Investors Discounting Positive Data Read-Out on Brain Cancer Studies; A Paradigm Shift in the Making

ImmunoCellular Therapeutics (NASDAQ:IMUC) and Northwest Biotherapeutics (NASDAQ:NWBO) both have cell-based therapeutic cancer vaccines in late stage clinical studies for Glioblastoma Multiforme (GBM), the most common and lethal from of brain cancer. Data are expected from IMUC by the end of 2013 and NWBO in 2014 that, if positive, will change the treatment paradigm in a difficult and treatment-limited cancer. Ahead of these pivotal read-outs, we see considerable opportunity in both companies predicated on compelling Phase I data, low efficacy and regulatory thresholds, and multiple signals of immunotherapy efficacy in brain cancer.  While hinging expectations on early-stage data carries risk, both companies have potential to double running into top-line data and provide returns in the 10-fold range on favorable results, with downside limited by their comparatively cheap valuations. ImmunoCellular and Northwest Bio carry market capitalizations of roughly $120-150M fully diluted, and could easily become billion dollar companies on positive data. Take a look at a few of PropThink's undervalued picks from 2012, many of which have panned out well in January - Keryx Biopharmaceuticals (NASDAQ:KERX) returned 200% to investors, while Organovo (OTCBB:ONVO) more than doubled since we covered the stock in early December. For both the run-up trader and long-term investor, IMUC and NWBO offer quality risk/reward profiles with increased visibility and price appreciation in 2013 as the probability of trial success is better understood by the investment community. Nevertheless, we remain more interested in ImmunoCellular for the time being due to its strong balance sheet. Looking at the totality of the immunotherapy data in GBM, we find it difficult to escape the view that IMUC and NWBO both have very real prospects for positive, late-stage trial outcomes.

Market Opportunity

In the United States, primary malignant brain cancers (neoplasms) are responsible for over 13,000 deaths annually. Glioblastoma Multiforme (GBM) (WHO grade IV astrocytoma) is the most common, representing more than 50% of all glial cell-derived neoplasms, with a US incidence of 12,000 new cases per year. Mortality from GBM accounts for 3-4% of all cancer deaths annually. Only modest advancements in treatment of these tumors has been made and despite standardized therapy of surgical resection, radiation therapy, and temozolamide chemotherapy, GBM continues to be uniformly lethal, with a median survival of 15 months and a 3-year survival of only 10%. Due to limited treatment options for GBM patients, there is an urgent need for novel treatments such as immunotherapies.

Immunotherapeutic Approaches and Dendritic Cell Vaccines

Immunotherapies have struggled to translate from bench to bedside as effective clinical treatments, but the zeitgeist has changed in the last two years with the landmark approvals of Dendreon’s (NASDAQ:DNDN) Sipuleucel-T (Provenge) and Bristol Myers Squibb’s (NYSE:BMY) Ipilimumab (Yervoy); these approvals have validated the use of immunotherapy in oncology. Our understanding of the central nervous and immune systems has evolved and undergone major revision, leading to a paradigm shift in opinion that has led to much enthusiasm for the potential role of immunotherapy in GBM

The Imunocellular vaccine ICT-107 and the Northwest Bio vaccine DCVax-L are both active cellular vaccines - Dendritic Cell Vaccines. Dendritic cells (human derived cells responsible for antigen processing and presentation to the immune system) play a central role in the body’s immune response. They act as first responders that trigger the systems that help the body fight infection or foreign bodies by initiating a T cell. The dendritic cells (DC) do this by recognizing, processing and presenting foreign antigens (substances that stimulate the production of antibodies and combine specifically with them) to the T cells, which then effectuate the immune response. The goal of a cell-based vaccine is to: 1) make use of and enhance the DC’s ability to trigger the T cell response; and 2) stimulate the DC to focus the T cell response to specifically target the cancer cells for destruction. DC therapy generally involves harvesting peripheral blood mononuclear cells from a patient, then culturing and processing them in a laboratory to produce more numerous and effective DCs. In the laboratory, the DCs are cultured with specific tumor antigens to enable the DCs to recognize cancer cells as targets for attack. When the newly cultured DCs are reintroduced into the patient, they seek out remaining tumor cells and signal T cells to destroy them.

