Poster discussions

Annals of Oncology, Jun 2012

Introduction The purpose of the present study was to examine the prevalence of sexual dysfunction in newly diagnosed sexually active colorectal cancer female patients prior to treatment and identify the characteristics and correlates of dysfunction.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

Poster discussions

Thomas Gruenberger 0 3 7 11 17 25 31 39 44 40 54 65 75 Friedrich Lngle 0 2 8 10 16 24 30 38 40 46 52 64 73 Josef Thaler 0 1 7 9 15 23 29 37 43 47 51 63 72 Wolfgang Eisterer 0 6 7 14 21 28 35 42 46 50 58 69 79 Alfred Lenauer 0 2 8 10 16 24 30 38 40 46 52 64 73 Stefan Stremitzer 0 3 13 20 27 34 41 45 49 57 68 78 Judith Stift 0 5 12 19 26 33 40 41 48 56 67 77 Birgit Gruenberger 0 4 18 32 38 55 66 76 0 Junichi Arita, Nobuyuki Takemura, Ryuji Yoshioka, Ono Yoshihiro, Fumitake Suzuki, Rintaro Koga, Takeshi Sano, Akio Saiura and Toshiharu Yamaguchi Department of Digestive Surgery, Cancer Institute Hospital of Japanese Foundation for Cancer Research , Tokyo 1 Department of Internal Medicine , Wels 2 Department of Surgery , Wr Neustadt 3 MUW, Department of Surgery , Vienna 4 Bamherzige Brueder Wien, Department of Internal Medicine , Vienna 5 MUW, Department of Pathology , Vienna 6 Department of Internal Medicine I, Medical University Innsbruck , Innsbruck 7 Rouen University Hospital , Rouen 8 La Piti Salpetriere, Paris 9 Department and Policlinics of Diagnostic Radiology , Dresden , Saxony 10 Department of Clinical Radiology, University Hospital Grosshadern , Munich , Bavaria 11 Department of Oncology, Klinikum Grosshadern and Comprehensive Cancer Center , LMU, Munich , Bavaria 12 European Cyberknife Center Munich , Munich , Bavaria 13 Department of Surgery, University Hospital Grosshadern , Munich , Bavaria 14 Department of Medical Oncology , Klinikum Grosshadern, University of Munich , Munich , Bavaria 15 University College London , London, UK 16 Mount Vernon Hospital Middlesex , UK 17 Imperial College, Department of Cancer and Surgery , London, UK 18 Duke University Medicine , Durham, NC 19 Imperial College London , London, UK 20 University College London, Cancer Research UK & UCL Cancer Trials Centre Cancer Institute , London, UK 21 Imperial College , London, UK 22 Memorial Sloan Kettering Cancer Center , New York, NY 23 Charles University, 1st Faculty of Medicine, Central Military Hospital , Prague 24 Masaryk University, Institute of Biostatistics and Analyses , Brno 25 Charles University, 1st Faculty of Medicine , Prague 26 Department of Cancer Prevention, Institute for Cancer Research, The Norwegian Radium Hospital, Oslo University Hospital , Oslo 27 Department of Gastrointestinal Surgery , Aker, Oslo University Hospital , Oslo 28 Charles University, 1st Faculty of Medicine, Central Military Hospital, Department of Medicine , Brno 29 Gabrail Cancer Center , Canton, Ohio 30 Pacific Oncology Hematology Associates , Encinitas, California 31 The University of Texas M.D. Anderson Cancer Center , Houston, Texas 32 Imperial College Hammersmith Campus , Cyclotron Building, London, UK 33 Oncologia Medica, Azienda USL-5, Istituto Toscano Tumori , Pontedera, Italy 34 Sarah Cannon Research Institute , Nashville, Tennessee 35 Daiichi Sankyo, Inc., Edison, New Jersey 36 Imperial College Healthcare NHS Trust Departments of Medical Oncology, Surgery and Radiology, Charing Cross Hospital , London, UK 37 St-Luc University Hospital , Brussels 38 Amgen Inc., Thousand Oaks, California 39 University Hospital Gasthuisberg , Leuven , Flemish Brabant 40 Antwerp University Hospital , Edegem, Antwerp 41 Amgen (Europe) GmbH, Zug 42 Ospedale Niguarda Ca' Granda, Milan 43 Ospedale Niguarda Ca' Granda, Milan 44 Institut Gustave Roussy , Paris 45 Academic Medical Center, University of Amsterdam , Amsterdam 46 ICO R. Gauducheau, Nantes 47 University Hospital Gasthuisberg , Leuven 48 Amgen, Thousand Oaks, CA 49 Amgen, San