Writing such a long post and knowing what are events and HR, you must be a biostatistics Ph.D. Give you a big thumb up for enlightening us. By the way, can you write down the ito calculus formula to show us why a big portion of metastatic cRPC patients previously treated with Docetaxel and (Abiraterone or MDV3100) can survive up to 21 months on prednisone. Someone from Norwegian Nobel committee is knocking your door.
Yeah, right, 2 months later, he can short at $30, why short at $4? Just wait, next year he can short at much better price, maybe, $50?
I doubt the ~3 moths benefits showed on healthier group can be fully enjoyed on our two branches. Probably more on placebo branch, less on cabo branch. Neither can add 3 mo. Just my guess.
I'll give you a big thumb down, smart guy. You are in a exceptional minority group, you must be genius -- so says your statistics.
Today can also labeled as the date 11th months after 9/26/2013. 2 months before that, at least 50% had been enrolled, I assume.
Comparing to EXAM, now my bet is a very fat tail. That might be a very distinct characteristics of cabo.
Look at EXAM, at 75% event(50% of population), the estimate of median OS was 26 v 23 something months. I forgot the actual data. Another 26 months passed(which equals the previous estimate of mean), cannot reach 100% of events(which is 66% of whole population).
I guess this distribution curve is very confusing for statisticians. Other scientists must give an explanation for the fat tail later on.
Your question perhaps can be rephrased like:
Assuming prednison branch will get +2 months(say, 8+2) OS benefits from X, how many +months(or weeks) OS benefits will cabo branch patients get from X after progress from cabo?
Comparing to China, all the rest of the world steel plants are midgets. If you are not able to do research on Chinese steel supply and demand, I sincerely suggest you stay away from these companies. P/E is nothing, for material producers, earnings are easily wiped out by price fluctuation.
That was my feeling too, holding up some data for the right time.
But if both Giesela and MMM are frank with disclosing, there must be something very special to account for. On 50% of whole population event count, the interim conclusion of OS is something like 20 mos v 26 mos. That number might be pretty precise estimate. And then, another 26 mos, not 65% yet! Sigma is multitude way above average?
Even the survival curve is gausssian, the waiting period is still too long.
Translate the number of 44% - 75% - 100% into 29% - 50% - 65% based on that scale. From 29% to 50%, it took 14 months. After 50%, that was the peak of bell, should take less than 14mos to 65%. That assumption obviously cannot hold. I tried to fit the numbers in exponential decay curve, with the assumption that the events happened at previous rate over current survival popution, it should be reached at 18mos or less. This assumption still was proved wrong.
Normal hypotheses are all rejected by big margin. So far, 26 months passed. My guess is:
1) if cabo is good for some patients, it is good for very very long time; and for patients didn't respond, they passed away sooner than those on placebo due to 140mg AEs. This is something opposite to bell curve.
2) as cabo becoming available in the US market, almost every one on placebo(it's easy for patients to figure out) moved to cabo. That will make reading of OS very tough.
Before cabo, the total annual sales of the world is $30M. Now cabo got more than $6M a quarter in the US. It may be very expensive to keep vandetanib's manufacturing, sales, medical monitoring, book keeping blah blah... teams. Or will AZN will keep this choice for patients open?
The abstract submission deadline for the ESMO Symposium on Immuno-Oncology 2014 has been extended to Monday 4 August 2014, 12:00 CEST.
They will not submit an abstract if they don't have anything at hand.