its odd that the other drugs in testing helping in bone mets seem to show bone flares and we are seeing none here? Some interpret that as not positive and eventually the bone scnas will not be as signifiant as originally thought.
That's all great and I hope it leads to great things for this XL184, but for what is currently the most important indication, it is vital to have all of the scientific ducks in a row. The less ambiguity there is about the efficacy of the drug, the greater the likelihood that a product makes it to market.
WOW You are all beat a dead horse!The biggest event for Exel is it expects to file an NDA for cabo for the treatment of medullary thyroid cancer,or MTC, in the second half of 2011.Prostrate and Ovarian have reached their limit of people for on going studies and in June there will be a huge amount of new updates.Addtionally:We are designing the cabo CRPC endpoint trial in collaboration with Dr. Howard Scher, and his colleague, Dr. Ethan Basch from Memorial Sloan-Kettering Cancer Center.You all know Dr. Scher is one of the world’s prostate cancer experts and a leading investigator in the recent (inaudible) development program. Dr. Basch has unique expertise and that he was a guest worker at the FDA, specializing in patient reported outcomes and was the FDA reviewer at the ODAC [ph] meeting.His expertise and insights have been critical to ensure that we are designing our composite endpoint trial in a rigorous fashion that is acceptable with the FDA, and with the unique insights that come from being an FDA staffer and ODAC reviewer.With the guidance of doctors Scher and Basch, we are collecting additional data in the CRPC NRE cohort,including prospectively collecting pain data using a validated instrument for pain reporting.We will use this initial data to initiate discussions with the FDA for SPA [ph],which most likely will take place in the second quarter of 2011.Last but certainly not least,we have achieved our enrolment goal of 315 patients for the exam trial, he phase 3 pivotal trial of cabo in patients with advanced MTC.This program continues to move ahead as planned,and we are on track to report top line results in the first half of 2011 and submit an NDA in the second half of 2011.
thanks, Clark, very helpful. I do agree that in such a scenario having unstable pain would result in an overestimate of the % who improved in a situation where the drug is having no effect...
but of course, defining such patients would be challenging also, would probably have to be done prior to such a study rather than in the analysis phase to rule out accusations of cherry picking the data, and of course would lead to a reduction in statistical power. I'm not too familiar with these types of trials and whether there are standard approaches to this or not, but Exel better be thinking about such issues as they move forward developing their composite endpoint.
My 2 cents on bone met flare: Since the reported incidence of flare on scintigraphy is between 6% and 25% (according to one source I've read), it seems that the majority of patients won't experience a flare in any case.
The cause of flare isn't well understood, but one theory postulates that increased blood flow due to an inflammatory response to therapy may be responsible for increased uptake on radionuclide scintigraphy. Given that one mechanism of action of XL184 is VEGFR2 TK inhibition (and VEGFR2 is important in angiogenesis), it is also theoretically possible that XL184 inhibits angiogenesis within bone, preventing or minimizing any potential flare.
Further study is required. As others have already noted, improvement in bone pain needs to be quantified more accurately. Finally, future evaluation of XL184 should also consider optimal imaging for assessment of response in bone mets--is radionuclide scintigraphy adequate? Would MRI or PET be better?
<<Since the reported incidence of flare on scintigraphy is between 6% and 25% (according to one source I've read), it seems that the majority of patients won't experience a flare in any case.>>
I kinda doubt this. The abiraterone trial, which is the largest trial I've found looking for flare, found it in about 40% of patients and they first looked at about 3 months. A good bet they would have found it in a bunch more if they had looked earlier. That said, obviously it is likely to be a function of the treatment efficacy as well. As I have noted earlier - if the bone is healing, as it ought to be for a regressed tumor, it should show as hot.
<<Given that one mechanism of action of XL184 is VEGFR2 TK inhibition (and VEGFR2 is important in angiogenesis), it is also theoretically possible that XL184 inhibits angiogenesis within bone, preventing or minimizing any potential flare.>>
Interesting hypothesis - would be interesting to see bone scan data for Avastin, although I'd be surprised if it existed.
<<I'd like to hear more about why you think the pain measurement was weak. >>
Since no one here knows the exact protocol they used to measure pain - including baseline control - we don't ***know*** it is a weak (i.e. unreliable) measurement. But pain is notoriously difficult to reliably measure even with a very rigorous protocol. And pain appears to have been a post hoc measurement for XL184 (thus unlikely to have been rigorously protocol'd) and when they had the opportunity to describe it in the poster they didn't mention anything like a threshold - which I would expect them to do if they used a threshold.
<<Wouldn't it be the case that some such patients would report an increase in pain and some would report a decrease...why would these not cancel one another out? And why should such patients be excluded from a study?>>
Because the metric they used was the % of patients that had improvement on either week 6 OR week 12. Imagine, for the purposes of illustration, that all the patients were with unstable pain (pain randomly oscillating around some central point - call it the baseline) then you could expect that even with no treatment 50% will be lower at 6 weeks and 50% of the remainder will be lower at 12 weeks making for 75% with 'improved' pain. Obviously not everyone will have unstable pain - but even 20 or 30% having such kind of pain will significantly inflate the percent having improvement.
Hope this helps.
I'd like to hear more about why you think the pain measurement was weak. I've heard this a few times here and on the other board but I'm not sure I understand the criticism correctly.
