Expert Financial Analysis and Reporting

Key Points

• In the third quarter 10-Q regulatory filing, it was disclosed that the draft form of the trial’s Statistical Analysis Plan (SAP) has been finalized for the phase 3 trial of DCVax-L in glioblastoma multiforme (GBM). This followed lengthy and extensive work on the part of independent statisticians and the distinguished Scientific Advisory Board.
• Preparing this draft has been the key hurdle to proceeding to the next steps of the regulatory process, locking down data and unblinding and then analyzing the data pursuant to the final SAP. While this draft is close to what the final form of the SAP will be, it is prudent, although not mandatory, to allow regulatory agencies to review it and offer comments.
• The CRO that conducted the trial has been and will continue to be completing the final compilation of data pursuant to locking the data base.
• After considering potential comments on the SAP, the Company could have complex decisions to make in finalizing a single SAP for regulators in the countries in which regulatory filings are planned (US, UK, Germany and Canada), given the potential variations in each regulator’s approach to such approvals and the rapidly evolving world of immune therapy.
• Prior analysis of blinded data strongly supports the hypothesis that DCVax-L treatment of GBM leads to prolongation of overall survival in the trial and a survival tail for some patients that is unprecedented. See this link.
• Key opinion leaders in the trial have been very encouraged by blinded results and believe that DCVax-L could be a major advance in the treatment of GBM. See this link.
• NWBO at this point is not giving precise guidance on when topline data will be available to investors, but one thing is crystal clear. In this marathon of a clinical trial, the finish line (release of top line data) is in sight and will be crossed.
• It is important to understand that while investors tend to center their attention on what the FDA will do, three other regulatory agencies will be independently reviewing the regulatory filings. It will be interesting to see how the Company handles what could be a complicated jig-saw puzzle on the road to various filings and potential approvals.
• I share the view of the two lead investigators in the trial that the blinded data and other evidence seen so far supports a hypothesis that DCVax-L could be a major advance in the treatment of GBM. If so, this could be as important to oncology as the development of checkpoint inhibitors. Like the checkpoint inhibitors, the technology on which DCVax-L is based is applicable to a broad range of solid tumors and the commercial potential could be staggering.

Important Disclosure in the 3Q, 2019 10-Q

Investors have been sitting on the edge of their chairs awaiting word that will give an insight into when we will finally see the top line results of the phase 3 trial. On page 22 of the 3Q, 2019 10-Q, there is some important information pertinent to this. The section reads as follows:

“As previously reported, the Company has been moving forward with the several stages of work that are needed to reach completion of this trial. These include finalizing the Statistical Analysis Plan, conducting the final data collection, data validation and data lock, and unblinding and analyzing the data. Each of these stages involves teams of outside experts as well as Company personnel.

The independent statisticians and the Company have completed the draft of the Statistical Analysis Plan (SAP).  Northwest is proceeding with the regulatory processes relating to the SAP.  It is continuing to consult with its Scientific Advisory Board and Board of Directors to move forward as prudently and expeditiously as possible to data lock, unblinding and top line data.

The independent contract research organization managing the trial has been moving forward on resolving queries to confirm the trial data, in parallel while the Company and the statisticians have been developing the SAP.  The Company’s understanding is that most of the queries have been resolved and only a small number remain outstanding.”

So, what does this mean?

Phase 3 Trial of DCVax-L Will Produce a Treasure Trove of Information

The phase 3 trial of DCVax-L started over 12 years ago. Median enrollment occurred in May 2014 and the last patient was enrolled in November 2015. While the length of the trial has been beyond exasperating for shareholders, this trial will be singularly unique in the amount and diversity of data that it will provide regarding the use of DCVax-L in the treatment of GBM and extremely important information on duration of its therapeutic effect. The trial employed a cross over design that allowed all patients in the trial (both those who started on standard of care (SOC) and those who received SOC plus DCVax-L) to receive or continue to receive DCVax-L when their cancer progressed, while remaining blinded about which treatment they had been receiving before their cancer progressed. Based on the Company’s presentations and announcements over time, here is an approximate summary of treatments received by patient groups in the trial:

• 331 patients were enrolled
• About 232 initially received DCVax-L plus SOC
• About 99 initially received SOC
• About 10% of the 331 total patients never received DCVax-L
• If all 10% were in the SOC arm of the trial, then about 66 of those 99 patients received DCVax-L when their cancer progressed.
• An unknown number of the 232 patients initially started on DCVax-L plus SOC continued to receive DCVax-L when their cancer progressed.

