• © 2009 by American Society of Clinical Oncology

Reply to X. Mirabel

We thank Dr Mirabel¹ for his support of our study. We agree that it would be useful to establish a standard stereotactic body radiotherapy (SBRT) fractionation schedule. We chose to evaluate a three-fraction regimen based on previously published and ongoing studies of liver and lung SBRT at the time of protocol development. We speculated that if tumor hypoxia were an important component to radiation resistance, then a three-fraction regimen would at least allow for some degree of reoxygenation; likewise, the technical issue of geographic miss associated with tight target margins might be less likely with three fractions as opposed to one. Furthermore, three-fraction regimens are more convenient than longer schedules.

We consider 60 Gy in three fractions to be the standard regimen at our institution as long as normal tissue dose constraints are met. Whether this schedule is the ideal regimen is open to debate, but the only way to resolve that question would likely be a multi-institutional randomized phase II comparison study, where sufficient numbers of patients could be enrolled to detect small differences or establish the noninferiority of other regimens. Of possibly higher priority is the need for structured large studies that refine our understanding of ideal patient selection so we can properly apply this therapy to populations of patients who will receive measurable clinical benefit in terms of disease-free and, we hope, overall survival. Stratification variables should include tumor volume and primary tumor site. Secondary end points should include disease-free and overall survival as well as quality-of-life measures. Mendez-Romero et al² showed that quality of life was not significantly affected by liver SBRT; it would be valuable to confirm those findings in a larger study. Designing such a study, however, is not necessarily a simple task and would require careful consideration of the study's target population, especially the question of how to sequence SBRT relative to other new molecular therapies now commonly used in the management of a variety of metastatic solid tumors. As Dr Mirabel¹ mentioned, we had well-defined normal tissue constraints, including stomach and small bowel, where the maximum total point dose could not exceed 30 Gy.

Dr Mirabel¹ wondered whether there are opportunities for broadening patient eligibility with the use of tracking technology that might allow for smaller margins around the tumor to account for intrafraction motion. In our study there were no patients who met the size criteria for eligibility who were subsequently excluded (after treatment planning) for violation of normal tissue constraints that would have been avoided with a smaller treatment-planning margin. While we support all efforts to refine treatment planning and delivery technology, we would caution that in particular interfraction variability of organ deformation and intrafraction breathing pattern variability offer some practical limits to just how small the margin can be even with tumor tracking,^3,4 and the minimum necessary margin is likely similar to what was used in our study.