Optimisation of rheumatology indices: dactylitis and enthesitis in E.G. Ferguson, L.C. Coates

Leeds Institute of Rheumatic and ABSTRACT sites of attachment to bone of tendon, Musculoskeletal Medicine, University Outcome measures are a key part of ligament, or joint capsule (4). This can of Leeds, Chapel Allerton Hospital, study design and clinical assessment. cause pain, tenderness, and swelling at Chapeltown Road, Leeds, United Kingdom. Enthesitis and dactylitis are typical these sites and is estimated from regis- Esme G. Ferguson, BSc, MBChB, MRCP features of psoriatic arthritis (PsA) try data that 30–50% of patients with Laura C. Coates, MBChB, PhD, MRCP and the spondyloarthritides but tradi- a diagnosis of PsA have enthesitis (5). Please address correspondence to: tionally scoring systems for enthesitis The use of clinical assessment tools for Laura C. Coates, have mainly been validated in ankylos- Leeds Institute of Rheumatic enthesitis has now become widespread and Musculoskeletal Medicine, ing spondylitis (AS). There are many in clinical trials despite the debate University of Leeds, scoring systems which are not validat- about which particular scoring system Chapel Allerton Hospital, ed used for dactylitis although newer is optimal. In particular, establishing Chapeltown Road, validated scores are now available. criterion validity has been difficult be- Leeds LS7 4SA, United Kingdom. Recently there have been advances in cause of a lack of gold standard. Ide- E-mail: [email protected] composite scores that include enthesi- ally a gold standard would have asso- Received on September 2, 2014; accepted tis and dactylitis to assess disease ac- ciated evidence of tissue abnormality in revised form on September 10, 2014. tivity. These are currently being vali- from histopathology studies. However Clin Exp Rheumatol 2014; 32 (Suppl. 85): dated further and have not yet been biopsy of tendons is neither safe nor S113-S117. tested in routine clinical practice. easy, and there are limited research © Copyright Clinical and data available. MRI has been shown to Experimental Rheumatology 2014. Introduction identify bone marrow oedema at tendon Psoriatic arthritis (PsA) is an immune insertions and abnormal signal around Key words: psoriatic arthritis, mediated disease in which there is the enthesis (6). Ultrasound scanning enthesitis, dactylitis, outcome heterogeneity in its presentation and using grey-scale and power Doppler measures, clinical assessment course, which contributes to the com- to identify increased vascularisation in plexity in diagnosis and assessment of and around the enthesis can identify PsA. The key clinical features of PsA abnormal findings in symptomatic and include joint, skin, nail, axial disease, asymptomatic entheses (7). Given that enthesitis, and dactylitis. Diagnosis of ultrasound identified power Doppler PsA based on the Classification of Pso- signal, and MRI bone marrow oedema riatic Arthritis (CASPAR) Study Group have been correlated with tissue evi- criteria has been well validated (Table dence of in other -rheu I) (1, 2). matic manifestations, it seems likely Outcome measures used in clinics and that imaging is the best gold standard research trials are well established in available at present (8). However cor- RA and AS. The Group for Research relation of imaging with clinically ap- and Assessment of and Pso- preciable tenderness or swelling is riatic Arthritis (GRAPPA) and the Out- limited particularly in enthesitis, which come Measures in Rheumatology clini- may have implications for validation cal trials (OMERACT) have identified of clinical measures (9). In addition to a core set of domains for PsA to be as- the soft tissue changes visualised, MRI sessed in clinical trials, including dac- scanning has identified the involve- tylitis and enthesitis (Fig. 1) (3). Over ment of bone adjacent to the enthesis. the last decade clinical outcome assess- No studies have addressed whether this ments for PsA have evolved. can be clinically identified and wheth- er it correlates with clinical enthesitis Enthesitis counts. Studies have also shown that Enthesitis is a recognised important ultrasound indices for enthesitis such as manifestation of spondyloarthropa- the Madrid Sonographic Enthesitis In- Competing interests: none declared. thies characterised by inflammation at dex (MASEI) can differentiate between

S-113 Optimisation of rheumatology indices / E.G. Ferguson & L.C. Coates

