A new report in the Journal of Clinical Oncology suggests that — at least between 2000 and 2005 — “excessive PSA screening in elderly men with limited life expectancies” was a significant problem, and may remain so today.
The role of PSA testing in older men with life expectancies of 15 years and less is controversial. The U.S. Preventive Services Task Force (USPSTF), when it last reported on this topic, recommended that prostate cancer testing was inappropriate in men aged 75 and older. By comparison, The “New” Prostate Cancer InfoLink has always felt that the Iowa Consensus position is more appropriately nuanced. Having said that, there clearly does come a point at which continuing to give PSA tests to elderly males with no prior evidence of prostate cancer becomes a pretty pointless exercise. Are we going to actively treat most of men of 85 years of age for localized prostate cancer if we actually find it? We sincerely hope not. Will there be rare exceptions? Probably.
Drazer et al. (with a full appreciation of the nature of the controversy) set out to describe the rates and predictors of PSA testing among older men in the United States.
They used as their database the population-based 2000 and 2005 National Health Interview Survey (NHIS). “PSA screening” was defined as the use of a PSA test as part of a routine exam within the preceding year. The NHIS collected demographic, socioeconomic, and functional characteristics, and a validated 5-year estimated life expectancy was calculated for all participants.
The results of the study showed the following:
- The PSA screening rate was 24.0 percent in men aged 50 to 54 years.
- The PSA screening rate increased steadily with age until a peak of 45.5 percent among men of age 70 to 74 years.
- Among men of > 74 years of age, screening rates then gradually declined by age.
- The PSA screening rate was 24.6 percent among men aged 85 years or older who reported being screened.
- Among men aged 70 years or older, screening rates varied by estimated 5-year life expectancy:
- For men with high life expectancies (≤ 15 percent probability of 5-year mortality) the screening rate was 47.3 percent.
- For men with intermediate life expectancies (16-48 percent probability of 5-year mortality), the screening rate was 39.2 percent.
- For men with low life expectancies (>48 percent probability of 5-year mortality), the screening rate was 30.7 percent.
- Estimated life expectancy and age remained independently associated with PSA screening in multivariate analysis.
Drazer and his colleagues conclude that — in the USA between 2000 and 2005 — rates of PSA screening were associated with age and estimated life expectancy.
Now there is very good evidence that, for men initially diagnosed with low- to intermediate-risk, truly localized prostate cancer, there is little to no clinical impact on their life within 5 years. This does tend to make one ask what the possible benefit could be to giving a PSA test to the vast majority of men with a life expectancy of 5 years or less? And yet (between 2000 and 2005) we were apparently giving PSA tests to 30-40 percent of such men and to about 25 percent of men of 85 years of age or more (most of whom, presumably, also had life expectancies of 5 years or less). Were most of these men actually asking for these tests, or were their doctors just giving them these tests as part of a routine? Are the same sorts of rates of PSA testing evident today in a similar patient cohort, or have the recommendation of the USPSTF and the results of the PLCO and ERSCP trials lowered these rates? We have no real idea.
The bottom line to all of this is that there is a major disconnect between what a PSA test can actually tell a doctor and his elderly patient and what a smart doctor and a smart elderly patient would actually do with that information. The question that all doctors (and their elderly male patients) probably need to be asking themselves is, “If you did or didn’t have PSA data in a man with a life expectancy of < 5 years (or perhaps even < 10 years) and no prior history of risk for prostate cancer, would it actually make any difference to your recommendations for his management?” The situation is clearly different for men who have shown prior risk for prostate cancer … but not a lot different if that history includes at least two prior negative biopsies.