Potential implications of the bystander effect on TCP and EUD when considering target volume dose heterogeneity

Michael J Balderson,Charles Kirkby

DOI: https://doi.org/10.3109/09553002.2014.942014

Abstract:Purpose: In light of in vitro evidence suggesting that radiation-induced bystander effects may enhance non-local cell killing, there is potential for impact on radiotherapy treatment planning paradigms such as the goal of delivering a uniform dose throughout the clinical target volume (CTV). This work applies a bystander effect model to calculate equivalent uniform dose (EUD) and tumor control probability (TCP) for external beam prostate treatment and compares the results with a more common model where local response is dictated exclusively by local absorbed dose. The broad assumptions applied in the bystander effect model are intended to place an upper limit on the extent of the results in a clinical context. Materials and methods: EUD and TCP of a prostate cancer target volume under conditions of increasing dose heterogeneity were calculated using two models: One incorporating bystander effects derived from previously published in vitro bystander data ( McMahon et al. 2012 , 2013a); and one using a common linear-quadratic (LQ) response that relies exclusively on local absorbed dose. Dose through the CTV was modelled as a normal distribution, where the degree of heterogeneity was then dictated by changing the standard deviation (SD). Also, a representative clinical dose distribution was examined as cold (low dose) sub-volumes were systematically introduced. Results: The bystander model suggests a moderate degree of dose heterogeneity throughout a target volume will yield as good or better outcome compared to a uniform dose in terms of EUD and TCP. For a typical intermediate risk prostate prescription of 78 Gy over 39 fractions maxima in EUD and TCP as a function of increasing SD occurred at SD ∼ 5 Gy. The plots only dropped below the uniform dose values for SD ∼ 10 Gy, almost 13% of the prescribed dose. Small, but potentially significant differences in the outcome metrics between the models were identified in the clinically-derived dose distribution as cold sub-volumes were introduced. Conclusions: In terms of EUD and TCP, the bystander model demonstrates the potential to deviate from the common local LQ model predictions as dose heterogeneity through a prostate CTV varies. The results suggest, at least in a limiting sense, the potential for allowing some degree of dose heterogeneity within a CTV, although further investigation of the assumptions of the bystander model are warranted.