A Slippery Preventive Slope

Intended as they are for large populations, preventive health measures must meet particularly high safety standards. The best such measures pose no risk at all. A few years ago it was estimated that so modest a measure as a reduction of salt intake could prevent a quarter of the strokes and perhaps a fifth of the heart attacks in the United Kingdom—doing no harm in the process. The drug proposed for the prevention of prostate cancer runs afoul of this standard.

The problem traces back to the landmark Prostate Cancer Prevention Trial, designed to determine whether finasteride, originally used to treat benign prostatic hyperplasia, could cut the incidence of prostate cancer. In the event, it did just that—by 25 percent. Incidentally, however, the PCPT demonstrated that the rigorous search for prostate cancer uncovers all too much of it. The rate of detected disease ran so high in the trial's intensively screened population—originally classified as low-risk—that even the suppressing effect of finasteride was not quite enough to bring it down to the lifetime risk of the general population: 17 percent. The study subjects underwent regular PSA (prostate-specific antigen) testing, a procedure that can detect even asymptomatic cancers and, therefore, results in overdiagnosis and overtreatment.

When PSA screening got going twenty years ago, no one could have foreseen or intended the ensuing epidemic of prostate cancer. With PSA we waded into a quagmire. No wonder some look to finasteride as the way out. The architect of the PCPT has been quoted as saying that, while no one knows whether finasteride reduces mortality, it "does reduce the frequency of prostate cancer treatment and the adverse effects the treatments carry." In the words of another proponent of finasteride, "It is a bit of a strange prevention argument. . . . We can't say we've avoided cancers that otherwise would have killed people. But we can say we've avoided cancers that people consider significant enough to treat."¹

Though it may soften the effects of the PSA revolution, finasteride is not risk-free. A third striking finding in the PCPT was that the rate of high-grade or aggressive cancer ran significantly higher in the finasteride group. Beginning with the original report on the trial in June 2003, many of the finasteride papers explicitly concede that the drug may induce high-grade cancer. On the other hand, the observed increase may be a mirage resulting from detection bias—that is, the greater likelihood of detecting cancer, especially high-grade cancer, in smaller glands on biopsy. Recent reports seem to support this suspicion, and as a result, the tide of opinion may be turning in favor of finasteride. Now some are recommending in print that all men undergoing screening—tens of millions—be "offered" the drug. Following the PCPT came the appropriately guarded recommendation that men at high risk of prostate cancer, especially by reason of family history, might consider finasteride. Later, the drug was recommended for the more nebulous category of men "at risk" of the disease, and more recently for patients "fearful of a diagnosis" of prostate cancer. Some now propose finasteride for any man "concerned" about prostate cancer, at which point the drug is effectively being endorsed for use across the board.

If not for mass screening, however, finasteride's proponents might not be able to cite an urgent need for a drug whose safety remains in question. According to Christopher Logothetis of M.D. Anderson Cancer Center, the virtue of finasteride is that it will reduce the overtreatment of prostate cancer. "Most of the time, treatment wouldn't help and may not be necessary," Logothetis told a reporter from the New York Times. "But the reality is that people are being operated on. We are trying to avoid a diagnosis to avoid a prevention whose value is disputed." With finasteride, "we're trying to overcome our other sins." But what is good medicine for the professional conscience may not be good medicine for the population.