arthritis 17

1.4.1. Ultrasound 17

1.4.2. Digital X-ray radiogrammetry 19

2. Aims of the thesis 21

3. Material and methods

22

3.1. Paper I – SLE study 22

3.2. Paper II – DXR study 24

3.3. Paper III – TIRA study 26

3.4. Paper IV – TIRx study 27

2

4. Results

29

4.1. Paper I 29

4.2. Paper II 30

4.3. Paper III 32

4.4. Paper IV 36

5. Discussion

40

5.1. Diagnostic potential of joint imaging in patients with anti-

citrullinated protein antibodies

40

5.1.1. Ultrasound used in the earliest stage of ACPA positive

arthralgia

40

5.1.2. Arthritis confirmed by ultrasound more common in ACPA

positive SLE-patients

41

5.1.3. Radiographic damage in rheumatoid arthritis is associated

with ACPA positivity

41

5.1.4. Higher rates of radiographic damage in ACPA- and anti-CarP


42

5.2. Prognostic potential of joint imaging to predict arthritis and joint

damage in patients with anti-citrullinated protein antibodies

43

5.2.1. Ultrasound detected erosions in ACPA positive patients are

predictive for arthritis

43

5.2.2. Digital X-ray radiogrammetry works as a prognostic tool for

radiographic damage in newly diagnosed rheumatoid arthritis

44

6. Conclusion

45

7. Future plans

46

8. Acknowledgements

47

9. References 49

3

POPULÄRVETENSKAPLIG SAMMANFATTNING

Reumatoid artrit (RA)är en kronisk inflammatorisk ledsjukdom.

Inflammationen kan i sin tur orsaka urkalkning av bentätheten med benskörhet

och ledskada som följd. Typiska symtom är, svullnad, värmeökning och

funktions nedsättning i lederna. Första symtomet är dock oftast värk som kan

finnas redan långt innan en patient visar typiska fynd som svullnad i lederna

eller förhöjda inflammatoriska värden. Förekomst av autoantikroppar som

reumatoid faktor (RF) eller antikroppar mot citrullinerade proteiner (ACPA)

kan hjälpa till att identifiera patienter med ökad risk att insjukna. Trots

förbättrade behandlingsmöjligheter med nya antireumatiska läkemedel finns det

i våra undersökningsresultat fortfarande tecken på ledskada (erosioner) på

röntgen hos en del av patienterna. Dessa erosioner är vanligare hos patienter

som är positiva för ACPA. Således finns det behov av att hitta

undersökningsmetoder som kan hjälpa att diagnosticera RA ännu tidigare så att

lämplig behandling med bromsmediciner kan initieras så tidigt som möjligt.

Vi ville därför undersöka om nyare undersökningsmetoder som ultraljud (UL)

och digital X-ray radiogrammetry (DXR) kan hjälpa att identifiera patienter

med ledvärk och ACPA positivitet. UL är en enkel undersökningsmetod som

används på reumatologkliniker för bedömning av inflammation i leder, senor

och slemsäckar. Vi fann att inflammation i lederna av patienter med värk och

ACPA positivitet är mycket vanligare jämfört med friska personer. De ACPA

positiva patienter som redan hade små erosioner vid första UL-undersökningen

hade klart högre risk att utveckla ledinflammationer senare i förloppet.

DXR är en automatiserad metod som beräknar benförlust i metacarpalbenen

(mellanhandsbenen) 2-4 utav digitala röntgenbilder. Vi kunde se, att en högre

benförlust under dem första 3 månader var kopplat till ökad förekomst av

röntgenologiskt påvisbara ledskador efter 1 år.

Våra studieresultat visar att UL kan bidra att identifiera patienter med ledvärk

och ACPA som har hög risk att utveckla inflammatorisk ledsjukdom. Tidig

behandling och rätt val av antireumatisk behandling hos dessa patienter kan

förbättra prognosen. DXR kan hjälpa reumatologen att identifiera patienter som

inom närmaste året kommer utveckla ledskada, men bidrar inte med prognostisk

information på längre sikt.

4

PAPERS IN THE THESIS

I Michael Ziegelasch, Myrthe A. M. van Delft, Philip Wallin, Thomas

Skogh, César Magro-Checa, Gerda M. Steup-Beekman, Leendert A.

Trouw, Alf Kastbom and Christopher Sjöwall

Antibodies against carbamylated proteins and cyclic citrullinated

peptides in systemic lupus erythematosus: results from two welldefined

European cohort.

Arthritis Research & Therapy (2016) 18:289

II Michael Ziegelasch, Kristina Forslind, Thomas Skogh, Katrine Riklund,

Alf Kastbom and Ewa Berglin

Decrease in bone mineral density during three months after diagnosis of

early rheumatoid arthritis measured by digital X-ray radiogrammetry

predicts radiographic joint damage after one year

Arthritis Research & Therapy (2017) 19:195

III Michael Ziegelasch, Antonia Boman, Klara Martinsson, Ingrid Thyberg,

Claudia Jacobs, Britt-Marie Nyhäll-Wåhlin, Anna Svärd, Ewa Berglin,

Solbritt Rantapää-Dahlqvist, Thomas Skogh, Alf Kastbom

ACPA-associated radiographic damage in early RA uncoupled from

inflammation and initial treatment response

Manuscript (submitted)

IV Michael Ziegelasch, Emma Eloff, Hilde Berner-Hammer, Jan Cedergren,

Klara Martinsson, Åsa Reckner, Thomas Skogh, Mattias Magnusson , Alf

Kastbom

Joint erosions visible on ultrasound predict arthritis development in

patients with anti-citrullinated protein antibodies and musculoskeletal

pain but no swollen joints

Manuscript

5

ABBREVIATIONS

ACPA – Anti-citrullinated peptide/protein antibodies

ACR – American college of rheumatology

bDMARD – Biologic disease modifying anti-rheumatic drug

BMD – Bone mineral density

CCP – Cyclic citrullinated peptides

CDAI – Clinical disease activity index

CRP – C-reactive protein

csDMARD – Conventional synthetic disease modifying anti-rheumatic drug

DAS28 – 28-joint Disease Activity Score

DMARD – Disease modifying anti-rheumatic drug

DXA – Dual energy X-ray absorptiometry

ESR – Erythrocyte sedimentation rate

EULAR – European league against rheumatism

GC – Glucocorticosteroids

GS – Grey scale

HAQ – Health assessment questionnaire

HLA – Human leukocyte antigen

IL - Interleukin

MSK – Musculoskeletal

MTX – Methotrexate

NSAID – Non steroidal anti-inflammatory drugs

PD – Power Doppler

RA – Rheumatoid arthritis

RF – Rheumatoid factor

SDAI – Simplified disease activity index

SJ – Swollen joints

SE – Shared epitope

SLE – Systemic lupus erythematosus

SLICC – Systemic Lupus International Collaborating Clinics

T2T – Treat to target

TJ – Tender joints

TNF-α – Tumor necrosis factor-alpha

US - Ultrasound

VAS – Visual analogue scale

6

1. INTRODUCTION

1.1. Rheumatic diseases and arthritis

Rheumatism is translated from old Greek and means “flow”. It was described

already 1591 in Liber de Rheumatismo et pleuritide dorsali by Guillaume de

Baillou. In his opinion, rheumatism was caused by any kind of cold secretion

flowing from the brain down to the extremities. Accumulated in hands and feet

it was associated with symptoms like swelling and tearing and pulling pains.

One of the first described rheumatic disorders was gout. Today, more than 100

diagnoses belong to the rheumatic diseases. They are all inflammatory and may

be localized in the skeleton, joints, muscles, connective tissue, vessels and other

organ systems. Arthritis is a common feature in a row of rheumatic diseases

such as rheumatoid arthritis (RA), but also in e.g. systemic lupus erythematosus

(SLE). Autoimmune processes start prior to the development of overt RA, a

period referred as ‘preclinical RA’ [1, 2].

1.2. The course of RA

1.2.1. Preclinical RA

The combination of genetic, autoimmune and environmental factors are

important for susceptibility to RA [3]. In RA, the genetic contribution of human

leucocyte antigen (HLA) has been estimated between 30-50 % [4]. In fact, class

II HLA is directly involved in the pathogenesis RA [5]. The shared epitope (SE)

hypothesis describes the presence of a specific amino acid sequence in the

protein molecule of HLA-DRB1 which is thought to facilitate the presentation

of arthritogenic peptides to T-cells [5]. In RA, SE alleles are risk factors

primarily for RA if antibodies to citrullinated peptide/proteins (ACPA) are

present [6].

ACPAs have proved to be powerful biomarkers with pathological effects

allowing the prediction of RA in patients with musculoskeletal (MSK) pain

(particularly arthralgia) but without clinical arthritis [7, 8]. ACPAs are

7

autoantibodies directed against citrullinated peptides and proteins. Typically,

the process of citrullination starts during inflammation [9, 10]. Arginine

deiminase converts arginine amino acid residues into citrulline residues in

proteins such as vimentin, fibrinogen, filaggrin, alfa-enolase etc. The post-

translational shapes can be seen as antigens generating immune responses which

results in the production of ACPAs with ≥95% specificity for RA [1]. Several

studies confirm that the presence of ACPAs associates with accelerated bone

erosions in RA [11-13]. To detect ACPAs, several assays have been developed,

for instance employing filaggrin-derived peptides (CCP-assay), mutated

citrullinated vimentin (MCV-assay), and viral citrullinated peptides (VCP-

assay).

Recently, antibodies directed against carbamylated antigens (anti-CarP) were

identified in RA patients as well as in healthy individuals before the onset of

clinical symptoms [14]. These antibodies may be helpful to predict RA mainly

in ACPA- and RF-negative cases [15-17]. Anti-CarP was found to have a high

specificity of 89% and a sensitivity of 44% for RA [18]. Anti-CarP can also be

detected in other rheumatic diseases, for instance SLE, where anti-CarP as well

as anti-CCP associate with erosive joint disease [19].

Carbamylation is mediated by a chemical reaction of cyanate mainly with lysine

residues in proteins [20]. Cyanate is present in the body in equilibrium with

urea. Inflammation, smoking and renal failure have been reported to increase

the non-enzymatic post-translational modification in which cyanate binds to

molecules containing primary amine or thiol groups and forms carbamyl groups

[20].

Environmental factors (smoking, periodontitis, gut microbiome) as well as

coexisting conditions (gender, insulin dependent diabetes mellitus, education,

overweight etc.) may finally trigger the onset of RA [21-23]. However, it can

take years until a healthy ACPA positive person with a genetic disposition will

develop overt RA [1] (Figure 1).

8

1.2.1. Clinical phase

Arthritis is defined as joint inflammation and is accompanied by typical joint

symptoms such as skin redness over the joint (rubor), swelling (tumor), heat

(calor), pain (dolor) and decreased joint function (functio laesa). The internal

membrane of the joint capsule, the synovium, is the primary site of

inflammation. Acute phase pro-inflammatory cytokines such as interleukin (IL)

-1 and -6, Tumor necrosis factor (TNF), and interferons are raised and poured

out with synovial fluid into the joint cavity [24]. Hyperemia in the synovium

occurs due to neovascularization and vasodilatation (Figure 2). This results in

development of the hyperplastic proliferative synovium [25]. If the

inflammation lasts more than 6 weeks, it is regarded as chronic. An increased

amount of local and circulating pro-inflammatory cells can be observed.

Ongoing, the remodeling of the synovium leads to the formation of pannus,

which is a layer of fibro-vascular tissue. In this phase, the pro-inflammatory

cytokines mediate activation of cartilage- and bone-degrading cells, i.e.

9

chondroclasts and osteoclasts [26]. The first sign of this is periarticular bone

loss [27], accompanied by development of erosions.

Peripheral arthritis is an obligatory part of the 1987 ACR classification criteria

as well as the 2010 ACR/EULAR criteria for RA.

10

1.3. Treat to Target

1.3.1. The Treat to Target concept

The introduction of novel therapeutic strategies during the new millennium set

new goals for the patients’ outcome, where the “Treat to Target “(T2T) concept

aims to achieve remission. This idea includes early diagnosis of RA and prompt

efficient pharmacotherapy in order to successfully achieve remission, with the

potential to prevent work disability and minimize socioeconomic consequences

for the patients and their families. Despite earlier treatment and the introduction

of biologic agents, the rate of disability pension results in high costs also in the

new millennium [28, 29].

1.3.2. Arthritis in the diagnosis of rheumatic diseases

1.3.2.1. Classification criteria for rheumatoid arthritis

In 1937, the Norwegian blood group serologist Erik Waaler discovered a

“serum factor” among patients with RA, and which caused agglutination of

human IgG bound to sheep red blood cells. This factor, which was later

designed as “Rheumatoid Factor” (RF), proved to be autoantibodies against

human IgG-Fc among RA patients. Waaler’s findings were published in 1940.

In 1948, the American scientist Harry M Rose ‘rediscovered’ the same auto-

antibody [30], which is still designated RF. The first classification criteria for

RA were published in 1958, and contained RF as well as two histological

factors among 11 components in total [31]. In 1987, a new set of modified RA

criteria was launched (Table 1). However, it was later concluded that the

performance of the 1987 criteria did not differ significantly from the 1958

criteria [32]. As radiographic damage is a sign of longstanding disease, the

patients identified by these criteria were still at great risk to develop disabilities.

11

To reach the T2T aims, it was necessary to maximize the sensitivity for the

diagnosis of RA, and new classification criteria were created [33]. Thus, in July

2010, the 2010 ACR/EULAR RA classification criteria were introduced [34]

(Table 2). These new classification criteria include ACPA testing. Also, the

2010 criteria were modified regarding the classification of joint involvement in

RA, contributing with up to 5 of the 6 needed points for RA diagnosis. Hence,

the 2010 criteria are constructed to enable identification of earlier RA as

compared to previous settings.

12

1.3.2.2. Arthritis in other rheumatic diseases

Arthritis is a common feature also in in a row of other rheumatic conditions.

Secondary to arthrosis, arthritis can be caused by mechanical stress, trauma or

osteoporosis and will then be termed “osteoarthritis”. Clinical arthritis may

include gout, spondylarthritis (including psoriasis arthritis), reactive arthritis,

sarcoidosis, Still’s disease, vasculitis syndromes and others with joint affection.

Furthermore, arthritis can be found as one criterion in the ACR-82 and 2012

revised SLICC-classification criteria for SLE [35, 36]. Articular involvement in

SLE is common and has been reported to be present in up to 80 % of the cases

[37]. However, the prevalence of erosive arthritis in SLE is low, but may

complicate the distinction between SLE and RA [38]. “Rhupus” has been

described as an overlap syndrome in which patients fulfill classification criteria

for both conditions - SLE as well as RA [39, 40].

13

1.3.2.3. Diagnosis by clinical features and the role of imaging

Patients with MSK pain are commonly referred from primary health care to the

rheumatology department if they test positive for ACPA. These patients may or

may not have elevated inflammatory variables such as C-reactive protein (CRP)

and/or erythrocyte sedimentation rate (ESR). When clinical arthritis is present at

the first visit to a rheumatologist, and the patient fulfills the ACR classification

criteria for RA, anti-rheumatic therapy with corticosteroids and a disease

modifying anti-rheumatic drug (DMARD) will be initiated.

Among imaging modalities, conventional radiographs of the hands and feet are

still the first choice; first to have a baseline status for later monitoring of

treatment efficacy, second to figure out information regarding the prognosis as

erosions at baseline are predictive of further joint damage [41]. There are some

known scoring instruments to describe the radiographic status. Sharp score

modified by van der Hejde (SHS) and Larsen score are the most commonly

used in clinical trials and research. In the SHS, a maximum score of 448 can be

reached, consisting of an erosion score range from 0 to 280 and joint space

narrowing score range 0 to 168 [42]. In Larsen score, 32 joints will be graded 0-

5 regarding structural changes and can result in a maximum score of 160 [43].

However, these methods are difficult to apply in clinical practice as they are

very time-consuming [44]. In an attempt to solve this, the author and coworkers

developed a modified scoring system according to Larsen, suggesting

assessment of present erosions (Yes/No) with a maximum score 32. However,

when testing the suggested simplified Larsen erosion score among local

radiologists, it was still found too time consuming (unpublished data). One way

to circumvent this is to describe the findings, instead of counting erosions.

In clinical practice as well as in research, disease activity is commonly

measured by the 28-joint count disease activity score (DAS28), including joint

swelling, tender joints, ESR and patients global VAS [45]. To assess functional

ability of the patients, the health assessment questionnaire (HAQ) is widely

used [46].

14

1.3.3. Antirheumatic treatment

Non-steroidal anti-inflammatory drugs (NSAIDs) are often initiated by the

general practitioner during the waiting period for the first rheumatologist

consultation. Once the diagnosis of RA is confirmed by the rheumatologist,

DMARDs should be started immediately [47]. Since DMARDs need weeks to

months to reach effectiveness, glucocorticosteroids (GCs) are used to achieve

rapid clinical improvement. When used systemically, GCs decrease the

migration of macrophages, B- and T-cells into injured tissue and inhibit

secretion of pro-inflammatory cytokines [48]. Already in low doses, GCs may

have protective effects regarding radiographic progression [49]. Furthermore,

GCs can be used for intraarticular injections. In these cases, GCs will be

directly effective in the joint capsules’ internal synovial membrane on the basis

of inhibited expression of pro-inflammatory cytokines (TNF, IL-1, IL-6, and

IFN) [50]. However, GCs used over a long time, have a risk of side effects such

as diabetes mellitus, osteoporosis, hypertension, edema, and muscle atrophy.

Osteoporosis prophylaxis with calcium, vitamin D, and bisphosphonate is

recommended [51]. The rheumatologist aims to taper and discontinue the

treatment with GCs as soon as the chosen DMARD takes over the control of the

disease [47].

According to EULAR recommendations, methotrexate (MTX) should be part of

the first line treatment strategies [47]. MTX is highly effective as monotherapy

as well as in combination with other DMARDs. MTX in monotherapy has been

shown up to 70 % improvement rates according to ACR criteria in DMARD-

naïve patients [52]. Adding an adequate dose of folic acid, MTX is generally

well tolerated [53]. However, hepatic, renal or pulmonary side effects can

hinder the treatment with MTX. In these cases sulfasalazine or leflunomide

should be considered as part of first treatment strategy.

Two decades ago, at the beginning of the new millennium, the first biologic

(b)DMARDs were introduced and has since then revolutionized the treatment

options for patients with highly inflammatory disease. Treatment with

bDMARDs are most often initiated as a second or third option, in case of

failures with conventional synthetic (cs)DMARD. At first, the tumor necrosis

factor-alpha (TNF) inhibitor etanercept was introduced, and a little later even

Infliximab and adalimumab. golimumab and certolizumabpegol followed a

15

couple of years later. Besides TNF inhibitors, also IL-1 and 6- inhibitors

(anakinra and tocilizumab respectively), inhibitor of T-cell activation

(abatacept) and depletion of CD20-positive B cells (rituximab) were introduced

as treatment options. Recently, oral inhibitors of Janus kinases (JAK inhibitors

baricitinib and tofacitinib) became further treatment options.

1.3.4. Monitoring

At RA onset and initiation of pharmacotherapy, a tight control of treatment

according to the T2T strategy has shown to be most effective. As recommended

in the T2T concept, the use of activity scoring instruments leads to more rapid

DAS28 remission and to a higher percentage of remission compared to “usual

care treatment” [54, 55].. DAS28 comprises inflammatory marker (ESR or

CRP), swollen and tender joint count, and a visual analogue scale (VAS) to

assess the patients’ general health. Initially, DAS included 44 joints

recommended by Ritchie [56]. This was later modified and reduced to 28 joints

defining the current DAS28 (or DAS28-CRP when CRP is used instead of ESR)

[45, 57] (Figure 3).

Other scores in use are the “Simplified Disease Activity Index” (SDAI) and

“Clinical Disease Activity Index” (CDAI) which are simpler instruments, but

more stringent in defining remission [58, 59] (Figure 3).

16

According to EULAR, remission is defined as DAS28<2.6. DAS28≤3.2 stands

for low, >3.2 - ≤5.1 for moderate and >5.1 for high disease activity. Response to

treatment according EULAR is based on both the change in DAS28 and the

absolute value at evaluation Figure 4).

1.3.5. Assessment of clinical remission

One of the ACR/EULAR definitions of remission in RA clinical trials is

Boolean-based and another is based on SDAI which has been described above

[60] . The Boolean based includes TJ count ≤1, SJ count ≤1, CRP≤1mg/dl, and

patients global (PG) assessment ≤1 [60]. In SDAI, the patient is considered to

be in remission when having a score of ≤3.3 [61]. The most used score for

definition of RA remission in clinical practice is the DAS28<2.6. However, it

can be difficult to reach the remission criteria as the PG assessment is often

high, particularly among elderly patients [62]. On the other hand, the

assessment of ankles and forefeet are partly limited. Tenderness and swelling in

the feet have proved to be greater compared to other joints, thereby resulting in

higher scores. Hence, it can be difficult to reach ‘true remission’, and

discussions are ongoing whether or not assessment of the feet should be

excluded when evaluating remission [60].

17

1.4. Alternatives of imaging for diagnosing and predicting

arthritis

Despite new revolutionized treatment options, work disability remains high

among patients with RA [63]. Disabilities are associated with functional

impairment and higher mortality rates [64]. The right timing for starting of anti-

rheumatic therapy is in a window of opportunity when clinical symptoms

appear [65]. Synovitis is the primary site of pathology in RA; therefore it seems

logical that imaging modalities to detect subclinical inflammation should be

used to ensure good response to treatment with DMARDs in an early stage of

the disease.

1.4.1. Ultrasound

Ultrasound (US) is defined as sound at frequencies above the audible frequency

range of humans. Grey scale (GS) US is an imaging modality in medicine since

a couple of decades. Doppler US is the compression or elongation of sound

waves due to changes in the distance between the transmitter and receiver. The

Doppler effect is named after the Austrian physicist Christian Doppler (1803-

1853), who thought that the colour of the stars is based on the change of

distance during light emission.

US has become a promising imaging tool in rheumatology. It allows the

assessment of soft tissues like joint capsules, tendons and bursae. US is

inexpensive and, used for real time scanning, it is possible to communicate with

the patient and to focus directly on the site of illness. It can also be conveniently

used as a bedside tool. Treatment strategies including intraarticular injections

can be decided at once during the patient’s consultation with the rheumatologist.

Further advantages are no radiation, and it is easy to use for monitoring.

However, limitations are that US cannot penetrate bone. The examiner has to

understand the multiplane examination technique, and to be skilled in anatomy

of the examined structures. Furthermore, the examiner has to be able to manage

instrument settings, particularly of Doppler, as positive adequate power Doppler

(PD) can crucially contribute to useful assessments of inflammation [66, 67].

18

Ultrasound detects effusions and oedema as signs of ongoing musculoskeletal

(MSK) inflammation. GS ultrasound visualizes thickening of the synovial

membrane in joints and tendons, effusions and structural bone changes, such as

erosions. Addition of PD to GS findings allows detection of hyperemia, which

is a sign of active inflammation [68] (Figure 5). Early diagnostic US to detect

ongoing/progressive inflammation has been shown to predict progression to

arthritis, both among ACPA-positive and/or RF-positive, and sero-negative

arthralgia patients [69-72]. The risk of progression has been reported to be

highest among patients with positive PD [73]. According to the 2010 criteria,

US may be used to confirm the clinical findings [34]. A Japanese study showed

that US may improve the performance of ACR-EULAR 2010 to identify

patients who will need DMARD treatment [74], but general US screening to

increase fulfilment of the 2010 criteria remains to be validated.

US is comparable with magnet resonance imaging (MRI) in terms of detecting

synovial inflammation [75, 76]. Regarding structural changes as erosions, US is

markedly site-dependent as it cannot penetrate bone. US is superior to X-ray in

easily accessible joints, but the sensitivity for detecting bone erosions is low in

anatomically complicated joints [77].

Besides the use of US as a diagnostic tool, it has discussed as an imaging

modality to monitor remission. There is evidence that patients who were

clinically assessed to be in remission, but where US actually showed signs of

ongoing inflammatory activity, faced a significantly increased risk of

radiographic progression [78, 79]. However, a randomized trial from

Haavardsholm et al. showed no additional benefit of ultrasound monitoring in

19

RA treated intensively according to T2T [80]. There are also indications that

patients in clinical remission with only minor signs inflammatory activity by PD

remain inactive for a longer period compared to those with high PD activity

[81]. Ultrasound used for follow up to assess remission and possibly to optimize

treatment, should be considered with reservation.

1.4.2. Digital X-ray radiogrammetry

Bone loss is a common feature in RA. It can appear as systemic osteoporosis

caused by treatment with GC’s, is influenced by sex and high age, and local as

periarticular bone loss due to arthritis [82-84]. Currently, the indication for

measurement of bone mineral density (BMD) is mostly to monitor systemic

osteoporosis. Several non-invasive methods are in use, where dual energy X-ray

absorptiometry (DXA) is considered the gold standard. Other methods include

quantitative computed tomography (QCT), quantitative ultrasound (QUS), and

radiographic absorptiometry (Table 3).

Digital X-ray radiogrammetry (DXR) is a technique to measure periarticular

BMD of the metacarpal bones 2-4 in hands, i.e. in close proximity to 2 of the

most affected joints in RA; MCP 2 and 3. Bone loss detected by this method has

repeatedly been shown to predict radiographic joint progression in early RA.

[85-88]. However, the majority of previous studies investigated 12-months

20

changes in DXR-BMD, and after this time, DXR-evaluation of joint damage

was not superior to conventional radiography for prediction of further radiologic

progression

Except QCT, which measures bone mineral mass per volume, the other methods

measure per projected area.

21

2. AIMS OF THE THESIS

General aim

The overall aim of this thesis was to assess the diagnostic and prognostic

potential of different joint imaging modalities to predict arthritis and clinical

outcomes in patients with anti-citrullinated protein antibodies.

Specific aims

Paper I

To investigate the relationship between RA-related autoantibodies and arthritis

in SLE.

Paper II

To evaluate whether BMD loss measured by DXR predicts radiographic joint

damage in patients with early RA.

Paper III

To determine the predictive value of anti-CCP on disease activity and

radiographic joint damage in contemporary early RA.

Paper IV

To investigate whether ultrasound findings predict arthritis development in

patients with anti-CCP and pain but no clinical arthritis.

22

3. MATERIALS AND METHODS

3.1. Paper I – SLE study

In September 2008, the prospective follow-up program KLURING (a Swedish

acronym for ‘Clinical Lupus Register in Northeastern Gothia’) at the

Rheumatology Clinic, Linköping University Hospital, Sweden, started to

include patients diagnosed with SLE according to ACR criteria and/or Systemic

Lupus International Collaborating Clinics (SLICC) classification criteria [36,

37]. At the time-point of this study, 236 patients were included, of which 83 %

were prevalent cases and 17 % had newly diagnosed SLE. The mean age was 54

years and 207/236 (88 %) of the patients were females.

16 of the 236 (7 %) KLURING patients were positive for ACPA. For the US

assessment of arthritis, we developed a semi-graded protocol including 36

clinically relevant synovial small joints in hands and feet. Also, 6 tendons

referred by Berner-Hammer et al. were subject of our examinations [89] (Table

4).

23

The findings were assessed regarding synovial hypertrophy in GS and

inflammatory activity with PD. The US examiner was blinded to the ACPA-

status of the patients. To grade synovitis, the semi-quantitative scoring system

by Szkudlarek et al. is commonly use [90]. Herein, synovial hypertrophy is

graded 0–3 (0 = no thickening, 1=minimal thickening not bulging over the line

linking the tops of periarticular bones, 2=thickening over but not extending the

diaphysis, 3 = synovial thickening bulging over the tops of the periarticular

bones and extension over the diaphysis of at least one side (Figure 6)) and

perfusion 0–3 (0 = no flow in the synovium, 1=single vessel signals,

2=confluent vessel signals in less than half of the synovial area, 3 = confluent

vessel signals in more than one half of the synovium (Figure 7)).

24

3.2. Paper II – DXR study

In this study, 176 patients from three Swedish centers were included

(Helsingborg, Linköping and Umeå with 60, 48 and 68 patients respectively).

All patients were >18 years old (range 19-87 years) and fulfilled the

ACR/EULAR classification criteria for RA. ACPA and RF were assessed at

baseline, and CRP, ESR DAS28, and HAQ were obtained at all scheduled visits

(baseline, 3, 6, 12, 24 months). For the evaluation of BMD loss, radiographs of

129 patients were available from baseline, 1 and 2 years. The radiographs were

assessed according to Larsen score by one reader at each participating center. In

Larsen score, 32 joints are assessed; metacarpophalangeal joints II-V, proximal

interphalangeal joints II-V, the wrists divided into four areas and the

metatarsophalangeal joints II-V. Each joint was graded 0-5 regarding structural

changes (Table 5) and could result in a maximum score of 160 [43].

25

The smallest detectable change (SDC) was calculated for the three readers

individually according to the method of Bruynesteyn [91]. Radiographic

progression was defined as a difference in Larsen score above the SDC of the

corresponding reader. The intra-rater and inter-rater reliability of the readers

was assessed by calculating the intraclass correlation coefficient (ICC). The

ICC was 0.903.

