The hottest theme at this year’s 2010 ASCO (American Society of Clinical Oncology) annual meeting was without a doubt cancer immunotherapy, an old paradigm that deals with redirecting the patient’s immune system against tumors. After decades of failures, this concept is finally proving itself useful. There are basically two approaches: 1) cancer vaccines that aim at eliciting an immune response against specific targets on cancer cells and 2) immunomodulatory drugs that aim at stimulating the immune system in a more general manner (not target specific). The undisputed leader in cancer vaccines is Dendreon (DNDN), which currently has the only FDA approved cancer vaccine. Following ASCO, it looks like the leader in immunomodulation drugs is BMS (BMY).
BMS presented data for two immunomodulatory agents at ASCO: Ipilimumab and MDX-1106. Both are monoclonal antibodies originally developed by Medarex, which was bought by BMS last year. The two agents are targeted against proteins that function as brakes on the immune system. By targeting these proteins, it is possible to abrogate their effect and unleash a potent immune response against tumors. Ipilimumab is much more advanced and could be in the market next year whereas MDX-1106 is an early stage candidate with impressive data. It is important to note that as with every disruptive approach, there is still some risk associated with these programs, especially with respect to side effects. Nevertheless, based on the available data, the two drugs put BMS in the sweet spot of cancer immunotherapy.
Ipilimumab (aka ipi) is an antibody targeting CTLA-4, a protein with an established role in suppressing the immune system. After a series of studies that were not robust enough, BMS, who was developing the drug in collaboration with Medarex, finally managed to come up with convincing evidence that ipi works. According to data from a phase III trial, this agent can prolong survival in patients with metastatic melanoma, perhaps the toughest oncology indication. This makes ipi the first ever treatment to show superiority in a large comparative trial for melanoma.
The trial, which was originally designed to evaluate ipilimumab and an experimental cancer vaccine (gp100) included three arms. Patients were randomized to receive ipilimumab+ placebo, gp100+placebo or ipilimumab+gp100. Therefore, the trial did not include a pure placebo arm. Patients on the two ipilimumab arms experienced a median survival of 10 months versus 6.4 months in the cancer vaccine arm. Extending overall survival by ~3.5 months in late stage patients is an impressive achievement by any means, especially when it comes to melanoma.
The lack of placebo arm does not appear to be a big issue, as the two ipilimumab arms were very similar in terms of activity and side effects, implying that the gp100 vaccine had no effect by itself. Some claim the absolute survival values for the control group are relatively low when compared with previous trials in melanoma. The most recent phase III trial that could serve as a historic control is the SYMMETRY trial with Synta’s (SNTA) elesclomol. According to an update at ASCO, the control arm in the trial that got chemotherapy had a median survival of 11.5 months, however, these patients were first line patients. The ipilimumab trial included pretreated patients, all of whom had received systemic therapy. Assuming a median progression free survival of 2 months and another month before randomization, one would expect the placebo arm in the ipi trial to have median survival of ~8.5 months, instead of the reported 6.5 months. In contrast, in a recent phase III study (E3695) in first line patients, the median survival was ~9 months, which is more “in agreement” with the ipi data. Therefore, the placebo arm seems to do worse than most historic controls, but this kind of variability between large trials is not unusual. The most important factor is the difference between arms within the same trial, which in this case is meaningful and statistically significant.
The only real issue with the ipi trial was side effects. Due to its mode of action, the drug leads to severe immune related side effects in up to 3% of patients, including a mortality rate of 1%. Nevertheless, this mortality rate is in the ball park of other treatments for melanoma that often include aggressive chemo regimens, and looks acceptable given the poor prognosis and lack of other options.
Therefore, despite the issues which have been brought up regarding this study, in my opinion the data still looks solid and justifies quick regulatory approval.
Another drug with a high likelihood of approval in melanoma is Plexxikon’s (in collaboration with Roche) PLX-4032, which has a completely different mechanism of action. What the two drugs do share in common is that both were rationally designed in light of the underlying science of the disease. Although there is still no data from a large randomized study for PLX-4032, early stage results were so overwhelming that its approval seems imminent. Although ipi and PLX-4032 basically compete for the same patients, physicians will probably use both sequentially, when possible. In addition, only ~60% of melanoma patients are suitable for PLX-4032, leaving the rest of the patient population to ipi. Down the road, a combination of the two agents seems reasonable, given the unrelated mechanisms of action.
In summary, ipi and PLX-4032 represent two landmark achievements. After decades of failures, physicians would finally be able to offer life extending off-the-shelf treatments to melanoma patients. Even more exciting is the potential use of those drugs in the adjuvant setting, where they might actually save lives of early stage patients. Ipilimimumab will probably be the first one to reach the market. If so, BMS will make history twice. Not only will it have the first approved immunomodulatory antibody, it will also have the first ever approved drug for melanoma based on a randomized trial.
By acquiring Medarex, BMS gained control of what seems to be another promising immunomodulatory antibody, MDX-1106. MDX-1106 binds PD-1 , a protein presented on immune cells and inhibit their function. MDX-1106 is being co-developed with Japan-based Ono Pharmaceuticals, which holds the ex-US rights for the product.