This puts forth a challenge in how to best analyze that data and present it to the four regulatory agencies in its most informative way. The Statistical Analysis Plan (SAP) provides the analytical structure and terms for analyzing the data.

Statistical Analysis of the Trial Results for DCVax-L Requires a Different Approach Than for Chemotherapy and Targeted Therapies

The original trial design intended to use median progression free survival (mPFS) as the primary endpoint with median overall survival (mOS) as the second primary endpoint. These are meaningful endpoints for chemotherapy drugs and targeted therapies which dominated oncology drug development at the time this trial was started. If they are effective, chemotherapy and targeted therapies work quickly to shrink the tumor or halt the progression so it is quite meaningful to use mPFS as an endpoint.

At the time the phase 3 trial started, there was very little experience with immunotherapies like DCVax-L and indeed there was a consensus opinion that they probably would not be effective. There was almost no clinical experience with this class of drugs. This changed dramatically with the advent of the checkpoint inhibitors led by Yervoy, Opdivo and Keytruda, the first medically and commercially important immunotherapies. The truly exciting aspect of these drugs was that a small but highly meaningful sub-group, perhaps 10% to 15% of patients, were living much longer than could be expected with chemotherapy and targeted therapy. This has come to be known as the survival tail. Also, it was seen that immunotherapy drugs took longer to act. In some cases, the tumor actually progressed, but patients nevertheless experienced improved survival.

The phase 3 trial of DCVax-L started before the immunotherapy revolution began, but also transcended the accumulation of knowledge that led to a much better understanding of how immunotherapy drugs worked and most importantly how to judge efficacy in a clinical trial. Capturing the survival tail may be the most critical endpoint along with median overall survival.

The Purpose of the SAP and What It Is Likely to Emphasize

The SAP is the document that companies submit to regulatory agencies that prospectively (before data is unblinded) explains what will be measured (i.e. the primary and other endpoints) and how the resulting data will be analyzed. The SAP can be prepared and submitted at any time prior to unblinding of the data. The original thinking over 12 years ago was that the SAP would feature mPFS and mOS as the key endpoints.

The median enrollment in the trial was reached in May of 2014 and the last patient was enrolled in November 2015. Half of the progressions in the trial and half of deaths in the trial were probably reached by 2017. If the SAP was to be based on mPFS and mOS, the trial could have been stopped then, data analyzed and topline results made available in 2017 or so. However, doing so would almost certainly not have allowed the determination of whether there was a medically important survival tail. Management, with strong recommendations from their Scientific Advisory Board of immune therapy experts from around the world, made the decision to not unblind the trial too early, in order to determine if and to what extent there is a survival tail.

The extraordinary length of this trial provides a wealth of information on duration of effect and survival. The last patient in the trial was enrolled four years ago so for that very last patient, we will know if they have or have not survived four years. For all other patients there will be a much longer observation period. For the median patient enrolled in May 2014, there will be five- and one-half years of observation.

In past reports I looked at six randomized GBM trials of other drugs in which 1,608 patients received SOC in the control arm.  See this link.  This data suggested that 50% of SOC patients are alive at 17 months, 30% at two years, 15% at three years, 8% at four years and 5% at five years. By this measure, the last patient treated in the DCVax-L trial more than four years ago, would have only an 8% chance of surviving to this point if they had received SOC. For all other patients, the odds would be less. Because we will have at least four years of data in the phase 3 trial on every almost every single patient (some could be lost to follow-up), we will have an extraordinary ability to compare the survival of patients in the trial to what would be expected if they had received just SOC. The survival tail, if any will should be crystal clear.

In the SAP, Northwest will prospectively identify what endpoints will be measured and the statistical analysis techniques that will be used to analyze the results. Here is what I think this SAP will deal with:

• Survival tail will be one of the most important endpoints. This may be the first trial in any cancer indication to do so and is only possible because of the unprecedented length of the trial.
• mOS will also be important
• mPFS will not be that important although it will be measured.
• The cross over design of the trial allowed patients on SOC who progressed to cross over to DCVax-L. This gives an insight into the effect of DCVax-L in recurrent glioblastoma for patients treated with SOC.
• Patients who suffered a progression while on DCVax-L were allowed to stay on the drug. This also could provide valuable information on the treatment of recurrent GBM tumors.
• There will be a subgroup analysis for all of the above parameters for the very important methylated and non-methylated sub-groups.
• Likewise, for degree of surgical resection and other risk factors.

This lengthy trial has been exasperating, but in the end, it does mean that there is a treasure trove of information that will give a strong insight into any survival benefits afforded by DCVax-L.  I sincerely doubt that there will ever be another trial that duplicates the length of this trial and yields the resultant information. It would be prohibitively expensive for the sponsor.

