NorthWest Biotherapeutics Inc (NWBO)

[gpb]

First, I'll say that like you I am discomforted by the lack of update on enrollment/site-activation. There are, however, factors outside the company's control which may be responsible for this, and it could (not saying this is probable, but it is possible) end up irrelevant with superb interim analysis which ends blinding on ethical grounds.

Second, I am slightly discomforted by the lack of dcvax-direct update. However, this could just be due to prudence and rigor on the company's part. Ensuring that all measurements of regression are unambiguous, finding out if the effect is dose-dependent before announcing it (because this will be the very first question posed by everybody from the investigators to the regulators to the shareholders), that sort of thing. At least we will get an update relatively soon when phase I dose-finding is sufficient to get started on the already-approved phase II. Or, heaven help us, if there is futility at every dose they'll have to announce the preclusion of phase II while they try and figure out what went wrong. Either way, based on the established timeline and relative sizes for the two trials, an update of some sort cannot be far off.

Third and most importantly, let's talk about AF. I'm inferring that you were referring specifically to his 12/17 article. If so, the core premise of his opinion is that all Dendritic Cell therapies will fail because they are all the same (and ICT-107 failed). However, they are not the same, as has been pointed out numerous times. The differences fall into two categories - intrinsic differences in the vaccine itself, and the relationship between the most recent trial and its predecessors.

With regard to the vaccine itself, here are three key differences - whether they work out in NWBO's favor is yet to be proven, but I'll address that too in a moment.
1) DCVax-L DCs are trained to recognize the full set of antigens expressed by the tumor cells. ICT-107 DCs are only trained to recognize six of the antigens possibly expressed by the tumor cells.
2) DCVax-L DCs are trained on the patient's own unique tumor, whereas the six antigens involved in ICT-107 don't change regardless of whether or not they are expressed by the patient's tumor.
3) DCVax-L DCs, when being trained, are exposed to the whole antigen as expressed by the tumor. ICT-107 uses peptide fragments of the antigen which are long enough to correspond to a particular antigen, but do not necessarily give the DC as much to work with when recognizing the antigen and distinguishing it from all the other cellular expressions. This is analogous to the blind date setup. If I tell you "I'm an adult white male" and I happen to be the only white guy in the restaurant, you could recognize me, or you could be stuck in a nightmare of "where's waldo" proportions. However, you will recognize me much more easily and robustly if I also tell you "I'll be the one wearing a rose on my lapel and a brown fedora on my head while reading today's edition of the wall street journal."

Now, about whether those differences are likely to favor NWBO, rather than cite any biased speculation from the company or its investors, I'll cite a scientific study by independent investigators who compared two Dendritic Cell vaccines - one based on autologous whole tumor lysate and the other based on a limited set of peptides - the results of which were presented at the American Association of Neurological Surgeons 79th annual meeting:
Vaccine Made of Glioblastoma Lysate Prolongs Survival
You really should read that whole article if you haven't already (John posted it a few days ago but it's worth repeating). The study data unambiguously shows superiority of the autologous whole tumor lysate vaccine (the vaccine more like DCVax-L) over the pre-selected peptide fragment vaccine.

Finally, about the trial histories:
1) NWBO's past experience with DCVax-L in GBM before the start of Phase III was slightly greater than IMUC's past experience with ICT-107 in GBM before the start of Phase II. The magnificent early phase results were based on a large sample population.
2) NWBO's past experience with DCVax-L in all indications before the start of Phase III was much greater than IMUC's past experience with ICT-107 in all indications before the start of Phase II.
3) DCVax-L's earlier results were much more internally consistent. ICT-107 showed impressive median OS, however, the increase in PFS was nowhere near proportional. In fact, PFS for ICT-107 was WORSE (by a small amount) than the PFS shown in a similar trial by AGEN in three times as many patients (for which median OS was still rising last I checked but was currently much lower than either DCVax-L or ICT-107). I viewed this lack of agreement between two measurements which should be and normally are highly correlated as an early red flag for ICT-107's follow-up performance. In contrast, while DCVax-L had almost identical median OS to ICT-107 (with 25% greater sample size), it ALSO had a correspondingly excellent median PFS (9.5 months or 56% higher than ICT-107). Regardless of whether or not the full extent of this benefit over SoC is maintained during Phase III (could be, the available dosing schedule is longer after all), at least DCVax-L's early results actually "made sense" with regard to internal consistency.