By David N. Leff

Now that new cases of lung cancer have leveled off as more and more people quit smoking, the commonest cancer among males, smokers and non-smokers alike, is carcinoma of the prostate.

And prostate cancer (PC) is becoming more and more common with every passing year. Of all the risk factors ascribed to this men-only malignancy, the riskiest of all is being black rather than white. The latest figures show 236 white males per million dying of prostate cancer, compared with 520 per million blacks.

In trying to nail down this dodgiest of all malignancies, oncologists have indicted every conceivable risk connection, from age and lifestyle to steroid hormone levels, to excessive height and birth weight to genomic CAG microsatellite repeat sequences, to PSA (prostate specific antigen). (See BioWorld Today, June 12, 1997, p. 1.)

Today's Science, dated Jan. 23, 1997, unveils a brand new prostate cancer risk factor, which is being hailed as the strongest predictor of contracting the disease yet found. The pertinent article's title: "Plasma insulin-like growth factor-I [IGF] and prostate cancer risk: A prospective study."

The paper's senior author is clinical oncologist Michael Pollak, of McGill University, in Montreal. "PSA is a tumor marker," Pollak told BioWorld Today. "A man with an elevated PSA serum level may well have prostate cancer. IGF," he continued, "is not like that. It's a normal hormone, not made by the cancer. IGF does a good job regulating cell turnover rates throughout the body. If you had no IGF, you'd be sick or dead."

What then, Pollak asked himself, was IGF doing in the prostate gland?

For epidemiological information on this question, he turned to the Harvard School of Public Health, in Boston. "It's well known," he observed, "that they have serum banks and follow-up data." For his purposes, the school tapped into its Physicians' Health Study, an ongoing database following the medical histories since 1982 of 20,017 healthy male U.S. doctors, age 40 to 82. Of that number, 14,916 donated blood plasma for Pollak's prostate project.

"We did a very simple study," he recalled, "where we measured the IGF levels in 152 men of the plasma cohort who got prostate cancer subsequent to 1992, and 152 matched control subjects who did not. When we looked at the IGF in this combined cohort of 304 men, we found that those in the highest 25 percent had levels 4.3 times higher than those in the lowest quartile.

"That finding," Pollak summed up, "indicated that IGF is actually the strongest predictor of getting PC. PSA is a marker of having PC. IGF, from our study, is a marker of PC risk."

But that's just the beginning of his IGF story.

PSA levels are to males over 50 years of age what mammograms and Pap smears are to women, namely, an early-stage diagnostic of cancer. "PSA is not just an antigen put there to help us detect PC," Pollak observed. "It is a proteolytic enzyme, with the probable physiological role of keeping seminal fluid from coagulating."

He added: "From an IGF point of view, PSA has an interesting property: It digests one of IGF's binding proteins, IGF-BP, so it serves to limit IGF's binding activity. Anything that digests that binding protein," Pollak went on, "will increase the IGF bioactivity, namely, increasing cell division and proliferation. So now we have an insight into why PC tumor cells like to make PSA.

"If they make PSA," he pointed out, "any IGF in their microenvironment that is biologically inert, because it's complexed to its binding protein, will be digested by PSA. This liberates free IGF to stimulate the tumor cells to grow."

So the payoff for men with very high levels of IGF -- such as those in the top 25 percent in the study -- Pollak suggested, "might be that they deserve closer clinical follow-up, more frequent PSA readings, or maybe we can even devise some risk-reducing strategies.

"There are drugs on the market," he noted, "that lower IGF levels, to treat such rare overgrowth conditions as acromegaly. These are analogues of the growth hormone inhibitor, somatostatin, but with longer half-lives than the native hormone, which is not practical to use. They may become candidate drugs now for PC prevention."

Pollak makes the point that "this is not the first report to link IGF to the prostate. But the majority of prior research was lab work with individual cells, or tissues or animals. Ours is the first paper," he remarked, "that says all that preclinical stuff actually is relevant to the human disease."

A month ago, the McGill oncologist began testing some transgenic mice genetically altered to develop PC. "We're treating them with somatostatin analogues," he said, "to see if we could lower the level of PC in this model."

IGF Control Could Work In Other Tumors, Too

Given IGF's body-wide activity, Pollak and his Harvard colleagues also reported recently -- to the 1997 meeting, in Edmonton, Alberta, of the American Epidemiological Society -- "on an apparent association of IGF with premenopausal breast cancer."

However, Pollak regards his findings as preliminary, and subject to validation by others. He is setting up international collaborations "to confirm and extend these observations," he said, "and I would hope that all kinds of other groups interested in prostate cancer would independently look at it with their own data sets."

Pollak mentioned in passing that apparently some individuals on the cusp of middle age are acquiring and using growth hormone, which stimulates IGF production, with the aim of mitigating the signs and symptoms of aging. "We think that's dangerous," he said. "It's not an FDA-approved indication. It's okay for growth-hormone-deficient kids. Subnormal to normal is perfectly appropriate. But going from normal IGF to abnormally high, they're asking for trouble, as far as our records are concerned."

Specifically, he concluded, "that trouble could follow increased IGF stimulation of cell division and proliferation, leading to cancer." *