With ImmunoCellular Therapeutics’ ICT-107, monocytes from the patient are obtained and matured to DCs and then pulsed with a standardised panel of six tumour antigens. Studies have shown that all GBM patients expressed three or more antigens, and 75% expressed all six. ICT-107 specifically targets cancer stem cells (CD133 positive). Cancer stem cells are considered a subset of cancerous cells (representing less than 5% of all cells in a tumor), which are responsible for the growth and re-growth of the primary and metastatic tumors. Complete eradication of tumor masses requires elimination of these cells, which are resistant to standard chemotherapy and radiation therapy. If the cancer stem cells are not destroyed, the tumor can regrow after treatment. Patients must be HLA-A1 or HLA-A2 (human leukocyte antigen) positive for the creation of an immune response, making 50-75% of patients eligible for treatment with ICT-107.

Northwest Bio’s DCVax-L utilizes a patient’s DC’s and an extract of the patient’s own tumor cells to produce an immune response. Resected tumor from surgery is shipped to a manufacturing facility, as are specific blood cells (monocytes) obtained by a process called leukopheresis. The monocytes are matured to DC’s and pulsed with tumor lysate (broken down tumor cells) to sensitize them against the tumor antigens of the patient’s own tumor. DCVax-L is then shipped frozen to the clinic for administration to the patient. DCVax-L can be manufactured in sufficient quantities for treatment of at least one year, and often for two or more years. In approximately 10% of patients, there might be insufficient tumor sample (need a minimum of 2g) to prepare vaccine for two to three years of treatment, suggesting that DCVax-L can be used in 90% of GBM patients

Which vaccine is better?

Much has been made of ICT-107’s ability to specifically target cancer stem cells. Although attractive as a differentiator between the two vaccines, it remains to be proven whether this translates to a clinical advantage. The dominant clinical differentiator is then the percentage of the GBM population that is eligible for treatment. DCVax-L can be used in up to 90% of patients, and according to IMUC, ICT-107 in 65-75%. The epidemiological studies of the frequency of HLA-A1 and HLA-A2 positivity give a combined frequency of 50-75% in the population depending on ethnicity, but in the Phase I and ongoing Phase IIb study, the trial figure is at the low end of the range of 50%. This lower figure in the clinical studies is due to the disease progressing in patients before they are randomised to treatment, as well as patient drop out for other “random” reasons.

Evidence Supports Clinical Effect and Positive Trial Outcomes

Oncology studies are notorious for Phase III trial failures following promising Phase I and Phase II data. The reasons are manifold, but we will focus on the available data, put it into clinical context, and draw on other strands of evidence to demonstrate why the Phase I data for ICT-107 and DCVax-L are so compelling.

  1. The best way to appreciate just how outstanding the efficacy signals are from the Phase I studies of DCVax-L and ICT-107 in newly diagnosed GBM patients is to compare them to the treatment arm of the 2005 paper that defined the current standard of care.

    Number of GBM Patients in study Standard of Care287 DCVax-L20 ICT-10716
    Median Progression Free Survival (months) 7 23 17
    Median Overall Survival (months) 15 35 38
    2 Year Survival % 27 - 80
    3 Year Survival % 16 55 55
    4 Year Survival % 12 30 50
    Median Age 56 51 52

  2. The Phase I studies are small, uncontrolled and single center, but it is very difficult not to be captivated by the profound increase in median overall survival in both the ICT-107 and DCVax-L studies. For both vaccines, the median overall survival more than doubled, and three times as many patients were alive at three years compared to the standard of care. The investment community have every right to be skeptical given the small patient numbers and usual Phase I trial shortcomings. We discussed the data with Dr. Laszlo L. Mechtler, who serves as Medical Director of the Neuro-Oncology Department and the Headache Center at DENT Neurologic Institute. Dr. Mechtler has a good understanding of immunotherapies, being a leading recruiter in Celldex’s (NASDAQ:CLDX) vaccine studies for EGFRvIII-expressing GBM tumors. As with neurosurgeons that we have spoken to, he is unable to attribute the outstanding survival from either study to age or confounding factors. He said, “I have never seen results like these in GBM.” What excited Dr Mechtler beyond the overall survival data was the fact that this is not chemotherapy; cancer vaccines maintain quality of life with a very favorable side-effect profile. He says that these vaccines “offer the vision of hope.”

  3. Even if the trial coordinators were biased and recruited the most favorable prognostic patients to the trials, something that is very difficult to do with the GBM population, it is most unlikely that these outcomes could have been achieved. Put another way, these are the most exciting data ever seen in GBM

  4. What is even more compelling is that two different DC vaccine approaches, led by two different teams at different clinical centres, have produced very similar results without any safety concerns.

  5. The IMUC Phase IIb trial is further de-risked by improving DC function, timing, and frequency compared to the phase I trial.

  6. One of the problems with oncology studies, and immuno-oncology in particular, are the long study time-frames, where-by the standard of care can change. The standard of care for newly diagnosed GBM patients was based on a study published in 2005 and has not changed over the course of the IMUC Phase I study, nor has it changed for the ongoing studies.