BMD was estimated on hand radiographs of the second, third and fourth

metacarpal bones using DXR (the online Pronosco X-posure System, SECTRA,

Linköping, Sweden) [85]. DXR is a technique based on computerized analyses

of standard hand radiographs to calculate peripheral bone mineral density

(BMD) of the three middle metacarpal bones [91, 92] (Figure 8).

26

3.3. Paper III – TIRA study

Two multicenter prospective observational early RA cohorts denoted ‘TIRA’ (=

a Swedish acronym for ‘‘early intervention in rheumatoid arthritis’’) enrolled

patients with recent-onset RA 10 years apart. Inclusions criteria were symptom

duration (defined as first observed joint swelling <12 months), and either

fulfilment of the 1987 ACR criteria [27] or suffering from morning stiffness

>60 min, symmetrical arthritis, and small joint engagement.

TIRA-1 enrolled patients 1996-1999, i.e. in the “pre-ACPA era”. During this

time, treatment with biologics was not used in clinical practice. After the

millennium, TIRA-2 enrolled patients 2006-2009, i.e when bDMARDs have

been introduced and become routine in rheumatology care. However, patients in

both TIRA cohorts were not treated with DMARDs prior to enrollment. The

follow up visits were scheduled at 6, 12, 24 and 36 months recording DAS28,

HAQ, and ongoing treatment. Laboratory variables were monitored according

to national Swedish guidelines at all visits. Yearly radiographs of hands and feet

were obtained up to 3 years in TIRA-2, and scored according to Larsen. To

27

confirm our findings, we used independent cohorts from Umeå in Northern

Sweden (TRAM-1 and TRAM-2, respectively), which were similar regarding

inclusion criteria, time-periods of enrollment, and follow-up procedures. Also,

radiographs were scored according to Larsen in both TRAM cohorts.

3.4. Paper IV – TIRx study

The prospective observational study designated ‘TIRx’ (Swedish acronym

meaning x-tra early intervention in RA) is a single center study conducted at the

University hospital in Linköping, Sweden. 116 anti-CCP-positive patients with

musculoskeletal pain and maximum 1 palpable synovitis were enrolled in this

study. Eligible patients referred from primary care centers were screened by one

of four participating rheumatologists between 2010 and 2013. Exclusion criteria

were >1 clinical arthritis, corticosteroid therapy during the preceding 6 weeks, a

prior diagnosis of inflammatory rheumatic disease or age <18. Due to various

reasons, 12 subjects discontinued (Figure 9).

28

Ultrasound examinations were performed by the author. All patients were

examined with the same settings for both B-mode (grey scale) and power

Doppler (PD). To grade synovitis, we used the semi quantitative scoring system

of Szkudlarek et al [90]. The rheumatologists as well as the patients were

blinded to the ultrasonography results. IgG anti-CCP and IgM-RF were

measured at baseline. DAS28 and Health HAQ were assessed at each visit. The

patients were treated as suggested by the physician and the patients’ acceptance.

Radiographs of hands and feet were obtained at baseline, 3 months (hands

only), and after 1 and 2 years. Hand bone mineral density (BMD) of the

metacarpal bones II-IV was measured with DXR online method. The

radiographs will be evaluated by the author according to the Larsen score, and

the findings will be related to BMD loss measured by DXR in a future separate

study.

Also, an age-matched control group consisting of 100 healthy blood donors (50

women and 50 men) was recruited for serum sampling and ultrasound

examination identical to that of the patients. This allowed us to characterize to

which extent ultrasound findings occur among healthy individuals, and interpret

patient findings accordingly.

29

4. RESULTS

4.1. Paper I

Antibodies against carbamylated proteins and cyclic citrullinated peptides

in systemic lupus erythematosus: results from two well defined European

cohorts

4.1.1. Patient characteristics

16 ACPA positive and 16 ACPA negative patients from the KLURING cohort

were examined with ultrasound regarding arthritis and tenosynovitis. The

patients in both groups matched regarding age (median 58 years), gender

(females n=14 and n=13 respectively), disease duration (median 10.5 years) and

were very similar regarding prednisolone dose (2.5mg). 7 patients were positive

for anti-CarP and 7 for RF.

4.1.2. Radiographic and ultrasound findings related to antibody status

Patients with ongoing clinical peripheral arthritis and/or MSK symptoms

underwent radiographs (102/236 (43 %)). Erosions were identified in 10

patients (9,8 % of those with radiographs available). All 16 ACPA positive

patients had radiographs available, 4 of whom had radiographic erosions

compared to 6 of the 86 ACPA negative (p<0.05).

Among the patients who were examined with ultrasound regarding arthritis and

tenosynovitis, none of the ACPA negative SLE controls (14/16 with

radiographs available) compared to 4 of ACPA positives had radiographic

erosions (p=0.103). Significantly more patients among the RF positives (n=7)

vs. negatives (n=23) (3 vs. 1; p=0.031) and anti-CarP positives (n=7) vs.

negatives(n=23) (4 vs. 0; p=0.001) had erosions.

The ACPA positive patients had significantly higher US mean sum scores both

in GS and PD compared to ACPA negatives (11,5 vs. 4.1; p=0.014 and 4.7 vs.

30

0.4, p=0.037 respectively). The same trend with numerically higher mean sum

scores in GS and PD among the antibody positive patients could be seen

regarding RF (12.0 vs. 6.6; p=0.32 and 5.4 vs. 1.8; p=0.30 respectively) and

anti-CarP (10.9 vs. 7.0; p=0.29 and 5.4 vs. 1.8; p=0.30 respectively) (Figure

10).

4.2. Paper II

Decrease in bone mineral density during three months after diagnosis of

early rheumatoid arthritis measured by digital X-ray radiogrammetry

predicts radiographic joint damage after one year

4.2.1. Patient characteristics at baseline

In this study, radiographs of 129 patients were available for scoring according

to Larsen and DXR analysis. The median age of the included patients was 58

years and 64 % were females. Among patients with BMD loss, the portion of

females was marginally higher compared to the patients without BMD loss

(70.4 vs. 55.1 %; p=0.05). Also, at baseline, patients with BMD loss were

31

significantly older compared to the others (60 vs. 56 years; p=0.042), had

significantly higher DAS28, (5.06 vs. 4.57; p=0.023), Larsen scores (4.7 vs. 3.2;

p=0.034), but lower BMD values (565 vs. 661mg/ccm; p= 0.008).

4.2.2. Development of Larsen score in relation to bone loss during the

first three months

In general, the higher the grade of BMD loss within the first 3 months, the

higher the total Larsen score (mean) at baseline (p=0.039), after 1 year

(p=0.024) and 2 years (p=0.056) (Figure 11). In simple regression analysis, sex,

ACPA, and baseline values of ESR, DAS28, and Larsen score were

significantly or trend-wise (p<0.2) associated with change in Larsen score

greater than SDC. Multiple regression analysis adjusting for these variables

showed a significant association between 3-month BMD loss and increase in

Larsen score > SDC after 1 year (p = 0.033, adjusted R-squared = 0.069).

However, 2-year increase in Larsen score was not significantly associated with

3-months BMD loss (p= 0.626, adjusted R-squared=0.034)

32

4.3. Paper III




The patients in TIRA-1 and TIRA-2 had baseline characteristics rather typical

for early RA cohorts. There were no significant differences between the 2

cohorts except for age (mean 55 vs. 59 years, p=0.003). ACPA was not used as

a diagnostic biomarker in TIRA-1, but the proportion of patients testing positive

were similar to TIRA-2 (65 vs. 68 % respectively; p=0.44).

4.3.2. Treatment and disease course according to baseline

autoantibody status in TIRA-1 and TIRA-2

In the pre-millennium cohort TIRA-1, the baseline levels for ESR and CRP

differed not significantly among ACPA-positives subjects compared to the –

negatives. However, the RF-positives compared to –negatives had significant

higher ESR (37 vs. 31 mm/h; p=0.012) and trend-wise higher CRP (31 vs. 25

mg/L; p=0.054) at baseline. Treatment with DMARDs was started in 54 % of

the patients (44 % of the RF positives and 52 % of the ACPA positives (Figure

12)). ACPA sassociated with higher CRP and ESR during the follow-up period

(p=0.001 and p=0.002, respectively).

33

In TIRA-2, ACPA positive patients had borderline lower ESR at baseline (31

vs. 35 mm/h; p=0.096), and significantly lower CRP (23 vs. 34 mg/L; p=0.007).

Treatment with DMARDs was initiated in 93 % among the ACPA positive

patients, and in 91 % among the ACPA negatives (p=0.715) (Figure 12). ESR

and CRP did not differ significantly during follow-up according to baseline

ACPA (Figure 13) or RF status in TIRA-2.

35

4.3.3. Radiographic outcome is associated with baseline ACPA status

In TIRA-2, baseline Larsen scores did not significantly differ between ACPA

positive and ACPA negative patients (3.0 vs. 2.4, p=0.28). At 36 months,

however, ACPA positive patients had significantly higher Larsen score than

ACPA negative patients (mean 5.4 vs. 3.5, p=0.027 (Figure 14)).

Similar results regarding radiographic damage were seen in the confirmation

cohorts from Northern Sweden, both in the historical and the modern cohort

(TRAM-1 and TRAM-2 respectively). Testing positive for ACPA at baseline

was associated with higher Larsen scores at baseline (TRAM-1; mean 5.8 vs.

2.7, p=0.01 and TRAM-2; mean 6.6 vs. 5.1, p=0.03) as well as follow up

(TRAM-1; mean 11.3 vs. 6.4, p=0.001 and TRAM-2; mean 8.9 vs. 6.7, p=0.009

(Figure 15)).

36

In multivariable analysis, Larsen score at follow-up was independently

associated with baseline Larsen score (p<0.001), but also baseline ACPA status

reached statistical significance (p=0.039). In multivariable linear regression

analysis in TIRA-2, baseline ACPA status remained associated with 3-year

radiographic damage also after adjusting for EULAR response to treatment at 6

months (p=0.050, regression coefficient = 1.045, 95 % CI 0.00-2.089). Also in

the contemporary confirmation cohort TRAM-2, ACPA remained significantly

associated with follow-up Larsen score after adjusting also for initial treatment

response (p=0.027, regression coefficient 1.109, 95 %CI 1,127-2.090).

4.4. Paper IV

Erosions on ultrasound predict arthritis development in patients with anti-

citrullinated protein antibodies and musculoskeletal pain but no swollen

joints


The included patients had a mean age 51.9 years (SD=15.1), and 82 (79.6 %)

were females. Further baseline characteristics are listed in Table 6. As controls,

we recruited 100 blood donors (50 females, 50 males; mean age 52 years) from

37

the Department of Transfusion Medicine at Linköping University Hospital. The

controls underwent serum sampling and US examination once.

Among the 104 patients in the TIRx study, 22 patients had clinical arthritis at

baseline.

The patients with clinical arthritis had significantly higher mean DAS28

compared to the 82 patients who had not (3.35 vs. 2.49; p=0.001). The other

baseline characteristics did not differ between these groups.

38

4.4.2. Initiation of treatment

As clinical arthritis was the variable to predict by ultrasound, the remaining 82

patients without baseline arthritis were further evaluated. Among them, 35 (43

%) received treatment with oral GC and/or DMARD within the first 2 years, (29

(35 %) received GC, 29 (35 %) received DMARD, and 32 (31 %) received

both. Seven out of the 35 patients initiated oral GC or DMARD therapy without

presenting with clinical arthritis. The remaining 28 patients were diagnosed

with clinical arthritis before or after starting treatment with GC and/or

DMARDs. Follow-up in TIRx is ongoing, but in this study we evaluate follow-

up until September 1st, 2017, resulting in a mean follow-up time of 68 months.

As expected, significantly more patients with clinical arthritis during follow-up

received treatment with DMARDs and/or GC (28 vs. 11 %; p<0.001).

4.4.3. Ultrasound findings in ACPA positive individuals with arthralgia

and healthy controls

At joint-level, significantly more MCP- and PIP-joints of the patients had signs

of synovial hypertrophy (GS≥1) (5.2 vs. 2.5 %; p<0.001 and 6.6 vs. 1.5 %;

p<0.001 respectively). In contrast, significantly more control MTP joints had

synovial hypertrophy as compared to patients (48.5 vs. 23.9 %, p<0.001). In

MTP joints, the GS findings were mostly seen in the MTP 1-4, especially

among the controls, but PD was found significantly more in the patients MTP

joints with synovial hypertrophy, as compared to control MTP joints (2.0 vs.

0.2 %; p=0.001). Therefore, we present MTP5 separately, and excluded the

MTP 1-4 joints regarding GS analyses, but included MTP1-4 regarding PD. PD

signals (PD≥1) occurred significantly more often also in the other joint regions

among patients as compared to controls (wrists 7.9 vs. 2.0 %, p<0.001; MCPs

0.7 vs. 0 %, p=0.026; PIPs 1.7 vs. 0 %, p=0.001).

4.4.4. Ultrasound findings and arthritis development

13 (15.6 %) of the 82 patients had ≥1 erosions verified by US at baseline. Ten

(77 %) of these patients developed clinical arthritis within the observation

39

period as compared to 29 (35 %) among those without US erosion (p=0.032). In

cox regression analysis, adjusting for RF, age, sex, ACPA level, and baseline

levels of ESR and CRP, US erosion predicted development of clinical arthritis

(hazard ratio 4.2, 95 % CI 1.7-10.4, p=0.002) (Figure 16).

40

5. DISCUSSION

Polyarthritis is the most typical clinical feature of RA, and is part of both the

1987 and 2010 ACR classification criteria for RA. Following the ‘treat to

target’ (T2T) concept, the 2010 criteria, which weight arthritis and include

ACPA, bring rheumatologists closer to an earlier diagnosis of RA than it was

possible using the criteria from 1987. However, work in this thesis shows that

ACPA is associated with arthritis and radiographic damage despite of earlier

diagnosis and treatment decisions. We aimed to investigate whether the early

use of different imaging modalities in ACPA positive patients can help to

predict arthritis development and outcome.

5.1. Diagnostic potential of joint imaging in patients with

anti-citrullinated protein antibodies

5.1.1. Ultrasound used in the earliest stage of ACPA positive arthralgia

In RA, US is already established as a clinical tool. Following the T2T concept,

algorithms for the use of US to aid diagnosis, monitoring and assessment of

remission of RA, has recently been elaborated [93]. However, the value of US

before RA can be diagnosed by the ACR classification criteria, has been

investigated in only a few studies which differ in methodology as well as

antibody status of the included patients [70, 72, 73]. In paper IV we studied US

findings in ACPA positive patients without clinical arthritis in a window

between the preclinical and clinical phase of RA. Our results show that,

compared to healthy blood donors, patients with MSK pain and ACPA

positivity in a preclinical phase have significantly more signs of inflammatory

joint changes detected by US. These findings are in line with the study from

Nam et al [73]. Previous MRI studies by Hoving et al. showed similar results in

a very early phase of RA development [94].

41

5.1.2. Arthritis confirmed by ultrasound more common in ACPA


To assess the occurrence of arthritis in ACPA positive SLE patients in paper I,

US was used. Arthritis detected by ultrasound occurred significantly more often

in ACPA positive SLE-patients as compared to ACPA negative. The same trend

was observed also in anti-CarP and RF-positive patients with SLE. Three of the

32 examined patients were positive ACPA, RF and anti-CarP. These patients

had significantly more US-verified arthritis and tenosynovitis as compared to

those who were not. However, with only three triple-positive patients, it is not

possible to draw any conclusions, although it is in line with a recent study by

Verheul et al demonstrating a high specificity for RA in triple seropositivity in

RA [95]

US is a valuable tool for the detection of arthritis, tenosynovitis and erosions in

patients with SLE [96, 97]. It can be assumed, that ACPA-positivity in SLE

indicates a phenotype characterized by arthritis. However, as there are only a

few studies, which in addition differ in the methodology, further validation of

US in SLE is necessary before it can be used as a tool in clinical practice.

5.1.3. Radiographic damage in rheumatoid arthritis is associated with

ACPA positivity

In paper III we compared the association between ACPA and radiographic

joint destruction in RA in two cohorts with very different prerequisites for early

RA management. First, a pre-Millenium cohort (TIRA-1) enrolled patients

before ACPA was introduced as diagnostic marker, and before the biologic

DMARDs were introduced, targeting pro-inflammatory cytokines as TNF-α, IL-

1, IL-6 , to decrease T-cell activation or to destroy B-cells. The second cohort,

TIRA-2, studied the ‘ACPA era’, where early aggressive treatment, T2T

strategies, and bDMARDs were no longer controversial. Introduction of the

2010 ACR/EULAR criteria allowed earlier diagnosis of RA, thereby facilitating

earlier treatment decisions.

In the pre-Millenium cohort TIRA-1, ACPA was analysed later from stored

samples. ACPA was shown to associate with higher disease activity (DAS28,

ESR, CRP, swollen joints) during follow up, as previously reported [98]. This

42

observation may definitely be a result of the low rate of patients treated with

early DMARDs. Notable in TIRA-1 is, that the RF status obviously did not

influence early treatment decisions as only 44 % of the RF positive patients

(and 58 % of the RF negatives) started treatment with DMARDs at baseline.

Since the beginning of the new Millennium, when ACPA became used as a

diagnostic tool and also became part of the new ACR classification criteria

2010, RA diagnosis can be made at an earlier stage, enabling earlier

introduction of DMARDs including potent biologics. These facts have resulted

in lower disease activity during follow up in the modern cohorts TIRA-2 and

TRAM-2. However, despite significantly improved and early instituted modern

anti-rheumatic pharmacotherapy, many of ACPA positive RA patients still face

an increased risk of developing radiographic damage with high Larsen scores.

5.1.4. Higher rates of radiographic damage in ACPA- and anti-CarP


Arthritis also occurs in other rheumatic diseases including SLE, and when RA-

related antibodies are present, an ‘overlap disease’ called “rhupus” may be

suggested [39, 40]. It was recently shown that ‘triple positivity’ for the RA-

related antibodies ACPA, RF and anti-CarP was highly specific for RA

development [96]. In paper I we found, that all three RA-associated antibodies

(ACPA, anti-CarP and RF) were significantly associated with radiographically

confirmed erosions in SLE-patients. We hypothesize that pathogenetic

mechanisms could be similar in RA and in a small group of patients with SLE

defined by a clinical phenotype dominated by arthritis. Interestingly though, 60

% of the studied SLE-patients with radiology proven erosions were not

identified by any of the antibodies.

43

5.2. Prognostic potential of joint imaging to predict arthritis

and joint damage in patients with anti-citrullinated protein

antibodies

5.2.1. Ultrasound detected erosions in ACPA positive patients are

predictive for arthritis

In paper IV we also studied the value of US to predict arthritis in ACPA

positive patients without clinical arthritis in a window between the preclinical

and clinical phase of RA. In this study we failed to show significant predictive

values at patient level of either GS or PD at baseline. However, erosions were

not rare and could be detected in 16% of the ACPA positive patients with MSK

symptoms, also after adjusting for potential confounders. The presence of

erosions on US predicts arthritis development, also after adjusting for potential

confounders.

These results are in line with the study by Tamas et al, where at least one

erosion could be detected by US in 66.7 % of patients with early RA [99].

Interestingly, some studies have shown that erosions can easily be detected by

US [100-102]. Also in the absence of radiographically detected erosions, US

can identify erosions as shown in the study by Dohn et al [103]. Apart from our

study, one more group (Nam et al) has described the predictive value of

erosions detected by US to foresay arthritis [73]. The results regarding US

erosions from this study are similar to our findings.

Only the presence of US erosions predicts arthritis development in our study,

despite the lack of predictive value concerning GS and PD scoring. Taken

together, US is a valuable tool to risk-stratify ACPA-positive patients with

MSK pain, but future studies will address whether US erosions also predict

long-term progression of joint damage as determined on conventional

radiographs.

44

5.2.2. Digital X-ray radiogrammetry works as a prognostic tool for

radiographic damage in newly diagnosed rheumatoid arthritis

In paper II we showed that BMD loss over 3 months in new onset RA

measured by DXR is predictive for radiographic damage after 1 year. To our

knowledge, this is the first study measuring BMD loss within such a short

interval. Using DXR as a prognostic tool, treatment decisions to avoid

radiographic damage and disability could be done in an earlier stage of the

disease. However, in our study we did not observe a significant difference in the

DXR value at baseline in the ACPA-positive compared with the ACPA-

negative patients. This fact is contrary to the findings in other studies in which

BMD loss was significantly more widespread in ACPA-positive patients [104-

106]. Furthermore, Forslind et al. showed that patients with early RA, who were

on a daily low dose of GC in addition to conventional DMARDS, had

significantly less DXR-BMD loss as compared with patients with RA who were

not receiving GC [85]. This finding was attributed to the anti-inflammatory

effect of prednisolone, hampering osteopenia induced by inflammation. Since

treatment with GC parallel to DMARDs at the time point of RA diagnosis is

common, the value of the DXR method to predict radiographic damage has to

be justified with more studies.

45

6. CONCLUSIONS

ACPA is associated with arthritis and is associated with higher risk for

radiographic damage not only in RA but also in SLE.

ACPA and anti-CarP positivity in SLE patients are associated with an arthritic

phenotype in SLE and correlate significantly with development of erosions.

ACPA in the era of treatment with biologics is still associated with increased

risk for radiographic damage.

US detected erosions in an early stage of RA development predict development

of arthritis. However, we could not confirm a clinical value of GS and PD in

non-arthritic patients with ACPA and MSK pain (Figure 16).

DXR-BMD loss in newly RA diagnosed patients predicts radiographic damage

after 1 year (Figure 17).

46

7. FUTURE PLANS

Does early introduction of treatment with biologics really decrease radiographic

changes? ACPA as a specific biomarker for RA and associate with radiographic

damage. In paper III, the collected TIRA data still provides a lot of potential

information to be evaluated. One project in planning is to read and score

radiographs in TIRA-2 after 10 years from inclusion with the aim to evaluate

radiographic changes and their relation to antibody status over a long time.

Do US findings, DXR-BMD loss and/or radiographic erosions in a preclinical

stage indicate a possibly more aggressive course of RA? And are these imaging

modalities associated with radiographic outcome after 10 years? All patients in

the TIRx study (paper IV) were examined regarding radiographic changes and

underwent DXR to measure BMD-loss. These data have to be assessed. Further,

we plan new X-rays in all patients after 10 years.

Can the combination of anti-CarP antibodies, ACPAs, and RF precede arthritis

in a near future? Triple seropositivity has a very high specificity for the

identification of patients with early RA. In order to further investigate whether

triple autoantibody positivity in arthralgia patients can precede arthritis and in

which period of time, a prospective study using ultrasound to detect early signs

of joint inflammation would be a suitable option.

47

8. ACKNOWLEDGEMENTS

I am not the first one saying, that it is not possible to thank everyone who

helped me during the years with these theses; but indeed, there are directly and

indirectly more people involved as I could imagine when starting this work.

Actually I never wanted to start new research after finishing my thesis in

Germany 1999. However, as I loved to work with ultrasound already as a

trainee in Switzerland, I have to thank first of all these persons, who were

involved in the acquisition process of the first old-fashioned ultrasound

machine1)

in our rheumatology department before my time in Linköping.

Whoever it was, you are responsible that I changed my mind and started with

research again.

This work would not have been possible without all the participating patients.

Also, I have to thank all the controls, the blood donors in the ‘TIRx’ trial who

directly after their blood donation sacrificed 30 minutes of their time to me for

an ultrasound examination.

During the years I had three main supervisors. I would like to thank Thomas

Skogh, you was aware about my interest for ultrasound and helped me to create

the ‘TIRx’ project which was the start of my research. Although just a short

time to retirement, you had and still have a lot of ideas keeping you involved in

my studies until now. Alf Kastbom, thank you for taking over after Thomas

and teaching me to think and interpret results critically, even when I was

convinced of the advantages of a diagnostic method. And my gratitude to

Mattias Magnusson, who jumped in the last months of my research, for your

patience and help to transform my ideas to interesting results.

Kristina Forslind, my co-supervisor in the especially in the DXR study, thank

you for teaching me scoring radiographs according to Larsen. Thank you, Ewa

Berglin, with a couple of inquiries every week, the DXR-paper is still very

popular!

Hilde Berner-Hammer, my co-supervisor, who helped me particularly

regarding ultrasound in the TIRx project. You had always time to discuss the

results when we were meeting on our ultrasound courses.

________________________________________________________________ 1) The mentioned ‘old-fashioned’ ultrasound machine was without Doppler and

of course not the used equipment in my research.

48

Jan Cedergren, my co-supervisor, I would like to thank you for your clinical

advices in my PhD work. Furthermore you have been my mentor and

introduced rheumatology to me when I started in the Rheumatology department

Linköping 2004.

Christopher Sjöwall, thank you for involving me in KLURING and all your

advices during my research.

Åsa Reckner, thank you for screening patients in the TIRx project. The follow

up of the patients is still going on.

Klara Martinsson at the AIR, and Marianne Pettersson at the Rheumatology

department, thank you for all the help regarding the blood samples.

Monika Nyberg, nurse in the radiology department, thank you for great

collaboration to schedule the radiographs of our patients at the days of screening

and follow up.

The staff in the Blood Centre of Linköping I would like to thank, permitting

me to occupy their conference room for the ultrasound examinations of blood

donors.

Karin Sjöstedt, head of the rheumatology clinic, and all colleagues in the

rheumatology department, thank you for helping me particularly the last

weeks of my PhD work to get all the time I needed. Of course I have to thank

also all nurses and care administrators for calling and taking care of the study

patients.

I would also like to thank my parents, who supported me with motivation but

also critically all my live that my dreams came true so far. My father surely has

still my words in mind that I never would start with any research again after my

first work in Germany. Obviously I was wrong.

If this PhD work should be successfully approved, so it is also thanks to the big

support of my wife Edna. You kept my back free when I had to work, even our

last short holidays where you explored the surrounding by your own, while I

was working at the hotel room. I am sorry for that, but thank you for all your

patience with me during the years! Lilly, du är största Lyckan i mitt liv! ♥ I love

you.

49

9. REFERENCES

1. Rantapaa-Dahlqvist, S., et al., Antibodies against cyclic citrullinated

peptide and IgA rheumatoid factor predict the development of rheumatoid

arthritis. Arthritis Rheum, 2003. 48(10): p. 2741-9.

2. Deane, K.D., Preclinical rheumatoid arthritis (autoantibodies): an

updated review. Curr Rheumatol Rep, 2014. 16(5): p. 419.

3. Deane, K.D., J.M. Norris, and V.M. Holers, Preclinical rheumatoid

arthritis: identification, evaluation, and future directions for

investigation. Rheum Dis Clin North Am, 2010. 36(2): p. 213-41.

4. Seldin, M.F., et al., The genetics revolution and the assault on

rheumatoid arthritis. Arthritis Rheum, 1999. 42(6): p. 1071-9.

5. Gregersen, P.K., J. Silver, and R.J. Winchester, The shared epitope

hypothesis. An approach to understanding the molecular genetics of

susceptibility to rheumatoid arthritis. Arthritis Rheum, 1987. 30(11): p.

1205-13.

6. Huizinga, T.W., et al., Refining the complex rheumatoid arthritis

phenotype based on specificity of the HLA-DRB1 shared epitope for

antibodies to citrullinated proteins. Arthritis Rheum, 2005. 52(11): p.

3433-8.

7. Sokolove, J. and D. Pisetsky, Bone loss, pain and inflammation: three

faces of ACPA in RA pathogenesis. Ann Rheum Dis, 2016. 75(4): p. 637-

9.

8. van Steenbergen, H.W., et al., Clinical factors, anticitrullinated peptide

antibodies and MRI-detected subclinical inflammation in relation to

progression from clinically suspect arthralgia to arthritis. Ann Rheum

Dis, 2016. 75(10): p. 1824-30.

9. Scherer, H.U., et al., The B cell response to citrullinated antigens in the

development of rheumatoid arthritis. Nat Rev Rheumatol, 2018. 14(3): p.

157-169.

10. van der Woude, D. and A.I. Catrina, HLA and anti-citrullinated protein

antibodies: Building blocks in RA. Best Pract Res Clin Rheumatol, 2015.

29(6): p. 692-705.

11. Forslind, K., et al., Prediction of radiological outcome in early

rheumatoid arthritis in clinical practice: role of antibodies to

citrullinated peptides (anti-CCP). Ann Rheum Dis, 2004. 63(9): p. 1090-

5.

12. Johansson, L., et al., Antibodies directed against endogenous and

exogenous citrullinated antigens pre-date the onset of rheumatoid

arthritis. Arthritis Res Ther, 2016. 18(1): p. 127.

13. Kastbom, A., et al., Changes in the anticitrullinated peptide antibody

response in relation to therapeutic outcome in early rheumatoid arthritis:

results from the SWEFOT trial. Ann Rheum Dis, 2016. 75(2): p. 356-61.

50

14. Conigliaro, P., et al., Autoantibodies in inflammatory arthritis.

Autoimmun Rev, 2016. 15(7): p. 673-83.

