"So are you saying down the road if these kind o trialf numbers hold up in a phase III trial the potential market is 1.5B."
You went off the tracks a bit on this one. Tarceva's $1.5 B includes 2 other indications. It is also a moving target as other companies are moving into the space with their drugs. All the PD1 and PDL1 entrants are going to shake things up a bit. Eventually it will settle out. The checkpoint inhibitors will not "cure" many patients and eventually the majority will progress into the latter stages of disease despite better treatments in earlier stages. Cabo will likely be a 2nd or 3rd line treatment in most of its eventual indications.
"Yeah but how is Cabozantinib going to compete with Rociletinib in this space."
Different space. 1512 was in wild type non-mutant EGFR NSCLC. BTW I like Clovis and pulled the trigger on it a few months ago.
"So if I am following you guys correctly the wtEGFR population is larger with a better OS, HR and PFS for Cabo over Erlotinib?"
That is correct.
The 1512 trial is in the refractory indication. The Erlotinib arm result was pretty consistent with what one would expect in this population. The beauty of this is that while it is a subgroup it comprises about 85% of the approved indication. In 2013 Erlotinib was a $1.5B drug in 3 indications, frontline EGFR pos NSCLC, 2nd line NSCL and Pancreatic cancer. I'm not sure of the split, but my guess is that 2nd line NSCLC makes up at least a third of that total. This is real partnership bait. It is the obvious next step in the clinical development program and the heavy lifting for trial design is already done.
"And did I mention this is in wtEGFR pts. Not exactly a small piece of the pie!!!"
Not everyone follows the significance of this so let me explain. Erlotinib (Tarceva) is a once a day pill that targets the epidermal growth factor receptor (EGFR). NSCLC patients are typically tested for EGFR mutations. Only a small minority of patients test positive for this mutation, somewhere in the 10-15% range. Those lacking the mutation are said to have wild type EGFR or wtEGFR. Tarceva is most effective in patients with the mutation. In the frontline setting patients are required to be EGFR positive. In the refractory setting Tarceva is approved for use in patients with both wt and mutated EGFR. It confers a benefit to the wt patients, but is not as great as that for the EGFR positive population.
I recall when the FDA finally approved Cabo for MTC. The pps actually dropped, and did so significantly. The pps was run up by uninformed speculators anticipating the approval without an appreciation for how small the indication is. The MTC approval serves multiple purposes, but profits from this indications may never cover the clinical and administrative costs of approval and marketing.
So what's RCC worth? Afinitor is the most widely used 2nd line treatment and it is in its peak years in this indication. It pulls in about $100 million per quarter. It is priced at $6k per month and likely treatment duration is about 5 months. Other 2nd line treatments account for about $150 million per quarter making 2nd line RCC and worldwide billion dollar market. If Meteor succeeds, It is not unreasonable to expect Cabo to assume Afinitors current market penetration. It should also be considered that Cabo is a higher priced drug and if superior to Afinitor, per patient treatment duration will be more lengthy. So, if Cabo were to be used over the same patient population, revenue would higher based on pricing and longer use.
So depending on how convincing the actual results are, approval based on Meteor could result in a revenue range of $400-$700 million annually. IMO this certainly would justify a boost from the present pps. Add in Cobi and renewed speculation about other indications and we could see some decent appreciation from present levels. That's the rosy scenario. A failure on Meteor would see us below $1 per share and Cobi asset would just about cover the cumulative corporate debt. We'll see.
Abstracts at 5pm.
'I am finally approaching my average price and am quite tempted to sell, but something keeps telling me this latest climb coupled with all the near term potential will pay off.'
