Tofacitinib Deaths in Context

Posted by Jeff Berk, BOLT International;


So I’m flying back from some meeting and I’m too brain dead to even fire up my IPad.  As is my custom, i peer between the seats and watch whatever the stranger sitting catercornered (yes.  I googled the spelling) from me is doing on his PC.  These peeps, by the way, are ALWAYS Type-A personalities; of which, if you haven’t figured it out by now, I’m not.  I’m Type-J.  Anyway… This particular flight was destined to change my life.  Because Mr. Type-A was watching .  Trust me, if you get nothing else of value from my blog today, you’ll thank me for passing along that link. There’s a lot of really cool stuff on there.  Check out Keith Barry.  He is just plain FREAKY!



Okay, so in one of the tofacitinib abstracts that we’ll see at EULAR 2011, the first one that reveals Ph-III results from the combination of tofacitinib + methotrexate, is the following:



J. Kremer 1,*, Z.-G. Li 2, S. Hall 3, R. Fleischmann 4, M. Genovese 5, E. Martin-Mola 6, J. Isaacs 7, D. Gruben 8, G. Wallenstein 8, S. Krishnaswami 8, S. Zwillich 8, T. Koncz 8, R. Riese 8, J. Bradley 8

1ALBANY MEDICAL COLLEGE, Albany, United States, 2Peking University People’s Hospital, Beijing, China, 3Cabrini Medical Centre, Victoria, Australia, 4Metroplex Clinical Research Center, Dallas, 5Stanford University, Palo Alto, United States, 6Hospital Universitario La Paz, Madrid, Spain, 7Newcastle University & Freeman Hospital, Newcastle-upon-Tyne, United Kingdom, 8Pfizer, Inc., New London, United States

792 pts (81.4% female) randomised and treated: 5 mg BID, n=315; 10 mg BID, n=318; PBO→5 mg, n=79; PBO→10 mg, n=80. Mean baseline values: age, 50.8-53.3 y; disease duration, 8.1-10.2 y; HAQ-DI, 1.24-1.45; DAS28-4 (ESR), 6.16-6.44; RF positive, 72.2-73.9%. Tofacitinib was statistically superior to PBO for primary efficacy endpoints (Table). Significant ACR20, ACR50, and HAQ-DI responses were seen vs PBO by Week 2. Most AEs were mild; most frequently reported (Mo 0-12) was infections and infestations. There were 4 deaths and 4 opportunistic infections. Deaths were: acute heart failure (10 mg BID) and respiratory failure (10 mg BID) during the study; traumatic brain injury (5 mg BID) and RA (5 mg BID) following discontinuation (22 and 42 days, respectively).  SIEs were reported for 2 (5 mg BID) and 4 (10 mg BID) pts in Mo 0-3 and 1 pt (each of 5 and 10 mg BID) in Mo 3-6. Decreases in neutrophils, increases in LDL and HDL, and small increases in serum creatinine were seen.

Conclusion:  This was the first Phase 3 study of tofacitinib in combination with background DMARDs in pts with active RA. Consistent with Phase 2 studies and a Phase 3 monotherapy study, tofacitinib demonstrated rapid, significant and clinically meaningful reductions in signs and symptoms of RA, and physical function. No new safety signals were detected.






This published today, and Pfizer’s stock promptly dropped about 3% on the news of the 4 deaths and 4 opportunistic infections.  (NOTE: I’m long on Pfizer, and thus today on paper I’m about 20 Benjamins lighter).  The questions that I’d like to address in this blog are 1) are the deaths caused by the drug? and 2) what will be the overall impact of these deaths on approval and uptake of tofacitinib.


The first question’s the easy one.  Nobody knows whether the deaths are caused by the drug, and 4 infections out of 318 patients on an immunosuppressive drug is no big whoop-de-doo.


As for the second question…  When I interviewed BOLT’s Rheumatology Thought Leader panel last month, nobody thought that tofacitinib should be used before anti-TNF failures until it had years of safety data behind it AND a demonstration that it reduces joint space narrowing.  So today’s news supports that indication, as there will almost certainly be a black box on the label to deal with the deaths.  My Panel had no intention of using tofacitinib upfront, and it’s highly unlikely that FDA / EMEA are going to allow such a positioning anyway.

Interestingly, the panel noted that they were much more impressed with tofacitinib monotherapy than the combination with methotrexate.  They also point out that Pfizer isn’t yet even running the trial that they want to see in TNF-inadequate responders.  So it’s going to be years before we know whether tofacitinib can be safely combined with an anti-TNF.


