Impressive DD you have done. The youtube Posadas snip about cabozantinib on liver mets is a must watch. that said, I think the early signal/activity on bone mets is what led the company to focus on bone-only population. But hey that's why one needs to pay attention to negative phase III trials and try to understand signal in subgroups, which is what Exel has done. I would say despite abi or enza in mCRPC there is still a tangible subset of patients that do not respond, in particular the kind that does not pump out PSA, as alluded to by Dr. Posadas.
Thanks so much Wildbiftek for sharing screenshot from the 2016 ASCO GU poster. I see how you get 0.52 hazard ratio for intermediate and poor risk subgroup, and I agree with the 0.73 extrapolated hazard ratio, which is clearly clinically meaningful.
From the prostate cancer debacle (which btw is not a total loss), based on how co handled E1512 and METEOR, EXEL mgt has learned it's better to underpromise but overdeliver to either avoid liability or a flat-out buyout. All speculations of course.
EMA tends to look upon data more favorably but it really depends on the magnitude of p-value and HR. ALLIANCE has good clinical trial infrastructure in place, but it will still take time to collect CAT scans and retrospectively review.
Sorry I missed the updated result. Was the poor risk and intermediate risk subgroup data also updated? The longer tail may be limited to the favorable risk subgroup
Agreed on the algorithm, but where did you get the HR of 0.52? The hazard ratio from NETEOR is 0.58 overall, 0.56 for intermediate risk subgroup, and 0.84 for poor risk subgroup, respectively.
E1512 PR messaging is a relevant comparison. It sounds as neutral if not slightly less bullish actually compared to the Cabosun announcement without commenting on whether the PFS improvement is clinically meaningful. Whatsmore it was muted about the strong survival trend (although it could simply have not been available). So I guess the million dollar question is what's a clinically meaningful PFS difference? From this FDA Axitinib ODAC, reading btw the line, FDA seems to have pinned it at two-months for a setting with low control median PFS.
Quote - Study E1512 Met Primary Endpoint, Significantly Improving Progression-Free Survival in the Cabozantinib Monotherapy and Cabozantinib/Erlotinib Arms Compared to the Erlotinib Arm
I would also subscribe the hypothesis that company leadership were caught by surprise by the CABOSUN result. Their statistician probably had advised at the design stage the true hazard ratio may be bigger than 0.8, hence Cabosun would barely miss the one-sided alpha of 0.12, by the HRs from RECORD-3 (1.4) and METEOR HR (0.58). Time will tell. The question is if the hazard ratio is much better than what historical data suggested, then why? Front-line patients being better able to tolerate hence longer treatment duration? Time will tell.
I trust your read of corporate psychology to be more accurate than mine. The key word to me is "clinically meaningful" in addition to "stat sig". HR
All good points. For CABOSUN to be registrational enabling, it must have a very nice p-value (which is possible); and additionally independent radiological review, which I assume is not part of the ALLIANCE trial design and would be difficult to retrospectively (not entirely unfeasible) collect.
Same here I got a 0.7 HR for a 2-sided p-value of 0.05 too. I would even guess that the HR is at least if not better than 0.7 because the company has clearly stated it to be both statistically and clinically meaningful. An ESMO late breaker sounds very reasonable too.
Well said. EXELIXIS characterized the data as "clinically meaningful improvement". A one-sided p-value of 0.12 equates about approx 0.8 hazard ratio, meaning 25% improvement. With a median PFS of approx 6-8 mo for the intermediate/poor risk for sunitinib, the median PFS for cabo would be 7.5-10 mo. Not trivial but certainly debatable whether a 1.5-2 mo mPFS increase is a meaningful difference, especially if the HR is only 0.8.
Quote from EXELIXIS PR The trial met its primary endpoint, demonstrating a statistically significant and clinically meaningful improvement in progression-free survival (PFS) for cabozantinib compared with sunitinib in patients with advanced intermediate- or poor-risk RCC.