to identify risk factors correlated to local recurrence: few data are available to identify risk factors for distant metastases. At the Universita ` Cattolica del Sacro Cuore of Rome a multidisciplinary rectal cancer database is available since 1990: we reviewed our clinical data to identify the metastases prediction risk factors in patients treated with preoperative radiochemotherapy. Materials/Methods: A total of 821 patients out of 1,421 were identified for the analysis. Selection criteria was: Stage cT3–4 N0-2 M0, preoperative treatment and a minimum follow-up of 2 years. The selected data for the analysis were: at the diagnoses, the cT and cN stage, the tumor length (#30 mm and .30 mm) and the tumor site (#30 mm and . 30 mm from the ano-rectal ring); after the preoperative radiochemotherapy: the pT and pN stage, the tumor length (#30 mm and .30 mm) and location (#30 mm and .30 mm from the ano-rectal ring) and TRG score. The radiation therapy dose ranged from 45 Gy to 55 Gy and chemotherapeutic agent(s) used included one or more of the following drugs: 5-fluorouracil (5-FU), mitomycin, cisplatin, oxaliplatin, raltitrexed, and capecitabine. All the patients underwent to a TME surgical resection; adjuvant 5-FU-based chemotherapy was administered in pa- tients with pathologic positive lymph nodes. Results: A rate of 20.6% of distant metastases was found. A total of 79.4% of patients was cT3 and 20.6% cT4; 17.3% was cN0 and 82.7% cN+. The rate of patients with a very low tumor was 41.2%, those with a mid-high lesion were 58.8%; the pretreatment tumor length was #30 mm in 9.5% of patients and .30 mm in 90.5% of patients. The distribution according to the pT-Stage was: 8% pT0, 8% pT1, 16% pT2, 61% pT3, and 7% pT4; the distribution according to the pN stage was: 52.6% pN0 and 47.4% pN+. The TRG stratification was: 11.2% TRG1, 14.3% TRG2, 29.6% TRG3, 42.9% TRG4, and 2% TRG5. By univariate analysis, among the pretreatment factors, only the cT stage was significantly correlated with the incidence of distant metastases (p = 0.043). In the group of post-treatment factors, there was a significant correlation between the pT (p \0.001), pN stage (p \0.001) and TRG (p \ 0.001) with the distant metastases rate. The tumor length and the location didn’t show a strong correlation (p = ns). By multivariate analysis the pT (p = 0.001), pN stage (p = 0.002), and the TRG (p = 0.001) predicted distant metastases. Conclusions: Our data show the presence of predictive risk factors of distant metastases namely related to the different response to the treatment. A score system to tailor the adjuvant treatment will be proposed. Author Disclosure: V. Valentini, None; M.A. Gambacorta, None; M.C. Barba, None; M. Balducci, None; B. Barbaro, None; A. Crucitti, None; C. Ratto, None; V. Papa, None; G.B. Doglietto, None; C. Coco, None. 1090 Pilot of Respiratory Gated Proton Beam Therapy for Liver Tumors T. S. Hong 1 , T. F. DeLaney 1 , H. J. Mamon 2 , A. Niemierko 1 , J. Adams 1 , B. Yeap 1 , L. S. Blaszkowsky 1 , D. P. Ryan 1 , C. G. Willett 3 , A. X. Zhu 1 1 Massachusetts General Hospital, Boston, MA, 2 Brigham and Women’s Hospital/Dana-Farber Cancer Institute, Boston, MA, 3 Duke University Medical Center, Durham, NC Purpose/Objective(s): Proton beam therapy may be an effective ablative strategy for liver tumors. However, proton-based gating strategies needed for treating moving targets with large doses have not been widely explored in the US due to the greater mechan- ical complexity of a cyclotron vs. linear accelerator. We report on the feasibility of gated proton beam therapy for liver tumors. Materials/Methods: Patients with hepatocellular carcinoma (HCC), intrahepatic cholangiocarcinoma (IHC), or hepatic metasta- ses, 1–3 lesion, 6 cm or less in maximum dimension, Childs A/B, and no extrahepatic disease were prospectively enrolled on an IRB-approved study. Prior therapy besides radiation was allowed. The study was planned as a dose/fraction escalation based on normal liver EUD using a 15 fraction schedule, with tumor dose constrained by normal tissue constraints to the chest wall (max- imum 60 Gy) and bowel (maximum 45 Gy). Treatment planning utilized 4D-CT: if the implanted fiducials moved greater than 1 cm peak-to-peak, gating was performed. The CTV was defined as a 0.5–1 cm expansion around the GTV at the physician’s discretion. The PTV was customized based on the amount of motion visualized during duty cycle. Gating was performed using an RPM in- terfaced with the cyclotron’s electric circuit. Prior to therapy, a film was taken during the duty cycle to ensure proper setup and functioning of the gating system. Treatments were performed on a 250 MeV cyclotron. Target coverage was verified 3 months post-treatment with MRI by radiation-induced T2 signal change surrounding the tumor. Results: Fifteen patients were enrolled from February 2006–March 2009, of which 14 are available for outcomes analysis. Eleven patients had HCC, 3 patients had IHC, and 1 patient had a solitary liver metastasis. Ten patients had 1 lesion, 3 patients had 2 le- sions, and 2 patients had 3 lesions for a total of 22 treated lesions. Planned dose escalation was not achieved because of non-liver normal tissue dose. Total dose over 15 fractions ranged from 52.5 Gy RBE–75 Gy RBE. Post-treatment MRI was performed in 12 patients and confirmed complete target coverage. With a median follow-up of 12 months, median survival is 12 months, and mDFS is 11 months. The LC is 100% in the treated lesions. Eight patients have died: 4 without recurrence, 2 with ‘‘elsewhere’’ hepatic recurrence, and 2 with extrahepatic recurrence. No patient developed radiation-induced liver disease. Following treatment, 2 pa- tients (both Childs B) developed Grade 3 bilirubinemia, 1 patient a Grade 3 GI bleed (resolved without surgery), and 1 patient with multiple severe medical comorbidities a Grade 5 stomach perforation. Conclusions: Gated proton beam therapy for liver tumors is feasible and well-tolerated. Tumor dosing is limited by chest wall or bowel constraints, rather than liver constraints. A Phase II study for HCC or IHC with expanded size criteria is currently being developed. Author Disclosure: T.S. Hong, None; T.F. DeLaney, None; H.J. Mamon, None; A. Niemierko, None; J. Adams, None; B. Yeap, None; L.S. Blaszkowsky, None; D.P. Ryan, None; C.G. Willett, None; A.X. Zhu, None. 1091 Patterns of Recurrence and Nodal Staging in Rectal Cancer Patients Undergoing Sentinel Lymph Node Mapping Compared to Conventional Surgery S. Saha 1 , 2 , M. Ghanem 1,2 , H. L. Gayar 1,2 , J. L. Nettleton 1,2 , W. Naill 1,2 , B. Chakravarty 1,2 , D. Wiese 1,2 , S. Sirop 1,2 , M. L. Arora 1,2 , T. Singh 1,2 1 McLaren Regional Medical Center, Flint, MI, 2 Michigan State University, Flint, MI Purpose/Objective(s): Nodal status remains one of the most important prognostic factors in rectal cancer. Sentinel Lymph Node Mapping (SLNM) has been found to upstage more patients with nodal disease compared to conventional surgery. Hence, S166 I. J. Radiation Oncology d Biology d Physics Volume 75, Number 3, Supplement, 2009