JAK Inhibitors for Rheumatoid Arthritis (Follow-Up)

Galapagos followed up strong 12 week data with equally impressive 24 week data for their two trials, DARWIN 1 and 2, testing the use of their selective JAK1 inhibitor filgotinib in rheumatoid arthritis (RA). In an earlier comparison, 12 week data from the DARWIN 1 trial compared well against the marketed incumbent JAK inhibitor tofacitinib, as well as the late stage competitor baricitinib. The following graph compares the 24 week data by focusing in on the best dose results for DARWIN 1 and 2 compared toIncyte’s newly released RA-BUILD and RA-BEACON data, along with tofacitinib data as a comparator (Study IV in the prescribing information). First is a comparison of the relative change in disease activity through the American College of Rheumatology (ACR) scale, and subsequently the absolute disease activity score (DAS28) index below 2.6, which is considered a sign of remission:

ACR70 and DAS28(CRP) < 2.6 (Placebo Adjusted)

Note: tofacitinib DAS28 data are DAS28-ESR, not DAS28(CRP); for DARWIN 2, the 12 week placebo responses were carried over to 24 weeks for the purpose of comparison.

As always, these comparisons between investigational drugs are inexact. Particularly, the exclusion criteria for the four referenced trials are not identical. Whereas Incyte accepted patients treated with prior anti-TNF drugs in the RA-BEACON study, RA-BUILD and DARWIN 1 and 2 all excluded prior anti-TNF therapy. And unlike the other 3 trials, DARWIN 1 did not require enrollees to have demonstrated an inadequate response to a prior conventional disease modifying agent (cDMARD). In effect, these trials likely have patient populations that, in order of RA severity, are DARWIN 1 < RA-BUILD < DARWIN 2 < RA-BEACON. This ranking, although not definitive, can be generally observed in the placebo response trend across these trials.

The general efficacy trend for these drugs, along with tofacitinib’s concern over gastrointestinal perforations, suggests that the investigational drugs baricitinib and filgotinib will have a chance to compete for market share. For prospective investors, the comparative outlook for baricitinib versus filgotinib is likely more interesting.

In terms of efficacy, the trend for filgotinib’s numbers appears superior. Galapagos has done a good job of identifying and focusing in on the doses that provide maximal response. However, the company has noted there was no significant difference in the efficacy between 100 and 200 mg total dose results, and in DARWIN 2, 100 mg filgotinib often had numerically larger ACR values than 200 mg. In effect, final data from prospective phase 3 trials with filgotinib are more likely to fluctuate within the range of efficacy presented by the 100 and 200 mg total doses rather than specifically replicating the highest values. In that regard, and given the differences in enrollment between the baricitinib and filgotinib trials, a final efficacy comparison between these two drugs will likely be a push. Neither drug is likely to prove superior efficacy over the other.

In regards to safety, some analysts are suggesting an advantage for filgotinib, primarily due to the hemoglobin level comparison for patients on filgotinib versus baricitinib. I suspect this factor is being overrated, particularly for drugs within the same class. Long time followers of the oncology space will know that carboplatin is perceived (and possibly even proven) to be more tolerable than cisplatin. Further in oncology, convenience due to a lack of concomitant steroid administration has been touted as an advantage for enzalutamide in its competition against abiraterone in prostate cancer. Although such conveniences and advantages can provide an upper hand in the long run, they’re not rapidly converted into market share absent a clear efficacy benefit. Carboplatin never squeezed out cisplatin from use despite comparable efficacy and improved adverse event profile (generally summarized here). Arguably, enzalutamide is now slowly taking over abiraterone due primarily to the mounting evidence for improved efficacy that has emerged since its approval; its advantage in dosing convenience did not create immediate and large dividends in the market. In effect, unless the adverse event profile significantly impacts management and practice, small changes in percentage one way or another are unlikely to be market-defining. In this regard, I don’t expect the hemoglobin effect of baricitinib versus filgotinib to be meaningful in the marketplace. Hemoglobin levels with baricitinib are primarily between 10 g/dL and the lower limit of normal, and do not fall under 8 g/dL. Many individuals with RA already present with anemia of chronic disease and have hemoglobin below the lower limit of normal. Nonetheless, the anemia in these patients is not targeted as a primary focus of care; rather, the underlying disease is the target. In that regard, mild changes in hemoglobin level with baricitinib versus filgotinib are unlikely to be a strong raison d’etre to prescribe one over the other, especially when these changes in hemoglobin levels do not require any intervention. Had either drug demonstrated a meaningful edge for infection risk, one could have made a meaningful case for an advantage to play out in the marketplace.


As an overall efficacy and safety proposition, there does appear to be a gap between baricitinib / filgotinib versus tofacinitib. If baricitinib is approved, it has the first shot at exploiting this gap but with no assurances that it will rapidly or meaningfully change prescriber preferences. If filgotinib is moved into registrational trials for the US market, I expect it to gain approval with a compelling efficacy package. However, the efficacy and safety proposition between baricitinib and filgotinib does not appear to be large. Looking forward, I expect first-to-market will be a larger factor in the competition between baricitinib and filgotinib.