  7. The Phase I studies were not powered for efficacy, had no control group, and were compared to historical controls. The use of historical controls is normally a serious cause for concern in oncology, but in the GBM population, it is not unreasonable given the plateau reached in treatment. This very significantly reduces the risk of the current trials failing for increased longevity in the trial control populations. At best, there has been no more than a 2-3 month increase in survival since the standard of care was defined.

  8. Without therapy, patients with GBM uniformly die within 3 months. Patients treated with optimal therapy have a median survival of approximately 15 months, with 27% of patients surviving 2 years, 16% to three years, and fewer than 10% of patients surviving to 5 years. An increase of two months in survival is considered to be clinically meaningful, particularly for immunotherapy, as it does not have the toxic side-effects of chemotherapy. Both vaccines increased median overall survival by 20 months, allowing a large margin of safety for the ongoing studies.

  9. Work on cellular-based immunotherapies in GBM has not been limited to IMUC and NWBO, with a number of academic groups conducting small clinical studies in newly diagnosed and recurrent GBM. Almost 15 studies have reported on a survival benefit of patients receiving various immunotherapies when compared to historical controls. It is very unlikely that small numbers, historical controls, single clinical sites, or trial bias could have led to so many trials demonstrating a survival benefit.

  10. One area that has proven to be difficult has been defining and measuring clinical correlates of vaccine-induced immune function. IMUC have encouraging data demonstrating improved progression free survival with antigen expression and lower numbers of cells positive for CD 133 (cancer stem cells) after treatment with ICT-107

In the context of oncology studies, looking at the totality of the immunotherapy data in GBM, we find it difficult to escape the view that ICT-107 and DCVax-L both have very real prospects for positive trial outcomes. 

Immunotherapy has shown itself to be very safe in GBM trials, with no evidence of autoimmunity or other major safety issues. None of the trials involving DCVax-L or ICT-107 demonstrated any grade IV events, and in the aforementioned fifteen studies that reported a survival benefit, only one grade IV event occurred in 316 patients. Furthermore, no dose-related toxicity has been demonstrated. This is in contrast to the very significant incidence and burden of complications from chemotherapy on the GBM population. It is not easy to foresee any major safety issue coming to light at this stage. To date, the safety profile of dendritic vaccines has been more than acceptable, not just in the GBM studies, but also from the thousands of patients who have received Provenge.

The Trials and How They Differ 

ImmunoCellular’s Phase IIb study of ICT-107 is a randomized, placebo controlled, double-blind trial and has a primary endpoint of overall survival, by far the most meaningful oncology endpoint (25 trial sites). In September 2012 IMUC announced that enrolment of 278 patients in the Phase IIb study had been completed, with a total of 124 patients being randomised, approximately two-thirds (83) treated with ICT-107, one-third with DC’s but no antigens as placebo. There will be an interim analysis for safety and futility after 32 events, which is predicted to happen in Q1 2013, and top-line data are expected towards the end of 2013. A more accurate prediction of a final read-out will be possible once the 32 events have occurred.

The DCVax-L Phase III study is a 300 patient, randomized (2:1) double blind placebo controlled study with a primary endpoint of progression free survival (70 trial sites). Overall survival is the most robust endpoint, as there can be a degree of variability in the radiological diagnosis of progression, and an increase in progression free survival does not always translate to an increase in overall survival. NWBO has been on financial life-support for a number of years, which may have influenced the choice of trial endpoint, as PFS provides an earlier read-out than overall survival (OS). The PFS endpoint will be satisfied by an extension of 6 months PFS, which is only 1/3 of the increase in PFS demonstrated in the Phase I trial.

The Phase I data from both the DCVax-L and the ICT-107 trials show a very good correlation between OS and PFS as well as 200 patients in the treatment arm should diffuse concern regarding the PFS end point. The two graphs below from the NWBO 2013 investor presentation illustrate the dramatic effect on both PFS and OS. It should be noted that the Phase I data have been combined with the graph from the treatment arm from the landmark study in 2005 of Temozolomide, so it is not a contemporaneous control group. Notwithstanding academic niceties, the combined data for the statistics and graphs are impressive. We think that NWBO’s choice of PFS will be acceptable to the FDA despite OS being the preferred primary endpoint.