PD-1 is similar to CTLA-4 in its inhibitory effect on the immune system, and can thus be viewed as another potential checkpoint to be blocked. However, there are several theoretical advantages in targeting PD-1 over CTLA-4. Targeting PD-1 should result in fewer side effects based on the observation that mice without CTLA-4 die within several weeks whereas mice that do not express PD-1 present relatively mild symptoms. Another property that distinguishes PD-1 from CTLA-4 is the presence of its cognate ligands on cancer cells. This could be used to select patients who are more likely to respond to MDX-1106, although this is yet to be validated.
At ASCO, data for 91 patients who received MDX-1106 was presented. Of the 91 patients, 11 (12%) achieved objective responses (reduction of at least 30% in tumor burden) and additional 10 (11%) patients achieved tumor shrinkage but were not considered responders. The most impressive part was duration of response, as all of the 21 patients (confirmed and unconfirmed responses) were still on study with response duration of 5-18 months. This response duration is quite unusual given the advanced stage patients that are usually enrolled into early stage clinical trials. All but one of the responses were observed in melanoma and renal cancer patients, so the response rate(both confirmed and unconfirmed) in these indications was 32.6% and 31%, respectively. This kind of benefit is unusually high for a phase I trial in solid tumors, especially for targeted therapies.
The traditional assessment of response rate might even under-represent the overall benefit of immunomodulatory drugs, which are sometimes associated with a subtle or delayed effect. For example, in its phase III trial, ipi had a response rate of under 10% but still managed to prolong overall survival. In addition, patients can eventually achieve an objective response with this type of drugs even after they initially experience disease progression. This phenomenon was elegantly demonstrated at ASCO with Pfizer’s(PFE) tremelimumab, another immunomodulatory antibody. Lastly, if physicians could identify patients who are more likely to derive benefit from MDX-1106 based on what their tumors express, it could boost chances of success while shortening the path to market.
The activity seen with MDX-1106 corroborates findings from a previous trial originally presented at ASCO 2009. This trial included a less intensive dosing regimen in which the majority of patients received a single injection of MDX-1106 in order to evaluate the safety profile of the antibody. Of the 39 evaluable patients, one patient with colorectal cancer achieved a complete response, with two additional patients with melanoma and renal cancer achieving a partial response. The low response rate in the trial could be attributed to the low number of administrations, but in any case, a clear long lasting effect was observed in this study as well. Interestingly, there seemed to be a correlation between PD-L1 (PD-1’s ligand) on the surface of tumor cells with clinical benefit which bodes well for developing predictive biomarkers for this agent.
Another company who indirectly benefited from MDX-1106 is Israel-based Curetech, which is also developing an anti-PD1 antibody (termed CT-011) in collaboration with Teva (TEVA). In contrast to MDX-1106, CT-011 was not rationally developed as a PD-1 antibody and its actual target has been identified only several years ago. CT-011 entered clinical trials already in 2004, two years before MDX-1106 reached the clinic. The only published results for CT-011 is a phase I trial in 17 patients with blood cancers. There was one complete response in a previously untreated lymphoma patient and sustained minor tumor shrinkage in a patient with AML.
CT-011 and MDX-1106 are the only PD-1 antibodies in the clinic and are currently being evaluated in phase II studies. It seems the two companies have a totally different strategy for reaching the market. Curetech is focusing primarily on blood cancers and combination regimens. Nevertheless, their largest trial is a randomized study in colorectal cancer in combination with chemotherapy. BMS is putting more emphasis on solid tumors, especially melanoma and renal cancer.
Based on the published clinical data, MDX-1106 looks better positioned with strong efficacy in refractory solid tumors despite the two-year head start CT-011 had (so far no data has been published for CT-011 in solid tumors). BMS’ focus on solid tumors addresses larger markets with higher unmet medical need and less competition in comparison to Curetech’s lean towards liquid tumors. BMS also has a lot of experience with immunomodulatory antibodies thanks to the ipi clinical program. Geographically, CT-011 is lagging behind MDX-1106 in Japan, where BMS’partner, Ono, is already conducting a phase I study. In addition, the largest (and probably most important) study for CT-011 (colorectal cancer) is conducted primarily in developing countries whereas MDX-1106 biggest trial is conducted in the US. Obviously, the picture will get clearer as more clinical data become available.
Strategically, BMS made two bets that paid off. The first bet was to acquire Medarex ahead of the clinical data for ipilimumab, which probably saved BMS a lot of money. The second bet (which was also Medarex’s bet) was the focus on immunomodulatory antibodies, ignoring the widespread skepticism within the industry. The Medarex acquisition brought BMS additional immunomodulatory antibodies which are still too early to evaluate, further strengthening its position in the field. It is important to note that despite the promising results, both programs are far from being risk free, especially due to the safety overhang which will remain even after ipi is approved, let alone for MDX-1106.
In retrospect, the acquisition of Medarex was a smart move solely based on ipi and MDX-1106, not to mention the full pipeline of antibodies and a potential royalty stream from multiple programs based on Medarex’s technology. The $2.4B paid for Medarex was money well spent, especially in comparison to the outrageous acquisition of Medimmune by AstraZeneca (AZN) for $15B three years ago.