Reminds me off Yogi Berra's "de ja vu all over again.." witticism. There's a difference this time, history and conventional wisdom are on our side. With the Comets, EXEL was bucking a trend. VEGF inhibitors had been tried and failed previously in CRPC. We wanted to believe that Cabo was special and had almost mystical effects on bone mets. The before and after scans were very impressive and the Comets did show a respectable 40% bone scan response rate, not quite equal to the 60-70% seen in the ph 2 program, but still decent. Problematic was effect was very temporary. I suspected Cabo would give about a 2-3 month average benefit and the PFS data seems to confirm this, but the full effect did not flow through to a sufficient OS benefit to give a give a statsig result. The pain data from Comet 2 was an important lesson, never underestimate the placebo effect. I will never again give any credence to unblended self reported outcomes.
Why do I say this is different? In RCC, VEGF inhibitors are the treatment of choice and have shown good results in pivotal trials. There was some doubt about using sequential TKI's, but the Cabo ph1B data and a recently published article "Metastatic RCC patients may benefit from sequential TKI strategy" suggest otherwise. Also important is the PFS vs OS primary endpoint. Cabo has not been compared head to head against other dirty VEGF TKI's (sunitinib, sorafenib, pazopanib, axitinib, et al) but its results in similar trials compare favorably and Cabo has shown very good data when used in sequence after prior TKI use. A statsig result in a head to head comparison with everolimus by any of the effective TKI's would be the anticipated result. So, that's why I feel as if history is on our side on this one as opposed to trying to establish a new paradigm.
"....analysis such as Ernies and said, 'hey, with 90% certainty the delay in events is caused by Cabo, we should put the wheels in motion" What do you think?"
Sorry, but that's not even close to what I posted. I don't mind being quoted or paraphrased as long as it's done accurately. Not sure what's behind the rally, but it is nice to see.
It does raise the issue of what happens to the sp after the topline results are released. Assuming they hit the primary endpoint, how high we go depends on how convincing the results are. Following the release of topline, expect either a quick secondary, a global partnership deal or maybe both. Then, there will likely be more news on Cobi in the next few months. The devil will be in the details.
"I think the answer about to whether ernie believes METEOR will be successful is he's not sure but he's hoping it is. With EXEL's recent history who can blame him for being cautious."
That's a good summation of my current thinking. What I said about power was not an opinion, approximation or prediction, it was strictly a pure statement of the definition of power and how it applies to this trial. If there were a way to state where this is going to go without qualifiers, conditions, assumptions, guesswork or equivocation then there wouldn't be any doubt or risk involved. If that's what anyone wants, I suggest an insured CD, not a developmental stage biotech.
"...the trial is powered such that Cabo needs to show 7.5 month vs Afinitor's 5 months."
Close, but not quite. Lets look again at power and the assumptions that go into it.
A clinical trial is a statistical sampling of what exists in the real world. Meteor is looking at a sample of 375 RCC patients out of many thousands and trying to answer a yes/no question, "is cabo PFS superior to everolimus PFS?" That yes/no is determined by applying a statistical test to the results of the trial and determining whether the probability of superiority based on those results meets a predetermined threshold, usually 95%. Two things affect that result, the margin of superiority (often represented as a comparison of medians) and size of the sample. When sponsors design a trial there are several considerations, large trials are more expensive then small trials but the larger the trial the more likely it is that a small advantage will achieve statistical significance. Power is a tool for sizing a trial. It provides a margin in case the drug is a little less effective than assumed and also protects against a poor draw in which through sampling error and bad luck a disproportionate number of poor prognosis patients are on the treatment arm. So 7.5 vs 5.0 is not a minimum threshold of superiority or a goal for the trial. It is the assumed level of efficacy in the real world on which the sample size is based that still provides some downside protection. So here is a word description for the power of the Meteor trial. Assuming that the real world superiority of cabo over ever is 7.5 vs 5.0 there is a 90% probability that the trial will meet its primary endpoint. Looking at it from the other direction...Even if the real world superiority of cabo vs ever is 7.5 vs 5.0 there is still a 10% chance the trial will fail due to a disproportionate number of poor prognosis patients on the treatment arm.
"METEOR is open label, it could be possible that patients move after they start on everolimus."
