Correlation of currently used radiobiological parameters with local control and acute and late mucosal toxicity in randomised studies of altered fractionation for locally advanced head and neck cancer

Abstract

Aims: There has been a resurgence in interest in radiobiological modelling in head and neck cancer. The aim of this study was to determine if currently used parameters accurately predict both tumour and toxicity outcomes.

Materials and methods: Trials were identified from a recent meta-analysis of altered fractionation. The tumour biologically effective dose (tBED; α/β=10Gy, t(k) [onset time of accelerated repopulation]=22 days, t(p) [average doubling time during accelerated repopulation]=3 days, α=0.3Gy(-1)), acute mucosal biologically effective dose (amBED; α/β=10Gy, t(k)=7 days, t(p)=2.5 days, α=0.3Gy(-1)) and late mucosal biologically effective dose (lmBED; α/β=3Gy) were calculated for each arm of each trial. The correlation between the absolute percentage difference in BED between treatment arms and the observed percentage difference in local control, acute grade 3 mucositis and late grade 3 mucosal reaction was then assessed.

Results: A strong correlation was observed between the percentage difference in tBED and the percentage difference in local control (P=0.006). A trend towards a correlation was seen between the percentage difference in amBED and the percentage difference in acute grade 3 mucositis (P=0.06). A significant correlation was observed between the percentage difference in lmBED and the percentage difference in grade 3 late mucosal toxicity (P=0.02). However, a 15% decrease in lmBED between control and experimental arms of the study was necessary for any sparing of late mucosal toxicity to be observed.

Conclusions: Currently used parameters for tumour accurately predict outcomes in randomised trials of altered fractionation. Although the relationship may be more complex for late mucosal reaction, the presence of a correlation is noteworthy given the infrequent reporting or occurrence of this toxicity. In the future, radiobiological modelling with the addition of volumetric parameters will be highly relevant, given attempts to dose escalate with intensity-modulated radiotherapy in poor risk patients and de-escalate in patients with an excellent prognosis.