Replies to post #559542 on NorthWest Biotherapeutics Inc (NWBO)

[Dr Bala]

Rubbish post.

[biosectinvestor]

There was no "cherry picking" and in fact not all patients are fully resected in DCVax-L's trial and the residual disease factors appear to be similar. The fact is, no surgery can remove all of the glioblastoma for any patient. That is the problem with this disease, so suggesting that this trial is the only trial to have "total resection", suggesting that it is the only trial where all the tumor tissue is removed is quite absurd given the disease itself and the issues involved in treating it. All that was required was that surgery with an intent for total resection or near total resection be a part of the patient's treatment. The meaning of that? That patients that got only a biopsy were not included, mainly because not enough tumor could be taken in that context, at that time, for the treatment. But the key issue was that patients have enough tumor removed to create the vaccine, which doesn't mean every patient got a total resection at all. If you're saying they remove 100% of every patient's brain and they still survived longer, I'd be quite impressed. But basically, all patients have remaining tumor tissue and cells somewhere remaining, which is why this disease is so terrible.

Here Dr. Liau discusses the ECA comparisons, and shows the tables comparing the ECA studies and DCVax-L. The external firm that did the statistical analysis specifically used the measure of residual disease for the selection of what studies to include for newly diagnosed GBM. Note that it is not relevant for recurrent GBM. "Residual disease", in the tables, is the reference to partial or total resection and note that it is really about minimal or significant. Again, one has all tumor tissue removed. That's why patients get additional treatment after surgery, to kill those additional remaining bits of tumor and cells. Obviously. But the comparison studies were explicitly selected to be comparable.

Newly Diagnosed:

um and then you know empty MGMT methylation was pretty uh similar uh as
18:17
well as kind of a residual disease for the um uh for for the trials that we had
18:22
that data for these were the Baseline demographics for the um recurrent glioblastoma patients

Recurrent patients:

that data for these were the Baseline demographics for the um recurrent glioblastoma patients
18:30
um again uh so in this population there are 10 different trials uh mounting to
18:36
640 patients and then the 64 patients that had crossed over uh constituted are
18:44
treatment arm for the recurrent nucleoplastoma group
18:49
um here's some additional uh Baseline characteristics um of these trials and and of our uh new
18:56
you know external control patients as well as our treatment patients and the newly diagnosed DC vacs group as well as
19:02
their current DC vacs group um I won't go through all the the details um other than to say that the
19:08
the comparisons were pretty similar um of note we also looked at idh
19:14
mutation uh because there's one you know a thought that perhaps if there were a lot of idh mutated patients those could
19:20
be why there was a increased survival um and the uh the percentage of idh mutated
19:25
patients is only three percent uh in in our treated patients which was similar
19:31
to the external controls so this this is the uh the data
19:38
um the uh there was um you know a statistically significant difference in media and overall survival
19:44
uh of our uh you know uh DC vaccinated patients compared to external controls

Of course recurrent patients did not get additional surgery and showed substantial increased survival.

https://youtu.be/zMK2CNnA_vk?t=1091

Note also that the ECA data on survival are only estimates and it is likely that the patients in those other trials may not have lived as long as the estimates suggest, IMHO. Just my opinion after listening to other experts talk about patients and their own experiences.

Residual Disease as a subgroup:

um then uh this one was actually very surprising to us uh this subgroup analysis showing that the there's actually a greater survival advantage in uh patients with significant residual disease compared to minimal residual disease um and as a surgeon we were always the thought or we were always taught and we thought that uh taking out as much of the tumor as is possible is uh leads to a better prognosis um and uh and you know that still is the case you know because if you look at you know Max minimal residual disease their survival is still longer uh than than the survival in these um significant residual disease cohort but the relative survival Advantage is greater in in this group and what that's kind of suggests is that you know perhaps like I said one thing that the dendricks of vaccination does that we know it does is it gets tumors into the or gets T cells into the tumors um and perhaps in order to have a more diverse repertoire of antigen presentation and epitope spreading there may need to be some residual tumor uh still there so that the T cells once they get into the tumor and cause um get activated there may need to be residual tumor there to enhance the immune response uh and and promote spreading again that's a hypothesis that
that probably needs further validation but I found that to be very interesting in this particular case

https://youtu.be/zMK2CNnA_vk?t=1413

Residual disease is not relevant for Recurrent patients. For newly diagnosed, here is the chart again below. Looks like the stats are generally comparable for the various trials used.



----
No, Dr. Ashkan did not participate in the statistical analysis of the external patients. Neither did Dr. Bosch or any of the NWBO employees. The statistical analysis was completed entirely by an external statistical analysis firm, blinded to these results. Your assertion again is absurd and counter to the facts. See below quotes further down below from the SAP for further substantiation.

