Replies to post #62748 on NorthWest Biotherapeutics Inc (NWBO)

[Rkmatters]

We know neuro-oncology image standards are changing for immunoncology. I’ve believed for a while that they are trying to incorporate iRANO changes into the trial or at least use it to advocate for a change that improves the trial's chances for success, yet again, some how.

http://investorshub.advfn.com/boards/read_msg.aspx?message_id=121329183

But to be clear I don't necessary think that they are having as big of an issue as some skeptics believe with misdiagnosing psPD as progression patients. A few cases of course, but not enough to ruin the primary PFS endpoint of the trial over it. Instead my main concern continues to be that the primary endpoint is in jeopardy to fall short due to the advancements procured to the standard of care, since the early 2000s. The issue I saw last summer and still do see with the PFS, is that when the study is unblinded there will be a closer difference between placebo and treatment arms; smaller than most think because we have never seen radiation and chemotherapy introduced to a placebo group presented with the same inclusion/exclusion variables that this trial is and their prior DCVax-L Phase II was privy to.

Today intra-operative MRI surgical gross total resection (GTR) procedure, as well as Fluorescence Guided Surgery (FGS), has become common practice, that was not the case during the STUPP landmark study testing, 2000-2003. UCLA was one of first in the country to use iMRI suite, 2005, and had access to that precision neurosurgery technology during the trial. As such, some of the success seen in Phase II study needs to be credited to the surgical improvements combined with drug therapy. The vaccine didn't pick-up after STUPP protocol left off, instead it picked up after iMRI and/or FGS + STUPP protocol left off, if that makes sense. Researchers today are finding that Chemotherapy alone may result in long term remission in some GTR patients. And, of course, that means that their first-ever placebo patient cohort will see longer progression free and overall survival than the historical studies that the vaccine's efficacy data is compared to. I believe that NW Bio recognizes this and they will attempt to use iRANO new rules somehow to try and fix those possible PFS shortcoming and at least try to get regulators to agree to allow OS as a co-endpoint. Sort of manipulate the system and use the new iRANO immunotherapy "image times need changing" card they've been dealt. Again, it is true there maybe some misdiagnosed PFS treatment arm patients and those psPD patients could skew results slightly, but the true reason their primary endpoint is at risk to fall short of significance is due to changes related to standard of care in their intent to treat enrollment population, as it relates to immunotherapy which UCLA spear-headed. And I'm convinced that NW Bio is trying to de-risks for all of these concerns, as if we know about them, then they certainly do too.

And so where does the advancements in surgical care leave this study as far as overall survival is concerned? We know that standard of care patients (placebo) are living longer but surgery combined with chemo+radiation is not a cure. GBM remains vastly an almost incurable disease despite therapeutic effort. The odds of GBM recurrence at the initial surgical site (local recurrence) in the past has been high. I read somewhere it used to account for 80 - 100% of all recurrence. But, as you may have guessed, that has been changing. IMRI and/or FGS patients typically see longer-progression free survival, as a byproduct of a successful surgery, along with responses to treatment, and that has the ability to keep local recurrence largely in check (surgical site). However this local tumor control does not necessarily translate to better OS rates in SoC patients, because there are higher rates of Multicentric (multiple lesions, no clear path of spread) distance failure. Distance lesion appears, as the cancer cells proliferate and it often returns with a vengeous.

http://www.ncbi.nlm.nih.gov/pubmed/20175026/

"Improvements in microneurosurgical techniques, radiotherapy and chemotherapy for the treatment of high grade gliomas resulted in better local tumor control and longer progression-free survival. Multicentric (MC) lesions located distant from the initial resection area contribute to treatment failure in a growing number of patients. These MC lesions may develop within the course of the disease (metachronous) or may already be present at the time of first tumor manifestation (synchronous). To look for mechanisms and regular patterns behind MC glioma manifestations and to investigate whether they are "a second primary tumor" or the result of continuous diffuse glioblastoma cell invasion, we retrospectively analyzed the initial and all follow-up MR studies of our high grade glioma (HGG) patients."