I’m going to share with you a few of the comments that they shared with me:

  • Tofacitinib is a dirty drug in that it inhibits many JAKs; at least JAK1, JAK2, JAK3 and TYK2.  It takes out some but not all of the important cytokines.  The big ones it is blocking are IL-2 and IL-9, which is coming from TH9 cells which are thought to be inflammatory, and IL-21.  Those are all involved ultimately in the development as well of TH17 cells.  So also downstream there is an indirect effect by inhibiting transcription of things like IL-6 and both the interferons, interferon alpha beta as well as interferon gamma.  Pfizer may argue, with some merit, that because their drug has such a short half-life, they are not oversuppressing the immune system by targeting such a broad spectrum.  As one Panelist put it, “since you don’t give such a high dosage since the half life is short and you don’t fully suppress all the JAKs but you take away the cream in the coffee, which is the inflammation”.  So this is similar to anti-TNF therapy; which never abolish all of the TNF.  And thus the rate of infection with anti-TNF therapy has not been as frequent as early dire predictions suggested that it would be.


“It is a dirty drug.  It is thought to block mostly JAK1 and JAK3.  The thing that is overall pretty nice about it is that it does take out some but not all of the important cytokines.  The big ones it is blocking are IL-2 and IL-9, which is coming from TH9 cells which are thought to be inflammatory, and IL-21.  Those are all involved ultimately in the development as well of TH17 cells.  So also downstream there is an indirect effect by inhibiting transcription of things like IL-6 and I think both the interferons, interferon alpha beta as well as interferon gamma.  So yeah, because it is not perfect I think that is a good thing.  My concern with the perfect sorts of agents, and maybe anti-CD20 is an example, you get something that works better, it binds tighter, it binds better to the FC receptors and what happens?  Your toxicity goes up.  I look at all these as still pretty blunt tools and if you have one that works and the toxicity profile is pretty reasonable just because you have one that might bind much more selectively to JAK1 or JAK3 or have a longer half life doesn’t necessarily make it any better”.



“I have to say I have to tell you the truth I was extremely hesitant with regard to oral drugs, the JAK inhibitors, because I thought they are completely dangerous suppressing the immune system and so on.  So I was very, very skeptical.  I am no longer a skeptic.  I am really becoming a believer.  I am always very open because not having been successful for many years doesn’t mean that the concept is entirely wrong and you may come up with a winner and the winner takes it all then like potentially tofacitinib”.



“When you look at the spectrum of targets that tofacitinib hits, it is hitting in my view all three of these JAKs. At least. There are additional JAK-like enzymes which do not have the JAK name in front of it.  Initially I thought it was a disadvantage because they are not very specific.  But maybe that is the reason why it is such a good drug because you don’t hit just one but you hit all of them.  Since the tofacitinib has let’s say a short half life, which I think may be an advantage, there is some possibility for the cells to recovery because otherwise you would say this must be very, very severely immunocompromised patient whose JAKs are all blocked,  but apparently at least that is what they say.  I think it is worth arguing that since you don’t give such a high dosage since the half life is short and you don’t fully suppress all the JAKs but you take away the cream in the coffee, which is the inflammation.  That makes some sense to me.  It is like with anti-TNFs you never abolish all the TNF but let’s say the inflammatory, the TNF and therefore there is still TNF in the regulation and it is not as dangerous as if you took TNF away completely”.


  • Monotherapy efficacy in TNF naive patients with long-time RA, has been very impressive with tofacitinib; in the range of 40% improvement vs placebo on ACR 20 and HAC disability.  One panelist questioned the wisdom of including the third primary endpoint, DAS remission, which is the one that the trial missed.  At week 12, the ACR 20 for both 5 mg and 10 mg was 61%.  ACR 50 was 40% for the 10 and about 37% for the 5 mg.  ACR70 is about 24% for the 10 mg dose and about 18 for the 5 mg.  So it fits the 60, 40, 20 rule for what is expected to be observed with an anti-TNF.  Another point of note is that the patients were generally severe, with a DAS 28 score on average going in at baseline of 6.7 (really high), and CRPs were quite elevated.  About between 15% and 20% of patients had been on previous anti-TNFs.  70%-80% had failed prior methotrexate.


  • Combination data with methotrexate is not bad, but not as robust as expected.  Tofacitinib met its primary endpoints in ORAL 1032 by showing statistically significant changes versus placebo in reducing signs and symptoms of RA, as measured by ACR20 response rates at six months in patients who were on background methotrexate; in improving physical function, as measured by mean change in HAQ DI at three months; and in reaching DAS28-4(ESR) <2.6 at six months.  Efficacy appears to be independent of CRP level (c-reactive protein).


