S46 Early and locally advanced NSCLC stage NSCLC investigates the age-adjusted impact of number and kinds of baseline co-morbidities upon overall survival. Methods: Between 2000 and 2007, 88 consecutive stage I primary NSCLC patients, 81 peripheral and 7 central lesions, were treated by linac-based SBRT. Forty-five gray in three fractions prescribed to the isocenter was given until 2005 when 67.5 Gy in 3 fractions was introduced for peripherally located tumors. Baseline co-morbidities were retrieved by consultation of a formal electronic registry of diagnoses as well as patients’ charts. The age-adjusted Charlson Co-morbidity index (CCI) was scored for each patient and subjected to univariate and multivariate survival analysis. Results: With a median follow-up of 44 months, 5 patients developed local failure. The local control rate at 4 years was 89% while the median overall survival was 22 months. The most frequent adverse event was asymptomatic pulmonary fibrosis. The median CCI score was 5. The CCI was a significant predictor of overall survival on both univariate (p = 0.002) and multivariate (p = 0.006) analysis. Patients with a CCI score of 3 or less had a median survival of 41 months versus only 11 months for those scoring 6 or more. Conclusions: Longer follow-up confirms the excellent local control of SBRT for inoperable early stage NSCLC and appears well tolerated. Much higher survival rates were observed in patients with a low CCI. The persistently low survival outcome observed with SBRT does appear to be largely attributable to the advanced age and array of competing co-morbidities characterising this patient population. 105PD SEQUENTIAL CHEMORADIATION AND RADIATION ALONE IN THE TREATMENT OF INOPERABLE, LOCALLY ADVANCED NON-SMALL CELL LUNG CANCER (NSCLC) AND CORRELATION OF TUMOUR ERCC1 WITH SURVIVAL R. Davda 1 , J.S.G. Griffith 1 , D. Smith 1 , H. Makker 1 , G.A. Read 1 , D. Ward 1 , V. Sundaresan 2 , M. Falzon 3 , J.M. Singer 4 1 Oncology, North Middlesex Hospital, London, United Kingdom, 2 Pathology, Princess Alexandra Hospital, Harlow, United Kingdom, 3 Pathology, University College Hospital London, London, United Kingdom, 4 Oncology, Princess Alexandra Hospital, Harlow, United Kingdom Background: There is evidence supporting induction chemotherapy prior to radical radiotherapy in the treatment of locally advanced, inoperable NSCLC. However, such patients often have significant co-morbidities and/or are of an age or performance status which make them unable to tolerate the toxicities associated with chemo- radiation. DNA damage induced by cisplatin used in chemoradiation may undergo nucleotide excision repair. The excision repair cross- complementation group 1 (ERCC1) enzyme is involved in this process and therefore tumour expression of ERCC1 may be associated with resistance to cisplatin and poorer survival following treatment. Methods: Patients treated with chemoradiation received 2 cycles carboplatin (AUC5) day 1, vinorelbine 25 mg/m 2 days 1, 8 followed by CT planned conformal radiotherapy 55 Gy in 20 fractions and depending on tolerance, 2 further chemotherapy cycles. Patients who were considered unable to tolerate chemoradiation but still PS < 2 received radiotherapy 55 Gy in 20 fractions. We are currently performing ERCC1 immunohistochemical analysis in tumour biopsies taken at diagnosis and we will correlate these results with survival. Results: 127 patients were treated; 82 received chemoradiation, 45 radiotherapy alone. In the chemoradiation group the age range was 40 85 yrs with a median age of 67 yrs. In the radiotherapy group the age range was 42 89 yrs with a median age of 74 yrs. Median survival in the chemoradiation group was 554 days (18.2 months). Median survival in the radiotherapy group was 533 days (17.5 months). Using a chi squared test there was found to be no statistically significant difference in survival between the two groups (p = 0.35). Conclusions: Patients with inoperable stages II/III NSCLC who are judged unable to tolerate chemoradiation can be treated with radiation alone, giving comparable overall survival. We await ERCC1 analysis to compare expression with survival. 