POSterS - $79 total radiation dose to the primary tumor site and neck lymph nodes was between 63,0 and 74,0 Gy fractionated at 2,0 Gylday, S day a week for up to 7 weeks. After loco-regional treatment, patients with high risk of distant metastasis received chemotherapy in adjuvant regime (Carboplatin plus 5-Fluorouracil, 3 - 4 cycles). Patients char- acteristics as follows: 74% male; mediana age 54 years; Karnofsky In- dex ~60%in all pts. Nodal status: N / - 29,1%; N2 - 41,1%; N3 - 16,4%. Histological classification and differentiated tumor: squamous cell - 85%; G / - 16,4%; G2 - 30,9%; G3 - 52,7%. Results: mediana follow-up was 35 months. Response rates for the 55 pts were: complete response - 21,8%; partial response - 56,4%; stable disease - 5,5% and 16,4% pts who progressed. The overall re- sponse rate (ORR) was 78,2%. 3-years recurrence-free survival and overall survival were 60,6% and 58,6% respectively. There was no- body with distant metastasis in our experience. Conclusions: IC with the Carboplatin plus 5-Fluorouracil followed by RT and adjuvant CT is feasible, well tolerated and active in locally ad- vanced carcinomas oropharynx. There results suggest, that the CRT regime (induction and adjuvant) is effective for locally advanced carcinomas oropharynx with high risk distant metastasis. This thera- peutically method can be considered as successful method for dis- tant metastasis control. 215 poster THE IMPORTANCE OF THE TREATMENTCENTER WITH OTHER PRO- GNOSTIC FACTORS IN 247 PATIENTS WITH LARYNX CANCERTREA- TED WITH POST-OPERATIVE RADIOTHERAPY WITH OR WITHOUT CHEMOTHERAPY F. Akman 1,O. Uruk Ataman ~,N. Dag ~, C. Ecevit2,A. Ikiz2, I. Arslan ~, S. Sar~oglu 3,E. Ada4,M. Sens,M. Kinay ~ IDoKuz EYLUL UNIVERSITY, Radiation Ontology, Izmir, Turkey, 2DoKuzEYLUL UNIVERSITY, ENT,Izmir, Turkey, 3DoKuzEYLUL UNIVERSITY, Pathology, Izmir, Turkey, 4DoKuzEYLUL UNIVERSITY, Radiology, Izmir, Turkey, 5COOKRIDGE HOSPITAL, Clinical Oncology, Leeds, England Purpose/Objective: The aim of this study was to identify the clinical and pathological factors that are associated with different types of failures (Ioco-regional and distant) in a wide group of patients with larynx cancer who underwent combined surgery plus postoperative irradiation with the decisions ofa multidisciplinary group. Materials/Methods: The study population consisted of the review of 247 patients retrospectively with locally advanced larynx can- cer who were treated between July 1991 and December 2004 by Dokuz Eyl~il Head and NeckTumour Group (DEHNTG).A multivariate analysis was performed, using the Cox proportional hazards model (forward LR), to determine the risk factors associated with Ioco-re- gional failure (LRF) and/or distant metastasis (DM). Two different risk groups were identified using clinical or pathological characteristics from the multivariate analysis for each failure type as follows: DM: Clinical risk factors: Clinical N positive disease and/or presence of clinical subglottic extension; Pathological risk factors: Number of pathological lymph nodes being 3 or more. LRF:Clinical risk factors: Clinical stage 4 disease and/or clinical presence of subglottic exten- sion; Pathological risk factors: Presence of positive surgical margins and/or positive pathologic lymph node and/or place of operation other than Dokuz EylUI University Hospital Results: The median follow-up was 47 (3-181) months. In total 43 (17.4%) patients were died related to larynx cancer, 24 patients (9.7%) with only LRE, 15 (6%) patients with only DM and 4 (1.6%) patients with both LRF and DM. The 5 and lO-years overall surviv- als for the whole patient group were 64% and 41%, Ioco-regional disease free survival was 88% and 75% and distant disease free sur- vivals were 90% and 80% respectively. Log-rank analysis was most significant for patients who did not have any clinical risk factors (94%) versus one (72%) or more (61%) risk factors for Ioco-regional recurrence free survival at 10 years (p=O.OOO7}.Thepresence of more than 3 lymph nodes pathologically (58%) was highly significantly as- sociated (p=O.O06) with distant metastasis free survival at I0 years compared to patients with less than 3 lymph nodes positive (84%). The presence of any of the pathological risk factors associated with LRF was also an important factor (p=O.04) between patients with no risk factors. Conclusions: In the presence of stage 4 disease and clinically detect- ed subglottic invasion the loco-regional treatment should be very intense with the use of tailored surgical approach and higher boost doses to the subglottic area postoperatively. The significance of the treatment center besides the well-known prognostic factors on the results of loco-regional failure free survival in our study proves the importance of the experience of the treatment team working in a multidisciplinary setting. 216 poster THE POTENTIAL UTILITY OF 3D POSITIONAL AND DOSIMETRICVE- RIFICATION IN PATIENTSUNDERGOING RADICAL RADIOTHERAPY FOR HEAD AND NECK CANCERS R. Height 1,M. Wada ~, A. Rolfo 1,D. Lim Joan 1, C. Mantle 1,C. Irvine ~, M. Feigen~,V. Khoo2 1AUSTINHEALTH, Department of Radiation Oncology, Melbourne, Aus- tralia, 2ROYALMARSDEN HOSPITAL, Department of Radiation Oncology, Sutton Surrey, UK Purpose/Objective: Radiotherapy treatment verification with cone beam CT scans will potentially improve the accuracy of treatment delivery. 3D patient set-up information acquired on treatment units will provide more accurate assessment of translational set-up errors. Furthermore there is scope for determining other positional varia- tions as rotation, more complex clinically important positional set up errors. When applied to the head and neck region, 3D informa- tion acquired will also yield potentially useful information regarding changes that occur in patients in response to treatment e.g. tumor growth / regression, oedema and weight loss, which may lead to changes in the external patient contour and positional/volumetric alteration in the target volumes and organs at risk. In anticipation of implementation of cone beam program in our department and to assess the potential utility of 3D information acquired during radical treatment, we used our repeat simulation CT acquired during the Sth week of treatment in the head and neck cohort, to assess the positional, volumetric and dosimetric information gained. Materials/Methods: Ten patients with primary mucosal SCCs en- tered in our re-scanning protocol were analyzed. Inclusion criteria included treatment with radical intent, either definitively or post operatively, with forward planned 3D CRT and where indicated with concurrently platinum based chemotherapy. Enteral nutritional support with PEG feeds were used in appropriate patients. Second CT scans were acquired at 42-50Gy on our simulation CT and IV contrast used to facilitate re-contouring where feasible. Available parameters as acute toxicity, clinical tumor response, and patient weight were documented at the time of the second CT for correla- tion. Simulation and repeat CT study-sets were co-registered, in or- der to assesscontour changes. Volumes of interest included external patient contour / shape, organs at risk and GTV primary and nodes. ,nitially optimized treatment plan was then overlaid on the second scan to assessthe"delivered dose"and dosimetric changes to target volumes and organs at risk. Results: Using image co-registration, set up variations not readily appreciable on 2D orthogonal assessment and with possible clinical and dosimetric consequences were apparent. These include slight variation in the jaw position in relation to the skull base, head at- titude, spinal flex and shoulder positioning, Volumetric and geo- graphical changes of the specified structures were assessed using absolute volume in ccs, centre of mass and dice coefficient analysis. Significant reduction in tumor volumes was seen at this stage. Ob-