Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Has there been any hold not lifted by fda? Way too many false alarms!
Azd9291 also had 5 ILD one of which verified. I will prefer metformin to icu/steroid/ventilator any time.
Igf1r getting hit is concerning though. Not much positive data exist when both EGFR and IGF1R were inhibited.
Hindsight is 20-20. No one has a crystal ball. You can bet there is a significant overlap in the advisors who helped Genentech and Arqule or Daiichi design their trials. Biotech investing is high risk and it's convenient to blame CEO, although I wonder the Medical executive should take on their fair shares.
From the crizotinib presentation today met amplification rate was reported to be 3.8%. So you are right it's a niche.
I also visited the attention poster yesterday. HGF level instead of met was shown as the subgroup that may be predictive of tivantinib activity in second-line. Not consistent with Marquee.
Correct MetMAB should have met tested via FISH, or small TKI works better than antibodies (including the HGF ab per the ficlatuzumab example) in NSCLC.
As stated in my very first post in this thread, the crizotinib data lend confidence that cabozantinib or met inhibitor as a class may have a role in met gene amplified NSCLC. Obviously it's a niche market, erlotinib generated close to half a billion dollar sale in for the mere10-15% of EGFR mutant lung pts. The frequency of met amplification in NSCLC only may be just as large as EGFR mutations. In addition the potential applicability for met is not limited to lung but prostate, HCC and RCC. So there is case for optimism here. What's unique about cabo is that it attacks more than just met but also vegfr2, ret, and axl.
Disclaimer I own both arql and exel and together they explain a huge huge chunk of of paper loss carnage my account has suffered.
Your failing to see the agreement and insisting on people agreeing 100% with a position which you have dug your feet is a common I hub symptom which only a few master posters here manage to avoid. I agree that met hi is a potential predictive bio marker for tivantinib based on MARQUEE. I plan to check out the ATTENTION poster/talk to see whether similar data will be revealed. I hope it does and if you still don't see the agreement then I don't know what else to say.
The micro tubule aspect of the MOA is nice but doesn't explain the EGFR mutant subgroup findings. EGFR mutation is not a bio marker for chemo that affects similar mitotic spindle such as taxane or vinca alkaloids for sure. We will just have to agree to disagree here.
We agree on the overall interpretation that MARQUEE and ATTENTION data were similar as both demonstrated a moderate trend in PFS but neither showed any survival benefit. The WILD-type subgroup OS HR of 1 is virtually the same as of .8ish in ATTENTION.
We disagree however on interpretation of the EGFR Mutant subgroup of MARQUEE. Why should tivantinib improve OS in EGFR mutant but not in EGFR wild type if it only works through CMET, and there is no data that suggests CMET is amplified more frequently in EGFR Mutant than in EGFR wild-type tumor? If we believe CMET amplification is independent of EGFR mutation status, then whatever magnitude of benefit attributable to cMET inhibition that moves the needle in mutant subgroup should have done the same in the WT but it didn't.
Therefore if one had to commit a sin of multiple comparison and seriously inflating type I error as in the case of subgroup analyses, I would choose the met hi subgroup over the mutant subgroup. This is fully defensible not only considering the MOA of tivantinib but also that the EGFR MUTANT subgroup is not even close to being stat sig.
Yes you are correct I believe 10-20% at most of the EGFR mutant tumors will be cMET amplified after erlotinib treatment. I believe the rate will be a lot lower for TKI naive EGFR tumors because after all why bother with CMET if it were perfectly happy with having just the activating mutations, an even under EGFR inhibitory pressure with an EGFR TKI, only 20% tumor resorts to cMET vs the majority (50-60%) T790M.
The METMAB failure is actually not that surprising from a development perspective. There are rather significant warning signs in the Met high subgroup analysis of the phase 2 trial: OS was improved despite no difference in PFS, and in addition the metmab subgroup appears to harm survival among met low patients. One has to ask whether the ph3 go decision was rushed.
On the contrary, one can't find many holes in the ARQL development strategy except the company should have focused on the met high subgroup although the MetMaB failure trial did back the comPany up that there was significant uncertainty in type of MetHigh test (IHC vs FISH) hence challenges in applying it to MARQUEE AND ATTENTION prospectively. The company should not get a free pass though. One should still ask whether it did all it could to data mine the phase 2 to make sure the OS Benefit has not been influenced by some imbalalnce in key baseline factors such as more EGFR mutants in the study arm. But if it's hard politically for GENENTCH to do it, it's hard for Arqule as well.
