@moorepd1 tweets, "@adamfeuerstein, your analysis of Delcath from ASCO was incorrect. Go listen to [CEO Eamonn] Hobbs' Jefferies presentation."
Along the same vein, Thomas O. complains, "The Delcath data at ASCO got better, survival is extended and patient lives are being saved, yet you still find ways to trash it. No doubt you're working hard to help your short-selling hedge fund buddies. Shame on you."
The Delcath data presented at ASCO appeared to get better, but it didn't.
Treatment with Delcath's CS procedure delayed the time before liver tumors started to grow or death by 6.5 months compared to patients treated with best alternative care, according to the ASCO presentation by lead investigator Dr. James Pingpank.
That data topped results reported by Delcath in an April 21 press release, which pegged the benefit for CS patients at just under 5 months compared to best alternative care.
Except Delcath pulled a data switcheroo. The results from the CS phase III study announced April 21 were confirmed and verified by an independent lab; while at ASCO, Pingpank presented his own analysis of the data, which was not independently confirmed.
The FDA, of course, only considers independently confirmed data, which means Pingpank's presentation at ASCO (6.5 month benefit) exaggerated the true efficacy of Delcath's CS procedure (about a five month benefit.) The independent data are still fine, but just not as good.
To address Thomas's point about survival, the true answer is no, Delcath's CS procedure did not result in a survival benefit. That's because more than half of the patients receiving best alternative care crossed over once their tumors started to grow and received treatment with CS.
Delcath says that patients randomized to the CS procedure initially reported a median overall survival of 298 days compared to just 124 days for patients treated with best alternative care who did not cross over to receive a CS procedure.
The problem with this survival analysis is that it skews in favor of the healthiest BAC patients likely able to cross over to CS, while under-counting very sick BAC patients who probably couldn't tolerate the CS procedure and therefore didn't cross over.