"...MTC was about 12 months from topline to approval."
MTC was 13 months from topline to approval with a priority review. CRPC should be similar. Some of the effort is duplicative in nature, but there will also be two pivotal trials to report on.
"Assuming statsig topline results, when can we expect the filing and subsequent approval from FDA?"
I didn't really answer "when" did I? If I remember correctly, the FDA gets 60 days to accept the sNDA and that starts the clock on the review. 6 months priority or 10 months std. It can take a few months to submit the NDA. If I remember, MTC was about 12 months from topline to approval.
"All the theoretic survival indicators to date have proven positive..."
Let's not forget that there are other indicators of toxicity that are negative, so I know what you are saying and I agree with the premise, but in the final analysis the drug has to overcome its own toxicity and still demonstrate a clinically meaningful OS improvement. Like you, I'm betting that it will, but we won't know for sure until the results are announced.
"...and we have yet to actually determine the lowest known effective dose, particularly given the potentials of synergy in AR combination at earlier stage of disease progression."
The monotherapy dose is pretty will nailed down. Less than 100 and more than 40 mg per day. Within that range each patient needs a little self experimentation to find an individual best dose. I will agree with you on the combination. There are a questions there that need to be nailed down in the clinic and it could take a significant amount of time to come up with answers.
"I am still very optimistic for a favorable review of Cabo benefit on the tumorous micro-environment, and feel that the support of the clinical community is still with us."
Yeah, but they can be fickle. I remember when Provenge was new and now its a "What have you done for me lately" attitude. When it gets down to crunch time, the NDA evaluation period, the lab research is almost irrelevant. Once they get the pivotal clinical data in their hands, the focus is on what the drug does that is of tangible benefit to the patient. It has to be measurable and make him feel better or live longer.
"Assuming statsig topline results, when can we expect the filing and subsequent approval from FDA?"
Absolutely and with a priority review. If there is an ODAC, it would be with the presumed approval, but some questions on labeling.
"Didn't the timing on this change shortly after Cabo approval in MTC?"
I'm not what you are asking. The timing of what? I'm thinking you're asking something about the regulatory process, but I'm coming up blank.
"Am I hearing you say here and now that should we not have a confidently robust data set as guidance to approval...that we may - in turn - have interpretive concerns due to inadvertent unblinding?"
The inadvertent unblinding thing is strictly a Comet 2 issue with a subjectively determined pain endpoint. I'm pretty sure (certain) that Comet 2 will have a statistically robust outcome, so I don't see a problem there. Comet 1 is a separate issue. I'm leaning toward expecting the interim to fail. This isn't a recent opinion. Right from the beginning I've felt the trial was likely to go to full completion and a final analysis. I'm not certain of this, I'm sorta 60/40 on it. The interim can fail and the final can still have a robust data set. My guess is that the interim requires somewhere in the neighborhood of p=.007 or less with a corresponding HR around .7. We could continue past the interim and still see a final p less than .01 with an HR=.75 at the final. No problem.
If Comet 1 succeeds (interim or final) then the pain label would be icing on the cake. The market isn't going to get hung up on the labeling details. If Comet 1 fails and Comet 2 succeeds, then what?? Could Comet 2 rescue the program?? Then we start splitting hairs and looking at all the details. A lot factors that would not have mattered suddenly become very important. Did Comet 1 at least show a trend toward a survival advantage. How robust is the pain data? What other QOL factors were there? Toxicity? Cost isn't supposed to factor into FDA determinations, but lurking somewhere in the background is the question of cost for palliation. Should Medicare/Obamacare pick up $12k per month when cheap opiates will suffice? Hopefully this scenario doesn't play out and we can just keep this as a theoretical discussion.
I'm not sure why Comet 2 is enrolling so slowly as compared to Comet 1. My guess is that Mito (control arm) is just not used very much any more and with 1:1 randomization, onc's would rather put their patients in other trials than have them get a 50/50 shot at Mito or Cabo. Perhaps it's the inconvenience of the infusion visits, I had expected Comet 2 to finish enrolment some time ago.
"...in view of all the Cabo CRPC P2 studies in place - what sort of info might eventually be included in an SNDA pain filing in that indication."
Let's assume that Comet 1 comes back with a statsig OS endpoint and that Comet 2 comes back positive on pain relief. Let's talk about the approval process and resulting label. First, the approval will encompass both trials. If necessary, FDA will drag its feet a bit to allow both trials to read out on their respective endpoints. Assuming a clean data set, both trials meet their endpoints without clear uncontroversial per protocol results, the approval will be a relatively straight forward process. There will be some consternation over inadvertent unblinding of the pain trial due to distinctive side effects assocciated with treatment, but a robust data set will set those concerns aside.
