Pfizer’s big blockbuster Xeljanz flunks its post-marketing safety study, renewing harsh questions for JAK class

Jason Mast

Editor

When the FDA ap­proved Pfiz­er’s JAK in­hibitor Xel­janz for rheuma­toid arthri­tis in 2012, they slapped on a black box warn­ing for a laun­dry list of ad­verse events and re­quired the New York drug­mak­er to run a long-term safe­ty study.

That study has since be­come a con­sis­tent headache for Pfiz­er and their block­buster mol­e­cule. Last year, Pfiz­er dropped the en­tire high dose co­hort af­ter an in­de­pen­dent mon­i­tor­ing board found more pa­tients died in that group than in the low dose arm or a con­trol arm of pa­tients who re­ceived one of two TNF in­hibitors, En­brel or Hu­mi­ra.

And on Wednes­day, Pfiz­er an­nounced that the six-year study had failed. Across 4,362 pa­tients, those who re­ceived ei­ther a low or high dose of Xel­janz ex­pe­ri­enced more ma­jor car­dio­vas­cu­lar events — such as stroke and heart at­tack — than those on Hu­mi­ra or En­brel. They al­so had high­er rates of can­cer, with Pfiz­er fail­ing to hit non-in­fe­ri­or­i­ty on both pri­ma­ry end­points.

Key ques­tions, such as dif­fer­ences in mor­tal­i­ty, have yet to be an­swered. Still, the study re­news dif­fi­cult ques­tions for a class of drugs that has gen­er­at­ed sig­nif­i­cant ef­fi­ca­cy in au­toim­mune con­di­tions but faced con­tin­u­al safe­ty con­cerns. The FDA put black box warn­ings on both Xel­janz and Eli Lil­ly’s Olu­mi­ant and last year re­ject­ed Gilead’s fil­go­tinib out­right, end­ing the big biotech’s plan to de­vel­op the drug in the US.

A Pfiz­er spokesper­son not­ed in an email that, un­like in pre­vi­ous Xel­janz tri­als, par­tic­i­pants in the lat­est study had to be over the age of 50 and have at least one car­dio­vas­cu­lar risk fac­tor. He said they were “con­tin­u­ing to eval­u­ate the find­ings of this study” and will work with the FDA to re­view the full re­sults.

“At this point we can­not spec­u­late on hy­po­thet­i­cal mar­ket im­pact,” he said.

An FDA spokesper­son sim­i­lar­ly de­murred but said they would re­view ad­verse event re­ports and any oth­er da­ta sent to the agency.

“We can­not spec­u­late on whether la­bel­ing changes will be made for Pfiz­er’s Xel­janz,” he said.

In­vestors, though, were quick to spec­u­late about the im­pact for Pfiz­er $PFE, shav­ing 2.3% off its stock Thurs­day morn­ing, or near­ly $5 bil­lion of mar­ket cap. Xel­janz has be­come es­sen­tial to Pfiz­er’s port­fo­lio as the New York phar­ma tries to dou­ble down on its pipeline, bring­ing $654 mil­lion in Q3 2020 alone.

It’s al­so not yet clear how the new find­ings will im­pact JAK in­hibitors more broad­ly. In a note to in­vestors, RBC’s Bri­an Abra­hams said physi­cians al­ready take safe­ty is­sues in­to ac­count when con­sid­er­ing JAK in­hibitors, but that the new da­ta could make the FDA “even more risk-averse on la­bel­ing for the class go­ing for­ward.” That could im­pact In­cyte as they await the FDA’s de­ci­sion on a top­i­cal for­mu­la­tion for Jakafi, one of the on­ly JAK in­hibitors that does not yet have a black box warn­ing.

Mean­while, Gilead’s de­ci­sion to bail on fil­go­tinib now ap­pears pru­dent, he said, and new­er, non-JAK ap­proach­es for au­toim­mune dis­eases could now gain steam.

“Greater bag­gage on Xel­janz and JAK class could keep door open for oth­er an­ti-in­flam­ma­to­ry ap­proach­es in dis­eases like RA, pso­ri­at­ic arthri­tis, and IBD, and ben­e­fit co’s pur­su­ing such ap­proach­es,” Abra­hams said, sin­gling out Gala­pa­gos’ Tole­do pro­gram of SIK in­hibitors.

The Pfiz­er study com­pared rheuma­toid arthri­tis pa­tients who re­ceived ei­ther 10mg Xel­janz, 5 mg Xel­janz or Hu­mi­ra, with just un­der 1,500 pa­tients as­signed to each group.

In raw num­bers, Xel­janz high and low dose pa­tients had 51 and 47 car­dio­vas­cu­lar events, com­pared with 37 for pa­tients on Hu­mi­ra. When cal­cu­lat­ed as a per­cent­age of per­son-years — a met­ric that sim­ply adds up the to­tal num­ber of years each pa­tient has been on the ther­a­py — pa­tients on Xel­janz saw 0.98 events per 100 years, com­pared with 0.73 events for Hu­mi­ra. That amounts to a 33% dif­fer­ence.

For ma­lig­nan­cies, Xel­janz pa­tients saw 1.13 events per 100 years, com­pared to 0.77 events for can­cer — a 47% dif­fer­ence.

Al­though Pfiz­er had al­tered the study last year over con­cerns about mor­tal­i­ty in the high dose, the com­pa­ny said there was no sta­tis­ti­cal­ly sig­nif­i­cant dif­fer­ence be­tween the two Xel­janz dos­es on ei­ther pri­ma­ry end­point. Pfiz­er, though, said they are still col­lect­ing and an­a­lyz­ing da­ta on dif­fer­ences in mor­tal­i­ty and an­oth­er ma­jor ad­verse event, pul­monary em­bolism.