Table I. CASPAR Criteria. other disease activity measures in AS (12). This index has not been used in Domains Description randomised control trials likely due to Essential Criteria Inflammatory (≥1 needed) burden of administration and concern 1. Inflammatory Joints relating to sites overlapping with fibro- 2. Inflammatory axial disease myalgia points. 3. Entheseal disease The Maastrict Additional Categories (≥3 points needed) Enthesitis Score (MASES) was devel- 1. Psoriasis oped during the validation of the MEI. • Current Skin/scalp psoriasis present During a 2 year period AS patients had History of psoriasis an MEI done and the 13 most specific • History History of psoriasis in first or second degree relative • Family history and sensitive sites were chosen to be Typical Psoriatic nail dystrophy, including onycholy- included in the reduced MASES score 2. Psoriatic nail involvement sis, pitting, and hyperkeratosis observed on current with a dichotomous 0/1 score for ten- examination derness. There was correlation between 3. A negative test for RF the enthesitis scores and disease activ- 4. Dactylitis Swelling of an entire finger ity measures (12). The MASES has not • Current A history of dactylitis recorded by a rheumatologist been validated in PsA, though in the • History Ill-defined ossification near joint margins -(but ex International Spondyloarthritis Interob- Radiological evidence of juxtaarticular new cluding osteophyte formation) on plain radiographs bone formation of a hand or foot. server Reliability Exercise (INSPIRE) in which a number of enthesitis indices were compared for use in AS or PsA there was moderate intraobserver reli- ability among PsA patients (ICC 0.56, 95% CI 0.34, 0.82) (13). Data from 24 weeks and 52 weeks in the golimumab PsA trials indicate MASES demon- strates discrimination and responsive- ness (14, 15). More recently, the Spondyloarthri- tis Research Consortium of Canada (SPARCC) created a new outcome measure for enthesitis in SpA (16) using information from ultrasound and MRI studies in PsA, healthy controls and AS patients. They identified the 16 most frequently affected entheseal sites that could be clinically assessed.(7). Inter- observer reliability was good and a sub- stantial correlation was seen between the enthesitis score and other disease activity measures. Generally enthesitis Fig. 1. Domains for Psoriatic Arthritis (3). is felt to improve with anti-TNF which was the case using this measure, though those with PsA and healthy controls (9). on a semi-quantitative score from 0 to the reduction in enthesitis score was US has been shown to be more sensi- 3 (0 = no pain, 1 = mild tenderness, 2 = not significant after 12 weeks of therapy tive than clinical examination from de- moderate tenderness, and 3 = wince or (16). Reduced versions of the SPARCC tecting enthesitis, but the significance withdraw) (11). A further study showed enthesitis index using more commonly of this is currently unclear and there are correlation of the Mander Enthesitis In- involved sites showed larger effect difficulties with which areas to scan as dex (MEI) with pain and stiffness VAS sizes and standardised response means. USS can be time consuming (10). scales and a reduction in the score with This would be useful in clinical practice The first enthesitis index was developed NSAID treatment. There was some as it would take less time but still iden- by Mander et al. A list of all entheses variability between different examiners tify improvement in enthesitis (16). easily accessible to clinical examina- performing the MEI, but intra-observer All of the entheseal outcome measures tion was created and this was tested on variability was not formally tested (11). discussed previously were developed 19 patients with AS, which resulted in Further validation provided evidence for and validated a measure of 66 entheseal sites graded of a correlation between the MEI and on patients with AS. The Leeds En-

S-114 Optimisation of rheumatology indices / E.G. Ferguson & L.C. Coates thesitis Index (LEI) is the only meas- Table II. Ehtheseal sites assessed in outcome measures*. ure developed specifically for PsA. The MASES SPARCC LEI 6 most commonly involved entheseal sites were identified using a step-wise First costochondral R, L process (Table II) (17). This index was Seventh costochondral R, L then compared to other entheseal indi- Supraspinatous insertion R, L Lateral epicondyle humerous R, L R, L ces in an open-label longitudinal study. Medial epicondyle humerous R, L The LEI showed closest correlation Posterior superior iliac crest R, L with other disease activity measures, Anterior superios iliac crest R, L a large effect size and the smallest Iliac crest R, L Fifth lumbar spinous process X floor effect when compared with the Proximal achilles R, L R, L R, L MEI. This low floor effect means that Greater trochanter R, L it can identify the majority of patients Medical condyle femur R, L Lateral condyle femur with