AND

"Most multicentric lesions correspond to known and long established migrational pathways of glioma cells. The topography of multicentric lesions and the time needed to develop meta- chronous multicentricity indicate an active migratory process for the development of seemingly isolated glioma centers. Metachronous manifestation is more common among patients with a long progression-free survival. With growing evidence in favor of radical tumor resection and better local control through individualized and specific adjuvant therapies in glioma disease, increased life expectancy will see more multicentric glioma manifestations. Once multicentricity is established, patients have a short median life expectancy of 72 (metachronous) to 110 (synchronous) days.
Distant satellite lesions thus are the hallmark of extensive glioma disease, their topographical and chronological patterns of manifestation closely linked to the biology of invasion. "

https://www.hirslanden.ch/content/global/en/startseite/gesundheit_medizin/mediathek_bibliothek/fachartikel/verschiedenes/multicentric_tumormanifestationsofhighgrade/_jcr_content/download/file.res/MulticentricTumorManifestationsofHighGrade.pdf

And so for a SoC patient, the time between PFS and OS will have to do with whether the recurrence is local or distance to the initial surgical site. Typically with local recurrence, the PFS will reveal itself earlier, but the OS will show up much later. With distance recurrence, PFS shows up later, but the OS distance is then bridged and shows up much sooner. The better the initial surgical resection, the increased chance that recurrence reveals itself as an multi centric lesion and be Metachronous.

For EXAMPLE:

A typical MGMT methylation TMZ responding patient (no immunotherapy), could end up recording PFS within a 9-14 mo (depending on overall health factors (blood counts, age, KPS score, tumor location) + type of GTR surgery - recording a later PFS in iMRI and FGS), yet still end up with OS in 19-24 month range.

BUT with IMMUNOTHERAPY, with the help of the DC vaccine, that same MGMT methylation responding patient is less likely to end up developing a full-blown MC lesion. (If one shows up, it will be a false PFS event early on - a response to therapy). If that TMZ-VACCINE GTR patient eventually progresses, the progression will has a higher likelihood to occur at the initial local surgical site, many months later (well after PFS event on a MC lesion TMZ responding SoC patient who undergoes a GTR). These are the hallmarks of Mesenchymal progression patients. And given the vaccine is going after cancer cells, it will be following the same pathways, these patients recurrent distance disease is less likely to develop. And thus all one needs to do is look at distance between long tail progression event and survival, to see the benefit of immunotherapy. There is less likely to be the spread of disease with an immunotherapy that is effective. If the cancer recurs, there's a higher likelihood it will be at the local site, and when recurrence is local, the distance between PFS and OS is greater. :)

In brief:

IMMUNOTHERAPY+TMZ+Radiation responder PFS event at distance lesion location is less likely to occur (typically this shows up late in Placebo patients who have iMRI or FGS). AND if a distance mass shows up early, it is likely due to immunotherapy working, and while that immunotherapy patient may record an early PFS, they will go onto live a very long time, as the tumor mass will disappear over time (early recorded PFS; yet long tail survivor).

IMMUNOTHERAPY+TMZ+Radiation responder PFS event at local initial tumor site, is where most recurrence will take place. And so recurrence did not show up in a distance MC lesion (as iMRI and/or FGS SoC patients have a higher likelihood too), that immunotherapy patient will end up recording a longer PFS. And once PFS does finally show up, it will most likely be due to local tumor recurrence; given the vaccine is going after migration cells, if local recurrence occurs later, well, so will OS recording.

And so in looking at the following data from a similar immunotherapy, Pyrr had asked me this:

Just gonna trust when they say they "matched" them with a historical control that it's equivalent to stratified randomization?

I never answered, but I will now. The STUPP stratification of MGMT methylation data recorded 10.8 months PFS and 21.3 months OS; so yeah, this trial did great!

The 5 patients median age was 67 years old. AGE and KPS scores is a significant prognosis factor. Here is the chart of the 4 patients (the unmethylated patient removed). Notice how this small group overall survival improves (OS: 27 mo, 27.5 mo, > 36, > 32)? Notice two were still alive (as of late 2012)?

Patients characteristics
Patient Age(yr) KPS(%) RPA Sex MMSE score Tumor volume (cc) MGMT promoter Tumor location PFS (mo) OS (mo)
1) 69 70 5 M 28 111.8 Methylated right frontal 19.5 27.0
3) 50 80 5 F 26 12.9 Methylated l
left temporal 3.2 > 36.0
4) 67 60 5 F 28 68.3 Methylated right frontal 16.1 27.4
5) 71 90 4 F 30 44.8 Methylated right frontal 20.3 > 32.0


MGMT methylation patients:
Patient #1, 19.5 mo. PFS (indicative of local progression, as it did not show up early (a distance lesion would)) and OS 27.0
Patient #3, 3.2 mo. PFS (early progressive event, indicative of distance mass) and OS > 36.0 (patient remains alive)
Patient #4 16.1 mo. PFS (indicate of local progression) and OS of 27.4
Patient #5 20.3 mo. PFS (indicative of local progression) and OS > 32.0 (though patient on Avastin care).

