NorthWest Biotherapeutics Inc (NWBO)

[flipper44]

Most will likely find this disjointed. Newbies may find it lacking in explanation regarding how I get from A to B. Regardless, most of it is information you already know, although perhaps presented from a slightly different lens.

IMHO

This may seem alike a strange place to begin, but I'd first ask you to look at one of the most important immunotherapy results in unmethylated GBM.



Unmethylated MGMT tumors are notoriously hard to treat, because the tumor's repair mechanism is intact, and thus even attacks on tumor cells are normally repaired before they can destroy a tumor cell.

To me, that graph is very critical in understanding dendritic immunotherapy. Let me try to give a very brief description. It is a progression free survival curve, based upon a specific subtype of patients. The important thing to understand here is that these were patients with unmethylated MGMT tumors that were thought targeted to respond to dendritic cells pulsed with six different antigens associated with glioblastoma. Unfortunately, it is a very narrow subgroup of patients but nonetheless, this study demonstrated that progression free survival could be extended with therapeutic intervention that focused on more than one antigen (here six) in a very difficult to treat subclass.

Notice the dotted oval line in the above graph. This area of the graph made the results appear slightly less impressive but provided a very important insight. IMUC could have designed their trial slightly better if they had excluded even more early progressors from initial enrollment. It is one thing to ask a dendritic therapy to defeat a cancer in patients with unmethylated MGMT tumors, it is yet a more difficult task to have it try to defeat that tumor if it is still actively progressing after chemoradiation therapy had concluded.

So, what would have happened if they had excluded those patients? Look at the graph again, and eliminate from your minds eye the part of the graph in the oval, and focus on the remaining part. You see that there is separation from beginning to end between control and treatment. More importantly, you'd basically remove the risk that the median might not be significant. As the curves are now, this result almost evaded science because of our regulators' dogged obsession with medians! You can see, the full graph barely separated in time at the 50% point. By excluding what were going to be non responders, they could have removed the drama and shown where the efficacy was and what patient subpopulation would benefit. In the future, patients similar to those in the oval will likely be provided stronger combination therapies, because monotherapy simply is not enough. This then would benefit all patients, because the stronger therapy, once proven, could be used on the whole populace if safety were not an issue. Full circle, as it were.

Boshie was correct to select the ICT-107 trial as a comparator for his statistical analysis regarding DCVax-L's enrollment and event data points. One primary thing I somehow neglected to observe, that Boshie probably caught, is that the extent of resection exclusion criteria in the phase II ICT-107 blinded trial was actually greater than that found in the Phase III ACT IV Rindopepimut blinded trial. I finally first noticed looking back at an SEC slide IMUC from 2014. Anyway, ICT-107 Phase II required the resection of tumor remaining to be 1cm or less. The Rindopepimut trial required 2cm or less. Meaning the ICT-107 trial had better overall surgeries on average than the Rindopepimut trial.

The other reasons the ICT-107 trial are a great comparator, is that, similar to DCVax-L, they actually used more than one antigen to target, they used ex-vivo dendritic cells pulsed with said antigens as their therapy, and they did not use parts of a limpet, (as Celldex did), to attach said antigens to. Probably most importantly, and I would guessed most overlooked, is that, imo, ICT-107 also had to obtain a sufficient number of PBMC to mature into dendritic cells. This wold require patients with higher lymphocyte counts (aka: often healthier patients). In other words, even though ICT-107's inclusion criteria had not listed absolute lymphocyte exclusion, the fact they had to get enough PBMC from leukapheresis means they could mostly only select higher absolute lymphocyte count patients, imo. Thus far more comparative with the DCVax-L trial.

On the other hand, IMUC used nonpulsed dendritic cells for their placebo. That, imo, was a huge mistake in the phase II ICT-107 trial, and imo it would be on the market today had they not done that, because the company now admits the "placebo" actually provided some efficacy. There is no question in my mind that there would have been more seperation in the OS and PFS KM graphs had IMUC simply used unmatured PBMC. Such errors are hard to perceive ahead of time, but NWBO was foresighted enough to conduct their phase III trial with PBMC placebo instead of active dendritic cells. (Note: In their new but currently suspended (for financial reasons) phase III trial IMUC used PBMC, just like NWBO is using in their phase III trial.

I can't tell you how angry I get when I think about the FDA's response to IMUC. I do not even invest in IMUC but ICT-107 should be on the market right now. It was significant for PFS in the ITT group, and as you say above, even effective in the very hard to treat unmethylated MGMT group. Even more, look at this graph:



My God! That is a 15.6 month benefit in the expected subgroup, EVEN WHEN SOME EARLY PROGRESSORS WERE ALLOWED IN! Had this not been, it is likely the separation would have been several years.

It is idiotic bureaucracy not to approve ICT-107 for HLA-A2*. The FDA could then follow it with market feedback from hospitals and clinics. It is such a narrow vulnerable subgroup. It's not too late FDA and IMUC. You could still get this done, and I'm not even invested in IMUC.

Anyway, back to DCVax-L. The only real advantage then that the Phase III DCVax-L trial has over the completed ICT-107 phase II trial is that NWBO likely excluded more (short lived) rapid progressors. However, this is likely a wash, because NWBO likely removed far more (long lived) pseudoprogressors. (although there are 32 in a separate blinded arm)

In the end, as Boshie notes in his article here, and I previously noted in my rudimentary chart and analysis here, DCvax-L (still) blinded data looks extremely promising. (Note: my chart uses the word "medium" instead of "median" which I missed when editing)

To sum up, The IMUC ICT-107 phase II trial is a very good example of a dendritic therapy using multiple antigens that worked extraordinarily well in a very narrow (DCVax-L is broad based) yet highly expected subclass of patients, and it is a very good trial to conduct (still) blinded comparisons with the phase III trial, wherein the differences in inclusion/exclusion criteria are verging on nominal, yet the apparent overall ITT results appear quite dramatic in favor of DCVax-L.

I'll try to add to this later, but that is it for now. Like I said, this is information already understood by most longs.

Here's to full parking lots!