enthesitis using just 6 sites, mak- Insertion plantar fascia R, L ing it far more feasible (17). The LEI Quadriceps insertion patella R, L has been used in a randomised control Inferior pole patella R, L study with certolizumab in PsA with Tibial tubercle R, L Estimated time to complete ~2-5 minutes ~2-5 minutes ~ 30 seconds significant improvements in the -treat ment group arms compared to the pla- *MASES: Maastrict Ankylosing Spondylitis Enthesitis Score; SPARCC: Spondyloarthritis Research cebo, indicating the ability of the LEI to Consortium of Canada; LEI: Leeds Enthesitis Index; X: single site; R: right; L: left. demonstrate responsiveness (18). In the INSPIRE study, in patients with PsA, len digits and the contralateral normal both the size and tenderness so that the both LEI and SPARCC showed excel- digits. Therefore it seems likely that score can differentiate between tender lent agreement (13). in normal clinical practice, there will and non-tender dactylitis. The tender- The other limitation of clinical enthesi- be some variation between observers ness scoring can be based on the RAI tis counts is the specificity of the find- resulting in lower agreement particu- with tenderness scored from 0-3 (LDI ing of tenderness in these areas. Many larly in grey cases, where digits may scoring) or can be simplified to a di- of the entheseal points are relatively be slightly swollen. This was later con- chotomous score of 0 for non-tender near to joints and accepted fibromyalgia firmed by a reliability study performed and 1 for tender (LDI basic) (22). points, raising the possibility that mis- in Canada. This showed a moderate The first study comparing dactylitis classification could occur. The key to agreement (kappa 0.57, 95%CI=0.34, outcome measures showed a relatively the reliability of these tools in clinical 0.82) between 10 experienced observers poor inter-observer reliability for identi- practice is the training provided to as- for number of digits with dactylitis (20). fying tender dactylitis and a poor agree- sessors in localising the correct points. Clinical measures of dactylitis have ment on non-tender dactylitis. This was been used as secondary outcome meas- improved significantly by using the Dactylitis ures in clinical trials but the majority LDI scoring system. Inter and intraob- Dactylitis describes a uniform swelling have used non-validated measures. The server reliability for the LDI score was of a digit with inflammation causing a simplest measure used is a simple count good, and was increased further using sausage digit and is a hallmark feature of dactylitic digits, though some stud- the LDI basic, suggesting that some of of PsA and is one of the items used ies have used non-validated physician the variability was due to the inaccu- to make a diagnosis of PsA using the graded severity which previously has racy of grading tenderness (22). CASPAR criteria (1). Dactylitis can be been shown to have poor inter-observer A longitudinal study with 28 patients further characterised as acute/tender reliability, which adds to the difficulties who were changing treatment was per- dactylitis where the digit is tender, often of devising a measure suitable in clini- formed to further investigate the use erythematous and warm, or as chronic/ cal practice (21). of this clinical tool and to compare it sub-acute/non-tender dactylitis where The Leeds Dactylitis Instrument was to other measures (tender dactylitis the digit is swollen but non-tender. It developed in response to this need for count, all dactylitis count, IMPACT1, has been hypothesised that the chronic a clinical, objective, validated outcome LDI, LDI basic). All measures showed form occurs following an episode of measure for dactylitis (see Figure 2). a change with treatment after 3 and 6 acute dactylitis in some patients but this Based on the evaluation of the median months. The majority of these corre- has not been confirmed. difference in digital circumference be- lated with other clinical disease activ- The definition and pathology of dacty- tween dactylitic digits and control dig- ity measures such as joint counts and litis remains problematic. Studies using its dactylitis was defined as an increase VAS for disease activity (23). Only the imaging have confirmed that physical in circumference of the digit of more count of all dactylitic digits performed examination can identify pathology than 10% compared to the contralateral badly probably due to the inclusion of in tender dactylitis (19). However this non-affected digit (22). The aim of the non-tender dactylitis which may not be study only assessed 12 obviously swol- LDI is to provide a quantification of thought to represent disease activity.

