I initiated coverage of Agenus (AGEN) with a Buy on July 17, 2013 based on the pending release of clinical trial data on five development programs over the next three quarters. Each of these events, some more than others, has the potential to boost the stock at the time the data is reported. Given the inherent risk of clinical trials, it is possible that one or more (possibly all) of these could produce disappointing results. However, this breadth of opportunity makes Agenus an intriguing biotechnology story. It is usually the case that the prospects for a company are determined by just one binary event that may be in the distant future (a year or two). Agenus has five in the next three quarters.
I start this report by going over the events to focus on over the balance of 2013 and early 2014. I have put a particular focus on the results of the MAGE A-3 vaccine as the DERMA Phase III trial in melanoma is likely to be reported in by the end of 2013 and should be the first of the string of reports. I am only going to touch briefly on other parts of the Company and for more details; I would suggest that investors read my basic report.
I see the upside potential for Agenus in the event of a success in the MAGE A-3 melanoma trial as being $10.00 and the downside as being $2.00. I know that the next question is what are the chances for success? I can't easily answer that question other than to say that the biological hypothesis behind the MAGE A-3 vaccine is strong and that the trial is extremely well designed and conducted. Let me just say that there is a reasonable chance for success.
If the MAGE A-3 trial in melanoma fails, it would be a major disappointment. However, there would remain significant promise in the Prophage program in glioblastoma, the HerpV vaccine for genital herpes and the 18 other vaccines that Glaxo (GSK) is developing using Agenus's vaccine adjuvant QS-21. On the negative side, Agenus would need to bring in more cash and this could entail an equity offering at depressed prices and/or possible partnering of the Prophage and HerpV programs.
Upcoming Clinical Results for Agenus
Of the five near-term clinical events four are likely to be reported in 2013 and one in early 2014. I discuss these in the order in which I expect they will occur.
3Q, 4Q 2013 Glaxo is developing the MAGE A-3 cancer vaccine for use as adjunctive therapy in surgically resected stage IIIb and IIIc melanoma patients. This vaccine uses Agenus's QS-21 in combination with other immunostimulants as an adjuvant. If this trial is successful, I am estimating sales of $500 million in 2019 and that Agenus would receive $18 million of royalties on this level of sales
4Q, 2013 Glaxo is developing RTS,S; this is a vaccine for malaria that uses QS-21 as an adjuvant. There are no vaccines for malaria, which is the cause of more human deaths than any other infectious disease. Interim results on the Phase III trial of this vaccine have been unclear with some positive and some not so positive data; investors are neutral to negative on potential for success. More definitive data will be reported in 4Q, 2013 or possibly in early 2014. If the data is positive, this vaccine could become part of the normal vaccination program in sub-Saharan Africa with this process beginning in 2015. Assuming success, I am estimating sales of $200 million in 2019 and that Agenus would receive a royalty of $3 million at that time.
4Q, 2013 Prophage is Agenus's cancer vaccine that is based on its heat shock protein technology. Encouraging investigator sponsored studies have led to the recent initiation of a Phase II trial in 222 recurrent glioblastoma patients. The full results of an earlier trial in recurrent glioblastoma that was the basis for this trial should be published in 4Q, 2013 allowing investors to see the full data set for the first time. This could spark more serious investor attention to this program. I think that we can assume that these results will be encouraging as the National Cancer Institute, already having seen the results, decided to fund most of the estimated $21 million cost of the new Phase II trial. Agenus is only responsible for manufacturing the vaccine.
The new Phase II recurrent glioblastoma trial could report topline results in 2015 and with highly positive results could be the basis of a registrational filing in 2015 or 2016. A separate and distinct Phase II trial in newly diagnosed glioblastoma patients could begin in early 2014. Agenus has full rights to Prophage in these indications. This could be a very important asset for the Company that could potentially allow it to transform into a commercial enterprise.
I believe that the recurrent glioblastoma market is currently about 2,000 patients and could represent an addressable market of $120 million or more in the US. Northwest Biotherapeutics (NWBO), ImmunoCellular (IMUC) and Celldex (CLDX) are each in Phase III or II programs in newly diagnosed glioblastoma and IMUC is beginning a Phase I trial in recurrent glioblastoma. This suggests that Agenus has a development lead in trials for recurrent glioblastoma, but is somewhat behind in newly diagnosed glioblastoma.
4Q, 2013 Agenus has developed HerpV, a vaccine for genital herpes that is based on its heat shock protein technology. Unlike Prophage, this product uses synthetic antigens and would be an "off the shelf" product. Results of a 65 patient Phase II should read out in 4Q, 2013. The purpose of this trial is to show that HerpV can reduce viral shedding and thereby sexual transmission of the disease, which is estimated to affect 60 million Americans. Agenus is at the forefront in developing this type of vaccine and promising results could lead to a partnering deal with a meaningful upfront payment in 1H, 2014.
