Cosmid,
Great post. I think most of what you have laid out has been hashed to some degree in many places already but it’s nice to see the summary of contrary points all in one place.
However, I would mention you shouldn’t underestimate what Bristol knows either. Don’t for one second think we have all the data. I’ve been on that side of the research world, you don’t show all your cards on the flop… we haven’t seen the river card just yet. However, I’m not foolish enough to believe that what we don’t know can’t be bad just as much as it could be good. Like you, I and many others have said, there is a lot of risk in NKTR.
Who said it was the largest deal ever for a drug? I must have missed that. I was never under that illusion. As for the abscopal effects… yes that’s been observed before. It was just an aside mention (and a deserved one) in my post and not a big part of my investment thesis.
Also, you said: Be careful, this can be very seductive. To my knowledge, 214 has yet to be tested with a control arm, so the appearance of increased efficacy is derived by comparison to historical trials. If I had a dime for every time that something performed better with each subsequent trial, I wouldn’t be investing, because I would be fully vested. In fact, I would interpret these findings to be a signal that they are getting better at selection against non-responders. This is good, right? Well, yes and no. An alternate interpretation (and my hunch) would be that there are more non-responders than anticipated, and they are getting better at NOT enrolling them. Not very holy grail-like.
I have some issues with this part of your post because some things need to be corrected and clarified. I think you may be misunderstanding what Saul, Chris and I summarized. Maybe that is my fault and I’ll try to do a better job this time. These are not “subsequent trials” or new enrollees in the trial. These are the same patients in the original trials who were not deemed responders after the initial follow up period but have become responders with a longer period of time on 214. That is not the same as the scenario you suggested where they pick better patient’s for 214 in later trials or subsequent phases to get better results. So the increasing efficacy has nothing to do with historical controls or lack of a control arm nor does it have anything to do with picking better patients for the trials or subsequent enrollees in a later phase of the trial.
Just so it’s clear, these are NOT new patients that NKTR selected because they felt there would be a higher percentage of responders in a more “selected" group nor is it because they are changing the divisor. These are the originally enrolled patients who remained on NKTR-214 despite not being a responder at initial follow up who then later became responders after a longer period of time taking the drug(i.e., patient 5 at 3 months follow up was not a responder but at 6 months of follow up patient 5 is now responding with much less cancer burden on imaging). That was almost unheard of in oncology until the PDL1 therapies. If it was just better patient selection I wouldn’t have even mentioned it in my post but it is the most impressive thing to me about 214 thus far. Many of them have become complete responders with no evidence of cancer remaining. What is the durability of the response? That’s the question now isn’t it? To be determined…
This is Phase 2 data, almost never compared with a control or placebo arm. It is observational data with endpoints/targets defined by the investigators that will be predominantly safety based followed by efficacy. Yes, historical comparisons will be made but that is NOT the aim of the study. The aim is two-fold, is it safe and does it work like the Phase 1 data suggested it would but on a larger scale? Did we miss something in the small scale that is now apparent on a larger scale? That’s what Phase 2 trials are intended to determine. The other point to be made here is that the patient’s being studied are in advanced stages of their disease for the most part. That’s the main reason there isn’t a control arm… what are you gonna do to the control? Watch them slowly waste away while offering them hospice? No. They are going to see how long 214 patients live and compare that to what we know already–> how long does a Stage 3/4 lung/breast/colon/melanoma cancer patient live on average? That’s the control and it’s the only one you need.
MC