Statistics from Altmetric.com
We thank the authors for the interest in our paper and are grateful for the opportunity to respond to the points raised.1 We agree that there is a clear distinction between composite measures of psoriatic arthritis such as disease activity in PsA (DAPSA) and composite measures of psoriatic disease such as minimal or very low disease activity (MDA/VLDA) and the PsA disease activity score (PASDAS). As the Vienna group rightly point out, these measures differ in terms of the components included, but not due to disagreement within the outcome measure community as suggested in the letter. The choice of components for each composite measure was decided using different methodology in the development of each one, thus resulting in different measures. We believe that this variation in scores is one reason for the need to compare such scores in different populations to establish the optimal measure or measures for psoriatic arthritis (PsA). Indeed when the DAPSA was originally suggested, it was because the same components used in the disease activity in reactive arthritis (DAREA) were identified in a principal component analysis (PCA) in PsA. Interestingly in this analysis, the variables tested were taken from the Outcome mesaures in Rheumatology (OMERACT) PsA domains and therefore DAPSA was not, a priori, designed specifically to be a unidimensional composite measure. The fourth component identified in the PCA was the psoriasis area and severity index (PASI) highlighting the importance of skin in PsA despite the fact that patients had a low baseline mean PASI of only 3.3. While PASI was not included in the DAPSA as the eigenvalue was 0.949 and therefore just under the threshold of 1,2 it is interesting to imagine how the results may have differed if it were developed in a group with slightly more active skin disease.
We agree with and highlighted the issues of residual disease activity in PASDAS remission within the paper. This perhaps relates to the number of measures included allowing one component to be high and the overall score to be in the remission range. It seems unlikely that PASDAS in its current form would be adopted into routine clinical practice but it has shown excellent discrimination in clinical trials with consistently superior effect sizes highlighting its potential use in this setting. As requested, we have performed further analysis on the PASDAS low disease activity (LDA) (≤3.2) state in our dataset. A total of 151/227 or 66.5% of patients met the criteria (23 had missing data prohibiting calculation). This is a similar proportion to MDA where the agreement between definitions was highest (kappa 0.638). However, agreement was reasonable with both c/DAPSA LDA (0.517 and 0.504, respectively). For example, 143 patients were in DAPSA and PASDAS LDA, 7 were in PASDAS LDA but not DAPSA and 38 were in DAPSA LDA but not PASDAS showing overlapping but slightly different definitions. While PASDAS includes different domains of PsA, it is relatively specific to musculoskeletal disease and was designed as a unidimensional measure based on treatment changes in PsA (using similar methodology to the disease activity score (DAS) in rheumatoid arthritis). It does not include a measure of psoriasis. The levels of residual disease activity were similar or slightly higher than the other definitions (we attach an updated table 1 from the paper here), particularly for skin psoriasis where 46% of patients had a PASI>1. This further highlights the need to assess psoriasis either within (MDA) or additional to (PASDAS/DAPSA) any target chosen.
We also agree with the point that stringency does not just relate to proportions achieving the outcomes but to the external validity of these outcomes. MDA/VLDA, PASDAS and DAPSA have all been shown to be associated with radiographic damage in previous studies.3–6 Unfortunately it was not possible to confirm in this cross-sectional study without radiographs. We did, however, use quality of life as an external validity marker which again highlighted the need to assess skin disease with poorer quality of life seen in patients in DAPSA remission with a PASI>1.7
As regards the PRESTA trial, we believe that this study highlights the need to choose an appropriate outcome measure to answer the question set by each individual study. The PRESTA trial was unusual as it recruited patients with severe psoriasis and PsA. In such a population, a composite measure such as the composite psoriatic disease activity index (CPDAI) may be appropriate to establish the overall improvement in psoriatic disease. While there was no difference in musculoskeletal outcomes, which would have been shown by secondary unidimensional outcomes, the CPDAI results would have highlighted the importance of considering a dose or treatment that is optimal in all domains. Thus, it may have allowed regulators to support the use of a higher dose in patients that were similar to the target population of the trial, that is, patients with severe skin and joint involvement. This more personalised and variable approach to treatment is now seen in the use of secukinumab where patients with PsA are routinely treated with 150 mg but for those with active skin disease, the 300 mg dose is recommended by both Food and Drug Administration (FDA) and European Medicines Agency (EMA).
We also agree that for rheumatologists the principal focus of treatment is the peripheral arthritis; however, we remain concerned that this is to the detriment of optimising therapy for patients. Many studies, in addition to ours, have highlighted the importance of active skin disease and its association with poorer quality of life for the patient. While we do not expect rheumatologists to specialise in the management of psoriasis, an awareness of skin disease and basic assessment using tools such as the body surface as included in MDA would support appropriate comanagement of patients with PsA to optimise their quality of life.
We also recognise the pressure on clinicians’ time but do not see how this differs between the measures proposed in the new T2T recommendations. Whether the VLDA/MDA criteria are assessed, or the DAPSA plus a separate assessment of skin and enthesitis are performed, the time taken is the same. MDA is a modular measure where the items are not ‘lumped’ into one numerical score but assessed individually with a cut point to achieve for each thus allowing appropriate treatment choices. With DAPSA plus individual measures, the DAPSA could be calculated and then the skin/enthesitis would have to be assessed separately although clear guidelines on measures or targets for this are not defined. Thus, the time for clinician assessment is identical and both scores are quick and feasible to calculate/assess. To further support busy clinicians, the Group for Research and Assessment of Psoriasis and Psoriatic Arthritis (GRAPPA) have recently developed a free ‘app’ for both Apple and Android that includes an MDA calculator allowing entry of patient reported outcomes and physician assessments to produce an easy to interpret summary of disease activity by domain (figure 1).
Providing additional domains are assessed alongside the DAPSA, opinion of disease activity should be similar and appropriate treatment is likely to be the same. The modular approach of VLDA/MDA allows clear visualisation of which domains are active to guide therapy. If patients do not achieve VLDA due to ongoing psoriasis activity, this does not mandate immediate escalation of biological therapy to risk patient safety or societal costs. This could simply suggest the need for concomitant topical therapy as highlighted by the authors with the DAPSA example given.
While there have been strong views on the positives and negatives of each target suggested in the recommendations, it is important to remember that our shared goal is to ensure that patients are assessed more appropriately, both in clinical trials and in routine practice allowing optimisation of treatment using a treat to target approach. We believe that encouraging physicians to briefly assess skin disease and other MSK domains in addition to the focus on peripheral arthritis is key to therapeutic optimisation in PsA given our goal to ‘to optimise long-term health-related quality of life and social participation through control of signs and symptoms’.8
Handling editor Gerd R Burmester
Contributors This reply was written by LCC and LJJvM and reviewed by the other authors prior to submission.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient consent Not required.
Provenance and peer review Commissioned; internally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.