15. Jiang, X., et al., Anti-CarP antibodies in two large cohorts of patients

with rheumatoid arthritis and their relationship to genetic risk factors,

cigarette smoking and other autoantibodies. Ann Rheum Dis, 2014.

73(10): p. 1761-8.

16. Koppejan, H., et al., Role of Anti-Carbamylated Protein Antibodies

Compared to Anti-Citrullinated Protein Antibodies in Indigenous North

Americans With Rheumatoid Arthritis, Their First-Degree Relatives, and

Healthy Controls. Arthritis Rheumatol, 2016. 68(9): p. 2090-8.

17. Shi, J., et al., Autoantibodies recognizing carbamylated proteins are

present in sera of patients with rheumatoid arthritis and predict joint

damage. Proc Natl Acad Sci U S A, 2011. 108(42): p. 17372-7.

18. Shi, J., et al., The specificity of anti-carbamylated protein antibodies for

rheumatoid arthritis in a setting of early arthritis. Arthritis Res Ther,

2015. 17: p. 339.

19. Lopez-Hoyos, M., et al., Anti-carbamylated protein antibodies in patients

with ageing associated inflammatory chronic disorders. Rheumatology

(Oxford), 2016. 55(4): p. 764-6.

20. Shi, J., et al., Carbamylation and antibodies against carbamylated

proteins in autoimmunity and other pathologies. Autoimmun Rev, 2014.

13(3): p. 225-30.

21. Klareskog, L., et al., Smoking, citrullination and genetic variability in the

immunopathogenesis of rheumatoid arthritis. Semin Immunol, 2011.

23(2): p. 92-8.

22. Reckner Olsson, A., T. Skogh, and G. Wingren, Comorbidity and

lifestyle, reproductive factors, and environmental exposures associated

with rheumatoid arthritis. Ann Rheum Dis, 2001. 60(10): p. 934-9.

23. de Hair, M.J., et al., Smoking and overweight determine the likelihood of

developing rheumatoid arthritis. Ann Rheum Dis, 2013. 72(10): p. 1654-

8.

24. McInnes, I.B. and G. Schett, The pathogenesis of rheumatoid arthritis. N

Engl J Med, 2011. 365(23): p. 2205-19.

25. Leblond, A., Y. Allanore, and J. Avouac, Targeting synovial

neoangiogenesis in rheumatoid arthritis. Autoimmun Rev, 2017. 16(6):

p. 594-601.

26. Hirayama, T., et al., Osteoclast formation and activity in the pathogenesis

of osteoporosis in rheumatoid arthritis. Rheumatology (Oxford), 2002.

41(11): p. 1232-9.

27. Arnett, F.C., et al., The American Rheumatism Association 1987 revised

criteria for the classification of rheumatoid arthritis. Arthritis Rheum,

1988. 31(3): p. 315-24.

51

28. Hallert, E., M. Husberg, and L. Bernfort, The incidence of permanent

work disability in patients with rheumatoid arthritis in Sweden 1990-

2010: before and after introduction of biologic agents. Rheumatology

(Oxford), 2012. 51(2): p. 338-46.

29. Hallert, E., et al., Rheumatoid arthritis is still expensive in the new

decade: a comparison between two early RA cohorts, diagnosed 1996-98

and 2006-09. Scand J Rheumatol, 2016. 45(5): p. 371-8.

30. Glynn, L.E., E.J. Holborow, and G.D. Johnson, The nature of the Rose-

Waaler factor. Proc R Soc Med, 1957. 50(7): p. 469-71.

31. Ropes, M.W., et al., 1958 Revision of diagnostic criteria for rheumatoid

arthritis. Bull Rheum Dis, 1958. 9(4): p. 175-6.

32. Bernelot Moens, H.J., M.A. van de Laar, and J.K. van der Korst,

Comparison of the sensitivity and specificity of the 1958 and 1987

criteria for rheumatoid arthritis. J Rheumatol, 1992. 19(2): p. 198-203.

33. Aletaha, D., F.C. Breedveld, and J.S. Smolen, The need for new

classification criteria for rheumatoid arthritis. Arthritis Rheum, 2005.

52(11): p. 3333-6.

34. Aletaha, D., et al., 2010 Rheumatoid arthritis classification criteria: an

American College of Rheumatology/European League Against

Rheumatism collaborative initiative. Arthritis Rheum, 2010. 62(9): p.

2569-81.

35. Ighe, A., et al., Application of the 2012 Systemic Lupus International

Collaborating Clinics classification criteria to patients in a regional

Swedish systemic lupus erythematosus register. Arthritis Res Ther, 2015.

17: p. 3.

36. Tan, E.M., et al., The 1982 revised criteria for the classification of

systemic lupus erythematosus. Arthritis Rheum, 1982. 25(11): p. 1271-7.

37. Petri, M., et al., Derivation and validation of the Systemic Lupus

International Collaborating Clinics classification criteria for systemic

lupus erythematosus. Arthritis Rheum, 2012. 64(8): p. 2677-86.

38. Richter Cohen, M., et al., Erosive arthritis in systemic lupus

erythematosus: analysis of a distinct clinical and serological subset. Br J

Rheumatol, 1998. 37(4): p. 421-4.

39. Fernandez, A., et al., Lupus arthropathy: a case series of patients with

rhupus. Clin Rheumatol, 2006. 25(2): p. 164-7.

40. Panush, R.S., et al., 'Rhupus' syndrome. Arch Intern Med, 1988. 148(7):

p. 1633-6.

41. van Nies, J.A., et al., Evaluating processes underlying the predictive

value of baseline erosions for future radiological damage in early

rheumatoid arthritis. Ann Rheum Dis, 2015. 74(5): p. 883-9.

42. van der Heijde, D., How to read radiographs according to the Sharp/van

der Heijde method. J Rheumatol, 2000. 27(1): p. 261-3.

52

43. Larsen, A., How to apply Larsen score in evaluating radiographs of

rheumatoid arthritis in long-term studies. J Rheumatol, 1995. 22(10): p.

1974-5.

44. Forslind, K. and B.S. Group, The usefulness of the Simplified Erosion

Narrowing Score (SENS) in clinical practice for estimating joint damage

in early rheumatoid arthritis. Scand J Rheumatol, 2011. 40(6): p. 497-8.

45. Prevoo, M.L., et al., Modified disease activity scores that include twenty-

eight-joint counts. Development and validation in a prospective

longitudinal study of patients with rheumatoid arthritis. Arthritis Rheum,

1995. 38(1): p. 44-8.

46. Josefsson, K.A., et al., Swedish version of the multi dimensional health

assessment questionnaire -- translation and psychometric evaluation.

BMC Musculoskelet Disord, 2013. 14: p. 178.

47. Smolen, J.S., et al., EULAR recommendations for the management of

rheumatoid arthritis with synthetic and biological disease-modifying

antirheumatic drugs: 2013 update. Ann Rheum Dis, 2014. 73(3): p. 492-

509.

48. Barnes, P.J., How corticosteroids control inflammation: Quintiles Prize

Lecture 2005. Br J Pharmacol, 2006. 148(3): p. 245-54.

49. Svensson, B., et al., Low-dose prednisolone in addition to the initial

disease-modifying antirheumatic drug in patients with early active

rheumatoid arthritis reduces joint destruction and increases the

remission rate: a two-year randomized trial. Arthritis Rheum, 2005.

52(11): p. 3360-70.

50. af Klint, E., et al., Intraarticular glucocorticoid treatment reduces

inflammation in synovial cell infiltrations more efficiently than in

synovial blood vessels. Arthritis Rheum, 2005. 52(12): p. 3880-9.

51. Phillips, K., A. Aliprantis, and J. Coblyn, Strategies for the prevention

and treatment of osteoporosis in patients with rheumatoid arthritis.

Drugs Aging, 2006. 23(10): p. 773-9.

52. Aletaha, D., J. Funovits, and J.S. Smolen, The importance of reporting

disease activity states in rheumatoid arthritis clinical trials. Arthritis

Rheum, 2008. 58(9): p. 2622-31.

53. Kinder, A.J., et al., The treatment of inflammatory arthritis with

methotrexate in clinical practice: treatment duration and incidence of

adverse drug reactions. Rheumatology (Oxford), 2005. 44(1): p. 61-6.

54. Naranjo, A., et al., Adjustment in the clinical practice of treat-to-target

guidelines for rheumatoid arthritis: Results of the ToARCan study.

Reumatol Clin, 2016. 12(1): p. 34-8.

55. Schipper, L.G., et al., A tight control treatment strategy aiming for

remission in early rheumatoid arthritis is more effective than usual care

treatment in daily clinical practice: a study of two cohorts in the Dutch

53

Rheumatoid Arthritis Monitoring registry. Ann Rheum Dis, 2012. 71(6):

p. 845-50.

56. Ritchie, D.M., et al., Clinical studies with an articular index for the

assessment of joint tenderness in patients with rheumatoid arthritis. Q J

Med, 1968. 37(147): p. 393-406.

57. Wells, G., et al., Validation of the 28-joint Disease Activity Score

(DAS28) and European League Against Rheumatism response criteria

based on C-reactive protein against disease progression in patients with

rheumatoid arthritis, and comparison with the DAS28 based on

erythrocyte sedimentation rate. Ann Rheum Dis, 2009. 68(6): p. 954-60.

58. Bentley, M.J. and G.W. Reed, Simplified composite disease activity

measures in rheumatoid arthritis: should they be used in standard care?

Clin Exp Rheumatol, 2008. 26(2): p. 358-66.

59. Rintelen, B., et al., Comparison of three rheumatoid arthritis disease

activity scores in clinical routine. Scand J Rheumatol, 2009. 38(5): p.

336-41.

60. Felson, D.T., et al., American College of Rheumatology/European League

against Rheumatism provisional definition of remission in rheumatoid

arthritis for clinical trials. Ann Rheum Dis, 2011. 70(3): p. 404-13.

61. Aletaha, D., et al., Remission and active disease in rheumatoid arthritis:

defining criteria for disease activity states. Arthritis Rheum, 2005. 52(9):

p. 2625-36.

62. Ito, H., et al., Global assessments of disease activity are age-dependent

determinant factors of clinical remission in rheumatoid arthritis. Semin

Arthritis Rheum, 2017. 47(3): p. 310-314.

63. Sokka, T., et al., Work disability remains a major problem in rheumatoid

arthritis in the 2000s: data from 32 countries in the QUEST-RA study.

Arthritis Res Ther, 2010. 12(2): p. R42.

64. Sokka, T., et al., Similar prediction of mortality by the health assessment

questionnaire in patients with rheumatoid arthritis and the general

population. Ann Rheum Dis, 2004. 63.

65. van Steenbergen, H.W., et al., EULAR definition of arthralgia suspicious

for progression to rheumatoid arthritis. Ann Rheum Dis, 2017. 76(3): p.

491-496.

66. Torp-Pedersen, S., et al., Power and color Doppler ultrasound settings

for inflammatory flow: impact on scoring of disease activity in patients

with rheumatoid arthritis. Arthritis Rheumatol, 2015. 67(2): p. 386-95.

67. Torp-Pedersen, S.T. and L. Terslev, Settings and artefacts relevant in

colour/power Doppler ultrasound in rheumatology. Ann Rheum Dis,

2008. 67(2): p. 143-9.

68. Porta, F., et al., The role of Doppler ultrasound in rheumatic diseases.

Rheumatology (Oxford), 2012. 51(6): p. 976-82.

54

69. Backhaus, M., et al., Prospective two year follow up study comparing

novel and conventional imaging procedures in patients with arthritic

finger joints. Ann Rheum Dis, 2002. 61(10): p. 895-904.

70. Freeston, J.E., et al., A diagnostic algorithm for persistence of very early

inflammatory arthritis: the utility of power Doppler ultrasound when

added to conventional assessment tools. Ann Rheum Dis, 2010. 69(2): p.

417-9.

71. Wakefield, R.J., et al., Should oligoarthritis be reclassified? Ultrasound

reveals a high prevalence of subclinical disease. Ann Rheum Dis, 2004.

63(4): p. 382-5.

72. van de Stadt, L.A., et al., The value of ultrasonography in predicting

arthritis in auto-antibody positive arthralgia patients: a prospective

cohort study. Arthritis Res Ther, 2010. 12(3): p. R98.

73. Nam, J.L., et al., Ultrasound findings predict progression to

inflammatory arthritis in anti-CCP antibody-positive patients without

clinical synovitis. Ann Rheum Dis, 2016. 75(12): p. 2060-2067.

74. Nakagomi, D., et al., Ultrasound can improve the accuracy of the 2010

American College of Rheumatology/European League against

rheumatism classification criteria for rheumatoid arthritis to predict the

requirement for methotrexate treatment. Arthritis Rheum, 2013. 65(4): p.

890-8.

75. Strunk, J., et al., Three-dimensional Doppler sonographic vascular

imaging in regions with increased MR enhancement in inflamed wrists of

patients with rheumatoid arthritis. Joint Bone Spine, 2006. 73(5): p. 518-

22.

76. Szkudlarek, M., et al., Ultrasonography of the metatarsophalangeal

joints in rheumatoid arthritis: comparison with magnetic resonance

imaging, conventional radiography, and clinical examination. Arthritis

Rheum, 2004. 50(7): p. 2103-12.

77. Ostergaard, M., S.J. Pedersen, and U.M. Dohn, Imaging in rheumatoid

arthritis--status and recent advances for magnetic resonance imaging,

ultrasonography, computed tomography and conventional radiography.

Best Pract Res Clin Rheumatol, 2008. 22(6): p. 1019-44.

78. Brown, A.K., et al., An explanation for the apparent dissociation between

clinical remission and continued structural deterioration in rheumatoid

arthritis. Arthritis Rheum, 2008. 58(10): p. 2958-67.

79. Brown, A.K., et al., Presence of significant synovitis in rheumatoid

arthritis patients with disease-modifying antirheumatic drug-induced

clinical remission: evidence from an imaging study may explain

structural progression. Arthritis Rheum, 2006. 54(12): p. 3761-73.

80. Haavardsholm, E.A., et al., Ultrasound in management of rheumatoid

arthritis: ARCTIC randomised controlled strategy trial. BMJ, 2016. 354:

p. i4205.

55

81. Gartner, M., et al., Persistence of subclinical sonographic joint activity in

rheumatoid arthritis in sustained clinical remission. Ann Rheum Dis,

2015. 74(11): p. 2050-3.

82. Haugeberg, G., et al., Reduced bone mineral density in male rheumatoid

arthritis patients: frequencies and associations with demographic and

disease variables in ninety-four patients in the Oslo County Rheumatoid

Arthritis Register. Arthritis Rheum, 2000. 43(12): p. 2776-84.

83. Haugeberg, G., et al., Bone mineral density and frequency of osteoporosis

in female patients with rheumatoid arthritis: results from 394 patients in

the Oslo County Rheumatoid Arthritis register. Arthritis Rheum, 2000.

43(3): p. 522-30.

84. Lodder, M.C., et al., Bone mineral density in patients with rheumatoid

arthritis: relation between disease severity and low bone mineral density.

Ann Rheum Dis, 2004. 63(12): p. 1576-80.

85. Forslind, K., et al., Hand bone loss measured by digital X-ray

radiogrammetry is a predictor of joint damage in early rheumatoid

arthritis. Scand J Rheumatol, 2009. 38(6): p. 431-8.

86. Guler-Yuksel, M., et al., Changes in hand and generalised bone mineral

density in patients with recent-onset rheumatoid arthritis. Ann Rheum

Dis, 2009. 68(3): p. 330-6.

87. Hoff, M., et al., Cortical hand bone loss after 1 year in early rheumatoid

arthritis predicts radiographic hand joint damage at 5-year and 10-year

follow-up. Ann Rheum Dis, 2009. 68(3): p. 324-9.

88. Rezaei, H., et al., Evaluation of hand bone loss by digital X-ray

radiogrammetry as a complement to clinical and radiographic

assessment in early rheumatoid arthritis: results from the SWEFOT trial.

BMC Musculoskelet Disord, 2013. 14: p. 79.

89. Hammer, H.B. and T.K. Kvien, Ultrasonography shows significant

improvement in wrist and ankle tenosynovitis in rheumatoid arthritis

patients treated with adalimumab. Scand J Rheumatol, 2011. 40(3): p.

178-82.

90. Szkudlarek, M., et al., Interobserver agreement in ultrasonography of the

finger and toe joints in rheumatoid arthritis. Arthritis Rheum, 2003.

48(4): p. 955-62.

91. Jorgensen, J.T., et al., Digital X-ray radiogrammetry: a new appendicular

bone densitometric method with high precision. Clin Physiol, 2000.

20(5): p. 330-5.

92. Rosholm, A., et al., Estimation of bone mineral density by digital X-ray

radiogrammetry: theoretical background and clinical testing. Osteoporos

Int, 2001. 12(11): p. 961-9.

93. D'Agostino, M.A., et al., Novel algorithms for the pragmatic use of

ultrasound in the management of patients with rheumatoid arthritis: from

diagnosis to remission. Ann Rheum Dis, 2016. 75(11): p. 1902-1908.

56

94. Hoving, J.L., et al., A comparison of magnetic resonance imaging,

sonography, and radiography of the hand in patients with early

rheumatoid arthritis. J Rheumatol, 2004. 31(4): p. 663-75.

95. Verheul, M.K., et al., Triple Positivity for Anti-Citrullinated Protein

Autoantibodies, Rheumatoid Factor, and Anti-Carbamylated Protein

Antibodies Conferring High Specificity for Rheumatoid Arthritis:

Implications for Very Early Identification of At-Risk Individuals. Arthritis

Rheumatol, 2018.

96. Iagnocco, A., et al., Wrist joint involvement in systemic lupus

erythematosus. An ultrasonographic study. Clin Exp Rheumatol, 2004.

22(5): p. 621-4.

97. Wright, S., et al., Hand arthritis in systemic lupus erythematosus: an

ultrasound pictorial essay. Lupus, 2006. 15(8): p. 501-6.

98. Kastbom, A., et al., Anti-CCP antibody test predicts the disease course

during 3 years in early rheumatoid arthritis (the Swedish TIRA project).

Ann Rheum Dis, 2004. 63(9): p. 1085-9.

99. Tamas, M.M., et al., Bone erosions in rheumatoid arthritis: ultrasound

findings in the early stage of the disease. Rheumatology (Oxford), 2014.

53(6): p. 1100-7.

100. Grassi, W., et al., Ultrasonography in the evaluation of bone erosions.

Ann Rheum Dis, 2001. 60(2): p. 98-103.

101. Magnani, M., et al., Ultrasonography detection of early bone erosions in

the metacarpophalangeal joints of patients with rheumatoid arthritis.

Clin Exp Rheumatol, 2004. 22(6): p. 743-8.

102. Wakefield, R.J., et al., The value of sonography in the detection of bone

erosions in patients with rheumatoid arthritis: a comparison with

conventional radiography. Arthritis Rheum, 2000. 43(12): p. 2762-70.

103. Dohn, U.M., et al., Are bone erosions detected by magnetic resonance

imaging and ultrasonography true erosions? A comparison with

computed tomography in rheumatoid arthritis metacarpophalangeal

joints. Arthritis Res Ther, 2006. 8(4): p. R110.

104. Bugatti, S., et al., Anti-citrullinated protein antibodies and high levels of

rheumatoid factor are associated with systemic bone loss in patients with

early untreated rheumatoid arthritis. Arthritis Res Ther, 2016. 18(1): p.

226.

105. Kleyer, A., et al., Bone loss before the clinical onset of rheumatoid

arthritis in subjects with anticitrullinated protein antibodies. Ann Rheum

Dis, 2014. 73(5): p. 854-60.

106. Orsolini, G., et al., Titer-Dependent Effect of Anti-Citrullinated Protein

Antibodies On Systemic Bone Mass in Rheumatoid Arthritis Patients.

Calcif Tissue Int, 2017. 101(1): p. 17-23.

RESEARCH ARTICLE Open Access

Antibodies against carbamylated proteinsand cyclic citrullinated peptides in systemiclupus erythematosus: results from two well-defined European cohortsMichael Ziegelasch1, Myrthe A. M. van Delft2, Philip Wallin1, Thomas Skogh1, César Magro-Checa2,Gerda M. Steup-Beekman2, Leendert A. Trouw2, Alf Kastbom1 and Christopher Sjöwall1,3*

Abstract

Background: Articular manifestations are common in systemic lupus erythematosus (SLE) whereas erosive diseaseis not. Antibodies to cyclic citrullinated peptide (anti-CCP) are citrulline-dependent in rheumatoid arthritis (RA),whereas the opposite is suggested in SLE, as reactivity with cyclic arginine peptide (CAP) is typically present.Antibodies targeting carbamylated proteins (anti-CarP) may occur in anti-CCP/rheumatoid factor (RF)-negative caseslong before clinical onset of RA. We analysed these antibody specificities in sera from European patients with SLE inrelation to phenotypes, smoking habits and imaging data.

Methods: Cases of SLE (n = 441) from Linköping, Sweden, and Leiden, the Netherlands, were classified according toAmerican College of Rheumatology (ACR) and/or Systemic Lupus Erythematosus International Collaborating Clinics(SLICC) criteria. IgG anti-CCP, anti-CAP and anti-CarP were analysed by immunoassays. Radiographic data from 102Swedish patients were available.

Results: There were 16 Linköping (6.8%) and 11 Leiden patients (5.4%) who were anti-CCP-positive, of whomapproximately one third were citrulline-dependent: 40/441 (9.1%) were anti-CarP-positive, and 33% of theanti-CarP-positive patients were identified as anti-CCP-positive. No associations were found comparing anti-CCP oranti-CarP with ACR-defined phenotypes, immunologic abnormalities or smoking habits. Radiographically confirmederosions were found in 10 patients, and were significantly associated with anti-CCP, anti-CarP and RF.Musculoskeletal ultrasonography scores were higher in anti-CCP-positive compared to anti-CCP-negative patients.

Conclusions: In the hitherto largest anti-CarP study in SLE, we demonstrate that anti-CarP is more prevalent thananti-CCP and that the overlap is limited. We obtained some evidence that both autoantibodies seem to beassociated with erosivity. Similar pathogenetic mechanisms to those seen in RA may be relevant in a subgroup ofSLE cases with a phenotype dominated by arthritis.

Keywords: Anti-CarP, Anti-CCP, Rheumatoid arthritis, Rheumatoid factor, Systemic lupus erythematosus,Ultrasonography

* Correspondence: [email protected]/AIR, Department of Clinical and Experimental Medicine,Linköping University, Linköping, Sweden3Rheumatology Unit, University Hospital, Linköping SE-581 85, SwedenFull list of author information is available at the end of the article

© The Author(s). 2016 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Ziegelasch et al. Arthritis Research & Therapy (2016) 18:289 DOI 10.1186/s13075-016-1192-x

http://crossmark.crossref.org/dialog/?doi=10.1186/s13075-016-1192-x&domain=pdf

http://orcid.org/0000-0003-0900-2048

mailto:[email protected]

http://creativecommons.org/licenses/by/4.0/

http://creativecommons.org/publicdomain/zero/1.0/

BackgroundIn clinical practice, the diagnosis of systemic lupuserythematosus (SLE) is often based on the involvementof at least two organ systems combined with a diversityof immunological abnormalities [1]. The presence ofantinuclear antibodies (ANA) and reduced levels of cir-culating complement proteins are typical immunologicalabnormalities in SLE, and both of these are included inthe most recent proposal of classification criteria [2].Over the years, several attempts to link specific autoanti-bodies to certain clinical phenotypes of SLE have beenmade. For instance, antibodies against double-strandedDNA (dsDNA) and C1q are commonly found in lupusnephritis [3, 4], anti-Ro/SSA antibodies often coincidewith lupus-related rash and photosensitivity [5], andanti-phospholipid antibodies are frequently found inpatients with SLE who have thromboembolic events [6].Detection of antibodies against cyclic citrullinated pep-

tide (anti-CCP) is an important diagnostic and prognos-tic tool in arthritis, as it is highly specific for rheumatoidarthritis (RA) and predictive of erosive disease. While apositive anti-CCP test in RA is typically citrulline-dependent [7], it has been suggested that anti-CCP inother conditions is generally not, and thus, also reactswith the corresponding cyclic arginine peptide (anti-CAP) [8, 9]. During the last years, it has been repeatedlyshown that antibodies targeting carbamylated proteins(anti-CarP) may occur in anti-CCP/rheumatoid factor(RF) negative cases [10–12]. Like RF and anti-CCP, anti-CarP antibodies can also be detected many years beforethe onset of RA [13–15].The process of carbamylation is mediated by a chem-

ical reaction of cyanate with mainly lysine residues inproteins [16]. Cyanate is present in the body in equilib-rium with urea. Inflammation, smoking and renal failurehave been reported to increase the non-enzymatic post-translational modification in which cyanate binds tomolecules containing primary amine or thiol groups andforms carbamyl groups [16]. Carbamylation of proteinscan lead to the loss of tolerance with formation ofantibodies directed against carbamylated proteins (anti-CarP antibodies) in susceptible individuals [10–12].Over the last years, this novel group of autoantibodies

has been intensively studied in patients before the onsetof clinical RA symptoms and in patients with establishedRA, in relation to prognostic factors such as anti-CCP/RF and to disease outcomes (i.e. radiological progres-sion) [10, 17, 18]. Anti-CarP has also been recentlyreported in primary Sjögren’s syndrome, and found to bestrongly associated with increased focal lymphocyte infil-tration, formation of ectopic germinal centre-like struc-tures, and to the degree of affected salivary glandfunction [19]. To our knowledge, only two previoussmall studies have addressed the occurrence of anti-

CarP antibodies in SLE. Thus, Scinocca et al. reportedthe occurrence of anti-CarP (homocitrullinated fibrino-gen) antibodies in 49% of 81 patients with RA, but innone of 37 patients with SLE, 37 patients with psoriaticarthritis, or 27 healthy controls [20]. In the study byLópez-Hoyos et al., 48% (16/33) of anti-CCP/RF-nega-tive patients with elderly-onset RA were judgedseropositive for anti-CarP compared to 39% (48/124) ofpatients with polymyalgia rheumatica, and 11% (4/37) ofpatients with SLE [21].Although articular involvement is common in SLE (at

least 80%) and often constitutes the presentingsymptom, it has historically received limited attention[22, 23]. A plausible reason hereto is that arthritis causessignificant problems to the patient at disease onset, butoften responds rather well to treatment with glucocorti-coids, hydroxychloroquine and methotrexate [22].Development of deformities is occasionally seen, but thevast majority of patients with SLE do not develop radio-graphically evident erosions [22]. Accordingly, severalinvestigators have estimated the frequency of erosivearthritis in SLE to be 2–5%, but the finding of erosionson radiography often complicates the distinctionbetween SLE and RA [22, 24]. As a result, the conceptof “rhupus” was introduced to describe patients whosimultaneously fulfil classification criteria for both con-ditions [25, 26]. It is still a matter of debate as towhether this group represents a true overlap of RA andSLE, rather than a subset of SLE [27, 28].Hitherto, only a few studies of lupus arthritis based on

imaging modalities other than conventional radiographyhave been published. A limited number of studies haveevaluated magnetic resonance imaging (MRI) in the as-sessment of patients with articular involvement in SLE[29–32]. Recently, Ball and colleagues demonstrated thatMRI is highly sensitive in identifying erosions, synovitisand bone oedema, independently of anti-CCP and RFantibody status [32]. In SLE, there are some reportsbased on musculoskeletal ultrasonography (US), but sofar this modality has seldom been used in clinicalroutine practice [31, 33].The primary aim of this study was to describe the

presence of citrulline-dependent anti-CCP and anti-CarPantibodies in 441 well-characterized patients within twoEuropean SLE cohorts. Second, in the Swedish dataset,we aimed to investigate the associations between auto-antibody status and articular involvement as defined byclassification criteria, conventional radiographic data andevaluation with musculoskeletal US.

MethodsDiscovery cohortPatients diagnosed with SLE (n = 236) were included inthe discovery cohort. All patients took part in the

Ziegelasch et al. Arthritis Research & Therapy (2016) 18:289 Page 2 of 11

prospective, structured follow-up programme “KLUR-ING” (the Swedish acronym for Clinical Lupus RegisterIn Northeastern Gothia) at the rheumatology outpatientclinic, University Hospital, Linköping, Sweden. Thepatient material has previously been described in detail[3, 34]. Patients were recruited consecutively and classi-fied as having SLE according to the 1982 AmericanCollege of Rheumatology (ACR) criteria and/or the 2012Systemic Lupus International Collaborating Clinics clas-sification criteria (SLICC)-12 [2, 35]. Most patientsrepresented prevalent cases (83%), but 41 patients (17%)had newly diagnosed disease at inclusion. Informationon smoking habits (current/former/never) and activitylimitations defined according to the validated Swedishversion of the health assessment questionnaire (HAQ)was recorded at the time of blood sampling [36]. Furtherpatient characteristics are summarized in Table 1.It should be emphasized that the definition of arthritis

slightly differs between the ACR-82 and SLICC-12 cri-teria. Whereas ACR-82 requires “non-erosive arthritisinvolving ≥2 peripheral joints, characterized by tender-ness, swelling, or effusion”, the presence of “synovitisinvolving 2 or more joints, characterized by swelling oreffusion OR tenderness in 2 or more joints and at least30 minutes of morning stiffness” is demanded to meetthe SLICC-12 arthritis criterion.