Also for Arqule, I wonder if one can blame bad luck as the metmab trial could have mopped up too many MET HIGH patients.
EGFR mutant subgroup OS HR=.72 in MARQUEE in ESMO 2013
http://files.shareholder.com/downloads/ARQL/0x0x693826/f59dd69a-fcd1-4a02-8a40-93346b731bb8/ESMO%20MARQUEE%20Final%20Sat%20am.pdf
That contrasts with a HR of 1 in the
Non-mutant subgroup. Due to the widely overlapping confidence interval, it would be a stretch to say MARQUEE (90+% Non-mutant) and ATTENTION (100% confirmed Wild type) were different, because after all the OS HR in the latter was >0.8 and not significantly different from 1.
The key difference between the two trials is not so much geography but the exclusion of EGFR MUTANT in the ATTENTION trial.
Aside from data maturity, mutant subgroup data also suffers from the weak biological rationale in that only 10-20% pts are expected to rely on cMET to drive acquired resistance. The only thing that lends comfort is the trial was stratified by EGFR MUTATION status so pts characteristics may be balanced.
However the drawback is that as soon as the focus is shifted to the mutation subgroup, the MET high (twice as big in terms of patient number as the mutant subgroup) will fall apart. Either one or the other but not both IMO.
Attention! ATTENTION ABSTRACT repost.
http://abstracts.asco.org/144/AbstView_144_128631.html
Yes caveat with cross trial comparison. But why do you think MARQUEE and ATTENTION results are different? They both showed stat "sig" improvement in PFS, and a trend of improvement in OS. The only difference would be the latter was terminate early due to ILD, a significant concern for Japanese investigators.
That said neither the cabo trial nor attention was enriched for c-met, and the c-met subgroup. I did read that tivantinib is not as active as cabo or crizotinib inhibiting c-met. They were all second third line- with cabo's being more heavily pretreated. In addition although the cabo trial did include a few EGFR mutant pts vs none in the ATTENtION trial, cabo is not expected to act on them any differently than the wild type. Considering the longer pfs (4.2 mo) shown in the cabo trial v the <3 mo for Tivantinib and erlotinib combined in the ATTENTION trial in predominantly wild type patient population, and higher response rate observed in the city of hope cabo+erl trial than in another tivantinib+erl trial (also reported in ASCO 14) in the EGFR mutant acquired resistance setting, I think cabo may be more potent than tivantinib.
In terms of cabo vs erlotinib, the ECOG Investigators must have found the median of 4 months compelling enough to add a third arm to compare erl and cabo head to head.
The phase 1/2 cabo+erl trial was reported in ASCO 2010 and the result /reference was described in the linked article below:
http://www.tlcr.org/article/view/523/1033#B20
Vegfr tki clearly has a role in ovarian cancer. Cabo does seem to be active there as well
Some more details from the PI regarding her cabo+erl EGFR mutant acquired resistance study, which used a 40mg dose as well. Bet the response rate will be higher if enriched for met-high patients.
http://breakthroughs.cityofhope.org/asco-2014-new-drug-lung-cancer
Actually a 10% response rate and a median PFS of 4.2-mo single-agent efficacy in a molecularly unselected patient population is decent. It compares favorably with erlotinib or arqule's tivantinib- another C-Met inhibitor (albeit weaker)combined with erlotinib see the ATTENTION abstract linked http://abstracts.asco.org/144/AbstView_144_128631.html.
As mentioned in the poster you linked, NCI/ECOG is running a large phase II study of cabo+erl vs cabo vs erl due to report next year.
http://clinicaltrials.gov/ct2/show/NCT01708954?term=cabozantinib+nsclc&rank=4.
Bone mets activity is nice. Just don't lose sight that cabo does affect tumor in soft tissue as well.
Note the withdrawn foretinib and erlotinib study. There was clearly interest from GSK in developing foretinib in lung.
Thanks for posting it. That's the egfrm acquired resistance setting I was referring to. Would be really interesting to see the median pfs.
ASCO 2014 turning point for cabozantinib as a nsclc drug?
