Expert Financial Analysis and Reporting

Background Information on Checkpoint Modulation

Let me provide background information that shapes my views on immuno-oncology before I get into my recommendation. The immune system has evolved over more than one billion years as a defense mechanism to protect humans and other life forms from pathogenic organisms that cause disease- notably bacteria, viruses, fungi and parasites. It also defends against mutations in cells that can cause cancers. Paradoxically, the immune system can also go awry and cause serious auto-immune diseases such as asthma, rheumatoid arthritis and many other maladies. Over the last three decades, we have seen the understanding of the working of the immune system increase exponentially and this has led to great medical and commercial advances.

The most notable success arising from this understanding has been the development of monoclonal antibodies for the treatment of cancer and auto-immune diseases. The first blockbuster product of this class was Rituxan (introduced in 1997) which is used to treat hematological cancers and certain auto-immune diseases; current sales are around $7 billion. More notably, worldwide sales of products based on the 60 approved products based on monoclonal antibodies are about $89 billion (roughly 16% of worldwide pharma sales of about $550 billion) and there are more than 575 monoclonal based products in clinical development worldwide. As future investors look back decades later however, they may conclude that monoclonal antibodies were just the tip of the iceberg.

In this report, I am focusing on immuno-oncology (a space in which monoclonal antibodies have made a huge impact) and the extremely promising product development work in that area. There are numerous other disease areas in which understanding of the immune system has and will lead to major medical breakthroughs. Still for now let’s just look at cancer drug development. The most exciting near term opportunity is the development of monoclonal antibodies that interact with receptors on T-cells to enhance or block T-cell activity and in doing so can bolster the immune system’s attack on a cancer. These products are called checkpoint modulators and this report is focused on this new area.

The first checkpoint modulator was Bristol-Myers Squibb’s (BMY) Yervoy which was introduced in 2011 and has since been joined by Bristol-Myers Squibb’s Opdivo, Merck’s Keytruda and Roche’s (RHHBY) Tecentriq. I estimate that sales of these four products could approximate $6 billion in 2016 which contrasts with no sales in 2010. More importantly, Street estimates are that sales of checkpoint modulators as a class could soar to $33 billion by 2022.

There are other emerging immuno-oncology areas beyond checkpoint modulation that are of great interest which I am not touching on in this report. Products based on engineered autologous T-cells (first generation is CAR-T cells) have caught the imagination of investors and the first two products of this group could be approved in 2017; these are Novartis’ CTL-019 and Kite’s KTE-C19. I think Wall Street is much too optimistic on these products in the near term, but over the coming decades, I see engineered autologous T-cells as a major growth area.

Next are therapeutic vaccines. This is a highly controversial area as there have been numerous product development failures over the last two decade and just two product approvals: Dendreon’s Provenge for prostate cancer and Amgen’s Imlygic for melanoma. Until recently, the discouraging development record made this area an investment pariah. Those of you who follow my research have seen this manifested in a  savage short attack on Northwest Biotherapeutics (NWBO) which is developing its cancer vaccine DCVax-L for glioblastoma, but this attitude is turning. There is now growing conviction and enthusiasm that checkpoint modulators can greatly enhance the efficacy of cancer vaccines. For example, NWBO’s DCVax-L is starting a trial in which it is combined with MRK’s Keytruda to treat glioblastoma. There is much more investment research from SmithOnStocks to come on cancer vaccines, but not in this report.

My Investment Recommendations in this Space Are Bristol-Myers Squibb (BMY) and Agenus (AGEN)

In this report, I have presented two names that I think are excellent opportunities for investing in the checkpoint modulator space. The first is the large pharma company Bristol-Myers Squibb (BMY, $57.14) which is the world leader in the checkpoint modulator space, both in sales and breadth of pipeline. A setback in the clinical development of Opdivo in first line lung cancer is, in my opinion, just a bump in the road and has led to an excellent buying opportunity.

The second investment idea is Agenus (AGEN, $4.28). This small company has brilliantly positioned itself as a force to be reckoned with owing to a very broad development pipeline of checkpoint modulators. This pipeline is just emerging from pre-clinical to clinical stage as four products will be in phase 1 development in 2017 and more than 10 others are in pre-clinical development. None of the Agenus products have yet created meaningful data in humans, which obviously increases the risk and has created a cautious attitude by investors.

I think that any of the several large biopharm companies who have been left standing at the checkpoint modulator gate would jump at the opportunity to partner with Agenus on certain products. There is already a broad ongoing collaboration with InCyte and a product development effort with Merck in place, but there are many other opportunities to come. I basically believe if Agenus can establish proof of concept with some of its products, that it will be acquired in its entirety by a larger company within the next two to three years.

I want to emphasize that Bristol-Myers and Agenus are not the only opportunities. They are just the ones in my coverage universe that now stand out the most. You can start by investing in these companies, but keep an open mind to other possibilities. The rising checkpoint modulation tide can lift a lot of boats.

