S6 ESTRO 33, 2014 Materials and Methods: A retrospective review was conducted on a prospectively assembled cohort of newly diagnosed non-metastatic N3 HNC treated with RT/CRT from 2000-2011. Tumor HPV status was ascertained by p16 staining on all available blocks for oropharyngeal cancer (OPC) and cervical nodal metastasis from 'unknown' primary (CUP). OPC and CUP without tissue blocks for p16 were excluded. Cancers of larynx and hypopharynx were analyzed as HPV(-). Overall survival (OS), cause-specific survival (CSS), local (LC), regional (RC), distant control (DC), and grade 3/4 late toxicity (LT) were compared between HPV(+) vs HPV(-) HNC. Multivariate analysis (MVA) identified survival predictors. Results: A total of 56 HPV(+) (55 OPC, 1 CUP) and 54 HPV(-) (18 OPC, 1 CUP, 13 larynx, 22 hypopharynx) N3 HNC were included. Compared to the HPV(-), HPV(+) patients were younger (median age: 57 vs 62, p<0.01), comprised fewer >10 pack-years smokers (35% vs 51%, p<0.01) but had similar proportion of T4 (18% vs 28%, p=0.11) and bilateral neck disease (55% vs 43%, p=0.18). N3 nodal features were similar between the HPV(+) vs HPV(-) cohorts: fixed (72% vs 80%, p=0.32), dermal involvement (14% vs 19%, p=0.55), and unresectable (42% vs 52%, p=0.34). 40 (71%) HPV(+) vs 30 (56%) HPV(-) received CRT (p=0.11). Neck dissections (ND) were performed in 22 HPV(+) (3 pre- and 19 post-RT) vs 11 HPV(-) (all post-RT) cases. 14/56 (25%) HPV(+) vs 9/54 (17%) HPV(-) (p=0.28) achieved radiological complete response (CR) at 8-12 weeks after RT/CRT (<=1.0 cm residual on CT/MR without adverse features), of which 0/14 HPV(+) and 2/9 HPV(-) had subsequent regional failure (RF). Median follow-up was 4.3 years. A total of 3 vs 6 local, 10 vs 17 regional and 18 vs 24 distant failures were identified in the HPV(+) vs HPV(-) cohorts respectively. HPV(+) HNC cohort has higher 3-year OS (62% vs 25%), CSS (67% vs 36%) (both p<0.01), and marginally better RC (84% vs 68%, p=0.09) and DC (71% vs 55%, p=0.14). The LC (95% vs 91%, p=0.27) and LT (31% vs 20%, p=0.61) rates were similar. MVA confirmed that HPV(+) status was predictive for better OS [HR 0.50 (95% CI: 0.30-0.84), p<0.01] and CSS [HR 0.47 (95% CI: 0.26-0.84), p=0.01] after adjusting for age (OS: p=0.51; CSS: p=0.31), smoking pack-years (OS: p=0.17; CSS: p=0.90), T category (OS: p=0.82; CSS: p=0.47), RT technique (IMRT vs non-IMRT) (OS: p=0.31; CSS: p=0.33), and use of chemotherapy (OS: p=0.37; CSS: p=0.81). Conclusions: This study indicates that N3 HPV(+) HNC patients have better survival compared to HPV(-) and most HPV(+) N3s are curable with high RC. ND seems unnecessary for HPV(+) with CR as no RF was observed. Distant failure is the main form of failure for both HPV(+) and HPV(-) cohorts. Future studies to optimize treatment in this disease are warranted. PO-0663 Delayed response assessment FDG PET-CT following (chemo)radiotherapy for head and neck squamous cell carcinoma F. Slevin 1 , S. Ramasamy 1 , M. Sen 1 , M. Subesinghe 1 , A. Scarsbrook 1 , R. Prestwich 1 1 St James Institute of Oncology, Clinical Oncology, Leeds, United Kingdom Purpose/Objective: We performed a baseline and delayed response assessment Fluorine-18 fluorodeoxyglucose (FDG) PET-CT according to a prospective protocol approximately 4 months following completion of (chemo)radiation for locally advanced head and neck squamous cell carcinoma (HNSCC). The aim of this analysis is to analyse the diagnostic accuracy of delayed response assessment FDG PET-CT. Materials and Methods: Eligible patients for analysis had head and neck squamous cell carcinoma of the oropharynx, oral cavity, hypopharynx or larynx. Patients with nasopharyngeal cancer was excluded. 