Accelerated repopulation during radiotherapy: Quantitative evidence for delayed onset

The time of onset of accelerated tumor-clonogen repopulation produced by radiotherapy is an important, but unresolved question. The suggestion by Withers et al. [Acta Oncol 27:131–146, 1988] and colleagues [Maciejewski and Majewski: Radiother Oncol 21:163–170, 1991] that onset is delayed has been questioned due to the variety of data sources and the methodology used to reach this conclusion. We describe an analysis of a single clinical data set, which allows a more credible estimate of the time of onset of accelerated repopulation. We analyze a recent report of tumor control of T2 and T3 laryngeal cancer as a function of dose and overall time. By virtue of their broad dynamic range, these results allow an estimate to be made of the time of onset of accelerated repopulation based on a single self-consistent data set. The analyses suggest that rapid proliferation in these tumors is delayed by an estimated 32 days (95% confidence limits of 21–36 days). For these laryngeal tumors, and within the framework of the linear-quadratic model, the hypothesis that accelerated repopulation begins at the start of radiotherapy can be rejected in favor of the suggestion that accelerated repopulation begins after 3–5 weeks. This conclusion, if generally true, would have significant ramifications for the design of optimized radiotherapeutic protocols. © 1993 Wiley-Liss, Inc.