I believe that BMY has the potential to regain most of the market capitalization it has lost since early August when it reported that the Checkpoint-026 trial of Opdivo in first line lung cancer had failed. The stock crashed from $75 on August 4 to $63 on August 5 and trailed down to a low of $49 on October 20; the current price is $57. By the end of 2017, I see the stock returning to or surpassing $80 and project 15% to 20% per annum growth in EPS and stock price in the period beyond 2017. Not insignificantly, the current dividend yield is 2.7%.

In my report on Agenus titled "Its Market Leading Position in Immuno-Oncology Makes it A Compelling Investment Story" I suggested that Agenus potentially could be acquired within the next three years at a price of as much as $25. Yes, I know the stock is now selling at about $4.

This report is quite lengthy and requires a good deal of time to read and then think through. I recognize that many investors just want to know the ley investment points and not get overly caught up in the details; hence the following section.

Key Investment Points

• Checkpoint modulators burst on the scene in 2011 with the introduction of Yervoy (ipilimumab). Since then sales have surged to an estimated $6 billion in 2016. There is the potential that sales could grow to $33 billion in 2022.
• Bristol-Myers Squibb was the first mover in this space and is likely to be the leading company in 2022 and beyond. Merck, Roche and Astra-Zeneca are the primary big pharma challengers to BMY. I believe that each of these companies and others have huge commercial opportunities as well.
• The initial use of checkpoint modulators has been as monotherapy or in combination with other cancer therapies. Very importantly, there is compelling evidence emerging that combinations of checkpoint modulators will be possibly twice as effective as monotherapies with comparable side effects. This has already been seen in some melanoma and non-small cell lung cancer trials.
• At this point, the only combination that has been approved is BMY’s Opdivo (anti-PD-1) and Yervoy (anti-CTLA-4) for the treatment of melanoma.
• I believe that the PD-1/ CTLA-4 combinations such as Yervoy and Opdivo will become the backbone of many, many cancer treatment regimens and will supersede monotherapies. The potential opportunity is staggering.
• Bristol-Myers Squibb with Opdivo and Yervoy and Astra-Zeneca with durvalumab (an anti-PD-L1 antibody) and tremelimumab (a CTLA-4 inhibitor) are the only companies that currently have checkpoint modulator combinations in advanced trials. They are positioned to enter the market in late 2017 or 2018, roughly two years ahead of potential competitors.
• BMY recently suffered a major blow when Opdivo failed as monotherapy in first line lung while Merck’s Keytruda was successful. BMY’s stock was pummeled in the aftermath.
• I think that MRK’s triumph and BMY’s bitter disappointment will be short lived. BMY has clinical trials underway in first line lung cancer with a combination of Opdivo and Yervoy which optimistically could lead to approval in 2H, 2017 or perhaps more likely in 2018. Astra-Zeneca’s similar combination of durvalumab and tremelimumab could be approved in 2018 or potentially slightly earlier. I think these combinations will quickly replace Keytruda in first line lung cancer cancer.
• Bristol-Myers has two possible strategies for gaining approval of the combination of Opdivo and Yervoy for lung cancer. The first would be to file for approval on the basis of the phase 1 Checkmate-012 trial which was very encouraging and the phase 2 Checkmate-568 trial that should complete in early 2017. These are open label trials, but there is precedent that indicates that the FDA might approve the combination based on these results assuming a positive outcome of Checkmate-568. If so, this could lead to approval in late 2017 or early2018. This would have an enormously positive impact on BMY’s stock and a highly negative impact on MRK.
• The second strategy is based on the possible (probable) success for the Opdivo/ Yervoy combination in the Checkmate-027 trial. This is a huge, 2000 patient phase 3 trial that should definitively establish the role of Opdivo combined with Yervoy in first line lung cancer. Topline results should be available in 1Q, 2018. Success in this trial would be enormously positive for BMY and devastating for MRK.
• There is also the possibility that the Checkmate-027 trial could be halted at an interim look in 2017 and result in approval in late 2017 or early 2018 if the Data Monitoring Board (DMB) detects a statistically significant survival advantage.
• Astra-Zeneca has thrust itself into the competition. It is conducting a trial of its PDL-1 inhibitor durvalumab and its CTLA-4 inhibitor tremelimumab that could report results in mid-2017. AZN could gain approval in early 2018 in first line lung cancer.
• Astra-Zeneca’s trial in first line lung cancer comes with some uncertainties. Neither of the drugs in its combination has yet been approved by the FDA as a single agent although it looks like durvalumab will gain approval for advanced bladder cancer in 2H 2017. Also, there is scant data on AZN's anti-CTLA-4 drug tremelimumab as a single agent. In contrast, Most oncologists have extensive experience with BMY’s Opdivo and Yervoy which meaningfully favors the BMY combination in the early months or years of commercial rollout.
• With the approval of the Bristol-Myers and Astra-Zeneca combinations, I see MRK’s Keytruda as being essentially replaced in first line lung cancer to the benefit of BMY and AZN. This is why I am buying BMY at this time. The first line lung cancer market could be turned upside down in next six to 18 months just as it was in the last four months when MRK’s Keytruda was successful in a first line lung cancer trial and BMY’s Opdivo failed. This will be discussed extensively in this report.
• Other drug companies are scrambling to find an anti- CTLA-4 to go with their anti-PD-1 or anti-PD-L1 drugs. Agenus has the next two anti-CTLA-4 drugs in early testing.
• I believe that other big drug companies will seek to license these Agenus drugs on very favorable terms to Agenus or perhaps acquire Agenus outright.

