Dosimetric Predictors of Clinician-rated and Patient-reported Radiation Dermatitis in Head and Neck Cancer: Exploratory Analysis of a Prospective Pilot Study

Purpose/Objective(s)There are no well-established dosimetric predictors of clinician-rated acute radiation dermatitis (ARD) in head and neck (HN) radiotherapy (RT). Data regarding the dosimetric associations with patient-reported outcome (PRO) measures of RD are even more scarce.Materials/MethodsIn this secondary analysis of a pilot study (NCT04173247) in which 24 patients were treated with at least 60 Gy, we examined dosimetric predictors of ARD. Dermatitis assessments were performed at baseline, weekly during RT, and at 1-month post-RT using clinician-rated (CTCAE v5.0) and patient-reported outcome measures (PRO-CTCAE v1.0 and Dermatologic Life Quality Index [DLQI]). The skin was defined as a ring-like structure comprising the inner 5 mm of the external contour generated by the treatment planning system, excluding bolus (if applicable). The volume of skin (cm3) receiving a specified dose (V[x] Gy) was calculated in 5-Gy increments between V30-V70 Gy. For each parameter, the volume of skin was compared between patients who did and did not develop each RD outcome: CTCAE G2+ RD, PRO-CTCAE G2+ radiation skin reaction (RSR), or clinically significant 4-point increase from baseline DLQI score (DLQI4+).ResultsMedian RT dose was 68 Gy (range 60-70) in 28-35 fractions with the target including the bilateral neck (n=19), unilateral neck (n=2), or primary site alone (n=3). Treatment was definitive in 16 patients and postoperative in 8; concurrent chemotherapy was delivered in 18 patients. CTCAE G2+ ARD was observed in 16/24 patients, PRO-CTCAE G2+ RSR in 18/24, and DLQI4+ in 13/24. Skin V40-V55 were associated with DLQI4+, and multiple skin dose parameters were identified to be associated with ARD at varying significance levels (Table 1). Skin dose was significantly associated with stage (I-II v. III-IV), nodal classification, extent of neck target, and concurrent chemotherapy. Skin dose-volume metrics were not associated with treatment group, intent (definitive v postoperative), age 65+, BMI category, diabetes, tumor classification, or primary site.ConclusionMultiple skin dose-volume parameters were associated with clinician-rated and PRO measures of ARD during HN RT. When feasible, the skin should be included as an organ at risk and avoided to limit the risk of moderate to severe ARD. These hypothesis-generating findings warrant further validation and may be helpful in identifying patients at high risk of ARD for inclusion in future ARD studies.