Avid Bioservices Inc (CDMO)

[Protector]

Correcting Abusive Interpretation of PPHM PRs

There was a lot to do, since end Feb begin March 2016 when Sunrise was stoped early about what such early stop meant.

In the FEB+MAR PR in which PPHM announced that Sunrise was stopped early based on a futility advice by the DMC, CEO King said this, based on a first look-in in a FRACTION of the data:

While we are deeply disappointed by this early outcome from the SUNRISE trial, we plan to take a deliberate and detailed approach in reviewing and verifying all available data from the trial in order to understand what subgroups or other patient characteristics may have impacted the performance of the study. While we perform this analysis, we plan to put our other chemotherapy combination studies on hold until we have a clear understanding of the SUNRISE study results."

Of course CEO King says this WITHOUT ANY VIEW on the Sunrise data, that still had to be unblinded, but just based on the fact that the GUIDELINES that PPHM gave to the DMC triggered and resulted in a STOP advice. At the best, since unblinding had to take place yet, they had some early largely incomplete overall survival median tables.

Out of last weeks PR we learn that CEO Kings statement was prudent but premature. The Sunrise trial resulted in a statistical significant improvement of overall survival (first end-point of the trial) while in FEB/MARCH CEO King did NOT ANTICIPATE such outcome in his statement.

As the text above shows, CEO King at that time and based on the limited data PPHM had available, was under the impression that a sub-group caused the triggering of the stop advice. Probably, and maybe we will now officially one day, PPHM has given a number of guidelines to the DMC saying that:

- when you see this...then tell us to stop
- when you see that...then tell us to stop to
- ...

And at least several of them must have had pre-defined sub-group related definitions otherwise CEO King would not by default assume it was such sub-group that must have cause the triggering of the DMC advice.

So in his mind he seemed to think that a specific sub-group triggered the DMC advice and he apparently did not start from the assumption that the complete trial was the cause. It PROVES that sub-groups were in any case defined UP-FRONT at trial design, just as Shan said before on a Q/CC and were approved by the FDA.

For some reason some now try to associate the word SUB-GROUP in the above PPHM PR fragment with an INVENTED/ASSUMED sub-group that would be at the basis of the statement that bavituximab showed statistical significant improvements in overall survival over the placebo control arm.

THE TWO HAVE NOTHING TO DO WITH EACH OTHER

The last PR of last week announces statistical significant results, something deemed impossible when I posted about the possibility over the past months, but does NOT in ANY DIRECT WAY or ANY INDIRECT WAY point to sub-groups. The blooper that the PR was talking about sub-groups being statistical significant improvements of overall survival is something that came alive on this board and started to live its own live UNTIL some went back to the original PPHM PR text and saw that there is no mention nor reason to believe that the statement was about sub-groups.

That doesn't exclude that the provisions that Dr. Garnick had included in the clinical trial in relation to PPHM's healthy fear that there might again be outperforming control arms, as observed in several other clinical trial and one in particular from PPHM itself - 1st ln NSCLC, may have resulted in censoring patients with 3rd line treatment or if censoring isn't the appropriate term here then say 'excluded' from the median tables on pre-negotiated terms with the FDA in relation to a specific and well defined up-front concern.

I think we all agree that the FDA, the Industry and everybody else involved wants to measure the performance of the drug and not of any treatment given after progression. So it might be that the sub-group about which CEO King is talking are the patients that had a 3rd ln treatment. Purely in terms of math theory one could say that if you exclude these that technically speaking the remaining patients are kind of a sub-group.

In that stretch the PR statement should be something like (RED text added):

Presented data will include a biomarker in the SUNRISE trial that correlated with a statistically significant improvement in overall survival for patients treated with bavituximab in combination with docetaxel compared to patients treated with docetaxel alone, after patients that received a 3rd line treatment were removed from the results. Peregrine will file a new patent application directed to the use of the biomarker prior to the presentation of results at the ESMO 2016 Congress, which is being held October 7-11, 2016 in Copenhagen, Denmark.

But still then it changes NOTHING! That is exactly what Sunrise wanted to measure, namely if Bavituximab+Docetaxel could improve on Bavituximab+Placebo. Sunrise did NOT want to measure if 2 treatments (1st ln and then 2nd ln with Bavi+Doce) could improve on 3 consecutive treatment (1st ln, 2nd ln and 3rd ln). And if the above reasoning is correct then it becomes clear what sub-groups as well CEO King in the text above and Shan in the Q/CC were talking about. These must have been the provision in the Sunrise trial protocol that Dr. Garnick had approved by the FDA in view of the possibility that the control arm would outperform (as PPHM EXPLICITLY COMMUNICATED).

There must be a SAME type of provision in relation to a change of SOC during the Sunrise trial. It didn't happen and Docetaxel is still the SOC, only a part of PD-L1 responders can be steered to Opdivo who has a companion test. But still, it proves that PPHM has pro-actively examined every scenario they could think of and included counter measures in the clinical trial protocol and had them approved by the FDA in case they would occur.

So the current state is that THE LAST COMMUNICATION of last week by PPHM about the results of Sunrise are that they have a statistical significant overall survival (Sunrise first end-point) for patients taking Bavituximab+Docetaxel vs patients taking Placebo+Docetaxel and that they ALSO managed to correlate this with a biomarker that will allow them in the future to early detect if bavituximab works.

This is not UNEXPECTED because CEO King ALSO said in the FEB/MARCH PR that Bavituximab performed as expected by the trial design and that it was a dramatically outperforming control arm that cause the triggering on the LIMITED 1st look-in data set.

MORE IMPORTANTLY : Last weeks PR statements are based on UNBLINDED and studied SUNRISE data vs CEO King's statement at the top of this post in MAR 2016 that was based on a summary letter of the DMC with very little data BECAUSE the DMC must make sure they do NOT communicate any data that would lead a defacto unblinding should PPHM (the sponsor) decide to IGNORE the advice.


PPHM took thousands of patients samples

I want to start by thanking the patients, investigators and scientists involved in the SUNRISE trial that have made possible the continuing collection and analysis of important data from the study. While the current interim analysis is still ongoing, it is exciting to already see that a biomarker associated with positive outcomes for patients receiving the combination of bavituximab with docetaxel has been identified.


The primary goal of the biomarker analysis is to identify a biomarker profile for patients that receive the most benefit from a bavituximab-containing therapeutic regimen. As specified in the study protocol, thousands of patient samples were collected to potentially identify biomarkers associated with improved outcome for patients receiving bavituximab.

As for those that claimed that PPHM never said that they took thousands of samples, well read again above. The FDA requires 5000 samples and PPHM cannot claim a biomarker correlation in non-compliance with FDA requirements. Seeing a biomarker in 500 samples is NOT conclusive, neither in 1000 because the FDA says 5000 minimum.