Francisco, CA 50 Ospedale Niguarda Ca' Granda, Milan 51 Graduate School of Medicine , Osaka 52 Fudan University Cancer Hospital , Shanghai, Shanghai 53 Novartis Pharmaceuticals Corporation , Florham Park 54 University of Texas MD Anderson Cancer Center , Houston 55 Hebei Medical University Fourth Hospital , Shijiazhuang, Hebei 56 Kyorin University Hospital , Tokyo 57 Clinico Anglo Americana , Lima 58 Shikoku Cancer Center , Matsuyama 59 Christie NHS Foundation Trust , Manchester 60 Asan Medical Center , Seoul 61 Seoul National University Hospital , Seoul 62 Novartis Pharma AG, Basel 63 Universittsklinikum Gttingen , Klinik fr Gastroenterologie und Endokrinologie, Gttingen 64 Stdtisches Klinikum Magdeburg, Klinik fr Allgemein- und Viszeralchirurgie, Magdeburg , Sachsen-Anhalt 65 University Medical Center , Hamburg-Eppendorf, Hamburg 66 Universittsklinikum Heidelberg , Chirurgische Klinik, Heidelberg , Baden-Wrttemberg 67 Krankenhaus St. Vinzenz, Innere Medizin, Zams, Tirol 68 Hubertus Wald Tumor Center, University Cancer Center Hamburg (UCCH) , Hamburg 69 Klinik fr Onkologie, Hmatologie, Knochenmarktransplantation mit Sektion Pneumologie, Hamburg 70 Klinikum rechts der Isar der TU Mnchen, Chirurgische Klinik und Poliklinik, Mnchen, Bayern 71 Klinik fr Innere Medizin, Berlin 72 Kantonsspital St. Gallen, St. Gallen 73 Krankenhaus Dresden-Friedrichstadt, Dresden , Sachsen 74 University Hospital Erlangen , Erlangen, Bayern 75 Surgical Department, University Hospital Erlangen , Erlangen, Germany 76 Universittsklinikum Leipzig , Leipzig , Sachsen 77 Universittsklinikum Leipzig, Department fr Operative Medizin , Leipzig , Sachsen 78 Klinikum Grohadern, Mnchen, Bavaria 79 University Hospital Frankfurt/Main , Frankfurt/Main, Hessen 80 DiaCura Coburg, Coburg, Bayern 81 Gray Institute for Radiation Oncology and Biology , Oxford Introduction: The purpose of the present study was to examine the prevalence of sexual dysfunction in newly diagnosed sexually active colorectal cancer female patients prior to treatment and identify the characteristics and correlates of dysfunction. Methods: Participants (n = 49) completed the Female Sexual Function Inventory (FSFI), menopause Symptom Scale (MSS) Medical Outcomes Study-Cohort Form (SF-36), abbreviated Dyadic Adjustment Scale (ADAS) and Dyadic Sexual Communication Scale (DCS). Results: Mean age of participants was 55 years (SD = 12.9). Thirty-six percent reported that they had been sexually active in the last month. Sexually active women were younger, had higher incomes and were more likely to be in a committed relationship (P < 0.05). Among those who were sexually active, mean FSFI score was 22.71 (SD = 7.3): 63% met criteria (total FSFI score less than 26) for sexual dysfunction. Dysfunction was more likely to occur in the domains of arousal, orgasm, pain and satisfaction but not desire of lubrication. Vasomotor symptoms and worse physical well-being were associated (P < 0.05) with dysfunction. Finally, poorer relationship quality and worse sexual communication were also associated ( p values < 0.05) with dysfunction; emotional distress was not (P > 0.05). These variables accounted for 36% of the variability in pretreatment sexual dysfunction. Conclusion: Nearly two-thirds of newly diagnosed sexually active colorectal cancer female patients experience significant dysfunction at diagnosis. Factors that may participate in dysfunction include select demographic factors, physical symptoms, and relationship factors. - SEXUAL DISORDERS IN FEMALE PATIENTS WITH COLORECTAL CANCER QUALITY CONTROL IN LYMPHADENECTOMY FOR LOCALIZED COLON CANCER Background: The most important determinant of prognosis for locoregional colon cancer is lymph node involvement. The American Joint Commission on Cancer and the College of American Pathologists have recommended examination of at least 12 lymph nodes to ensure adequate and accurate staging. We sought