ESTRO 38 Abstract book

S366 ESTRO 38

A new model for patient-reported moderate-to-severe xerostomia at 6 months (XER M6 ) after treatment was previously developed in a unique, high quality dataset of 750 patients who received definitive radiotherapy for head and neck cancer (HNC) and externally validated in datasets from two other institutions. The aim of the current research was to validate this NTCP-model in a cohort of HNC patients who received postoperative radiotherapy (PORT) with or without chemotherapy (CT). Material and Methods The population of this prospective cohort study was composed of 218 HNC patients treated with PORT ± CT. Baseline patient- and treatment characteristics, as well as acute and late toxicity and patient-rated outcome measures (i.e. EORTC QLQ-C30 and EORTC QLQ-HN35) were prospectively scored and linked to dose-volume parameters from both parotid and submandibular glands. The salivary glands were delineated according to the international guidelines (Brouwer et al. Radiother Oncol 2015) similar to those used in the development set. Multiple imputation was used to deal with missing data to avoid the bias of a complete case analysis. The original xerostomia model comprises baseline patient-reported xerostomia complaints, the summed square root transformation of the mean dose in both parotid glands (SSQ-D mean ) and D mean of both submandibular glands. A closed testing procedure was used to validate and, if necessary, update the xerostomia model for use in the postoperative setting. Results The prevalence of XER M6 was 27%, as compared to 41% (95%CI: 34.4% - 47.5%) as predicted by the model. A closed testing procedure to validate this model revealed that a model revision was necessary for use in the postoperative setting (Figure 1). Particularly, the association with the submandibular glands was weaker than described by the original model. In the current dataset, in contrast to the development dataset, patients often underwent neck dissections and had zero, one or two remaining submandibular glands post-surgery in 26%, 53% and 21% of the cases, respectively. Therefore, we tested whether a revised model including a variable for the number of post- surgery remaining submandibular glands or interaction terms including this factor, would better fit to the current dataset. However, this was not the case. Subsequently, we reduced our initial model, by excluding the mean dose in the combined submandibular glands, which did not change significantly the goodness of fit. The closed testing procedure then revealed that an update of the model intercept, to compensate for the lower response rate, was sufficient for this reduced model (ROC-AUC = 0.66, Nagelkerke R 2 = 0.11) to be used in the postoperative setting (Figure 2).

Conclusion A new model for XER M6 was externally validated in the postoperative setting and established after excluding the variable describing the D mean of both submandibular glands and an update of the model intercept. PO-0713 PDRN-based cream in the prevention and treatment of radiodermatitis in Head and nech cancer: our experience. F. Pastore 1 , A. Rese 1 , G. Panelli 1 , A. Pepe 1 , D. Toledo 1 , V. Iorio 1 1 Emicenter, Radiation Oncology, Mondragone, Italy Purpose or Objective This study was designed to assess the efficacy and tolerability of a PDRN (Polideoxyribonucleotides)- based cream formulation in the prevention and treatment of radiodermatitis in patient with head and neck cancer treated with CTRT (concomitant chemo-radiotherapy). PDRN is a well-known registered drug with tissue repairing, anti-ischemic and anti-inflammatory properties and we already tested it in a previous study on breast cancer patients. Material and Methods Patients with head and neck cancer has been randomized to use a cream, standard of care in our institution or a PDRN-based cream, throughout the duration of radiotherapy and for two weeks afterwards. Patients were monitored before therapy, weekly during therapy, and for 2 weeks after radiotherapy was completed, using RTOG skin toxicity grading scale. Patients were treated with Helicoidal Tomotherapy, choosing appropriate RT dose and fractionation according to disease presentation and with concomitant chemotherapy, using both conventional than hypofractionation schedules. We assessed on these patients several clinical variables as tumor histology, skin phototype, concomitant chemotherapy and medical conditions. Results Between december 2016 and march 2018, 102 consecutive patients, were enrolled in this study. Median age of the patients was 62.5 (range 46-84). 50/60 (83%) patients in the PDRN group vs. 57/62 (91%) patients in the control group had some grade of skin toxicity at the end of the treatment. At 4th week, 19/60 patients (31%) in the PDRN group vs. 37/62 (60%) patients in the control group had G2 grade of skin toxicity (p=0.0095). After 15 days from the end of the treatment, 5/60 patients (8%) in the PDRN group vs. 11/62 patients (18%) in the control group had some grade of skin toxicity (p=0.049). After 60 days from