Overview and Purpose of This Report

Bristol-Myers has declined from $75 per share on the day before results of Checkmate 026 were released to a current price of $57; this represents a loss in market value of $23 billion. Checkmate-026 was a study of Opdivo in first line, metastatic lung cancer. In that trial Opdivo as a monotherapy failed to show superiority to platinum based chemotherapy which is the current standard of care. Clearly, the market felt that the results in this trial represented an extremely disappointing (nearly catastrophic) change in the potential for Opdivo. This note examines the reasons why the market has reacted so negatively and presents the point of view that this may ultimately prove to have been a massive over reaction and a great buying opportunity for investors in Bristol-Myers stock.

I think that the market reaction is partially a consequence of the difficulty in figuring out what is going on in the T-cell checkpoint modulation corner of the immuno-oncology space. What is so exciting for investors is that these products can potentially be used as a component of treatment regimens for almost every type of cancer; the potential applications are vast beyond anything I have seen. Some respected Wall Street analysts are estimating that sales of these drugs can increase from about $6 billion in 2016 to over $33 billion in 2022. The leading company in this space is Bristol-Myers Squibb (judged by current sales and breadth of pipeline) while Merck, Roche and Astra-Zeneca should also emerge as leaders. Many other companies are scrambling to gain a small niche, which in a market of this size can be very meaningful commercial opportunity.

The number of clinical trials ongoing and the data coming out of these trials are far too complex for any one person (group) to understand. Indeed, the conclusions of analysts, key opinion leaders and investors shift like a weathervane in a windstorm as a plethora of new data spews forth. It takes an IBM Watson type of artificial intelligence to collate and interpret this information. However, it is obvious to me that the leading position of Bristol-Myers Squibb promises exceptional growth for the Company even as Merck, Roche, Astra-Zeneca and others gain share in the market and achieve blockbuster sales.

With great humility and admittedly even greater over-simplification, in this note let me try to paint the picture as I see it. I want to emphasize that I am going to give you my conclusions (impressions) rather than trying to support each of my conclusions with detailed analysis of clinical trial data. As I just discussed, the amount of clinical data is overwhelming and expanding at a lightning speed; to try to discuss just a small part of this data would result in an encyclopedic, incomprehensible report.

Bristol-Myers Squibb’s Guidance for 2017

In 2016, BMY, I estimate that Opdivo sales will reach $3.3 billion (remember, it was just launched in 2015 and represent 18% of BMY’s sales. Opdivo is heavily dependent on sales in non-small cell lung cancer. About 30% to 40% of Opdivo sales are in lung cancer in the U.S., another 20% to 30% is in other indications in the U.S and about 25% to 35% of sales are outside of the US. Most of Opdivo sales are in the second line, non-small cell lung cancer setting where it has a 60% to 70% share of the overall market (chemotherapy regimens take most of the remaining share). Investors are concerned that as Keytruda is used in patients in the first line setting that these patients are no longer potential Opdivo patients in the second line. BMY agrees, but also points out that only about 25% of patients in the front line setting are eligible to receive Keytruda according to the label.

Merck has now gained parity with Opdivo on its label in the second line setting and can compete on an equal footing. Also, Roche’s Tecentriq has a comparable label. Hence, the competition in second line lung cancer will be more intense in 2017 than in 2016. With all of this competition, the question that investors have is whether Opdivo can grow. BMY has issued guidance that it will grow on a global basis. My estimate is that Opdivo sales will increase from $3.3 billion in 2016 to $4.2 billion in 2017. However, some investors look for little growth or even a decline in 2017.

During the 3Q conference call, BMY increased 2016 non-GAAP EPS guidance to between $2.80 and $2.90 and guided to 2017 non-GAAP EPS to be between $2.85 and $3.05. Hence the stock is selling at roughly 18.5 times using BMY’s 2017 EPS guidance. As previously mentioned, it also has a 2.7% dividend yield.

