Age-based (<65 vs ≥65 years) incidence of infections and serious infections with tofacitinib versus biological DMARDs in rheumatoid arthritis clinical trials and the US Corrona RA registry

A randomised, open-label, blinded endpoint post-authorisation safety study (Study A3921133; NCT02092467; database not locked and subject to change) evaluated the safety of tofacitinib 5 mg and 10 mg twice daily (BID) versus tumour necrosis factor inhibitors (TNFi) (adalimumab/etanercept) in rheumatoid arthritis (RA) patients aged ≥50 years with ≥1 cardiovascular risk factor. An ad hoc interim safety analysis of Study A3921133 reported incidence rates (IRs) per 100 patient-years (95% CIs) for fatal infections (within 28 days of treatment) and non-fatal serious infection events (SIEs), respectively: tofacitinib 5 mg BID, 0.18 (0.07 to 0.39) and 3.35 (2.78 to 4.01); tofacitinib 10 mg BID, 0.22 (0.09 to 0.46) and 3.51 (2.93 to 4.16); TNFi, 0.06 (0.01 to 0.22) and 2.79 (2.28 to 3.39).¹ SIEs risk (fatal/non-fatal) was further increased with tofacitinib in patients aged >65 years versus younger patients; therefore, the European Medicines Agency recommended that older patients should receive tofacitinib when there is no suitable alternative treatment.²

Further to these recommendations, we sought to assess age-based (<65 vs ≥65 years) SIE risk in RA patients receiving tofacitinib in Phase 2, 3 and 3b/4 tofacitinib studies with a TNFi control/comparator arm,^3–5 and in the US Corrona RA registry.

The clinical data set included 2180 patients (tofacitinib 5 mg BID, n=1064 (943.4 patient-years); tofacitinib 10 mg BID, n=306 (236.6 patient-years); adalimumab, n=643 (554.3 patient-years); placebo, n=167 (108.1 patient-years)). Overall, 1841 (84.4%) patients were aged <65 years and 339 (15.6%) ≥65 years. Crude IRs (patients with events/100 patient-years) and HRs were calculated for all first infections and first SIEs, overall and by age.

For all infections (online supplemental figure S1), IRs and infection risk (by HRs) were higher with active treatments versus placebo, and similar across active treatments and age groups. For SIEs (figure 1), IRs were higher in older versus younger patients for active treatments, and similar among younger patients for all treatments. For older patients, versus adalimumab, SIE IRs were similar for tofacitinib 5 mg BID and numerically higher for tofacitinib 10 mg BID, though few events occurred in this group (n=4), and the HR 95% CI was wide and included 1. Importantly, HRs revealed similar SIE risk between older and younger patients for tofacitinib 5 mg BID and adalimumab (consistent with previous studies, which report higher absolute SIE risk in older patients but similar relative risk with TNFi in older vs younger patients),^{6 7} while the risk was significantly greater in older versus younger patients with tofacitinib 10 mg BID.

Figure 1.

(A) IRs (95% CI) and (B) HRs^a (95% CI) between treatment groups, overall and stratified by age (<65 years or ≥65 years) and (C) HRs^b (95% CI) between age groups for each treatment group, for SIEs in pooled Phase 2, 3 and 3b/4 studies (months 0 to 12).^c Pooled data from Phase 2 (A3921035; NCT00550446), Phase 3 (ORAL Standard; NCT00853385) and Phase 3b/4 (ORAL Strategy; NCT02187055) studies. IR=unique patients with events/100 PY. ^aCox proportional hazards model includes treatment as the only factor. ^bCox proportional hazards model includes treatment, age group (<65 years or ≥ 65 years) and treatment by age group interaction terms. ^cFor Study A3921035, only data within the first 3-month randomised parallel treatment period were included (before patients who were receiving ADA were switched to tofacitinib 5 mg BID after month 3). ^dCould not be defined as there were 0 events in the placebo group. ADA, adalimumab; BID, twice daily; IR, incidence rate; N, number of treated patients; n, number of patients with event; PY, patient-years; SIE, serious infection event.

In the registry data set (total, n=10 357; tofacitinib, n=1999; biologic disease-modifying antirheumatic drug (bDMARD), n=8358), age-/gender-standardised SIE IRs were higher in older versus younger patients, and similar between tofacitinib and bDMARD initiators for both age groups (online supplemental figure S2).

Our results are consistent with a real-world analysis of >130 000 RA patients, which reported similar adjusted SIE HRs for tofacitinib versus six of seven bDMARDs, including adalimumab.⁸ Limitations of the present analyses should be considered. The clinical data set included variations in sample size and patient-years of exposure between treatment and age groups, and low numbers of older patients and events in some treatment groups which led to wide 95% CIs or undefined relative risk estimates. Additionally, registry data were not matched for baseline variables beyond age/gender.

In conclusion, as would be expected, SIE incidence was higher in older versus younger patients. SIE risk was similar between age groups with tofacitinib 5 mg BID and adalimumab but higher in older versus younger patients with tofacitinib 10 mg BID, suggesting an effect modification by age for this dose. Real-world data showed similar SIE risk for patients initiating tofacitinib or bDMARDs despite limited baseline matching. These data support the globally recommended dose of 5 mg BID for RA.