A BiTE (Bispecific T Cell Engager) antibody is a bi-specific antibody (bsAb) which directs T-cells to attack cancer cells, by simultaneously binding the two cells. Upon binding, a physical link is created between the two cells, which in turn triggers the T cell to attack the target cell. Every BiTE antibody has two binding arms, the first binds the CD3 receptor present on T-cells and the second binds a specific element on a cancer cell. The T-cell binding arm provides the activity while the cancer binding arm provides the specificity. By changing the cancer binding arm, the BiTE antibody can be adapted not only from one type of cancer to another, but also from one target to another in the same type of cancer. Therefore, BiTE represents a universal and modular platform for producing bsAbs for an unlimited number of targets.
As previously stated, bi-specific antibodies are aimed at recruiting immune cells against cancer. Therefore, one of the first decisions to be made concerns the type of immune cells to be recruited. The first attempts to develop bi-specific antibodies, mainly included recruiting T-cells, which are considered the most potent cells of the immune system. T cells play a critical role in the body’s efforts to eliminate malfunctioning cells such as cancer or virally infected cells, making them even more obvious candidates. T cells attack target cells by bombarding them with toxic proteins that eventually force the cell to commit programmed-cell death (apoptosis), and are considered very efficient in fighting cancer for several reasons. First, when properly activated, T cells act as serial killers, killing multiple cells one after the other. Second, upon activation, T cells start to multiply at the site of activation (which is hopefully next to the tumor), increasing the number of effector cells at an exponential rate. Third, T cells are constantly scanning tissues in order to find cancer cells. Many cancer cells manage to avoid being recognized by a T cell, even if it passes in proximity, however, a bsAb might provoke the T cell to recognize and attack even the most evasive of cancer cells.
Most of the efforts to trigger T cells against cancer cells utilized the CD3 receptor because of its involvement in the activation of T cells’ attacking mechanisms, but this concept remained invalidated for over twenty years, since no one was able to decipher exactly how T cells are triggered through this receptor. Today, an increasing body of scientific evidence, including one clinical trial, implies that Micromet has found the recipe for the activation of T cells via the CD3 receptor. So how did Micromet succeed where others have repeatedly failed?
It turned out that in order to achieve maximal stimulation of T-cells, they must be as close as possible to the cancer cell. While larger antibodies manage to bind a T cells and a cancer cell simultaneously, the distance between the two is not short enough for the T cell to “realize” it is standing in front of a cancer cell. Micromet’s BiTE antibodies are unique because they are much smaller and flexible and consequently bring T cells close enough to cancer cells to generate a powerful response.
The extremely high potency of BiTE antibodies was already observed in early pre-clinical trials, where BiTE antibodies demonstrated up to a 1000 fold increase in potency compared to prior bispecific antibodies that target CD3. BsAbs are commonly assessed by their E:T ratio. This number represents the ratio between the number of effector cells ( in our case, T cells) to the number of target cells (cancer cells) needed in order to show a substantial anti-cancer activity. Obviously, a more potent bsAb will have a lower E:T ratio, since it implies each immune cell kills more target cells at a given time. Most bsAbs have an E:T ratio of 40-100:1, whereas some which use CD3 have a ratio of 2:1. MT103, the first BiTE antibody to enter the clinic, is the only bispecific antibody I am familiar with which achieved a lower than 1 E:T ratio: A staggering 1:10 ratio. In other words, a potent anti-cancer activity is observed even if cancer cells outnumber T cells by a factor of 10!
MT103 targets CD19, a receptor which is expressed in many blood cancers including Non-Hodgkin’s Lymphoma [NHL] as well as several leukemias. Because CD19 is expressed in the majority of blood cancers, it is often described as “The next CD20″. Throughout the years there were several attempts to target CD19 with naked antibodies, but none of them bore fruit. Today, CD19 is receiving a great deal of attention in the pharma industry as several companies, including Sanofi-Aventis (NYSE:SNY) and Genentech (Private:DNA) are developing Antibody-drug conjugates based on anti-CD19 antibodies. Although none of the other companies who target CD19 reported clinical results, the CD19 market is going to be highly competitive.
In the first clinical trials for MT103, the antibody was given in a series of infusions similarly to the way standard antibodies are administered. These trials failed to show any significant clinical response and were associated with severe side effects that led to the early termination of the trials. Then, in 2004 Micromet started another phase I trial in NHL patients, in which MT103 was to be administered via continuous infusion. This is very unusual, as the typical dosing of antibodies for cancer is once a week. Conventional antibodies tend to stay in patients’ bloodstream for more than two weeks, and unlike MT103, are very safe at high doses. However, the company did not have much of a choice, and it hoped that by lengthening the infusion time, the side effects could be mitigated and the overall amount of administered MT103 could be increased. It is important to mention that patients do not need to be hospitalized throughout the course of treatment, but simply carry a small pump the size of a cell phone that constantly infuses them with the drug.
This strategy proved successful as side effects decreased substantially compared to the short infusion trials, and the company’s CEO boasts the fact that patients actually gained weight during the treatment, a great unofficial indicator for safety. More importantly, signs of clinical activity could be seen very early in the dose escalation process.
The primary objective in phase I trials is verifying the safety of the drug and not necessarily its clinical activity. As a result, these trials typically start from very low doses in order to make sure the drug is safe, so nobody really expects the smallest doses in phase I trials to have a therapeutic effect. Amazingly, the first patient that was enrolled to the trial achieved Stable disease [SD], followed by only 4 weeks of treatment. The study has enrolled over 30 patients in 6 different cohorts. Out of the nine Patients who received the next two doses (cohorts 2 & 3 ), eight had SD as well. The trend continued, as in cohorts 4-6 multiple partial and complete responses were observed.
It is still too early to reach conclusions as to the efficacy of MT103 in this trial since dose escalation is still ongoing, however, with 7 out of 20 patients in the three highest doses achieving clinical responses, it definitely looks like that MT103 has a great potential for the treatment of several subtypes of NHL. Moreover, all three patients who received the highest dose, had an objective response ( One had a CR and the other two had PRs). It won’t be reasonable to expect that this dose can actually achieve 100% response rate in larger populations, but this is indeed a great sign of a dose dependent activity.
While the response rate of this trial is very encouraging, the actual doses which led to the responses are nothing short of sensational. MT103 was effective at doses at which most antibody-based platforms are not even evaluated. The highest dose in the trial was 0.06 mg/m2 per day, which translates to 0.42 mg/m2 per week. Because MT103 is a small antibody, the actual number of antibody molecules per mg is three times higher, so in order to level the playfield, the dose of MT103 should be regarded as 1.26 mg/m2 per week. Rituxan, which is approved for the treatment of NHL, is administered at 375 mg/m2 per week, an almost 300 fold difference.
This comparison is brought just to give an idea for how groundbreaking the MT103 data really is. The two antibodies have never been compared and there are still a lot of data missing in order to reach any definitive conclusions. It also does not imply that MT103 is a direct competitor of Rituxan, since they target two different antigens and seem to have a different activity spectrum for different subtypes of NHL. Moreover, Rituxan was found to augment the activity of antibody platforms against CD19 in animal models, so combining the two antibodies is certainly an option. What we can safely say, though, is that MT103 showed a strong clinical effect in a small group of patients, 90% of whom had been previously treated with Rituxan. In addition, we can also say that Rituxan is very unlikely to have any effect when dosed anywhere near the MT103 levels.
NHL is a collection of several types of lymphomas, which may vary in their response to a given treatment. The phase I trial patient population was rather heterogeneous but several observations can be made about MT103’s activity among several subtypes of NHL.
The most evident effect was shown in patients with Mantle cell lymphoma [MCL]. This aggressive subtype of NHL represents about six percent of all NHL cases in the United States. The MT103 phase I trial included 15 MCL patients, but if we look again at the three highest doses, there were two complete responses and one partial response out of eight patients (response rate of 42%). Rituxan, for instance, has 27%-37% response rate in MCL as mono-therapy, but when added to chemotherapy regimens, response rate increases to 60%-90%. Despite these favorable response rates, most patients relapse after initial treatments, making MCL one of the most challenging lymphomas.
Another subgroup of patients who seemed to benefit from MT103 treatment is patients with Follicular lymphoma [FL], which is the second most frequent lymphoma worldwide. Of the eight FL patients in the three highest doses, one achieved complete response while the other two achieved partial response (response rate of 42%). FL has been proven to be very sensitive to Rituxan, with around 50% response rate in pre-treated patients. Patients whose disease relapse after the treatment with Rituxan are still responsive to the antibody, but with a lower response rate.
Micromet launched a Phase II trial in Acute Lymphoblastic Leukemia [ALL] patients in October 2007. ALL is an extremely aggressive disease characterized by high mortality rates. ALL patients are typically responsive to chemotherapy, but in many cases, a small number of cancer cells manage to survive in the bone marrow, leading to a very quick relapse after less than 6 months. In this phase II trial, MT103 is given to patients who achieve remission with standard treatments, but still have remnants of cancer in their bone marrow. The high activity of MT103 in targeting cancer cells inside the bone marrow, might prove useful in maintaining ALL patients in remission for longer periods.