Anticipating Investor Questions About the SAP

There are a few questions that investors may have about the SAP.

• Did the FDA and other regulatory agencies participate in drafting the SAP? They may or may not have made suggestions.
• Who put the SAP together? This was the collective input of expert outside statisticians, the Company’s Scientific Advisory Board, and management.
• Do the regulatory agencies have to approve the SAP? Now that the draft is completed, it will be submitted to regulatory agencies for their review. They may or may not make suggestions, but NWBO wants to give them the opportunity.
• Which regulatory agencies will the SAP be filed with? The regulators for the countries where the trial was conducted: US, UK, Germany and Canada.
• Some expected the SAP to be completed in August or September. Why was it delayed until November? This is a very complex and critical document and NWBO only has one chance to get it right. They erred on the side of caution.

What Happens Now?

The draft of the SAP will now be submitted to regulatory agencies in the US, UK, Germany and Canada. The CRO that handled the clinical trial will now finish the collection of data and then lock the data base. This could take some time with the holiday season coming up. When the data is finalized and collected, it will then be unblinded and analyzed in accordance with the SAP. In an information rich trial like this one, this top line data analysis will be extensive. Thereafter, top line data will be released. The full trial results, assuming a positive outcome, would then be published in a major medical journal and/or a major medical conference.

Thoughts on Regulatory Approval

Of course, the regulatory decisions in the four countries will depend on the results of the trial. Critics might question that the FDA may have issues with the cross over design that led to only 33 patients or so receiving just SOC. This may not be a meaningful control arm. I think that this argument might be appropriate if we were dealing with a drug in a disease state in which there was other effective therapy and/or that the disease was not immediately life-threatening, say a new cholesterol lowering or anti-hypertensive agent.

It is important to take into account that DCVax-L is being developed for treating a deadly cancer for which there has been no new drug approved since 2005 whose initial approval was based on extending survival. I think there is precedent that shows that the FDA is eager to approve drugs in a circumstance like this and will not be rigidly dogmatic in evaluating trials dealing with such aggressive cancers. Consider the case of the CAR-T drugs Kymriah and Yescarta. These were tested in desperate r/r DLBCL and r/r ALL patients who had short survival expectancies. Glioblastoma multiforme is comparably deadly by this measure. It is important to note that both CAR-T drugs were approved on the basis of objective responses with no information on survival and there was no control group. The trial of Yescarta was based on only 112 patients with r/r DLBCL and the Kymriah trial in r/r ALL enrolled only 77 patients.  In sharp contrast, the DCVax-L trial enrolled 331 patients and will have extremely meaningful results on overall survival and duration of survival (the long tail).

How might regulatory agencies consider the data? Obviously, this will be importantly determined by the trial results. Let’s assume that the unblinded data confirms what has been suggested by the blinded data; i.e., there is a survival tail comparable or better than has been seen with checkpoint inhibitors in lung cancer, melanoma and other cancers and there is a therapeutically meaningful improvement in mOS as judged against historical data for SOC. If any of the four regulatory agencies conclude that DCVax-L will have a particularly great impact on the treatment of GBM, they could designate DCVax-L for a priority review.

Dr. Keyoumers Ashkan is one of the two lead investigators in the DCVax-L trial and has said publicly that he believes that DCVax-L is potentially a major advance in treating GBM.  Dr. Ashkan is the clinical lead for neuro-oncology at King's College Hospital in the UK. He has an active research interest in brain tumors and heads the Neuroscience Clinical Trial Unit. Kings College is the premier teaching hospital in the UK. Dr. Ashkan is one of the most respected neurosurgeons in the UK and has the respect of British regulators.

At the 2019 ASCO meeting Dr. Ashkan mentioned that under a UK compassionate use program, he has continued to treat a few patients each month with DCVax-L. He further said that results in these patients seems to confirm that there is an important therapeutic effect. I don’t know what a few patients means, but the importance is that Dr. Ashkan has developed real-life experience with DCVax-L and has described that experience as quite encouraging. This may or may not affect the UK regulatory review.

Final Word

Completion of the SAP is a major milestone on the road to unblinding, and top line data for this very long and potentially very significant GBM cancer vaccine immunotherapy trial is now very close.

Given all the variables on this final stretch of road described above, it is dangerous to be precise as to an exact date for release of topline data. However, we must remember that the seemingly endless “delays” that have frustrated all of us, may very well be the result of extended patient survival in the phase 3 trial.