Abstract
The Group for Research and Assessment of Psoriasis and Psoriatic Arthritis (GRAPPA)–Outcome Measures in Rheumatology (OMERACT) Psoriatic Arthritis (PsA) Core Set working group is focused on the development of a core set of instruments used to assess the domains described in the 2016 PsA Core Domain Set. At the 2021 annual meeting, the group presented an update on the domain of structural damage. In this report, we discuss the steps taken to assess the domain match and feasibility of plain radiographic instruments in the assessment of structural damage in PsA.
- GRAPPA
- outcome assessment
- psoriatic arthritis
Introduction
Structural damage can be defined as abnormalities in the structure or integrity of a joint, bone, or tendon likely to be attributable to psoriatic arthritis (PsA). Structural damage is in the middle circle of the 2016 PsA Core Domain Set and should be measured at least once in the development of a new therapeutic.1,2 It is consistently prioritized by individuals with PsA, in part due to its negative effect on physical function.1,3,4
Despite advances in diagnosis and therapeutics, radiographic structural damage remains common.5,6,7,8 It is associated with disease activity and mortality, and is known to precede clinical damage.9,10,11 There is also a discordance between disease activity and radiographic progression in the treatment arms of biologic randomized controlled trials (RCTs), suggesting structural damage is not a redundant outcome.12,13,14
In this report, we summarize progress made in advancing a radiographic instrument for the assessment of structural damage through the Outcome Measures in Rheumatology (OMERACT) Framework Instrument Selection Algorithm and report the working group’s (WG) vote on the domain match and feasibility of plain radiographic instruments.
Domain Definition
The domain definition describes (1) the target population, (2) intended use for the domain (eg, clinical trials), (3) target domains (eg, joint space narrowing), (4) qualitative or literature support, and (5) sources of variability. It is modality specific, as different modalities have different sensitivities, specificities, and reliabilities for different target domains. Further, operational definitions include image acquisition variables and joint positioning, which are also modality specific.
The domain definition was drafted (AA, WT), discussed in 2 virtual WG meetings, and revised. The WG were cognizant of the importance of content-expert radiologist opinions in drafting these documents, and thus the document was subsequently reviewed by 5 expert radiologists who provided feedback by a survey (Supplementary Material 1, available with the online version of this article). A revised domain definition was next circulated to members of the WG for feedback, then finalized (Supplementary Material 2).
The target population for this domain definition was adults aged ≥ 18 years with peripheral PsA. Its intended use was for RCTs comparing a disease-modifying antirheumatic drug to a placebo or active comparator. Specifying the population and intended use ensures that when an instrument is eventually endorsed, it is used in the right population and setting.
Potential target domains were elicited from a review of the literature.15,16 The target domains selected were joint space narrowing and joint erosion due to the frequency of these features, associations with disease activity and functional outcomes, and progression over time in RCTs (Supplementary Material 3, available with the online version of this article). Other features are uncommon and/or progress slowly when assessed by plain radiography (Supplementary Material 4). Although it is discriminative enough to warrant inclusion in the classification criteria for PsA, new bone formation including shaft or tuft periostitis and juxtaarticular osteoproliferation has not shown to progress significantly in RCTs utilizing tumor necrosis factor inhibitors up to a follow-up of 2 years.17,18 Therefore, while it may be desirable to compare new bone formation in head-to-head trials assessing different therapeutic pathways, plain radiography may not be the appropriate modality.
The domain definition focuses on the assessment of structural damage in the hands, wrists, and feet. There is a paucity of data to support the use of plain radiography to assess structural damage in other joints in peripheral PsA. The assessment of large joints is a gap given the frequency of this phenotype; however, plain radiography may not be adequately sensitive to assess progression of structural damage in large joints in the timeframe of an RCT, and the relevant target domains may differ.
Domain Match
Candidate instruments identified from a recent literature review were assessed against the domain definition for content validity.16 Seven instruments were included: modified Larsen, modified Steinbrocker, Ratingen score, modified total Sharp score version B (mTSS-B), modified Sharp/van der Heijde score (mSvdHs), Reductive X-Ray Score for Psoriatic Arthritis (ReXSPA), and Simplified Psoriatic Arthritis Score (SPARS).16
The domain match questionnaire was adapted from the OMERACT handbook and administered to participating WG members by online platform (n = 13 participants, inclusive of 1 patient research partner [PRP])19 with reference material including radiologist feedback and a summary of the instruments (Supplementary Material 1 and Supplementary Material 5, available with the online version of this article). The results of the initial vote and available distribution data were discussed at an online WG meeting to optimize consensus. A second round of voting was undertaken (n = 14) where respondents were asked if the instrument was “Good to Go” (GREEN), “Some Cautions, but Okay” (AMBER) or “Not Right for This Application” (RED). A 50% majority response was accepted as supportive (agreement), whereas a 70% majority response was accepted as consensus. A > 15% RED response was considered a hurdle for domain match and feasibility (Supplementary Material 6).
