Purpose: To assess the long-term outcome of conventional external beam radiation therapy in the management of clinically confined prostate cancer and to examine the proposition that radiation accelerates tumor growth in those who fail treatment.
Methods and materials: One thousand and forty-four men with T1-4NxM0 prostate cancer treated by conventional external beam radiation therapy at the Massachusetts General Hospital between 1977 and 1991 were analyzed. Median follow-up was 49 months. Failure was defined as: two sequential rises in serum prostate specific antigen (PSA) level; or a PSA > 1 ng/ml 2 or more years after radiation; or any clinical failure. Kaplan-Meir actuarial analyses were used to assess outcome.
Results: At 10 years only 40% of the T1-2 group remained disease free. When subdivided by grade, the well-differentiated tumors (Gleason 1-2) exhibited a 53% actuarial 10-year disease-free survival, moderately differentiated (Gleason 3) 42%, and poorly differentiated (Gleason 4-5) 20%. The corresponding values for the T3-4 men were 33% for Gleason 1-2, 20% for Gleason 3, and 10% for Gleason 4-5. Overall the value for T3-4 tumors was 18% at 10 years. On relapse the median PSA doubling times for the T1-2 patients were predicted by histology: 18.8 months for Gleason 1-2 patients; 11.1 months for Gleason 3; and 9.6 months for Gleason 5. Significant differences were found between the Gleason 3 and the Gleason 4-5 groups (p = 0.04) and the Gleason 1-2 and the Gleason 4-5 groups (p = 0.03). A wide range of doubling times was seen within each grade group. When compared with recently reported data on selected T1-2 patients who were managed by expectant observation there was no advantage over the first decade (and certainly no disadvantage) in terms of metastasis-free survival or disease-specific survival for the irradiated Gleason 1-3 patients. However, a gain was seen for those with Gleason 4-5 tumors.
Conclusion: Less than half of the T1-2NxM0 and less than one-fifth of the T3-4NxM0 patients receiving conventional radiation therapy were biochemically disease free at 10 years. The PSA doubling times on relapse show a wide variation. Grade was important in determining the rate of relapse suggesting that radiation does not induce a homogeneous acceleration of prostate tumors. A metastasis-free and disease-specific survival advantage was found for the poorly differentiated tumors when compared with similar patients reported in the literature who were managed initially by observation.