Replication cohortA total of 205 consecutive patients with SLE wereincluded in the replication cohort. All patients wereselected from the Leiden NPSLE clinic, at the rheuma-tology department, Leiden University Medical Center(LUMC), the Netherlands. The LUMC serves as anational referral centre for patients with SLE suspectedof having neuropsychiatric involvement. All patientsincluded were admitted for one day and underwent acomplete examination that has previously beendescribed in detail [37]. All patients were classified ashaving SLE according to the 1997 updated ACR-82 and/or the SLICC-12 criteria [2, 38]. Data on gender, smok-ing status and ethnicity were retrospectively recorded.Detailed patient characteristics relevant to the presentstudy are summarized in Table 1.The following definitions of smoking habits were used

in both cohorts [39]. “Former smoker” means to haveever been a regular smoker; occasional smokers werenot included. “Ever smoker” constitutes the sum of“former smoker” and “current smoker”.

Laboratory analysesIn the discovery cohort, IgG-class anti-CCP and anti-CAP antibodies were analysed by enzyme-linked im-munosorbent assays (CCPlus and anti-CAP respectively,Euro-Diagnostica, Malmö, Sweden). Levels ≥25 units/

mL defined a positive test as suggested by the manufac-turer. Citrulline dependency was defined as a higherantibody level obtained in the anti-CCP test than in theanti-CAP test. RF was detected by nephelometry at theclinical immunology department, University Hospital,Linköping, Sweden.For the replication cohort, anti-CCP and anti-CAP

antibodies were analysed using an in-house CCP2 assaywith a defined cutoff ≥25 units/mL [40]. Healthy con-trols (n = 193) who were living in the Leiden area, wereincluded for comparison. Briefly, biotinylated CCP2citrulline and corresponding arginine control peptidewere coupled to streptavidin-coated enzyme-linked im-munosorbent assay (ELISA) plates. Serum samples werediluted 1:50 and bound IgG was detected using rabbitanti-human IgG horseradish peroxidase (HRP) (DAKO,Glostrup, Denmark).Analysis of IgG anti-CarP antibody in both cohorts

was performed in Leiden, using a previously describedimmunoassay detecting antibodies against carbamylatedfetal calf serum proteins [10]. Other laboratory analyses,including IgM RF quantified by fluoroenzyme immuno-assay, were performed at the routine diagnostic labora-tory of the Leiden University Medical Center, Leiden,the Netherlands.

RadiologyThrough meticulous chart review of each patient, weidentified all conventional radiographs of the hands,wrists and/or feet, obtained at the University Hospital,Linköping (discovery cohort). The results were scruti-nized by an experienced rheumatologist (MZ) who wasblinded to the patients’ antibody status. In cases whereerosions had been identified by the radiologist and in allcases with indistinct findings, the radiographs werefinally evaluated by MZ to determine whether erosionswere present.

Musculoskeletal USIn the discovery cohort, all anti-CCP-positive cases plusjust as many anti-CCP-negative cases, matched accord-ing to sex, age, disease duration and present prednisol-one dose, were systematically investigated withmusculoskeletal US by an experienced examiner (MZ)blinded to the patients’ antibody profiles and conven-tional radiography results. Characteristics of the 32patients, representing a subgroup of the discoverycohort, are shown in Table 2.US examination was performed using the LOGIQ-E9

(GE Healthcare, Milwaukee, WI, USA) with a linear scan-ner 6–15 MHz. All patients were examined with the samesettings for both B-mode (grey scale) and power Doppler(PD). The protocol included dorsal assessments of thefollowing 36 joints: bilateral radiocarpal, intercarpal, distal


Table 1 Clinical and immunological features of the 441 included patients

Patient characteristics Discovery cohort(n = 236)

Replication cohort(n = 205)

Mann–Whitney Utest or chi-squaretest, p value

Background variables

Female, % (n) 88% (207) 89% (183) N.S.

Age, years, median (range) 54 (19–94) 42 (13–80) <0.001

Disease duration, years, median (range) 15 (0–52) 8 (0–32) <0.001

Caucasian ethnicity, % (n) 93% (219) 69% (142) <0.001

Conventional radiology available, % (n) 43% (102) N.A.

Erosions on x-ray, % (n) 4.2% (10)a N.A.

HAQ score (median, range) 0.13 (0-3) N.A.

Ever smoker (former or current), % (n) 45% (107) 43% (88) N.S.

Meeting ACR-82, % (n) 85% (201) 91% (188) 0.038

Meeting SLICC-12, % (n) 99% (233) 99% (203) N.S.

Clinical phenotypes (SLICC-12 definitions on criteria 1 − 11), % (n)

1) Acute cutaneous lupus 45% (106) 53% (108) N.S.

2) Chronic cutaneous lupus 15% (36) 18% (36) N.S.

3) Oral ulcers 12% (29) 32% (66) <0.001

4) Non-scarring alopecia 22% (51) 19% (39) N.S.

5) Synovitis 76% (180) 70% (143) N.S.

6) Serositis 37% (87) 23% (48) 0.003

Pleuritis 35% (83) 18% (37) <0.001

Pericarditis 14% (34) 14% (28) N.S.

7) Renal 28% (66) 26% (53) N.S.

8) Neurologic 11% (26) 22% (44) 0.004

Seizures 4.2% (10) 5.4% (11) N.S.

Psychosis 1.7% (4) 4.4% (9) N.S.

Mononeuritis multiplex 0.4% (1) 0% (0) N.S.

Myelitis 0.4% (1) 2.9% (6) N.S.

Peripheral or cranial neuropathy 5.1% (12) 8.3% (17) N.S.

Acute confusional state 0.8% (2) 2% (4) N.S.

9) Haemolytic anaemia 4.7% (11) 5.9% (12) N.S.

10) Leukopenia and/or lymphopenia 52% (122) 30% (62) <0.001

11) Thrombocytopenia 12% (28) 18% (37) N.S.

Raynaud 26% (61) 40% (82) 0.002

Interstitial lung disease 3.5% (8) 3.4% (7) N.S.

Immunological features (SLICC-12 definitions on criteria 1 to 6), % (n)

1) Antinuclear antibody (ANA) 100% (236) 100% (205) N.S.

2) Anti-dsDNA antibody (anti-dsDNA) 50% (118) 56% (115) N.S.

3) Anti-Smith antibody (anti-Sm) 8.1% (19) 10% (21) N.S.

4) Antiphospholipid antibody 59% (139) 44% (90) 0.002

Lupus anticoagulant 35% (69)b 32% (65) N.S.

Anti-cardiolipin antibody 34% (80) 24% (49) 0.027

Anti-β2-glycoprotein I antibody 26% (62) 18% (27)c N.S.

5) Low complement 53% (124) 50% (102) N.S.

6) Direct Coombs test 56% (59)d 23% (38)e <0.001


radioulnar, metacarpophalangeal (MCP) I–V, interphalan-geal (IP) thumb joints, proximal interphalangeal (PIP II-V),and metatarsophalangeal (MTP) joint I-V. In addition, sixtendons (extensor carpi ulnaris in the wrist, tibialis poster-ior and flexor digitorum longus in the feet) were evaluated.To grade synovitis, we used the scoring system ofSzkudlarek et al. in which synovial hypertrophy is graded0–3 (0 = no thickening, 3 = synovial thickening bulging overthe tops of the periarticular bones and extension over thediaphysis of at least one side) and perfusion 0–3 (0 = noDoppler signals in the synovium, 3 = confluent Dopplersignals in more than one half of the synovium) [41, 42].

StatisticsPotential associations between antibody status and clin-ical characteristics were tested by Fisher’s exact test forcategorical variables and by the Mann-Whitney U testfor numerical variables. The Mann-Whitney U test orchi-square test was used to evaluate differences betweenthe cohorts. Statistical analyses were performed usingSPSS v23.

For analyses where we had prior hypotheses, a signifi-cance level of 5% was regarded as statistically significant(two-sided p values <0.05). For all other tests performedin a more exploratory manner, the exact p values (if pwas <0.05) are reported as the reference.

ResultsComparison between cohortsAs shown in Table 1, the size of the two cohorts wassimilar, whereas in some instances there were significantdifferences in the clinical phenotypes according to theclassification criteria that were fulfilled (oral ulcers, sero-sitis, neurological involvement, Raynaud). Significantlymore patients in the discovery cohort were older, hadlonger disease duration, and were Caucasian than in thereplication cohort. In addition, laboratory criteria suchas the presence of leukopenia/lymphocytopenia, anti-phospholipid antibody, anti-snRNP antibody, anti-La/SSB antibody, RF and the direct Coombs test differedbetween the cohorts.

Table 2 Characteristics of the 32 discovery cohort patients evaluated with musculoskeletal ultrasonography

Anti-CCP-positive(n = 16)

Anti-CCP-negative(n = 16)

P value

Age, years, median 58 58 N.S.

Disease duration, years, median 10.5 10.5 N.S.

Female, number 14 13 N.S.

Fulfilled ACR-82 criteria number, median 4 4 N.S.

Meeting arthritis criterion according to ACR-82, number 13 14 N.S.

Meeting nephritis criterion according to ACR-82, number 4 2 N.S.

Conventional radiographs available (hands/wrists/feet) 16 16 N.S.

Erosions on radiography, number of individuals 4 0 N.S.

HAQ score, median 0.5 0.19 N.S.

Ever smoker, number of individuals 9 3 N.S.

Rheumatoid factor positive, number of individuals 6 1 N.S.

anti-CarP antibody level, median 275 102 0.04

anti-CarP positive, number of individuals 6 1 N.S.

Daily dosage of prednisolone, median 2.5 2.5 N.S.

ACR American College of Rheumatology, HAQ health assessment questionnaire, anti-CarP antibodies targeting carbamylated proteins, N.S. not significant

Table 1 Clinical and immunological features of the 441 included patients (Continued)

Anti-small nuclear ribonucleoprotein antibody (anti-snRNP) 37% (88) 17% (35) <0.001

Anti-Ro/Sjögren’s syndrome A antibody (SSA) 37% (88) 42% (87) N.S.

Anti-La/Sjögren’s syndrome B antibody (SSB) 28% (66) 13% (27) <0.001

Anti-cyclic citrulline peptide antibody (anti-CCP)f 6.8% (16) 5.4% (11) N.S.

Anti-carbamylated protein antibody (anti-CarP) 9.8% (23) 8.3% (17) N.S.

Rheumatoid factorf 25% (26)d 15% (31) 0.046aCalculated in 236 patients. bData available on 195 patients. cData available on 148 patients. dData available on 107 patients. eData available on 164 patients. fNotperformed with identical assays. HAQ health assessment questionnaire, ACR American College of Rheumatology, SLICC Systemic Lupus Erythematosus InternationalCollaborating Clinics, ANA antinuclear antibodies, dsDNA double-stranded DNA, N.A. not applicable, N.S. not significant


Presence of anti-CCP/CAP/CarP antibodies in SLEIn the discovery cohort, 16 patients (6.8%) were anti-CCP-positive, 9 (56%) of whom were also anti-CAP-positive using Euro-Diagnostica kits; however, only oneof the 9 anti-CCP/anti-CAP-positive patients had ahigher antibody level for anti-CAP than for anti-CCP inthe assays: 4 of the 7 patients with a positive citrulline-dependent anti-CCP test had a history of biopsy-provenlupus nephritis. There were 23 anti-CarP-positivepatients (9.8%); only 6 (26%) of the anti-CarP-positivepatients were identified as anti-CCP-positive (Fig. 1a).In the replication cohort, 11 patients (5.4%) were anti-

CCP-positive, of whom 8 (73%) were also positive foranti-CAP using in-house assays; in 4 of the 8 anti-CCP/

anti-CAP-positive patients, there was a higher opticaldensity signal for anti-CAP than for anti-CCP. Therewere 17 anti-CarP-positive patients (8.3%), of whom 2(12%) were also identified as anti-CCP-positive (Fig. 1b).A limited number of samples from the discovery cohortwere also analysed with the anti-CCP in-house assay inLeiden. The agreement between the assays was fair (rho= 0.75; concordance 90%).

Associations between anti-CCP, anti-CarP, or RF and clin-ical or other laboratory featuresIn the discovery cohort, neither anti-CCP nor anti-CarPwere associated with arthritis by classification accordingto physical examination. However, the presence of anti-

Fig. 1 a-b Distribution of anti-carbamylated protein (anti-CarP)-positive, anti-cyclic citrullinated peptide (anti-CCP)-positive and anti-cyclic argininepeptide (anti-CAP)-positive patients in the discovery cohort (n = 236) (a), and in the replication cohort (n = 205) (b)


CarP was associated with neurological involvement asdefined by SLICC-12 (p = 0.028); this association wasparticularly driven by the presence of cranial/peripheralneuropathy (p = 0.021). On the contrary, anti-CarP wasless common among patients with a positive lupusanticoagulant test (p = 0.012). Positive anti-CCP testswere inversely related to anti-La/SSB (p = 0.007).Smoking habits and HAQ score were not associatedwith anti-CCP, anti-CarP or RF.In the replication cohort, the presence of anti-CarP

was not significantly associated with neurologicalinvolvement as defined by SLICC-12 (p = 0.126); andthere was no association with specific neurological mani-festations included in the SLICC-12 criteria. Anti-CarPwas not associated with any other SLICC-12 definitionor SLE autoantibody. On the contrary, anti-CCP was morecommon among patients with positive anti-Sm (p = 0.017).There was no association between anti-CCP positivity andanti-La/SSB. Smoking habits were not associated with anti-CCP, anti-CarP or RF.

Associations between anti-CCP, anti-CarP or RF, andradiographyIn the Swedish cohort, conventional radiographs of thehands, wrists and/or feet were available in 102 patients(43%), and erosions (mostly demonstrated in the inter-carpal, MCP and MTP joints) were identified in 10patients. This corresponds to 4.2% calculated for theentire Swedish cohort, and to 9.8% calculated strictlyfor patients with radiographs available. Radiological

erosions were significantly associated with a positiveanti-CCP (p < 0.05), anti-CarP (p < 0.05) or RF (p < 0.05)test, respectively.All patients with a positive anti-CCP test underwent

radiography of the hands, wrists and feet. Thus, patientswith radiographs available were more often anti-CCP-positive (16/102; 16%) as compared to patients withoutradiographs (0/134; p < 0.001). Similarly, “ever”’ smokerswere significantly overrepresented among patients withradiographs (55/102; 54%) compared to those without(52/134; 39%; p = 0.025). There were no significantdifferences in age, sex or fulfilment of the SLICC-12arthritis criterion. Neither did the occurrence of anti-CarP, anti-CAP or RF significantly differ according tothe availability of radiographic data.

Associations between anti-CCP, anti-CarP or RF, and mus-culoskeletal USIn anti-CCP-positive patients, US examination of thejoints and tendons resulted in significantly higher arth-ritis scores, but not tenosynovitis scores (Fig. 2a). Therewere similar trends for anti-CarP and RF, but thesewere not statistically significant (Fig. 2b–c). Patientswith radiological erosions had significantly higher arth-ritis scores, but not tenosynovitis scores (Fig. 2d).Individuals with ongoing prednisolone medication hadhigher arthritis, but not tenosynovitis scores (Fig. 2e),whereas fulfilment of the ACR criterion for arthritis didnot separate the groups (Fig. 2f ).

Fig. 2 a-f Standardized musculoskeletal ultrasonography grading of arthritis and tenosynovitis with grey scale (B-mode) and power Doppler (PD)in 32 patients divided according to the presence of anti-cyclic citrullinated peptide (anti-CCP) (a), anti-carbamylated protein antibody (anti-CarP)(b), rheumatoid factor (c), radiographically confirmed erosions (d), daily intake of prednisolone (e), and with regard to the fulfilment of the 1982American College of Rheumatology (ACR) classification criterion 5 (arthritis) (f) [35]. Median and interquartile range are illustrated. * p < 0.05


Associations between anti-CCP or anti-CarP, and the fulfil-ment of RA classificationBesides meeting the classification criteria for SLE, weevaluated to what extent patients with a positive anti-CCP and/or anti-CarP antibody test simultaneouslyfulfilled the 2010 ACR/European League AgainstRheumatism (EULAR) criteria for RA [43]. As illustrated(Fig. 3), the likelihood of meeting the RA criteria washigher with a positive anti-CCP compared to a positiveanti-CarP antibody test. However, when taking citrullinedependency into account for the anti-CCP test thisdifference was less pronounced, particularly in thediscovery cohort.

DiscussionThis is hitherto the largest evaluation of anti-CarP anti-bodies in SLE, and the first study on anti-CarP in aEuropean SLE population. Herein, we demonstratedsimilar frequencies of anti-CarP-positive SLE patients inthe two cohorts (9.8% vs. 8.3%) and that the overlap withanti-CCP antibodies is limited. Our findings are in linewith what has been reported by López-Hoyos et al., butclearly higher than observed by Scinocca and co-workers[20, 21]. The latter may be explained by a difference inthe antigen used for the detection of anti-CarP anti-bodies (fibrinogen vs. fetal calf serum). Furthermore, wefound significant associations between all three RA-asso-ciated antibodies (anti-CCP, anti-CarP and RF) andradiographically confirmed erosions in the Swedish

dataset. Based on the results, we hypothesize that patho-genetic mechanisms could be similar in RA and in asmall group of patients with SLE with a clinical pheno-type dominated by arthritis [44]. Interestingly though,60% of the patients with radiology proven erosions werenot identified by any of the antibodies.Articular manifestations affect a majority of patients

with SLE, at least at some time during the disease course(73% in the present study). However, only a minority ofthe patients with SLE who have an arthritic phenotypesimultaneously meet RA classification criteria [24, 25,31]. The presence of anti-CCP antibodies is consideredhighly specific for RA, but can also be found in otherconditions, including SLE, where frequencies from 2–17% have been described [9, 32, 45–51]. Whether or notthere is a true association between a positive anti-CCPtest and erosive arthritis in SLE remains an open ques-tion, as several investigators have reported this [9, 46–50], whereas others have not [31, 45]. Kakumanu et al.reported a prevalence of 17% for anti-CCP positivityamong 329 patients with SLE but that citrulline-dependent anti-CCP was mainly found in patients witherosive arthritis, which involved only 26 patients [9].Pooled data from the present study indicate that 10 of27 anti-CCP-positive patients (37%) had citrulline-dependent anti-CCP, which corresponds to 2.3% in thewhole study population.As previously mentioned, smoking affects the carba-

mylation process [16] and smoking is also strongly

Fig. 3 Percentage of patients who, besides being classified as having systemic lupus erythematosus (SLE), also fulfilled the 2010 American Collegeof Rheumatology/European League Against Rheumatism classification criteria for rheumatoid arthritis (RA) [43] provided they were: (1) anti-cycliccitrullinated peptide (anti-CCP) antibody-positive, (2) identified as citrulline-dependent anti-CCP antibody-positive (i.e. higher anti-CCP than cyclicarginine peptide (anti-CAP) antibody level), and (3) anti-carbamylated protein (anti-CarP) antibody-positive


associated with anti-CCP-positive RA, especially inpatients with the HLA-DRB1/shared epitope [52]. Thus,we anticipated that smoking would also be overrepre-sented among anti-CarP-positive and anti-CCP-positivepatients with SLE. To our surprise, we did not detectsuch associations in this evaluation. In line with ourfindings, however, recent studies from Stockholm didnot identify any correlation between smoking habits andanti-CCP in SLE [51], and anti-CarP was not signifi-cantly associated with smoking in patients with RA [11].Of potential interest in relation to these results, thefrequency of “ever” smokers in the discovery and replica-tion cohorts was similar (approximately 45%), butwas clearly lower than in two early RA cohorts (TIRA-2and EIRA-1) from southern and central Sweden, wherethe frequency reached approximately 60% [52].Evaluations with musculoskeletal US were performed

in 32 patients with SLE (whereof 16 were anti-CCP posi-tive) in the discovery cohort, showing significantlyhigher arthritis scores among anti-CCP-positive cases(Fig. 2a). Previous studies on musculoskeletal US in SLEwith relevant comparators are scarce and a considerablestrength in the present study is the matched controlgroup (Table 2). Some studies have reported a highernumber of affected tendons in SLE, compared to whatwe found [31, 53–55]. Interestingly, the effects of corti-costeroids on B-mode/PD arthritis scores seemed to belimited (Fig. 2e).We also investigated which RA-associated

autoantibody-positive patients with SLE who simultan-eously met the classification criteria for RA. This groupcould be referred to as “rhupus”, although there is noagreed definition for this overlap condition [25–27].Although anti-citrullinated protein antibodies (ACPA)are indeed included in the 2010 ACR/EULAR criteriafor RA, we used this most recent RA classification [43].We found the risk of meeting RA criteria was highergiven a positive anti-CCP test as compared to a positiveanti-CarP antibody test (52% vs. 20%); when takingcitrulline dependency into account for the anti-CCP test,this difference appeared to decrease. We consideredusing the 1987 ACR RA criteria (which does not includeACPA), but we were unable to retrieve sufficient data todetermine fulfilment of this classification.Apart from the statistically significant connection

between a positive anti-CarP test and erosive arthritis inour study, the presence of anti-CarP was associatedsignificantly with neurological involvement only in thediscovery cohort, whereas a non-significant trend wasfound in the replication cohort. This may be due to thesmall number of patients with this clinical symptom.However, it is tempting to speculate that carbamylationmay have an effect on neural tissue. Of relevance to thisis the fact that supplementation of sodium cyanate to

patients suffering from sickle cell anaemia disrupts theequilibrium between urea and cyanate in body fluids,resulting in increased carbamylation and severe poly-neuropathy as side effects [56]. Furthermore, in animalmodels, cognitive impairment and loss of memoryfunctions have been described as a result of increasedcarbamylation [57].Whether or not anti-CarP antibodies can also be

found in patients receiving sodium cyanate is unknown.It should be emphasized that both irreversibleneuropathy and cognitive impairment are recorded asSLE damage accrual in the neuropsychiatric domain ofthe SLICC/ACR damage index [58]. Although renalfunction also affects the carbamylation process, we didnot find any association between anti-CarP and renal in-volvement. The reason hereto is most likely that renalflares in well-controlled SLE are identified early andtreated aggressively, resulting in a very small number ofpatients having end-stage renal disease [59]. In addition,carbamylation alone may not be sufficient for a break oftolerance against carbamylated proteins [60].Although radiology data were at hand for all cases

with a positive anti-CCP test, this was the case only fora minority of patients in the discovery cohort and not atall in the replication cohort, which is a limitation of ourstudy. Furthermore, in the musculoskeletal US evalua-tions, it is likely that a larger number of evaluatedpatients with different autoantibody profiles would haveenabled statistically significant and meaningful differ-ences in anti-CarP and RF to be identified. Finally, anti-CCP, anti-CAP and RF were analysed with differentassays in the two cohorts. The agreement with anti-CCPwas fair, but we cannot exclude small methodologicaldifferences that might have influenced the results tosome extent.

ConclusionsIn summary, we demonstrated that anti-CarP is morecommon than anti-CCP in well-characterised SLE intwo European cohorts, whereas the overlap of these anti-body specificities is limited. In the Swedish dataset weobtained some evidence that both autoantibodies areassociated with erosive joint disease.

AcknowledgementsThe authors thank research nurse Marianne Peterson and all the clinicians atthe Rheumatology clinic, University Hospital, Linköping, for their work. Inaddition, we thank Marije K. Verheul, LUMC, Leiden, the Netherlands for helpand advice in running the autoantibody assays.

FundingThe work was supported by grants from the County Council of Östergötland,the Swedish Society for Medical Research, the Swedish RheumatismAssociation, the Swedish Society of Medicine, the Professor Nanna Svartzfoundation and the King Gustaf V 80-year foundation. We also acknowledgethe financial support from the Dutch Arthritis Foundation and the IMI JU


funded project, BeTheCure, contract no 115142-2. LAT is supported by aZON-MW Vidi grant.

Availability of data and materialsAll data generated or analysed during this study are included in thispublished article.

Authors’ contributionsMZ was involved in conception and design of the study, substantial datacollection (imaging) and manuscript writing. PW was involved in substantialdata collection, design of figures and critical review of the manuscript. TSwas involved in conception and design of the study, and manuscript writing.MAD was involved in substantial data collection and critical review of themanuscript. CMC was involved in data collection, acquisition of data andcritical review of the manuscript. GMSB was involved in data collection,acquisition of data and critical review of the manuscript. LAT was involved inconception and design of the study, data collection (laboratory), acquisitionof data and critical review of the manuscript. AK was involved in conceptionand design of the study, data collection, acquisition of data and manuscriptwriting. CS was involved in conception and design of the study, datacollection, acquisition of funding and data and manuscript writing. Allauthors read and approved the final manuscript.

Competing interestsLAT is listed as an inventor on a patent application for the detection of anti-CarPantibodies in RA. The other authors declare that they have no conflicts of interest.

Consent for publicationNot applicable.

Ethical approval and consent to participateOral and written informed consent was obtained from all subjects with SLE.The study protocol was approved by the Regional Ethics Review Board inLinköping (M75-08/2008) and the local medical ethics committee of theLeiden University Medical Center (P11.191).

Author details1Rheumatology/AIR, Department of Clinical and Experimental Medicine,Linköping University, Linköping, Sweden. 2Department of Rheumatology,Leiden University Medical Center, C1-R, LUMC, PO Box 9600, Leiden 2300, RC,The Netherlands. 3Rheumatology Unit, University Hospital, Linköping SE-58185, Sweden.

Received: 17 September 2016 Accepted: 22 November 2016

References1. Fries JF, Holman HR. Systemic lupus erythematosus: a clinical analysis. In:

Smith LH, editor. Major Problems in Internal Medicine. Philadelphia, London,Toronto: W.B. Sauners; 1975. p. 8–20.

2. Petri M, Orbai AM, Alarcón GS, et al. Derivation and validation of theSystemic Lupus International Collaborating Clinics classification criteria forsystemic lupus erythematosus. Arthritis Rheum. 2012;64:2677–86.

3. Enocsson H, Sjöwall C, Wirestam L, Dahle C, Kastbom A, Rönnelid J, WetteröJ, Skogh T. Four anti-dsDNA antibody assays in relation to systemic lupuserythematosus disease specificity and activity. J Rheumatol. 2015;42:817–25.

4. Trouw LA, Daha MR. Role of anti-C1q autoantibodies in the pathogenesis oflupus nephritis. Expert Opin Biol Ther. 2005;5:243–51.

5. Frodlund M, Dahlström Ö, Kastbom A, Skogh T, Sjöwall C. Associationsbetween antinuclear antibody staining patterns and clinical features ofsystemic lupus erythematosus: analysis of a regional Swedish register.BMJ Open. 2013;3, e003608.

6. Burgos PI, Alarcón GS. Thrombosis in systemic lupus erythematosus: risk andprotection. Expert Rev Cardiovasc Ther. 2009;7:1541–9.

7. van Venrooij WJ, van Beers JJ, Pruijn GJ. Anti-CCP antibodies: the past, thepresent and the future. Nat Rev Rheumatol. 2011;7:391–8.

8. Kakumanu P, Yamagata H, Sobel ES, Reeves WH, Chan EK, Satoh M. Patientswith pulmonary tuberculosis are frequently positive for anti-cycliccitrullinated peptide antibodies, but their sera also react with unmodifiedarginine-containing peptide. Arthritis Rheum. 2008;58:1576–81.

9. Kakumanu P, Sobel ES, Narain S, et al. Citrulline dependence of anti-cycliccitrullinated peptide antibodies in systemic lupus erythematosus as amarker of deforming/erosive arthritis. J Rheumatol. 2009;36:2682–90.

10. Shi J, Knevel R, Suwannalai P, et al. Autoantibodies recognizing carbamylatedproteins are present in sera of patients with rheumatoid arthritis and predictjoint damage. Proc Natl Acad Sci USA. 2011;108:17372–7.

11. Jiang X, Trouw LA, van Wesemael TJ, et al. Anti-CarP antibodies in twolarge cohorts of patients with rheumatoid arthritis and their relationshipto genetic risk factors, cigarette smoking and other autoantibodies.Ann Rheum Dis. 2014;73:1761–8.

12. Koppejan H, Trouw LA, Sokolove J, Sokolove J, Lahey LJ, Huizinga TJ, Smolik IA,Robinson DB, El-Gabalawy HS, Toes RE, Hitchon CA. Role of Anti-carbamylatedprotein antibodies compared to anti-citrullinated protein antibodies inindigenous north Americans with rheumatoid arthritis, their first-degreerelatives, and healthy controls. Arthritis Rheumatol. 2016;68:2090–8.

13. Shi J, van de Stadt LA, Levarht EW, Huizinga TW, Hamann D, vanSchaardenburg D, Toes RE, Trouw LA. Anti-carbamylated protein (anti-CarP)antibodies precede the onset of rheumatoid arthritis. Ann Rheum Dis.2014;73:780–3.

