In many ways, the pain is worse than the impending death. When prostate cancer spreads, it settles in the bones, along with other organs such as the lungs. But it's the bones that ache and break and dominate one's life.

Prostate cancer grows slowly, but in more than 40 percent of cases -- some 37,000 men each year -- it escapes the gland and begins spreading long before the physician finds the tumor.

Studies at Stanford showed that there is a direct correlation between the size of the tumor and the likelihood that it will spread. "Until the cancer reaches 1 1/2 cubic centimeters {slightly larger than a pea}, it doesn't begin to invade outside the prostate," said Dr. Thomas Stamey, chairman of urology at the Stanford University Medical Center. When it reaches 3 cc, it moves to the seminal vesicles. At 4 cc, about the size of a marble, the tumor begins to spread to the first line of lymph nodes, the portals to the rest of the body.

"Everything is predictable in this cancer by its volume," Stamey said.

Killing prostate cancer once it spreads has proved difficult. Surgery is useless at this point. Chemotherapy, the application of toxic drugs, fails to help, though some are occasionally used. Radiation can provide some relief.

Hormone therapy, however, has been the mainstay for slowing, sometimes even stopping and reversing, the spreading disease. Prostate cells, including prostate cancer cells, even when they've spread elsewhere in the body, need the stimulation of testosterone, the male hormone, to grow. Blocking testesterone slows the growth. Deciding which hormone therapy works best, however, has been controversial.

There are two basic approaches: removing the testes, the prime sources of testosterone, or using drugs that block testosterone in the body, a so-called chemical castration.

Surgically removing the testes, an orchiectomy, is the simplest.

"Orchiectomy is still commonly done," said Dr. John Lynch, a urologist at Georgetown University School of Medicine. "A lot of men don't want to have an orchiectomy for psychological reasons. I don't think they understand all the issues involved. It is done under local anesthesia. It takes 20 to 30 minutes. It is not disfiguring," since the scrotum remains. It does, however, leave the man impotent and infertile.

Unlike the hormones that can cause such side effects as increased risk of heart attack and stroke, hair loss, breast development, hot flashes and other changes, orchiectomy has no side effects, Lynch said. "Of all the hormonal treatments, it is the best one."

Orchiectomy is irreversible; the use of hormones such as diethylstilbestrol is not. And that's why many men find it more attractive. There are a number of hormone preparations that block testosterone's action in the body. With the male hormone blocked, the prostate cancer cells stop growing, at least for a time.

A controversy arose earlier this year when syndicated columnist Ann Landers wrote about a "breakthrough" hormone treatment in Canada.

Dr. Fernand LaBrie in the department of molecular endocrinology at Center Hospital, the University of Laval in Quebec, concluded that the small amount of testosterone produced by the adrenal glands was enough to stimulate prostate cancer even if the testes were removed. His group developed an approach in which an anti-testosterone drug called flutamide, not yet approved for use in the United States, was combined with a drug that blocks a brain hormone needed to stimulate the production of male hormones.

For patients who previously had not received a hormone therapy, said Dr. Lionel Cusan, a LaBrie colleague, the results are impressive. "We get a 90 percent response or even greater in those who have never been treated before." Response, however, is defined as stopping the tumors from growing any larger.

"If they have had hormones before," Cusan said, "there is a resistance to the treatment." For those patients, the response to the Canadian treatment was 30 percent or less.

Part of the controversy about the Canadian study, Cusan said, is that they did not use a placebo group who received no treatment. All the patients were treated and then compared to so-called historical controls -- what would be expected based on what happened to patients in the past.

The National Cancer Institute is now completing a study of 617 prostate patients comparing the Canadian two-drug approach with more traditional anti-androgens alone. Although the final data is not available, Dr. Andrew Dorr, NCI's coordinator for the study, said that the response in Americans was less than half of that reported by the Canadians and that the treatment did not prolong survival but did delay the progression of the disease by about two months.

None of the hormonal therapies cure the disease, Lynch said. "You can either stabilize the disease and get some improvement, but it is a temporary improvement." There is, however, some evidence that hormone therapy prolongs survival, but it is limited, and the length of survival varies with the individual.