GIVEN AGE and KPS are significant prognosis factor, I say these patient DID amazingly well on IMMUNOTHERAPY:

Patient #1 was 69 years old male, and had a KPS of 70.
Patient #3 was 50 years old woman, and had a KPS of 80
Patient #4 was 67 year old woman, and had a KPS of 60
Patient #5 was 71 year old woman, and had a KPS of 90

Here’s the study I referenced:

http://investorshub.advfn.com/boards/read_msg.aspx?message_id=122652331

The study’s conclusion:

“Despite the small number of cases, the survival time is clearly unexpected for a group with these characteristics. MGMT promoter methylation was the only positive prognostic factor, and yet the cohort with methylated MGMT promoter in the EORTC-NCIC shows a median OS of 23 mo, while all of our 4 methylated patients have lived longer than 27 mo, with two still alive. The median OS in our group exceeds the OS expected using the EORTC nomograms by almost 15 mo (27.0 to 12.3). Even with the intrinsic limitations of the small sample size, we conclude that this DC-based immunotherapy is likely to have added to the results of standard of care and provided an OS clearly superior to the expectations, suggesting a strong benefit from immunotherapy in an unfavorable group of patients. “

And to explain just how unexpected such results are, I also want to add this SNO abstract. It show just how AGE is a factor in the proliferation of disease. It shows just how impressive the DC data above is!:

251. THE SHORTER SURVIVAL OF OLDER PATIENTS WITH GLIOBLASTOMA IS PARTLY DUE TO MORE RAPID TUMOR CELL PROLIFERATION
?Junck, L., McKeever, P.E., Li, L., Blaivas, M., and Tkaczyk, A.; University of Michigan, Ann Arbor, MI, USA 
Objective: To determine the relationship among tumor cell proliferation, patient age, and prognosis in glioblastoma. Background: Many studies have shown advancing age to be a strong predictor of shorter survival in malignant gliomas. It is unknown whether this is because the tumors grow more rapidly in older persons or because older persons are more greatly affected by these tumors. The MIB-1 antibody has been shown to identify cells in the G1, S, G2, and M phases of the cell cycle. Methods: Paraffin-embedded tissue from 33 patients with glioblastoma was stained with the MIB-1 antibody. The proliferation index was calculated as the proportion of labeled nuclei among 1000 or more nuclei in microscopic fields containing the largest numbers of positive nuclei. For statistical analysis, the MIB-1 was modified with a square root transformation to normalize the dis- tribution. Results: The patients’ ages ranged from 18 to 78 years (median 60), and the proliferation index ranged from 1.5% to 80.8% (median 26.4%). The linear relationship between MIB-1 proliferation index and age was highly significant (R-squared=0.59, p<0.0001). Univariate analysis disclosed that the relative risk of death was 2.03 for each decade increase in age (p=0.002) and 1.63 for each 10% increase in proliferation index (p=0.0001). Multivariate analysis indi- cated that the relative risk of death was 1.73 per decade of age with control for proliferation index (p=0.04) and 1.38 per 10% increase in proliferation index with control for age (p=0.04). In the causal path- way from age to death, 22.7% of the relationship between age and death can be accounted for by the association of proliferation index and age. Conclusion: The relationship of survival with age in glioblastoma is at least partly due to more rapid tumor cell proliferation in older patients. Study supported by: NIH grants CA RO1- 68545 and 3P30 CA 46592-12 

Anyway, net net, OS should be fine even as a secondary endpoint. The issue is that this Phase III trial is very far along into the treatment regimen, so it will is anyone’s guess on how regulators will respond their request to accommodate a change due to iRANO.

http://investorshub.advfn.com/boards/read_msg.aspx?message_id=122274315

It has been suggested that NW Bio is trying to change the statistical analysis plan to include OS as a co-primary endpoint. Doing so would likely be their first choice as covers both the aspect of false PFS and it de-risks the trial. The issue is that OS might end up being considered a confounded co-endpoint in their Phase III, as no two patients treatment will be exactly uniform after progression. The same can't be said about their current PFS primary endpoint. All patient’s SoC would be the same (surgery, radiation, chemo (Stupp protocol)). They need to test DCVax-L on it’s own merits, and so I have my doubts that regulators will allow OS as a co-endpoint, as the crossover care is not standard and thus confounds it.

They may also be advocating for an iRANO change that gives the trial greater chance of success on their PFS primary endpoint, there are two possibilities:

Looking at MRI scans and reclassifying events:

The issue is they can't take back PFS events, as they can't forward date events. Why? Crossover. The patients crossed over when PFS was recorded and their treatment changed. They were confirmed to be on vaccine. The patients would be eligible for to deviate off of STUPP protocol, physician choice on care. So they can't just go back and redo the timeline of patients and keep their data, as the false progression patients deviated off the controlled aspect of the trial.

