The new data on prostate cancer screening: What should we do now?
THE US SCREENING TRIAL
What about the US trial? Unfortunately, it was beset by limitations that make its interpretation extremely difficult.3
Between 1993 and 2001, 76,693 men were randomized to prostate cancer screening with PSA testing and digital rectal examination, or else to usual care.
The problem is that in the United States “usual care” often includes PSA testing. Thus, 34% of men participating in the trial had had a PSA test within 3 years prior to enrolling on the trial, and 52% of the control group had PSA testing during the trial. In the group randomized to screening, 85% complied with PSA testing. This trial thus compared one group in which most were screened at least once against another group in which 85% were screened regularly. Rather than asking whether screening is effective, the trial compared two different PSA screening schedules.
Thus, it was no surprise that there was less than a 25% increase in the cancer detection rate and less than a 30% reduction in the likelihood of having detectable metastatic disease at the time of diagnosis. And after 7 years of follow-up, the two groups showed no statistically significant difference in the likelihood of dying of prostate cancer.
SHOULD MEN BE SCREENED FOR PROSTATE CANCER?
The European trial provides strong evidence that PSA testing reduces prostate cancer mortality rates,2 while the US trial sheds little light on the subject. But does this mean that men should be screened routinely?
It’s not that simple. The 75% false-positive rate of PSA testing and the high number needed to treat (n = 48) to save one life represent significant harmful effects of prostate cancer screening that must be factored into the decision-making process. And we know from other studies that half or more of men undergoing prostate cancer treatment will report erectile dysfunction, while a smaller number will experience urinary incontinence. More and more men without detectable meta-static disease are being treated with medical or surgical castration, which is associated with loss of libido, osteoporosis, weight gain, loss of muscle, and an increased risk of diabetes and death from cardiovascular disease. Prostate cancer treatments also result in large medical bills, which are a source of hardship for the increasing number of Americans with inadequate health insurance.
The benefit of PSA testing is limited by several key facts:
- It is an inaccurate test with a high false-positive rate
- The treatment of prostate cancer results in serious adverse effects
- Most men will develop prostate cancer if they live into their 70s
- Most prostate cancers are not life-threatening.
Whereas cervical cancer screening typically detects precancerous lesions that can be treated superficially and colon cancer and breast cancer screening often detect precancerous lesions or small tumors that can be removed with relatively minor surgery, prostate cancer treatment is radical and often results in significant long-term adverse effects. The benefits of PSA screening must be balanced against the harm.
One way out of this dilemma, as discussed in Dr. Klein’s article, is to eliminate the reflex progression from PSA elevation to biopsy and from positive biopsy to treatment. As Dr. Klein discusses, variables other then PSA help predict the likelihood that a biopsy would detect a clinically significant cancer and can reduce the likelihood of performing unnecessary biopsies.1
Similarly, there is growing interest in active surveillance for clinically localized low- or intermediate-grade prostate cancers, thus sparing men unnecessary and aggressive treatment.8 The challenge is determining which cancers are indolent and which are aggressive. Until we have accurate tools to make such a distinction, overtreatment will remain a problem as men and their doctors opt for aggressive treatment in the face of uncertainty about a cancer’s true danger.
MOVING FORWARD
This year has brought strong evidence that PSA screening lowers the risk of dying of prostate cancer, but at a cost of overdiagnosis, overtreatment, and a significant burden of treatment side effects and costs. Moving forward will depend on a more sensitive and specific screening test, tools for better predicting which cancers actually need treatment, and treatments that result in fewer long-term side effects. Progress on all these fronts can be expected in the future.
