Purpose: For patients with locally-advanced lung cancer treated with fractionated RT (60-66 Gy), toxicities, such as radiation pneumonitis, are driven by volume-effect dose indices (e.g. V20). If, during ART, daily lung volumes are aligned to the reference CT, it is unlikely that a significant reduction in V20 or mean-lung-dose (MLD) will be realized since large discrepancies will require mismatch of large volumes of lung. Here, we investigated strategies beyond those used traditionally to minimize OAR dose to understand the potential benefit of daily ART.
Methods: Retrospective data were evaluated from 10 locally-advanced, NSCLC patients previously treated on the Varian Halcyon. Adaptive workflow simulation was performed using an Ethos emulator on a virtual machine. Patients were originally planned using target and OAR criteria outlined in RTOG 1106. Separate plans were calculated for a set of pre-defined treatment geometries (IMRT or VMAT). Planning strategies compared in this study included use of standard RTOG 1106 optimization constraints (OPT_S) alone, and combined with ipsilateral lung constraints (OPT_IL) and ring constraints (OPT_RING). Plans were adapted for every fraction.
Results: Target coverage was maintained within RTOG 1106 constraints. Reduction of 1.11±0.78% (p=0.03) and 0.26±0.13Gy (p=0.01) was observed in V20 and MLD, respectively for OPT_IL relative to OPT_S. Corresponding reduction for OPT_RING was 0.45±0.54% (p=0.03) and 0.14±0.19Gy (p=0.05) relative to OPT_S. One patient exceeded the V20 constraint for OPT_S but met constraints when using OPT_IL. Relative to OPT_S, significant but minor improvement was observed in heart V30 and V50 (0.65±0.57% (p=0.01) and 0.46±0.55% (p=0.03), respectively) for OPT_RING and in esophagus D2cc (-0.31±0.25Gy, p=0.05) for OPT_IL.
Conclusion: Results suggest that a planning strategy incorporating ipsilateral lung constraints may offer the best option for sparing normal lung tissue doses and gleaning the most benefit from daily ART for patients with locally advanced lung cancers.
Funding Support, Disclosures, and Conflict of Interest: This project has been supported by a Varian Research grant.