Found this info in an article recently:
The following example illustrates the practical implications of the delayed separation on study statistical power, number of events, and study duration. Figure 5 illustrates the mathematical form of a delayed separation, with a hazard ratio of 0.7 after the separation. The control arm has an exponential survival distribution with median survival of 18 months (red dashed curve). The form of the delayed separation is specified by a hazard ratio function (HR(t), solid grey line) that has value 1.0 for the first 3 months and then decreases linearly between 3 and 6 months to become 0.7 after 6 months. The experimental arm survival distribution (solid blue line) is the result of mathematically blending the control arm survival distribution function and the hazard ratio function. The survival distributions in Figure 5 can be used to illustrate the consequences of delayed separation. The following additional specifications are made: 1:1 random assignment, N = 600, accrual time 18 months, two-sided type I error probability .05, statistical power of 90%, and use of the log-rank statistic. Under proportional hazards (no delayed separation), 331 events are required for final analysis and are projected to occur at 2.85 years. With a delayed separation to 3 months as specified, 331 events would result in a statistical power estimate of only 62.7%. Using simulations that take into account delayed separation, it is found that 90% statistical power requires 510 events and would not be realized until 5.66 years.
I think that you're right, not sure how I feel about this. They won't comment on any discussions with the FDA but have met with them recently I'm assuming.
They're not commenting on if or how they're applying this potential new analysis. That's what the questions on the CC are about right now.
I don't have any research notes but you can find summaries of them by using Google and searching for "analyst coverage NewLink" or going to Twitter and typing in "$NLNK" in the search box. The peak sales multiplier depends on factors such as clinical relevance of the drug, competition, and profit margins.
HyperAcute should be standard-of-care in the adjunct pancreatic cancer setting if they stop the trial for efficacy and I can't think of many other drugs in the clinic that are going for resected pancreatic cancer right now. I believe they said on a conference call that they would have standard profit margins for the drug based on their off-the-shelf manufacturing and potential price tag of around $90,000. They've also said that they don't want to partner with anyone for this drug and want to sell it themselves which can be risky but also potentially lucrative.
That's a good question. My take is that you can multiply the projected revenues for this indication by a standard multiple that many companies receive with a cancer product close to approval. Analysts project around $850 million in peak sales for resected pancreatic cancer with a near 100% probability of approval if the trial is stopped. Multiply this by around 4x and you get close to $3.5 billion market cap or $120 a share. This assumes only HyperAcute pancreas and nothing for other HyperAcute products, IDO inhibitors, or Ebola.
The market will likely price in a higher probability for other HyperAcute products like locally advanced pancreatic cancer, lung, melanoma, renal and thus the value of the stock after an IMPRESS stoppage may be somewhere between $120 and $200 in my opinion.
You're right about no one having inside information. Most individual investors believe that those with more money have more information and to an extent that is true, but not regarding outcomes to clinical trials. That's why investors buy when stocks go up and sell when they go down. Most of the time the smart money is on the other side of that trade. In regards to your IMPRESS odds, I believe it's closer to 60:40 for success at this interim and around a 20:1 payoff for the April $60's, but it's anyone's guess.
If NewLink has a drug that has a greater than 12 month improvement in survival and that's not good enough for a stoppage then they've designed one of the worst trials of all time
It's difficult for me to understand the controls contributing less than 4 events per month right now but I'll admit I don't have a spreadsheet or any algorithms set up. So I guess it's possible that we don't actually see the 2nd interim in the 1st quarter despite what the company is projecting. They've been off before.
That makes sense but I guess I'm asking why it's taking almost a year in between interims? It seems to me that if there were 6 months between the 1st and 2nd interim (18.5 events per month) then the average survival wouldn't have changed. At 9 months between interims (12.3 events per month) then there begins to be a difference between survival with enrollment early in the trial vs. later. At 11 months between interims so far (10 events per month), I can't wrap my head around why it's taking so long. Enrollment has been at least 20 patient per month since around the beginning of 2012 so even if the control group are the only ones getting us closer to 333, that would the entire 10 events per month. So I guess my question is 'Are the controls the only ones dying?'
Can someone help me understand why the 2nd interim is taking so much longer than expected? If you calculate the average overall survival after the 1st interim, with no survival benefit between groups it was about 27 months (222 enrolled around Dec 2011 and 222 events around Mar 2014). However, if we had the 2nd interim tomorrow it would now be around 33 months combined survival (333 enrolled around May 2012 and 333 events around Feb 2015). I would think that the combined survival numbers would remain consistent throughout the trial but that isn't happening. NewLink has in the past even suggested that the 2nd interim would happen around 6-9 months after the 1st, but it's been almost 1 year. They've also recently said both the 2nd interim and final (if nec) analysis would take place this year but I'm now questioning this.
I know that enrollment picking up over time has a little to do with it but that wouldn't explain a 6 month difference in overall survival between the 2 interim analysis. I can honestly say that this is messing with my projections but if you assume that the control group isn't the one changing then it's good news with the potential for a 2nd look stoppage.
That's a good explanation. You could look at it by saying that if no treatment group patient had died by the 222 event, it's still possible that the trial would have continued due to the limited survival benefit of treatment vs. control if the control OS was close to the average follow-up time for all patients in the trial. I'm not saying that this happened obviously, but it could explain why we could have a big survival difference between groups and have the trial continue at 222 but stop at 333.
I don't think anyone would want to risk going to jail to save their friends a few bucks by leaking information. And even if they would, I'm pretty sure they're not going to send you an email response telling you that they have done so.
Since enrollment picked up over time, I would imagine that the both groups death rates would be going up over time, not down. At some point it the rate would hit a plateau but I would imagine it would be close to the MOS + the end of enrollment. So perhaps 22-24 months (control) + Sept. 2013 which would be around Sept. 2015. Maybe I'm missing something.
They've been enrolling at least 20 people per month in their trial since around Sept. 2011 and around 25 per month near the end of enrollment on Sept. 2013. This means that it has been 40 months since those enrolled in Sept. 2011 (160 total at the time).
Based on 222 events in Feb. 2014 and probably 333 in Jan. 2015 the rate of deaths is just under 10 per month. Assuming that half of these patients enrolled is the control group I can't see how the rate of deaths is so low. To me, it should be closer to 12-14 deaths per month. I guess at least one of the groups (guess which one) is living way longer than anyone has ever seen in this patient population.