Assessment of relative tumor burden in patients with clinical T1c prostate cancer treated with either external beam or radical prostatectomy.

The choice between external beam radiation therapy (EBRT) or retropubic radical prostatectomy (RPX) as potentially curative treatment for localized carcinoma of the prostate gland (CaP) has not been delineated in randomized studies. Both treatments are more effective if tumor burden is low. We sought to compare these two treatments in patients who had clinical stage T1c (cT1c) lesions and who were thought to have limited tumor burdens pretreatment. Sixty cT1c patients referred to the Department of Radiation Oncology received 66 Gy in 33 sessions of EBRT to localized prostate ports and 59 cT1c patients had RPX. No neoadjuvant nor early adjuvant therapies were prescribed. Radiotherapy success was defined biochemically as a nonrising prostate-specific antigen (PSA) of +/- 1.5 ng/ml. RPX success required a postoperative PSA that was undetectable (PSA <0.2 ng/ml by the Hybritech or Abbott IMx technics). Analysis for nonrising posttreatment PSA levels was performed using Kaplan-Meier and Cox regression methods. Mantel-Haenszel methods were used to determine odds ratios for treatment groups adjusting for potential confounders. We ultimately assessed the relative tumor burden by histologic examination of the RPX specimens. The two treatment groups, although not randomized, were statistically similar in biopsy Gleason Scores, transrectal ultrasonography calculated gland volumes, number of positive biopsy cores, and estimated amount of cancer identified on initial biopsies. Pathologic stage T3 was identified in 25% of RPX patients. Fifty to 60% of RPX specimens histologically had substantial tumor burden and by inference also the EBRT patients. At a median follow-up (F/U) of 36 months, 76% of RPX patients maintained an undetectable PSA, whereas 62% of EBRT patients had a PSA < 1.5 ng/ml at a median F/U of 29 months. The pretreatment PSA values significantly affected EBRT patients' risk of a rising posttreatment PSA level. Twenty-four months after treatment, RPX patients were 3.7 times more likely to maintain a nonrising PSA level (RPX patients posttreatment PSA < 0.2 ng/ml), than EBRT patients (posttreatment PSA < or = 1.5 ng/ml) (p = 0.006). Sixty-six gray in 33 sessions to localized EBRT ports is not sufficiently aggressive therapy for one third or more of patients with cT1c CaP. RPX alone is insufficient therapy for one fourth of cT1c patients. Analysis of the RPX specimens showed that many cT1c tumors have a significant tumor burden. Selection methodologies to separate out patients who require more than conventional dose or type of radiotherapy or more than RPX as monotherapy are needed. Pretreatment PSA and number of positive biopsies may assist this selection process.