Prognosis and Survival in Prostate Cancer

While the histopathologist has relatively little difficulty in making the diagnosis of prostatic adenocarcinoma, the clinician may have considerable problems in deciding what this diagnosis implies for the patient. A pro­ portion of adenocarcinomas do not seem to progress and are thus "latent". This has been demonstrated from step-sectioning of prostates examined after autopsy which shows that, irrespective of the incidence of invasive prostatic cancer in the population of origin, the frequency of small "latent" cancers does not appear to vary much across populations (Breslow et al. 1977). This implies that varying proportions of these tumours do not pro­ gress to the clinically manifest form. Any series of prostate cancer contains a proportion of tumours which do not behave in a truly malignant fashion. While a proportion of cases of cancer of the prostate present with clinical symptoms, for an increasing number the diagnosis is made following histo­ logical examination of tissue removed to relieve the symptoms of benign prostatic hyperplasia. Transurethral resection of the prostate (TURP) has become more widespread, and this is strongly correlated with the rising incidence of prostatic carcinomas (Potosky et al. 1990). Figure 20.1 illustrates the age-standardised rates (European Standard Population) of transurethral resections in Scottish men, of all ages, over the period 19751992. There was a greater than threefold increase in the rate for this procedure during this period, with the rate of increase most pronounced in men aged over 65 years. Further, the rising prevalence of prostate-specific antigen (PSA) estimations also leads to biopsy of the prostate and the diagnosis of adenocarcinoma and, more recently, prostatic intraepithelial neoplasia. Most population-based cancer registries publish the proportion of cancer, including prostate cancer, diagnosed on the basis of histology (Parkin et al. 1992). Table 20.1 illustrates the proportions of prostate cancers for which the diagnoses were histologically verified for populations with con­ trasting incidence rates. These data indicate that, Shanghai and Osaka apart, the proportion of diagnoses with histological verification is very high irrespective of the level of incidence. However, such tabulations do not, and cannot, distinguish between invasive and "latent" forms of the disease. The