A Prospective Study of Daily Adaptive Radiotherapy to Better Organ-at-Risk Doses in Head and Neck Cancer

IMRT with daily cone-beam computed tomography (CBCT) set up is the current recommended radiation technique for head and neck squamous cell carcinoma. The conformality of IMRT has made it possible to reduce dose to normal tissues, but it also predisposes the treatments to drastic delivered dose deviations due to changes in the patient (e.g. weight loss, muscle atrophy, or normal tissue edema) or in the tumor (e.g. treatment response or treatment related edema). These changes can create poor coverage and/or cold spots in the tumor and increased coverage and/or hot spots in the normal organs at risk. Image guidance with CBCTs can help reduce set up error between treatments, but image guidance can't adjust for these patient or tumor changes. Adaptive radiotherapy (ART) is the process of replanning the patient's radiation plan by adjusting the target structures, organs at risk structures, and optimal radiation delivery during the radiation therapy course. ART allows treating physicians to account for all the patient or tumor changes that can't be currently resolved with image guidance. Previously, ART was a prohibitively resource and time-consuming process that could only be performed once or twice during a conventional 6.5 to 7-week head and neck radiation treatment course in select patients with substantial structural changes. However, with the recent advent of adaptive software, ART can now be performed on a daily basis. The data on the possible advantages of ART in head and neck cancer is still limited to small, mostly pre-clinical studies. In a feasibility study, five patients with locally advanced HPV-positive oropharynx cancer undergoing definitive chemoradiation underwent an MRI every two weeks during their treatment. The results showed an average decrease in the primary tumor GTV volume of 44%, 90%, and 100% and an average decrease in the nodal GTV volume of 25%, 60%, and 80% by at weeks 2, 4, and 6 weeks, respectively. Another similar study on eight patients saw an average 70% GTV volume reduction by week 6. In addition to tumor coverage, parotid glands are of particular interest in ART because of their radiosensitivity and association with xerostomia and decreased quality of life. Toxicity of salivary glands are related to dose as a continuous variable, so even modest decreases in dose can have a clinical impact. The average volume of the parotids has been reported to decrease as much as 14.7%, 37%, and 48% by the end of weeks 2, 4, and 7. The parotid glands also appear to shift superiorly and medially during treatment. Feasibility ART studies have shown that the dose to the parotid may decrease by as much as 5.5 Gy with ART. Therefore ART holds the promise of preserving the expected dose distribution to salivary OARs, improving long-term patient outcomes. The ability to ensure coverage of the target structures and proper avoidance of organs at risk on a daily basis also greatly improves the confidence of the daily set-up and targeting. Current recommended radial PTV margins with daily CBCT are 3-5 mm, with even larger margins recommended for patients at risk of larger inter-fraction variability. These PTV margins are built to ensure the disease is not missed by the treatment due to daily setup error. However, these margins also significantly increase the irradiated volume, since the volume increases exponentially with each additional millimeter. Therefore, decreasing margins, even by a few millimeters, can results in large changes in treatment volume and improvement in toxicity. In a retrospective Danish analysis of patients treated with 5 mm versus 3 mm PTV margins, Navran et al. noted that acute grade 3 toxicity (65% vs 53.8%, p = 0.008) and late grade 3 dysphagia (20.4% vs 11.1%, p = 0.012) were significantly lower in the smaller margin group. There were no differences in disease outcomes between the two groups. In a retrospective study analyzing the benefit of changing the margin on gross disease from 8 mm to 5 mm, just a 3 mm difference, Samuels et al. found that plans with the smaller PTV margin had a clinically meaningful improvement in the normal tissue complication probability for the ipsilateral parotid gland and contralateral submandibular gland when compared to conventional PTV margin plans. In fact, 40% of patients would have theoretically experienced an improvement in the function of both of these glands, just by modestly shrinking the PTV.