Draisma G et al. (2006) Gleason score, age and screening: modeling dedifferentiation in prostate cancer. Int J Cancer 119: 2366–2371

Since PSA screening became widespread, most prostate cancers detected have low Gleason scores, a characteristic associated with a favorable prognosis. This observation raises an interesting question: does PSA screening detect more tumors with low Gleason scores than would be detected clinically, or does the early introduction of treatment for prostate cancers detected by PSA screening improve the course of disease, by preventing the dedifferentiation that results in an increased Gleason score?

Draisma et al. have addressed this question by assessing the Gleason scores of 2,204 prostate cancers detected at a single Rotterdam center (data obtained from the population-based European Randomized Study of Screening for Prostate Cancer, which involved >40,000 participants). Diagnoses of prostate cancer in unscreened men, and in participants who developed cancer after a negative PSA result (i.e. who were not diagnosed by PSA screening) were confirmed by referring to Dutch cancer-registry data. The authors constructed two semi-Markov models: in one, dedifferentiation only occurred before PSA screening, whereas in the other it could also occur during the PSA-screening phase. They found that the model that permitted dedifferentiation during PSA screening fit the data markedly better, which indicates that prostate cancers dedifferentiate throughout the preclinical phase of their development, and can progressively increase in Gleason score.

Early detection and treatment of prostate cancers could potentially prevent tumor dedifferentiation, say Draisma et al., although the available 5 years of follow-up data were insufficient to show this effect directly. They speculate that PSA screening might also reduce prostate cancer mortality.