14. Brink M, Verheul MK, Rönnelid J, Berglin E, Holmdahl R, Toes RE, Klareskog L,Trouw LA, Rantapää-Dahlqvist S. Anti-carbamylated protein antibodies inthe pre-symptomatic phase of rheumatoid arthritis, their relationship withmultiple anti-citrulline peptide antibodies and association with radiologicaldamage. Arthritis Res Ther. 2015;17:25.

15. Gan RW, Trouw LA, Shi J, et al. Anti-carbamylated protein antibodies arepresent prior to rheumatoid arthritis and are associated with its futurediagnosis. J Rheumatol. 2015;42:572–9.

16. Shi J, van Veelen PA, Mahler M, et al. Carbamylation and antibodies againstcarbamylated proteins in autoimmunity and other pathologies. AutoimmunRev. 2014;13:225–30.

17. Montes A, Regueiro C, Perez-Pampin E, Boveda MD, Gomez-Reino JJ,Gonzalez A. Anti-carbamylated protein antibodies as a reproducibleindependent type of rheumatoid arthritis autoantibodies. PLoS One.2016;11, e0161141.

18. Ajeganova S, van Steenbergen HW, Verheul MK, et al. The associationbetween anti-carbamylated protein (anti-CarP) antibodies and radiographicprogression in early rheumatoid arthritis: a study exploring replication andthe added value to ACPA and rheumatoid factor. Ann Rheum Dis. 2016.doi:10.1136/annrheumdis-2015-208870.

19. Bergum B, Koro C, Delaleu N, et al. Antibodies against carbamylatedproteins are present in primary Sjögren’s syndrome and are associated withdisease severity. Ann Rheum Dis. 2016;75:1494–500.

20. Scinocca M, Bell DA, Racapé M, et al. Antihomocitrullinated fibrinogenantibodies are specific to rheumatoid arthritis and frequently bindcitrullinated proteins/peptides. J Rheumatol. 2014;41:270–9.

21. López-Hoyos M, Álvarez-Rodríguez L, Mahler M, Torices S, Calvo-Alén J, VillaI, Seaman A, Yee A, Martínez-Taboada V. Anti-carbamylated proteinantibodies in patients with ageing associated inflammatory chronicdisorders. Rheumatology (Oxford). 2016;55:764–6.

22. Grossman JM. Lupus arthritis. Best Pract Res Clin Rheumatol. 2009;23:495–506.23. Petri M. Musculoskeletal complications of systemic lupus erythematosus in the

Hopkins Lupus Cohort: an update. Arthritis Care Res (Hoboken). 1995;8:137–45.24. Richter Cohen M, Steiner G, Smolen JS, Isenberg DA. Erosive arthritis in

systemic lupus erythematosus: analysis of a distinct clinical and serologicalsubset. Br J Rheumatol. 1998;37:421–4.

25. Panush RS, Edwards NL, Longley S, Webster E. ‘Rhupus’ syndrome. ArchIntern Med. 1988;148:1633–6.

26. Fernandez A, Quintana G, Rondon F, et al. Lupus arthropathy: a case seriesof patients with rhupus. Clin Rheumatol. 2006;25:164–7.

27. Amezcua-Guerra LM. Overlap between systemic lupus erythematosus andrheumatoid arthritis: is it real or just an illusion? J Rheumatol. 2009;36:4–6.

28. McGonagle D, Tan AL. What magnetic resonance imaging has told us aboutthe pathogenesis of rheumatoid arthritis – the first 50 years. Arthritis ResTher. 2008;10:222.

29. Ostendorf B, Scherer A, Specker C, Modder U, Schneider M. Jaccoud’sarthropathy in systemic lupus erythematosus: differentiation of deformingand erosive patterns by magnetic resonance imaging. Arthritis Rheum.2003;48:157–65.

30. Sa Ribeiro D, Galvao V, Luiz Fernandes J, et al. Magnetic resonance imagingof Jaccoud’s arthropathy in systemic lupus erythematosus. Joint Bone Spine.2010;77:241–5.


http://dx.doi.org/10.1136/annrheumdis-2015-208870

31. Tani C, D’Aniello D, Delle Sedie A, et al. Rhupus syndrome: assessment of itsprevalence and its clinical and instrumental characteristics in a prospectivecohort of 103 SLE patients. Autoimmun Rev. 2013;12:537–41.

32. Ball EM, Tan AL, Fukuba E, McGonagle D, Grey A, Steiner G, Bell AL, RooneyMR. A study of erosive phenotypes in lupus arthritis using magneticresonance imaging and anti-citrullinated protein antibody, anti-RA33 and RFautoantibody status. Rheumatology (Oxford). 2014;53:1835–43.

33. Zayat AS, Md Yusof MY, Wakefield RJ, Conaghan PG, Emery P, Vital EM. Therole of ultrasound in assessing musculoskeletal symptoms of systemic lupuserythematosus: a systematic literature review. Rheumatology (Oxford).2016;55:485–94.

34. Ighe A, Dahlström O, Skogh T, Sjöwall C. Application of the 2012 SystemicLupus International Collaborating Clinics classification criteria to patients ina regional Swedish systemic lupus erythematosus register. Arthritis ResTher. 2015;17:3.

35. Tan EM, Cohen AS, Fries JF, et al. The 1982 revised criteria for the classificationof systemic lupus erythematosus. Arthritis Rheum. 1982;25:1271–7.

36. Ekdahl C, Eberhardt K, Andersson SI, Svensson B. Assessing disability inpatients with rheumatoid arthritis. Use of a Swedish version of the StanfordHealth Assessment Questionnaire. Scand J Rheumatol. 1988;17:263–71.

37. Zirkzee EJ, Steup-Beekman GM, van der Mast RC, et al. Prospective study ofclinical phenotypes in neuropsychiatric systemic lupus erythematosus;multidisciplinary approach to diagnosis and therapy. J Rheumatol.2012;39:2118–26.

38. Hochberg MC. Updating the American College of Rheumatology revisedcriteria for the classification of systemic lupus erythematosus. ArthritisRheum. 1997;40:1725.

39. Stolt P, Bengtsson C, Nordmark B, Lindblad S, Lundberg I, Klareskog L,Alfredsson L, EIRA study group. Quantification of the influence of cigarettesmoking on rheumatoid arthritis: results from a population basedcase-control study, using incident cases. Ann Rheum Dis. 2003;62:835–41.

40. Kerkman PF, Fabre E, van der Voort EI, et al. Identification andcharacterisation of citrullinated antigen-specific B cells in peripheral bloodof patients with rheumatoid arthritis. Ann Rheum Dis. 2016;75:1170–76.

41. Szkudlarek M, Court-Payen M, Strandberg C, Klarlund M, Klausen T,Østergaard M. Power Doppler ultrasonography for assessment of synovitisin the metacarpophalangeal joints of patients with rheumatoid arthritis: acomparison with dynamic magnetic resonance imaging. Arthritis Rheum.2001;44:2018–23.

42. Szkudlarek M, Court-Payen M, Jacobsen S, Klarlund M, Thomsen HS,Østergaard M. Interobserver agreement in ultrasonography of the fingerand toe joints in rheumatoid arthritis. Arthritis Rheum. 2003;48:955–62.

43. Aletaha D, Neogi T, Silman AJ, et al. 2010 Rheumatoid arthritis classificationcriteria: an American College of Rheumatology/European League AgainstRheumatism collaborative initiative. Ann Rheum Dis. 2010;69:1580–8.

44. Boeters DM, Nieuwenhuis WP, Verheul MK, Newsum EC, Reijnierse M, ToesRE, Trouw LA, van der Helm-van Mil AH. MRI-detected osteitis is notassociated with the presence or level of ACPA alone, but with thecombined presence of ACPA and RF. Arthritis Res Ther. 2016;18:179.

45. Mediwake R, Isenberg DA, Schellekens GA, van Venrooij WJ. Use of anti-citrullinated peptide and anti-RA33 antibodies in distinguishing erosivearthritis in patients with systemic lupus erythematosus and rheumatoidarthritis. Ann Rheum Dis. 2001;60:67–8.

46. Amezcua-Guerra LM, Márquez-Velasco R, Bojalil R. Erosive arthritis insystemic lupus erythematosus is associated with high serum C-reactiveprotein and anti-cyclic citrullinated peptide antibodies. Inflamm Res.2008;57:555–7.

47. Chan MT, Owen P, Dunphy J, Cox B, Carmichael C, Korendowych E,McHugh NJ. Associations of erosive arthritis with anti-cyclic citrullinatedpeptide antibodies and MHC Class II alleles in systemic lupuserythematosus. J Rheumatol. 2008;35:77–83.

48. Skare TL, Nisihara R, Barbosa BB, da Luz A, Utiyama S, Picceli V. Anti-CCP insystemic lupus erythematosus patients: a cross sectional study in Brazilianpatients. Clin Rheumatol. 2013;32:1065–70.

49. Budhram A, Chu R, Rusta-Sallehy S, Ioannidis G, Denburg JA, Adachi JD,Haaland DA. Anti-cyclic citrullinated peptide antibody as a marker of erosivearthritis in patients with systemic lupus erythematosus: a systematic reviewand meta-analysis. Lupus. 2014;23:1156–63.

50. Amaya-Amaya J, Molano-González N, Franco JS, Rodríguez-Jiménez M,Rojas-Villarraga A, Anaya JM. Anti-CCP antibodies as a marker of rhupus.Lupus. 2015;24:892–4.

51. Gustafsson JT, Gunnarsson I, Källberg H, et al. Cigarette smoking,antiphospholipid antibodies and vascular events in systemic lupuserythematosus. Ann Rheum Dis. 2015;74:1537–43.

52. Svärd A, Skogh T, Alfredsson L, Ilar A, Klareskog L, Bengtsson C, Kastbom A.Associations with smoking and shared epitope differ between IgA- andIgG-class antibodies to cyclic citrullinated peptides in early rheumatoidarthritis. Arthritis Rheumatol. 2015;67:2032–7.

53. Wright S, Filippucci E, Grassi W, Grey A, Bell A. Hand arthritis in systemiclupus erythematosus: an ultrasound pictorial essay. Lupus. 2006;15:501–6.

54. Zoma A. Musculoskeletal involvement in systemic lupus erythematosus.Lupus. 2004;13:851–3.

55. Delle Sedie A, Riente L, Scirè CA, et al. Ultrasound imaging for therheumatologist. XXIV. Sonographic evaluation of wrist and hand joint andtendon involvement in systemic lupus erythematosus. Clin Exp Rheumatol.2009;27:897–901.

56. Peterson CM, Tsairis P, Onishi A, Lu YS, Grady R. Sodium cyanate inducedpolyneuropathy in patients with sickle-cell disease. Ann Intern Med.1974;81:152–8.

57. Crist RD, Grisolia S, Bettis CJ, Grisolia J. Carbamoylation of proteins followingadministration to rats of carbamoyl phosphate and cyanate and effects onmemory. Eur J Biochem. 1973;32:109–16.

58. Gladman D, Ginzler E, Goldsmith C, et al. The development and initialvalidation of the Systemic Lupus International Collaborating Clinics/American College of Rheumatology damage index for systemic lupuserythematosus. Arthritis Rheum. 1996;39:363–9.

59. Mohammad AJ, Weiner M, Sjöwall C, Johansson ME, Bengtsson AA,Ståhl-Hallengren C, Nived O, Eriksson P, Sturfelt G, Segelmark M. Incidenceand disease severity of anti-neutrophil cytoplasmic antibody-associatednephritis are higher than in lupus nephritis in Sweden. Nephrol DialTransplant. 2015;30 Suppl 1:i23–30.

60. Verheul MK, van Erp SJ, van der Woude D, et al. Anti-carbamylated proteinantibodies: a specific hallmark for rheumatoid arthritis. Comparison toconditions known for enhanced carbamylation; renal failure, smoking andchronic inflammation. Ann Rheum Dis. 2016;75:1575–6.

• We accept pre-submission inquiries

• Our selector tool helps you to find the most relevant journal

• We provide round the clock customer support

• Convenient online submission

• Thorough peer review

• Inclusion in PubMed and all major indexing services

• Maximum visibility for your research

Submit your manuscript atwww.biomedcentral.com/submit

Submit your next manuscript to BioMed Central and we will help you at every step:


RESEARCH ARTICLE Open Access

Decrease in bone mineral density duringthree months after diagnosis of earlyrheumatoid arthritis measured by digitalX-ray radiogrammetry predicts radiographicjoint damage after one yearMichael Ziegelasch1* , Kristina Forslind2,3, Thomas Skogh1, Katrine Riklund4, Alf Kastbom1 and Ewa Berglin5

Abstract

Background: Periarticular osteopenia is an early sign of incipient joint injury in rheumatoid arthritis (RA), but cannotbe accurately quantified using conventional radiography. Digital X-ray radiogrammetry (DXR) is a computerizedtechnique to estimate bone mineral density (BMD) from hand radiographs. The aim of this study was to evaluatewhether decrease in BMD of the hands (BMD loss), as determined by DXR 3 months after diagnosis, predictsradiographic joint damage after 1 and 2 years in patients with early RA.

Methods: Patients (n = 176) with early RA (<12 months after onset of symptoms) from three different Swedishrheumatology centers were consecutively included in the study, and 167 of these patients were included in theanalysis. Medication was given in accordance with Swedish guidelines, and the patients were followed for 2 years.Rheumatoid factor and antibodies to cyclic citrullinated peptides (anti-CCP) were measured at baseline, and 28-jointDisease Activity Score (DAS28) was assessed at each visit. Radiographs of the hands and feet were obtained atbaseline, 3 months (hands only) and 1 and 2 years. Baseline and 1-year and 2-year radiographs were evaluated bythe Larsen score. Radiographic progression was defined as a difference in Larsen score above the smallestdetectable change. DXR-BMD was measured at baseline and after 3 months. BMD loss was defined as moderatewhen the decrease in BMD was between 0.25 and 2.5 mg/cm2/month and as severe when the decrease wasgreater than 2.5 mg/cm2/month. Multivariate regression was applied to test the association between DXR-BMDloss and radiographic damage, including adjustments for possible confounders.

Results: DXR-BMD loss during the initial 3 months occurred in 59% of the patients (44% moderate, 15% severe):32 patients (19%) had radiographic progression at 1 year and 45 (35%) at 2 years. In multiple regression analyses,the magnitude of DXR-BMD loss was significantly associated with increase in Larsen score between baseline and1 year (p = 0.033, adjusted R-squared = 0.069).

Conclusion: DXR-BMD loss during the initial 3 months independently predicted radiographic joint damage at1 year in patients with early RA. Thus, DXR-BMD may be a useful tool to detect ongoing joint damage andthereby to improve individualization of therapy in early RA.

Keywords: Digital X-ray radiogrammetry, Bone mineral density, Disease progression, Early rheumatoid arthritis

* Correspondence: [email protected] of Rheumatology, Department of Clinical and ExperimentalMedicine, Linköping University, Linköping, SwedenFull list of author information is available at the end of the article

© The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Ziegelasch et al. Arthritis Research & Therapy (2017) 19:195 DOI 10.1186/s13075-017-1403-0

http://crossmark.crossref.org/dialog/?doi=10.1186/s13075-017-1403-0&domain=pdf

http://orcid.org/0000-0002-1307-440X

mailto:[email protected]

http://creativecommons.org/licenses/by/4.0/

http://creativecommons.org/publicdomain/zero/1.0/

BackgroundEarly stages of rheumatoid arthritis (RA) are characterizedby gradually developing joint swelling, stiffness and pain,and the patients often have a history of several months ofsymptoms when first presenting to the rheumatologist.Periarticular bone loss may already be present at thisstage, representing an early radiologic manifestationvisible on plain radiographs [1, 2]. The disease course inRA shows considerable inter-individual variation, rangingfrom mild and self-limiting to severe erosive disease,sometimes with extra-articular manifestations. Earlytreatment with disease-modifying anti-rheumatic drugs(DMARDs) is known to improve disease outcome [3–5],and may limit disease-associated bone loss [6]. However,further improved individual prediction of the diseasecourse and outcome remains an important issue in orderto optimize anti-rheumatic therapy.Digital X-ray radiogrammetry (DXR) is a technique

that uses computerized analyses of standard hand radio-graphs to estimate peripheral bone mineral density (BMD)of the three middle metacarpal bones (DXR-BMD) [7, 8].DXR-BMD loss has repeatedly been shown to predictradiographic joint progression in early RA [9–14]. How-ever, the majority of previous DXR-BMD studies havebeen based on 12-month change, and by that time, con-ventional X-ray assessments of joint damage are at least asinformative about disease progression [9, 11–14]. A Dutchstudy addressing DXR-BMD change after 4 months, re-ported an independent association between DXR-BMDloss and subsequent radiographic damage [10]. This studywas part of a clinical trial with selected patients, and thetreatment regimens were slightly different from standardcare in Sweden. Therefore, we wished to investigatewhether 3-month change in DXR-BMD predicts radio-graphic joint damage after 1 and 2 years in “real-world”patients with recent-onset RA.

MethodsPatientsPatients (n = 176) with early RA (64% women, symptomduration < 12 months), fulfilling the inclusion criteria(see subsequent text) and giving their informed consent,were consecutively included from three Swedish regions(one in Northern and two in Southern Sweden) in2008–2014 and were followed for 2 years. All patientsfulfilled the 2010 American College of Rheumatology(ACR)/European League Against Rheumatism (EULAR)[15] and/or the 1987 ACR [16] classification criteria.Pharmacotherapy was prescribed as found appropriate bythe treating rheumatologist, according to Swedish guide-lines. Baseline characteristics are detailed in Table 1.At baseline, 83% of patients were prescribed oral pred-

nisolone, 49% received osteoporosis prophylaxis withlow-dose calcium phosphate and vitamin D, and 6% with

bisphosphonates. 91% received conventional syntheticDMARDs (csDMARD) (88% methotrexate, 2.4% othercsDMARDs and 0.6% combination therapy). One patient(0.6%) was started on a tumor necrosis factor (TNF) in-hibitor at baseline. During the follow-up period 14.4%received biologic therapy (bDMARDs) as displayed inTable 2.

Radiographic assessment and digital X-ray radiogrammetry(DXR)Radiographs (posterior-anterior projection) of the hands,wrists and forefeet were performed at baseline, 3 months(hands only) and 1 and 2 years. The baseline and 1-yearand 2-year radiographs were read in chronological orderand evaluated according to the Larsen score [17] by oneinvestigator at each center (MZ, KF and EB). Thescoring system included 32 areas; metacarpal-phalangealjoints II −V, all proximal interphalangeal joints, thewrists divided into four areas and the metatarsophalan-geal joints II–V. Each joint and joint area was graded0–5. The maximum total score was 160. The smallestdetectable change (SDC) was calculated for the threereaders individually (EB, 2; KF, 1; MZ, 3) according tothe method of Bruynesteyn [18]. Radiographic progres-sion was defined as a difference in Larsen score abovethe SDC of the corresponding reader. The intra-raterand inter-rater reliability of the readers was assessed bycalculating the intraclass correlation coefficient (ICC).The ICC was 0.903.BMD was estimated on hand radiographs of the sec-

ond, third and fourth metacarpal bones using DXR (theonline Pronosco X-posure System, SECTRA, Linköping,Sweden), a computerized version of the traditional tech-nique of radiogrammetry measuring cortical bone thick-ness [8, 19]. DXR-BMD was assessed at inclusion andafter 3 months. The mean value of both hands was usedin all analyses. DXR-BMD values are given in mg/cm2.DXR-BMD loss was categorized either as a moderate de-crease in DXR-BMD (≥ 0.25 but < 2.5 mg/cm2 permonth) or a severe decrease (≥ 2.5 mg/cm2 per month),as defined by the provider (Sectra) [20]. To ensure con-sistent image acquisition, the images for each patientwere always taken in a frontal position of the handsusing the same X-ray machine. The images were sentunprocessed to Sectra for DXR analysis.

Clinical and laboratory assessmentsThe erythrocyte sedimentation rate (ESR, mm/1 h) andC-reactive protein (CRP, mg/L) were determined at base-line and after 3, 6, 12 and 24 months. At the same timepoints, the 28-joint Disease Activity Score (DAS28) wascalculated by the patient’s regular physician [21].Therapy response was determined according to EULARresponse criteria [22]. Functional status was evaluated


using the Swedish version of the Stanford Health Assess-ment Questionnaire (HAQ) [23]. Rheumatoid factor (RF)and antibodies to cyclic citrullinated peptides (anti-CCP2)were analyzed in baseline serum samples at the clinicalimmunology units of the local hospitals.

StatisticsStatistical calculations were performed using SPSS soft-ware (version 23, IBM Corporation, Armonk, USA).Linear regression analyses were used to explore the ef-fect of DXR-BMD alone, and in combination with vari-ous clinical and laboratory variables, chosen with respectto clinical assumptions, for associations with change in

Larsen score after 1 and 2 years. After testing each vari-able in a simple regression analysis with change inLarsen score at 1 and 2 years as the dependent variable,multiple regression analyses were performed includingvariables with p < 0.2. Radiographic progression wasdefined as a difference in Larsen score above theSDC between baseline and 1 and 2 years, respectively.Thus, for example, radiographs assessed by KF with adifference > 1 between the two time points weregraded as the actual difference, otherwise as 0. Assensitivity analyses, we also performed linear regres-sion with the same variables, but with absolutechanges in Larsen, i.e. not considering the SDC. Inaddition, logistic regression analysis was performed, inwhich radiographic progression was defined as changein Larsen score greater than SDC. The Pearson chi2

test or Fisher’s exact test were used for categoricalvariables, and the independent samples t test wasused for continuous variables. All p values are two-sided, and p values less than 0.05 were consideredstatistically significant.

ResultsIn total, 176 patients without previous DMARD expos-ure were included in the study. Table 1 shows the pa-tient characteristics and Table 2 the anti-rheumatictherapy including glucocorticoids (GC), and treatmentsthat influence BMD, initiated at baseline. There were nosignificant differences in these background characteris-tics between the participating sites (data not shown).Nine patients were lost because of missing radiographsat 3 months. Thus, the evaluation included 167 patientswith radiographs at baseline, 3 months and 1 year

Table 2 Treatment characteristics

Treatment Number (percentage)of patients

csDMARDs/biologics started atbaseline, n (%)

152 (91.0)

- MTX, n (%) 147 (88.0)

- Other csDMARDs, n (%) 4 (2.4)

- csDMARD triple therapy, n (%) 1 (0.6)

Anti-TNF at baseline, n (%) 1 (0.6)

Anti-TNF ever, n (%) 18 (10.8)

Other bDMARDs 6 (3.6)

Oral glucocorticoids, n (%) 138 (82.6)

Calcium supplements, n (%)a 79 (49.1)

Bisphosphonates, n (%)b 7 (6.0)

Abbreviations: MTX methotrexte, cDMARDs conventional disease modifyinganti-rheumatic drugs, TNF tumor necrosis factor, bDMARDs biologic diseasemodifying anti-rheumatic drugsaData available for 161 patientsbData available for 124 patients

Table 1 Baseline characteristics by radiographic progression at 1 year and bone mineral density loss after 3 months

Total Radiographic progression BMD loss

No Yes p value No Yes p value

Age 58 (14.5) 57.9 (14.4) 60.3 (14.7) 0.41 55.7 (15.5) 60.3 (13.4) 0.042

Women, n (%) 107 (64) 84/135 (62.2) 23/32 (71.9) 0.41 38/69 (55.1) 69/98 (70.4) 0.05

Symptom duration (months) 6 (3.7) 5.5 (3.3) 6.4 (5.0) 0.35 6 (3.2) 5.5 (3.9) 0.31

Anti-CCP2 positive, n (%) 107 (64) 88/135 (65.2) 19/32 (59.4) 0.54 42/69 (60.8) 65/98 (66.3) 0.51

RF positive, n (%) 103 (63) 83/131 (63.4) 20/32 (62.5) 1.0 40/68 (58.8) 63/95 (66.3) 0.41

ESR (mm/h) 29 (21.4) 27.3 (21.1) 35.4 (21.8) 0.055 25.4 (18.2) 31.3 (23.2) 0.066

CRP (mg/ml) 22 (24.4) 21.6 (25.0) 24.7 (22.2) 0.51 18.2 (23) 25 (25.1) 0.074

DAS 28 4.86 (1.30) 4.79 (1.3) 5.12 (1.4) 0.21 4.57 (1.27) 5.06 (1.29) 0.023

HAQ 0.94 (0.60) 0.91 (0.59) 1.08 (0.59) 0.15 0.86 (0.55) 1 (0.62) 0.16

Larsen total 4.1 (4.9) 3.57 (4.63) 6.25 (5.46) 0.005 3.16 ((4.23) 4.73 (5.24) 0.034

DXR-BMD 579.5 (86.1) 584.7 (84.2) 558.0 (91.7) 0.115 600.6 (84.2) 564.8 (84.7) 0.008

DXR-BMD loss 0.68 (1.81) −0.52 (1.65) −1.37 (2.31) 0.058 0.83 (1.05) −1.74 (1.45) < 0.001

Values are mean (SD) unless otherwise statedAbbreviations: CCP cyclic citrullinated peptides, RF rheumatoid factor, ESR erythrocyte sedimentation rate, CRP C-reactive protein, DAS28 Disease Activity Score,HAQ Health Assessment Questionnaire, DXR-BMD bone mineral density measured by digital X-ray radiogrammetryp values in italics are statistically significant


(Helsingborg, n = 38; Linköping, n = 65; Umeå, n = 64).Of the 167 patients, 129 also underwent radiography at2 years (Fig. 1). Compared with the patients with 2-yearradiographs available, patients without 2-year radio-graphs had lower mean Larsen score at baseline (2.1(SD = 3.17) vs 4. 7 (SD = 5.17); p < 0.001) and at 1 year(2.6 (SD = 3.34) vs 5.8 (SD = 5.66); p < 0.001), and wereto a lesser extent RF positive (48.6% vs 67.5%; p = 0.037).Other baseline characteristics, as detailed in Tables 1and 2 did not significantly differ from those in patientswith radiographs available at 2 years.At 3 months, 105 (63%) patients had low disease

activity (DAS28 ≤ 3.2) and 80 (48%) had reached EULARremission (DAS28 ≤ 2.6): 46 (28%) of the patients hadmoderate and 6 (4%) high disease activity (DAS28 > 5.2).DAS28 values from the 3-month visits were missing for10 (6%) patients. After 1 year, 108 (65%) of the patientshad low disease activity, 36 (22%) patients had moderateand 7 (4%) high disease activity: 88 (53%) of the patientshad reached EULAR remission. DAS28 values from the1-year visit were missing in 16 patients (10%).The mean age of the male patients (n = 60) at baseline

was 61 years (SD = 14.5) and the mean age of the female

patients (n = 107) was 57 years (SD = 14.3). Comparingour DXR-BMD values with a Danish reference cohort ofhealthy individuals [24], 92 patients (55.1%) in our co-hort had bone loss in the hand exceeding the age-relatedbone loss in the hand among the Danish controls.Of the 167 patients, 32 (19%) had radiographic pro-

gression at 1 year and 45 of 129 patients (35%) hadradiographic progression at 2 years. The change in DXR-BMD over 3 months showed BMD loss in 98 patients(59%). The DXR-BMD loss was moderate in 73/167 pa-tients (44%) and severe in 25/167 patients (15%). Radio-graphic joint damage was significantly different across thethree categories of DXR-BMD loss at baseline and at 1 year(p = 0.039 and p = 0.024, respectively) and there was a trendtowards statistical significance after 2 years (p = 0.056)(Table 3). Categorizing DXR-BMD loss according to theage-related reference material presented by Ornbjerg et al.[24] yielded very similar results (data not shown).Patients with change in Larsen score greater than the

SDC after 1 year had significantly higher Larsen scores(mean) at baseline (3.6 vs 6.2; p = 0.005). Compared withpatients without DXR-BMD loss, patients with DXR-BMD loss after 3 months were significantly older(60.3 years vs 55.7; p = 0.042), had significantly higherbaseline DAS28 (5.1 vs 4.6; p = 0.023) and significantlyhigher Larsen scores at baseline (4.7 vs 3.2; p = 0.034).Also, the proportion of women was significantly higher(70.4% vs 55.1%; p = 0.05) among patients with BMD loss(Table 2). There was no significant difference in theDXR value at baseline in the anti-CCP2-positive com-pared with the anti-CCP-negative patients (576 vs587 mg/cm2; p = 0.426).Simple regression analyses with change in Larsen score

greater than the SDC at 1 year as the dependent variablewere performed, including the following covariates: age,sex, oral corticosteroid treatment, DXR-BMD loss/month, baseline DAS28, CRP, ESR, Larsen score, anti-CCP2 status, and RF status. Also, DAS28 > 2.6 at3 months (yes/no) was included. Covariates with a pvalue < 0.2 in these analyses were included in a multipleregression model (Table 4). This model, adjusting for sexand baseline values of ESR, DAS28, Larsen score andanti-CCP2 status, showed a significant association be-tween 3-month BMD loss and increase in Larsen scoreabove the SDC after 1 year (p = 0.033, adjusted R-squared = 0.069) (Table 4). No significant associationwas observed between early bone loss and increase inthe Larsen score above the SDC at 2 years (p = 0.604).When using the same covariates but with change in Lar-sen score without considering the SDC as the dependentvariable, DXR-BMD loss was significantly associatedwith the 1-year Larsen score (p = 0.048), but not the Lar-sen score at 2 years (p = 0.491). In logistic regressionanalysis, there was no significant association between

Fig. 1 Availability of radiographs at baseline and follow up. Therewere 176 patients included in the study. aNine patients were lostbecause of missing radiographs at 3 months, and thus, the evaluationincluded 167 patients with radiographs at baseline, 3 months and1 year. bOf the 167 patients included in the evaluation, 129 alsounderwent radiography at 2 years


DXR-BMD loss and 1-year radiographic progressiondefined as a change in Larsen score above the SDC(p = 0.158). Treatment with bisphosphonates, calciumand vitamin D did not influence the DXR-BMD loss(data not shown).When analyzing only patients without erosions at

baseline (n = 123), using the same multivariate model(adjusted for the same variables but not for the Larsenscore at baseline), 3-month DXR-BMD loss remainedassociated with radiographic progression after 1 year(p = 0.021, R-squared = 0.07). Also, when analyzing the1-year outcome among the 129 patients with 2-yearradiographs available, the association between DXR-BMDloss and radiographic progression remained statisticallysignificant (p = 0.039, adjusted R-squared = 0.08).

