PROSTATE-CANCER TRENDS IN CANADA - RISING INCIDENCE OR INCREASED DETECTION

Objectives: To analyse trends in the incidence and mortality rates of prostate cancer in Canada according to age distribution, temporal patt ern and provincial variation; to determine any association with the ra te of prostatectomy; and to determine whether any observed increase in the rate of prostate cancer was due to an increase in the detection r ate. Design: Descriptive epidemiologic study based on Canadian populat ion data from 1959 to 1989 and chart review from one Canadian hospital . Setting: The chart review was conducted at the Ottawa Civic Hospital . Subjects: The data on prostate cancer trends were obtained from the Canadian population. Charts were reviewed for two groups of patients: (a) men discharged from inpatient care during 1976 and 1986-87 with pr ostate cancer first diagnosed in the same year and (b) men who underwe nt transurethral resection of the prostate (TURP) during 1976 and 1986 . Outcome measures: Incidence and mortality rates of prostate cancer, rates of prostatectomy and TURP, and correlations between them. From t he hospital data, changes between 1976 and 1986-87 in distribution of cancer stages, distribution of cases detected incidentally after surge ry for suspected benign prostatic hypertrophy and average number of sl ides analysed per gram of tissue obtained from prostatectomy. Results: The epidermiologic data showed that the age-adjusted incidence rates increased by 72% overall, an increase seen in all age groups over 60 y ears. The mortality rates increased by 29% overall, primarily in men o ver 85 years old. The prostatectomy rate increased by 55%. There were significant linear correlations between the national and provincial in cidence rates of prostate cancer and the TURP rates. The chart review revealed that during 1976, 53% of the cases of prostate cancer diagnos ed were localized, as compared with 75% in 1986-87 (p < 0.01). The pro portion of tumours diagnosed incidentally in men undergoing TURP incre ased by 11%, whereas the number of procedures did not increase. Signif icantly more slides per gram of tissue were analysed in 1986-87 than i n 1976 (p < 0.01). Conclusions: The correlations between the incidence rates of prostate cancer and those of TURP suggest that increased tre atment of benign prostatic disease has led to increased detection of p rostate cancer. Extrapolation of the data obtained from the chart revi ew indicates that the increase in observed incidence rates can be attr ibuted to an increase in the rate of localized disease and thus primar ily to early detection rather than to elevated risk. However, because the rate of death from prostate cancer was elevated in elderly men, in creases in exposure to unestablished risk factors cannot be ruled out.