JAK or TNF Inhibitor as First Targeted Drug for Rheumatoid Arthritis?

A Janus kinase (JAK) inhibitor outperformed tumor necrosis factor (TNF) blockers for substantially reducing rheumatoid arthritis (RA) symptoms in patients for whom conventional anti-rheumatic drugs had failed, a pragmatic head-to-head trial found.

Patients randomized to baricitinib (Olumiant) for 12 weeks achieved 50% reductions in RA symptoms according to American College of Rheumatology criteria (ACR50) at a rate of 42%, compared with just 20% of those assigned to any of several TNF inhibitors, according to Celine J. van de Laar, PhD, of Erasmus University in Rotterdam, the Netherlands, and colleagues.

As well, the group reported in RMD Open, 28-joint Disease Activity Score with C-reactive protein (DAS28-CRP) values below 2.6 were seen at week 12 in 75% of the baricitinib group versus 46% of patients taking TNF inhibitors in the PERFECTRA study.

“All PROMs [patient-reported outcome measures] showed a clear improvement for both strategies, generally in favour of the baricitinib strategy,” van de Laar and colleagues wrote. “PERFECTRA suggests that the strategy to start baricitinib is a feasible alternative to starting with TNF [inhibitors] with respect to disease activity and PROMs,” for RA patients in whom a targeted therapy is indicated.

Current guidelines from the ACR and its European counterpart, the European Alliance of Associations for Rheumatology, recommend that, in RA patients not achieving clinical targets with conventional drugs such as methotrexate, either a JAK inhibitor or TNF blocker should be tried next. The choice largely depends on patient and clinician preference, which may be influenced by payer policies, convenience (JAK inhibitors are oral drugs taken daily whereas anti-TNF agents are injected at longer intervals), and side-effect profiles that differ between classes.

Effectiveness has generally been considered equivalent — yet, as van de Laar and colleagues observed, the evidence has come mainly from randomized, placebo-controlled trials in which patients were selected according to strict criteria. “[T]here is a need for real-world evidence of these treatment options within the T2T [treat-to-target] framework,” the researchers said.

To get a better handle on how these drug types stack up, PERFECTRA was started in 2019 with the aim of randomizing 200 patients to an inhibitor of either JAK or TNF. After conceiving the study, van de Laar and colleagues applied to Eli Lilly & Co., manufacturer of baricitinib, for funding to test its product against any of four then-approved TNF inhibitors. Fourteen centers participated, one in Belgium and the rest in the Netherlands, seeking to enroll consecutive patients for whom methotrexate or similar first-line drugs had failed.

Only minimal exclusion criteria were applied: disease duration had to be less than 5 years, patients couldn’t have a clear contraindication for either drug type, and could not have used either previously. Each center was given a randomization list, but each patient’s actual treatment was determined through “shared discretion” of the treating physician and patient.

In the end, 201 patients were enrolled from 2019 to 2022, with 98 assigned to the JAK inhibitor. Among the others receiving a TNF inhibitor, about two-thirds were given adalimumab (Humira); most of the rest got etanercept (Enbrel), while a handful took either golimumab (Simponi) or infliximab (Remicade).

Patients were typical of those seen in clinical practice: mean age about 55, two-thirds women, with median disease duration of 2 years. Methotrexate and corticosteroid doses averaged about 20 mg/week and 20 mg/day, respectively. Baseline DAS28-CRP scores averaged about 4.1 (SD 1.0); the mean CDAI score was just under 20 (SD 9). Patients rated their pain at a mean of approximately 60 (SD 25) on a 100-point scale.

Treatments were given for 48 weeks, with an evaluation at week 24 to identify patients not in full remission who would then switch to the other drug type. ACR50 response at week 12 was the primary outcome, with DAS28-CRP, the Clinical Disease Activity Index, physicians’ global assessment, and patient-reported pain, fatigue, and health status serving as secondary measures.

Baricitinib’s advantages at week 12 were maintained through week 48 for most outcomes. No measure showed even a small advantage for TNF inhibitors at any time point.

Patients were also more likely to stay on baricitinib for the full 48 weeks. About 30% of the anti-TNF group had stopped their original medication by the study’s end, compared with roughly 20% of the baricitinib arm (P=0.04).

PERFECTRA was designed primarily as a noninferiority trial, and the results indeed showed that baricitinib was definitely noninferior to anti-TNF agents. But the JAK inhibitor also met prespecified criteria for superiority, the researchers said, both for intention-to-treat and per-protocol analyses.

Eleven serious adverse events were noted. All but three were considered by the treating physicians to be unrelated to the study drugs. Those three, “possibly” related, were all infections, and all occurred in the baricitinib group. Also, a herniated diaphragm seen after bariatric surgery in one patient on etanercept “could be attributed” to the drug, the patient’s physician reported.

Limitations to the study included the potential for bias in the randomization process, insofar as the individual physicians and patients ultimately chose which drug type to initiate; also, it’s possible that other JAK inhibitors and TNF blockers could have yielded different results. Additionally, the funding from baricitinib’s manufacturer must be considered a limitation, although the study was developed at the investigators’ initiative.

Overall, according to the report, “[t]he pragmatic nature of PERFECTRA, including real-world, non-selected patients, generates results that are more easily generalizable [than data from conventional trials] and provides insight into results that can be achieved in daily clinical practice.”

Please enable JavaScript to view the

comments powered by Disqus.