Checkpoint Modulation Briefly Explained (If you feel that you don’t need or want this background, skip to the next section)

As a starting point, let’s go back to the period before the first checkpoint modulator (inhibitor) Yervoy (ipilimumab) was introduced in 2011. In the years preceding the launch of Yervoy there was pervasive skepticism about immune-oncology among key opinion leaders and almost no awareness of checkpoint modulation by investors. Now just five years later, checkpoint modulation is probably the hottest area of drug development on the planet. So what exactly is checkpoint modulation? Let me approach this from my layman’s viewpoint.

Checkpoint modulation is a critical component of the adaptive immune response to cancer that regulates T-cell activity, upregulating and downregulating it like a thermostat. Of course, T-cells play a leading role in the immune system’s attempt to eradicate a cancer. Let me provide a brief overview of how checkpoint modulation works.

Let’s begin with discussing antigens, which are foreign substances (they usually don’t occur naturally in the body) that induce an immune response which leads to the activation of killer T-cells and the production of antibodies by B-cells. The immune system uses both of these mechanisms and others to attack and attempt to destroy cells that display those antigens. Checkpoint modulation controls the intensity of the activity of killer T-cells. I am not going to go into a detailed explanation, but when the immune system detects a foreign antigen, killer T-cells are created that specifically recognize that antigen. These antigen-targeted killer T-cells are leading participants in the immune response against cancer cells that display that particular antigen. They seek out and attach to the antigen on the cancer cell surface and release cytokines that assassinate the cancer cell.

Killer T-cells are very effective killing instruments but the number of these cells circulating in the body is not always optimum. Sometimes their numbers and effect need to be bolstered by encouraging them to differentiate and at other times dampened because too many can lead to extensive damage of normal cells in addition to cancer cells. The immune system needs a way to turn on and off the differentiation of T-cells. This function is done through actions orchestrated by regulatory T-cells (whose function is like their name) as well as antigen presenting cells. These cells produce cytokines (proteins) that interact with receptors on the surface of killer T-cells.

There are many, many types of T-cell receptors and the cytokines that interact with them. Some of these receptor/ cytokine pathways encourage T-cell differentiation and others dampen it. Also, these pathways (there are tens or hundreds) interact with each other. It is far beyond my ability to understand and characterize all of these. Indeed, even leading scientists in the checkpoint modulator space are just beginning to put the first pieces of this exceedingly complex puzzle into place in a discovery process that will take decades.

The Most Advanced Checkpoint Modulation Drugs

The very early advances in the checkpoint modulation field have been in developing drugs that interfere with the PD-1 and the CTLA-4 pathways. Both PD-1 and CTLA-4 are receptors that occur on the cell surface of killer T-cells. They are activated by ligands (proteins) produced by other cells of the immune system. The most widely studied ligand for the PD-1 receptor is PDL-1 (there is also a PDL-2). There are also two ligands for CTLA-4: these are CD80 (B7-1) and CD86 (B7-2). The ligands attach to these receptors and this results in a message being sent to the DNA in the nucleus of the cell that stops differentiation, thereby reducing the intensity of the T-cell attack.

Very importantly, cancer cells can sometimes highjack these inhibitory processes by producing ligands that activate the PD-1 or CTLA-4 receptors. By doing so, they turnoff or dampen the differentiation (and number) of killer T-cells and their ability to attack that cancer.

The first four checkpoint modulator drugs developed by the pharmaceutical industry and approved for commercial use are monoclonal antibodies that bind to either PD-1, CTLA-4 or PDL-1 (a ligand as opposed to a receptor). They bind to these receptors or but do not activate them; in doing so, they effectively block the ligands produced by cancer cells from binding to these receptors. Hence, the method used by cancer cells to block T-cell activity is rendered ineffective. Bristol-Myers” Yervoy (ipilimumab) was approved in 2011; it targets CTLA-4. Bristol-Myers’ Opdivo (nivolimumab) and Merck’s Keytruda (pembrolizumab) were approved in 2015; they target PD-1. More recently, Roche’s Tecentriq (atezolizumab) was approved. It is an antibody against the PDL-1 ligand. Instead of blocking the PD-1 receptor and preventing PDL-1 from binding, it knocks down the number of PDL-1 ligands that can bind to PD-1.

Clinical Results with Yervoy, Opdivo, Keytruda and Tecentriq

The initial indication of Yervoy as a monotherapy was in late stage, metastatic melanoma patients. Yervoy (targeting CTLA-4) produced 15% to 20% objective response rates in metastatic cancer patients who no longer responded to chemotherapy and other drug therapies. What so impressed investigators was that the responses were very long lasting and close to cures for some cancer patients who had exhausted all therapeutic options. Results with Yervoy were impressive but later clinical trials showed that Opdivo (targeting PD-1) and Keytruda (targeting PD-1) used as monotherapy resulted in better clinical outcomes than Yervoy so that both Bristol-Myers and Merck focused their clinical development on these drugs initially in melanoma and then progressing rapidly to almost all other cancer types.

