By David N. Leff

It's common knowledge that cancer of the prostate (CaP) is the commonest lethal malignancy known to man (but not woman). In the U.S. this year, an estimated 179,300 men will be diagnosed with CaP, and 39,200 will die of the disease.

Yet, observed oncologist David Agus, at the Memorial Sloan-Kettering Cancer Center in New York, "The actual cause of CaP death remains a great question. The real answer is: We have no idea.

"In general," he continued, "patients die from androgen-independent prostate disease. That is, when they fail the therapy of hormone withdrawal, and the disease comes back. And it isn't much treated at that point. Chemotherapy in some patients has a benefit; in many, none. (See BioWorld Today, Sept.30, 1999, p. 1.)

"As their disease progresses," Agus went on, "the patients die. But what they die of really hasn't been catalogued well at all. It's different from some other cancers, where you know the tumor invades a certain vascular system, and there could be bleeding or infection, a very low blood count, etc. In prostate cancer it probably has something to do with many of the cytokines that the tumor itself is secreting, or is caused to secrete. But what the cause of death is we don't know exactly."

"In fact," he suggested, "a great deal about CaP isn't known exactly, or isn't at all predictable. The medical oncology community has created a little bit of a problem in prostate cancer, just categorizing the disease as such. In reality there are prostate cancers - plural - with multiple subtypes."

Agus cited hypothetical examples: "There's that one subtype that's going to happen in an elderly gentleman - and it won't affect him at all. A biopsy of his prostate gland shows that, yes, he has prostate cancer - and he's going to go 10 or 15 years with no problems. Then there's another patient who'll have a virtually identical prostate cancer that will be very aggressive and life threatening.

"Right now we don't have the tools to predict accurately who's going to be aggressive and who won't. So I think the future is going to be subcategorizing prostate cancer into CaP-1 to CaP-15, each representing a different aspect of the disease."

A Medical Dogma Challenged

Agus is lead author of a research paper in the twice-monthly Journal of the National Cancer Institute, dated Nov. 3, 1999. It's headed: "Prostate cancer cell cycle regulators: Response to androgen withdrawal and development of androgen independence." This seemingly noncommital report in fact challenges a long-held dogma as to the cause and course of CaP.

"The current thought in prostate cancer," Agus told BioWorld Today, "is that when you withdraw androgens - that is, remove testosterone from treatment - 99 percent of tumor cells will die of apoptosis; only a few will survive. Then those residual cells will grow up and bring on androgen-independent disease, which is responsible for the death of the patient.

"What our study shows, on the contrary," he went on, "is that when you withdraw androgen, very few cells die. The majority of them go instead into growth arrest. They stop multiplying, and lie dormant in a phase of the cell cycle. The cells themselves get smaller, and that's responsible for the tumor shrinking, but very few of the cells, if any, die. Prostate-specific antigen - PSA - goes down to zero, because PSA is driven by testosterone, so it gives you a new paradigm: If you want to cure this disease, you're going to have to attack the growth arrest itself.

"Right now," he pointed out, "we wait for CaP to come back as an 'androgen-independent' disease, and then we try palliative chemotherapy. What this study addresses is that when you have this growth arrest itself, now is the time to go for cure. Try to attack these growth-arrested cells with drugs that seem to be sensitive to the arrest part of the cell cycle."

In pursuit of this strategy, Agus and his co-authors employ what they call an entirely new animal model of human prostate cancer. "It's different from most people's nude mice, implanted with human CaP tumor cells," he observed. "Ours didn't start out growing in cell culture dishes. Rather, they were taken from a patient and transplanted into mice. And because of that, they aren't cells that were selected to grow on plastic, or in a tissue culture dish. They grow into tumors, the way human tumors do, because they are human tumors."

In their mouse models, and currently in a Phase II human trial as well, the Sloan-Kettering team is attacking these CaP tumors with a humanized monoclonal antibody called Herceptin. "It attacks HER2, a receptor on the surface of cells that allow them to grow," Agus explained. "This growth factor receptor is overexpressed in about one-third of breast cancers," he added, "and probably 20 percent of prostate cancers."

Model Mice Prefigured Clinical Trial

"So what we were able to show in our mice," he recounted, "is that when we gave Herceptin to androgen-dependent tumors, it worked. The tumors stop growing, got smaller, and PSA went up, because Herceptin seems to signal down an androgen receptor pathway. And that aberrant signaling is responsible for this cell arrest, as well as increasing PSA."

Agus observed, "These animal studies helped us design the ongoing clinical trial of Herceptin, which started in its present form about a month ago." The open-label study will enroll 40 androgen-dependent and -independent CaP patients, of whom 17 or 18 have already entered treatment.

"The trial has really two angles to it," Agus said." One is giving Herceptin as a stand-alone drug. And the other is when Herceptin itself doesn't work, to add Taxol to it, because of the synergy between them. Herceptin is administered the same as in breast cancer, a weekly intravenous dose. And its side effects are rather small."