In the final vote (Figure 1A), there was agreement from the WG that the mTSS-B and the mSvdHs had content validity (GREEN), with both target domains being assessed in the joints of the hands, wrists, and feet. This vote may reflect the recognition of instruments utilized in RCTs. There was a lack of agreement on the Ratingen score, which scores for osteoproliferation and the composite domain of destruction but not joint space narrowing. The final rating for the Ratingen was RED, given that a significant proportion of respondents (n = 6/14 [43%]) rated the instrument RED. There was consensus from the WG that the modified Larsen, modified Steinbrocker, SPARS, and ReXSPA were not appropriate for use in this setting (RED). The key WG discussion points are summarized in Table 1.
Feasibility
Feasibility was assessed among PRPs and the WG (Table 1). A live webinar followed by a Q&A session was held with PRPs. PRPs were invited to attend and/or watch a recording of the webinar at their convenience and complete a feasibility survey adapted from the OMERACT handbook with PRP input. The webinar discussed the types, importance, and effect of structural damage; the ways in which it can be measured; factors influencing the modality chosen in routine care and clinical trials; and the potential risks and benefits of plain radiography.
Of the 9 PRPs who participated (Supplementary Material 7, available with the online version of this article), 8 had had plain radiography of their hands and feet. Seven PRPs felt it was easy to undergo and that the time required was reasonable, although it was noted that it could be inconvenient to wait for the imaging. Three PRPs (33%) were uncertain about the potential for harm of plain radiography, opining that the risks of radiation should be explained and that those risks may not be prohibitive. Three PRPs (33%) opined that there was a potential for discomfort in individuals with underlying joint pain or deformities. Two-thirds of PRPs felt that the costs were acceptable. Eight PRPs (89%) felt that there may be benefits to having plain radiography, but some commented that this might be the case in RCTs rather than at an individual level.
In the WG survey, 11 of 13 respondents felt it was easy to access plain radiography. After a first round of voting (n = 13), the results were discussed in an online meeting to optimize consensus. Many members who had utilized the instruments assessed felt that they were feasible; however, some expressed concern regarding the lack of a training platform or image atlas for the instruments. There was concern that this might affect the long-term fidelity of all instruments and negatively affect equitable access to radiographic scoring in centers where training is unavailable.
Following a second round of voting (n = 14, inclusive of 1 PRP), there was consensus (Figure 1B) that most instruments were AMBER based on the results of the initial survey balanced against the absence of training material (Supplementary Material 8, available with the online version of this article). The exceptions were the modified Steinbrocker and modified Larsen score, where 21% and 29% of respondents, respectively, felt that the instruments were not feasible, whereas > 70% of respondents felt both instruments could proceed with caution. It is important to note that the modified Steinbrocker has been utilized for decades to assess long-term structural damage in observational cohorts.
Conclusion
In the assessment of structural damage in peripheral PsA in RCTs evaluating a therapy against a comparator or a placebo, plain radiography is feasible and acceptable to individuals with PsA. The mSvdHs and mTSS-B have content validity (GREEN) and all instruments are potentially feasible (AMBER). Gaps highlighted are the need for instruments for large joints and the need for a training platform to optimize instrument feasibility. The next step will be to evaluate the measurement properties of the mSvdHs and mTSS-B in RCTs.
ACKNOWLEDGMENT
The authors thank Drs. David Salonen, Iris Eshed, Charles Peterfy, Robert Lambert, Jyoti Panwar and for their valuable contributions as content-expert radiologists, and participating patient research partners for participating in the feasibility exercise. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health.
Footnotes
As part of the supplement series GRAPPA 2021, this report was reviewed internally and approved by the Guest Editors for integrity, accuracy, and consistency with scientific and ethical standards.
AA is supported by an Arthritis Australia-Australian Rheumatology Association fellowship. AJM is funded by the PARTNER psoriatic arthritis fellowship. LCC is funded by a National Institute for Health Research Clinician Scientist award. The work was supported by the National Institute for Health Research Oxford Biomedical Research Centre. RC acknowledges that Section for Biostatistics and Evidence-Based Research, the Parker Institute, Bispebjerg and Frederiksberg Hospital is supported by a core grant from the Oak Foundation (OCAY-18-774-OFIL). YYL is supported by the National Medical Research Council, Singapore (NMRC/CSA-Inv/0022/2017). The funding sources had no role in the views expressed in this work.
The authors declare no conflicts of interest relevant to this article. This paper does not require institutional review board approval.
- Accepted for publication December 7, 2021.
- Copyright © 2022 by The Journal of Rheumatology
References
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.
- 8.
- 9.
- 10.
- 11.
- 12.
- 13.
- 14.
- 15.
- 16.
- 17.
- 18.
- 19.