1Q, 2014 The MAGE A-3 vaccine is also being developed for surgically resected stages Ib, II or IIIa non-small cell lung cancer. If this trial is successful, I am projecting sales of $1.6 billion in 2019 and Agenus would receive a royalty of $59 million at that time.
There is also another trial that will read out data in about one year, 4Q, 2014, on a Glaxo vaccine for herpes zoster, commonly known as shingles. The vaccine is being developed under the premise that it will be superior to the current market leader Merck's Zostavax and replace much of that product's sales. In that case, I think that sales of $750 million could occur in 2019 and Agenus would receive $12 million of royalties.
A Close Look at MAGE A-3 Vaccine
The results of the MAGE A-3 trial in melanoma are widely anticipated by investors following Glaxo, Agenus and biotechnology in general. The trial is a large, landmark study as adjuvant therapy for newly resected melanoma patients. If successful, the MAGE A-3 vaccine would quickly become standard of care in the adjuvant setting where there are no approved products. This would be a striking advance in cancer therapy. The impact with success on Glaxo would be enormous and much more enormous. Glaxo is also conducting a Phase III trial in non-small cell lung cancer and Phase II trials in bladder cancer, head and neck, gastric, hepatocarcinoma, and multiple myeloma. Investors might leap to the conclusion that the MAGE A-3 vaccine could be incorporated in the adjuvant settings in these cancers as well and look at MAGE A-3 as being a multi-billion sales opportunity for Glaxo. In this event, investor excitement would be enormous I would expect a dramatic move, based on both the potential for huge royalty payments, fervent investor psychology and short covering that could drive Agenus to $10.00 or more.
Biotechnology investors know that nothing is for sure in clinical trials so what happens if the MAGE A-3 trials fail in melanoma and non-small cell lung cancer? The immediate effect of a failure in melanoma might be to knock the stock down, but this decline might be ameliorated by the prospects for Prophage in glioblastoma, some small hope for the malaria vaccine RTS S and the HerpV genital herpes vaccine. Also longer term, Agenus has the potential to receive royalties on 18 other Glaxo vaccines that are in various stages of development. I see these as modifying the impact on the stock although I think that failure could cause the stock to decline with the release of headline news to perhaps $2.00 per share just as a kneejerk reaction.
Investors are also concerned that although Agenus has gone through a financial restructuring that has cleaned up the balance sheet, it is still in an uncomfortable cash position. It had $13 million at the end of the 2Q, 2013 and at the current burn rate of $4 to $5 million per quarter will be down to $2 million at year end. Agenus does not burn much cash as the Prophage and Glaxo vaccines require little capital. Most of its spending is on the HerpV Phase II program for genital herpes.
The Company clearly needs to bring more cash into the Company. With success in the MAGE A-3 melanoma trial, I think the market would be very receptive to an equity offering at substantially higher prices than current. With failure in MAGE A-3, the Company could be forced into an equity offering at depressed prices although it might be able to monetize the royalty stream of the malaria vaccine or partner HerpV and/or the Prophage glioblastoma program.
What Are the Chances for Success in the MAGE A-3 Melanoma Trial?
The following section lists the reasons to be hopeful of positive results in the Phase III MAGE A-3 melanoma trial called DERMA
- All too often cash strapped biotechnology companies cut corners in terms of timelines and patient enrollment in clinical trials, but DERMA is a large and well conducted study.
- The trial was started in December of 2008 and enrolled 1,349 patients with two thirds on the MAGE A-3 vaccine and one third on supportive care. Large well powered and designed trials significantly enhance the chances for success.
- The trial was done in patients who had been rendered disease free by surgery prior to randomization. Researchers have come to believe that immunotherapy works best when used in the early stages of cancer. The long record of failures of earlier cancer vaccines may have been due to their being used to treat patients with advanced disease. If this hypothesis is true, MAGE A-3 may have a much higher chance for success than the earlier cancer vaccine trials.
- The treatment in this study was sustained over a long period of time as patients received 13 intramuscular injections over 27 months. It is believed that immunotherapy works slowly so that providing this long course of therapy could allow the treatment to take effect and remain in effect.
- About 2/3 of melanoma patients express adequate amounts of MAGE A-3. The DERMA trial selects for adequate MAGE A-3 expression rather than taking on all comers. Again, this enhances the chances for success.