to determine if an appropriate lymphadenectomy was being performed routinely in a large community based metropolitan health system, and whether or not the results of lymphadenectomy could be improved through education and observation of health care providers and their individual practices. Methods: A total of 1111 patients comprising AJCC stage I (n = 323), stage II (n = 386), and stage III (n = 402) colon cancer were evaluated in this study. Patients underwent resection by either open (n = 665) or laparoscopic (n = 418) approach. Lymphadenectomy results were obtained from postoperative pathology reports and tracked across four separate time periods: era 1 (1/1/02 1/31/04), era 2 (1/1/05 1/ 31/06), era 3 (1/1/07 1/31/07), and era 4 (1/1/08 12/31/08). Results were compared against a study endpoint measurement (1/1/09 12/31/10). Results: A significant improvement in lymphadenectomy results (12 nodes) was noted for all stages: AJCC stage I (34.2% vs. 59% vs. 65% vs. 69% vs. 61%, P = 0.0001), stage II (62% vs. 76% vs. 90% vs. 91% vs. 84%, P = 0.0004), and stage III (55% vs. 72% vs. 84% vs. 86% vs. 82%, P Surgical approach was not a significant factor. Neither age nor sex was significant. Conclusions: By raising awareness of optimal staging practices for colon cancer within a large metropolitan health care system, the standard of care being practiced in the community was significantly improved and this elevated standard has been maintained over a study observation period >5 years. Introduction: Two imaging modalities dedicated for liver specific phase, namely Gd-EOB-DTPA enhanced MRI (EOB-MRI) and contrast-enhanced intraoperative ultrasound (CE-IOUS) using Sonazoid have been recently developed. Both are reportedly useful for diagnosis and staging in patients undergoing surgery for colorectal liver metastasis. The aim of this study is to compare diagnostic ability of these modalities. Methods: A prospective clinical trial has been undertaken (UMIN-CTR ID: 000005537). Consecutive 84 patients with colorectal liver metastasis were scheduled for liver resection between January 2011 and November 2011. Of these, 66 patients underwent contrast-enhanced CT, EOB-MRI, preoperative contrast-enhanced ultrasound, fundamental intraoperative ultrasound, CE-IOUS, and liver resection of curative intent. Eighteen patients were excluded from this trial because they underwent laparoscopic liver resection (n = 7), did not undergo contrast-enhanced CT due to a history of allergy (n = 1), underwent probe laparotomy due to peritoneal metastasis (n = 1), had more than ten liver metastases (n = 9). Hepatobiliary phase was referred for diagnosis of liver metastasis in EOB-MRI. After laparotomy and exploration including fundamental IOUS, 0.5 ml of Sonazoid was injected through peripheral vein. Thorough liver screening on CE-IOUS was performed 10 to 15 minutes later under harmonic mode, where hypoechoic liver lesions were regarded as liver metastases. A nodule which was temporary diagnosed as metastasis in any imaging modality was resected. Radiofrequency ablation was not performed. All possible liver metastases were resected as long as it was identified by IOUS. Intraoperative needle biopsy was allowed alternative to resection for indeterminate nodules. Results: A total of 190 nodules were temporary diagnosed as liver metastases in at least one imaging modality in the 66 patients. Of 190, 177 were diagnosed as adenocarcinoma in histological examination, 8 were adversely diagnosed as benign lesion in histological examination, and the other 5 could not be identified during operation and were finally diagnosed benign through postoperative follow-up. Fifteen were newly found during IOUS and 11 of them were finally diagnosed as adenocarcinoma. Thirteen were newly found during CE-IOUS and 11 of them were finally diagnosed as adenocarcinoma. Sensitivity and positive-predictive value were 83 and 