Both companies saw the enormous potential for Opdivo and Keytruda because their mechanism of action applies to almost all types of solid and hematological tumors. Most currently used drugs (chemotherapy, targeted therapy and monoclonal antibodies) are effective in only a few cancers. In contrast, the number of potential applications for Opdivo and Keytruda, are staggering. The opportunity in part was to replace currently used drugs in certain cancers, but perhaps more importantly was the opportunity to combine Opdivo and Keytruda with other drugs in almost all types of cancers.

It is an axiom that the most effective way of combatting cancer is to combine drugs with different mechanisms of action. Hence, cocktails of drug therapy evolve over time. Moreover, one cocktail might be applicable as first line treatment while others are used after first line, second line or additional lines have failed. The number of such cocktails is virtually limitless and very importantly the exciting aspect of Opdivo and Keytruda is that they can potentially be used in the great majority of cocktails developed for solid and hematological cancers.

The mechanism of action of Keytruda and Opdivo are pretty much the same and almost all of the clinical data suggests that they are the same drug for all intents and purposes. Although the mechanism of action of Tecentiq is somewhat different the end result is likely the same. There is little reason to believe that one of these three drugs is meaningfully better or worse than another. It is also likely to be the case that numerous other drugs in development that block the PD-1 receptor are going to produce similar results. If this is the case, the thing that probably jumps to your mind is that if these drugs essentially produce the same clinical effect won’t they be interchangeable and won’t that ultimately mean that usage will depend primarily on price. I think that the answer to this is paradoxically no.

Clinical oncologists are extremely data driven as they should be. While they might believe, as I do, that these drugs are all pretty much equivalent, most will only use a drug that is backed up with convincing clinical trial data. This confers a huge strategic advantage to Bristol-Myers and Merck as first movers in the space and they are taking every advantage of their position. They are pursuing as many trials in as many cancers as possible that combine Opdivo and Keytruda with both existing drugs and drugs in development. As importantly, other companies in the cancer space recognize that Opdivo and Keytruda are likely to be the backbones of most cancer therapy regimens and are anxious to do trials that combine their drugs with Opdivo and Keytruda. If this hypothesis is correct, Opdivo and Keytruda would be parts cancer cocktails used in most solid and hematological tumors.

Bristol has to date gained the leading position in the PD-1 antibody race. It has approval in five different tumor types (melanoma, lung, renal, head and neck, and Hodgkin’s lymphoma) as compared to three for Keytruda (melanoma, lung and head and neck cancer). Opdivo initially had a broader label (all levels of PDL-1 expressers and non-expressers) in second line lung that allowed usage without PD-L1 testing. It is also the only company with an approved combination of checkpoint inhibitors with Yervoy and Opdivo in melanoma. Generally, it has stronger clinical data that has focused on survival benefits. This has all translated into a much better launch for Opdivo as it achieved 3Q, 2016 sales of $920 million versus $356 million for Keytruda.

While BMY was the leader in developing this technology, MRK employed a quick to market strategy that allowed them to gain the first commercial approval for Keytruda in metastatic melanoma in September 2014. However, BMY has been able since then to cast a much wider web across cancer indications. Both are also engaged in a staggering amount of clinical activity. Merck has said that the Keytruda clinical development program includes more than 30 tumor types in more than 360 clinical trials of which roughly 200 trials combine Keytruda with other cancer treatments. Bristol-Myers hasn’t provided a similar statistic, but I would guess that its reach is even broader.

The Battle between Opdivo and Keytruda in Non-Small Cell Lung Cancer

Lung Cancer is a Huge Commercial Opportunity

The most important cancer target from both a medical and a commercial standpoint is non-small cell lung cancer. Its annual US incidence is 224,000 and annual death rate is 158,000. This is put in perspective by looking at the incidence and death rate of the three next deadliest cancers; colon cancer has an incidence of 135,000 and deaths of 49,000: pancreatic cancer is 53,000 and 42,000 and breast is 245,000 and 40,000. BMY scored the first major victory in this space by gaining approval in second line lung cancer and so far has dominated this clinical indication. Second line treatment is used after a first round of platinum based chemotherapy has failed and the patient has relapsed.

PDL-1 is the Key Biomarker in Lung Cancer

Cancer drugs are more effective if biomarkers can be identified that identify those patients who are most likely to benefit. In the case of non-small cell lung cancer, the currently used biomarker is the percentage of cancer cells that produce the PDL-1 (called PDL-1 expression) ligand. The idea behind this marker is that the more cancer cells producing PDL-1, the more likely that a PD-1 receptor inhibitor like Opdivo and Keytruda will produce improved clinical benefit versus other treatment regimens.

Using PDL-1 Expression to Select Patients; Merck and Bristol-Myers Took Different Approaches

The biology behind the whole area of checkpoint modulation is in a very early stage of discovery. Intuitively it would seem to be the case that these drugs would be relatively more effective against cancer cells with higher PDL-1 expression and early trials suggest this is the case indicating that PD-L1 overexpression is an important facilitator for tumor growth and metastasis. Bristol-Myers and Merck took quite different approaches in planning trials in lung cancer using PDL-1 expression as the key biomarker to define patient characteristics.