DiscussionTo our knowledge, this is the first study addressing thepredictive value of 3-month DXR-BMD in patients withrecent-onset RA compared with radiography and clinicaldata. In clinical practice, evaluation of prescribedDMARDs is commonly performed 3 months after initi-ation. At this occasion, particularly in early disease, addi-tional information on the patient’s radiographic prognosis

would be highly valuable in order to optimize therapy de-cisions. In this study, we found DXR-BMD loss during thefirst 3 months to independently predict radiographic jointdamage at 1 year and the 1-year progression frombaseline.Our results on metacarpal bone loss among patients

with early RA are in line with previous reports [9–14].The shortest interval of DXR-BMD assessments amongprevious studies was 3 months in the study fromBøyesen et al. [25], addressing 3-month change in DXR-BMD as a predictive factor for erosive progression iden-tified on magnetic resonance imaging (MRI) in patientswith early RA. In the 53 patients completing that studythere was only a trend towards higher MRI synovitisscore and 3-month DXR BMD loss in patients developingMRI erosions, and no significant changes. Wevers-de Boerand coworkers [10] presented 4-month data with similarfindings on the 1-year radiographic outcome as in thecurrent study. Thus, early DXR-BMD assessments seemto be of clinical value, in order to optimize early institu-tion of anti-rheumatic pharmacotherapy and thereby di-minish the risk of future disability [4, 26–30]. However,we found no predictive value of DXR-BMD loss in relationto the 2-year radiographic outcome. This was somewhat

Table 3 Mean Larsen scores and 3-month bone mineral density loss in early rheumatoid arthritis

Mean (SD) Larsen score p value

No BMD loss Moderate BMD loss Severe BMD loss

Baseline (SD) 3.2 (4,23) 4.3 (4.81) 6.0 (6.28) 0.039

1 year (SD) 4.0 (4.68) 5.3 (5.45) 7.3 (6.38) 0.024

2 years (SD) 5.4 (5.1) 6.9 (5.6) 8.8 (6.04) 0.056

BMD bone mineral density

Table 4 Regression analyses with change in Larsen score between baseline and 12 months as dependent variable

Simple regression Multiple regression

Variable Adjusted R-square β-coeff p value Adjusted R-square β-coeff p value

DXR decrease (mg/cm2/month)a 0.049 −0.235 0.002 0.069 −0.181 0.033

DAS28b 0.015 0.148 0.070 0.028 0.767

Larsen scoreb 0.017 0.153 0.049 0.068 0.404

Ageb 0.001 0.082 0.291

Gender 0.007 0.114 0.143 0.071 0.381

anti-CCP2 statusb 0.010 −0.125 0.107 −0.128 0.114

RFb −0.001 −0.069 0.385

Corticosteroid treatment −0.002 0.042 0.417

Disease duration (months)b −0.006 0.026 0.739

CRP (mg/L)b −0.004 0.045 0.566

ESR (mm/h)b 0.024 0.173 0.025 0.133 0.165

DAS28 > 2.6 at 3 months −0.002 0.065 0.422

Abbreviations: DXR digital X-ray radiogrammetry, DAS28 Disease Activity Score in 28 joints, CCP cyclic citrullinated peptides, RF rheumatoid factor, CRP C-reactiveprotein, ESR erythrocyte sedimentation rate, BMD bone mineral densityaDecrease in DXR-BMD between baseline and 3 monthsbBaseline value


surprising, since existing radiographic damage oftenpredicts radiographic progression. One possible explan-ation for the disparate 1-year and 2-year findings in ourstudy could be that potent instituted pharmacotherapyattenuated radiographic differences over time. Also,missing data from 2-year radiographs (n = 38) need tobe considered, but the association between DXR-BMDloss and radiographic damage at 1 year remained statis-tically significant, also after excluding the 38 patientswithout 2-year radiographs. Thus, difference in statis-tical power appears to be an unlikely explanation forthe discordance between 1-year and 2-year data. A pre-vious study by Forslind et al. [9] showed that patientswith early RA, who were on prednisolone 7.5 mg perday in addition to conventional DMARDS, had signifi-cantly less DXR-BMD loss as compared with patientswith RA who were not receiving corticosteroids. Thisfinding was attributed to the anti-inflammatory effectof prednisolone, hampering osteopenia induced by in-flammation. Although not primarily designed to ad-dress this, our study did not identify a significantimpact of oral corticosteroid on BMD loss or radio-graphic progression. Similarly, treatment with bispho-sphonates, calcium and vitamin D did not significantlyimpact BMD loss.In our study we did not observe a significant difference

in the DXR value at baseline in the anti-CCP2-positivecompared with the anti-CCP-negative patients. This factis contrary to the findings in other studies [31–33] inwhich BMD loss was significantly more widespread inanti-CCP-positive patients. Different methods for estimat-ing BMD loss (comparative micro computed tomography(micro-CT) analysis and dual-energy X-ray absorpti-ometry) may influence the differing results. Also, it needsto be pointed out that our study was not primarily de-signed to address the influence of anti-CCP on BMD loss.Whenever a patient with RA presents with erosions at

the time of diagnosis, an aggressive disease course isassumed, and treatment is chosen accordingly. Thus, in-formation on the radiographic prognosis is even moreimportant in the large category of patients withoutevidence of erosive disease at diagnosis. Interestingly,we found that 3-month DXR-BMD loss also predicted1-year radiographic damage in this subgroup of pa-tients. However, the comparative value of DXR-BMDcompared with other imaging modalities needs to beassessed.Limitations of this study are the missing 2-year radio-

graphs in 38 patients (23%), and the fact that the base-line Larsen score and RF status differed betweenpatients with radiographs available at 2 years and thosewithout. Nevertheless, analyzing the 129 patients withall radiographs available up to 2 years did not substan-tially alter the findings.

ConclusionIn this real-world study of patients with early RA, wefound that DXR-BMD loss during the initial 3 monthsindependently predicted radiologic damage at 1 year.However, DXR-BMD loss predicts only a minor part ofthe variation in radiographic damage, and an associationwas not established after 2 years of disease. Futurestudies should compare the value of DXR-BMD withother imaging modalities.

AbbreviationsACR: American College of Rheumatology; bDMARD: Biologic diseasemodifying anti-rheumatic drug; BMD: Bone mineral density; CCP: Cycliccitrullinated peptides; csDMARD: Conventional synthetic disease modifying anti-rheumatic drug; CRP: C-reactive protein; DAS: Disease Activity Score; DAS28: DASwith 28 joints; DXR: Digital X-ray radiogrammetry; DXR-BMD: Bone mineraldensity measured by digital X-ray radiogrammetry; ESR: Erythrocytesedimentation rate; EULAR: European League Against Rheumatism;GC: Glucocorticoid; HAQ: Health Assessment Questionnaire; ICC: Intraclasscorrelation coefficient; MRI: Magnetic resonance imaging; MTX: Methotrexate;RA: Rheumatoid arthritis; RF: Rheumatoid factor; SDC: Smallest detectablechange; TNF: Tumor necrosis factor

AcknowledgementsThis work was financially supported by the Swedish Rheumatism Association,the Norrbacka-Eugenia foundation, the King Gustav V 80-year Foundation, theSwedish Medical Society, ALF Grants from Region Östergötland, the LinköpingUniversity Hospital Research Fund and the Foundation for Assistance to DisabledPeople in Skane (Stiftelsen för Bistånd åt Rörelsehindrade i Skåne).

FundingNot applicable.

Availability of data and materialsNot applicable.

Authors’ contributionsKF, AK, TS, EB and MZ designed the study and participated in patientenrollment and characterization. MZ, KF and EB scored radiographs. MZ andEB analyzed the data. All authors contributed to the interpretation of data,drafting the article and revising it critically. All authors approved the finalversion of the manuscript.

Ethics approval and consent to participateAll patients gave their written informed consent. The regional ethicscommittees in Linköping (DNR 2012/273-31), Lund (DNR Lu 464/2008), andUmeå (DNR 09-095 M) approved the study protocol, which was performedin accordance with the Declaration of Helsinki.

Consent for publicationNot applicable.

Competing interestsThe authors declare that they have no competing interests.

Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.

Author details1Department of Rheumatology, Department of Clinical and ExperimentalMedicine, Linköping University, Linköping, Sweden. 2Section ofRheumatology, Department of Medicine, Helsingborg Hospital, Helsingborg,Sweden. 3Department of Clinical Sciences, Section of Rheumatology, LundUniversity, Helsingborg, Sweden. 4Department of Diagnostic radiology, Umeåuniversity Hospital, Umeå, Sweden. 5Department of Public Health and ClinicalMedicine/Rheumatology, Umeå university Hospital, Umeå, Sweden.


Received: 28 March 2017 Accepted: 10 August 2017

References1. Brook A, Corbett M. Radiographic changes in early rheumatoid disease. Ann

Rheum Dis. 1977;36(1):71–3.2. Grassi W, et al. The clinical features of rheumatoid arthritis. Eur J Radiol.

1998;27 Suppl 1:S18–24.3. Darawankul B, et al. The good EULAR response at the first year is strongly

predictive of clinical remission in rheumatoid arthritis: results from theTARAC cohort. Clin Rheumatol. 2015;34(1):43–9.

4. Markusse IM, et al. Long-term outcomes of patients with recent-onsetrheumatoid arthritis after 10 years of tight controlled treatment: arandomized trial. Ann Intern Med. 2016;164(8):523–31.

5. Schneider M, Kruger K. Rheumatoid arthritis–early diagnosis and diseasemanagement. Dtsch Arztebl Int. 2013;110(27-28):477–84.

6. Dolan AL, et al. Does active treatment of rheumatoid arthritis limit disease-associated bone loss? Rheumatology (Oxford). 2002;41(9):1047–51.

7. Jorgensen JT, et al. Digital X-ray radiogrammetry: a new appendicular bonedensitometric method with high precision. Clin Physiol. 2000;20(5):330–5.

8. Rosholm A, et al. Estimation of bone mineral density by digital X-rayradiogrammetry: theoretical background and clinical testing. Osteoporos Int.2001;12(11):961–9.

9. Forslind K, et al. Hand bone loss measured by digital X-ray radiogrammetryis a predictor of joint damage in early rheumatoid arthritis. Scand JRheumatol. 2009;38(6):431–8.

10. Wevers-de Boer KV, et al. Four-month metacarpal bone mineral density losspredicts radiological joint damage progression after 1 year in patients withearly rheumatoid arthritis: exploratory analyses from the IMPROVED study.Ann Rheum Dis. 2015;74(2):341–6.

11. Guler-Yuksel M, et al. Changes in hand and generalised bone mineraldensity in patients with recent-onset rheumatoid arthritis. Ann Rheum Dis.2009;68(3):330–6.

12. Hoff M, et al. Cortical hand bone loss after 1 year in early rheumatoidarthritis predicts radiographic hand joint damage at 5-year and 10-yearfollow-up. Ann Rheum Dis. 2009;68(3):324–9.

13. Rezaei H, et al. Evaluation of hand bone loss by digital X-rayradiogrammetry as a complement to clinical and radiographic assessmentin early rheumatoid arthritis: results from the SWEFOT trial. BMCMusculoskelet Disord. 2013;14:79.

14. Forslind K, et al. Does digital X-ray radiogrammetry have a role inidentifying patients at increased risk for joint destruction in earlyrheumatoid arthritis? Arthritis Res Ther. 2012;14(5):R219.

15. Aletaha D, et al. 2010 Rheumatoid arthritis classification criteria: an AmericanCollege of Rheumatology/European League Against Rheumatismcollaborative initiative. Arthritis Rheum. 2010;62(9):2569–81.

16. Arnett FC, et al. The American Rheumatism Association 1987 revised criteriafor the classification of rheumatoid arthritis. Arthritis Rheum. 1988;31(3):315–24.

17. Larsen A. How to apply Larsen score in evaluating radiographs ofrheumatoid arthritis in long-term studies. J Rheumatol. 1995;22(10):1974–5.

18. Bruynesteyn K, et al. Deciding on progression of joint damage in pairedfilms of individual patients: smallest detectable difference or change. AnnRheum Dis. 2005;64(2):179–82.

19. Barnett E, Nordin BE. The radiological diagnosis of osteoporosis: a newapproach. Clin Radiol. 1960;11:166–74.

20. de Rooy DP, et al. Loss of metacarpal bone density predicts RAdevelopment in recent-onset arthritis. Rheumatology (Oxford). 2012;51(6):1037–41.

21. Prevoo ML, et al. Modified disease activity scores that include twenty-eight-joint counts. Development and validation in a prospective longitudinalstudy of patients with rheumatoid arthritis. Arthritis Rheum. 1995;38(1):44–8.

22. van Gestel AM, Haagsma CJ, van Riel PL. Validation of rheumatoid arthritisimprovement criteria that include simplified joint counts. Arthritis Rheum.1998;41(10):1845–50.

23. Ekdahl C, et al. Assessing disability in patients with rheumatoid arthritis. Useof a Swedish version of the Stanford Health Assessment Questionnaire.Scand J Rheumatol. 1988;17(4):263–71.

24. Ornbjerg LM, et al. Establishment of age- and sex-adjusted reference datafor hand bone mass and investigation of hand bone loss in patients withrheumatoid arthritis treated in clinical practice: an observational study fromthe DANBIO registry and the Copenhagen Osteoarthritis Study. Arthritis ResTher. 2016;18:53.

25. Boyesen P, et al. Prediction of MRI erosive progression: a comparison ofmodern imaging modalities in early rheumatoid arthritis patients. AnnRheum Dis. 2011;70(1):176–9.

26. Goekoop-Ruiterman YP, et al. Clinical and radiographic outcomes of fourdifferent treatment strategies in patients with early rheumatoid arthritis (theBeSt study): A randomized, controlled trial. Arthritis Rheum. 2008;58(2Suppl):S126–35.

27. Breedveld F. The value of early intervention in RA − a window ofopportunity. Clin Rheumatol. 2011;30 Suppl 1:S33–9.

28. Ahlstrand I, et al. Pain and activity limitations in women and men withcontemporary treated early RA compared to 10 years ago: the Swedish TIRAproject. Scand J Rheumatol. 2015;44(4):259–64.

29. Kudo-Tanaka E, et al. Early therapeutic intervention with methotrexateprevents the development of rheumatoid arthritis in patients with recent-onset undifferentiated arthritis: A prospective cohort study. Mod Rheumatol.2015;25(6):831–6.

30. Kyburz D, Finckh A. The importance of early treatment for the prognosis ofrheumatoid arthritis. Swiss Med Wkly. 2013;143:w13865.

31. Kleyer A, et al. Bone loss before the clinical onset of rheumatoid arthritis insubjects with anticitrullinated protein antibodies. Ann Rheum Dis. 2014;73(5):854–60.

32. Bugatti S, et al. Anti-citrullinated protein antibodies and high levels ofrheumatoid factor are associated with systemic bone loss in patients withearly untreated rheumatoid arthritis. Arthritis Res Ther. 2016;18(1):226.

33. Orsolini G, et al. Titer-dependent effect of anti-citrullinated proteinantibodies on systemic bone mass in rheumatoid arthritis patients. CalcifTissue Int. 2017;101(1):17–23.

• We accept pre-submission inquiries

• Our selector tool helps you to find the most relevant journal

• We provide round the clock customer support

• Convenient online submission

• Thorough peer review

• Inclusion in PubMed and all major indexing services

• Maximum visibility for your research

Submit your manuscript atwww.biomedcentral.com/submit

Submit your next manuscript to BioMed Central and we will help you at every step:


1



Michael Ziegelasch1)

, Antonia Boman2)

, Klara Martinsson1)

, Ingrid Thyberg1)

, Claudia Jacobs1)

,

Britt-Marie Nyhäll-Wåhlin3)

, Anna Svärd3)

, Ewa Berglin2)

, Solbritt Rantapää-Dahlqvist2)

,

Thomas Skogh1)

, Alf Kastbom1)

Author affiliations 1)

Department of Rheumatology and Department of Clinical and

Experimental Medicine, Linköping University, Linköping, Sweden, 2)

Department of Public

Health and Clinical Medicine, Rheumatology, Umeå University, Umeå, Sweden, 3)

Clinic of

Rheumatology, Falun, Sweden and Center for Clinical Research Dalarna, Uppsala University,

Uppsala, Sweden

Corresponding Author Dr Alf Kastbom, Department of Rheumatology, Linköping University

Hospital, Linköping SE 581 85, Sweden; [email protected]

Keywords Rheumatoid arthritis, anti-citrullinated protein antibodies, disease activity,

radiographic damage

2

ABSTRACT

Objective: The discovery of anti-citrullinated protein antibodies (ACPA) and the introduction of

new therapeutic options have had profound impact on early rheumatoid arthritis (RA) care. Since

ACPA status, most widely assessed as reactivity to cyclic citrullinated peptides (CCP),

influences treatment decisions in early RA, we aimed to determine whether anti-CCP remains a

predictor of disease activity and radiographic joint damage in contemporary ‘real-world’ early

RA.

Methods: Two observational early RA cohorts from southeast-mid Sweden enrolled patients

1996-1999 (TIRA-1; n=239) and 2006-2009 (TIRA-2; n=444), respectively. Clinical and

radiographic data were consecutively collected up to 3 years. As confirmation, we used two

cohorts from Northern Sweden, denoted TRAM-1 (patient enrollment 1996-2000; n=249) and

TRAM-2 (2006-2011; n=528), with 2 years follow-up. Baseline IgG anti-CCP antibodies were

related to disease activity, pharmacotherapy, and radiographic joint damage as scored according

to Larsen.

Results: In the TIRA-1 cohort, anti-CCP positive patients had significantly higher levels of C-

reactive protein (CRP) and 28-joint disease activity score (DAS28) during follow-up compared

with anti-CCP negative patients (p<0.01). Contrastingly, anti-CCP negative patients in the

TIRA-2 cohort had higher baseline CRP (34mg/L vs. 23mg/L; p=0.007) and higher swollen joint

count as compared to anti-CCP positive patients (10.3 vs 7.3, p<0.001), whereas no significant

differences were seen during follow-up. In TIRA-2, anti-CCP positivity was associated with

higher mean 3-year Larsen score (5.4 vs. 3.5, adjusted p=0.039). In TRAM-2, anti-CCP also

predicted radiographic damage (8.9 vs. 6.7, adjusted p=0.027), without any significant

differences in disease activity during follow-up.

3

Conclusion: Although modern management of early RA abolishes the association between

baseline anti-CCP status and disease activity over time, the radiographic damage remains

increased among anti-CCP positive patients. These findings call for close radiographic

monitoring in early anti-CCP positive RA also in cases with limited disease activity over time.

The findings also support the hypothesis of direct bone degrading effects of antibodies to

citrullinated proteins.

4

INTRODUCTION

Management of early rheumatoid arthritis (RA) has improved substantially in recent decades

thanks to treat-to-target strategies and novel pharmaceutical options (1, 2). Also, major

breakthroughs were made in the late 1990s by the identification of citrulline residues as key

antigenic determinants for RA-specific autoantibodies (3, 4), and the development of easily

available detection assays, such as the anti-cyclic citrullinated peptide (anti-CCP) test.

Subsequently, occurrence of IgG-class anti-CCP was shown to precede the onset of RA (5) and

associate with a worse prognosis both in terms of disease activity (6-8), extra articular

manifestations (9-11) and radiographic joint damage (12-14). Furthermore, anti-CCP positive

patients in clinical remission appear to have a higher risk of radiographic progression compared

with anti-CCP negative patients (15).

The anti-CCP antibody test may react with a number of different citrullinated antigens (16-19),

among which the relative importance remains unclear (20, 21). In Sweden, clinical use of the

anti-CCP test increased rapidly following its introduction around 2003, and the awareness of its

diagnostic and prognostic properties became widespread among rheumatologists shortly

thereafter. To a large extent, the extensive body of evidence stating that a serum test for anti-

CCP is associated with a more severe course of early RA originates from the era before the

widespread introduction of early aggressive ‘traditional’ anti-rheumatic pharmacotherapy, the

introduction of biologics, and the use of anti-CCP status as a factor to consider in treatment

decisions. Therefore, we aimed to study anti-CCP as a predictor in contemporary ‘real-world’

recent-onset RA, and relate the findings to historical cohorts.

5

PATIENTS & METHODS

Patients

Two multicenter prospective observational early RA cohorts denoted ‘TIRA’ (= a Swedish

acronym for ‘‘early intervention in rheumatoid arthritis’’) were enrolled 10 years apart and

followed up prospectively over 3 years (henceforth denoted primary cohorts). The inclusion

criteria were identical in the two cohorts: symptom duration (defined as first observed joint

swelling <12 months), and either fulfilment of the 1987 ACR criteria (22) or suffering from

morning stiffness >60 min, symmetrical arthritis, and small joint engagement. TIRA-1 enrolled

239 patients 1996-1999 in southeast-mid Sweden (rheumatology clinics of Eskilstuna,

Jönköping, Kalmar, Linköping, Västervik, and Örebro). During this time, neither anti-CCP nor

treatment with biologics was used in clinical practice. TIRA-2 enrolled 444 patients 2006-2009

using identical inclusion criteria and similar geographic coverage as in TIRA-1 (Falun replaced

Örebro). Follow-up visits were scheduled after 6, 12, 24 and 36 months. To confirm our findings,

we used independent cohorts (henceforth denoted confirmation cohorts) from the Northern

Region of Sweden (rheumatology departments at Umeå, Luleå, Örnsköldsvik, Sundsvall, and

Östersund) during similar time-periods; TRAM-1 (enrollment 1996-2000) and TRAM-2 (2006-

2011), respectively (TRAM = a Swedish acronym for “early rheumatoid arthritis clinic”). These

cohorts had similar follow-up routines as in the TIRA cohorts, except that follow-up were done

regularly for 24 months. TRAM-1 and TRAM-2 enrolled 249 and 528 patients respectively, with

early RA (symptoms < 12 months) and fulfilling the 1987 ACR criteria for RA (22).

At baseline and scheduled follow-up visits, we recorded ongoing medication, tender (TJC) and

swollen joint count (SJC), patient’s global assessment (100 mm visual analogue scale),

6

erythrocyte sedimentation rate (ESR) (mm/h), 28-joint disease activity score (DAS28)(23) and

response to treatment at 6, 12 and 24 months according to the European League against

Rheumatism criteria (24). Patient-reported pain (visual analogue scale), physician’s global

assessment of disease activity, and functional ability according to the Swedish version of the

Stanford Health Assessment Questionnaire (HAQ)(25) were also registered.

Laboratory analyses

Serum samples were drawn at baseline and stored at -70- -80°C until analyzed. Anti-CCP

antibodies were detected using 2nd

generation immunoassays (Immunoscan CCPlus,

EuroDiagnostica AB, Malmö, Sweden). We used the cut-off level recommended by the

manufacturer (25 AU/mL). IgM-class rheumatoid factor (RF), ESR (mm/h) and C-reactive

protein (CRP, mg/L) were analyzed according to routine methods at the local hospitals.

Radiographic evaluation

In TIRA-2, radiographs of the hands and feet taken at baseline and after 36 months were

evaluated in chronological order by one experienced reader (MZ), and graded according to the

Larsen score (26). In TIRA-2, 419 (94%) of the patients had radiographs available at baseline

and 265 (60%) after 36 months. In total, 254 (57%) of TIRA-2 patients had radiographs available

from both time-points. Patients with complete radiographic data were significantly younger

compared to those without (mean 56.8 years vs. 61.0, p=0.002), had significantly lower baseline

7

ESR (29.8 vs. 35.3 mm/h; p=0.014) and were more often anti-CCP positive (72.5% vs. 62.9;

p=0.040). Radiographs were not available from TIRA-1.

In TRAM-1 and TRAM-2, radiographs of hands and feet from baseline and 24 months follow-up

(available from 96.8% and 89.5% of the patients, respectively), were scored according to the

Larsen method (25) in chronological order by two experienced readers (SRD and EB). Patients

with and without available radiographic data did not show any significant difference concerning

DAS28, ESR, sex or age in TRAM-1 (data not shown). In TRAM-2, there were significant

differences between patients with radiographs compared with those without regarding age at

inclusion (mean 56.0 vs. 61.1, p=0.001) and baseline DAS28 (mean 4.6 vs. 4.8, p=0.036), but

not regarding the separate components of DAS28 (SJC, TJC, patients global assessment), CRP,

HAQ, or pain VAS between the patients.

Statistical analyses

Statistical calculations were performed using SPSS software (version 23, IBM Corporation,

Armonk, USA). Two-sided p-values <0.05 were considered significant. In the TIRA cohorts,

missing data points for CRP, ESR, TJC and SJC were considered to occur at random, and except

for baseline values, we adopted the last observation carried forward (LOCF) method.

Radiographic data was not subject to LOCF. In TRAM-1 and TRAM-2 missing values were

imputed using chained equations (27).

Measures of disease activity, function and pain were compared between anti-CCP positive and

negative patients by the general linear model for repeated measurements. We separated the

assessments for baseline and follow-up and calculated p-value separately for each. Student’s t-

8

test was used for baseline analyses, and general linear model for repeated measurements

regarding the follow-up period. Categorical data was compared using Chi-square method.

The Mann Whitney U test was used to compare baseline and follow-up radiographic joint

damage according to anti-CCP status. To evaluate anti-CCP as an independent predictor of

radiographic damage, we used general linear models that included baseline clinical and

laboratory factors that trend-wise (p<0.1) correlated with Larsen score in univariable analysis in

the discovery cohort (TIRA-2).

RESULTS

Demographic and clinical characteristics of the cohorts

Baseline characteristics of all cohorts are shown in table 1. Of note, mean age at inclusion had

increased in both contemporary cohorts, while the proportion of anti-CCP positive patients

remained similar.

9

Anti-CCP status and clinical course

In the primary historical cohort, TIRA-1, there were no significant differences between anti-CCP

positive and anti-CCP negative patients regarding baseline ESR and CRP (Figure 1). During

follow-up however, both CRP and ESR were higher among anti-CCP positive cases (p=0.001

and 0.002 respectively). Similar findings were made concerning DAS28 and SJC, with

significantly higher values throughout follow-up, but not at baseline (Figure 1). TJC and HAQ

did not differ significantly in relation to anti-CCP status in TIRA-1.

Table 1: Patient characteristics at baseline

Primary cohorts Confirmation cohorts

TIRA-1

N=239 TIRA-2

N=444

p-value TRAM-1

N=249 TRAM-2

N=528

p-value

Period of enrollment

1996-1999 2006-2009 - 1995-2000 2006-2011 -

Mean age, years (SD)

54.9 (15.4) 58.7 (14.5) 0.003 54.0 (14.6) 57.9 (14.1) <0.001

Females, n (%)

163 (68.2) 297 (66.9) 0.52 168 (67.5) 368 (69.7) 0.53

Anti-CCP pos, n (%)

156 (65.3) 303 (68.2) 0.44 186 (74.6) 367 (69.5) 0.15

Rheumatoid factor

pos, n (%)

155 (64.9) 264 (59.5) 0.88 202 (81.0) 385 (73.0) 0.02

Proportion of patients

fulfilling ACR87 at

inclusion, n (%)

231 (96.7) 421 (94.8) 1.0 249 (100) 528 (100)

ACR, American College of Rheumatology; CCP, Cyclic Citrullinated Peptides; RF, Rheumatoid

factor; ESR, Erythrocyte sedimentation rate; CRP, C-reactive protein; DAS28, Disease Activity

Score; HAQ, Health assessment questionnaire.

