Immunotherapeutic Niches, Competition and Collaboration. Part I: Vaccines.
(Version 3.0. Sentiment provided editing assistance. I was going to submit this to Seeking Alpha, but it required I belong to a firm with financial analyst(s). Anyway, I think you'll find this more informative than my prior drafts. The next installment (Part II) will not be completed until Monday or Tuesday of next week -- this series will take a bit longer than I previously suspected.)
This is Part 1 of 5 on DCVax and other immunotherapeutic approaches against cancer. Because there are more treatments in various company pipelines than one can shake a stick at, this discussion is limited to most major treatments recognized recently in the Journal of Neuro-Oncolgy, plus I am including another group called immunostimulants. I know Pyrr and L. Smith covered much of this territory previously, and my only thought is to simply log this exercise/self-imposed assignment to enhance my own due diligence and perhaps provide yet another reference set of old posts to keep in your personal files.
In all there are 17 therapies that can be placed into four approaches that treat cancer. The four Immunotherapeutic Approaches herein are:
I. Vaccination
II. Cellular
III. Immunomodulation
IV. Immunostimulants
This message/post (Part I) will discuss Vaccination Therapeutic Approaches. Each therapy is broken down into three components: Niche, Competition and Collaboration. Below are 4 vaccines presented in this discussion.
A. IMA950 Rindopepimut (EGFRvIII) is a vaccination immunotherapeutic approach using Tumor-specific antigen vaccine (TSA).
B. ICT-107 is a vaccination immunotherapeutic approach using Tumor-associated antigen vaccine (TAA).
C. HSPPC-96 (Aka: Prophage) is a vaccine therapeutic approach using HSP96 vaccine (Heat Shock Protein Technology).
D. DCVax-L is a vaccine approach using dendritic cells partially matured and activated outside the body (ex vivo) with whole tumor lysate, interferon gamma plus an attenuated virus. DCVax-Direct also partially matures dendritic cells outside the body with interferon gamma and an attenuated virus, it is then directly injected intratumorally inside the patient’s body (in vivo).
____________________
A. IMA950 Rindopepimut (EGFRvIII).
Coincidentally, Celldex came out with updated phase II interim updates November 13-16, 2014, and from reading through the message board and Smith on Stocks, there is already a great deal of understanding regarding any potential ramifications.
i. Potential Niche for Rindopepimut.
Rindopepimut is owned by Celldex Therapeutics. It uses a peptide vaccine to target one particular antigen found on some tumors -- EGFRvIII. EGFRvIII is expressed in tumors in about 20-25% of glioblastoma patients. J Neurooncol. 2006 Jan;76(1):23-30. EGFRvIII is a functional and constantly activated mutation of the epidermal growth factor receptor EGFR, a protein that contributes to cell growth and has been well validated as a target for cancer therapy. Cells expressing the EFGRvIII mutation are constantly signaled to divide, resulting in very aggressive tumors and a poor prognosis.
On their website, Celldex Therapeutics states, "Unlike EGFR variant, EGFRvIII has not been detected at a significant level in normal tissues, but it has been identified in glioblastoma; therefore, targeting this tumor-specific molecule is not likely to impact healthy tissues.” That is in theory, in practice, Celldex’s Rindopepimut therapy already logged a “potentially related” severe case of cerebral edema in a 42 patient compassionate use study -- according to an abstract in the Journal of Neuro-Oncology. (2014) 16 (suppl 5):v115-v116.
Conclusion:
Glioblastoma patients that have tumors expressing this particular antigen present the potential therapeutic niche for Rindopepimut. If this therapy were to become first line standard of care for patients with both newly diagnosed and recurrent GBM expressing the EGFRvIII antigen, it would encompass 20-25% of the total GBM patient population. However, as discussed below, this percentage is actually illusory when considering competition, because once tumors expressing EGFRvIII are eliminated, Rindopepimut is no longer efficacious against GBM progression.
ii. Potential Competition between DCVax and Rindopepimut.
As some people already know, the competition between Rindopepimut and DCVax is likely modest at best. Let's take a look at the main scientific reasoning behind this. In addition to tumor EGFRvIII expression being limited to about 20-25% of the GBM population, a major problem with EGFRvIII targeting is immunologic escape after progression-free survival with epidermal growth factor receptor variant III (EGFRvIII) peptide vaccination in patients with newly diagnosed glioblastoma. J Clin Oncol. 2010;28(31): 4722–4729.
In other words, tumor cells express hundreds of antigens, and even within the patients whose tumors express EGFRvIII, there are other tumor cells that do not. Once a tumor specific antigen therapy like Rindopepimut eliminates the EGFRvIII antigen it is targeting, tumor cells not expressing that mutation rapidly increase and subsequently the cancer returns. Moreover, as time passes, more mutations arise due to the accelerated mutation rate in malignant cells. Consequently, the problem with tumor specific antigen therapies is that cancer inevitably escapes.