97% for contrast-enhanced CT, 83 and 97% for EOB-MRI, 76 and 96% for preoperative contrast-enhanced ultrasound, 89 and 96% for fundamental IOUS, and 98 and 97% for CE-IOUS. Conclusion: EOB-MRI and contrast-enhanced CT are complementary to each other. CE-IOUS scored high diagnostic value for diagnosis of colorectal liver metastasis. PERIOPERATIVE CHEMOTHERAPY INCLUDING BEVACIZUMAB IN POTENTIALLY CURABLE METASTATIC COLORECTAL CANCER; A MULTICENTER, SINGLE ARM PHASE I/II TRIAL: ASSO LM1 Introduction: Patients with colorectal cancer (CRC) and liver metastases may benefit from pre- and postoperative chemotherapy and disease resection. Best treatment sequence, length and combination still needs to be determined. We evaluated Bevacizumab and XELOX in a multicenter approach. Resectabilty was the primary endpoint, periOP morbidity, response rate, recurrence-free and overall survival the secondary end points. Methods: Patients with liver metastases from CRC potentially curable by resection were eligible for this trial. Patients received Bevacizumab 5mg/kg (day 1), oxaliplatin 85mg/m2 (day 1) and Capecitabine 3000mg/m2/day (days 17) of a 2-week cycle for six cycles. The sixth cycle did not include Bevacizumab, resulting in 5 weeks between PD-0002 Table 1 the last Bevacizumab dose and surgery. Therapy with Bevacizumab plus XELOX was restarted 5 weeks after surgery for another six cycles. Response rates were assessed by the PIs and after central review, pathological response was assessed centrally according to Rubbia-Brandt. Results: 43 patients ( pts) with a median age of 66 years (38-80) were enrolled into this multicenter trial. 78% of these pts had synchronous disease, 51% of these started treatment without primary tumor resection. The median number of metastases was 2 (1-10) and median Fong score was 3 (0-4). In total of 8 pts were ineligible for resection, 1 due to PD, 7 due to protocol violation or personal reason. Resectability of operated patients was 97% with 34 R0 resections and one R1 resection. Postoperative morbidity occurred in 22% of the pts, 3 operative revisions were related to the primary tumor resection. Efficacy results for response in 38 eligible patients confirm an ORR of 66%, 31% SD and 3% PD according to RECIST. Preoperative grade 3/4 adverse events were 17% diarrhea, 5% HFS, 2% neuropathy and 5% thromboembolic events. Histological assessment of resected metastases revealed a major histological response in 56%, a partial histological response in 28%, and no histological response in 16%. Recurrence-free (RFS) and overall survival (OS) significantly differed dependent upon the fulfillment of the adjuvant treatment cycles in curative resected patients (24 pts completed periOP CTx vs 10 pts neoCTx only): median RFS was 26.2 months in patients receiving postsurgical Bev + Xelox vs. 7.4 months in those without postOP therapy ( p <.05); similarly median OS was significantly prolonged in pts after postOP CTx (47.9 mts vs 27.7 mts; p <.05). Conclusion: These data demonstrate that Bevacizumab and XELOX provide substantial response rates with well-known adverse events in this potentially curative mCRC setting. In patients eligible for resection after 3 months of neoadjuvant therapy perioperative morbidity is not increased compared to studies without Bevacizumab. Long-term recurrence-free and overall survival was significantly prolonged in our study in patients receiving postoperative chemotherapy plus Bevacizumab. A STEP TOWARDS NOTES TOTAL MESORECTAL EXCISION FOR RECTAL CANCER: ENDOSCOPIC TRANS ANAL PROCTECTOMY (ETAP) Introduction: NOTES is an emerging concept which has been recently applied to the field of rectal excision. The aim of this study was to assess the outcome of NOTES proctectomy. iv | poster discussions Methods: From Feb 2010 to Oct 2011, 32 patients (25 males) underwent ETAP for adenocarcinoma. Surgical technique: a full thickness