Merck took the conservative stance of testing Keytruda in cancers in which PDL-1 expression was ≥50%. It is intuitive that the relative therapeutic effect would be somewhat proportional to PDL-1 expression. By choosing high PDL-1 expression, the clinical trials would seem to be more likely to show effectiveness. Hence, Merck enrolled only patients with PDL-1 expression ≥50 %.

In contrast, Bristol-Myers took a much more risky approach setting a cutoff for patient enrollment at a much lower PDL-1 ≥ 1%. In second line lung cancer, both Opdivo and Keytruda successfully met their endpoints and were approved, but Keytruda had a major disadvantage. Opdivo was ultimately approved for patients with any level of PDL-1 expression and also in non PDL-1 expressers as well; it could be used in any patient. In contrast, Keytruda can only be in patients for whom a test has shown PDL-1 ≥50 % and a cumbersome test for this biomarker is required to show that PDL-1 expression is ≥50 %. Recently, Merck gained an equivalent label.

It is probably the case that Keytruda and Opdivo are equally effective in all patients groups. Still, as I mentioned earlier oncologists are very data driven and most only used Keytruda if tests showed PDL-1 expression ≥50%. On the other hand, Opdivo could be used in virtually all cancers. In practice, performing the PDL-1 test is an annoyance for physicians and most preferred to use Opdivo because they wouldn’t have to test. As a consequence Opdivo came to dominate the second line setting.

Going on to First Line Lung Cancer; Merck’s Focus on Patients with ≥50% PDL-1 Expression Produced a Big Victory

Bristol-Myers and Merck then went on to do trials in first line, non-small cell lung cancer which compared Opdivo and Keytruda to chemotherapy which is standard of care. Both companies followed the same approach as in second line. Merck’s enrollment criteria was again based on ≥50% PDL-1 expression and BMY again used ≥1% PDL-1 expression. Bristol and most investors felt that clinical trials in first line would play out like second line, but shockingly it didn’t. Keytruda was shown to have a statistically significant benefit over chemotherapy in both overall survival (OS) and progression free survival (PFS). However, Opdivo was not shown to be better than chemotherapy in this setting.

Even though most key opinion leaders intuitively feel that Keytruda and Opdivo are pretty much equivalent and the different outcomes in first line lung were due to trial design, this is a major win for Merck (at least in the short term). As I mentioned before oncologists are very much driven by data and labeling. Based on the data, Keytruda should be used in all first line lung cancer patients with PDL-1≥50% which is generally believed to be about 25% of first line lung cancer patients. Current standard of care chemotherapy regimens would continue to be used in the remaining 75% of first line patients. There would be no role for Opdivo as monotherapy in the first line setting.

I am not going to try to analyze in detail why the BMY trial failed. Sometimes it just happens that the sample size is not large enough to offset some unforeseen characteristics. It is my guess that results for Keytruda and Opdivo in this setting will ultimate turn out to be about the same. However, for the immediate future Keytruda will be the only drug used in the front line setting. Keytruda has an unchallenged clear path forward as the standard of care in those 25% of first line lung cancer patients who have PDL-1 expression ≥50%.

Many investors now total sales of Keytruda matching or exceeding quarterly sales on Opdivo by 2H, 2017. It is this abrupt change of fortune that has decimated Bristol Myers.

Keytruda Results in First Line Lung Cancer

Keytruda (pembrolizumab) reduced the risk of progression or death by 50% vs standard of care which is platinum based chemotherapy (HR=0.50, CI 0.37-0.68) in patients with advanced NSCLC expressing PD-L1 > 50%. mPFS was 10.3 months for pembrolizumab and 6.0 months for chemotherapy. For death alone, Keytruda reduced the risk by 40% (HR=0.60, [0.41-0.89]) vs chemotherapy. This was all the more impressive because half of the patients given chemotherapy were crossed over Keytruda so that in some patients Keytruda was being compared to Keytruda. Even with the crossover, the one year survival rate for Keytruda was 70% versus 54% for chemotherapy. The response rate was 45% for Keytruda versus 285 for chemotherapy.

The Keytruda trial excluded patients with brain metastasis, autoimmune disease, prior radiotherapy, EGFR/ALK mutations and PD-L1 expression below 50%.

Opdivo Results in First Line Lung Cancer

In contrast Opdivo (nivolumab) did not reduce the risk of progression or death vs platinum based chemotherapy (HR=1.05 [0.91-1.41]) in patients with advanced NSCLC expressing PD-L1 > 5%. Note that patients in the Keytruda trial were expressing PD-L1 > 50%. Opdivo did not reduce the risk of death (HR=1.02, CI 0.80-1.30). mPFS was worse for Opdivo at 4.2 months versus 5.9 months for chemotherapy. Even more discouraging was that there was statistically significant difference in survival and disease progression in patients expressing PD-L1>50%.There was no clear explanation for difference this trial from earlier phase ½ trials.

