ESTRO 38 Abstract book

S104 ESTRO 38

weekly cisplatin). A total of 484 patients were managed with ART due to: age >70 years (n=216), comorbidity/poor PS (n=102), or declined chemotherapy (n=166). Patients and treatment characteristics are summarized in Table 1 . The pattern of failure among the whole cohort is shown in Figure 1a . With a median follow up of 4.3 years, 5-yr LC following CRT, A-Mod (n=128), A-Hypo (n=141), and A- Hyper (n=215) were 81% (95%CI:74-85%), 70% (95%CI:60- 78%), 68% (95%CI:59-75%) and 65% (95% CI:58-71%) respectively. Outcomes following propensity score matching are shown in Figure 2 ; the 5-yr LC following CRT was 80% (95%CI:72%-86%) vs 71% (95%CI:61%-78%) compared to A-mod (p=0.08), 82% (95%CI:72%-88%) vs 71% (95%CI:61%-79%) compared to A-Hypo (p=0.07), and 81% (95%CI:74%-86%) vs 66% (95%CI:58%-72%) compared to A- Hyper (p<0.001). There was no significant difference in 5- yr LC between different ART schedules (p>0.05, Figure 2b ). Among patients who had locoregional recurrence (LRR) without synchronous distant metastases (n=202), 113 (56%) had salvage surgery (5-yr OS: 40% [95%CI:32%- 51%]), while 89 were not surgical candidates (5-yr OS: 5% [95%CI:2-15%]). Late toxicities are summarized in Figure 1B .

Conclusion Approximately

two III-IVB laryngeal/hypopharyngeal SCC patients who declined or were unfit for concurrent cisplatin achieved LC following ART. There was no significant difference in LC among the different ART regimens. PV-0201 Development and validation of prediction models for salivary dysfunction in HN cancer patients L. Van den Bosch 1 , A. Van der Schaaf 1 , F.J.P. Hoebers 2 , H.P. Van der Laan 1 , E. Schuit 3 , E. Bakker 1 , O.B. Wijers 4 , A.M. Van der Wel 4 , R.J.H.M. Steenbakkers 1 , J.A. Langendijk 1 1 University of Gronigen - University Medical Center Groningen, Radiation oncology, Groningen, The Netherlands ; 2 Maastro Clinics, Radiation Oncology, Maastricht, The Netherlands; 3 Utrecht University - University Medical Center Utrecht, Julius Center for Health Sciences and Primary Care, Utrecht, The Netherlands; 4 Radiotherapeutic Institute Friesland, Radiation Oncology, Leeuwarden, The Netherlands Purpose or Objective NTCP-models for salivary dysfunction and other toxicities in head and neck cancer (HNC) patients can be used to optimize treatment plans and select the best RT technique. However, currently used NTCP-models have several shortcomings, such as retrospective assessments, being developed on limited data for only a single time point or grade without properly addressing missing data, non-linearity, and multicollinearity, no external validation and a limited number of organs at risk (OARs) investigated. Our aim was to address these shortcomings and to develop and externally validate an NTCP-profile based on a comprehensive set of NTCP-models for patient-rated (PR) and physician-rated (PhR) toxicities. Here we present the results for salivary dysfunction. thirds of stage