10

In the primary contemporary cohort, TIRA-2, anti-CCP positive patients had lower baseline

levels of CRP (23 vs. 34 mg/L, p=0.007), DAS28 (5.0 vs. 5.3; p=0.007), SJC (10.3 vs. 7.3,

p<0.001), and borderline significance regarding ESR (31 vs. 35 mm/h, p=0.096) compared with

anti-CCP negative patients. In contrast, no significant differences remained during follow-up

concerning these variables (Figure 1). We found no baseline differences regarding the self-

reported outcomes ‘patients global assessment’ (PGA), pain, and HAQ, but during follow-up,

anti-CCP positive patients scored higher regarding PGA (p=0.006) and pain (p=0.034) (Figure

1).

In the confirmation historical cohort, TRAM-1, there were significantly higher values of ESR

amongst anti-CCP positive patients during the follow-up (p=0.049), although there were no

significant difference recorded at baseline (additional file 2). There were no significant

differences associated with anti-CCP status concerning CRP, DAS28, SJC, TJC, PGA and pain

VAS (additional file 2).

In the confirmation contemporary cohort, TRAM-2, ESR was significantly higher among anti-

CCP positive cases at baseline (p=0.019), but not at follow-up (additional file 2). There were no

significant differences according to anti-CCP status regarding DAS28, CRP, SJC, TJC, PGA and

pain VAS. HAQ was borderline significant for the anti-CCP positive patients at follow-up, but

not at baseline, in both the TRAM 1 and 2 cohorts (additional file 2).

Baseline anti-CCP status and radiographic joint damage

In TIRA-2, baseline Larsen scores did not significantly differ between anti-CCP positive and

anti-CCP negative patients (3.0 vs. 2.4, p=0.276). At 36 months, however, anti-CCP positive

11

patients had significantly higher Larsen score than anti-CCP negative patients (mean 5.4 vs. 3.5,

p=0.027 (Figure. 2)). In the confirmation cohorts, anti-CCP was associated with higher Larsen

scores at baseline (TRAM-1; mean 5.8 vs. 2.7, p=0.01 and TRAM-2; mean 6.6 vs. 5.1, p=0.03)

as well as at follow-up (TRAM-1; mean 11.3 vs. 6.4, p=0.001 and TRAM-2; mean 8.9 vs. 6.7,

p=0.009 (Figure 2)). In multivariable analysis, Larsen score at follow-up was independently

associated with baseline Larsen score in all cohorts, while baseline anti-CCP status reached

statistical significance only in the two contemporary cohorts (Table 2). In the multivariable

analysis, baseline anti-CCP status was associated with follow-up radiographic damage also after

adjusting for EULAR response to treatment at 6 months.

Table 2. Multivariable linear regression analysis with radiographic damage at

follow-up as dependent variable. The models include baseline factors that trend-

wise (p<0.1) correlated with radiographic damage at follow-up in the primary cohort

using univariable analysis. Significant p-values are shown in bold.

Baseline

data

TIRA-2

(N=237) TRAM-1

(N=92) TRAM-2

(n=252)

B p-value B p-value B p-value

Larsen score 1.134 <0.001 1.163 <0.001 0.978 <0.001

Anti-CCP

status

1.054 0.039 1.322 0.393 1.109 0.027

Age

0.005 0.793 0.078 0.114 0.004 0.801

ESR

0.013 0.216 0.028 0.371 0.016 0.107

12

Anti-rheumatic treatment and response according to anti-CCP status

In TIRA-1, the pattern of treatment was significantly different between anti-CCP positive and

anti-CCP negative patients at all time-points of the study period except for baseline (p=0.057),

consistent with a more aggressive treatment approach among the anti-CCP positive patients

(Figure 3a). In TIRA-2, the pattern of treatment was significantly different from year 2 and

onwards, and borderline from 12 months (p=0.087), again with the appearance of a more

aggressive pharmacotherapy among anti-CCP positive patients (Figure 3b). The treatment was

generally more intense in TIRA-2 compared to TIRA-1 (Figure 3a+b). For instance, 52% anti-

CCP positive patients from TIRA-1 and 93% in TIRA-2 were prescribed conventional DMARDs

at baseline (p<0.001).

The frequencies of treatment with conventional DMARDs were similar in both TRAM-1 and -2

during the 24 months (Figure 3c+d). In TRAM-1, only 17% among anti-CCP positive patients

and 14% among anti-CCP negative were not prescribed DMARDs at baseline. In TRAM-2, the

corresponding baseline percentages were 12% and 9%, respectively. These numbers were fairly

consistent throughout the 24 months. From 12 months in TRAM-1 the number of patients treated

with combination therapy of DMARDs increased, whilst in TRAM-2 the number of patients

treated with biologics also increased. In TRAM-2, treatment with biologics had increased to 13%

in anti-CCP positive patients, compared with 6% in the anti-CCP negative patients (p<0.05).

EULAR responses rates evaluated at 6, 12 and 24 months were significantly superior (p<0.0001

for all occasions) in the contemporary cohorts (TIRA-2 and TRAM-2) versus their historical

counterparts (TIRA-1 and TRAM-1, respectively) (additional file 1). The frequencies of good

response were between 39-57% for TIRA-2/TRAM-2 vs. 20-35% for TIRA-1/TRAM-1

13

(additional file 1). Differences between historical and contemporary cohorts remained

statistically significant at all time-points also after stratifying for anti-CCP antibodies (data not

shown).

DISCUSSION

In this study we present data showing that modern management of early RA attenuates the

association between baseline anti-CCP status and disease activity over time, but despite this,

patients with anti-CCP antibodies remain at increased risk of radiographic joint damage. These

findings highlight the need for careful radiographic monitoring of anti-CCP positive early RA

patients also when disease activity is brought under control and treatment response is achieved.

These findings also provide support to the hypothesis that antibodies to citrullinated

proteins/peptides (ACPA) can mediate bone degrading effects.

At the time of inclusion in our historical cohorts, neither anti-CCP tests nor the therapeutic use of

biologics were established in clinical routine. Also, the concept of early aggressive therapy and

the ‘window of opportunity’ had not been fully implemented. Thus, it is not surprising that the

contemporary early RA patients were subject to more aggressive pharmacotherapy and more

often achieved EULAR responses than in the historical cohorts. However, we had expected the

differences in anti-rheumatic therapy according to baseline anti-CCP status to be more

pronounced and to appear earlier in the disease course than was the case in the contemporary

cohorts. In TIRA-2, the treatment pattern was not significantly different until 2 years after

inclusion, and in the contemporary replication cohort TRAM-2, no significant differences were

found except for the proportion of biological treatment at 24 months. Possibly, this was because

14

disease activity measures among anti-CCP positive patients did not exceed those of the anti-CCP

negative neither at baseline nor during follow-up in TIRA-2 or TRAM-2.

The current study describes two independent cohorts of early RA patients diagnosed 2006-

2009/2011 where baseline anti-CCP status independently associated with radiographic joint

damage after 2 and 3 years, respectively. This association emerged despite rheumatologists’

awareness of anti-CCP status, and very similar disease activity measures during follow-up. Also,

the association remained after adjusting for response to treatment after 6 months. It should be

noted that initial treatment with bDMARDs was rare, and hypothetically, a more frequent early

bDMARD use could have attenuated the radiographic damage associated with anti-CCP.

Although radiographic damage in the contemporary cohorts was less pronounced compared to

the historical cohort TRAM-1 (Fig 2), we believe that our findings call for a careful radiographic

monitoring of anti-CCP positive patients also in the case of limited disease activity. These

findings also imply a need for drugs with other modes of action than directed against pro-

inflammatory activity.

The radiographic results in the current study are in line with previous work showing bone-

specific effects of ACPA. For instance, Bugatti et al showed an association between ACPA, RF

and low bone mass (28). Llorente et al described similar results, but they could not find any

association with RF, strengthening the particular involvement of ACPA in the pathogenesis of

bone destruction in early RA and undifferentiated arthritis (29). The importance of anti-CCP

antibodies as well as receptor activator of nuclear factor kappa-B ligand (RANKL) for the

radiographic progression has been shown by us and others (30-32). Also, other studies have

supported that ACPA occurrence is associated with bone remodeling (31, 32). Carpenter et al

15

investigated radiographic damage in RA in two large cohorts enrolling patients 1986-2001 and

2002-2013 respectively (33). Instead of ACPA, that study assessed the contribution of RF, which

associated with greater radiographic progression in each cohort, although the newer cohort had a

lower absolute difference regarding mean annual change in Sharp/van der Heijde Score (SHS).

Identical to our study, more patients were treated earlier and partly with new more aggressive

anti-rheumatic drugs resulting in improvement of disease outcome. These observations are in

line with our results as well as the results from previous studies (34-39).

Interestingly, despite lower baseline disease activity variables among anti-CCP positive patients

in the TIRA-2 cohort, anti-CCP positivity was also associated with significantly worse patient-

reported global health and pain during follow-up. This could either be a reflection of the doctor’s

awareness and communication combined with the feeling of being treated with more

“aggressive” combinations of drugs. More speculatively, it could reflect more specific ACPA-

mediated effects, such as pain induction in sensory neurons (40). However, we could not detect

significant differences in global health or pain in the contemporary confirmation cohort.

The well-characterized ‘real-world’ patients, including replication and historical cohorts, are

strengths of the current study. Limitations include missing radiographic data, since differences

were seen in the contemporary cohorts regarding baseline characteristics between those who had

radiographic data and those who did not. Also, patients were not enrolled based on the 2010

ACR/EULAR criteria and therefore may not completely align with current practice in the clinics.

16

CONCLUSIONS

Although modern management of early RA abolishes the association between baseline anti-CCP

status and disease activity over time, the radiographic damage remains increased among anti-

CCP positive patients. This calls for careful radiographic monitoring of early RA patients with

anti-CCP antibodies also when disease activity is brought under control. Further, these findings

support the hypothesis of direct bone-degrading effects of ACPA.

LIST OF ABBREVIATIONS

ACPA – Anti-citrullinated protein antibodies

ACR – American College of Rheumatology

AKA – Anti-keratin antibodies

APF – Antiperinuclear factor

AU – Arbitrary units

CCP – Cyclic citrullinated Peptides



DAS28 – Disease Activity Score


17


JSN – Joint space narrowing

LOCF – Last observation carried forward

PG-VAS – Patient’s global visual analogue scale


RANKL – Receptor activator of nuclear factor kappa-B ligand


SJC – Swollen joint count

SHS – Sharp/van der Heijde score

TJC – Tender joint count

TIRA – Tidiga insatser för reumatoid artrit (acronym for ‘‘early intervention in rheumatoid

arthritis’’)

TRAM – Tidig reumatoid artrit mottagning (acronym for “early rheumatoid arthritis clinic”)

VAS –Visual analogue scale

18

DECLARATIONS

Ethics approval consent to participate

The study protocol was approved by the regional ethics review board in Linköping (TIRA-1 and

TIRA-2) and in Umeå (TRAM-1 and TRAM-2). All participating patients gave a written informed

consent.

Consent for publication

Not applicable

Data availability

The datasets used and analysed during the current study are available from the

corresponding author on reasonable request

Competing interests

All the authors declare that they have no competing interests.

19

Funding

This study was supported by grants from the Swedish Medical Society, King Gustaf V’s 80-Year

Fund, Östergötland County Council (ALF-grants), and the Swedish Rheumatism Association.

SRD was supported by grants from Västerbotten county council (ALF) and the Swedish

Foundation for Strategic Research, Sweden.

Author contributions

MZ, AB, SRD, TS, and AK made substantial contribution to conception, design,

interpretation of data, and drafting the manuscript; MZ, AB and SRD performed statistical

calculations; MZ, IT, CJ, B-M N-W, AS, EB, SRD, and AK were involved patient

recruitment and characterization; KM performed laboratory analyses and were involved in

data acquisition; MZ, SRD, and EB scored radiographs; all authors revised the manuscript

critically for important intellectual content

Acknowledgements

We would like to acknowledge all TIRA participants, the staff at the Department of

Rheumatology, University Hospital, Umeå, and our colleagues at the Departments of

Rheumatology at the hospitals at Sunderbyn-Luleå, Sundsvall and Östersund, Sweden.

20

REFERENCES

1. Balduzzi S, Scire CA, Sakellariou G, Benaglio F, Bugatti S, Montecucco C, et al. In early

inflammatory polyarthritis more intensive management according to the 2010 ACR/EULAR

criteria leads to higher rates of clinical remission: comparison of two cohorts treated according to

different treat-to-target protocols. Clin Exp Rheumatol. 2017;35:401-5.

2. Smolen JS, Aletaha D, McInnes IB. Rheumatoid arthritis. Lancet. 2016;388:2023-38.

3. Girbal-Neuhauser E, Durieux JJ, Arnaud M, Dalbon P, Sebbag M, Vincent C, et al. The

epitopes targeted by the rheumatoid arthritis-associated antifilaggrin autoantibodies are

posttranslationally generated on various sites of (pro) filaggrin by deimination of arginine

residues. J Immunol. 1999;162:585-94.

4. Schellekens GA, de Jong BA, van den Hoogen FH, van de Putte LB, van Venrooij WJ.

Citrulline is an essential constituent of antigenic determinants recognized by rheumatoid

arthritis-specific autoantibodies. J Clin Invest. 1998;101:273-81.

5. Ten Brinck RM, van Steenbergen HW, van Delft MAM, Verheul MK, Toes REM, Trouw

LA, et al. The risk of individual autoantibodies, autoantibody combinations and levels for

arthritis development in clinically suspect arthralgia. Rheumatology (Oxford). 2017;56:2145-53.

6. Haschka J, Englbrecht M, Hueber AJ, Manger B, Kleyer A, Reiser M, et al. Relapse rates

in patients with rheumatoid arthritis in stable remission tapering or stopping antirheumatic

therapy: interim results from the prospective randomised controlled RETRO study. Ann Rheum

Dis. 2016;75:45-51.

7. Willemze A, Trouw LA, Toes RE, Huizinga TW. The influence of ACPA status and

characteristics on the course of RA. Nat Rev Rheumatol. 2012;8:144-52.

21

8. de Rooy DP, van der Linden MP, Knevel R, Huizinga TW, van der Helm-van Mil AH.

Predicting arthritis outcomes–what can be learned from the Leiden Early Arthritis Clinic?

Rheumatology (Oxford). 2011;50.

9. Lopez-Longo FJ, Oliver-Minarro D, de la Torre I, Gonzalez-Diaz de Rabago E, Sanchez-

Ramon S, Rodriguez-Mahou M, et al. Association between anti-cyclic citrullinated peptide

antibodies and ischemic heart disease in patients with rheumatoid arthritis. Arthritis Rheum.

2009;61:419-24.

10. Zhu J, Zhou Y, Chen X, Li J. A metaanalysis of the increased risk of rheumatoid arthritis-

related pulmonary disease as a result of serum anticitrullinated protein antibody positivity. J

Rheumatol. 2014;41:1282-9.

11. Turesson C, Jacobsson LT, Sturfelt G, Matteson EL, Mathsson L, Ronnelid J.

Rheumatoid factor and antibodies to cyclic citrullinated peptides are associated with severe

extra-articular manifestations in rheumatoid arthritis. Ann Rheum Dis. 2007;66:59-64.

12. Johansson L, Pratesi F, Brink M, Arlestig L, D'Amato C, Bartaloni D, et al. Antibodies

directed against endogenous and exogenous citrullinated antigens pre-date the onset of

rheumatoid arthritis. Arthritis Res Ther. 2016;18:127.

13. Kastbom A, Forslind K, Ernestam S, Geborek P, Karlsson JA, Petersson IF, et al.

Changes in the anticitrullinated peptide antibody response in relation to therapeutic outcome in

early rheumatoid arthritis: results from the SWEFOT trial. Ann Rheum Dis. 2016;75:356-61.

14. Forslind K, Ahlmen M, Eberhardt K, Hafstrom I, Svensson B, Group BS. Prediction of

radiological outcome in early rheumatoid arthritis in clinical practice: role of antibodies to

citrullinated peptides (anti-CCP). Ann Rheum Dis. 2004;63:1090-5.

22

15. Koga T, Okada A, Fukuda T, Hidaka T, Ishii T, Ueki Y, et al. Anti-citrullinated peptide

antibodies are the strongest predictor of clinically relevant radiographic progression in

rheumatoid arthritis patients achieving remission or low disease activity: A post hoc analysis of a

nationwide cohort in Japan. PLoS One. 2017;12:e0175281.

16. Vossenaar ER, Despres N, Lapointe E, van der Heijden A, Lora M, Senshu T, et al.

Rheumatoid arthritis specific anti-Sa antibodies target citrullinated vimentin. Arthritis Res Ther.

2004;6(2):R142-50.

17. Pratesi F, Tommasi C, Anzilotti C, Chimenti D, Migliorini P. Deiminated Epstein-Barr

virus nuclear antigen 1 is a target of anti-citrullinated protein antibodies in rheumatoid arthritis.

Arthritis Rheum. 2006;54:733-41.

18. Pratesi F, Tommasi C, Anzilotti C, Puxeddu I, Sardano E, Di Colo G, et al. Antibodies to

a new viral citrullinated peptide, VCP2: fine specificity and correlation with anti-cyclic

citrullinated peptide (CCP) and anti-VCP1 antibodies. Clin Exp Immunol. 2011;164:337-45.

19. Bang H, Egerer K, Gauliard A, Luthke K, Rudolph PE, Fredenhagen G, et al. Mutation

and citrullination modifies vimentin to a novel autoantigen for rheumatoid arthritis. Arthritis

Rheum. 2007;56:2503-11.

20. van Heemst J, Trouw LA, Nogueira L, van Steenbergen HW, van der Helm-van Mil AH,

Allaart CF, et al. An investigation of the added value of an ACPA multiplex assay in an early

rheumatoid arthritis setting. Arthritis Res Ther. 2015;17:276.

21. Schellekens GA, Visser H, de Jong BA, van den Hoogen FH, Hazes JM, Breedveld FC,

et al. The diagnostic properties of rheumatoid arthritis antibodies recognizing a cyclic

citrullinated peptide. Arthritis Rheum. 2000;43:155-63.

23

22. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The

American Rheumatism Association 1987 revised criteria for the classification of rheumatoid

arthritis. Arthritis Rheum. 1988;31:315-24.

23. Prevoo ML, van 't Hof MA, Kuper HH, van Leeuwen MA, van de Putte LB, van Riel PL.

Modified disease activity scores that include twenty-eight-joint counts. Development and

validation in a prospective longitudinal study of patients with rheumatoid arthritis. Arthritis

Rheum. 1995;38:44-8.

24. van Gestel AM, Prevoo ML, van 't Hof MA, van Rijswijk MH, van de Putte LB, van Riel

PL. Development and validation of the European League Against Rheumatism response criteria

for rheumatoid arthritis. Comparison with the preliminary American College of Rheumatology

and the World Health Organization/International League Against Rheumatism Criteria. Arthritis

Rheum. 1996;39:34-40.

25. Wells G, Becker JC, Teng J, Dougados M, Schiff M, Smolen J, et al. Validation of the

28-joint Disease Activity Score (DAS28) and European League Against Rheumatism response

criteria based on C-reactive protein against disease progression in patients with rheumatoid

arthritis, and comparison with the DAS28 based on erythrocyte sedimentation rate. Ann Rheum

Dis. 2009;68:954-60.

26. Larsen A. How to apply Larsen score in evaluating radiographs of rheumatoid arthritis in

long-term studies. J Rheumatol. 1995;22):1974-5.

27. White IR, Royston P, Wood AM. Multiple imputation using chained equations: Issues

and guidance for practice. Stat Med. 2011;30:377-99.

24

28. Bugatti S, Bogliolo L, Vitolo B, Manzo A, Montecucco C, Caporali R. Anti-citrullinated

protein antibodies and high levels of rheumatoid factor are associated with systemic bone loss in

patients with early untreated rheumatoid arthritis. Arthritis Res Ther. 2016;18:226.

29. Llorente I, Merino L, Ortiz AM, Escolano E, Gonzalez-Ortega S, Garcia-Vicuna R, et al.

Anti-citrullinated protein antibodies are associated with decreased bone mineral density: baseline

data from a register of early arthritis patients. Rheumatol Int. 2017;37:799-806.

30. Boman A, Kokkonen H, Arlestig L, Berglin E, Rantapaa-Dahlqvist S. Receptor activator

of nuclear factor kappa-B ligand (RANKL) but not sclerostin or gene polymorphisms is related

to joint destruction in early rheumatoid arthritis. Clin Rheumatol. 2017;36:1005-12.

31. Harre U, Georgess D, Bang H, Bozec A, Axmann R, Ossipova E, et al. Induction of

osteoclastogenesis and bone loss by human autoantibodies against citrullinated vimentin. J Clin

Invest. 2012;122:1791-802.

32. Malmstrom V, Catrina AI, Klareskog L. The immunopathogenesis of seropositive

rheumatoid arthritis: from triggering to targeting. Nat Rev Immunol. 2017;17:60-75.

33. Carpenter L, Norton S, Nikiphorou E, Jayakumar K, McWilliams DF, Rennie KL, et al.

Reductions in Radiographic Progression in Early Rheumatoid Arthritis Over Twenty-Five Years:

Changing Contribution From Rheumatoid Factor in Two Multicenter UK Inception Cohorts.

Arthritis Care Res (Hoboken). 2017;69:1809-17.

34. Fiehn C, Belke-Voss E, Krause D, Wassenberg S, Rau R. Improved radiological outcome

of rheumatoid arthritis: the importance of early treatment with methotrexate in the era of

biological drugs. Clin Rheumatol. 2013;32:1735-42.

25

35. Finckh A, Choi HK, Wolfe F. Progression of radiographic joint damage in different eras:

trends towards milder disease in rheumatoid arthritis are attributable to improved treatment. Ann

Rheum Dis. 2006;65:1192-7.

36. Sokka T, Kautiainen H, Hakkinen A, Hannonen P. Radiographic progression is getting

milder in patients with early rheumatoid arthritis. Results of 3 cohorts over 5 years. J Rheumatol.

2004;31:1073-82.

37. Emery P, Kvien TK, Combe B, Freundlich B, Robertson D, Ferdousi T, et al.

Combination etanercept and methotrexate provides better disease control in very early (<=4

months) versus early rheumatoid arthritis (>4 months and <2 years): post hoc analyses from the

COMET study. Ann Rheum Dis. 2012;71:989-92.

38. Maillefert JF, Combe B, Goupille P, Cantagrel A, Dougados M. Long term structural

effects of combination therapy in patients with early rheumatoid arthritis: five year follow up of

a prospective double blind controlled study. Ann Rheum Dis. 2003;62:764-6.

39. Smolen JS, Van Der Heijde DM, St Clair EW, Emery P, Bathon JM, Keystone E, et al.

Predictors of joint damage in patients with early rheumatoid arthritis treated with high-dose

methotrexate with or without concomitant infliximab: results from the ASPIRE trial. Arthritis

Rheum. 2006;54:702-10.

40. Wigerblad G, Bas DB, Fernades-Cerqueira C, Krishnamurthy A, Nandakumar KS, Rogoz

K, et al. Autoantibodies to citrullinated proteins induce joint pain independent of inflammation

via a chemokine-dependent mechanism. Ann Rheum Dis. 2016;75:730-8.

26

FIGURE LEGENDS

Figure 1. Disease activity over time in relation to baseline anti-citrullinated protein antibody

status in early rheumatoid arthritis patients enrolled 1996-1998 (TIRA-1) and 2006-2009 (TIRA-

2). Circles show mean values and error bars show standard error of the mean.

Figure 2. Radiographic damage in different cohorts of early rheumatoid arthritis patients versus

baseline anti-citrullinated protein antibody status. * indicate p<0.05, ** p<0.01, and *** p<0.001

Figure 3. Treatment of early rheumatoid arthritis patients enrolled in the primary cohorts (a)

TIRA-1 (1996-1998) and (b) TIRA-2 (2006-2009) and in the confirmation cohorts (c) TRAM-1

(1996-2000) and (d) TRAM-2 (2006-2011). bDMARD: biologic disease-modifying anti-

rheumatic drug; csDMARD: conventional synthetic DMARD.

ADDITIONAL FILES

Additional file 1. Table showing EULAR responses after 6, 12 and 24 months in TIRA-1 and

TRAM-1 compared to the modern cohorts of TIRA-2 and TRAM-2

Additional file 2. Figure showing disease activity over time according to baseline anti-CCP

status in early rheumatoid arthritis patients enrolled 1996-2000 (TRAM-1) and 2006-2011

(TRAM-2).

0 6 12 18 24 30 362

3

4

5

6

DAS28

CCP -

CCP +p=0.777

p=0.005

months

0 6 12 18 24 30 360

2

4

6

DAS28

CCP -

CCP +p=0.007

p=0.131

months

0 6 12 18 24 30 360

10

20

30

40

CRP

CCP -

CCP +p=0.232

p=0.001

months

mg

/l

0 6 12 18 24 30 360

10

20

30

40

CRP

CCP -

CCP +p=0.007

p=0.544

months

mg

/l

0 6 12 18 24 30 360

10

20

30

40

50

ESR

CCP -

CCP +p=0.110

p=0.002

months

mm

/h

0 6 12 18 24 30 360

10

20

30

40

50

ESR

CCP -

CCP +

p=0.096

p=0.532

months

mm

/h

TIRA-1 TIRA-2

0 6 12 18 24 30 360

5

10

15

Swollen joint count

CCP -

CCP +p=0.281

p=0.001

months

no

.join

ts

0 6 12 18 24 30 360

5

10

Swollen joint count

CCP -

CCP +p<0.001

p=0.748

months

no

.jo

ints

0 6 12 18 24 30 360

2

4

6

8

10

Tender joint count

CCP -

CCP +

p=0.474

p=0.690

months

no

. jo

ints

0 6 12 18 24 30 360

2

4

6

8

10

Tender joint count

CCP -

CCP +p=0.010

p=0.649

months

no

.join

ts

0 6 12 18 24 30 360.0

0.2

0.4

0.6

0.8

1.0

1.2

HAQ

CCP -

CCP +p=0.230p=0.137

months

TIRA-1 TIRA-2

0 6 12 18 24 30 360.0

0.2

0.4

0.6

0.8

1.0

1.2

HAQ

CCP -

CCP +p=0.351

p=0.725

months

0 6 12 18 24 30 3625

30

35

40

45

50

Patient's global VAS

CCP -

CCP +p=0.764

p=0.747

months

mm

0 6 12 18 24 30 3620

30

40

50

60

Patient's pain VAS

CCP -

CCP +p=0.404

p=0.576

months

mm

0 6 12 18 24 30 360

20

40

60

Patient's pain VAS

CCP -

CCP +p=0.838

p=0.034

months

mm

0 6 12 18 24 30 360

20

40

60

Patient's global VAS

CCP -

CCP +p=0.789

p=0.006

months

mm

Baseline 3 years0

5

10

15

La

rse

n s

co

re

TIRA-2

CCP neg

CCP pos*

Baseline 2 years 0

5

10

15

La

rse

n s

co

re

TRAM-1

CCP neg

CCP pos

** ***

Baseline 2 years 0

5

10

15

La

rse

n s

co

re

TRAM-1

CCP neg

CCP pos

** ***

Baseline 2 years 0

5

10

15

La

rse

n s

co

re

TRAM-2

CCP neg

CCP pos

* **Baseline 3 years

0

5

10

15

La

rse

n s

co

re

TIRA-2

CCP neg

CCP pos*

a b

c d

Supplementary table 1. EULAR responses in historical cohorts TIRA-1 and TRAM-1 compared contemporary cohorts TIRA-2 and TRAM-2,

respectively.

Months EULAR response

TIRA-1 TIRA-2 p-value TRAM-1 TRAM-2 p-value

6 None 75 (34.1%) 58 (15.1%) 103 (46.0%) 170 (33.0%)

Moderate 82 (37.3%) 130 (33.9%) <0.001 78 (34.8%) 145 (28.2%) <0.001

Good 63 (28.6%) 196 (51.0%) 43 (19.2%) 200 (38.9%)

12 None 65 (30.1%) 56 (14.4%) 85 (37.9%) 28 (5.4%)

Moderate 81 (37.5%) 106 (27.3%) <0.001 63 (28.1%) 226 (43.9%) <0.001

Good 70 (32.4%) 226 (58.2%) 76 (33.9%) 261 (50,7%)

24 None 62 (29.2%) 57 (14.7%) 79 (35.3%) 41 (8.0%)

Moderate 77 (36.3%) 103 (26.6%) <0.001 70 (31.2%) 234 (45.4%) <0.001

Good 73 (34.4%) 227 (58.7%) 75 (33.5%) 240 (46.7%)

Additional file 2. Disease activity over time according to baseline anti-citrullinated protein

antibody status in early rheumatoid arthritis patients enrolled 1996-2000 (TRAM 1) and 2006-

2011 (TRAM 2).

