7th ICHNO Abstract book

page 10 7 th ICHNO Conference International Conference on innovative approaches in Head and Neck Oncology 14 – 16 March 2019 Barcelona, Spain __________________________________________________________________________________________ 7th ICHNO

recurrences. Reirradiation and surgery are the main “curative” options for local recurrences. However the risk of severe toxicity following reirradiation is high, and a nomogram predicting the risk of relapse and grade 5 toxicity after reirradiation was recently published (Li et al, J Clin Oncol 2018). The goal of this analysis was to review the outcomes and toxicity following NPC local reirradiation performed at Gustave Roussy Cancer Campus since 2005, and externally validate the published nomogram. Material and Methods The charts of all patients having received radiotherapy for a non-metastatic NPC between 2005 and 2018 were reviewed. Patient and tumor characteristics, as well as outcomes, were retrospectively reviewed. A leave one out cross validated receiver operator characteristics curve was performed to evaluate the discriminatory performance of the nomogram. Results Out of 346 NPC patients treated with radiotherapy over this time period, 35 had received a second course of radiotherapy and were included in the analysis. They were mostly male (75%), median age was 54 years, and 20% had retained a chronic grade 3 or higher toxicity as a result of the first course of radiotherapy. Local recurrences were classified as rT3T4 in 50% of the cases. The median interval between the first and second courses of radiotherapy was 78 months. 43% of the patients received concurrent chemotherapy and 37% induction chemotherapy. The median dose of reirradiation was 62 Gy. Two third of the patients had received IMRT or proton therapy as reirradiation technique. After a median follow-up of 54 months after the end of reirradiation, 7 patients had suffered a grade 5 toxicity, including 6 with a vascular blowout. These treatment- related deaths occurred early after reirradiation, with a median time to occurrence shorter than 6 months. 85% of patients presented at some point in follow-up a grade 3 or higher toxicity altering their quality of life, including cranial nerve palsies, hearing impairment or temporal brain necrosis. Overall survival rates were 56% at 2 years and 40% at 5 years. The published nomogram was not predictive of overall survival or grade 5 toxicity in our patient cohort (areas under the ROC curves of 0.58 and 0.55, respectively). Conclusion Reirradiation is an effective treatment for local recurrence of NPC,but is associated with a high rate of severe toxicity, including 20% of treatment-related deaths. Nomograms predicting survival and toxicity would be helpful in selecting patients for reirradiation, although they should be externally validated prior to clinical use in non-endemic areas. OC-011 New insights from the De-ESCALate HPV trial H. Mehanna 1 , M. Robinson 2 , A. Hartley 3 , A. Kong 1 , B. Foran 4 , T. Fulton-Lieuw 5 , M. Dalby 5 , P. Mistry 5 , M. Sen 6 , L. O'Toole 7 , J. Dunn 5 1 University of Birmingham, Institute of Head and Neck Studies and Education, Birmingham, United Kingdom; 2 University of Newcastle, Pathology, Newcastle, United Kingdom; 3 University Hospitals Birmingham, Oncology, Birmingham, United Kingdom; 4 Weston Park Hospital, Oncology, Sheffield, United Kingdom; 5 University of Warwick, Clinical Trials Unit, Late breaking news: Late breaking news

Coventry, United Kingdom; 6 St James's Institute of Oncology, Oncology, Leeds, United Kingdom; 7 Castle Hill Hospital, Queen's Centre for Oncology, Hull, United Kingdom Purpose or Objective To elucidate the risk factors associated with worse survival from cetuximab, compared to cisplatin. In De-ESCALate HPV, an international, multi-centre, randomised, controlled trial, patients with low-risk HPV+OPSCC were randomised to receive radiotherapy (70G in 35F) and either cisplatin (3 doses of 100 mg/m 2 ) or cetuximab (400 mg/m 2 loading dose followed by weekly 250 mg/m 2 ). There were no differences between the cisplatin and cetuximab arms in the reported severe or all grade toxicity. Importantly, however, there was a significant difference in the 2-year overall survival and in the time to recurrence between cisplatin and cetuximab. Material and Methods We undertook new proportional hazard modelling analyses of the survival data to examine factors associated with poor survival from cetuximab. Results Overall, there was a statistically significant difference in the 2-year overall survival between cisplatin and cetuximab (97.5% vs 89.4%, log-rank p=0.001; hazards ratio=4.99, 95% CI 1.70-14.7). All patient factors (age, T/N stage, smoking, performance status, HPV-DNA-ISH) appear to be associated with worse survival with cetuximab compared to cisplatin (Figure 1). In particular, patients with TNM8 stage III (T4 or N3) disease showed larger 2-year OS detriment when treated with cetuximab (67.1% (42.5-83.1%), compared to cisplatin (93.3% (95%CI 75.9-98.3%, Log rank p=0.03), HR=4.83 (95%CI 1.00-23.31, Fig 2). When considering the cetuximab-treated group only, all patient factors (age, T/N stage, performance status, HPV- ISH), appeared to be associated with progressively worse survival, except smoking. Smokers appeared to have better 2-year OS (92.0% (82.9% to 96.3%)) than non- smokers (87.0% (77.1% to 92.8%)) when treated with cetuximab. This may be due to smoking resulting in more EGFR upregulation. Conclusion There was significant detriment from the use of cetuximab instead of cisplatin in terms of tumour control across all patient characteristics and groups, and especially TNM8 Stage III patients. Patients treated with cetuximab appeared to have better survival if they smoked than non- smokers. Cisplatin and radiotherapy remain the standard of care in this setting.