Bristol-Myers Strategy for Checkpoint Modulators Going Forward

Therapy Will Evolve Toward Combinations of Checkpoint Modulators

BMY believes that monotherapy was just the first step for the anti-PD-1 drugs. The next major advance will be to combine PD-1 monotherapy with other checkpoint modulators. The activity that the PD-1 monotherapy has shown in the first line lung cancer will probably limit it to the 20% or 25% of the high PDL-1 expressers. This is critical to the investment outlook for Bristol-Myers and Merck.

The role of PD-1 monotherapy in the second-line lung cancer setting was clearly answered with CheckMate-057 and CheckMate-017, both of which showed pretty significant improvements in overall survival versus chemotherapy. Opdivo and other anti PD-1 drugs have clearly been established as standard of care in the second-line setting.

The second question was the role of PD-1 monotherapy in the front-line setting. Based on results in CHECKMATE-026 and KEYNOTE-24 their view is that PD-1 monotherapy in the front-line setting is likely to be used in the greater than 50% expressers, which is roughly 25% of the population so that there are still 75% of patients with significant unmet medical need who are now treated with chemotherapy doublets.

Bristol-Myers has made an earlier and much bigger bet on combination checkpoint modulation treatment in first line lung cancer. It has completed a phase 1 trial called Checkmate-012 that produced an exciting signal that a combination of Opdivo and Yervoy is much more effective than Opdivo with comparable side effects. It is now wrapping up a larger phase 2 study that is a still larger study of Opdivo and Yervoy in this setting. And most importantly, it has a landmark phase 3 study Checkmate-027 that will give a comprehensive view of the safety and efficacy in the first line lung cancer setting of: (1) Opdivo and Yervoy combined, (2) Opdivo, Yervoy and chemotherapy combined, (3) Opdivo used as monotherapy and (4) chemotherapy. I will next discuss each of these trials.

Checkmate-012 

This was an open label trial that looked at the combination of Opdivo and Yervoy in about 80 patients. It was labeled a phase 1 trial, but its size was basically that of a phase 2. Patients enrolled in the trial covered a broad range of PDL-1 expression from >1% on up to >50%.

At > 1% the response rates were about 60% which compares to Opdivo monotherapy in other trials of roughly 30%; this was a doubling of the response rate. At > 50% PDL-1 expression the response rate was 92% for Opdivo plus Yervoy compared to about 50% for Opdivo monotherapy in other trials. This trial suggests that the combination would be effective at all levels of PDL-1 expression. It also indicates that the response rates with the combo are almost double the response rates with monotherapy across whole levels of PD-L1 expression. It also shows that the level of response in absolute terms grows with the level of PD-L1 expression.

Checkmate-568

This is a single-arm study now in progress that is looking at the combination of Yervoy and Opdivo in first-line lung. The primary endpoint here is response rate. It is intended to replicate the data seen in Checkmate -012. At ASCO in 2016, management indicated that the initial purpose of this trial was to provide information for the critical Checkmate-027 trial. Since then and probably due to the failure of Checkmate-026, management sees a more meaningful role for this trial.

The size of 568 has been doubled from 170 patients to 340. This is the same patient size as Keynote-024 trial of Merck in first line lung cancer. However, in Keynote-024 Keytruda was compared to chemotherapy while Checkmate-568 is an open label trial. The recruitment of the first 170 patients is almost completed and BMY expects to complete the second 170 by early 2017. The increase in enrollment will provide more data for the spectrum of PDL-1 expression, in particular in the non-expressers and in the highly expressing patients.

CheckMate -568 is complementary to CheckMate -012 and will read out in the early part of next year. This will allow them to confirm whether the results in Checkmate-012 are real. Importantly, there will be longer term follow-up from 012 at ASCO 2017 that will give a valuable insight into duration of response.

.Checkmate-027

This is a large phase 3 study or more appropriately it is actually two large phase 3 studies in one protocol. It is a 2000 patient study that will enroll about 1,000 PD-L1 positive patients and 1,000 patients PD-L1 negative patients. As previously noted this study will have several arms: (1) Opdivo and Yervoy combined, (2) Opdivo, Yervoy and chemotherapy combined, (3) Opdivo used as monotherapy and (4) chemotherapy.

The PDL-1 expresser part of the trial is close to being fully enrolled. The non-expresser part started a little bit later and is still recruiting. Here they are looking at: (1) Opdivo and Yervoy and (2) Opdivo and Yervoy combined with chemotherapy with both arms compared against chemotherapy.

