Background Interstitial lung disease (ILD) is associated with a number of rheumatoid arthritis (RA) therapies but the impact of exposure to T cell (abatacept [ABA]), B-cell (rituximab [RTX]) and IL-6 (tocilizumab [TCZ]) inhibitors is unclear.
Objectives To compare biologics used for RA with respect to ILD incidence and risk factors.
Methods RA patients (pts) (≥1 ICD9 diagnosis of RA: 714.0 or 714.3) aged ≥18 were selected from the MarketScan® databases (January 2010-June 2012) if they had ≥ one new prescription/administration of ABA, RTX), TCZ or an anti-TNF after having discontinued a different biologic agent, met continuous enrollment criteria and did not have baseline evidence of IBD, psoriatic arthritis, axial spondylarthritis, psoriasis or cancer or ILD. The index event was defined as the first exposure to the study eligible biologic. Study outcome was the incidence of ILD. Cox proportional hazards modeling estimated the risk of ILD while adjusting for age, gender, biologic cohort, comorbidities, methotrexate and steroid exposure. A total of 4 claims-based ILD case definitions were used with varying sensitivity and specificity.
Results We identified 13,296 episodes of exposure to eligible biologics (mean age [SD] 52.6 [12.5], 81.9% female) in 10,800 pts. Mean (SD) follow-up was 0.7 (0.5) years. Mean (SD) number of unique biologic agents received prior to index ranged from 1.4 (0.7) in the TNF cohort to 2.1 (1.1) in the TCZ cohort. TCZ users were significantly more likely (p<0.0001) to have exposure to higher dose glucocorticoids (≥7.5 mg/day) and a prior history of pulmonary disorders other than ILD (p<0.01).The ILD incidence rate (IR) per 1000 PY (unadjusted) ranged from 1.8 (95% CI 1.0 – 2.9) to 6.4 (95% CI 4.9 – 8.3) depending on ILD definition. Using the most sensitive definition, the unadjusted IR ranged from a low 4.0 (95% CI 1.6 – 8.2) in the ABA cohort to a high of 12.2 (95% CI 5.6 – 23.2) in the infliximab cohort. The aggregate IR for the TNF group was 7.1 (95% CI 5.0 – 9.6). Cox model results using the sensitive definition show significantly increased ILD incidence (HR 3.5; 95% CI 2.1-6.0) in pts ≥65 years of age (vs. <65). Exposure to glucocorticoids in the 6 months prior to index was marginally associated with increased incidence (HR 2.0; 95% CI 0.98-4.1; p=0.0586) in this model as well. Cox model results using the most specific definition suggests that being male (HR 3.1; 95% CI 1.2-8.4) and having a history of another pulmonary disorder (HR 4.8; 95% CI 1.7-13.7) were associated with significantly higher ILD incidence. Adjusted HRs in both models found no significant increase in the risk of ILD in pts exposed to ABA, RTX and TCZ as compared with anti-TNF therapies.
Conclusions Older pts, men, and pts with a history of other pulmonary disorders or recent exposure to glucocorticoids may be at increased risk for developing ILD. There was no significant increase in the risk of ILD in pts exposed to RA biologics with newer mechanisms of actions as compared with anti-TNF therapies. Limitations of this study include short duration of follow up, lack of information in claims about disease severity, channeling of pts with pulmonary conditions to TCZ, and confounding by indication.
Disclosure of Interest J. Curtis Grant/research support: NIH, AHRQ, UCB, Centocor, CORRONA, Amgen, Pfizer, BMS, Crescendo, Abbott, Genentech, Inc, Consultant for: UCB, Centocor, CORRONA, Amgen, Pfizer, BMS, Crescendo, Abbott, Genentech, Inc, K. Sarsour Employee of: Genentech, Inc, P. Napalkov Employee of: Genentech, Inc, L. Costa Employee of: Outcomes Research Solutions, Inc, K. Schulman Employee of: Outcomes Research Solutions, Inc