That still should not be a confounding factor. I would think most would wait until progression to try something else and those who drop out prior to progression are censored at their last observation, so subsequent treatments do not really have much effect on the final analysis.
"So when you say, "So, if the everolimus arm is performing consistent with historic results, then the Cabo arm outperforming the control has contributed at least an additional 2-4 months prolongation to the trigger date", the way I interpret this, and I hope I'm wrong, is that if you are right, then Cabo will only show a 2-4 month greater PFS than everolimus. Is that the conclusion, again assuming your calculations are correct?"
Your asking for a level of certainty that just can't be justified. There are just too many unknowns and variables, the biggest is whether the everolimus arm will repeat what was observed in the phase 3 Record-1 trial. I started the count forward from 5/1/2014. An argument can be made that it should start further in the past as some patients were enrolled several months earlier. We'll see what happens.
"Despite these good #'s & the PDUFA almost a certainty, how would you explain the persistent large short position which is obviously holding down the SP?"
I don't worry much about what other people do with their money. I do suspect a chunk of that short position is offset by convert hedging.
"Do you think that Cabo might achieve its end point or beat by good margin?"
I hope it makes the primary endpoint and that is my expectation. By how much, who knows, I'll be happy with any statsig result. I do not expect the 13 month PFS from the phase one to be repeated. I'm thinking more in terms of a 3-4 month improvement. Even a 1-2 month improvement would likely achieve significance.
"Second what is the historic OS in everolimus treated patients and shouldn't we have to wait a little longer to get OS nos?"
I think it was around 13 months. Meaningful OS data is pretty far into the future
"ncbi.nlm.nih.gov/pmc/articles/PMC4413536/ that looks at Everlomius. This is the most recent one to be reported. And the median PFS was 6.9 months (95% CI, 5–9 months). going with the KM plot seems like 70% patients progressed at 12-13 months. please correct me if I'm wrong."
Some important differences. This was an observational study with no control arm. There was no central blinded reading of the imaging. Progression timing was determined by the local physicians. Uniformly these tend to be more lenient than centrally determined measurements and physicians will give the patients more of a break to keep them on medication. Also the large majority had only one prior treatment. In the everolimus ph 3 the vast majority had more than one prior treatment. The Meteor entry criteria is very similar to the everolimus ph 3 so I expect the patient profile to be similar, but obviously there are no guarantees.
"While we have noted that the event rate has slowed over the last few months we are very close to having achieved the required 259 PFS events and are in the final steps of data collection, data cleaning and source data verification before running the final analysis."
Let's look at this. We have a definitive statement that as of the end of April, 259 events had not occurred, but we are very close to that milestone. Some other things we know. The event trigger is 259 of the first 375 enrolled, which is a rather high 70%. This was done intentionally to capture data from that significant minority of patients who have been shown to do very well on a second VEGF TKI like Cabo. A smaller percentage would capture data only from the early progressors. The full 375 enrolment was reached in June and the median enrolment for the 375 occurred in April 2014. The mean time on trial for the 375 has not been discussed, but as the trial first opened in 2013, I suspect that mean (numeric average) time on trial is somewhat longer than the reported median. From its phase 3 pivotal trial, everolimus patients experienced a median PFS of 5 months. Looking at the KM graph for that trial, 70% of the patients had progressed by 8 months, however the curve had flattened out such that 80% progression extended out to 11.5 months. So, statistically we can expect some variability on duration of treatment before 70% of the everolimus patients would progress. If the Meteor result is consistent with the everolimus pivotal trial, 8 months is likely, but 10 months is still plausible.
So lets back up and conservatively use 5/1/2014 for a starting point. The everolimus arm had 8 months on trial by 12/31/2014 and 10 months on trial by 2/28/2015. So, if the everolimus arm is performing consistent with historic results, then the Cabo arm outperforming the control has contributed at least an additional 2-4 months prolongation to the trigger date.