DNA ploidy, Gleason score, pathological stage and serum PSA levels as predictors of disease-free survival in C-D1 prostatic cancer patients submitted to radical retropubic prostatectomy.

We report our experience with 85 prostatic cancer patients aged 51-79 years, who underwent radical retropubic prostatectomy from 1989 to December 1994 (mean follow-up 35 months). In order to get a more relevant analysis we chose to describe in detail only pathological C-D1 cases and to subdivide the patients, according to the Gleason sum, into G2-G5 and G6-G10 groups. Means of pre- and postsurgery PSA levels were ranked by DNA ploidy and presence or absence of recurrence: aneuploid patients showed lower levels of PSA production that may be due to cell dedifferentiation. However, in patients who developed recurrence, postsurgery PSA levels were higher (p < 0.005). The influence of DNA ploidy on disease-free survival was evaluated: the cumulative survival proportion was better in diploid (0.3581) than in aneuploid patients (0.2996). Using the Cox proportional hazard model with age, Gleason sum, DNA ploidy and presurgery PSA levels as covariates, we demonstrated that, in our series, only the presurgery PSA level was an important and significant predictor of recurrences (p < 0.005). Considering global recurrences with age, Gleason sum and presurgery PSA levels kept fixed, DNA aneuploidy conferred a relative risk 2.3 times higher than diploidy. When, in the same analysis, we introduced postsurgery PSA levels, only DNA ploidy and the latter variable kept statistical significance with a relative risk of 2.5. Considering only local and distant recurrences (with exclusion of those identified by elevated PSA levels) the relative risk was 3.9 and 3.8, respectively. These data support the critical role of nuclear DNA analysis as predictor of outcome after surgery even in this discussed subset of patients (C-D1).