BMY has expressed a lot of confidence for a successful outcome based on the results from Checkmate-012, the data for which was shown ASCO 2016. Checkmate -012 clearly demonstrated that response rates with the combo are almost double as the response rates with monotherapy across all levels of PD-L1 expression. It also shows that the level of response in absolute terms grows with the level of PD-L1 expression. That is all consistent with the trial design of Checkmate -227

They are contemplating enrolling more than 1,000 patients in the PDL-1 expresser arm. They are right at the end of the enrollment period and patients continue to come in pretty quickly. Enlarging the trial would provide more statistical power. They don’t believe that this would cause a significant delay in the study because of where they are in enrollment and the speed with which patients are enrolling.

According to ClinicalTrials.gov Checkmate-027 is scheduled to complete in January 2018 and BMY says that is what analysts should assume even if the trial is enlarged. However, many of the Opdivo trials that they stopped early at either planned interim analysis or because the Data Monitoring Committee saw a survival benefit that led them to recommend that the trial should be stopped early. Barring this, the trial is likely to be completed and topline data made available in 1Q, 2018.

BMY says they are confident that they have chosen a regimen that maximizes the probability of demonstrating efficacy while addressing concerns with toxicity. This is based on the side effect profiles and discontinuation of therapy that were seen in Checkpoint-012, side effect data that is occurring in the open label Checkpoint-568 trial and no concerns raised by the DMC in the Checkpoint-027 trial.

BMY’s Regulatory Strategy for Gaining Approval of the Opdivo/ Yervoy Combination in First Line Non-Small Cell Lung Cancer

Early Filing Based on Checkpoint-012 and Checkpoint-568

At ESMO in November 2016, BMY hinted at a possible early filing strategy based on the combination of 012 and 568. This of course would be dependent on data and the agreement of the regulatory authorities. BMY is not talking further about this for competitive reasons other than to say they do have a strategy for the potential approval of the combination in 2017. This is not included in any guidance. Several of Opdivo’s approvals have been based on single-arm Phase 2 studies such as the approval in Hodgkin’s.

Another avenue to be to obtain a listing in the NCN Compendium. This would require publication of the data from 012 and 568 in at least two major medical journals. Even though this is not an approval, this is a stamp of approval that would gain widespread reimbursement even if the FDA does not formally approve the combination of Opdivo and Yervoy.

Checkmate-027 is the Critical Trial

This trial which I previously described will be the trial that defines the role in first line non-small cell lung cancer of the combination of Opdivo and Yervoy. Management has designed this trial in part on the findings of Checkmate-012 and if these results are replicated, BMY will have dominant treatment platform. Astra Zeneca could also have a combination approved and could be the second leading company in the space. Merck would be rendered an also ran.

This trial is scheduled to complete in early 1Q, 2018 and we should see topline data in 1H, 2018. With success, I would think that it could be the basis of approval of the combination in late 2018. Astra-Zeneca’s combination is probably slightly ahead of this timeline. However, the much greater familiarity of the oncology universe with Opdivo and Yervoy would likely mean that they would dominate usage.

There is some chance that Checkmate-027 could be stopped at some interim point if the Data Monitoring Board detects a statistically significant survival advantage. This has occurred in some Opdivo studies. If so, this could lead to approval in 2H, 2017 or early 2018.

BMY’s Strategy in Immuno-Oncology

BMY has eight I-O assets in various stages of clinical development where. In the FRACTION study, they are looking at all of those assets in combination with Opdivo or in combination with other drugs across a range of tumors. They expect three additional assets to go into the clinic in 2017. The success with Yervoy and Opdivo in the last couple of years got a lot of people excited. However, BMY says that some of these assets need more study and may not be as promising as earlier thought. They are adamant that combination regimens are going to be the path to get the greatest benefit to the most patients.

Management is not that excited about the combination of Opdivo and Yervoy with chemotherapy. They have cited as a reason their experience with Opdivo combined with Alimta. They saw an increase in response rates that to about 45% in patients across lung cancer, but durability was short lived at about six months. They did this study over three years ago so that they have excellent long term follow-up and concluded that the response is not long lasting.

They also have a phase 3 study 370 that is looking at sequencing chemotherapy. Patients are started on a dose or two of chemotherapy. If they get a response, Opdivo is then layered on with the objective if increasing the durability of response.

Astra-Zeneca’s Checkpoint Modulator Combination

Astra also believes that combinations are the key to the future and they are studying durvalumab (an anti-PD-L1 antibody) in combination with tremelimumab (a CTLA-4 inhibitor). Durvalumab is also being studied in broad range of other cancers as monotherapy. MYSTIC is a phase 3 trial of the combination of durvalumab and tremelimumab in first line non-small cell lung cancer. Results are expected in 1H, 2018

 

Tagged as AGEN, Agenus, BMY, Bristol-Myers Squibb, checkpoint-027 clinical trial, Checkpoint-568 clinical trial, Keytruda, Opdivo, Yervoy + Categorized as Company Reports, LinkedIn