Of more concern is that the study is a cross-over design, such that after documented disease progression, patients will be offered treatment. Most patients with GBM are likely to cross-over to DCVax-L given the lack of any other good second-line treatment. This will have the effect of making it difficult to determine what the overall survival of the control group would have been compared to the primary treatment arm, given the small number remaining not receiving DCVax-L. In essence, the trial will compare survival post standard of care treatment and early DCVax-L treatment, with subjects who are given DCVax-L late after progression on the standard of care treatment. This is the same design as the Phase III Provenge trial and the reason why the beneficial effect was lower than if there had been a standard control group, as there can still be benefit from immunotherapy even if a disease has progressed. Top-line data are expected in 2014

In GBM, Regulatory Path is a Path of Least Resistance

The efficacy threshold in GBM required by the FDA is low. Median survival of patients treated with Gliadel wafers was 13.9 months vs 11.6 months in the control group, and this 2.3 month difference was sufficient for FDA approval in 2003. Two years later, Temozolomide received FDA approval with a median survival of 14.6 months (radiotherapy plus temozolomide) and 12.1 months with radiotherapy alone, a 2.5 month increase in overall survival. Avastin was approved in 2009 for recurrent GBM on the basis of two open label Phase II studies, neither of which demonstrated an improvement in overall survival, just a short-lived tumor response in a minority of patients.

We do not consider the use of PFS instead of OS in the DCVax-L study as posing a huge regulatory risk given the trial size, good correlation between PFS and OS in the Phase I GBM studies and precedence from the approval of Avastin for recurrent GBM (decreased tumor size in less than 30% of patients) without evidence of improved survival. Given the size and multi-center nature of both studies, if the results of either trial are even half as good as the Phase I studies, it will be difficult for the FDA to deny patients a life-prolonging treatment when they have no other options. Therefore, we do not think that Phase III confirmatory studies will be required by FDA before the vaccines are made commercially available. However, the ICT-107 Phase IIb study is significantly smaller that the DCVax-L Phase III study with there being approximately 83 and 200 patients in the treatment arms respectively. If ICT-107 improvement in overall survival is less than six months then this significantly increases the likelihood of the FDA requiring a Phase III study.

Like what you see? PropThink's Email Subscription offers up periodical reports like the one above. And our Premium Subscription gives you exclusive content and actionable articles ahead of their public release. Right now, try a free 30-day trial.

Importantly, DCVax-L and ICT-107 are not Provenge

DCVax-L and ICT-107 are next generation dendritic vaccines and should not be viewed as suffering from the Provenge shortcomings. Both vaccines have the advantages of significantly lower cost and complexity:

  • Shorter production cycle time

  • Cryostorage

  • Enough vaccine (20-30 doses) from one manufacturing run to produce 2-3 years of vaccine compared to 1 dose per run for Provenge

  • Greater purity and potency of dendritic cells (80% v 20%)

  • Greater efficacy and targeting by multiple tumor antigens rather than a single antigen (Provenge single antigen)

  • ICT-107 has a COGS of 5-10% compared to 70% for Provenge.

  • DCVax-L and ICT-107 can be injected under the skin rather than having to be infused intravenously like Provenge

For Many, Financials Will be Key Factor

With just under $30M in cash after a recent $19M financing and a cash burn of $3.5M per quarter, ImmunoCellular has a healthy balance sheet with no outstanding debt; a capital raise ahead of the Phase IIb data is unlikely.

Long-term NWBO investors have suffered repeated dilution from numerous financings. At the end of 2012, NWBO completed a $13.8M offering, a 16-for-1 reverse split, and up-listed to the NASDAQ. Given a quarterly burn rate of $5.7M, Northwest Bio’s current $12.6M cash balance is insufficient to fund operations through the mid-2014 Phase III data. Slow trial recruitment would further stress finances, and another offering is to be expected this year. Despite the need to raise further capital, NWBO has transformed its balance sheet and removed most of the $38M current and long-term liabilities. An up-front payment from the partnering of DCVax-Prostate could provide non-dilutive financing now that the data are available for Xtandi and Zytiga in metastatic prostate cancer, but immunotherapy partnerships are unpredictable. That said, Linda Powers, NWBO’s CEO, has demonstrated a determination to get over the Phase III data goal line. Ms. Powers also acts as a director of Cognate Bioservices, a DCVax-L manufacturing facility, and is a managing director of Toucan Capital, a small venture fund that recapitalized NWBO in 2004 and ultimately went on to became a major shareholder. So intimate is this relationship that in 2009, Cognate agreed to convert accounts payable to equity in NWBO. Ms Powers and Toucan Capital were majority shareholders before NWBO’s most recent offering, and through creative funding arrangements, Powers has kept the small immunotherapy developer afloat; with so much at stake personally, we expect Powers and Toucan will make a concerted effort to keep NWBO in the black.

