doubling survival at 1 year is beyond any AML study in the last 20-30 years. They are looking for a single agent to take a huge leap forward, where virtually no drug has been able to take even an incremental step forward. daco improved survival across the curve by 18 percent and about the same at 1 year. And is used off label in the US and was approved in EU.
Is it worthwhile? sure... but perhaps they should consider combos to achieve that result. Or lower the 'doubling of survival' criteria.
Any drug that hits the goal of doubling of survival at one year is probably approvable in an accelerated way (AA). I'm guessing no single agents have met that criteria yet.
From what I can read, again, daco no where near doubled 1 year survival in its phase 3.
Control arm had 55 alive and active arm had 65. And remember, dac approved in EU and used off label for AML in US.
Doubling 1 year survival of cytarabine is a home run. Perhaps this study is seeking a single agent 'AML home run', rather than perhaps a combo of drugs to do the same.
My bad... 1 year survival is what was required to be doubled. Not mOS. Still a relatively high bar. Will see if can find 1 year survival stats.
Sapa did not make the cut for this study.
The company PR'd that in early 2013.
Can find the Pr on Cycc's web site dated Jan 7 2013.
"...Recently, the study's Data Monitoring & Ethics Committee did not recommend sapacitabine for the next phase of the LI-1 Trial based on the criterion that sapacitabine as a single agent was not likely to double one year survival versus low-dose cytarabine..."
Recall this would be pretty much what daco was up against in its phase 3 (cytarabine) and had mOS = 7.7 v 5. Nowhere near double.
Doubling mOS is a pretty high bar for a single agent in AML.
did do some comparing.
each set of cohorts had 3 arms. two arms kept the same dosage ; one arm changed between the two sets of cohorts going from a higher dose, which did fairly well in the first study relative to other arms to a lower dose which didn't do as well.
They appeared to be a little more selective. first group: "84% were scored as intermediate-2 or high risk" but that was an entrance criteria to the 2nd study.
blast count was limited to 20% in the 2nd study where it did not appear to be limited in the first study.
The first study had about 10 patients that also used revlimid in addition to an HMA which doesn't appear to be the case in the 2nd study.
The selection criteria helped mOS by almost 2 months in the two arms that were repeated. Don't know if cycc feels they know why... but with the solid mOS result, likely will keep the entrance criteria for phase 3.
Both studies were age 60 and up.
they ran about 60 patients through the first time and similar number the 2nd time. Going to check to see if the tweaked the dosage or entrance criteria. mOS did improve.
Did Spiro or Judy ever mention the motivation? They have run 120 patients through MDS phase 2 studies now.
mOS = median overall survival
BSC = best supportive care
HR = hazard ratio (eg .86 would imply a 14% chance LOWER likelihood of measured event - for SEAMLESS the event is death)
p value is a value that indicates how likely the HR value is NOT correct. You want a low p value
great weekend to all
my reply disappeared.
agree that a possibility is a control arm other than bsc. ontx study could still affect analysis...
du: i believe anyone who stops receiving treatment for cancer directly is pretty much on bsc. A patient not receiving bsc is one either being actively treated or somehow taking drugs to lower their life expectancy (euthanasia for example).
google best supportive care and there are some good efinitions. Key word i see being used is anyone NOT receiving antineoplastic agents. (ie cancer fighting agents)
one more quick point. And Adam F needs to publish an article on this. mOS does not equal overall survival benefit. Its an indicator of benefit for sure...
The key to the study success is OS throughout the whole study. The HR value with a low enough p value. He scans topline mOS and draws a conclusion.
My view of the ONTX study is the mOS was more than enough to gain approval. Had the K-M curve continued to show that same improvement throughout the study, the HR value would have been much lower. We'll likely see that when data is publsihed, what has to be the case is the control arm gained on RIG in terms of percentage of patients still alive later in the study.
At the risk of being redundant, Celgene's pancreatic cancer study first reported topline mOS improvement of 1.7 months and had HR value of .72. So less mOS with a very solid HR. Rig had 2.4 month mOS with .86 HR.
l2l: good post. i think this highlights where sapa might be able to surprise. I think this study is likely going to be similar to RIG/ONTX. An HR of .86 with a pretty decent mOS advantage tells me once the patient failed RIG, they didn't do well quickly. While the BSC arm was steady. My belief is we see a curve like daco v low dose cytarabine. Where daco had a decent mOS advantage, but the k-m curves converged rather than diverged after mOS.
Sapa could have a more modest mOS advantage, but with a divergent k-m curve, could post an approvable HR and p value. If not in the US, then in EU.
Plugging the presentation slide data comparing sapa + dac in the pilot study versus dac in their phase 3 study into a HR calculator and making some conservative assumptions, yielded an HR of ~ .86. And that is comparing dac study with 1/4 of the patients under 70 who lived 2 months longer on average than the others.