I need to look back, but I remember seeing them say that it was a validated measure. Was it substandard in some way?
You mention that in the Tax 327 study, they removed patients with unstable pain at baseline and state that "this will substantially lower the result". Wouldn't it be the case that some such patients would report an increase in pain and some would report a decrease...why would these not cancel one another out? And why should such patients be excluded from a study? thanks in advance.
I am glad to see others that appreciate the points that Clark is making. All of those issues deserve investigation to enhance our understanding. But that is the science angle.
The investment angle is quite different. Clark's issues suggest the risk that the investor takes in buying EXEL. But there is also a large opportunity cost to NOT buying EXEL here. IF Cabo turns out to offer the benefits that I think are obvious (not proven yet but clear) then EXEL will carry a MC of $4B-$8B if not bought out early. When? Say 2-3 years.
So - follow the science closely, and
buy the dips.
(Arthritic hot spots) <<That would indeed be interesting/useful data - is it in one of their posters or papers?>>
Covered during the Dec 2 R&D day. Still available on company website under Event Calendar.
<<(Or, as I noted on iHub, data from a mouse with a broken bone would be equally interesting)>>
The mouse scans they showed were of the poor little guy's forearm before and after PC cell line transplant and Cabo treatment.
<<How is it doing that without creating a hot spot?>>
I don't know. I will note that the per protocol scans were done with 6 weeks of interval. A lot can happen during that interval that is missed. It sounded as if Dr. Smith had some misgivings regarding what the true meaning of the bone scan resolution phenomena and used an alternative test (CT) to further confirm the apparent effect. Again, those observations are not published, but noted during the R&D day presentation.
<<As we've discussed, bone scan data may not be representative of clinical progression if the drug is, for instance, interfering with the osteoblast activity.>>
Yes, but scans will show fractures and erosion. The absense of those effects combined with the osteoblastic arrestment is further evidence of the consistency of the effect.
<<Robust pain data vs time would be much more telling IMO.>>
We have to deal with what is available to us now. The pain observations don't meet normal phase 3 standards. There are anecdotal accounts from the clinicians of striking improvements, but it is up to us as individual investors to determine how much or how little credence to attach to those reports.
<<But when was the last time you heard a speaker invited to an R&D day say something unglowing?>>
You are absolutely correct on this point. Again, it is up to us as investors to determine how much credibility we attach to what they say and also listen for what they don't say. These guys understand that they are not being invited to present an unbiased viewpoint. I doubt that DNDN would have invited Dr. Scher to their R&D day.
<<Thanks, I think we are in general agreement that, at this point, the pain data is not (to use your word) profound. (Not ripping you - just agreeing)>>
Again, I will point out that it wasn't the pain data that I described with that adjective. My statement was "when the evidence of patient improvement is so profound?" The key observations that justify the termed profundity are the bone scan resolutions. You can take each of the underpinnings that support the importance of the bone scans and find flaws; no flares, PSA all over the map, small sample KM curve, lack of rigor measuring pain, markers are unvalidated, etc. etc. I understand that and that is the sort of exercise I would expect the FDA to do when they reject the notion of a subpart H filing. As an investor, I get to do an analysis that the FDA would consider inappropriate. I get to look at the big picture and decide for myself, does this make sense, what are the likely (perhaps even obvious) conclusions that can be drawn from the totality of the data?
<<First, I will point out that Cabo patients with arthritic joints showed normal tracer uptake in their knees and ankles while simultaneously showing bone scan resolution of their mets.>>
That would indeed be interesting/useful data - is it in one of their posters or papers?
(Or, as I noted on iHub, data from a mouse with a broken bone would be equally interesting)
<<Dr. Smith (their lead investigator at Mass Gen) took it a step further and did CT scans on resolved bone mets and saw what he described as sclerotic activity consistent with remineralization of osteolytic lesions (healing).>>
How is it doing that without creating a hot spot?
<<I’ll next note that they have been watching the first 20 bone scan patients for several months now and out of the total group of 62 Cabo bone scan patients, only one has progressed by bone scan.>>
As we've discussed, bone scan data may not be representative of clinical progression if the drug is, for instance, interfering with the osteoblast activity. Thus I think your first point (about arthritis) is most interesting - but I take the bone scan data with a large grain of salt whether 10 weeks or 40. Robust pain data vs time would be much more telling IMO.
<<Couple these observations with the durable marker improvements comparable to those seen with bone drugs Zometa and Denosumab>>
Not much question that XL184 has more cancer killing activity that zometa etc. And I'd expect the profile to be different because biphosphonates are anti-osteoclastic - thus not going to directly effect bone scans or ALP.
As for the R&D day speakers - I like R&D day speakers for the background they provide. The insight into the medical community. But when was the last time you heard a speaker invited to an R&D day say something unglowing? What I'd want to hear is the nay sayers (and I have no doubt that there are some prominent members of the PCa community who have doubts) - or ideally from some member of the medical community who has a history of being even-handed (e.g. like Nissen in cardiac issues)
<<The RDT did not incorporate a rigorous tool to measure pain. I recall mention of a post hoc questionaire. Cross trial comparisons are pretty much meaningless.>>
Thanks, I think we are in general agreement that, at this point, the pain data is not (to use your word) profound. (Not ripping you - just agreeing)