circumferential rectal transaction was performed above the dentate line and perirectal fat identified. Proximal dissection along the plane of perirectal fat is carried out and the rectum is mobilized as high as possible. An endorec Trocar (Aspide France) was inserted transanally. The dissection was then extended to achieve circumferential rectal mobilization. In male the dissection could be done in front or behind the Denonvilliers fascia. Finally, the peritoneal reflection was identified and perforated to enter the peritoneal cavity at the level of the Douglass pouch. Laparoscopy was then performed in a standard fashion. The rectosigmoid was exteriorized transanally and a lateroterminal handsewn anastomosis fashioned. Results: Mesorectum was intact in all cases. The duration of the procedure lay between 3 and 5 hours. 12 to 16 lymph nodes were retrieved. There were no perioperative deaths. With our limited experience, ETAP could be performed in all cases of low and mid rectal cancer when a TME with coloanal anastomosis is indicated. However this approach has a greatest advantage in the more difficult cases such as: - bulky tumor of the middle rectum - narrow pelvises especially in male patient - obese patients - patients in which laparoscopic or large laparotomy is to be avoided (e.g. patients with cardio-pulmonary insufficiencies) in which case the procedure could be completed by NOTES and small Pfannenstiel incision. Conclusion: There is a definite desire to advance the borders of NOTES within the arena of rectal surgery. Our preliminary experience shows that this procedure is safe and effective in the short term, but we lack data on the oncological outcome in the long term. Certainly, more data will be required on the safety and effectiveness of this technique and the long term oncological outcome which are far more important aspects of treatment than the scar less surgery. THE PROGNOSTIC UTILITY OF METABOLIC INFORMATION PROVIDED BY FDG PET IN THE EVALUATION OF SOLITARY COLORECTAL HEPATIC METASTASES Calvin Park1, Michael Lin1, Thomas Hugh2 and Weng Ng1 1University of New South Wales, Sydney, New South Wales, 2University of Sydney, Sydney, New South Wales Introduction: Background: Hepatic metastasis is the most common site of recurrence in colorectal cancer (CRC). Early detection and resection improve survival. There have been many attempts to establish a reliable pre-operative clinical scoring system for selecting suitable candidates for surgery. Recently positron emission tomography (PET) using the glucose analogue FDG has made a significant clinical impact in the assessment of these patients. FDG PET reduces futile hepatic resection by selecting more appropriate patients for surgery, and excluding patients with more extensive hepatic disease or extrahepatic metastases. Semi-quantitative measurements using standardized uptake value (SUV) evaluates the degree of metabolic activity of the metastasis, has also shown potential as a prognostic indicator. Methods: Patient and Departmental PET data-bases were reviewed and those meeting the following criteria were included: (1) Patients with resected primary colorectal cancer, (2) first CRC recurrence in liver (3), solitary hepatic metastasis as reported on PET or PET/CT. Patients with extra-hepatic metastasis as reported on PET or PET/CT and patients who received neo-adjuvant chemotherapy or other forms of local hepatic therapy prior to PET are excluded. A total of 71 usable patient data was collected. SUV measurements, glycolytic tumor volumes and clinopathological data were collected. Overall survival (OS) and disease free survival (DFS) were used as outcomes and analysed using the Kaplan-Meier analysis. Results: Five-year overall survival was 57.2%. OS was strongly associated with clinicopathological factors such as the interval between primary resection and PET scan with an interval <420 days demonstrating better survival ( p < 0.06) and also the number of recurrences (single vs multiple, p < 0.034). The size of the hepatic metastasis as measured both on PET and histopathology were predictive. 