1066P - Hyperfractionated twice daily re-irradiation (bid re-RT) and chemotherapy (CT) for locoregionally recurrent head and neck squamous cell carcinoma (...

Date 10 September 2017
Event ESMO 2017 Congress
Session Poster display session
Topics Cytotoxic agents
Cancers in Adolescents and Young Adults (AYA)
Head and Neck Cancers
Surgical oncology
Biological therapy
Radiation oncology
Presenter Daphne Day
Citation Annals of Oncology (2017) 28 (suppl_5): v372-v394. 10.1093/annonc/mdx374
Authors D. Day1, G. Locke2, K. Chan3, S. Cheng3, L. Wang4, B. O'Sullivan2, A.R. Hansen1
  • 1Department Medical Oncology And Hematology, Princess Margaret Cancer Centre, M5G 2M9 - Toronto/CA
  • 2Department Of Radiation Oncology, Princess Margaret Cancer Centre, University of Toronto, M5G2M9 - Toronto/CA
  • 3Medical Oncology, Sunnybrook Odette Cancer Centre, M4N 3M5 - Toronto/CA
  • 4Biostatistics Department, Princess Margaret Cancer Centre, M5G 2M9 - Toronto/CA



Re-RT +/- CT is a salvage option for patients (pts) with LR HNSCC in a previously radiated field, although efficacy and toxicity, and the optimal treatment regimen remains undefined due to a lack of randomized trials. Hyperfractionated bid re-RT, by reducing the dose per fraction may improve radiation (RT) therapeutic ratio and is increasingly used in LR HNSCC. The aim of this SR is to assess the treatment outcomes of bid re-RT + CT in LR HNSCC.


We conducted a SR of MEDLINE, EMBASE and the Cochrane library up to Nov 2016 for clinical trials of bid re-RT + CT in pts with LR HNSCC. Paired reviewers selected studies for inclusion and extracted data. Individual patient level overall survival (OS) data were extracted where possible from published Kaplan-Meier (KM) curves to construct an aggregate KM curve.


We identified 10 clinical trials (all were phase 1 or 2) with 404 pts. Median (of reported medians) prior RT dose was 64Gy, and median time from prior RT was 30.9m. Seventy-three and 156 pts respectively had CT and surgery as part of 1st line treatment. Median re-RT dose was 60Gy administered as continuous or split courses. The re-RT fields consisted of gross tumor volume plus a minimum margin (range 1-2cm). All CT regimens were combinations either with cisplatin (n = 6) or 5-FU (n = 4), given concurrently with (n = 9) or prior to (n = 1) re-RT. Twenty-eight (7%) pts had debulking surgery prior to re-RT. In pts who were analyzable for toxicities, acute events (>/=grade 3) were reported in 252 of 377 (67%) pts and late events (>90d post re-RT) in 87 of 333 (26%) pts. Treatment-related deaths occurred in 26 (6%) pts, mostly due to infection or vascular events. Of the 5 trials with extractable KM curves, estimated median OS was 10.2m (95% CI 8.7-12.6m); 1- and 3-y OS rates were 46.8% and 11.2% respectively. No differences were observed in median OS and toxicity rates based on CT type (Wilcoxon test).


This is the 1st aggregate analysis of bid re-RT and CT in LR HNSCC. Long-term OS was observed in a subset of pts, however treatment-related morbidity was apparent. The optimal re-RT and CT regimen is still undefined and further study is required.

Clinical trial identification

Not applicable

Legal entity responsible for the study





All authors have declared no conflicts of interest.