Background/Purpose: Growing evidence supports the efficacy of triple therapy (methotrexate + sulphasalazine + hydroxychloroquine) for controlling disease activity in patients with early rheumatoid arthritis (ERA), but it requires multiple pills and appears to be less effective than methotrexate + anti-TNF therapy for preventing joint damage. The GRADE approach for developing treatment recommendations has been adopted by the ACR and requires that treatment recommendations integrate the estimated treatment effects for key outcomes and patients’ preferences, but methods for doing this have not been defined. The objective of this study was to apply a novel approach for jointly considering the comparative benefits and harms of treatment options and patients’ preferences to estimate the preferred treatment for patients with ERA.

Methods: We determined the preferred treatment for patients with moderate-severe ERA (<2 years since diagnosis) separately for methotrexate-naïve patients (choice of triple therapy versus methotrexate) and for patients who have had an inadequate response to methotrexate (choice of triple therapy versus methotrexate + anti-TNF therapy).  We used patients’ preferences that we elicited through a discrete choice experiment to apply weights to 3 key outcomes from a network meta-analysis (ACR50 response, radiographic progression, withdrawals due to adverse events) and other important considerations (dosing, rare adverse events). The preferred treatment was the one that had the highest total value, which was calculated as a sum of the preference-weighted treatment characteristics. We incorporated uncertainty in both the treatment effects and individual patients’ preferences by averaging the results across 10,000 samples from Bayesian models for both the network meta-analysis outcomes and the patients’ preferences.

Results: Triple therapy was the most preferred treatment in methotrexate-naive patients (78% preferred triple therapy to methotrexate) or after an inadequate response to methotrexate (58% preferred triple therapy to methotrexate + anti-TNF therapy). In methotrexate-naïve patients, triple therapy provided a higher chance of benefit (ACR50 response) than methotrexate alone based on the network meta-analysis, which was most important to patients and outweighed the additional pills or other inconveniences. After failing methotrexate, triple therapy provided a similar chance of ACR50 response to anti-TNF therapy based on the network meta-analysis, but did not require injections or have an increased risk of infection/ possible increased risk of skin cancers as with anti-TNF therapy. For most patients, these issues outweighed any benefit of anti-TNF therapy on preventing joint damage.

Conclusion: Based on estimates of the comparative benefits and harms of methotrexate-based treatment options for ERA and patients’ preferences, many patients would prefer triple therapy to methotrexate as initial therapy and triple therapy to methotrexate + anti-TNF therapy after an inadequate response to methotrexate. Explicitly considering patients’ preferences may affect treatment recommendations made using GRADE.

« Back to 2015 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/many-patients-with-early-rheumatoid-arthritis-want-triple-therapy-an-analysis-combining-comparative-effectiveness-research-and-patients-preferences-to-inform-treatment-recommendations/