105 consecutive patients who underwent a baseline and response assessment FDG PET-CT scan for HNSCC following (chemo)radiation between August 2008 and April 2013 were identified retrospectively. Data were obtained from institutional electronic case note records. Clinico-pathological findings and serial clinical follow up provided a reference standard. Results: 73 out of 105 patients were male. Median age was 57 (range 25- 75). 86 out of 105 (82%) patients had stage IV disease. None had metastatic disease on baseline FDG PET-CT. 87 out of 105 (83%) patients had received concurrent chemoradiation and 14 of 105 (13%) had radiation therapy alone. Radiation was delivered at a dose of 70 Gray in 35 fractions in 103 out of 105 patients. Median follow up was 25 months (Range 7-73 months). Response assessment FDG PET-CT was performed at a median 17.4 weeks following completion of (chemo)radiation (inter- quartile range 16.3 -18.4 weeks). 79 out of 105 (75%) response assessment scans demonstrated a complete metabolic response. 19 out of 92 (20%) for assessable primary tumours were equivocal or positive. 16 out of 93 (17%) for patients with pre-treatment nodal disease were equivocal or positive. Based upon clinical-pathological outcome and follow up, the sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) for primary disease and nodal disease were 90%, 88%, 47%, 99% and 92%, 94%, 69% and 99% respectively (Table 1). Distant metastases were detected on response assessment FDG PET- CT in 9/105 patients (9%). Conclusions: The diagnostic accuracy of delayed response assessment with FDG PET-CT performed at approximately 16 weeks post- (chemo)radiotherapy is good. The very high NPV of a complete metabolic response on delayed response imaging can be used to guide management decisions. The PPV is limited for loco-regional residual disease, and requires clinico-pathological correlation. FDG PET-CT is a powerful screening tool for the detection of interim metastatic disease. PO-0664 DNA detection and p16 immunohistochemistry as alternative methods for the diagnosis of HPV in oropharyngeal cancer R. Autorino 1 , F. Miccichè 1 , F. Bussu 2 , N. Dinapoli 1 , G.C. Mattiucci 1 , J. Galli 2 , M. Rigante 2 , G. Almadori 2 , G. Paludetti 2 , V. Valentini 1 1 Polyclinic University A. Gemelli Catholic University, Institute of Radiotherapy, Rome, Italy 2 Polyclinic University A. Gemelli Catholic University, Institute of Otorhinolaryngology, Rome, Italy Purpose/Objective: HPV16 infection has been proven to be associated with oropharyngeal squamous cell carcinomas and is probably the main reason of the reported increase in the incidence in Western countries. With the aim to evaluate the prevalence of HPV infection in SCCs of the oropharynx and the reliability of different diagnostic tools, 50 consecutive primary oropharyngeal SCCs were enrolled with the collection of FFPE tumor samples. Materials and Methods: All the patients were primarily treated by radio+ chemotherapy. HPV DNA was detected in fresh samples and in FFPE slides by Digene Hybrid Capture 2 (HC2). p16 expression was evaluated by immunohistochemistry. Statistical analysis, also in relation with oncological outcome, was performed by JMP software, version 7.0.1 Results: When considering the whole series, the frequency of HR HPV DNA detection was 32%. The agreement rate between p16 IHC and HPV DNA detection in FFPE samples was fair but not excellent (kappa= 0.618). HPV DNA detection displayed a highly significant impact on DSS at Wilcoxon test (p=0.001). Also p16 IHC retained a prognostic value but with a lower statistical significance (p=0.0475) and a smaller difference between the 2 populations (5yr DSS: 85% in HPV DNA+ vs. 33% in HPV DNA-; 72% in p16+ vs. 38% in p16-). HPV DNA detection, but not p16 IHC, displayed an impact also on locoregional control (p=0.0461). In a multivariate analysis only HPV DNA detection showed a statistically significant correlation with DSS (see table).