Understanding PSA Utility
Study Underscores Negative Predictive Value
By Bill Malone
Recent reports from randomized trials have not agreed on the benefits of prostate specific antigen (PSA) screening, adding to the uncertainty with which many clinicians and lay people approach the test. This issue of Strategies explores a recent study that found a strong predictive value for PSA.
Although prostate specific antigen (PSA) screening can lead to over-diagnosis and over-treatment of prostate cancer, the disease remains the second leading cause of cancer death in American men, leaving clinicians in a dilemma about when to screen men using PSA and how to interpret results. Now new research recommends refining the utility of PSA, highlighting its negative predictive value (BMJ 2010;341:c4521). The nested case-control study analyzed PSA in archived frozen EDTA-plasma from 1,167 men participating in the Preventive Project in Malmö, Sweden during 1981–1982. The men were all 60 years old at the time of original blood draw and were followed for 25 years, until either the age of 85 or death. A total of 126 men were diagnosed with prostate cancer, and there were 43 cases of metastasis and 35 deaths from the disease.
The investigators found PSA levels to be highly accurate in predicting long-term risk of death or metastases. Participants with a PSA concentration ≥2 ng/ml at age 60 had, on average, 26 times the odds of dying from prostate cancer compared with those who had concentrations <2 ng/ml. In contrast, most of the deaths from prostate cancer came from men with the highest PSA concentrations—66% in the top 10% and 95% in men above the median. Still, only one in six participants at the 95th percentile of PSA (5.2 ng/ml) died of the disease before age 85.
One of the study’s distinctive contributions is that previous reports have not followed men long enough to determine the benefit of screening, said Hans Lilja, MD, PhD, lead author of the study. “Since the introduction of the PSA test, it’s notable that the kind of information that we have provided in this paper has not yet been gathered,” he said. “We know that there is no specific cut point or level of PSA at which there is no risk of having prostate cancer at biopsy. So, the question has remained whether this prostate cancer in the gland translates into something that is significant for the individual, meaning metastases or death. In this context, we uniquely show that for individuals with a PSA below the median at age 60, the risk of dying from prostate cancer is very, very low.” Lilja is -attending research clinical chemist in the departments of clinical laboratories, surgery, and medicine at -Memorial Sloan-Kettering Cancer Center in New York City. Unlike this study, others have only had a median follow-up of 7-10 years.
Because of when and where the blood samples were taken, the study also benefited from unusually clean controls. In 1981 in Sweden, PSA screening was uncommon, whereas recent randomized trials have been criticized for having contaminated controls because they took place in environments in which PSA screening was common. “This study clearly shows that PSA can do what we need it to do—predict cancer deaths,” said Bernard Cook, PhD, an immunodiagnostics consultant for Beckman Coulter. “This study is superior to other reports in several ways. It’s a very good experiment because the blood was collected and stored 25 years ago, so it’s a very clean study in terms of no previous screening. And, it’s looking at death or metastases for outcomes instead of biopsy results, which makes this study one-of-a-kind.”
The authors of the study suggest that at least half of men age 60 and older could potentially forego frequent follow-up PSA testing, reducing the chances of over-diagnosis and over-treatment. “For men over 60, focusing screening on those at increased risk—that is men with concentrations greater than the median at age 60—is likely to shift the ratio of harms to benefits for screening,” the authors wrote.
At the very least, these low-risk men might safely opt out of yearly testing, as some groups have recommended for men over 50, Cook said. “I wouldn’t go so far as to say we should never screen these men again, but perhaps every three or five years could be considered,” he said. “With less frequent testing for these men, not only will money be saved, but it’s less likely that we will see one of those PSA elevations that turns out to be prostatitis, and not cancer.”
Having a solid grasp of the negative predictive value of PSA can allow clinicians and researchers to put more focus on those men who need closer monitoring, Lilja emphasized. “We’ve shown here that the association between the PSA level at age 60 and outcome measured as metastases or death from prostate cancer is very strong. That does not eliminate the problem of the lack of specificity, but it does demonstrate that PSA is a basic tool which is very valuable for a first risk stratification through which we can eliminate a large cohort of men who have low risk, and instead focus on a smaller proportion of men who harbor 90 percent of the events,” he said. “This way, we could focus on this high-risk group and try to understand how to assess their risk better.”
In an accompanying editorial, Gerald Andriole, MD, also suggested that PSA testing should be tailored to individual risk. He recommended that young men at high risk of prostate cancer, such as those with a strong family history and higher baseline PSA concentrations, should be followed closely, while elderly men and those with a low risk of disease could be tested less often, if at all. “Approaches such as these will hopefully make the next 20 years of PSA based screening better than the first 20,” he said. Andriole is chief of urologic surgery at the Washington University School of Medicine in St. Louis.
Both Lilja and Cook noted that more research is needed to confirm and refine the findings. “One extremely important matter now is to replicate this data,” said Lilja. “Many comments have been made about the paper questioning whether our results are unique to the environment and ethnic group studied.” Another limitation is the age of the cohort: most screening paradigms begin at age 50, or even earlier. To address this, the team of researchers at Memorial Sloan-Kettering Cancer Center plan to conduct a similar analysis based on the approximately 25-year follow-up of a large representative cohort of men who were age 50 or older at the original blood draw.
Even with these limitations, the results of this study can help guide clinicians, Cook emphasized. “We’ve always known in general that a higher PSA is more dangerous, but this study offers clinicians some solid numbers to help put things into perspective to help counsel individual patients,” he said.
Join us for a discussion of this article, and interact with your peers, on LinkedIn.