Replies to post #53458 on Avid Bioservices Inc (CDMO)

[Thurly]

Entdoc,

The information generated in early trials can provide valuable signals for new trials, agreed. Maybe the trials in India and Georgia were enough. But regardless of the rigor (and I've heard they were rigorous -- inspections every month or two), unblinded trials have a built in bias.

I think that the safety data looks good considering the number of trials and ISTs done to date. And I agree, the biostaticians will pour over those numbers and design the trial that optimizes PIIb's chances for success. Dendreon did something similar, adopting the Cox regression model for Provenge and then shortening the length of the last PIII perfectly. They met their SPA numbers with very little room to spare. But they had two previous double blind randomized placebo control PIIIs to draw from (2nd trial aborted due to reasons unrelated to AEs in the trial). So they had singularly reliable data to use while modifying the trial design. Despite all of the positive signals, despite the fact that the FDA apparently likes what they saw in our data and allowed us to design of a registrational trial, the CI is based on trial data that is inherently uncertain because of the lack of randomized control in the testing environment.

I am excited about the survival numbers (i.e., the lack thereof). Survival can't be faked. Dendreon unfolded similarly. While their survival numbers beat taxotere by 4-5 months, 31% of patients lived longer than 3 years and I think it was 20% or 28% (I'd have to go back and check my notes) that were alive 6 years out.

Chalk me up as a cautious investor -- not pessimistic, as still I see a significant opportunity here -- but it's based on more than the NSCLC trial data we've seen to date. I like management. I understand the ATM and accept the necessity of dilution. AVID's going well and I expect the TMTI HFV contract to be renewed. I like how focused they are in bringing bavi to market. I hope they make it with the PIIb. Personally, I think they're more likely to succeed in anti-viral before oncology (IND Q1-2012). If they don't succeed in the PIIb but do in the anti-viral, with a stockpile order we should still be in great shape to run the full PIII without significant dilution.

[azure107]

entdoc

You said, "In some ways PPHM DID miss statistical significance, by reporting data that was only good at the 05% confidence level."

Do you mean that the uncertainty of the experimental data is limited to 0.05% confidence level? i.e. that 95.95% of the data is significant and under the normal distribution of the gaussian curve?