Reply to editorial “From Probability to Practice: Demystifying When Lung Shunt Fraction Matters in 90Y Selective Internal Radiation Therapy”

We appreciated the favorable commentary on our recent article1 by Li et al.2 Nevertheless, we would like to offer some points of note. Our focus on whole-liver dose thresholds and whole-liver volume instead of perfused volume (PV) and dose may seem misaligned with modern treatment planning. However, we aimed to develop a method employing case-specific parameters that would be readily available for the largest number of patients. Whole liver volume will always be known prospectively, while PV estimation before any angiographic procedures remains speculative. Redefining the boundary lung shunt fraction (LSFbound) metric in terms of PV and PV dose, as outlined by Li et al., may seem intuitive, but this approach explicitly requires measured or estimated PV. We contend this will amplify uncertainty and reduce clinical applicability, rather than simplify our original framework in a meaningful way. The other simplification, a nominal 1-kg lung mass instead of a patient-specific lung mass, also raises concerns. This approach mirrors historical 90Y-SIRT treatment planning with generalized assumptions for all patients. Instead, patient-specific data can often be easily acquired and employed.3,4 The use of 1-kg lung mass for most or all patients will produce non-linear changes in computed LSFbound that may affect appropriate patient triage for LSF assessment. There would be a particular risk for patients with a lung mass well below 1 kg. Such patients cannot always be identified through other clinical factors. Finally, applying the original or modified LSFbound method to new clinical data, as outlined by Li et al., was not especially informative. The cases presented were from a “no-MAA” workflow, maintaining clinical characteristics (small HCC, small PV) associated with minimal risk for high LSF.5 The observation that the LSFbound approach worked ideally for such cases, with essentially no risk of inappropriate triage for LSF assessment, was predictable. While reassuring, these findings do not substantially extend the scope of validation beyond what would be anticipated for such cases. We instead encourage the field to consider a prospective application of the LSFbound approach across the full spectrum of 90Y-SIRT eligible patients at their institution, if deemed appropriate and of interest. Supplementary material is available at Radiology Advances online. No funding support was associated with this study. Please see ICMJE form(s) for author conflicts of interest. These have been provided as supplementary materials. M.A.T. receives research support from Boston Scientific. C.D.M. receives research support from Boston Scientific, is an advisor for AstraZeneca Pharmaceuticals LP, Boston Scientific, and Eisai, and receives grant funding from the Society of Interventional Oncology and the American Cancer Society.