1

Joint erosions visible on ultrasound predict arthritis development in

patients with anti-citrullinated protein antibodies and

musculoskeletal pain but no swollen joints

Michael Ziegelasch1*, Emma Eloff1, Hilde Berner-Hammer2, Jan Cedergren1, Klara Martinsson1, Åsa

Reckner1, Thomas Skogh1 , Mattias Magnusson1 , Alf Kastbom1

*Corresponding Author

Author affiliations 1) Department of Rheumatology, and Department of Clinical and Experimental

Medicine, Linköping University, Linköping, Sweden; 2)Department of Rheumatology, Diakonhjemmet

Hospital, Oslo, Norway

Keywords

Anti-citrullinated protein antibodies, rheumatoid arthritis, ultrasonography

2

Abstract

Background: Anti-citrullinated protein antibodies (ACPA) are associated with an increased risk of

developing rheumatoid arthritis (RA), and in particular erosive disease. Detection of joint

inflammation prior to clinical synovitis may improve treatment decisions in early disease. We sought

to determine the value of ultrasound (US) to predict arthritis development among ACPA positive

patients with musculoskeletal (MSK) pain.

Methods: We prospectively followed 104 ACPA-positive patients with MSK pain and ≤1 clinical

arthritis for mean 68 months (range 23-91). At baseline, 22 (21%) had one, arthritis upon clinical

examination, while 82 (79%) had none. US at baseline assessed joint erosions, synovial hypertrophy

in grey scale (GS), and inflammatory activity judged by power Doppler (PD) in 36 small joints and 6

tendons. The US and PD results were blinded to patients and treating rheumatologists during the

initial 3 years. US findings among patients without baseline arthritis were compared to 100 age-

matched controls. The association between US findings and arthritis development was analyzed by

Cox regression analysis.

Results: Significantly more patient joints had synovial hypertrophy (GS score >0) compared to

control joints in metacarpophalangeal (MCP) (5.2 % vs. 2.5 %; p<0.001) and proximal interphalangeal

(PIP) joints 2-5 (6.6 vs. 1.5 %; p<0.001). In contrast, metatarsophalangeal (MTP) joints 1-4 of the

controls were more often scored GS>0 compared as compared to patient joints (59% vs. 28%;

p<0.001). Positive PD (>0) occurred significantly more often in patient joints compared to the

controls in all joint areas. At subject level, the mean sum score of all investigated joints was higher

among patients than controls, regarding GS as well as PD (p<0.001 for both). 13 patients (16%), but

none of the controls, had erosions on US (p<0.001). Out of the 82 patients without clinical arthritis at

baseline, 39 (48%) developed at least one during follow-up. Neither GS nor PD findings predicted

arthritis development, but patients with baseline US erosions (n=13) had increased risk of arthritis

3

development during follow-up also after adjustment for potential confounders (Hazard ratio =4.2,

95% CI 1.7-10.4, p=0.002).

Conclusions: Arthritis-related US findings are more common among patients at increased risk of RA

compared to healthy controls, but with site-specific differences. At patient level, erosions detected

on US predicted arthritis development. Thus, US assessment of erosions improves risk-stratification

of ACPA-positive patients without swollen joints, and potentially identifies patients eligible for very

early pharmacotherapy.

4

Introduction

Subclinical autoimmune manifestations, such as circulating rheumatoid factor (RF) and/or anti-

citrullinated protein antibodies (ACPA), typically precede the onset of clinically manifest rheumatoid

arthritis (RA), as defined by the 1987 American College of Rheumatology (ACR87) and/or the 2010

Euro-American classification criteria [1, 2]. Neither of these two RA classification criteria allow an RA

diagnosis in patients suffering from musculoskeletal (MSK) pain without the presence of palpable

synovitis. However, given the benefits of modern early immunomodulating therapies in RA [3], and

the high diagnostic specificity of ACPA [4], patients within this category may benefit from anti-

rheumatic drug therapy prior to fulfilling classification criteria for RA. On the other hand, considering

the substantial risk for over-treatment with potent immunomodulation in this clinical setting, there

is a great need for predictors of disease development and progression. An imaging modality allowing

detection of subclinical joint inflammation could bring additional value in identifying those who may

benefit from early treatment.

Magnetic resonance imaging (MRI) has a high sensitivity for detection of preclinical synovitis, [5].

However, a low synovitis score on MRI may not always associate with subsequent development of

synovitis, and can also be found in healthy controls [6]. Also, the cost and access to MRI may limit its

use in many clinical settings. Ultrasound (US) is a promising imaging tool allowing assessment of soft

tissues like joint capsules, tendons and bursae. US is cheaper than MRI and, used for live scanning, it

can detect effusions and edema as signs of ongoing musculoskeletal inflammation. Grey scale (GS)

ultrasonography visualizes thickening of the synovial membrane in joints and tendons, effusions and

structural bone changes as well as erosions. Addition of Power Doppler (PD) to GS findings allows

detection of hyperemia, which is a sign of active inflammation [7]. Early diagnostic ultrasound to

detect ongoing/progressive inflammation was shown to predict progression to arthritis both among

seropositive and seronegative arthralgia patients [8-11], and the risk of progression was highest

5

among those with positive PD [12]. Furthermore, in experienced hands, US appear more sensitive

than radiography for the detection of minimal structural changes located at bone surfaces [13, 14].

US is becoming increasingly used in clinical practice. In patients with RA-related autoantibodies and

arthralgia, but no clinical arthritis, it is used for risk-stratification and possibly also when deciding on

DMARD initiation. Still, the available evidence remains limited regarding such use of US in this

patient category. Our study aimed to compare US findings between ACPA-positive patients with MSK

pain and healthy controls, and to investigate the value of US to predict arthritis development among

patients with ACPA and MSK pain without clinical arthritis at baseline.

Patients & Methods

Patients & controls

We set up a prospective observational study designated ‘TIRx’ (Swedish acronym for “X-tra early

rheumatology follow-up”), which enrolled 116 patients between 2010 and 2013 at the University

Hospital in Linköping, Sweden. The patients were referred from primary care centers within the

Östergötland County in Southeastern Sweden to the Rheumatology clinic, based on ACPA positivity

and MSK symptoms. Screening, enrollment, and follow-up were performed by 4 experienced

rheumatologists (AK, JC, TS, or ÅR).

Inclusion criteria were:

1. A positive ACPA test in clinical routine

2. MSK pain of any sort and duration.

3. Maximum one arthritis upon clinical examination.

6

Exclusion criteria were:

1. >1 palpable synovitis

2. Fulfilment of ACR87

3. Oral or intraarticular corticosteroid therapy within 6 weeks prior screening

4. Previous diagnosis of inflammatory rheumatic disease

5. <18 years of age.

Due to various reasons, 12 patients (10%) discontinued participation, and thus, 104 patients were

eligible for analysis. (Fig.1). Baseline characteristics are shown in Table 1.

Follow-up visits were scheduled at month 3, 12, 24, and 36, and thereafter every other year.

Importantly, patients were instructed to contact the clinic without delay in case of increased

symptoms between scheduled visits. At each visit we obtained a 28-joint disease activity score

(DAS28) [15], functional status by a Swedish version of the health assessment questionnaire (HAQ)

[16], and registered ongoing treatment. Pharmacotherapy and non-pharmacological interventions

were instituted as suggested by the physician and the patient’s acceptance. Follow-up in TIRx is still

ongoing, and the present study includes data collected until September 1st 2017, resulting in a mean

follow-up time of 68 months (SD 16, range 23-91).

As controls, we recruited 100 blood donors (50 females, 50 males; mean age 52 years) from the

Department of Transfusion Medicine at Linköping University Hospital. The controls underwent

serum sampling and US examination once, according to the procedure described below.

7

Ultrasound

All US examinations were performed by an experienced rheumatologist (MZ) at baseline. A

“ProFocus system” from BK Medical (BK Global Headquarters, Peabody, MA) with a linear scanner 6–

15 MHz was used. All patients were examined with the same settings for both GS and PD. The

protocol included dorsal assessments of the following 36 joints: bilateral radiocarpal, intercarpal,

distal radioulnar, metacarpophalangeal (MCP) joints 1-5, interphalangeal (IP) thumb joints, proximal

interphalangeal (PIP) joints 1-5, and metatarsophalangeal (MTP) joints 1-5.

PIP 1 is the end joint of the thumb. Its special anatomy compared to PIP 2-5 allows more stability in

sideways and twisting movements. Therefore, the PIP of the thumb is more prone to stiffness and

other problems. Secondary osteoarthritis is common in the thumb; therefore, we excluded PIP 1

from our calculations.

To grade synovitis, we used the semi quantitative scoring system by Szkudlarek et al. [17] in which

synovial hypertrophy was graded 0–3 (0 = no thickening, 1=minimal thickening not bulging over the

line linking the tops of periarticular bones, 2=thickening over but not extending the diaphysis, 3 =

synovial thickening bulging over the tops of the periarticular bones and extension over the diaphysis

of at least one side) and perfusion 0–3 (0 = no flow in the synovium, 1=single vessel signals,

2=confluent vessel signals in less than half of the synovial area, 3 = confluent vessel signals in more

than one half of the synovium). At least grade 2 synovial hypertrophy and/or PD signals are defined

as arthritis detected by US [18-20].

In addition, six tendons (extensor carpi ulnaris in the wrist, tibialis posterior and flexor digitorum

longus in the feet) were evaluated. Also, easy assessable and typical sites regarding erosions (MCP 2

and 5, ulnar head, and MTP 5) were examined bilaterally and reported as present or not.

8

The US investigator (MZ) did not participate in clinical management of the patients, and US results

were blinded to both patients and their respective physician during the first 3 years of the study.

Laboratory analyses

Serum samples were drawn at baseline and stored at -80°C until analyzed. IgG anti-CCP and

agglutinating rheumatoid factor (RF) were measured at baseline. Anti-CCP was analyzed by the 2nd

generation enzyme-immunoassay (Immunoscan CCPlus, EuroDiagnostica AB, Malmö, Sweden).

Reference values were ≥25 IU for being positive in a low titer, and ≥75 IU in a high titer.

Agglutinating RF was analyzed by nephelometry and we used the cut-off level recommended by the

manufacturer (30 U/mL). Erythrocyte sedimentation rate (ESR, mm/h) and C-reactive protein (CRP,

mg/L) were analyzed according to clinical routine at the clinical chemistry department, Linköping

university hospital.

Statistical analyses

Statistical analyses were performed using IBM SPSS Statistics 23. Continuous data were summarized

with mean values and standard deviation, and non-normally distributed data with median values

and interquartile range (IQR). To calculate the predictive value of US for developing of arthritis, we

performed Cox regression analysis, and significant findings were adjusted for baseline variables of

potential importance for arthritis development (age, sex, symptom duration, RF, ACPA levels,

smoking habits, ESR, and CRP).

9

Ethics

The study protocol was approved by the Regional Ethics Committee in Linköping (DnR 220-09 and

2015/236-32), and all participants gave written informed consent to participation.

Results

Ultrasound findings in patients and controls

The US findings in patients without clinical arthritis at baseline (n=82) were compared to the findings

among controls (n=100). Table 2 shows the results at joint level. Among patients, significantly more

joints had synovial hypertrophy (GS>0) in the MCP- and PIP-joints as compared to the controls. In

contrast, compared to the patients, the controls showed significantly more synovial hypertrophy of

the MTP1-5 joints (48.5% vs. 23.9%, p<0.001). In MTP joints, the GS findings were mostly seen in

MTP 1-4, especially among the controls. Therefore, we present MTP5 separately, and excluded MTP

1-4 joints regarding GS analyses. PD signals (PD>0) occurred significantly more often in all joint areas

among patients than controls (wrists 7.9 vs. 0.2%, p<0.001), MCPs 0.7 vs. 0%, p=0.026), PIPs 1.7 vs.

0% p=0.001), MTPs 2.0 vs. 0.2%; p=0.001). Since PD scores of all MTPs were more common in

patient MTPs, in contrast to GS MTP 1-4, we included PD from MTPs in further analyses.

13 (16%) of the patients had at least one detectable erosion on US, whereas none was found among

the controls (p<0.001). On conventional radiographs, 4 out the 13 US erosive abnormalities could be

detected.

Table 3 summarizes the results of US findings on patient level. GS and PD sum scores in hands and in

total were higher in patients versus controls, while GS results in MTP5 did not differ significantly.

However, PD scores in MTP1-5 were significantly higher among the patients.

10

Comparing differences of US findings between the sexes among patients, females had significantly

higher GS scores (mean 1.1 vs. 0.2; p=0.015) and PD scores (mean 0.23 vs. 0; p=0.013) in the PIP

joints 2-5. No significant differences in other joint regions or tendons were observed. Male controls

had significantly higher GS sum scores in the wrists (1.9 vs. 0.9; p=0.011) and MCPs (0.54 vs. 0.06;

p=0.002) compared with female controls, whereas in PIP joints no differences could be observed.

However, the occurrence of PD signals were rare in healthy subjects; none detected in MCP- and

PIP-joints, and in the wrists there was no significant differences between the sexes (0.1 vs. 0;

p=0.103).

Arthritis at baseline and follow-up in the TIRx cohort

Among the 104 patients participating in follow-up, 22 (21%) had one arthritis upon clinical

examination at baseline, and hence 82 (79%) did not (Figure 1). The baseline characteristics are

listed in Table 1.

Among the 22 patients with 1 clinical arthritis at baseline, 18 had arthritis in joints covered by the

standardized US examination protocol. 6 had no US findings in these joints, 3 had synovial

hypertrophy grade 1, and hence the remaining 9 patients had US defined arthritis (7 by positive PD

and 2 by GS grade ≥2). Of the 82 patients without clinical arthritis at baseline, 39 (48%) developed

clinical arthritis after median 25 weeks (range 5-302 weeks).

Ultrasound findings and arthritis development

We tested each US modality in Cox regression with future arthritis development as outcome (table

4). Neither GS nor PD scores associated with arthritis development. However, 10 out of the 13 (77%)

patients with ≥1 baseline erosion on US developed arthritis during follow-up, as compared to 29/69

(42%) among those without erosions (p=0.032). In Cox regression analysis, baseline erosion

11

predicted progression to arthritis (Hazard ratio (HR) 2.8, 95% CI 1.4-5.8, p=0.005; table 4). After

adjusting for potential confounders (sex, age, symptom duration, smoking habits, ESR, CRP, RF, and

ACPA levels) in the Cox regression model, the association between US erosion and arthritis

development remained statistically significant (HR=4.2, 95% CI 1.7-10.4); p=0.002) (Figure 2). Since

seven of the patients started treatment with DMARDs or oral corticosteroids during follow-up,

without arthritis being confirmed on clinical examination, we also performed Cox regression

analyses excluding these patients. In these analyses, US erosions remained significantly associated

with arthritis development (HR 4.2; 95% CI 1.7-10.0; p=0.001), while GS and PD were still not.

Discussion

The aim of this prospective observational study was to investigate the value of US in predicting

arthritis development among ACPA positive patients with musculoskeletal (MSK) pain. As a second

aim, since this is a patient category without findings on clinical examination, we wished to compare

the US findings in patients with those from healthy age-matched individuals. We found that,

compared to healthy blood donors, ACPA-positive patients with MSK pain in total had more

inflammatory joint changes on US examination, indicating ongoing inflammation in this patient

category. Contrastingly, MTP1-4 of controls had significantly more signs of synovial hypertrophy on

GS (59% of the examined joints) as compared to the patients (28%). This important finding is in line

with Nam et al [12], who found GS>1 in at least one MTP joint in 42% of control subjects. Taken

together, GS scores of MTPs should be interpreted with caution in at-risk individuals as it is also

common in the general population.

Evidence regarding US prediction of arthritis development in at-risk populations is limited, and more

data is needed. Van der Stadt et al demonstrated a predictive value of US abnormalities at a joint

level, but not at a patient level [11]. Nam et al found PD, US erosions, and GS (when excluding MTPs)

12

to be predictive of arthritis development at patient level in ACPA positive cases [12]. In the current

study, we failed to show significant predictive values at patient level of either GS or PD at baseline.

This finding is in line with van de Stadt et al from the Netherlands, but discordant to Nam and

coworkers in the UK. Autoantibody differences are unlikely to explain these disparities, since we and

the UK group enrolled ACPA-positive patients only, while the Dutch cohort consisted of one third

ACPA-/RF+ patients. Also, symptom definition and duration appear fairly similar between the British

and the present study.

The presence of erosions on US predicts arthritis development in our study, also after adjusting for

potential confounders. Since erosions were not rare in this population (16%), these results suggest

that US is a valuable tool to risk-stratify ACPA-positive patients with MSK pain, despite the lack of

predictive value concerning GS and PD scoring. Future studies will address whether US erosions also

predict long-term progression of joint damage as determined on conventional radiographs.

A strength of the current study is the long follow-up period of almost 6 years in median, which

increases the chances to identify those at risk of developing arthritis after several years. Also, the

large control population puts the US data in perspective, for instance demonstrating that GS scores

in MTP 1-4 must be interpreted with caution. The US examinations were performed by only one

investigator, which in fact can be considered positive, since US findings were consistently assessed

and scored. Thus, a possible low inter-observer reliability and disagreement can be avoided.

Furthermore, the results were blinded to patients and the treating physicians, thereby removing the

bias of clinical judgement and treatment decisions influenced by US results, which could have

influenced the primary outcome.

A limitation of the study is that treatment was not defined in the protocol. Importantly, however,

analyses excluding the 7 patients who were subjected to corticosteroid and/or disease-modifying

anti-rheumatic drug therapy without having developed arthritis did not alter the results in any

substantial way. Another potential drawback is the fact that the US investigator was not blinded to

13

participant status (patient versus control). We believe, however, that the risk of over-rating findings

in patients seems limited given the increased GS scores in control MTPs. The controls were matched

for age but not for sex. Indeed, there were some sex differences in US findings among controls, but

they were limited to GS. Finally, due to practical reasons, arthritis development was not confirmed

by a second investigator or compared with US findings in the same joint. However, clinical

investigators were experienced rheumatologists, and patients were seen by the same doctor at the

vast majority of visits. Having previously examined the patient in an arthritis-free status likely

reduces the risk of misclassifying a developing arthritis.

Conclusions

ACPA positive patients with musculoskeletal pain show significantly more signs of inflammatory

changes on baseline US compared to healthy controls. Joint erosions detected by US at baseline

were predictive for arthritis development, and therefore, US can aid in assessing prognosis and

tailoring treatment in ACPA-positive patients with MSK pain.

14

Abbreviations

ACPA – Anti-citrullinated protein antibodies

ACR – American College of Rheumatology

AU – Arbitrary units

CCP – Cyclic citrullinated Peptides



DAS28 – Disease Activity Score


GS – Grey scale


IP – Interphalangeal

MCP – Metacarpophalangeal

MRI – Magnetic resonance imaging

MSK – Muskoloskeletal

MTP – Metatarsophalangeal

PD – Power Doppler

PIP – Proximal interphalangeal

PPV – Positive predictive value



15

References

1. Rantapää-Dahlqvist, S., et al., Antibodies against cyclic citrullinated peptide and IgA rheumatoid factor predict the development of rheumatoid arthritis. Arthritis Rheum, 2003. 48(10): p. 2741-9.

2. Verheul, M.K., et al., Triple Positivity for Anti-Citrullinated Protein Autoantibodies, Rheumatoid Factor, and Anti-Carbamylated Protein Antibodies Conferring High Specificity for Rheumatoid Arthritis: Implications for Very Early Identification of At-Risk Individuals. Arthritis Rheumatol, 2018.

3. Donahue, K.E., et al., in Drug Therapy for Early Rheumatoid Arthritis: A Systematic Review Update. 2018: Rockville (MD).

4. Avouac, J., L. Gossec, and M. Dougados, Diagnostic and predictive value of anti-cyclic citrullinated protein antibodies in rheumatoid arthritis: a systematic literature review. Ann Rheum Dis, 2006. 65(7): p. 845-51.

5. Gent, Y.Y., et al., Three-year clinical outcome following baseline magnetic resonance imaging in anti-citrullinated protein antibody-positive arthralgia patients: an exploratory study. Arthritis Rheumatol, 2014. 66(10): p. 2909-10.

6. Parodi, M., et al., How normal are the hands of normal controls? A study with dedicated magnetic resonance imaging. Clin Exp Rheumatol, 2006. 24(2): p. 134-41.

7. Porta, F., et al., The role of Doppler ultrasound in rheumatic diseases. Rheumatology (Oxford), 2012. 51(6): p. 976-82.

8. Backhaus, M., et al., Prospective two year follow up study comparing novel and conventional imaging procedures in patients with arthritic finger joints. Ann Rheum Dis, 2002. 61(10): p. 895-904.

9. Freeston, J.E., et al., A diagnostic algorithm for persistence of very early inflammatory arthritis: the utility of power Doppler ultrasound when added to conventional assessment tools. Ann Rheum Dis, 2010. 69(2): p. 417-9.

10. Wakefield, R.J., et al., Should oligoarthritis be reclassified? Ultrasound reveals a high prevalence of subclinical disease. Ann Rheum Dis, 2004. 63(4): p. 382-5.

11. van de Stadt, L.A., et al., The value of ultrasonography in predicting arthritis in auto-antibody positive arthralgia patients: a prospective cohort study. Arthritis Res Ther, 2010. 12(3): p. R98.

12. Nam, J.L., et al., Ultrasound findings predict progression to inflammatory arthritis in anti-CCP antibody-positive patients without clinical synovitis. Ann Rheum Dis, 2016. 75(12): p. 2060-2067.

13. Camerer, M., et al., High-resolution ultrasound of the midfoot: sonography is more sensitive than conventional radiography in detection of osteophytes and erosions in inflammatory and non-inflammatory joint disease. Clin Rheumatol, 2017. 36(9): p. 2145-2149.

14. Wakefield, R.J., et al., The value of sonography in the detection of bone erosions in patients with rheumatoid arthritis: a comparison with conventional radiography. Arthritis Rheum, 2000. 43(12): p. 2762-70.

15. Prevoo, M.L., et al., Modified disease activity scores that include twenty-eight-joint counts. Development and validation in a prospective longitudinal study of patients with rheumatoid arthritis. Arthritis Rheum, 1995. 38(1): p. 44-8.

16. Wells, G., et al., Validation of the 28-joint Disease Activity Score (DAS28) and European League Against Rheumatism response criteria based on C-reactive protein against disease progression in patients with rheumatoid arthritis, and comparison with the DAS28 based on erythrocyte sedimentation rate. Ann Rheum Dis, 2009. 68(6): p. 954-60.

17. Szkudlarek, M., et al., Interobserver agreement in ultrasonography of the finger and toe joints in rheumatoid arthritis. Arthritis Rheum, 2003. 48(4): p. 955-62.

16

18. Horton, S.C., et al., Ultrasound-detectable grey scale synovitis predicts future fulfilment of the 2010 ACR/EULAR RA classification criteria in patients with new-onset undifferentiated arthritis. RMD Open, 2017. 3(1): p. e000394.

19. van der Ven, M., et al., No clear association between ultrasound remission and health status in rheumatoid arthritis patients in clinical remission. Rheumatology (Oxford), 2017. 56(8): p. 1276-1281.

20. Witt, M., et al., Relevance of grade 1 gray-scale ultrasound findings in wrists and small joints to the assessment of subclinical synovitis in rheumatoid arthritis. Arthritis Rheum, 2013. 65(7): p. 1694-701.

17

Table 1 Baseline characteristics of ACPA positive individuals with musculoskeletal pain and maximum 1 synovitis

Characteristic Without clinical arthritis (n=82)

With clinical arthritis (n=22)

p

Age (SD) 51.8 (14.1) 51.9 (18.6) 0.987 Gender, females 66 (80.5%) 17 (77.3%) 0.739 Symptom duration 0-6 months

6-12 months >18 months

15 (18.3%) 37 (45.1%) 30 (36.6%)

6 (27.3%) 7 (31.8%) 9 (40.9%)

0.471

ACPA-level Low High

32 (39.0%) 50 (61.0%)

8 (36.4%) 14 (63.6%)

1.000

RF Negative Positive

58 (70.7%) 24 (29.3%)

12 (54.5%) 10 (45.5%)

0.201

CRP (SD) 6.4 (6.0) 10.8 (12.5) 0.118 ESR (SD) 11.9 (9.5) 16.3 (15.4) 0.216 DAS28 (SD) 2.49 (1.05) 3.35 (0.93) 0.001 Smoking Nonsmoker

Ex-smoker Current smoker

43 (52.4%) 26 (31.7%) 13 (15.9%)

14 (63.6%) 3 (13.6%) 5 (22.7%)

0.234

HAQ 3.46 (4.38) 0.55 (0.40) 0.139

ACPA Anti-citrullinated protein antibodies; CRP C-reactive protein; DAS Disease activity score; ESR Erythrocyte sedimentation rate; HAQ RF Rheumatoid factor; SD Standard deviation

18

Table 2 Ultrasound findings in patients with musculoskeletal pain and anti-citrullinated protein antibodies but without arthritis compared to controls. The scores are bilateral sum scores of the different joint areas.

GS PD Joint Score Patients (n=82) Controls (n=100) Patients (n=82) Controls (n=100)

Wrist 0 78.7% (387/492) 83.5% (501/600) 92.1% (453/492) 98.0% (588/600) 1 12.4% (61/492) 9.5% (57/600) 6.1% (30/492) 1.2% (7/600) 2 8.5% (42/492) 6.7% (40/600) 1.6% (8/492) 0.8% (5/600) 3 0.4% (2/492) 0.3% (2/600) 0.2% (1/492) - p=0.243 p<0.001 MCP1-5 0 94.8% (777/820) 97.5% (975/1000) 99.3% (814/820) 100% (1000/1000) 1 1.7% (14/820) 2.0% (20/1000) 0.6% (5/820) - 2 3.0% (25/820) 0.5% (5/1000) 0.1% (1/820) - 3 0.5% (4/820) - 0% (0/820) - p<0.001 p=0.026 PIP2-5 0 93.4% (613/656) 98.5% (788/800) 98.3% (645/656) 100% (800/800) 1 1.5% (10/656) 0.9% (7/800) 1.2% (8/656) - 2 4.6% (30/656) 0.6% (5/800) 0.5% (3/656) - 3 0.5% (3/656) - - - p<0.001 p=0.001 MTP1-4 0 72.1% (473/656) 41.3% (330/800) 97.9% (642/656) 99.8% (798/800) 1 5.8% (38/656) 21.4% (171/800) 1.2% (8/656) - 2 20.3% (133/656) 37.3% (298/800) 0.9% (6/656) 0.3% (2/1000) 3 1.8% (12/656) 0.1% (1/800) - - p<0.001 p=0.001 MTP5 0 92.1% (151/164) 92.5% (185/200) 98.8% (162/164) 100% (200/200) 1 3.0% (5/164) 6.0% (12/200) 0.6% (1/164) - 2 4.9% (8/164) 1.5% (3/200) 0.6% (1/164) - 3 - - - - p=0.079 p=0.293 Tendons 0 94.5% (465/492) 98.7% (602/610) 98.8% (486/492) 100% (610/610) 1 3.3% (16/492) 0.8% (5/610) 0.4% (2/492) - 2 1.8% (9/492) 0.5% (3/610) 0.6% (3/492) - 3 0.4% (2(492) - 0.2% (1/492) - p=0.001 p=0.058

GS, Grey scale; MCP, metacarpophalangeal joint; MTP, metatarsophalangeal joint; PIP, proximal interphalangeal joint of the finger; PD, power Doppler

19

Table 3. Baseline ultrasound sum scores in patients without clinical arthritis at baseline compared to

controls. Values are mean (standard deviation) unless otherwise indicated

Patients (N=82) Controls (N=100) p

Hands GS PD

4.0 (4.7) 1.0 (1.8)

2.1 (2.5) 0.04 (0.2)

0.001 <0.001

MTP1-4

GS PD

4.1 (4.1) 0.2 (0.7)

7.7 (4.3) 0.04 (0.3)

<0.001 0.016

MTP5

GS PD

0.3 (0.8) 0.04 (0.2)

0.2 (0.6) 0

0.462 0.181

Total (Hands + MTP5) Total (Hands + MTP1-5)

GS PD

4.0 (4.2) 1.3 (2.0)

2.0 (2.3) 0.1 (0.4)

<0.001 <0.001

Tendons GS PD

0.5 (1.1) 0.1 (0.6)

0.1 (0.4) 0 (0)

0.005 0.063

Erosions present GS 13/82 (15.9%) 0 <0.001

Hands include wrists, metacarpophalangeal joints 1-5, and proximal interphalangeal joints 2-5. GS Grey scale; MTP Metatarsophalangeal joint; PD Power Doppler

Table 4 Cox regression analysis of ultrasound findings in patients without baseline arthritis (n=82)

with arthritis development as dependent variable.

US modality/score Hazard ratio 95% CI p

Grey scale 0-1 2-3 ≥4

Reference 1.53 1.48

0.65-3.60 0.70-3.13

0.33 0.31

Power Doppler 0 ≥1

Reference 1.68

0.89-1.15

0.11

Erosions 0 ≥1

Reference 2.82

1.37-5.82

0.005

20

Figure 1.

21

Figure 2.

No erosions (n=69)

Erosions (n=13)

adjusted p= 0.002

22

Figure legends

Figure 1. Distribution of patients during follow-up

Figure 2. Survival plot illustrating progression to arthritis during follow-up in relation to the presence

of ultrasound erosions at baseline among patients with anti-citrullinated protein antibodies

and musculoskeletal pain.

diagnostic and prognostic potential of joint imaging in ... · decrease in bone mineral density...

Documents