Gleason score and radiation dose
The radiation oncology community quickly embraced this statement and adopted this definition. Since its near universal adoption within the radiation oncology community, several points have remained controversial. To counter these arguments, several groups correlated this definition of biochemical failure with clinical outcome demonstrating the robustness of this definition. Some advocated an absolute posttreatment PSA nadir instead of PSA trend to assess bNED control, 31 although this approach is not widely accepted.
Others pointed out the difficulties introduced with backdating the DOF and suggested alternative definitions. However, in the end, it is not just PSA which predicts tumor control. The most important factors which consistently predict bNED control rates on multivariate analysis are pretreatment PSA, T Stage, Gleason score and radiation dose. Based on these variables, patients can be divided into prognostic groups that can be used to predict outcome as well as direct treatment. At Fox Chase, risk groups are defined as follows. Low risk group patients include those with PSA ≤10 ng/ml, Gleason score 2-6 and T1c/T2a disease. The high risk group consists of patients with Gleason score 8-10, PSA >20 ng/ml or T3/T4 disease. Intermediate risk patients are with PSA >10 ng/ml or Gleason score 7 or T2b/T2c disease. As our data has matured, the presence of perineural invasion is no longer an independent risk factor. Conventional RT results A direct comparison of results for older series of patients treated with conventional doses of radiation from the PSA era is difficult due to differing definitions of bNED control, unequal distribution of critical prognostic factors between series, and varying lengths of followup. However, as pre treatment PSA levels increase, rates of bNED control consistently decrease for institutions using conventional doses of RT.24, 25, 41 bNED control for 1 044 men with stage T1-T4 prostate cancer treated at Massachusetts General Hospital was reported by Zietman et al. Five hundred and four men had T1-T2 disease and had rates of bNED control at 5 and 10 years of 60% and 40%, respectively.
For the 540 men with stage T3-T4 disease treated with conventional doses of radiation, bNED control rates at 5 and 10 years were 32% and 10%, respectively. Similar results were reported by Horwitz et al. for 480 patients treated with conventional doses of ra- diation at William Beaumont Hospital. The 5- year rate of bNED control ranged from approximately 80% for T1 disease to 25% for patients with T3 disease. Identical trends in outcome were observed when patients were stratified by pretreatment PSA and Gleason score. As pre-treatment PSA and Gleason score increased, bNED control decreased.24 At the Cleveland Clinic Keyser et al.42 reported the results for a group of patients with T1 and T2 disease treated with EBRT or prostatectomy.
The 607 patients in this series all had pretreatment PSA levels ≤10 ng/ml and T1 or T2 disease. For the 253 patients treat- ed with conventional doses of radiation, bNED control rates stratified by pretreatment PSA <4 or between 4-10 ng/ml were 100% and 65%, respectively. No statistically significant different rate of bNED control was observed between the radiation and surgery groups. 3D conformal and IMRT results Results of treatment with 3DCRT with longterm (8-10 years) PSA followup are available and demonstrate superior bNED control rates using this technique compared with conventional techniques.
Because higher doses of radiation can be delivered to the prostate using 3DCRT (without substantially higher rates of normal tissue complications), evidence indicates that bNED control rates are significantly improved. Long-term data from institutions including FCCC, Memorial Sloan-Kettering Cancer Center (MSKCC), the Cleveland Clinic and MDACC show increased rates of bNED control, especially for patients with pretreatment PSA levels >10 ng/ml (Table I 43-45 and Figures 7, 8).5, 18, 46, 47 In 1999, the results using EBRT alone in 1 765 men with T1 and T2 prostate cancer from 6 institutions was summarized and reported. The biochemical durability of RT as well as the effect of prognostic factors on outcome was described.48 As a follow up to this experience, 9 institutions pooled data on nearly 5 000 patients treated with EBRT alone between 1986 and 1995 with a median follow-up of more than 6 years. The results of this collaborative effort were first reported at the 44 Annual Meeting of ASTRO in November 2002.