S-115 Optimisation of rheumatology indices / E.G. Ferguson & L.C. Coates

Research agenda • Further research is needed in en- thesitis to investigate the usefulness and validity of using these scoring systems in clinical practice as well as clinical trials. Further correlation with imaging is likely to be interest- ing in future studies. • Further validation of the composite indices will indicate if these may be useful in practice as well as clinical trials. • Further evaluation of dactylitis measures in clinical practice and whether using a specific outcome measure is more useful than using a tender dactylitis count.

Conclusion There has been much progress in out- come measures in PsA in the last dec- Fig. 2. A dactylitic digit being measured using the dactylometer (30). ade. There are now validated scoring systems for enthesitis and dactylitis A subgroup of patients in the above trials and in the clinical setting, this is along with composite measures that study also had MRI scans performed currently an area that is on the research include these elements. Many are used at baseline and 6 months to assess the agenda for GRAPPA. in clinical trials and have shown good inflammation in the dactylitic digits. The Composite Psoriatic Disease Activ- sensitivity to change, however their use Similar to the Olivieri study (19), this ity Index (CPDAI) assesses 5 domains in clinics may be limited in part by time showed that clinically tender dactylitic (joints, skin, entheseal, dactylitis, and and lack of knowledge and education digits had significant MRI abnormali- spinal manifestations) with a measure about these tools. Studies of enthesitis ties compared to non-involved digits of disease activity and impact on the have shown a discord between findings or non-tender dactylitis. However the patient for each domain (25). This was on MRI or USS and clinical assessment, correlation between the level of inflam- recently compared to the Disease activ- however whether this is clinically rel- mation on MRI and clinical evaluation ity in Psoriatic Arthritis (DAPSA) (26) evant needs to be further investigated. was moderate at best (0.37 for LDI lo- in data sets from a randomised control Though some of the entheseal tools are cal score and MRI score) (24). trial using etanercept.Both were effec- quick many assessment points are near Although the LDI and LDI basic meas- tive in determining treatment response to joints and may be positive if there ures do take longer to perform, par- however CPDAI could distinguish is active joint inflammation or chronic ticularly if multiple digits are involved, response between the two etanercept damage, reducing sensitivity. Dactylitis these measures perform better in terms doses suggesting it may be a more assessments such as the LDI are likely of both truth and discrimination when sensitive tool (27). More recently the useful in research trials, however their considering the tool in the context of GRACE project for development of use in clinical practice may be limited the OMERACT filter (23). Thus, it is psoriatic arthritis indices led to the Pso- due to time constraints. Many tools the most validated clinical outcome riatic Arthritis Disease Activity Score have shown to be strongly associated measure available for dactylitis. In a (PASDAS) which measures physician with other measures of disease activity randomised control trial where dacty- and patient VAS, swollen and tender such as VAS pain and stiffness scores, litis was a secondary outcome the LDI joint count, CRP, enthesitis, dactylitis which identifies the question of the ad- was able to identify improvements in count, and the physician component ditional benefits