106PD DOES ELASTIC REGISTRATION ALTER FDG-PET BASED GTVS FOR RADIOTHERAPY PLANNING OF LUNG CANCER? U. Nestle 1 , A. Grgic 2 , N. Moca 2 , S. Kremp 3 , A. Schaefer 2 , J. Fleckenstein 3 , C. Kirsch 4 1 Klinik f¨ ur Strahlenheilkunde, Universit¨atsklinikum Freiburg, Freiburg, Germany, 2 Klinik f¨ ur Nuklearmedizin, Universit¨ atsklinikum des Saarlandes, Homburg/Saar, Germany, 3 Klinik f¨ ur Radioonkologie, Universit¨atsklinikum des Saarlandes, Homburg/Saar, Germany, 4 Klinik f¨ ur Strahlenheilkunde, Universit¨ atsklinikum Freiburg, Homburg/Saar, Germany Background: FDG-PET and PET/CT is increasingly used for radio- therapy (RT) planning in non-small-cell lung carcinoma (NSCLC). For registration of PET and CT data, non rigid (“elastic”) methods have been proposed. As by elastic fusion errors concerning lesion size or depiction of tracer uptake might occur and lead to false delineation of the radiotherapy target volume (GTV), we compared FDG-based GTVs (P-GTV) in rigidly (LR) and elastically (ER) fused PET-CT-datasets. Methods: 16 patients with histologically proven NSCLC underwent FDG-PET acquisition and 3 CT acquisitions in different breathing positions (expiration, inspiration, mid-breathhold) all in the same RT position on the same day. The PET-scans were registered to all CT-scans using a linear and a elastic method. In all resulting fused datasets, P-GTVs of the primary tumors were delineated using the same contrast oriented contouring algorithm. Results: The mean P-GTV determined after LR was 86.1±106.2 ml, whereas after ER it was 93.2±110.1 ml (r = 0.98). The mean maximum standardized uptake values (SUVmax) were 10.1 (±4.1) after LR and 10.1 (±3.9) after ER (r = 0.98). In intra- and interindividual comparison, there were no significant differences in size and SUVs of the tumors depending on the registration method. Conclusions: Our results indicate no clinically significant error concerning size and/or uptake depiction of lung tumors after elastic registration of FDG-PET-data to CT data acquired in different breathing positions. The question of tumor localisation is subject to further studies. 107PD DOSE-PER-FRACTION ESCALATION OF ACCELERATED HYPOFRACTIONATED THREE-DIMENSIONAL CONFORMAL RADIOTHERAPY IN LOCALLY ADVANCED NON-SMALL CELL LUNG CANCER L. Kepka, D. Tyc-Szczepaniak, M. Kolodziejczyk, K. Bujko Radiation Oncology, M. Sklodowska-Curie Memorial Cancer Center and Institute of Oncology, Warsaw, Poland Background: To determine the efficacy of accelerated hypofraction- ated three-dimensional conformal radiotherapy (3D-CRT) with dose- per-fraction escalation for treatment of stage III non-small cell lung cancer (NSCLC). Methods: Between 2001 and 2007, 173 patients with stage III NSCLC were treated using accelerated 3D-CRT and the simultaneous boost technique. Initially, the total dose of 56.7 Gy (including 39.9 Gy to the elective area) was delivered over 4 weeks in fractions of 2.7 Gy (1.9 Gy to the limited elective area). The dose-per-fraction escalation study commenced after the outcomes of 70 patients had been evaluated. The dose per fraction was increased from 2.7 Gy through 2.8 Gy to 2.9 Gy. The total dose increased from 56.7 Gy through 58.8 Gy (level 1 escalation) to 60.9 Gy (level 2 escalation). The dose to the elective area and the overall treatment time