COMPARISON OF THE INCIDENCE OF LATENT PROSTATE CANCER DETECTED AT AUTOPSY BEFORE AND AFTER THE PROSTATE SPECIFIC ANTIGEN ERA

Publisher Website
Google Scholar
Abstract
Most data regarding the prevalence of latent prostate cancer found only at autopsy are from old reports. To determine if significant differences exist in the prevalence of latent prostate cancer between periods before and after the advent of screening for prostate cancer, we compared 2 groups of men undergoing autopsy during the 2 periods. Our institutional autopsy record database was searched to identify all men found to have prostate cancer before or after death between 1955 and 1960 (total 3,307 men and 1,578 men older than 40 years), and between 1991 and 2001 (total 2,938 men, 1,380 men older than 40 years). We calculated the age based incidence of latent prostate cancer detected only at autopsy in an at risk population of men (older than 40 years). We also compared Gleason grade distribution and proportion of stage cT3 or greater cancers between the 2 periods. Between 1955 and 1960 the prevalence of latent prostate cancer detected only at autopsy in men older than 40 years was 4.8% compared to 1.2% (p < 0.0001) between 1991 and 2001. A significant decrease in the prevalence of latent, autopsy detected cancers was observed in men 70 to 89 years old at death. Autopsy detected cancers were found to be grossly invading adjacent structures (stage cT3 or greater) in 17 of 76 (22%) cancers discovered between 1955 and 1960, while none of the latent prostate cancers found in the 1991 to 2001 period were found to extend grossly beyond the prostate. Autopsy rates are decreasing at our institution. With the more widespread use of screening, the prevalence of latent prostate cancer has decreased 3-fold. The decrease in the prevalence of latent prostate cancer is especially dramatic in men older than 70 years. Further study will determine the significance of many of the tumors currently detected clinically, which may have been latent and found at autopsy if not for screening.

This publication has 13 references indexed in Scilit: