NEW YORK (TheStreet) -- NewLink Genetics (NLNK - Get Report) and Northwest Biotherapeutics (NWBO) announced results from interim analyses of their respective phase III cancer studies Friday. Both studies continue as planned. Good news? Not really. Let's discuss.
The NewLink phase III study of its HyperAcute Pancreas (HAP) vaccine is designed with overall survival as the primary endpoint. The first interim analysis was conducted after the 222nd patient death was reported, or half the number of deaths needed to conduct the final analysis.
To stop the study successfully at the first interim analysis, HAP needed to demonstrate a 40-45% improvement in median overall survival compared to placebo. As with all interim analysis, the statistical bar was set high.
NewLink announced Friday that the HAP pancreatic cancer study would continue unchanged, based on a review of patient data by independent monitors. Onward to the second interim analysis, planned for 330 deaths.
"As we have previously emphasized, continuation of this study was an anticipated outcome considering the high statistical threshold assigned to this first interim analysis under the special protocol assessment," said NewLink CEO Charles Link, in a statement.
Really? Sounds like a lot of spin. Investors who have met with NewLink (and talked to me) say management was much more confident in HAP demonstrating a statistically significant survival benefit at the first interim than they let on today. Remember, NewLink originally told investors the first interim analysis would take place last summer. The delay (patients are taking longer to die!) suggested HAP was working because control arm patients couldn't be living longer, or so the NewLink bulls claimed.
The math says otherwise.
In January, NewLink told investors that the median overall survival for all patients in the study -- both arms -- was 25 months. It's now March, so let's assume the blended median overall survival is 27 months.
NewLink says the phase III study was designed with the assumption that pancreatic cancer patients enrolled into the control arm of the study would have a median overall survival of 17 months.
If that's true, then it suggests median overall survival for HAP-treated patients should be around 37 months, or more than double the control arm and enough to stop the study early for positive efficacy.
Well, that didn't happen, so NewLink's assumption about the median overall survival of pancreatic cancer patients must be wrong.
Using more conservative assumptions doesn't help NewLink either. Let's say pancreatic cancer patients in the control arm should live about 20 months. Using the same blended median 27-month overall survival, this would imply HAP median overall survival of 34 months, or a 70% improvement.
Again, the study should be stopped early for positive efficacy. It wasn't.
None of these back-of-envelope analyses guarantees a HAP study failure, but they do take a serious whack at the optimism of NewLink bulls because the underlying assumptions used to design the study are incorrect. The pancreatic cancer patients in the study are living longer than the company expected, which bodes poorly for the ultimate outcome.
Moving on to NorthWest Biotherapeutics and the phase III study of its brain cancer vaccine DCVax. I read today's press release and thought it must be a joke. Here's what the company said:
Northwest Biotherapeutics, a biotechnology company developing DCVax personalized immune therapies for solid tumor cancers, announced today, in response to shareholder inquiries, that the Data Safety Monitoring Board (DSMB) has made an unblinded review of the safety data for the Company's ongoing international Phase III GBM Trial, and has recommended that the trial continue as planned. The DSMB's review of the efficacy data is still pending. [Emphasis mine.]
Wow, there's a lot of bamboozling in that statement. Let's work through it all.
"In response to shareholder inquiries." No! Northwest Bio issued a press release on Dec. 10 informing investors that the progression-event trigger had been reached for the first interim analysis.
The analysis was supposed to take 6-8 weeks, which stretched to 8-10 weeks, and then almost 13 weeks.
"Review of the safety data." What? The interim analysis was based on efficacy data, not safety. In this case, independent monitors are looking at patient scans to determine when their brain tumors progress.
"The DSMB's review of the efficacy data is still pending." Oh my. How do independent monitors analyze patient scans for safety but not efficacy? Impossible. Also, how can Northwest Bio claim the trigger event for the first interim analysis was reached last December based on the required number of progression events, but tell us today that the review of those scans is still pending?
Does not compute.
What's likely going on here? At best, Northwest Bio has so botched the DCVax study given all the delays and changes made over the years that patients have been lost and their data (scans, etc.) are unreliable. Worst case, DCVax is just a placebo and today's press release is another stalling tactic.
By the way, Northwest Bio still hasn't told investors if the DCVax study is fully enrolled, nor has the company explained how or why FDA will be copacetic with a primary endpoint of progression-free survival.
Follow Adam Feuerstein on Twitter.