of specific measures. dactylitis in the treatment groups (18). summary of the short-form 36 (28). Ini- In busy clinical practice, generic meas- tial comparisons with other composite ures such as VAS scales for disease ac- Composite measures of psoriatic measures such as the CPDAI suggested tivity and measures of disease impact arthritis involving enthesitis and the PASDAS was better able to discrim- such as the HAQ can be very useful, dactylitis measurements inate between high and low disease ac- but in such a heterogenous disease phy- Given the complexity of PsA and the tivity. This was further compared using sicians assessments should consider all multiple areas that can be affected data from a golimumab study and the aspects of disease to ensure that a holis- composite scores may be useful in pro- PASDAS was better able to distinguish tic approach to treatment is taken. It is viding a tool that can be used both in treatment effect (29) likely that in clinical practice composite

S-116 Optimisation of rheumatology indices / E.G. Ferguson & L.C. Coates measures could have great benefit and arthritis from psoriasis alone and healthy Dactylitis in patients with seronegative spon- more needs to done to further investi- controls? J Rheumatol 2014; 41: 466-72. dylarthropathy. Assessment by ultrasonogra- 10. BALINT PV, KANE D, WILSON H, McINNES phy and magnetic resonance imaging. Arthri- gate their use in long-term observation- IB, STURROCK RD et al.: Ultrasonography tis Rheum 1996; 39: 1524-8. al cohorts which would indicate which of entheseal insertions in the lower limb in 20. GLADMAN DD, COOK RJ, SCHENTAG C et ones may be useful in clinical practice. spondyloarthropathy. Ann Rheum Dis 2002; al.: The clinical assessment of patients with 61: 905-10. psoriatic arthritis: results of a reliability 11. MANDER M, SIMPSON JM, McLELLAN A, study of the spondyloarthritis research con- References WALKER D, GOODACRE JA, DICK WC: Stud- sortium of Canada. J Rheumatol 2004; 31: 1. TAYLOR W, GLADMAN D, HELLIWELL P et ies with an enthesis index as a method of 1126-31. al.: Classification criteria for psoriatic arthri- clinical assessment in ankylosing spondyli- 21. FUCHS HA, BROOKS RH, CALLAHAN LF, tis: development of new criteria from a large tis. Ann Rheum Dis 1987; 46: 197-202. PINCUS T: A simplified twenty-eight-joint international study. Arthritis Rheum 2006; 12. HEUFT-DORENBOSCH L, SPOORENBERG quantitative articular index in rheumatoid 54: 2665-73. A, VAN TUBERGEN A et al.: Assessment of arthritis. Arthritis Rheum 1989; 32: 531-7. 2. COATES LC, CONAGHAN PG, EMERY P et enthesitis in ankylosing spondylitis. Ann 22. HELLIWELL PS, FIRTH J, IBRAHIM GH, al.: Sensitivity and specificity of the classi- Rheum Dis 2003; 62: 127-32. MELSOM RD, SHAH I, TURNER DE: Develop- fication of psoriatic arthritis criteria in early 13. GLADMAN DD, INMAN RD, COOK RJ et al.: ment of an assessment tool for dactylitis in psoriatic arthritis. Arthritis Rheum 2012; 64: International spondyloarthritis interobserver patients with psoriatic arthritis. J Rheumatol 3150-5. reliability exercise--the INSPIRE study: II. 2005; 32: 1745-50. 3. GLADMAN DD, MEASE PJ, STRAND V et al.: Assessment of peripheral joints, enthesitis, 23. HEALY PJ, HELLIWELL PS: Measuring dacty- Consensus on a core set of domains for psori- and dactylitis. J Rheumatol 2007; 34: 1740-5. litis in clinical trials: which is the best instru- atic arthritis. J Rheumatol 2007; 34: 1167-70. 14. KAVANAUGH A, McINNES I, MEASE P, et al.: ment to use? J Rheumatol 2007; 34: 1302-6. 