Shiori Nishimura, Toshitaka Hirano, Soko Setoguchi, Hiroki Matsui, Yoshinori Takeuchi, Yasushi Inoue, Yoshiya Tanaka, Kunihiro Yamaoka, Yoichi Ii, Shintaro Hiro, Masato Hoshi, Naonobu Sugiyama, Yutaka Matsuyama, Hiraku Kumamaru
In the nationwide Japanese cohort, tsDMARDs did not show increased malignancy risk compared to bDMARDs, and PCP risk had no clear difference at 2nd and 3rd LOTs, whereas the risk of HZ was higher with tsDMARDs.
AIM: To compare risks of malignancy, opportunistic infections, and venous thromboembolism between the targeted synthetic disease-modifying antirheumatic drugs (tsDMARDs) and biologic DMARDs (bDMARDs) at the 2nd and 3rd lines of therapy (LOT).
METHODS: We conducted a cohort study using the National Database of Health Insurance Claims (2013-2021). Patients with rheumatoid arthritis (RA) initiating tsDMARDs or bDMARDs after conventional synthetic DMARDs (csDMARDs) were identified at the 2nd LOT or 3rd LOT. Hazard ratios (HRs) with 95% confidence intervals (CIs) were estimated using Cox models weighted by inverse probability of treatment weighting.
RESULTS: Among 498 813 csDMARD initiators, 66 638 and 4177 initiated bDMARDs and tsDMARDs at 2nd LOT; 14 085 and 1757 initiated bDMARDs and tsDMARDs at 3rd LOT. For malignancy, HRs (95% CIs) for tsDMARDs versus bDMARDs were 0.92 (0.60-1.43) at 2nd LOT and 0.94 (0.53-1.65) at 3rd LOT. HRs (95% CIs) for pneumocystis pneumonia (PCP) and deep vein thrombosis (DVT) were as follows: PCP, 0.98 (0.75-1.29) at 2nd LOT and 1.12 (0.81-1.54) at 3rd LOT; DVT, 1.13 (0.64-1.99) at 2nd LOT and 0.96 (0.52-1.78) at 3rd LOT. In contrast, herpes zoster (HZ) risk was consistently higher with tsDMARDs (2nd LOT: HR, 2.61 [95% CI, 2.24-3.05]; 3rd LOT: HR, 2.90 [95% CI, 2.38-3.53]).
CONCLUSIONS: In the nationwide Japanese cohort, tsDMARDs did not show increased malignancy risk compared to bDMARDs, and PCP risk had no clear difference at 2nd and 3rd LOTs, whereas the risk of HZ was higher with tsDMARDs.