DCVax, on the other hand, uses all the antigens from the tumors in the body. DCVax theoretically additionally targets tumors that express EGFRvIII biomarkers (in addition to hundreds of other tumor expressed antigens). Rather than scientists determining which antigens to target, the body’s own personalized dendritic cells uptake abnormal tumor antigens, migrate to nearby lymph nodes and instruct t and b cells to target those abnormal biomarkers expressed on those tumor cells. DCVax Dendritic cells are properly matured ex vivo and selected to ensure the patient's DCs have the best chance to do their job correctly and overcome tumor induced immune suppression. DCVax dendritic cells are also preactivated with attenuated bovine tuberculosis bacillus and interferon gamma; because dendritic cells uptake and express attenuated virus fragments prior to in vivo administration, the attenuated virus itself is not injected into patients. Still, the inclusion of dendritic cells expressing these particular viral fragments assists in initiating an extremely reliable immune response.
Rindopepimut was only tested on patients with EGFRvIII specific tumor expression, which again, only accounts for 20-25% of the global GBM patient population. In Celldex's ongoing phase II open label trial for recurrent GBM patients with EGFRvIII, updated interim result improvements, while better than standard of care, are modest. PFS and OS are demonstrating 2-4 month improvement over SOC. Although significant for patients, if these results continue to manifest themselves in this manner, it is at best an incremental advancement. Three phase II Rindopepimut trials on newly diagnosed GBM with EGFRvIII resulted in 5-7 months OS advantage over historical/contemporary controls. A larger ongoing 700 patient phase III Rindopepimut trial in newly diagnosed GBM patients expressing EGFRvIII tumors is expected to demonstrate efficacy; however, tumor escape (as noted previously) will likely shorten the durability of response significantly. The primary results for Celldex’s phase III trial in newly diagnosed GBM patients are not due until late 2016 or early 2017.
So what do we know about DCVax efficacy in the EGFR variant tumors, and more specifically, in the EGFRvIII expressing tumors? Remarkably, serendipitously and somewhat curiously, as of November 13-16, 2014, we now have a maurine (mouse) trial at our disposal that resulted in poorer response against EGFRvIII tumors than the remaining tumor types when treated with either 1. DCVax(L?) 2. Novolomib and 3. DCVax(L?) with Novolomib. No one else, from my humble review of the literature, has delved into whole tumor lysate pulsed dendritic cells' effect upon EGFRvIII tumors, but I suspect the phase III DCVax trial results will include pathology data regarding EGFRvIII. Upshot? I would not place too much stock in said ‘well timed’ maurine study without further confirmation.
In my opinion, a. due to the realization that the DCVax-L trial will likely provide EGFRvIII data and b. due to the fact that soley EGFRvIII targeted tumors ultimately escape Rindopepimut therapy once this mutation is eliminated and c. Celldex's sudden rush to move forward phase II open label results for FDA approval based upon otherwise premature incremental results from a small study, Celldex is concerned, and should be.
While this article focuses on the biological aspects for treatment, it should also be noted that Northwest Biotherapeutics already has FDA and German (PEI) manufacturing approvals. NWBO is already commercially producing DCVax-L for patients in Germany, it received unprecedented regulatory designations in Great Britain and Germany, it developed a patented manufacturing process to lower therapeutic costs for patients, and NWBO is years ahead of competitors regarding European regulatory product advancement. In Newly Diagnosed GBM, NWBO is at least one full year ahead of Celldex in their phase III trials, and Celldex is potentially two to three years behind NWBO in regulatory progress within Europe.
EGFRvIII is found in many other types of solid tumor cancers; but to date, I am not aware that Celldex is initiating additional trials.
Conclusion: Rindopepimut’s potential 0 to 25% of GBM market is uncertain until the phase III DCVax-L and Rindopepimut trial results in newly diagnosed GBM patients are published. Potential competition from Rindopepimut is modest at best.
There are many reasons to believe the overall response rate against all GBM patients eventually treated with DCVax-Direct will be higher than the 80% response rate reported for DCVax-L. Each percent over 80% for Direct likely leaves less therapeutic range and market share for Rindopepimut. Recently, DCVax-Direct results in other cancer types confirmed it can overcome the local and systemic immunosuppressive effects presented by cancerous tumors. It can switch back on the patient’s correct immune system response to intracellular diseases like cancer — called the THI response -- while restoring balanced CD4/CD8 ratios, and like DCVax-L, it utilizes all abnormal antigens from tumors that are selected, engulfed and thereafter potently expressed by dendritic cells to T and B cells within the lymph nodes.
iii. Potential Collaboration between DCVax and Rindopepimut.
