Predictors of cancer progression in T1a prostate adenocarcinoma

BACKGROUND. The biologic behavior of T1a prostate adenocarcinoma is variabl e. A critical issue in the management of patients with Tla prostate adenoca rcinoma is to distinguish those who will develop cancer progression from th ose who will not. Predictive factors that identify those at high risk of ca ncer progression are needed to stratify patients for treatment. In the curr ent study the authors attempted to identify such predictors of cancer progr ession in a large series of untreated patients with lengthy follow-up. METHODS. The authors studied 102 patients who were diagnosed with Tla prost ate adenocarcinoma (incidental tumor involving less than or equal to 5% of the resected prostatic tissue) at the time they underwent transurethral res ection of the prostate (TURP) at the Mayo Clinic between 1960-1970. None of these patients were treated. Patient ages ranged from 48-91 years (mean I standard deviation, 69 +/- 7 years). The average weight of the resected pro state tissue was 24 +/- 18 g (range, 3-115 g; median, 18 g). Tumor volume w as measured by the grid method. Cox proportional hazards models were used t o identify factors associated with cancer progression. Survival curves were estimated using the Kaplan-Meier method. RESULTS. Five-year and 10-year progression free survival rates were 93% and 87%, respectively. During the mean follow-up of 9.5 +/- 6.8 years (range, 0.3-31 years; median, 9.0 years), 14 patients developed clinical cancer pro gression, including 5 patients with systemic progression (1 with distant me tastases and 4 who died of prostate adenocarcinoma). The interval from diag nosis to clinical cancer progression ranged from 1-23 years (mean, 7.3 year s). The amount of resected prostate tissue (TURP weight) was associated wit h progression (P = 0.04). Patients with a TURP weight greater than or equal to 30 g had 100% progression free survival at 10 years compared with a pro gression free survival rate of 73% in patients with a TURF weight < 12 g. G leason score, tumor volume, number of chips involved by tumor, number of tu mor foci, and the presence of high grade prostatic intraepithelial neoplasi a were not significant in predicting cancer progression. There was a trend toward a worse prognosis with the increasing number of chips involved by ca ncer (P = 0.16). Patients with < 3 chips involved by cancer had a 88% 10-ye ar progression free survival rate compared with 73% in patients with greate r than or equal to 3 chips involved by cancer. CONCLUSIONS. The clinical course of Tla prostate adenocarcinoma is variable . If left untreated, a small but significant proportion of patients are at risk for disease progression and death. However, the current study found th at patients with a TURF weight greater than or equal to 30 g have an excell ent prognosis and can be managed conservatively. Cancer 1999;85:1300-4. (C) 1999 American Cancer Society.