The Pill That Could · A Different Cancer
A Different Cancer, part 4
Genentech's bet was that humanized monoclonal antibodies, against the right targets, would be effective as cancer drugs. The bet required them to find the right targets.
Slamon's paper provided one.
* * *
The collaboration between Slamon at UCLA and Genentech in South San Francisco began around 1987 or 1988. The Genentech side included an antibody chemist named Axel Ullrich, who had been working on the HER family of receptors before Slamon's paper appeared and whose group had cloned the human HER2 gene. (Ullrich had, in fact, been the first to publish the cloning of human HER2; the gene appeared in the literature with two names — HER2 from his group, neu from Weinberg's group — which is why the chapter has so far been writing it as HER2 with a parenthetical or hyphenated reference to neu, in the way the early literature did.) Other key Genentech people on the project included an immunologist named Mike Shepard, who would lead the antibody development effort, and eventually a clinical scientist named José Baselga, who would later run much of the clinical program.
The plan was straightforward. Make a humanized monoclonal antibody that bound the HER2 receptor on the surface of breast cancer cells. Test the antibody in the dish for whether it slowed the growth of HER2-amplified breast cancer cells. If it did, test it in animals, then in patients. If it worked in patients, develop it as a drug.
Each step was harder than the previous chapter would suggest. Generating the original mouse antibody against HER2 was a months-long campaign. Identifying which of the many candidate antibodies actually killed HER2-amplified cells in culture took further months. Humanizing the lead antibody — figuring out which mouse residues to graft onto a human framework, then making sure the humanized version still bound its target tightly — was a piece of careful molecular engineering that took, in this case, several years. The eventual humanized antibody was called 4D5 in early publications and would become, in 1990, the molecule that Genentech designated as the lead clinical candidate.
It is worth saying briefly how a monoclonal antibody, given as a drug, can kill cancer cells. There are several mechanisms, and the relative contributions of each in any given case can be hard to disentangle. The antibody can simply block the receptor's function — binding to the outside portion in a way that prevents dimerization or activation. It can cause the receptor to be pulled into the cell and degraded, reducing the number of receptors on the surface. It can, by virtue of being an antibody, recruit immune cells to the cell it has bound: natural killer cells, in particular, recognize antibody-coated cells and kill them, in a process called antibody-dependent cellular cytotoxicity. It can activate complement, a separate arm of the immune system that punches holes in the cell membrane of antibody-coated targets. The antibody Genentech eventually developed appears to do most of these things to some extent. The relative importance of each is still, decades later, a subject of research.
Imatinib, by contrast, has one mechanism. It fits into the ATP pocket of BCR-ABL, prevents the kinase from doing its job, and the cancer cells die because the cancer is dependent on the kinase being active. The mechanism is clean. The mechanism of an antibody is messier, in the sense that more than one thing is happening at the same time. Both kinds of drug can be effective. They are different kinds of drug, doing different kinds of work, with different practical implications.
* * *
The Genentech HER2 program proceeded through the late 1980s and early 1990s with a quiet steadiness that, in retrospect, looks similar in shape to the Ciba-Geigy program from chapters four and five. There were internal champions and internal skeptics. The cancer was uncommon enough as a single subtype — a quarter of all breast cancers, but only the HER2-amplified ones — that the commercial case was not obvious. The technology — humanized monoclonal antibodies as cancer drugs — had no approved precedent. The compound itself had real challenges; making large enough quantities of a humanized antibody to support clinical trials was, in 1990, a non-trivial bioengineering problem.
What sustained the program through this period, more than anything else, was Slamon. He was, throughout, the external advocate — the academic clinician who kept showing up at meetings and conferences with new data on HER2 amplification, who kept calling Genentech to push the project forward, who kept the molecule from being shelved during periods when the company's internal priorities shifted. He was, in his way, the Druker of the Herceptin story.
There was also, on the patient side, a movement that did not exist in CML. Breast cancer, by the early 1990s, had a deep and well-organized advocacy infrastructure. Organizations like the Susan G. Komen Foundation, founded in 1982 by Nancy Brinker after her sister's death, had built constituencies of patients and survivors who were active politically and visible publicly. Some of those advocates, hearing about the HER2 antibody, took up the cause of accelerating its development. Marti Nelson, a California physician with HER2-positive breast cancer who could not get into the clinical trial, became one of the visible faces of the campaign for expanded access. She died in 1994, before the drug was approved. Her name became, for many people involved in the program, a reminder that the cost of moving slowly was paid by particular people in particular timelines that were not always generous.
All 13 chapters of The Pill That Could · Library · All courses