The label will mostly concern itself with the clinical aspects that represent approvable endpoints. There is a section that discusses MOA and effects on the bone microenvironment and cabo's effects on markers of bone resorbtion will make it into this section. There will be no mention of bone scan improvement in the indication. However, because it is a secondary endpoint, it will make it into other sections of the label. The Indications and Usage" is always on the front page of the label and it is very important, because it determines what constitutes on-label prescription criteria and reimbursement. The Cabo indication would include the requirement for CRPC with bone metasteses previously treated with docetaxel and abiraterone or enzalutamide. It would also include a mention of palliative pain relief similar to Mito's label "NOVANTRONE in combination with corticosteroids is indicated as initial chemotherapy for the treatment of patients with pain related to advanced hormone-refractory prostate cancer."
"Until then, RET is another marker among many to test for which adds to logistical complexity and turn around time which would undoubtedly rise if all lung cancer patients were signed up for a test all of a sudden. It will take some time for all this to get rolling."
There is already of coalition of 16 of the top thoracic centers calling itself the The Lung Cancer Mutation Consortium which does multiplex testing of all new NSCLC patients. The test costs about $5K but it identifies specific mutations and also the occasional multiple mutation. So, depending where you live, you are probably already within driving distance of a treatment center that offers a comprehensive gene test. So in a sense the "ball" is already rolling if you are an educated consumer. If not and you don't show up as FISH positive for KRAS, EGFR, or ALK at the local hospital, you probably fall into the "other" category and get chemotherapy.
One of the things MMM mentioned yesterday is that there will be an update on Cabo in RET positive NSCLC presented at ESMO at the end of this month. Abstract should be out around the 20th of this month.
"...and until then how likely is it that a course of Cabo will be used without some confidence that he/she has the RET-fusion mutation? Last time this was brought up, it was mentioned that tests which detect RET-fusion mutations weren't commonly available..."
They are commonly available, just not commonly used. Most of the large service labs will perform the test for about $200 and return the results is less than a week. However there are a handful of clinics that take it a step further and do a comprehensive test on all new NSCLC patients for a full panel of a few dozen possible mutations. That is the direction the practice is moving and the medical care available will catch up to the diagnosis capability eventually.
"...it might be cheaper to start dosing advanced patients with Cabo and then check for a response."
Not so much and this kind of flies in the face of the "targeted therapy" approach embraced by pretty much all of the stakeholders.
"However, as with many of the new cancer drugs, crizotinib comes with very high price tag. In addition, because the biomarker is found in only 3% to 5% of patients with lung cancer, all patients need to be screened.... Therefore, the economic burden of crizotinib on the healthcare system might be too high, Canadian researchers report.
Approximately 50% of Caucasian NSCLC can be accounted for by EGFR, KRAS, and ALK mutations. ROS1 another 1-2% and RET another 1-2% and other oncogenes are being identified on a regular basis. The excuse that "all patients need to be screened" is almost specious and I suspect that the guy who said it realizes that is the case. Anyone with an ounce of insight realizes that in a relatively short time all newly diagnosed NSCLC patients will be given a comprehensive test panel, not a series of FISH kit tests as is still commonly done. Now the other and more relevant issue is the "economic burden of crizotinib on the healthcare system might be too high." Regulators don't seem to get excited about the $5 per month for Avastin, but the $10K that all the new drugs are commanding is causing regulators worldwide to look hard at cost/benefit for these drugs, including Cabo. I think Obamacare has moved the U.S. on step closer to moving away from the "whatever the market will bare" model we currently have.
"According to the slides, looks like MMM will be addressing potential revenues from Cabo..."
Come on people. Those are not Cabo revenue estimates. They are the entire size of the market for the indications they describe. In all cases there are competing drugs and in most cases EXEL is aiming a subgroup indication that even with 100% market penetration would not equal the size of the quoted statistic.
"I haven't looked in depth on METMAB. I have seen the hi MET subgroup analysis regarding PFS doubling in METMAB and OS tripling. I didn't think too much about the OS being so much better than PFS because I have a predisposed bias that MET inhibitors may have more downstream metastatic preventative features than a typical VEGFR inhibitor has."
Let's flesh this out a bit more. In nearly every trial I have ever looked at that simultaneously measures both PFS and OS in terms of HR, PFS has shown a lower (better) HR than OS. The only exception I recall is Provenge, but immunotherapy has a delayed response, so that exception has a rationale. The benefit from small molecules is upfront, early in the disease progression. The patient enjoys a period of tumor shrinkage or stability and then at progression, goes off medication and the benefit generally stops at that point. Let's look at an example. In a controlled trial Cabo patients enjoy a 4 month improvement in PFS, say 6 months vs ten months, in percentage terms patients experience an improvement of 4/6 or 67%. Even if that 4 month improvement is maintained into the survival analysis, it is less marked statistically. Lets say median control OS is 10 months and Cabo maintains its 4 month edge, so we have 10 mos vs 14 mos. In percentage terms, the Cabo patients enjoy 40% better (4/10) survival. The HR statistics will also reflect the dilutive effect of patients no longer receiving medication. I follow your consideration of "downstream" effects, but any benefit should end with progression and the halt of treatment. The circumstance that the control arm in the metmab trial experienced a better OS HR than the PFS HR has only one good explanation, there was a built in imbalance in patient prospective performance.
