ESTRO 37 Abstract book

S138

ESTRO 37

characteristics including treatment arm, timing of neck dissection, chemotherapy regime, age, sex, T stage, N stage, performance status, smoking status, and alcohol status were not significantly different between the two treatment technique cohorts and between the three fractionation cohorts. However, there were significantly fewer 3D-CRT patients with oropharyngeal cancer (79% vs 88% p=0.014) and there were significantly more patients in the 55Gy/20# cohort when compared with the 6 and 7 week cohorts (48% vs 16% vs 30%). 2-year local control was examined by each of the technique and fractionation cohorts, and by p16 status which was available in 424 (80%) of the 532 patients with available data on technique and 417 (79%) of the 525 patients with available data on fractionation. Results 2-year primary local control in p16+ve patients was 95% (95% confidence intervals: 91-99%) vs 93% (89-97%) for 3D-CRT (n=118) vs IMRT (n=206). The corresponding figures for p16-ve patients were 78% (66-91%) vs 76% (63- 88%), (n=45 vs n=55). In p16+ve patients receiving 4 week (n=25) vs 6 week (n=193) vs 7 week (n=98) fractionations, the 2-year primary local control was 87% (74-100%) vs 95% (92-98%) vs 92% (86-97%). The corresponding figures for p16-ve patients were 85% (69-100%) vs 80% (66-93%) vs 74% (59-88%) (n=23 vs n=37 vs n=41). No significant differences were observed in any of these comparisons. Conclusion The 2-year primary local control outcomes seen in this study support the continued examination of accelerated hypofractionation in the ongoing CompARE and HYPNO prospective randomised trials. This data may also be useful in further modifying radiobiological models for the addition of synchronous chemotherapy to radiotherapy. OC-0275 Comprehensive toxicity profiles in HN cancer patients treated with radiotherapy: a benchmark study L. Van den Bosch 1 , A. Van der Schaaf 1 , H.P. Bijl 1 , J.G.M. Vemer-van den Hoek 1 , O. Chouvalova 1 , M.C.A. Kramer 1 , R.J.H.M. Steenbakkers 1 , J.A. Langendijk 1 1 University Medical Center Groningen, Department of Radiation Oncology, Groningen, The Netherlands Purpose or Objective The purpose of this prospective cohort study was to determine comprehensive toxicity risk profiles (CITOR) in head and neck cancer (HNC) patients treated with primary radiotherapy with or without systemic treatment. Material and Methods This population-based prospective cohort study was composed of 750 HNC patients treated between 2007 and June 2016. All patients were enrolled into a prospective data registry in which physician-rated toxicities (CTCAE v4.0) were prospectively collected weekly during treatment and at regular intervals up to 5 years thereafter. Toxicity prevalence for all patients as well as for 4 subgroups was analyzed. Subgroups were defined based on stage (early: stage I-II vs. advanced: stage III-IV) and tumor location (proximal: nasopharynx, oropharynx, oral cavity vs. distal: hypopharynx and larynx). Toxicity patterns of individual patients were identified as persistent low (grade 0-1), decreasing, persistent high (≥ grade 2) or progressive, by comparing the toxicity grade near the end of treatment with the last recorded toxicity grade during follow up. The recovery percentage was calculated as the number of patients with a decreasing pattern divided by the number of patients with a decreasing, persistent high or progressive pattern. Results The compliance rate was high, i.e. 96% for acute (week 1 to 12) and 88% for late (month 6 to 60) toxicity scoring. For all endpoints scored during and after treatment, similar toxicity patterns were observed: there was a marked increase during treatment, reaching a maximum

at the end of treatment, followed by a gradual recovery starting in de first 6 months after treatment with almost no further recovery beyond 24 months. Weight loss showed a similar pattern, however the maximum loss was reached at 6 months after treatment. For late toxicities, including skin fibrosis, hypothyroidism and late mucosal toxicity, a gradual increase in prevalence was seen over time (fig. 1). In the four different subgroups, more or less similar toxicity patterns were observed, however with different severities, as shown for dysphagia in figure 2. The percentage of patients recovering, derived from the individual toxicity patterns, differed between toxicity endpoints. For endpoints that showed a general trend of recovery, the highest recovery was seen for sore throat (99%) and oral pain (95%), while the lowest recovery was seen for xerostomia (73%), aspiration (54%) and weight loss (39%).

Conclusion This is the largest prospective population-based cohort study providing comprehensive toxicity risk (CITOR) profiles in HNC patients treated with definitive radiotherapy with or without systemic treatment. These CITOR profiles can be produced for numerous patient-, tumor- and treatment-related subgroups, serving as a toxicity benchmark of the current treatment, allowing comparison with new treatment approaches. When combined with NTCP-models, predicted CITOR profiles enable a comprehensive assessment of the potential benefit of new radiation technologies.

OC-0276 Combining different genomic signatures to improve the prognostic power for LRC after PORT-C in HNSCC