If the study yields an HR value under .8, while not reaching the SPA, I think CYCC would still submit for approval in EU + US. And why not , as daco is being used in the US and was approved in EU with HR of .82 over low dose cytarabine.
Monday will be 1 month (4 weeks) since Spiro threw out the nugget of registration pathway announcement.
Others may say no, but I think the ONTX data release could have CYCC tweaking parameters. I'm guessing CYCC modeled lower bsc survival. So it could have caused some level of statistical re-eval. Which could take a little longer, but is wise given knowing a little bit more from ONTX study.
I could be wrong :)
Q4 cc is just 2+ weeks away. Maybe the question will be raised.
l2l: its groundless?
Its the most recent set of data known for RMDS BSC. Its patients being classified as higher risk. I agree, we don't have detail data, but it certainly isn't groundless. There's also more than blast count that defines higher risk... why did CYCC exclude 20-30 and RIG include?
I don't know... its a difference... And why you can't draw 100% confident conclusions...
Had RIG had more patients, 2.5 months of mOS might have been alot more statistically significant. They expected 4 months and designed their study accordingly. I'm saying its a good lesson... Studies often underestimate the control arm, as much as overestimating the active arm.
yahoo mbs stink :) Doesn't like less than signs... and truncated my message.
In reviewing the ontx study, it does appear blast counts were allowed to be higher in the RIG p3 than in the sapa p2 study. They both seemed to be targeting what is considered medium to high risk MDS patients, but sapa study limited blast count to less than 20% where I see RIG study allowing patients with blast counts up to 30% (do others see that too?)
Makes me think sapa in the same set of patients might not do as well as the latest p2 data might indicate. Perhaps performing closer to the RIG outcome at least as far as mOS.
Which leads me back to study design. RIG had a nice 2.5 month advantage on mOS of 5.8 months for BSC. That's pretty solid. It seems probable that the study wasn't powered enough and also probably that the K-M curve was not diverging past mOS.
The hurdle for CYCC got a little tougher imo for their mds p3. But without the results of the ONTX study, they could easily have designed a trial that wasn't powered enough (like ONTX). CYCC was quoting that BSC for patients was likely ~ 4 months and now they see it fall in the 5.5-6 months. They need a trial design that has enough patients to beat 6 months (IMO).
The bar is higher, but having a lot better idea of where the bar is should help cycc design a study that gives sapa a good chance of succeeding in mds.
In reviewing the ontx study, it does appear blast counts were allowed to be higher in the RIG p3 than in the sapa p2 study. They both seemed to be targeting what is considered medium to high risk MDS patients, but sapa study limited blast count to
l2l and onc: from the PR, they are already planning a trial targeting low risk MDS.
Quoting the PR...
"...Ongoing efforts include trials of oral rigosertib in transfusion-dependent lower risk MDS patients, where, after consultation with regulatory agencies, we are planning to initiate a Phase 3 trial as soon as possible," continued Dr. Kumar..."
you r right... "...who had progressed on, failed or relapsed after prior therapy with hypomethylating agents (HMAs)..." is the way the PR describes the study.
ONTX dug into the data removing those who relapsed after HMAs.
Either way, the trial targeted patients who had tried a frontline HMA (daco or vidaza) where that treatment had succeeded or immediately failed.
ONTX seems to try to say their drug works when the HMA quickly fails in the front line. I guess implying their drug doesn't work if there is prolonged exposure to an HMA in the front line and then perhaps a lag in treatment and a later relapse.
Another good point for sapa who had patients who tried and failed both Vidaza and daco in their p2 study.
ONTX could have a point, but retrospective analysis probably won't aid in approval. Might aid in a redesigned p2 and follow on p3 study.
ontx had 2 studies fail since their IPO. But they did sell off before that (October).
In Dec, they had a pancreatic cancer study halt after first set of patients in a phase 3 study were analyzed for futility. They did start to recover
ONTX has a huge cash pile. Think they are only 70M net of cash for market cap. Probably higher than CYCC because of a chunk of their stock is held by Baxter and likely not turning over.
it was not front line.
Quoting the PR... "...in patients who had progressed on or failed prior treatment with hypomethylating agents..."
Similar to CYCCs MDS phase 2 for sapa. I didn't see an age criteria in the ONTX study. CYCC study had age 60 and up.
with ONTX failure, it would seem they've provided a blueprint for a successful sapa p3 study.
Their trial randomized about 300 patients 2:1. Seems like you could add ~ 50-75 patients to each arm and run the same trial with a slightly different ratio. Or add 100 to the sapa arm and 50 to the control arm and keep the randomization 2:1.
The ONTX trial was run under a SPA, so FDA would likely accept it pretty readily.
Knowing how well BSC performed over the past 12-24 months is a huge advantage for CYCC. What apparently was the biggest detriment to the study's success was significantly underestimating BSC. Rig performed about as expected.