6th ICHNO Abstract Book

page 66 6 th ICHNO Conference International Conference on innovative approaches in Head and Neck Oncology 16 – 18 March 2017 Barcelona, Spain __________________________________________________________________________________________ PO-137 Multivariate oral rinse models predict Head and Neck squamous cell carcinoma (HNSCC) compared to a historical series of patients (pts) treated with three-dimensional conformal radiotherapy (RT3D). Material and Methods 6th ICHNO

M. Donovan 1 , I. Reis 2 , K. Curtis 3 , F. Khan 1 , E. Franzmann 4 1 Icahn School of Medicine at Mt. Sinai, Pathology, New York City, USA 2 University of Miami, Public Health Sciences and Division of Biostatistics, Miami, USA 3 Vigilant Biosciences, Clinical, Ft. Lauderdale, USA 4 University of Miami, Otolaryngology, Miami, USA Purpose or Objective Background: Head and neck squamous cell carcinoma (HNSCC) is the 6th most common cause of cancer mortality throughout the world affecting some 50,000 people in the US and 600,000 worldwide each year. The ability to detect the disease in a potentially malignant phase and earlier stage could have significant impact on overall outcome. Previous studies have demonstrated that a combined salivary CD44, a tumor-initiating marker, and total protein assay was able to aid in the diagnosis of HNSCC. We sought to better understand the cut-point performance characteristics of these biomarkers in an expanded training cohort. Material and Methods Methods: Oral rinse specimens from 310 patients (107 HNSCC cases; 203 controls) were obtained from biorepositories. A training cohort was generated, divided equally and demographically balanced. Levels of CD44 and total protein (TP), +/- clinical variables were evaluated using the AUC, sensitivity, specificity; support vector machine (SVM) multivariate models were also generated. Results Results: 95% HNSCC patients (cases) were >/=40 years of age, 72% male, 97% white and 65% smokers vs. 23%, 36%, 60% and 44% respectively, for controls. Predicted cut- points for CD44 and TP yielded an AUC of 0.72 for discriminating cancers. A logistic regression model which included continuous levels of CD44 and TP combined with sex, and race produced an of AUC 0.83; while an SVM model which incorporated age, sex, race, smoking history with continuous CD44 and TP yielded a sensitivity of 87% and a specificity of 94%. Conclusion Conclusions: Elevated and combined levels of both CD44 and TP continue to perform well for discriminating HNSCC from control patients. The incorporation of clinical factors including smoking status, sex and race appears to improve overall performance of the assay. Additional validation studies are underway to further confirm these results. PO-138 IMRT of sino-nasal cancer: improved results compared to 3D radiotherapy. T. Frédéric-Moreau 1 , L. Piram 1 , J. Miroir 1 , N. Saroul 2 , C. Millardet 3 , F. Kiakowski 4 , M. Lapeyre 1 , J. Biau 1 1 Centre Jean Perrin, Radiotherapy, Clermont Ferrand, France 2 CHU Gabriel Montpied, Otorhinolaryngology Surgery, Clermont Ferrand, France 3 Centre Jean Perrin, Medical Physics, Clermont Ferrand, France 4 Centre Jean Perrin, Biostatistics, Clermont Ferrand, France Purpose or Objective Prospective evaluation of the results of intensity- modulated radiation therapy (IMRT) in sino-nasal cancer Poster: Salivary gland, skull base, skin and thyroid cancers

Between 2011 and 2015, 42 pts were treated with IMRT and integrated boost (Rapidarc®). Doses in low-risk, intermediate and high volumes were 54, 60 and 66-70 Gy respectively. They were retrospectively compared with a historical series of 30 pts treated with RT3D (50 to 66Gy), from 2007 to 2011. There were 28 ethmoid, 27 maxillary sinuses and 17 nasal cavities. There were 34 squamous cell carcinomas, 27 adenocarcinomas and 4 other histological types. There was no significant difference between the 2 groups (histology, location, stage and surgery). The efficacy and toxicity results were evaluated. Results Median follow-up was 30 months (range, 1.4-112 months). Two-year local control was 81% in IMRT vs 57% in RT3D (p=0,05). In multivariate analysis, the prescribed dose at high risk volume was predictive of local control regardless technique (p=0,05). Two-year overall survival was 92% in IMRT vs 55.5% in RT3D (p = 0.002). In IMRT, ocular acute toxicity rate of grades 1, 2 and 3 were 48%, 28% and 0% respectively. Mucosa, skin and salivary acute toxicity rate of grade ≥ 3 were 14%, 2% and 0% respectively. At one year, ocular toxicity rate of grades 1, 2 and 3 were 36%, IMRT significantly improves local control and overall survival in sino-nasal cancer allowing to bring a higher dose to the target volume. Acute and late toxicities remain low. chemo- Tomotherapy for nasopharyngeal cancer: 2-year treatment outcomes C.T.K. Fong 1 , C.S. Boon 1 , P. Sanghera 1 , A. Hartley 1 1 Queen Elizabeth Hospital, Hall-Edwards Radiotherapy Research Group, Birmingham, United Kingdom Purpose or Objective To report outcomes for nasopharyngeal carcinoma (NPC) patients treated with selective neoadjuvant docetaxel, cisplatin and 5FU (TPF) chemotherapy and IMRT delivered by helical TomoTherapy® (HT) using a hypofractionated accelerated schedule plus concurrent platinum in a UK institution. Material and Methods Since December 2011, patients with NPC receiving chemoradiotherapy were treated using HT. Three dose levels were used: 65Gy/30 fractions to GTV; 60Gy/30 fractions to high-risk local site plus retropharyngeal nodes; 54Gy/30 fractions to uninvolved bilateral nodal levels Ib, II, III, IV, V, VIIb. All patients received concurrent platinum chemotherapy. Patients with T≥3 or N≥1 or M1 also received neoadjuvant TPF. Data collected included locoregional control, survival and late radiation toxicity using CTCAE v4. Results 19 patients met the inclusion criteria (median age 46 years, range 19-68). WHO histology: keratinising squamous = 4; non-keratinising = 14 (undifferentiated = 11, differentiated= 3); basaloid squamous = 1. TNM-7 staging was: I =3; II =3; III =4; IVa =4; IVb =2; IVc =3. Fourteen patients received neoadjuvant TPF (median 3 cycles, range 1-5). All patients completed the prescribed radiation dose with median overall treatment time of 39 days (range 37-45). Concurrent chemotherapy used were cisplatin (n=15) or carboplatin (n=4). At a median follow-up of 31 months (range 13–52) post- chemoHT, none of the 19 patients have developed locoregional relapse. 2 patients with stage IVc disease developed further distant metastases at 3 and 6 months 18% and 0%. Conclusion PO-139 Hypofractionated accelerated