Could Celldex and NWBO form a cocktail to treat patients with EGFRvIII tumors? While GBM ultimately escapes the attack on EGFRvIII by Rindopepimut via loss/termination of this particular mutation, DCVax — especially in the case of DCVax Direct — would most likely prevent further escape because it addresses the remaining/hundreds of tumor antigens that would remain unaddressed by Rindopepimut. Despite a serious event of cerebral edema in a 42 patient compassionate use study possibly related to Rindopepimut therapy, overall safety appears excellent in both therapies.
Conclusion: Potential Collaboration by combining the two treatments is theoretically plausible (in the EGFRvIII tumor expressing population) and may only be ruled out (or ruled in) through additional formal studies for combination therapy (aka: cocktails). If successful, this might provide earlier remission with longer durability in that patient population than was previously achievable.
B. ICT-107
Coincidentally, like Celldex/Rindopepimut, Immunocellular/ICT-107 came out with a phase II trial update at the same time — November 13-16, 2014. In the case of ICT-107 however, the trial was blinded, and the phase II primary endpoint was already reached.
i. Potential Niche for ICT-107
Immunocellular owns the right to ICT-107. ICT-107 is an autologous (patient-derived) dendritic cell vaccine that targets six different tumor antigens associated (TAA) with glioblastoma multiforme (GBM): AIM-2, MAGE-1, TRP-2, gp100, HER-2, IL-13Ra2. At least four of these tumor-associated antigens are highly expressed on cancer stem cells (CSCs). ICT-107 is designed for use following surgical tumor resection, and is used in combination with standard treatment to include radiation and chemotherapy.
In clinical trials, ICT-107 is potentially effective against 30% of the overall GBM tumor population. HLA-A2 positive unmethylated MGMT patients comprise approximately 15-20% of the broad GBM population, and HLA-A2 positive methylated MGMT patients comprise approximately 10-15% of the broad GBM population. Scientists have grown accustomed to various oncology treatments working against tumors expressing “methylated MGMT.” A simple explanation is due to the fact that normally, when tumor DNA is partially destroyed, a gene known as MGMT repairs the tumor’s DNA, thereby rendering the therapeutic agent ineffective. However, patients with tumors that have “methylated” MGMT “promoter” genes cannot effectively repair tumor DNA when it’s damaged; and therefore, cancer therapies are likely to be more effective in those patients. Separately, "overrepresentation of the HLA-A2 phenotype was observed in both ovarian and prostate cancer in Swedish patients compared to the normal population. The higher frequency of this allele found in Scandinavian countries significantly decreases moving further south in Europe. The relationship of HLA and cancer further extend the correlation of single alleles to cancer survival and prognosis.” Journal of Nucleic Acids Investigation 2010; 1:e9. On a personal note, I am primarily Scandinavian; that said, because ICT-107 works better against patients with HLA-A2, one can extrapolate the slippery slide ancestry and medicine might play if human discretion unduly interfered with antigen/genetic targeting selection for cancer therapies. DCVax avoids all that by using each individual patient's personal tumor(s) to provide immunotherapeutic targets. Hence, DCVax’s therapeutic spectrum is not unnecessarily burdened by a patient’s particular ancestry.
Still, for patients expressing all six of the targeted antigens (which is a very small percentage of the global GBM population), the phase I study demonstrated very strong results. Unfortunately for patients during enrollment, the phase II trial did not specifically select for patients whose tumors expressed most or all of these these six antigens, and it’s no coincidence that the survival response median amongst the entire group fell dramatically. There are two very important lessons that were confirmed due to the protocol difference between the phase I and II ICT-107 trials. First, it is important to target a higher number of abnormal tumor antigens with immunotherapy. Second, it is also important that the targeted tumors have the antigens intended to be targeted by the therapeutic intervention. Comparatively, DCVax-L already uses all tumor antigens, and because they are from the patient’s own tumors, a perfect match is inherently assured. If ICT-107 initiates a phase III trial, it will almost assuredly enroll patients with tumor profiles that match most or all of the six antigens it targets.
Conclusion:
One might be tempted to conclude that HLA-A2 allele targeting by ICT-107 plus EGFRvIII antigen targeting via Rindpepimut covers up to 55% of tumors within the global GBM population. This is incorrect for at least three reasons.
1. HLA-A2 and EGFR-vIII can overlap in tumor patient populations.
2. Tumors expressing both attributes are vulnerable to EGFRvIII peptide treatment.
3. Only a subset of the HLA-A2 population demonstrates clinically remarkable efficacy, and this subset only constitutes 10-15% of the GBM population. Also, as with Rindopepimut, ICT-107's niche is predominantly transient due to tumor escape — although complete longterm remission has been achieved in a very small patient population. (Notice tumor escape appears more difficult when more antigens are targeted by the therapeutic vaccine. ICT-107 targets six antigens, but Rindopepimut only targets one antigen.)
ii. Potential competition between DC-Vax and ICT-107.
