In this study, we show a significant response to anti-TNF therapy in routine clinical practise with DAS28 remission in 58% PsA patients compared to 44% of RA patients. There were significant differences in single variables and the DAS28 scores between PsA and RA patients at baseline. This may reflect differences in pattern of joint involvement and/or lower CRP levels frequently noted in PsA compared with RA patients [17
]. To reduce the possible bias due to differences at baseline between PsA and RA, we analysed a subgroup matched for baseline DAS28, this still showed a significantly greater response in PsA patients compared to RA. While response rates in PsA have been looked at before, these studies looked into EULAR response rates [20
] and did not comment on DAS28 response, or remission, as we do here.
Individual variables including tender and swollen joint counts, CRP, patient global VAS and the HAQ showed significant improvement in both patient groups, most parameters showed the greatest response within the first 3 months, however significant improvements were seen between 3 and 12 months, residual tender and swollen joints were more common in RA patients. The mean CRP fell in both PsA and RA patients, it fell more sharply in the PsA group overall, however, in the matched PsA and RA subgroup analysis CRP changes were comparable. The patient reported outcomes such as patient global VAS, pain VAS and HAQ also showed significant improvement in both patient groups. PsA patients had less tender and swollen joints and stiffness at 12 months, but also a significantly lower DAS 28 and HAQ.
Fifty-eight percent of PsA patients were in DAS28 remission at 12 months compared to 44% of RA patients. When the two groups were analysed as regards EULAR good response, that is, a DAS28 of <3.2 at endpoint and and improvement of more than or equal to 1.2, then 73% of PsA, but only 51% of RA patients, met this criteria. It was interesting to note that male PsA patients attained significantly lower DAS28 scores than females 1.88 vs 2.65 at one year (P < 0.05). Since the initial DAS28 in PsA was lower than RA patients when biologic treatment started we compared remission rates in PsA vs RA, by analyzing a subset of patients matched for disease activity at baseline (n = 41, in each group). Remission rates defined by DAS28 were significantly greater in PsA patients (P = 0.01), 63.5% PsA patients vs 41.4% RA patients at 12 months even in this particular group.
This is the first study to examine remission rates in a cohort of PsA patients from routine clinical practise studied prospectively in a dedicated biologic clinic. The finding of DAS28 remission in almost two-thirds of patients is significantly higher than the observed level in RA patients, even when matched for baseline disease activity. Remission has become an attainable goal in the treatment of RA, the recent Combination of Methotrexate and etanercept in Active Early Rheumatoid (COMET) study suggesting high remission rates in RA patients with high levels of disease activity commenced on treatment early [21
]. The aim of remission in the treatment of RA has been adopted as desirable and feasible by EULAR in the most recent recommendations [22
]. In a recent editorial, the potential of remission in PsA was considered to be a realistic goal [15
]. There are several possible criteria which may be used in the definition of remission, including single clinical variables, response criteria, pooled indices such as the DAS and ACR and then imaging technology criteria. In an important study of the performance of response criteria for assessment of peripheral arthritis in PsA, Fransen et al, compared a detailed analysis of individual items and pooled indices in dicriminating change in two clinical trial data sets [19
]. The authors concluded that response criteria and pooled indices, specifically the EULAR response criteria, performed better than the ACR or the PsARC in discriminating active from placebo drugs. In addition, they found that the DAS and DAS28 performed better than single core-set measures in PsA. Furthermore, two studies have reported that the DAS28 is a valid instrument for measuring disease activity with respect to response to biologic therapies [19
]. In measuring remission in this prospective cohort study we have applied the DAS28, as a validated measure of disease activity in PsA patients treated with biologics, and in the knowledge that such a pooled index developed for RA has been shown to be useful for assessment of the peripheral arthritis of PsA. The results of this study suggest therefore that, using a single measure of disease activity with an agreed level defined as remission, biologic therapies result in high remission rates in PsA patients, greater than a comparator group of RA patients even when matched for baseline disease activity.
In this study we found a number of individual baseline parameters were associated with remission examining independent correlations. In particular, in PsA patients, male gender and the patient-derived indices including HAQ, patient global VAS and early morning stiffness appeared to be associated with remission. Linear regression analysis however failed to confirm all these variables as predictors of remission and suggested the association was strongest between baseline HAQ and remission. A previous report of PsA patients treated with anti-TNF agents had identified improvement in the DAS28 score as the best predictor of improvement in quality of life (QoL) measured using the SF36 [24
]. Interestingly, the authors of this paper also found significantly higher QoL responses in their PsA cohort compared to the RA cohort studied, a result which confirmed an earlier report from the Norwegian registry [25
]. In a recent international multicenter study of RA patients [26
], measures of HAQ also behaved differently in men and women. We therefore reanalysed the data in our RA cohort and found there are differences in men and women in relation to HAQ response; however, the size of the response in male PsA and RA patients is similar. The association of male gender with remission was unexpected and raises an intriguing question: Why do male PsA patients show a better response to therapy? One possible explanation is related to testosterone levels, which have previsouly been reported to be higher in HLAB27 positive subjects [27
], and may have a protective effect in seronegative spondyloarthopathies [28
]. Indeed, it is intriguing to hypothesize that testosterone levels may augment the response of PsA patients to biologic therapies.