LONDON — Baricitinib (Olumiant) failed to show noninferiority to tumor necrosis factor (TNF) inhibitors for venous thromboembolism (VTE) risk in FDA-required postmarketing studies, a researcher reported here.
In two large randomized trials with up to 6 years of follow-up in patients — ordered by the FDA as a condition of baricitinib’s approval in 2018 for rheumatoid arthritis (RA) — those treated with the Janus kinase (JAK) inhibitor suffered VTE events at a rate of 2.5%, compared with 1.7% among those assigned to adalimumab (Humira) or etanercept (Enbrel), according to Torsten Witte, MD, of Hannover Medical School in Germany.
That yielded a hazard ratio of 1.61 (95% 0.969-2.660) for time to first VTE event. While not a statistically significant difference, the FDA-approved protocol set a maximum upper bound of 1.8 for the confidence interval to declare noninferiority, which was far exceeded, Witte told attendees at a packed late-breaking abstract session at the European Alliance of Associations for Rheumatology’s (EULAR) annual meeting.
Baricitinib already carries a boxed warning about VTE risk (as well as risks for major cardiovascular events and certain infections and malignancies). But the VTE, heart, and cancer risks were based on a presumed class effect stemming from the famous ORAL Surveillance study with tofacitinib (Xeljanz), the first JAK inhibitor to win FDA approval. The manufacturer of that drug, Pfizer, was required to conduct ORAL Surveillance after its clinical program suggested higher risks relative to TNF inhibitors such as adalimumab.
Some good news for baricitinib maker Eli Lilly did come from the new studies. There was no hint of extra risk for major cardiovascular events, overall mortality, or opportunistic infections, Witte reported. The drug did show a modest and statistically insignificant elevation in cancer rates (HR 1.270, 95% CI 0.853-1.890), relative to the TNF inhibitors, and serious infections were clearly increased (HR 1.323, 95% CI 1.040-1.681). Witte explained, however, that the latter were heavily driven by COVID-19 infections, insofar as the study was conducted during the global pandemic. (Baricitinib was actually authorized temporarily as a COVID-19 treatment during the pandemic.)
It’s worth observing too that cardiovascular risks are, unfortunately, part of the RA package. The disease involves systemic inflammation and it’s well known that RA patients carry a heavier vascular disease burden than the same-age general population, with increased overall mortality in consequence.
Study Details
Lilly organized two separate but very similar trials in response to the FDA mandate: one in the U.S. called RA-BRANCH and another with international enrollment dubbed RA-BRIDGE. Both randomized patients 1:1:1 to 2 mg/day or 4 mg/day of baricitinib or to one of the two TNF inhibitors; the latter selection was made at individual study sites (etanercept is more popular in Europe). Because baricitinib is an oral product and anti-TNF biologics are given by injection, no attempt at blinding was made. Anti-TNF drugs were dosed at the levels recommended in country-specific labeling.
As well as a requirement to show active RA, enrollment criteria included particular risk factors for VTE such as a previous event, age 60 or more, obesity, or overweight status plus age of 50-59. This was intended to keep the enrollment size practical while still ensuring adequate power to detect differences in VTE risk.
In total, 3,640 patients were enrolled and treated. The protocol called for the trials to be stopped when 123 primary VTE events occurred, but the FDA agreed to let the study end when 82 events were recorded, at which point it was determined that continuing would probably not change the main findings. Data from the two trials were pooled in Witte’s report, and for the overall adverse event risk calculations, the two baricitinib dosage groups were pooled as well.
Notably, no difference in VTE risk was seen between the two baricitinib doses. Witte and colleagues calculated hazard ratios of 1.704 for the lower dose and 1.688 for the higher dose (95% CI 0.943-3.079 and 0.968-2.943, respectively), both with respect to the TNF inhibitor group.
It may also be reassuring that VTE episodes were rare, despite the samples being enriched for VTE risk. Events totaled 60 across both baricitinib doses (2,433 participants, 7,694 patient-years of exposure) and 20 with TNF inhibitors (1,207 participants, 3,830 patient-years of exposure).
The international study also had some “exploratory” efficacy outcomes included in the protocol, and those results should gladden Lilly’s corporate heart. The 4-mg dose, though not the lower dose, provided significantly greater benefit than did the TNF inhibitors, at P<0.001 for each of the following outcomes:
- ACR70 responses (70% reduction in symptoms by American College of Rheumatology criteria): 11.7% with 2 mg, 21.6% with 4 mg, 14.5% with anti-TNF
- Clinical Disease Activity Index score ≤2.8, indicating remission: 9.0% with 2 mg, 17.6% with 4 mg, 11.8% with anti-TNF
- 28-joint Disease Activity Score <2.6, indicating remission: 22.2% with 2 mg, 34.4% with 4 mg, 26.5% with anti-TNF
Witte concluded by saying that these data should “contribute to individualized benefit-risk assessments.”