Pricing, Reimbursement, and Valuation

DCVax-L and ICT-107 are likely to be high-premium products, although final product pricing will be heavily influenced by efficacy. Even if efficacy is half as good as the previous Phase I studies, these vaccines are likely to raise the benchmark for pricing of a new therapy. Provenge costs $93,000; Avastin is about $100,000 a year; and Yervoy (increase of two months overall survival) is $120,000 per course of treatment for metastatic melanoma. These figures set the lower boundary for pricing per patient for the first year of treatment. DCVax-L is given seven times in the first year and then two shots per year for the second and third years. ICT-107 is initially given as a primary course of four vaccinations and then four maintenance doses during the rest of the first year, followed by further maintenance doses afterwards. If the Phase I studies are emulated, then 50% of patients survive to three years and many beyond, thus adding a significant maintenance therapy upside.

Given the fixed manufacturing cost (single run produces 2-3 years supply even though only 50% of patients survive to three years) it is reasonable to assume that the first year of treatment will be priced significantly more than subsequent maintenance years. The first year of treatment could be in the order of $120,000 - $150,000, and yearly maintenance at $50,000 - $80,000. Further, the much greater beneficial effect (increase in median OS of 20 months compared to Provenge’s four month increase) on GBM survival with no real alternative therapeutic options suggests that these figures are eminently acceptable. If FDA approval is obtained, we see little reimbursement resistance, but the focus will instead be on off label use. Off label use might extend to Grade III astrocytomas (27% four year survival) or recurrent GBM.

In the IMUC and NWBO scenarios, we are looking to conservatively estimate the lower end of the payback multiple range. The downside calculation is easy, in that NWBO goes bankrupt and IMUC will trade at cash or less.

U.S. incidence of GBM is 12,000 per year

Population suitable for treatment with DCVax-L (75% conservatively) is approximately 9,000 patients

Revenue per patient estimated at $120,000 annually

DCVax-L revenue per year is then 9,000 x $120,000 ≈ $1.1B

HLA-A1 HLA-A2 restrictions mean that only 50% of 12,000 patients are suitable for ICT-107 (non-trial figure will be >50%)

Revenue per patient estimated at $120,000 annually

ICT-107 revenue per year 6,000 x $120,000 ≈ $720M

As a rule of thumb, the global market is 1/3 U.S., 1/3 E.U., and 1/3 rest of the world, so it would not be unreasonable to double the above figures considering ex-U.S. markets, giving NWBO $2.2B or IMUC $1.5B in potential revenue. In addition, this calculation does not include maintenance revenue, therefore by no stretch of the imagination there is blockbuster potential for both companies. 

We can now project a basic revenue range for each company: $250M-500M for IMUC and $750M-1B for NWBO. These projections are very conservative and we expect revenues could be well in excess (double), but the purpose was to define the lower end of the range. The fully diluted market capitalization of NWBO is approximately $120M, and IMUC $150M ($85M and $106M non-diluted). After another capital raise, Northwest Bio’s fully diluted market cap will be approximately $150M. As can be seen, these are potentially billion dollar companies in the making. Both companies then offer the opportunity for a 5-10 fold return at the very least, with a much shorter holding period than the typical venture capital type of investment, as data are due towards the end of 2013 and mid-2014.


For ImmunoCellular, the announcement of trial continuance after the 32 events that trigger the interim is expected this quarter and could see IMUC trading in the $2.50 - $3.00 dollar range, followed by a further run into data towards the end of the year. The same strategy is more difficult to apply to NWBO because it will dilute in the near-term, and average trading volume is considerably lower. Positive IMUC data in late 2013 should validate the immunotherapy approach to GBM and cause a pop in NWBO. On positive data, NWBO is highly likely to be up for sale given its finances and that management and Toucan are majority owners. IMUC, on the other hand, will likely partner ICT-107 or commercialize alone. Overall, we see considerable opportunity in these two companies based on compelling Phase I data, low efficacy and regulatory thresholds, and multiple other immunotherapy efficacy signals in GBM. In short, the market significantly underestimates the likelihood of positive trial outcomes as reflected in current valuations. We remain far more focused on IMUC in the near-term based on the company’s quality financials and management team. Nevertheless, NWBO and DCVax-L hold interest because of the immunotherapy’s larger eligible patient population - 90% vs. 65-75% and Phase III trial designation. No studies have ever produced such strong data in brain cancer, and GBM presents a huge unmet medical need, with no change in standard of care for eight years. ICT-107 and DCVax-L both have blockbuster potential, and their respective developers, ImmunoCellular and Northwest Biotherapeutics, multi-bagger prospects.

In connection with IMUC, PropThink has taken a long position.