Maytansine is a natural product of fungus. In the 1970s, it captured the interest of cancer researchers because of its impressive ability to kill a variety of cancer cells. The problem was that it proved adept at killing a variety of other types of cells as well—as long as they were trying to divide, the kind of cells useful in keeping a person alive.

 

But maytansine has captured the interest of Genentech, which about seven years ago embarked on a program that could transform the company’s collection of monoclonal antibodies into a drug delivery platform that may be able to turn broad-acting and toxic agents such as maytansine into highly targeted cancer drugs.

 

In 2000, Genentech entered a licensing deal with Cambridge, Massachusetts-based Immunogen for DM1, a tumor-activated form of maytansine. Genentech is now in a mid-stage clinical trial for HER2-positive breast cancer with T-DM1, Immunogen’s drug linked to Genentech’s monoclonal antibody trastuzumab, which is better known as Herceptin. This antibody-drug conjugate binds to the HER2 receptors that are overabundant on cancer cells in women with this form of breast cancer and delivers the maytansine directly into the tumor cells. Genentech is now working with a broader range of Immunogen candidates as part of a subsequent agreement. 

 

Antibody-drug conjugates, or ADCs as Genentech calls them, consists of an antibody, a drug, and a linker that binds one to the other. The linker is crucial—not only because it allows the antibody to deliver the cytotoxic agent in a targeted manner, but it must also be able to release the drug at the right place at the right time. 

 

“The third part is the understudied step, the underappreciated step,” said Mark Sliwkowsky, a staff scientist and Genentech’s director of translational oncology. “The way you link these molecules to these antibodies is just hugely important and has not been studied maybe in as much detail as it should have been.”

 

So far an early-stage clinical study of T-DM1 has provided encouragement. It showed that the drug could be delivered in at a high enough dose to have an effect in women who had already exhausted all other therapies and were no longer responsive to Herceptin. The dose-limiting toxicity came at the highest dose used in the early-stage trial, and it caused the patients’ platelet count to drop, although these levels were quickly restored to normal when the drug was discontinued. Sliwkowsky said that the conjugate does not inhibit Herceptin from acting as it normally does by itself. That raises the possibility of not only getting the targeting of the monoclonal antibody, but the therapeutic effect of it as well.

Using monoclonal antibodies as a drug delivery mechanism is not a new idea. Since they were discovered in the 1970s there has been talk of linking them to drugs to target their delivery. But, so far, it's been an idea that was easier to conceive than execute. Genentech is by no means alone in the efforts. In fact, Wyeth’s Mylotarg, a treatment for certain types of acute myeloid leukemia and approved in 2000, is an example of an antibody-drug conjugate already in use.

 

But Genentech may be in an unusually strong position to develop and exploit this technology. Its long work with finding cell surface targets in cancer cells and developing monoclonal antibodies for them has given it a collection of monoclonal candidates with which to work. Some of these antibodies hit compelling targets, but offered no therapeutic effect. Now, they could be combined with potential chemotherapeutic agents.

 

At the same time that Genentech has been working to develop ADCs, the company has had a separate effort to build up its small-molecule capabilities in-house. This has happened in an effort unrelated to its push to develop ADCs, but instead from a growing number of targets that Genentech had become interested in within tumor cells. These targets could not be attacked with protein therapeutics, but instead needed small molecules that could penetrate the cell. That growing in-house chemistry capability has proven useful in helping to construct the linkers needed to marry the monoclonal antibodies to drugs.

 

“Some of us have training in chemistry and backgrounds in chemistry, but there’s nothing like having a critical mass of card-carrying chemists,” said Sliwkowski. “We can go down the hallway and say, ‘can you whip up something that has these properties?’ and they generally say, ‘yeah.’ And after you bring them some liquor, they stay late one night, and the next morning they bring you a vile.”

 

For Genentech, ADCs represent a powerful new drug delivery platform that can lead to the development of a range of new cancer therapeutics, but it is also another indication of the blurring line between biotechnology companies and small-molecule drug makers. Such labels today say more about market multiples, growth prospects, and corporate culture than technology a company uses.