Issues to be concerned about include the following;
- Cancer vaccines are a paradigm shift in technology. With that comes great promise, but also great risk and uncertainty. It seems highly plausible that boosting the immune response to cancer will result in a therapeutic benefit, but this remains hypothesis rather than established science. There may be some unknown biological factors that arise to thwart the hypothesis and the trial that only become apparent after the fact.
- Many scientists believe that cancer vaccines should target multiple antigens unlike this vaccine that targets just MAGE A-3. Over time, it may be the case that this vaccine destroys cancer cells with MAGE A-3 expression, but has no effect on non-MAGE A-3 expressing cells so that they become the dominant population of cancer cells and the cancer progresses. This happens frequently with chemotherapy and targeted therapy agents.
How Does the MAGE A-3 DERMA Trial Differ From the Failed Allovectin Trial?
The disappointment with Vical's (VICL) Phase III trial of Allovectin in melanoma has caused some investors to fear that this increases the potential for failure in the DERMA trial of MAGE A-3. However, there are significant differences in trial designs and patient population as follows:
- The Allovectin study included later stage melanoma patients while the DERMA trial was in earlier stage patients.
- The DERMA trial was in patients who had their tumors surgically removed and were considered disease free. The Allovectin patients did not have their tumors removed,
- Allovectin accepted all comers while DERMA accepted only patients with adequate MAGE A-3 expression. DERMA is more targeted.
- The DERMA trial enrolled 1,349 patients as compared to 390 in the Allovectin trial.
- The DERMA trial used a powerful adjuvant that contains QS-21 while Allovectin did not.
- The DERMA study was designed and enrolled by Glaxo, the number one vaccine developer in the world, while the much less experienced and less well funded Vical conducted the Allovectin study
Glaxo's MAGE A-3 Cancer Vaccine
MAGE-A3 Is the Basis for Glaxo's Cancer Vaccines
The body's innate and adaptive immune systems have the inherent ability to target and kill cancer cells, but sometimes this fails to prevent a cancer from occurring. The hypothesis behind cancer vaccines is that they can enhance the immune response and in doing so can effectively treat cancer. This is done through the inoculation of tumor specific antigens that are recognized by the immune system and which then seek out and kill cells that express that antigen.
Glaxo has developed a vaccine for cancer that uses MAGE-A3 as the sole antigen to elicit an immune response. It is a fusion protein of MAGE-A3 and Hemophilus influenzae protein D that is produced recombinantly in Escherichia coli; it is then combined with AS02, a proprietary mixture of adjuvants that includes Agenus's QS-21. It is being tested in Phase III trials for non-small cell lung cancer and melanoma and Phase II trials in bladder cancer, head and neck, gastric, hepatocarcinoma, and multiple myeloma.
MAGE A-3 has some compelling properties for use as an antigen in cancer vaccines. The normal function of MAGE-A3 in healthy cells is unknown. Importantly, it is only expressed in the testes and the placenta, which do not have HLA molecules that display proteins to the immune system. However, it is widely expressed in a variety of tumors. As a result, the immune system only sees MAGE A-3 on cancer cells and only launches an immune response specific to cancer cells. It is a genuinely tumor-specific target for tumor specific immunotherapy. MAGE-A3 is expressed in up to 76% of melanomas, 35% to 50% of non-small cell lung cancers, 30% to 58% of bladder cancer and 24% to 78% of liver cancers.
Glaxo initially chose to conduct Phase II trials in non-small cell lung cancer and melanoma. Results in these trials encouraged GSK to move to large Phase III trial. In the following sections, I discuss the Phase II trial results.
Phase II Results in Surgically Resected Non-Small Cell Lung Cancer
Design of the Trial
A Phase II trial in stage Ib and II non-small cell lung cancer was conducted in surgically resected patients who were confirmed as having MAGE-A3 expression based on a sample of removed tumor tissue. The enrollment criteria required that patients had to have meaningful MAGE A-3 expression in their cancer cells. They also must have been surgically resected four to eight weeks earlier and to have recovered from their surgery with a good performance status.
The vaccine was given by intramuscular injection with the first vaccination given 4 to 8 weeks after surgery. This was followed by four injections at three week intervals and eight more at three month intervals, Hence, there were roughly 8 doses in the first year and 5 in the second
Computer tomography was used to determine if the tumors regrew. The first scan was performed 4 to 8 weeks after surgery, at three-month intervals for the next 18 months and then at six-month intervals for the next 42 months. The scans were evaluated by the investigator or a radiologist blinded to the treatment assignment. Disease free progression was defined as: (1) the time from the date of surgical resection to the date of recurrence or (2) to the development of a second lung neoplasm. Standard definitions of disease free survival and overall survival in accordance with RECIST criteria were used.