5yr OS was 67% and 39.6% for lesions >48mm and > 48mm respectively on PET ( p > 0.03) and 66.1% and 25.4% for lesions C < 60mm and C > 60mm for surgical specimens ( p < 0.007). SUV ratio (SUV of lesion/ normal liver) of the hepatic metastasis on PET predicted OS with a threshold of 4.1 (<4.1= 60.3%, and >4.1 =26.7%, p = 0.039). Glycolytic volumes at SUV thresholds of 3 and 3.5 also demonstrated statistically significant differences in survival (cut-off volumes of 80cc and 150cc). Five-year overall disease free rate was 54.3%. Disease free interval was associated with the number of nodes examined in primary CRC (cut-off 17, less is better, p < 0.052). The Duke stage of the primary CRC and the site of the primary (colon vs rectum) were not predictive of OS or DFS in this group of patients. Conclusion: We have found that PET had strong utility in predicting overall survival in patients with CRC and solitary hepatic metastasis. Metabolic information provided by PET in terms of SUV ratios and glycolytic volumes were prognostic and may be useful as an adjunct to clinicopathological factors in selecting patients for adjuvant chemotherapy or targeted therapy. 18F-FDG POSITRON EMISSION TOMOGRAPHY (FDG-PET) AND RESPONSE TO NEOADJUVANT CHEMORADIOTHERAPY IN RECTAL CANCER Encarnacin Gonzlez-Flores1, Julia Ruiz2, Beatriz Gonzalez-Astorga3, Vernica Conde4, Javier Garcia Garcia4, Aranzazu Gonzalez-Vicente3, Jesus Soberino4, Cynthia Gonzalez5, Luque-Caro Raquel6, Carmen Sanchez-Toro7, Joaquina Martinez-Galan3 and Juan Ramn Delgado8 1H Virgen de las Nieves, Granada, 2"Virgen de las Nieves", Granada, 3"Virgen de las Nieves" Hospital, Granada, 4"Virgen de las Nieves" Hospital, Granada, 5"Virgen de las Nieves" Hospital, Granada, 6Hospital Universitario Virgen de las Nieves, Granada, 7"Virgen de las Nieves" Hospital, Granada, 8"Virgen de las Nieves" Hospital, Granada Introduction: Prediction of rectal cancer response to preoperative, neo-adjuvant chemo-radiation therapy (CRT) provides the opportunity to identify patients in whom a major response is expected and who may therefore benefit from alternative surgical approaches. Traditional morphological imaging techniques are effective in defining tumour extension in the initial diagnostic and staging work-up, but perform poorly in distinguishing residual neoplastic tissue from scarring post CRT, when restaging the patient before surgery. Fluorine-18 fluorodeoxyglucose positron emission tomography (FDG-PET) is a promising tool for monitoring the effect of anti-tumour therapy. The aim of this study was to prospectively assess the value of sequential FDG-PET scans in predicting the response of locally advanced rectal cancer to neo-adjuvant CRT. Methods: 42 patients with locally advance rectal cancer who undergo a neo-adjuvant therapy are included in this study. All patients received chemoradiotherapy treatment. PET-TAC was performed before and six weeks after completion of preoperative neo- adjuvant protocol. Surgical resection included of total mesorectal excision. PET- TAC parameters included maximum standard uptake value (SUV max) and the mean percentage decrease in SUV max. Results: were correlated with pathological response, assessed both by histopathological staging of the surgical specimens ( pTNM). In all patients the primary lesions showed high SUV max (SUV max 13.14 +/-5). Following neo-adjuvant CRT, of the 42 patients submitted to surgery, 37 patients (88.1%)were classified as responders with relevant decreasing in FDG uptake (75% +/- 25%) while the remaining 5 patients (11.9%) were non-responders. After neo-adjuvant therapy 5 patients (11.9%) showed negative PET-TAC, with complete pathological remission and tumor absence in primary local