Preliminary observations on biochemical relapse-free survival rates after short-course intensity-modulated radiotherapy (70 Gy at 2.5 Gy/fraction) for localized prostate cancer

Purpose: To compare the preliminary biochemical relapse-free survival rates between short-course intensity-modulated radiotherapy (SCIM-RT) delivering 70 Gy in 28 fractions and three-dimensional conformal radiotherapy (3D-CRT) delivering 78 Gy in 39 fractions. Methods and Materials: Between January 1998 and December 1999, 166 patients were treated with SCIM-RT and 116 with 3D-CRT. The SCIM-RT cases were treated to 70 Gy (2.5 Gy/fraction) using 5 intensity-modulated fields using a dynamic multileaf collimator. The BAT transabdominal ultrasound system was used for localization of the prostate gland in all SCIM-RT cases. The 116 3D-CRT cases were treated to 78.0 Gy (2.0 Gy/fraction). The study sample therefore comprised 282 cases; 70 Gy in 28 fractions is equivalent to 78 Gy in 39 fractions for late-reacting tissues, according to the linear-quadratic model. The median follow-up for all cases was 25 months (range 3-42). The median follow-up was 21 months for the SCIM-RT cases (range 3-31) and 32 months for the 3D-CRT cases (range 3-42). The follow-up period was shorter for the SCIM-RT cases, because SCIM-RT was started only in October 1998. Biochemical relapse was defined as 3 consecutive rising prostate-specific antigen levels after reaching a nadir. The analysis was then repeated with a more stringent definition of biochemical control: reaching and maintaining a prostate-specific antigen level of less than or equal to0.5 ng/mL. Radiation Therapy Oncology Group toxicity scores were used to assess complications. Results: For the 282 patients, the biochemical relapse-free survival rate at 30 months was 91% (95% confidence interval 88-95%). The biochemical relapse-free survival rate at 30 months for 3D-CRT vs. SCIM-RT was 88% (95% confidence interval 82-94%) vs. 94% (95% confidence interval 91-98%), respectively. The difference was not statistically significant between the two treatment arms (p = 0.084). The multivariate time-to-failure analysis using the Cox proportional hazards model for clinical parameters showed the pretreatment prostate-specific antigen level (p <0.001) and biopsy Gleason score (p <0.001) to be the only independent predictors of biochemical relapse. Clinical T stage (p = 0.66), age (p = 0.15), race (p = 0.25), and neoadjuvant androgen deprivation (p = 0.66) were not independent predictors of biochemical failure. SCIM-RT showed only a trend toward a better outcome on multivariate analysis (p = 0.058). Late rectal toxicity was limited; the actuarial combined Grade 2 and 3 late rectal toxicity rate at 30 months was 5% for SCIM-RT vs. 12% for 3D-CRT (p = 0.24). Grade 3 late rectal toxicity (rectal bleeding requiring cauterization) occurred in a total of 10 patients. The actuarial Grade 3 late rectal toxicity rate at 30 months was 2% for the SCIM-RT cases and 8% for the 3D-CRT cases (p = 0.059). Late urinary toxicity was rare in both groups. Conclusion: With the currently available follow-up period (less than or equal to30 months), the hypofractionated intensity-modulated radiotherapy schedule of 70.0 Gy delivered at 2.5 Gy/fraction had a comparable biochemical relapse profile with the prior 3D-CRT schedule delivering 78.0 at 2.0 Gy/fraction. The late rectal toxicity profile has been extremely favorable. If longer follow-up confirms the favorable biochemical failure and low late toxicity rates, SCIM-RT will be an alternative and more convenient way of providing dose escalation in the treatment of localized prostate cancer. (C) 2002 Elsevier Science Inc.