Enrollment in the Trial
A total of 1,089 patients with completely resected non-small cell lung cancer were screened for expression of the MAGE-A3 antigen. Significant MAGE-A3 expression was observed in 363 tumors or 33% of those screened. Of these, 183 patients were enrolled in a trial that started in May 2002. The trial was randomized to 122 actively treated patients and 61 control patients who were just given placebo and observed.
Results in the Trial
Following a median follow-up period of 44 months, there was cancer recurrence in 43 (35%) of the 122 actively treated patients and in 26 (43%) of 61 placebo treated patients. This was a proof of concept trial that was not powered to achieve statistical significance. The p-value was 0.122 and the hazard ratio was 0.75 indicating a 25% reduction in risk of disease recurrence for the drug group. The 25% reduction in the risk along with the drug's good safety profile was considered encouraging and was the basis for Glaxo's decision to move forward into Phase III.
The choice to conduct this trial in the adjuvant setting (just after surgery) in the early 2000s was unusually prescient on the part of Glaxo. At the time, most cancer vaccines were tested in the same way as chemotherapy and targeted therapy drugs. They were given late in the disease state and in combination with standard of care. Glaxo elected to conduct studies in earlier stage patients who were not receiving drug therapy. Subsequently, other studies have shown that earlier stage patients with less tumor burden are more likely to benefit from immunotherapy and presumably cancer vaccines.
A late divergence was seen between the drug and placebo groups in this study which was in line with studies with Yervoy (ipilimumab) in melanoma and Provenge in prostate cancer. The delay could be the result of a longer time required to induce an immune response and subsequent clinical activity. Alternatively, it could be that only a subset of patients may respond to immunotherapy because of differences in characteristics of the tumor or its immune microenvironment.
A possible reason for tumor cells to escape immune recognition could be due to their loss of tumor antigen expression or antigen drift. This is a significant concern in cancer vaccines that use one or just a few antigens. In this study, the authors said that only a few biopsies (no number given) of recurrent tumors could be obtained and in this group only one patient had lost the expression of MAGE-A3.
Phase II Study in Melanoma
Design of the Trial
Previous studies conducted with the MAGE-A3 vaccine in melanoma showed only limited immunological and clinical responses. To improve the immunogenicity of the MAGE-A3 vaccine, Glaxo developed a proprietary adjuvant AS02 which combined Agenus's QS-2 with monophosphoryl lipid A (a TLR-4 agonist) in an oil/ water mixture. The combination with AS02 resulted in a more efficient elicitation of humoral and cellular responses, but the clinical benefit remained limited.
Glaxo then went on to further develop an even more effective immunostimulant AS15 which was a combination of QS-21, monophosphoryl lipid A and CpG 7909 (a TLT agonist) in a liposomal formulation. This adjuvant was shown to be more effective in preclinical models. A randomized Phase II trial was then conducted in first line metastatic melanoma to evaluate the safety, clinical activity and immunogenicity of MAGE-A3 vaccine combined with the adjuvants AS02 or AS15. The objective was to find the most effective adjuvant for subsequent Phase III clinical development.
In earlier clinical studies in metastatic melanoma, immunization against the MAGE-A3 antigen was followed by a few, but long-term objective responses almost all in patients with cutaneous or lymph node metastases. This was the population included in the Phase II study. The need for a potent adjuvant to strengthen the immune response against MAGE-A3 had previously been shown in non-small cell lung cancer trials. The first six doses were given at two-week intervals. This was followed by six doses at 3-week intervals, a dose every three months during year two and then every six months during years 3 and 4.
This randomized Phase II trial evaluated the two different immunostimulants combined with MAGE-A3 vaccine to determine if a more robust and persistent immune response would be associated with increased clinical benefit. The primary endpoint of the trial was safety with tumor response according to RECIST criteria being a secondary endpoint. Tumor response was assessed at 13, 23, 32 and 54 weeks and then every six months for the next three years by physical examination, lesion photography, and ultrasound and computed tomography. Patients were allowed to continue immunization even if the tumor progressed if the status was not aggressive.
Patients Enrolled in the Trial
Patients with MAGE-A3 positive stage III or IV M1a melanoma were randomly assigned to receive the MAGE-A3 protein combined with either the AS02 or AS15 adjuvant. Patients entered into the trial had to have significant MAGE-A3 expression as assessed by reverse transcriptase polymerase chain reaction on a tumor biopsy taken less than or equal to 12 weeks before enrollment. Between May 2005 and January 2007, 165 patients were screened and 59% had MAGE-A3 positive tumors resulting in 72 eligible patients being enrolled in the trial.