4. McGONAGLE D, LORIES RJ, TAN AL, BENJA- Golimumab, a new human tumor necrosis 24. HEALY P, GROVES C, CHANDRAMOHAN M, MIN M: The concept of a “synovio-entheseal factor alpha antibody, administered every HELLIWELL PS: MRI changes in psoriatic complex” and its implications for under- four weeks as a subcutaneous injection in dactylitis - extent of pathology, relationship to standing joint inflammation and damage psoriatic arthritis: Twenty-four-week efficacy tenderness and correlation with clinical indi- in psoriatic arthritis and beyond. Arthritis and safety results of a randomized, placebo- ces. Rheumatology (Oxford) 2008; 47: 92-5. Rheum 2007; 56: 2482-91. controlled study. Arthritis Rheum 2009; 60: 25. MUMTAZ A, GALLAGHER P, KIRBY B et al.: 5. GLADMAN DD, CHANDRAN V: Observational 976-86. Development of a preliminary composite cohort studies: lessons learnt from the Univer- 15. KAVANAUGH A, MEASE P: Treatment of psori- disease activity index in psoriatic arthritis. sity of Toronto Psoriatic Arthritis Program. atic arthritis with inhibi- Ann Rheum Dis 2011; 70: 272-7. Rheumatology (Oxford) 2011; 50: 25-31. tors: longer-term outcomes including enthesi- 26. EBERL G, STUDNICKA-BENKE A, HITZEL- 6. MARZO-ORTEGA H, McGONAGLE D, tis and dactylitis with golimumab treatment HAMMER H, GSCHNAIT F, SMOLEN JS: O’CONNOR P, EMERY P.: Efficacy of etaner- in the Longterm Extension of a Randomized, Development of a disease activity index for cept in the treatment of the entheseal pathol- Placebo-controlled Study (GO-REVEAL). J the assessment of (DAREA). ogy in resistant spondylarthropathy: a clinical Rheumatol 2012; 89 (Suppl.): 90-3. Rheumatology (Oxford) 2000; 39: 148-55. and magnetic resonance imaging study. [see 16. MAKSYMOWYCH WP, MALLON C, MORROW 27. FitzGERALD O, HELLIWELL P, MEASE P et comment]. Arthritis Rheum 2001; 44: 2112-7. S et al.: Development and validation of the al.: Application of composite disease activ- 7. D’AGOSTINO MA, SAID-NAHAL R, HAC- Spondyloarthritis Research Consortium of ity scores in psoriatic arthritis to the PRESTA QUARD-BOUDER C, BRASSEUR JL, DOUGA- Canada (SPARCC) Enthesitis Index. Ann data set. Ann Rheum Dis 2012; 71: 358-62. DOS M, BREBAN M: Assessment of periph- Rheum Dis 2009; 68: 948-53. 28. HELLIWELL PS, FitzGERALD O, MEASE PJ: eral enthesitis in the spondylarthropathies by 17. HEALY PJ, HELLIWELL PS: Measuring clini- Development of composite measures for ultrasonography combined with power Dop- cal enthesitis in psoriatic arthritis: assess- psoriatic arthritis: a report from the GRAPPA pler: a cross-sectional study. Arthritis Rheum ment of existing measures and development 2010 annual meeting. J Rheumatol 2012; 39: 2003; 48: 523-33. of an instrument specific to psoriatic arthritis. 398-403. 8. McQUEEN FM, GAO A, OSTERGAARD M et Arthritis Rheum 2008; 59: 686-91. 29. HELLIWELL PS, KAVANAUGH A: Compari- al.: High-grade MRI bone oedema is common 18. MEASE PJ, FLEISCHMANN R, DEODHAR AA son of composite measures of disease activ- within the surgical field in rheumatoid arthritis et al.: Effect of certolizumab pegol on signs ity in psoriatic arthritis using data from an in- patients undergoing joint replacement and is and symptoms in patients with psoriatic ar- terventional study with golimumab. Arthritis associated with osteitis in subchondral bone. thritis: 24-week results of a Phase 3 double- Care Res (Hoboken) 2014; 66: 749-56. Ann Rheum Dis 2007; 66: 1581-7. blind randomised placebo-controlled study 30. COATES LC, HELLIWELL PS: Disease meas- 9. EDER L, JAYAKAR J, THAVANESWARAN A (RAPID-PsA). Ann Rheum Dis 2014; 73: 48- urement--enthesitis, skin, nails, spine and et al.: Is the MAdrid Sonographic Enthesi- 55. dactylitis. Best Pract Res Clin Rheumatol tis Index useful for differentiating psoriatic 19. OLIVIERI I, BAROZZI L, FAVARO L et al.: 2010; 24: 659-70.

S-117