“To date the monotherapy data has been very impressive with close to 40 percent improvement over placebo.  That was both the phase II data and the most recent phase III data from the ACR.  The combination data with methotrexate has been a little less robust but still quite good.  It is close to the neighborhood of what we see with TNF inhibitors.  I think most of us look at this as efficacy profile which is pretty comparable to the TNF inhibitors or any of the biologics.  So to that end, people are going to think that this is, at least for the data we have, fairly equal efficacy”.


“They are just like the anti-TNFs.  You can see both ways.  First of all, of course I am impressed because for years there haven’t been any oral drugs that have any apparent progress or promise so here is one that shows similar action as a TNF inhibitor so you can be impressed because this is finally one oral that makes it to the market potentially”.


“Tofacitinib doesn’t seem to have the issue of what the CRP level is.  Yeah, that is correct.  I would say I am much more bullish on the Pfizer product.  I agree with that.  I think that if we had to choose amongst agents I would choose this one because I think it is probably going to be more effective.  Tofacitinib? The tofacitinib, exactly.  I am much, much more bullish about that.  I think that their data are just a lot more solid”.




  • There is inadequate data on efficacy of tofacitinib in the TNF inadequate responder population, but this is the group where the drug is actually going to be used first.  This niche, adding tofacitinib instead of methotrexate to anti-TNF, is one where the risk/benefit ratio should most favor tofacitinib.


Does Actemra have a place in your practice? Yes. Who is getting Actemra? People who have failed Enbrel, failed anti-TNF therapy.  So when tofacitinib is available how are you going to choose between those two? I don’t know.  Good question.  Don’t know.  I think if you are failing an anti-TNF I will have to see how that data compares in someone who fails…we don’t have that data yet.  I don’t know how the drug is going to work if you fail a biologic because they haven’t done those studies yet.  Everything they have done is TNF-naïve.  That is a very different population then perhaps where the drug may be labeled.  I am sure they don’t want a label on a TNF-alpha failure market.  That is not why they are spending hundreds of millions of dollars”.

“And most importantly data in the TNF inadequate responder population because that is the population where the drug is really going to be used and where the risk benefit ratio is really in most favor of the drug”.


“I would like to see the results with the dual therapy more.  I say that because even if they use monotherapy, and I might use monotherapy in my practice, no one is going into remission on monotherapy.  Even the data with the JAK inhibitor doesn’t have very impressive remission rate.  So it suggests to me that even if it is good and safe as a monotherapy that is fine and well but I am going to need to add more drugs to it to get a higher magnitude of response.  I might as well know up front what are the ramifications good and bad for using it as a combination therapy, particularly in the refractory groups”.



“Tofacitinib has not looked at patients who failed TNF-alpha blockers.  They have never reported that study”.


  • Thus far there is no radiologic data showing that tofacitinib halts disease progression.  The first data is expected to be presented at EULAR 2011 (May 2011), but the whole trial is still a year away.  There is speculation that there will be a moderate effect on joint space narrowing.  None of this data will be in patients who progressed on a biological, so because this is the setting in which tofacitinib will first be considered, this will be a limitation in the product label at launch.

“Has there been any public release of joint space data? No, they don’t have that.  When? A year.  A year away? Yeah.  I honestly can’t see FDA giving them a front line labeling without joint space narrowing data. But they are going to have that within the year.  So it will be a change in the label? Potentially”.

“What we are really missing as we talked about before is radiographic data that demonstrates the drug will slow structural progression”.


“There are no radiological data.  We don’t really know if radiological progression is halted by the oral drugs because neither fostamatinib nor tofacitinib currently has any data that are relating to radiologic progression and we all like to see that, that this is halted”.


“I would say the biggest concern I have is what happens at EULAR when they start presenting some of their data about joint space narrowing.  I guess there has been some concern just in talking with people – it is unclear if those data are going to come back positive or not and is that going to change the likelihood somebody is going to prescribe the drug?  So if the joint space narrowing data comes back ambivalent or negative? I think that is going to be a big negative.  And is it still going to be prescribed?  Yes, but does it have the potential to just completely crush TNF inhibitors?  No.  I think that will be gone because all the TNF inhibitors, every one of them, all of them reduce joint space narrowing.  Some of them claim to do it faster than others and subtle differences and things like that but if you don’t have that then it makes it really hard for a physician to want to prescribe it.  And that comes up over and over in SAB meetings that I go to of TNF inhibitor companies.  If they miss in their Sharp scores that is going to be a huge problem for tofacitinib”.


Categories: IMMUNOLOGY, Rheumatoid Arthritis
Tags: , , , , , , , , ,
Bookmark the permalink.

Post a comment

Comments on are monitored.

You may use these HTML tags and attributes:
<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>