Biotech: PFE's High-Dose Jak Setback Positive for GILD, GLPG, CELG; Mixed for INCY
February 20, 2019
RBC Capital Markets, LLC
Brian Abrahams (Analyst)
Yesterday PFE released a statement that based on increased rates of PE and overall mortality in the 10mg BID vs. 5mg BID (or the TNF-inhibitor group) observed in the ongoing post-marketing study of Xeljanz in older, high-risk RA pts, they are discontinuing that dose in the study. We note that, per the label, 5mg BID is the max. recommended dose for RA and 10mg BID is only recommended for UC. However, we believe that this new development for PFE highlights potential safety risks for the Jak inhibitor class overall and the importance of dose and selectivity profile in determining each drug's therapeutic window - and ultimate potential safety profile. Given our view that GILD/GLPG's filgotinib has the best CV risk profile (lowest thromboembolic event risk) to date, this could give GILD/GLPG an opportunity to capitalize on this commercially even as a later entrant - assuming better safety is maintained in larger FINCH 1 and 3 (~1,300 pts each) and the U.S.-gating and slowly-enrolling MANTA looks clean on testicular tox. Any additional concerns about Jak safety could also benefit CELG, whose key pipeline drug ozanimod is an S1P modulator targeted for overlapping indications. On the flip side, the translatability of Jak-related thromboembolic events to increased mortality in a high-risk population further reduces the likelihood INCY's partner LLY will be able to convince FDA to approve the higher 4mg dose of baricitinib, though this is less of a key driver for INCY.
New data highlighting toxicity of long-term, high-dose Xeljanz could potentially open a door for GILD/GLPG's filgotinib, which seems to have less thrombotic risk compared to other Jak inhibitors. As we await results of the FINCH 1 and 3 studies of filgotinib 100mg and 200mg in RA (MTX-IR and MTX-naive, respectively; expected 1Q19), the news highlighting that PFE's high-dose Jak inhibitor led to a "statistically and clinically important" increase in PE risk and overall mortality will likely increase the already high focus on CV and other AEs seen with the Jak class. Based on our deep dive leading into the FINCH 2 readout last year, and our subsequent analysis of the the FINCH 2 data, we believe that the ph.II and ph.III data so far suggest filgotinib carries less AE risk than the other marketed Jaks and those in development (e.g., upadacitinib; PDUFA 3Q19). In the DARWIN 3 long-term safety study of filgotinib in RA representing 1,700 patient years of exposure, the VTE rate for filgotinib was 0.1, compared to rates of 0.5-0.7 reported for upadacitinib, 0.5 for baricitinib and 0.2 for tofacitinib. We note that in DARWIN 3 the only VTEs (1 DVT and 1 PE) occurred in the same patient, and in the ph.III FINCH 2 study the only thrombosis-related event was a retinal vein occlusion and no deaths were reported (out of 448 pts treated with filgotinib). For comparison, in the ph.III SELECT-BEYOND study of upadacitinib in RA, in a similar pt population to FINCH 2, 4 pts had PE and there was 1 death due to cardiac failure and suspected PE (out of 329 pts treated with upadacitinib). GILD management has expressed cautious optimism about the competitive prospects for filgotinib as they await results from the large FINCH 1 and 3 trials before they feel confident claiming superior safety, and the MANTA testicular toxicity study remains a significant hurdle (timeline-wise) required for U.S. filing - given the challenges to recruitment discussed here. That being said, we believe today's development does highlight a potentially increasingly important commercial advantage that lower thromboembolic risk could be if filgotinib were ultimately approved.
Relative Jak1 selectivity may be key in the balance of anti-inflammatory activity and AE profile. The reduction in AEs observed thus far with filgotinib may be due to increased selectivity for Jak1 vs. Jak2 .