Results of the Trial
In the MAGE-A3 combined with AS15 arm, clinical activity was higher and the immune response was more robust than in the MAGE-A3 combined with AS02 arm. Therefore the AS15 immunostimulant was selected for the combination with the MAGE-A3 protein in the Phase III trials.
Median progression free survival was 3 months in both treatment arms. The one year progression free survival was 19% in the AS15 arm versus 3% in the ASo2 arm. Median overall survival was 31.1 months in the AS15 arm and 19.9 months in the AS02 arm. The hazard ratio was 0.55 suggesting that AS15 reduced the risk of recurrence by 45% as compared to AS02. After a median observation period of four years, the two year overall survival rate was 59% in the AS15 arm and 37% in the AS02B arm and the four year overall survival rates were 40% and 16% respectively.
The injections were well tolerated and the rates of adverse events considered related to the treatment were comparable in both arms. Almost all of the adverse events were of grade 1 or 2 and resolved within 2 to 3 days and most were related to injection site reactions (pain, redness and swelling) and systemic symptoms (fever, fatigue, chills and nausea), No patient left the study for safety reasons. No immune related adverse events were observed.
The unmet medical need in melanoma is for more effective postsurgical therapy in stage III disease. High dose interferon alpha remains the most active immunologic adjuvant evaluated to date and is indicated for high risk melanoma. However, meta-analysis has suggested that interferon alpha does not improve overall survival as compared with observation.
As in non-small cell lung cancer trials, there was a late divergence of the overall survival curves. This indicates that there was a slow onset of clinical benefit in several patients, some of whom initially experienced disease progression. The authors hypothesized that this might be due to time required to mount an effective immune response. This is in contrast to chemotherapy or small molecules targeted therapies such as B raf inhibitors that can cause immediate tumor shrinkage. Delayed clinical benefit has been described for other immunotherapeutic approaches as well, including Yervoy (ipilimumab) and Provenge.
The MAGRIT Phase III Trial in Non-Small Cell Lung Cancer
The MAGRIT phase III trial started in October of 2007 and the estimated primary completion date is July 2013 according to ClinTrials.gov. The patients will be followed until September 2022. It enrolled approximately 2,289 patients with surgically resected stage Ib, II or IIIa NSCLC who have tested positive for sufficient expression of MAGE-A3. This is the largest trial ever conducted in non-small cell lung cancer.
Patients were randomized 3:1 between the MAGE-A3 vaccine, MAGE-A3 and chemotherapy and placebo. The new and more effective AS15 adjuvant was added to the MAGE-A3 vaccine. As in Phase II, each group received 13 intramuscular injections over 27 months. The primary endpoint is progression free survival. Some key secondary endpoints are; overall survival, lung cancer specific survival and disease free survival at 2, 3, 4 and 5 years.
There are 300,000 newly diagnosed NSLC patients each year. Their disease is graded by the AGCC staging system. Staging describes anatomical distribution of the tumor, its size and the extent to which it is contained. Then each stage has various grades that describe the pathology or aggressiveness of the tumor.
There are 90,000 NSCL patients who fall in the Ib to IIIa stages of the disease. Of these 60,000 can be resected and 35% of these have adequate MAGE-A3 expression which qualifies them for inclusion in the trial. They will receive 13 doses over a 27-month period. This indicates that the addressable market of patients who fit the entry criteria for this trial is about 21,000 patients in the US and Europe. The topline results will be announced in late 2013 or early 2014
The DERMA Phase III Trial in Melanoma
The DERMA trial was started in December 2008. The enrollment criteria for the trial is patients with stage IIIb/ IIIc cutaneous melanoma who have been rendered disease free by surgery prior to randomization and whose tumors show meaningful expression of MAGE-A3.
There were approximately 1,349 patients enrolled and they were randomized 2:1 between the vaccine and placebo. The primary endpoint is progression free survival. Key secondary endpoints are overall survival, disease specific survival and distant metastasis free survival. Each group will receive 13 intramuscular injections over 27 months. ClinTrials.gov states that the final collection of data for the primary outcome measure will be October 2016. However, Glaxo has indicated that there will be topline data in 2H, 2013.
There are 11,000 patients in the US and Europe who are diagnosed each year with stage IIIb/ IIIc stage melanoma. Of these 10,000 are resected and 66% of these patients have adequate MAGE-A3 expression. This indicates that the addressable market of patients who fit the entry criteria for this trial is about 6,600 patients in the US and Europe. The topline results will be announced in 3Q or 4Q of 2013.