----

Accusing Dr. Ashkan of dishonestly, as some have Dr. Linda Liau and then effectively the 70 doctors that signed the study, as well as effectively accusing the JAMA reviewers of incompetence seems not just over the top, but not within the bounds of vague reality. It is very unlikely that all of these parties would 1) be involved in such a ridiculous conspiracy and 2) the suggestion that their patients having received the treatment or their participation in the study was such great compensation that they'd advocate for a failed treatment and commit academic fraud, for what in many cases were just a few patients, that's just so not credible as an allegation except in the recesses of internet chat boards that it really deserves no comment. But the reality is, lies do travel and modern discourse suggests that many people are quite gullible and vulnerable to such baseless allegations.

More from the SAP.

STATISTICAL ANALYSIS PLAN, Quantics Consulting, reviewed by James J. Digham, of University of Chicago, Professor of Biostatistics

The design of this SAP has been developed by multiple independent experts, and reviewed by independent, leading neuro-oncology experts who were not involved in the Trial, and all of the analyses called for under this SAP will be carried out by independent experts. This SAP was developed prior to data lock and unblinding of the Trial data.

EXCLUSION CRITERIA
Patients may not be enrolled in the trial if they have any of the following exclusion criteria, at the indicated time points:

. . .

Post operative MRI scan evidence of biopsy only without significant tumor resection.

It doesn't say total there. Total resection is the intention for any patients getting surgery in this context, for the Standard of Care, and of course it would apply if you need a sample of the tumor to generate the vaccine. For most patients, that is often not possible given the need to preserve brain function and because some tumors are inaccessible.

TREATMENT PLAN OVERVIEW
After the informed consent for the collection of the tumor has been signed (Appendix D), all subjects will undergo surgery at a center participating in the trial. As much as possible, or approximately 2-5 grams of the tumor material obtained from surgical resection will be placed in a sterile container containing a mixture of tissue- dissociating enzymes. This is the source for the tumor lysate material that is used as antigen in this study. The container is shipped to Cognate, Sunnyvale, California for manufacturing of the tumor lysate antigen preparation.

...

Visit 2 (Post-surgery MRI): MRI scans pre- and post-surgery need to document extent of resection against enrollment criteria. Patients must have a surgically accessible tumor for which surgical resection, with intent to perform a gross total or near gross total resection, is indicated. Biopsy only does not satisfy eligibility rules. MRI scan (post-surgical) must demonstrate that a significant resection was performed. All MRI scans will be reviewed by independent radiology review.
The post-surgery MRI will be used to quantitatively determine the extent of resection for purposes of the statistical analyses in section 13.2.3 and 13.3.3. Extent of resection will be defined as the amount of residual tumor volume measured in mm3.
The amount of tumor lysate protein is determined as soon as possible after surgery by Cognate. Determination that sufficient tumor lysate protein was generated is required for a patient to remain eligible and to undergo leukapheresis.

In Criteria for endpoint review it notes that not all patients received total resection as taking note of that is required for measuring progression.

? In the case of incomplete resection during primary therapy: a 25% increase or greater in the residual tumor if the recurrent portion of the tumor is at least 1 cm or greater in its longest diameter, measured by MRI and confirmed by scans above as attributable to tumor growth;

13.7. SECONDARY ANALYSIS OF COVARIATES
Associations of 1) age, 2) KPS, 3) extent of resection, 4) clinical site, and 5) Temodar® usage with the primary endpoint and first secondary endpoint will be assessed. Proportional hazards regression modeling will be done to determine the sets of covariates independently associated with the endpoints and also to evaluate interactions of these covariates with the effect of treatment with DCVax-Brain. All randomized patients will be included in the analysis.

These independent biostatisticians indicate:

In the 15 years since TMZ was approved for GBM, scientific knowledge about GBM has advanced greatly. Today, many prognostic factors are well known (e.g., MGMT gene promoter status, IDH mutation status, age, extent of resection, performance status, etc.) and molecular sub-types of GBM have been identified and analyzed [3]. However, hundreds of clinical trials testing a wide range of diverse treatments have failed to achieve meaningful extension of survival. A survey of the trials conducted over more than a decade following TMZ approval (2005-2017) found that, out of 417 clinical trials involving 32,000 patients, there were only 16 Phase III randomized trials and only 1 of those trials was found to extend survival or provide other material benefit [4]. In the last 2 years, more Phase III clinical trials for GBM were reported, some including treatment with checkpoint inhibitor drugs (CPI), and all of these Phase III trials also failed to extend survival.
The only two new treatments approved for GBM in the 15 years since approval of TMZ in 2005 were both approved for commercial use with no extension of survival. The first is Bevacizumab, which received accelerated approval followed by full approval based on progression-free survival and improvement of quality of life, but no improvement in survival [5]. The second is the Optune tumor treating fields (TTF) device which was approved for recurrent GBM based upon non-inferiority with standard of care followed by a label extension to newly diagnosed GBM, based on prolongation of progression free survival and overall survival [6]. For all of the foregoing reasons, there is a severe unmet medical need for new and better treatments for GBM that extend survival. This Phase III trial of DCVax®-L, with an intent to treat (ITT) population of 331 patients, is one of the last remaining Phase III trials for GBM that is currently still active. To better analyze the data based on the current scientific advances and the recent understanding of the GBM biology, its natural history and response to therapy over the past 13 years, we have developed this SAP.
The design of this SAP has been developed by multiple independent experts, and reviewed by independent, leading neuro-oncology experts who were not involved in the Trial, and all of the analyses called for under this SAP will be carried out by independent experts. This SAP was developed prior to data lock and unblinding of the Trial data.

Further, they considered extent of resection in their analysis.

The identification of the most appropriate external controls from other clinical trials in newly diagnosed GBM, based upon comparable patient populations, time period of trial conducted, extent of resection and standard of care RT/TMZ treatments consists of a structured literature search and has been conducted by the independent expert firm, York Health Economics Consortium, UK (see Appendix A). This independent expert analysis was checked and confirmed with a panel of leading, neuro-oncology experts who were not involved in the Trial. In addition, the statistical analyses include methodologies to ensure appropriate matching of these identified external controls to the DCVax®-L trial population based on known prognostic factors, applied by independent statisticians.
To support the first secondary endpoint for survival analysis in recurrent GBM, a systematic literature search was also conducted similarly, to identify the most appropriate external controls from clinical trials in recurrent GBM, based upon comparable patient populations and the time period of the trial conducted (see Appendix B).

Analysis #3: Survival in pre-specified sub-groups: It has become well recognized that GBM patients may be categorized into distinct sub-groups based upon certain patient characteristics and/or tumor characteristics, and these distinct sub-groups have materially different survival with GBM. Such characteristics include the methylation status of the MGMT gene in the tumor, IDH mutation status, extent of surgical resection of the tumor, and others. As explained above, in this SAP a portion of the statistical alpha is separately assigned to analysis of pre-specified sub-groups. This alpha will be applied to the first sub-group analysis. After the first one, additional sub-group analyses will be conducted in a hierarchy – i.e., each successive sub-group analysis after the first one will be reached if the one before it is positive. The first sub-group analysis will be survival in patients with methylated MGMT gene. Thereafter, additional analyses will include, minimal residual disease, age <65, significant residual disease, unmethylated MGMT, and age ≥ 65.

Secondary Endpoint #5: Tumor Response – Comparison between subjects randomized to DCVax®-L and those randomized to Placebo within Study 020221.
Resection of GBM typically focuses on removing the enhancing tissue as that is supposed to be the area of active tumor growth. Incomplete resections result in residual enhancing disease post-surgery and most patients also have non-enhancing disease areas surrounding the enhancing lesion. Tumor response will use pre- and post-randomization measurements of enhancing and non-enhancing disease. Modified RANO criteria will be applied per Cloughesy et al. [7], allowing for evaluation of non-measurable disease to determine response rates.

The parameters that will be evaluated include but are not limited to the following:

Therapy:
? Extent of resection (complete/ partial and significant residual /minimal residual)

These are not all references, as some are not easily transferable other than by screen shots and I don't have the patience to do screen shots. What is apparent is that they extensively took into account issues of partial or total resection in their comparison with other trials, and they did statistical analysis to address measuring between the trials on the basis of various prognostic factors including resection, total and partial.

[HappyLibrarian]

GBM is such a deadly disease there’s very little to cherry pick among the patients. If there were then the survival rate would be a hell of a lot higher. Medically speaking the cherries are rotten as it were.

[biosectinvestor]

The ECA takes this all into account, and there was no cherry picking. Further, patients that had more residual disease in the DCVax-L trial did better than the same patients in the comparable trials. Dr. Liau discusses this in her presentation with Al Musella. The term there is residual disease. And no, none of the patients get ALL of their disease removed, that is the problem with GBM. You can't remove all of the tumor. The numbers of patients with minor or significant residual disease is comparable with the other trials they used for nGBM. For rGBM, it's not relevant. And the independent group that did the SAP and ECA actually calculated for these differences in patients and between trials.