An in silico comparative planning study investigating the predicted benefit of dysphagia-optimised volumetric arc therapy (Do-VMAT), relative to standard VMAT (S-VMAT), to reduce radiotherapy-associated dysphagia (RAD) in oropharyngeal cancers (OPC).

Abstract

e18540 Background: RAD is a significant treatment-related functional morbidity in OPC and is associated with irradiation of the pharyngeal constrictor muscle (PCM). This study investigated whether a novel PCM-sparing Do-VMAT technique reduced the risk of RAD, compared to S-VMAT. Methods: Twenty OPC patients who previously underwent radiotherapy (RT) to bilateral neck were selected for this retrospective study. For each patient, S-VMAT and Do-VMAT plans were created. In both arms, the mucosal and nodal tumour received 65 Gy while the remaining oropharynx and at-risk nodal levels received 54 Gy; delivered in 30 fractions over 6 weeks. For Do-VMAT, a mandatory mean dose constraint of < 50 Gy to the volume of the combined superior (SPCM) and middle PCM (MPCM) together with an optimal dose constraint of < 20 Gy to the volume of inferior PCM (IPCM), lying outside the high-dose clinical target volume, were additionally defined and optimised to minimise PCM dose. Normal tissue complication probabilities for physician-scored RAD at 6 months (RAD6M) were determined as per the model by Christianen et al in which the 2 variables were mean dose to SPCM and supraglottic larynx (SGL). Statistical analysis was performed using paired t tests. Results: All plans had adequate target volume coverage, and dose to critical organs were within acceptable parameters. Dose to the parotid and submandibular glands were similar between the 2 RT techniques. The mean integral dose was 110.8 Gy-litre (standard deviation 17.1) with S-VMAT and 103.4 Gy-litre (15.4) with Do-VMAT, p < 0.001. With Do-VMAT, the average of the mean dose to PCM, SPCM, MPCM, IPCM, and SGL were 9.3 Gy (1.4), 4.7 Gy (1.7), 7.4 Gy (3.6), 22.7 Gy (4.1), and 2.6 Gy (2.3) respectively lower than S-VMAT (p < 0.001 for all). The mean predicted risk for RAD6M was 31.0 % (4.4) with S-VMAT, and 23.9 % (3.8) with Do-VMAT, p < 0.001. The absolute reduction in estimated toxicity risk with Do-VMAT was > 5 %, the minimally clinically important difference, in every patient. Conclusions: Do-VMAT demonstrated clinically relevant reductions in the predicted risk of RAD, compared to S-VMAT, by reducing dose to the PCM in OPC. The clinical benefits of Do-VMAT is being prospectively evaluated in DARS (CRUK/14/014), a randomised phase III study.