Absolute serious infection rates were low. However, across the JAKinibs, the incidence of HZ is higher than expected for the population. While the risk was numerically greatest with BARI, indirect comparisons between the drugs did not demonstrate any significant difference in risk.
How JAKinibs increase the risk of HZ reactivation is unclear, but how different JAKs interact in the immune response suggest that there may be differences in safety profiles between JAKinib drugs, underpinned by their differential selectivity profiles. This systematic literature review and meta-analysis aimed to evaluate the risk of serious infection and HZ in RA patients receiving JAKinib. Since no head-to-head studies have been undertaken with the three JAKinibs, each agent was compared directly with placebo and a network meta-analysis was employed to allow indirect comparisons. Patient-exposure was calculated using data from 21 Phase II and III RCTs. In total there were 11 TOF studies, 6 BARI, and 4 UPA – representing 5,888, 3,520, and 1,736 patients, respectively.
The primary outcome of interest was serious infection. The estimated incidence risk ratios (IRRs) compared with placebo in per protocol analyses were not statistically significant. Secondary outcomes of interest included the number of opportunistic infections, including rates of HZ. For HZ infection, the estimated IRR compared with placebo was 1.38 for TOF, 2.86 for BARI, and 0.78 for UPA. Across JAKinibs, the rate of VZV reactivation was 3.23 per 100 patient-years – higher than that the 1.6 seen with TNFi.
Overall, the absolute serious infection rates were low. Although the IRRs were numerically different between TOF, BARI and UPA, it would be inappropriate to use this as evidence of a differential risk between agents due to the rarity of serious infections. Across JAKinibs, HZ incidence is higher than expected. While the risk was numerically greatest with BARI, indirect comparisons between the drugs did not demonstrate any significant difference in risk. The results of this network meta-analysis cast doubt over whether the small differences between JAKinibs are clinically meaningful.