Letters Section Editor: Stephen J. Lurie,
MD, PhD, Senior Editor.
In Reply: We agree with Drs Hakama and Auvinen
that our data are not ideally suited to a definitive study of the validity
of the PSA test because no biopsies were performed and patients were not specifically
evaluated for the development of prostate cancer over time. Nonetheless, our
study provided an unusual opportunity to observe natural variations in levels
of PSA over time, which were largely unaffected by clinical interventions.
Our results do not address the validity of the PSA test, but strongly suggest
that the amount of circulating PSA varies markedly over time in an individual
patient. This variability challenges the recommendation for further invasive
testing, including prostate biopsy, solely on the basis of a single PSA result.
The fluctuations in PSA levels we observed are not related to "laboratory
problems," as suggested by Hakama and Auvinen. As we detailed in our article,
all PSA tests were performed in a research laboratory on properly stored and
prepared samples. We focused our attention on the frequency with which elevated
levels of PSA returned to the normal range, because men with such fluctuating
levels are at greatest risk of unnecessary biopsy.
Eastham JA, Scardino PT, Begg CB. Reliability and Validity of Prostate-Specific Antigen. JAMA. 2003;290(13):1705-1706. doi:10.1001/jama.290.13.1705-a