If avastin can add 7 months to pfs in 1st line egfr mutant, and crizotinib is associated with 33% response rate in met amplified, what will cabozantinib do considering it also targets vegfr and met? So instead of a niche agent for ret-fusion positive, met amplified, it might have a role in front line egfr mutant in addition to setting of acquired resistance? Furthermore if you were to throw in axl. That's just about 30-40% of adeno nsclc!!!
Should Gsk be kicking itself for returning xl880?
That I agree - ask the statistician.
Lol. Reminds of onty which till this day no one knows heck what spending function the company used. OBF or haybittle peto.
The counter example of course dndn.
If you could share did you assume a conservative alpha spending or poor treatment effect?
That may be the case. David Miller's post gave her the benefit of doubt. He wrote not long ago that management seemed never to have ascribed much chance of success for the interim.
The probabilities you gave for components 1 and 3 should be switched, as most trials were not adequately powered for the interim.
But you are correct I presume she was speaking of the conditional power at final analysis to be 90%.
Why would a conservative alpha chosen you asked. Many reasons. Designed by a different statistician. Prostate cancer more challenging than other tumor types and so more prone to be confounded. Scared by how closely failed the final analysis and had a change of heart that every bit of alpha counts.
Time will tell.
Just to clarify. I am not saying interim analysis affects the overall power. The study was designed to have 90% power with an interim analysis built in and therefore the overall power is fixed at 90%.
What I am pointing out is the CMO remark that 'the full power at the final analysis is 90%', which to me means 0 power was wasted on the interim analysis, hence a very conservative alpha spending at the interim analysis which the CMO characterized as being 'very little' earlier in the call.
I believe these remarks would contradict an OBF boundary which we know will spend at least 20-30% power (at an alpha of 0.013ish), leaving only 60-70% power instead of 90% for the final analysis.
It was odd as one would typically not comment on power allocated to final, unless she was trying to emphasize preceding remark that "vast majority" of alpha was reserved for the final. Obf at the IA with such a high information fraction would preserve a majority but one would not quantify it as vast majority, the spending is small but not very little. Perhaps I am reading too much, it does sound like CMO is making an effort to defend the trial.
Final analysis fully powered at 90%. Highly likely only 0.0001 was spent.
Exel comet1 interim alpha
36 min into the earnings call the CMO said very little alpha was spent. It can't be obf can it?
Correct OBF is fairly common although I wonder there are exceptions. Did anyone ever figure out what group seq boundaries were used for the 50% and 75% IA conducted for the Ph3 Stimuvax START trial? If I am not mistaken 0.001 was used for both.
Let me not forget to say thank you David for the refresher.
Yes it would if a conservative alpha spending function were used for the interim.
OK 0.05 alpha it is - both you and David verified it. Are you sure it's an O'Brien Fleming function? Very little inflation in event goal which is atypical of a hefty alpha spending.
Just curious how do you know alpha is 0.05? Mgt never disclosed it. Did you back calculate from the total event goal, delta, beta and randomization ratio, the info fraction and confirm that's the case?
Any HR of less than 0.8 is still more than decent.
Yeah with an OS hr of 1. Different tumor type and endpoint. Not applicable.
By my calculation the hr could be as low as 0.72 yet still miss the interim.
By 0.75 You are assuming an overall alpha of 0.05. One can't rule oout 0.01 is being used instead.
Good results from those trials could support compendium listing /reimbursement for off label use and increase valuation of ponatinib obviously.
Read the ct gov trial page. It states the trial is enrolling patients per a feb 2014 update. The risk tolerance for lung cancer patients is higher than hemonc patients.
Interesting mention of iclusig fgfr bio marker selected lung indication on yahoo finance msg bd. I do undervalued indication so long as they could circumvent the risk associated with vegfr inhibition. Interesting this lung trial is already recruiting since fda lifted hold.
http://clinical trials.gov/ct2/show/NCT01935336?term=Ponatinib+nsclc&rank=2
Good timing. Temp data cleaner jus posted today. Temp programmer job last month. Suggesting database not yet locked.
Does feel like AMLN déjà vu all over again. The key question is what the peak iclusig sale would be. It appears to me even with a limited label, and the so called risk of thrombus, the significantly longer than expected treatment duration (flat line at 2 years and extrapolable to up to 10 years) will more than offset it being limited to last line or t315i. Good luck!