Looking at MRI scans and censoring misdiagnoses PFS data:

In hindsight they can see which patients were not true progression and reinterpret scans. They may be able to do censor the data of patients who were improperly diagnosed. And then enroll new patients in their place, adhering to new imaging standards. But that will take a bit of work to figure out a way to do this without disturbing the trial statistical design. Remember the data is being analyzed many way. I’m honestly not even sure it they can do this, as it may mean unblinding of patients. Also not sure how they would determine how many more placebo and vaccine events to enroll in the study’s place. And then, there is also the issue that they may be removing some of their best responding patients if they adopt this approach (the Brad's of the study). Therefore, I highly doubt they will consider doing this. Beyond a few obvious scans, it may be difficult to determine which scans remain questionable. Let's not forget that the blinded treatment changes at crossover, and so the follow-up scan is confounded in all patients. The Compassionate Use Arm struggled to determine how to classify the comparison scans, hence came up with "Indeterminates". There could be many of these patients, in the placebo arm, who may have been true progression, but then they crossed over to vaccine and their next scan is an improvement. Of course, any "Indeterminates" would automatically qualify as patients to censor. I imagine it would not be a much smaller group of the trial, but what if it isn't. If the vaccine works, follow up scans in theory would change. It just seems to complicated, so I just don't know if it's possible. Perhaps Flip knows more about that. He seems to think it's a possibility.

Whatever is being done behind the scenes, in my view, NW Bio is always trying to improve the Phase III for the better. The last change regarding lymphocyte count and the statistical significance hurdles certainly improved DCVax-L’s chances for success.

http://www.nwbio.com/nw-bio-obtains-approvals-for-enhancements-of-phase-iii-trial-of-dcvax-l-for-gbm-brain-cancer/

If they are not able to amend the trial to account for iRANO and further de-risk the trial somehow, then I still think that they will attempt to add more patients to the trial. And they may need to add more to get to 248 events in a reasonable timeframe. Reason being is they are screening out potential immunotherapy non-responders and they will only be left with a healthier immune patient. They may have not told us outright in 2014, but I believe they changed the lymphocyte count “inclusion” criteria after six weeks of chemoradition. Those blood cell counts can make a difference of 6 months in OS even before immunotherapy is introduced. And that kind of "inclusion" change will affect both arms before and after crossover. After crossover is where it will be seen most in immunotherapy. The less comprised and immunotherapy patient is going into an immunotherapy treatment, the better off they will do.

"The overall cohort has not yet reached its number of predetermined events. It seems like EVERYONE is living longer than we expected. Hopefully if patents didn't get anything the differences would be bigger." — Linda Liau in October 2015

"It seems like everyone is living longer than we would be expecting, so in reality what we are really comparing now is actually early DC vaccination verse later DC vaccinations” - Linda Liau in October 2015.

I believe now that she was talking about before and after various protocol changes were made in the trial design. Waiting for a secondary endpoint to hit with 348 patients might take a very long time if they don't as this trial does many things that earlier reiteration did not, including vaccine patients continuing therapy after PFS.

http://investorshub.advfn.com/boards/replies.aspx?msg=122273000

It may seem strange to consider ALL patients are living longer as a reasoning to add patients to the trial. Remember removing PFS misdiagnosed patients means the therapy was working. Therapy doesn't end in this trial as it did in prior Phases. It could be those patients never see an OS event. Hopefully that is the case. And remember if the vaccine therapy works there will be crossover patients who are in a healthier WBC state. I've shown example of studies where 20% of rGBM (which placebo patients are) could go on and have long tail survival (12 mos; 30 mos) once they go onto a DC therapy. That's essentially being a newly diagnosed patient. So here would be an issue as if some 25 -35 percent of vaccine patients are not eventing. And then another 20% of placebo patients are also long tail survival the study would not reach it’s endpoints for a while. They will not want to unblind the trial before patients receive at least 2 years of treatment therapy and risk PFS coming in short. They may need to be able to add patients to the overall trial so that they can reach the secondary endpoint sooner. (More patients enrolled than more potential events, some always occur early). So I think they actually may attempt to add more patients to account for a possible long-tail issue if they are seeing responses to immunotherapy. Anyway, lots of guesses. I’m going to have to wait along with everyone else on the reported outcome.

BTW, sorry for replying a day or so later. Life has been busy. I hope it helps :)