Background Little is known about the influence of specific comorbidities on response to biologic therapy.

Objectives To examine the impact of comorbidities on anti-TNF therapy efficacy in rheumatoid arthritis (RA) patients (pts) within the US CORRONA RA registry.

Methods A retrospective cohort of adult RA pts with moderate-to-severe RA (CDAI>10) who initiated anti-TNF biologics after June 1, 2009 and had a 6 month follow up (in a window of 3–9 months) were included. Comorbidity assessments included a modified Charlson comorbidity index (CCI) [HIV status was not assessed]) and diagnoses of diabetes, cardiovascular disease (CV) and history of depression. Propensity score (PS) method was employed to match pts with and without select comorbidities. Variables used for PS matching were selected if standardized differences were ≥0.1 between groups. PS were stratified by line of therapy. The primary outcome was achievement of low disease activity (LDA) at 6 months (CDAI ≤10). Secondary outcomes included DAS28(ESR) LDA (DAS<3.2), modified ACR 20/50/70. Post hoc analysis was performed on the CDAI components to estimate pt and MD reported outcomes.

Results A total of 4064 anti-TNF initiations were included. For CCI, diabetes and CV, there were no significant differences in response in the matched populations. For patients with a history of reported depression (hxdep), LDA rates were significantly lower (29% for hxdep vs 39% for no hxdep, p<0.001 OR=0.65 [0.56, 0.75]). In the matched population (n=1810) LDA rates remained significantly different: 31% vs 36% (hxdep vs no hxdep), p=0.027, OR=0.80 [0.66, 0.98]. Matched patients differed in mean (SD) patient fatigue (VAS 0–100): hxdep: 61.4 (26.2) and no hxdep: 54.7 (27.9) p<0.001. In pts matched on factors including patient fatigue (n=1252) there was no significant difference in response rates: 31.8% vs 31.8% OR=1.0 [0.79, 1.27]. Similar results were found in the secondary outcomes. Estimates of the CDAI components [Table 1] showed the Pt Global and Tender Jts were the largest contributors to CDAI change.

Conclusions Of the comorbidities examined, a history of reported depression in RA patients is an indicator of lower response rates to TNFi. CDAI was primarily influenced by patient responses. Patient fatigue may be a manifestation of depression and explained response rate differences. Further analyses are needed to explore this relationship.

Acknowledgement This study is sponsored by Corrona, LLC. The Corrona RA registry has been supported through contracted subscriptions in the last two years by AbbVie, Amgen, BMS, Crescendo, Genentech, Horizon Pharma USA, Janssen, Eli Lilly, Novartis, Pfizer, and UCB.

Disclosure of Interest D. Parenti Employee of: Janssen Scientific Affairs, LLC, G. Reed Shareholder of: Corrona, LLC, Employee of: Corrona, LLC, S. Kafka Employee of: Janssen Scientific Affairs, LLC, L. Ellis Employee of: Janssen Scientific Affairs, LLC, J. Greenberg Shareholder of: Corrona, LLC, Consultant for: Janssen Scientific Affairs, LLC, AstraZeneca, Celgene, Genentech, Novartis, Pfizer, Employee of: Corrona, LLC, R. DeHoratius Employee of: Janssen Scientific Affairs, LLC