Let's talk about the metmab ph 2 study, OAM4558g. There is a power point on that study from ASCO and there are a few hints that there is something hinky with the data set. In the overall population the PFS was negative for T+M and positive for OS. In the hi MET subgroups, the PFS was good and the OS was very good favoring the T+M arm. PFS should always amplify a signal of activity. That the OS signal was stronger than the PFS signal (patients go off meds at progression) is a strong indicator of an imbalance in the trial arms. For whatever reason, the T+M hi met arm was very much predisposed to do better than the rest of the randomized groupings. The trial also measured the objective response rates, but there was no mention of this in any of the results publications. PFS and OS can be subject to normal variance, like tossing heads three times in a row. Objective responses almost never occur spontaneously. In looking at an overall picture of potential efficacy, one wants to see both. With no mention (remember what's left unsaid is sometimes more important) of ORR, I assume there were no responses.
"Could you find any press release at the Jevtana interim?"
No. I didn't even know one was scheduled until I read about it in the accessdata medical review.
"Was Exel's mgmt privy to the pace of enrollment during Comet-1? Or were they just alerted when it reached target enrollment? Or were they given notice to certain enrollment milestones?"
Management has made it clear on more than one occasion that they have full knowledge of the enrollment statistics and they have shared some of those details. The large majority of enrolments took place in the period from Jan-Sep 2013. I think it was on Sep 26 they announced that the per protocol target of 960 was reached. Recently disclosed was that there was a slight overshoot and the last patient was randomized in November. This is not unusual and should have little effect on the trial outcome.
"If reported OS from EXAM is less than statsig (or marginal), how could the P4 trial possibly recruit that high dose -.particularly given the obvious probability of higher dose off-target toxicity compromising the OS signal?"
First, there is a good reason for the foot dragging. They are waiting to see if there are any EMA post approval requirements. It would be good to do one trial to satisfy both major regulators. Even if the EXAM trial fails on OS statsig, not to worry. It won't affect the approval. The EXAM trial is not properly powered to demonstrate a statsig OS signal, if it does then that is icing on the cake. Remember, even if the OS is not statsig, the interim done in June 2012 showed a strong numeric advantage, and potential trial participants won't care so much about whether or not a p value was reached as long as the drug extended survival by 6 months. The phase 4 protocol will incorporate an aggressive dose reduction program. The trial can be run in countries without drug approval or concentrate on uninsured and underinsured participants. It wouldn't surprise me if the EMA helps direct patients into the trial to avoid buying the drug for them.
"...but to be nit-picky, you could imagine a weird distribution.....where the resulting difference distribution doesn't have 7 months of median survival...."
You are right. Subtracting a median from another median isn't going to necessarily yield another median. It is an approximation and frankly good enough. From a powering perspective, the assumed HR is more important than nailing down the median performance of the control arm. The trial is adequately powered. If there is a clinically meaningful advantage to treatment, it will reach significance.
From our perspective, it becomes a factor in trying to figure out when to expect the interim and final analysis triggering events, but there is enough uncertainty in some of the other variables to keep it all rather murky.
"(FWIW I found a total of 5 pilot studies comparing sequential ABBY after Enzalutamide and vice versa and all OS were under 7.2 months with an active drug) "
Thx, that is helpful and a bit confidence building. It also takes out a confounding factor. I think where EXEL got the 7 month estimate was by taking the abi arm median OS of 14.8 months and subtracting the median time on treatment of 32 weeks from that nstatistic giving an approximate 7 month survival after drug discontinuation. Or perhaps there is a simple stated 7 month post-treatmetn survival stat in the document I missed or they may have gotten the stat from one of the trial clinicians. The confounding factor is that Comet participants are willing and able to enroll in a clinical trial, Checking the sequential treatment trials for OS removes that filter because obviously those patients participated in a trial post-abi/MDV.
"The trial excluded patients with a history visceral metastases which Comet-1 doesn't."
Good point. The Cabo NRE had 30% patients with visceral mets and the Abi post-chemo trial also had 30% with visceral mets.
I looked up the Jevtana trial result in its pivotal pos-chemo trial. It enrolled 755 patients and had a single interim analysis scheduled. It had to continue to the final analysis and the ultimate HR was .70. It looks to me as if that HR needed to pass a single interim is going to fall right near that .70 inflection point, very close to that 2.5 month OS advantage Hbomb is looking for.