The company stated it regards this trial as a proof-of-concept trial for MT103 in aggressive blood cancers. If proven effective, MT103 can find itself in a phase II trial for DLBCL, which is the most common aggressive lymphoma and therefore represents one of the biggest commercial opportunities for MT103. This trial is also used to validate the concept of using MT103 for decreasing chances of disease relapse after conventional therapy.
MedImmune (Now part of AstraZeneca) partnered with Micromet for the development of MT103 and is expected to launch two trials in the US. The first will enroll NHL patients, similarly to the original phase I trial reported at ASH. The second trial will focus on CLL, the most common leukemia in western countries. CLL cells are characterized by relatively low expression of CD20, Rituxan’s target, which can serve as an explanation why CLL is less sensitive to the antibody. In addition, A retrospect analysis among CLL patients showed that CD19 expression may be associated with higher risk of disease progression and death.
Despite MT103’s promising activity in blood cancers, we must not forget that most patients are treated with a combination of chemo/radio–therapy and monoclonal antibodies. There is no way to predict whether MT103 can be co-administered with standard therapies, despite the good safety profile. This combination might even prove to be somewhat counterproductive, as the potential proliferation of T cells can be offset by chemotherapy and radiation treatments.
Going forward, BiTE’s real potential lies in the treatment of solid tumors, which represent more than 90% of cases of cancer worldwide. Unfortunately, naked antibodies against solid tumors suffer from very poor success rates in the clinic, and even the approved ones have modest efficacy, when added to chemotherapy. Companies who develop antibodies for cancer prefer to focus on blood cancers, which are far more responsive and easy to treat. This will ultimately turn the market of antibodies for blood cancers into a very crowded and competitive one. The market of solid tumors will be characterized by higher entry barriers and lower success rates on the one hand, but also by a lack of competition and great demand even for slightly effective antibodies.
The main problem with conventional antibodies such as Herceptin and Erbitux
is that only a tiny fraction of the total injected antibody molecules actually reaches the tumors, and when they do, they have a very subtle effect, as each antibody affects one cancer cell, not necessarily killing it. In addition, conventional antibodies affect primarily the more external cells of the tumor, since they cannot penetrate the dense tumor mass.
These are some of the issues that should be confronted by next-generation antibody platforms, such as the BiTE platform. Micromet’s approach looks very appealing because it actually deals with several issues at once. First, BiTE antibodies have a better chance of killing a cancer cell upon binding, assuming they succeed in recruiting T cells. Second, BiTE antibodies are smaller so they might be able to penetrate solid tumors more easily. Third, the use of T cells as effector cells may lead to a very strong amplification, so it might take a small amount of antibodies to trigger a powerful immune response.
One of the most impressive cases in the MT103 trial was an MCL patient whose liver was infiltrated with cancer cells. This patient had a CR after four weeks of treatment, as MT103 managed to wipe out all detectable tumors in the liver, according to a biopsy. This is certainly not a case of metastatic cancer but it implies that MT103 can penetrate and generate a potent immune response also in less accessible sites, in addition to its activity in the bloodstream and bone marrow.
Nevertheless, we are dealing here with drug development, where for every reason why a certain drug should work, there are ten reasons why it shouldn’t. There is still no clinical data about BiTE antibodies for the treatment of solid tumors and historical success rates for solid cancers average around a mid single digit. Therefore, Micromet involves tremendous risks and uncertainties as well as great potential.
The company has more than five additional BiTE antibodies in pre-clinical studies, all of which target solid tumors. This strategy is a breath of fresh air in an industry where most companies prefer to focus on blood cancers. It is a great example for how the next generation of antibody platforms should be developed: Proof of concept in blood cancers and then a strong shift to solid tumors, where market opportunities are huge. It would be unrealistic to expect most of Micromet’s clinical programs, including MT103, to succeed, but Still, if Micromet gets it right with only one of its BiTE molecules for solid cancers, the financial reward will be enormous.
Micromet intends to advance its first BiTE antibody for the treatment of solid tumors to the clinic in the coming weeks. The antibody, MT110, targets EpCAM, which is one of the most known and well characterized tumor-associated antigens. A recent scientific paper published by Micromet, presents very impressive results in animal models regarding the efficacy and safety of MT110. It remains to be seen whether these results can be, at least, partially duplicated in clinical trials.
Disclosure: Author is long MITI