A Priori Knowledge of Maximum Achievable Organ at Risk Sparing: Impact on Organ Doses and Patient Reported Outcomes

Abstract

IMRT planning is challenging, at least in part, due to the essentially-infinite number of possibilities, and since the planner is never sure that they have gotten "the best", or at least "nearly the best" achievable plan. We prospectively assessed the impact of providing IMRT planners with a priori knowledge of the maximum achievable dose sparing for organs at risk (OARs) for patients on both dosimetric and clinical endpoints. We hypothesized that a priori knowledge of maximum achievable sparing would improve dosimetric and clinical (i.e., patient reported outcome) measures.We examined patients receiving IMRT for oropharynx cancer on prospective clinical trials from 2012-2019 at our institution. A tool that generates estimates of maximum achievable dose sparing for organs at risk was used starting July 2016. Patients treated pre-implementation (i.e., baseline) and post-implementation (i.e., experimental) were compared for the endpoints: (1) Doses received by various OARs MINUS doses achievable (per the tool), termed the 'excess of feasible dose'; (2) Plan quality metrics (e.g., target homogeneity and conformity), and (3) Patient reported outcomes (PRO) per questionnaires completed at 3-, 6- and 12-months post-treatment.139 patients were analyzed (60-baseline cohort, 79-experimental). (1) The excess of feasible dose was lower in the experimental (vs baseline) cohort, for the contralateral parotid (9.6 vs 5.5 Gy; P < 0.001), the ipsilateral parotid (17.6 vs 6.8 Gy; P < 0.001), larynx (12.5 vs 8.2 Gy; P = < 0.001), oral cavity by (15.8 vs 14.3 Gy; P = 0.002), and contralateral submandibular gland (17.4 vs 17 Gy; P = 0.40). Lesser variation of excess of feasible dose (i.e., more consistent sparing) was seen in the experimental (vs. baseline) cohort for the parotid and contralateral submandibular glands (P < 0.05). (2) Some small differences in target homogeneity and conformity were observed but not felt to be clinically significant. (3) The average post-RT PROs were significantly better in the experimental cohort vs baseline at 6 months post-RT. Dry mouth severity, sticky saliva, meal enjoyment, severity of pain, and EAT10 composite (swallowing) were all significantly improved; P < 0.05).Clinical plans aided by an a priori estimate of maximum achievable dose sparing for OARs was associated with improved (and more consistent) OAR sparing and better clinical outcomes (i.e., better post-RT PROs). Clinical implementation of this tool can yield quantifiable dosimetric and patient benefits.