Is α/β for prostate tumors really low?

Purpose: Brenner and Hall's 1999 paper estimating an alpha/beta value of 1.5 Gy for prostate tumors has stimulated much interest in the question of whether this ratio (of intrinsic radiosensitivity to repair capacity) is much lower in prostate tumors than in other types of tumors that proliferate Faster. The implications for possibly treating prostatic cancer using fewer and larger fractions are important. In this paper we review updated clinical data and present somewhat different calculations to estimate alpha/beta. Methods and Materials: Seventeen clinical papers published from 1995 to 2000 were reviewed to obtain estimates of biochemical control from radiotherapy alone using external beam, I-125 implants, or Pd-103 implants, The focus was on intermediate risk patients. Three methods of estimating alpha/beta were employed. First, a simple two-step graphical comparison of isoeffective doses from external beam and implant modalities was made, to see which value of alpha/beta predicted the observed identity of biologic effect. Second, the same data were subjected to Direct Analysis (maximum likelihood estimation), from which an estimate of alpha/beta and also of the T1/2 of repair of sublethal damage in the tumors (both with confidence intervals) were obtained. Third, preliminary clinical data comparing two different sizes of high-dose boost doses were analyzed in which significantly different bNED was observed at 2 years. Results: The second method gave the definitive result of alpha/beta = 1.49 Gy (95% CI 1.25-1.76) and T1/2 = 1.90 h (95% CI 1.42-2.86 h). The first method gave a range from 1.4 to 1.9 Gy and showed that if mean or median dose were used instead of prescribed dose, the estimate of alpha/beta would be substantially below 1 Gy. The third method, although based on early follow-up, was consistent with low values of alpha/beta in the region of 2 Gy or below. The estimate for T1/2 is the first value reported for prostate tumors in situ. Conclusions: All the estimates point toward low values of alpha/beta, at least as low as the estimates of Brenner and Hall, and possibly lower than the expected values of about 3 Gy for late complications. Hypofractionation trials for intermediate-risk prostatic cancer appear to be indicated. (C) 2001 Elsevier Science Inc.