Analysis of a large number of clinical studies for breast cancer radiotherapy: estimation of radiobiological parameters for treatment planning

Numerous studies of early-stage breast cancer treated with breast conserving surgery (BCS) and radiotherapy (RT) have been published in recent years. Both external beam radiotherapy (EBRT) and/or brachytherapy (BT) with different fractionation schemes are currently used. The present RT practice is largely based on empirical experience and it lacks a reliable modelling tool to compare different RT modalities or to design new treatment strategies. The purpose of this work is to derive a plausible set of radiobiological parameters that can be used for RT treatment planning. The derivation is based on existing clinical data and is consistent with the analysis of a large number of published clinical studies on early-stage breast cancer. A large number of published clinical studies on the treatment of early breast cancer with BCS plus RT (including whole breast EBRT with or without a boost to the tumour bed, whole breast EBRT alone, brachytherapy alone) and RT alone are compiled and analysed. The linear quadratic (LQ) model is used in the analysis. Three of these clinical studies are selected to derive a plausible set of LQ parameters. The potential doubling time is set a priori in the derivation according to in vitro measurements from the literature. The impact of considering lower or higher T-pot is investigated. The effects of inhomogeneous dose distributions are considered using clinically representative dose volume histograms. The derived LQ parameters are used to compare a large number of clinical studies using different regimes (e.g., RT modality and/or different fractionation schemes with different prescribed dose) in order to validate their applicability. The values of the equivalent uniform dose (EUD) and biologically effective dose (BED) are used as a common metric to compare the biological effectiveness of each treatment regime. We have obtained a plausible set of radiobiological parameters for breast cancer: alpha = 0.3 Gy(-1), alpha/beta = 10 Gy and sub-lethal damage repair time T-rep = 1 h (mono-exponential behaviour is assumed). This set of parameters is consistent with in vitro experiments and with previously reported analyses. Using this set of parameters, we have found that most of the studies, using BCS plus whole breast RT and a boost to the tumour bed, have EUDs ranging from 60-70 Gy. No correlation is found between BED and the local recurrence rate. The treatments of BCS plus brachytherapy alone have a wide range of EUD (30-50 Gy), which is significantly lower than the treatments with whole breast EBRT plus a boost of the tumour bed. The studies with different fractionation schemes for whole breast EBRT also show a significant variation of EUD. Carefully designed clinical studies with large numbers of patients are required to determine clinically the relative effectiveness of these treatment variations. The derived LQ parameter set based on clinical data is consistent with in vitro experiments and previous studies. As demonstrated in the present work, these radiobiological parameters can be potentially useful in radiotherapy treatment planning for early breast cancer, e.g., in comparing biological effectiveness of different radiotherapy modalities, different fractionation schemes and in designing new treatment strategies.