With the latest results from the phase two ICT-107 trial, it is quite clear that patients with HLA-A2 positive methylated tumor expression can achieve impressive PFS survival — 15+ months beyond standard of care. OS survival has not matured yet, but they are likely to demonstrate equally striking results within the same subpopulation. However, this patient population only amounts to 10-15% of the global GBM population. An encouraging finding is that the more matching antigens a tumor expresses, the longer people avoid tumor recurrence. This confirms not only a qualitative confirmation for immunotherapy, but an additional quantitative relationship as well.
There does not appear to be any current literature discussing any possible efficacy data relating to DCVax therapeutic effects upon HLA-A2 tumors positive methylated MGMT, but theoretically, DCVax also targets the abnormal biomarkers associated with this mutation. Once again, one hopes that the DCvax-L phase III trial will test and report any and all HLA-A2 tumors, both positive and negative MGMT. Regardless, ICT-107 can only encroach upon (at most) 10-15% of DCVax bandwidth restricted to Glioblastoma. If one takes into account Rindopepimut, their combined market share is not likely to exceed 30% of the GBM market, and DCVax is likely to be utilized at some point in the treatment regimen for this subclass.
Importantly, neither Rindopepimut nor ICT-107 are targeted toward other glioma types/grades. NWBO technology can be used against all glioma types, and all solid tumors. Finally, ICT-107 has not yet initiated its phase III trial for Newly Diagnosed patients with GBM. In terms of potential regulatory approval, this places it 3 or more years behind DCVax-L and 2 or more years behind Rindopepimut.
iii. Potential Collaboration between DCVax and ICT-107.
Unlike Rindopepimut, there may be more concern in directly combining DCVax and ICT-107 during the same vaccination. Tolerance from excessive dosage can occur when too many dendritic cells are administered to a patient. It may be difficult, especially in the very carefully dosed and timed DCVax-Direct, to incorporate ICT-107. First and second line successive treatment may prove more valuable.
C. Prophage Series G-100
i. Potential Niche for Prophage series G-100
Agenus own the rights to Prophage series G-100. Prophage is a protein peptide complex consisting of a 96 kDa heat shock protein (Hsp), gp96, and an array of gp96-associated cellular peptides. The response to Prophage seems to be more pronounced in those patients with less expression of the checkpoint ligand PD-L1 on the white blood cells, suggesting that combinations of Prophage with checkpoint modulators like PD-antagonists might make Prophage even more effective in a greater percentage of patients with GBM. PD-L1 is expressed in approximately 25% of the GBM patient population. PD-L1 and PD-antagonists will be discussed further in Parts III and IV of this series entitled, “Immunomodulators."
Conclusion:
Theoretically, Prophage vaccines can target all cancers; however, no other marketable products are on the near horizon. Prophage is also currently being studied against melanoma in a phase I trial. Consequently Prophage's only niche appears to target GBM, perhaps with some focus on tumors not expressing PD-L1.
ii. Potential Competition.
Unlike the previous two therapies discussed, Prophage is not designed to target any preset tumor antigen(s). Instead, the proprietary process selects specific antigens in a patient’s specific tumor(s) that increase expression when cells are exposed to elevated temperatures or other stresses.
This process ultimately targets more antigens than Rindopepimut and ICT-107, and is personalized to the particular patient’s tumor(s); however, unlike DCVAX, Prophage does not use all the antigens from the tumor, and the personalized selection/activation process is not accomplished by the dendritic cell; instead the antigens are selected through a manmade proprietary process.
The results in the July phase II open label trial demonstrated 9 month overall survival advantage over historical controls, and PFS surpassed historical comparisons by 10 months. (Notice, again, as we move up the chain from 1, 6 and now many tumor antigens, the overall results and durability improve). Pragmatically, it is hard to imagine that Prophage will perform at this level in a phase III controlled trial, because larger randomized double blinded studies typically result in lower efficacy results than phase I/II trials. This phenomenon is normally attributed to the stricter enrollment protocol which prevents “cherry picking” patients that are more likely to do well regardless of therapeutic intervention. DCVax-L's (which utilizes all the tumor antigens) Phase II trials results demonstrated 21 months OS against the same historical control in their phase II trial, and PFS surpassed the above historical comparison by 17 months. Notice DC-Vax uses all the antigens and returns selection control and expression to the dendritic cell. While Prophage therapy actually offers a more straight forward head-to-head comparison, published data demonstrates that DCVax-L is superior.
A phase III GBM trial has not been initiated for Prophage. Like ICT-107, this places Prophage approximately three or more years behind DCVax-L.
iii. Potential Collaboration
IMHO, the two antigen targeting selection processes utilized by DCVax and Prophage are far too complex to integrate with each other.
