Avid Bioservices Inc (CDMO)

[Protector]

Sorry, my wrong. I understood it differently.

I will extend on the simulations but I need better models. Thing is we work in the dark and I try to limit the impact of cumulative errors by tuning ranges better or by refining. However, there isn't always data available for that.

Halve way, in the Best case, was 30 days ago, and in the worst case it will be in 60 days.

Now just consider this:

- we know there is competitive enrolment within a same country (at least for Europe). So if hospital A does slow then Hospital B may enrol more. That simplifies things. But if a country is FULLY enrolled the maximum number of authorized patients by the protocol then in the simulation I should do as if ALL centres in that country stop enrolling. MY SIMULATION DOESN'T HAS THAT because we don't know WERE the patients are enrolled. I could add this on a COUNTRY bases (because we know exactly what centre open when and how many patients it could have enrolled under the theoretical assumptions between 0.3/0.5 per month/centre but I ma not going because of the point below who smooths that).

- We know PPHM assigned more then 345 patients to the overall EEA clinical trial (429 and we don't know about France, lets assume that France given its size and population gets at least as much patients as Germany allotted then we would be at 429+43=472 patients or 472-345=137 over assignment).

- We know there are still 26 centres to open but MAYBE, while the protocol doesn't ask for a MAXIMUM, it could be allowed to open LESS centres then the 100 planned but NOT more.

- There is the Greek exception. If the protocol is correct PPHM assumed that in Greece 120 patients can be enrolled in only 9 centres (8 currently open) in 18 months (European Centres were opened as of JUN/JUL/AUG). 120/9=13.3 patients per centre OR 13.3/18=0.74 per centre per month. That is MUCH higher then my best 0.5 and worst 0.3 patients per centre per month.

- We have no idea how the remaining 237 patients are going to be recruited in the US+World. Did they also have over allotment?

- Although by looking at the centres list most of the treatment centres are centres should not have lack of patients we do not know how many 1 or 2nd ln patients to get in. That may be VERY dependent per country because in some countries detection early and well organized in others not. And there is the economical aspect. In Europe A Docetaxel treatment fall under social security so I suppose the only thing not covered will be Bavituximab because it isn't approved yet, but Doce and treatment will be paid back. We know from several protocols that PPHM pays the hospitals for the treatment which should make the trial realy accessible to a large number of people from the economical point of view.

- For the two early look ins we are in a FAR better position now to speculate because the first 11 months of the trial are behind us. BUT while we know that they are event triggered and that there is a data monitoring committee we still have no idea on how many events. We can now try to combine events with time window and get some more precise estimates and we can make the assumptions as if no more centres are going to be opened. If they open more it can only accelerate things.

- Last but not least estimations are estimations. In the best case I assumed centres get 0.5 patients per month/centre and in the worst 0.3 per month/centre. Any less (Eg: 0.25 would bring us to OCT 2015 for full enrolment and PPHM's DEC 2015 would match 0.22 patients per month per centre. That is about 1 patient every 5 months per centre CONSIDERING the currently open centres and the dates they were opened and AGAIN it would mean that PPHM took ZERO margin which I never believe.

So I believe we will be fully enrolled in the range of my best and worst case simulation that I posted here. So possibly between 6 to 10 months ahead of schedule. That will then automatically move DEC 2016 also 6 to 10 months closer as well as both look-ins.