in 3 patients and 2 patients had partial pathological remission. After neo-adjuvant therapy 32 patients (76.1%) showed decrease in Suv max and16 patients presented partial pathological remission, with a correlation of 84.2% between decrease in Suv max and partial pathological remission of total patients. Conclusion: The results suggest the potential utility of FDG-PET as a complementary diagnostic and prognostic procedure in the assessment of neo-adjuvant CRT response of locally advanced rectal cancer. SUV(max) seem the best predictor of CRT response. RADIOFREQUENCY ABLATION (RFA) VERSUS ROBOTIC RADIOSURGERY (RRS) IN THE TREATMENT OF UNRESECTABLE COLORECTAL LIVER METASTASES Introduction: Radiofrequency ablation (RFA) and stereotactic radiation are treatment possibilities for unresectable colorectal liver metastases. In this analysis the efficacy of RFA and robotic radiosurgery (RRS) was compared in a total of 60 patients. Methods: The authors prospectively followed 30 patients with colorectal liver metastases not qualifying for surgery who were treated in curative intent with RRS. To compare efficacy of both treatment modalities, patients treated with RFA during the same period of time were matched according to number and size of the treated lesions. Local tumor control, local disease-free survival (local DFS), and disease-free survival (DFS) were analyzed to evaluate the efficacy. Treatment related side effects were recorded for comparison. Results: The median maximum diameter of the matched lesions was 33 mm (7-53 mm). Baseline characteristics did not differ significantly. Local control rate showed a significant difference and favored RRS (76.6% vs. 40.0%; p = 0.008). This related to a significantly longer local DFS of patients treated with RRS compared to RFA (34.4 months vs 7.1 months for RFA; p < 0.001). Both, median DFS (11.4 months for RRS vs 5.9 months for RFA p = 0.16) and the overall recurrence rate (70% for RRS and 80% for RFA, p = 0.55) were comparable between both treatment groups. Conclusion: Single session robotic radiosurgery (RRS) is a safe and effective method to treat colorectal liver metastases. In this analysis, local tumor control achieved by RRS was superior to RFA treatment. PRIMARY TUMOR LOCATION IS A MAJOR INDEPENDENT PROGNOSTIC FACTOR FOR MCRC PATIENTS Introduction: Previous attempts to address the question of whether primary tumor location affects outcome of mCRC were limited by small sample size, a high degree of heterogeneity in received treatments, and/or limited information regarding molecular and pathologic features. We performed the present analysis to clarify the prognostic impact of primary tumor location in mCRC patients ( pts). Methods: Cancers proximal or distal of the splenic flexure were classified as right- or left-sided respectively. Firstly, we tested the association of tumor location with outcome in a cohort of 200 pts treated with FOLFIRI plus bevacizumab within a prospective 1st-line pharmacogenetic study. For these pts data on mucinous histology and BRAF mutational status were available. Secondly, data from two large randomized phase III trials (AVF2107 and NO16966) of first-line chemotherapy plus or minus bevacizumab were used for validation. Primary endpoint was OS. Results: In the initial cohort 56 (28.0%) and 144 (72.0%) were right-sided and left-sided respectively. Right-sided cancers had a worse outcome both in terms of PFS and OS (PFS, left vs right: HR = 0.43[95%CI: 0.28-0.66], p = 0.0001; OS: HR = 0.37 [95%CI:0.22-0.63], p = 0.0003). A multivariate model confirmed right-sided location as an adverse prognostic variable independent of histology and BRAF mutational status. In the subgroup of non-mucinous and BRAF wild-type tumors (N = 155), pts with right-sided achieved median PFS and OS of 10.0 and 28.9 months compared to 13.0 and 47.6 mos of left-sided (PFS, left vs right: HR = 0.54[95%CI:0.34-0.84], p = 0.003; OS: HR = 0.52[95%CI:0.30-0.93], p = 0.022). In the AVF2107 study, data regarding primary tumor location were available for 559 pts (68.8%). Of these, 353 (63.2%) were left-sided and 206 (36.8%) right-sided. Left-sided cancers achieved a better OS and PFS compared to right-sided (median OS 20.4 vs 14.6 months, HR = 0.55[95%CI: 0.43-0.70], p < 0.0001; median PFS 8.5 vs 7.1 months, HR = 0.68 [95%CI: 0.55-0.83], p = 0.0003). In NO16966 trial, analyses were conducted on 1268 pts (62.3%). Of these, 935 were left-sided (73.7%) and 333 (26.3%) right-sided, and the left-sided cancers achieved a better OS and PFS compared to right-sided (median OS 23.0 vs 18.0 months, HR = 0.71[95%CI: 0.62-0.82], p < 0.0001; median PFS 8.9 vs 7.6 months, HR = 0.90[95%CI: 0.79-1.03], p = 0.12). In both studies, the efficacy of bevacizumab was independent of tumor location. Conclusion: These data suggest a strong prognostic impact of primary tumor location on mCRCs outcome. Tumor location did not impact the effect of bevacizumab. Recent data on mRNA signatures suggest a different molecular background for left vs right-sided CRC. Given the consistency of the data across an exploratory set and two confirmatory phase III studies, side of origin should be considered as a stratification factor in future randomized clinical trials in mCRC. PD-0009 Table 1 QUALITY OF COLONOSCOPY PERFORMANCE AMONG GASTROENTEROLOGISTS AND SURGEONS IN SINGLE TERTIARY CENTER IN QUEBEC Hugo Bernard1, Feriel-Lylia Messekher1, Vronique Lussier2 and Charles Mnard2 1Universit de Sherbrooke, Sherbrooke, Qubec, 2CHUS-Hotel-Dieu, Sherbrooke, Qubec Introduction: High quality endoscopy is a prerogative for an optimal colorectal cancer screening1. Adenoma detection rate, ceacal intubation rate, withdrawal time, and complication rate are objective measures of colonoscopy performance. However, previous studies have shown variable colonoscopy performance among different endoscopy specialties. Therefore, we used endoscopy quality markers to compare gastroenterologists (GI) and surgeons (GS) in our institution. Methods: This study was conducted in a single tertiary care teaching hospital, in Sherbrooke, Quebec. Data were prospectively collected from June 2009 to June 2010 from 475 consecutive screening colonoscopies conducted by either gastroenterologists or general surgeons. Quality parameters data were recorded for each colonoscopy and subsequently compared between the two groups. Results: A total of 475 patients were included, 52% female, with a mean age of 58. Twelve gastroenterologists and ten surgeons performed respectively 228 (48%) and 247 (52%) colonoscopies. The completion rate was 98.2% for gastroenterologists and 94.3% for surgeons ( p = 0.03). Withdrawal time over 6 minutes was achieved in 86.1% of gastroenterologists and 71.7% of surgeons ( p < 0.001). The polyp detection rate was 44.4 %. The adenoma detection rate (ADR) for gastroenterologists was 26.8% and for surgeons 23,5% ( p = 0.46). Endoscopist-specific ADR ranged between 5.9-50.0 % for GI and 11.4-41.2% for GS. Procedure complications rate was 0.44% for GI and 0.81% for GS ( p = NS). There were 2 post-polypectomy bleedings treated with supportive care and 1 unspecified complication. No perforation occurred. Sedation complications rate was 4.39% for GI and 9.72% for GS ( p = 0.02). Conclusion: Surgeons and gastroenterologists in this tertiary care teaching hospital performed equally well in regard to ADR. Gastroenterologists had a significantly better completion rate, less sedation related complication and an average longer withdrawal time. Regardless of these differences, both specialties offered quality colonoscopies, respecting the recommended standards of care5. There is still a large inter-endoscopists variability in ADR in both specialty. This issue needs further study and a larger population to draw appropriate conclusion. METABOLIC SYNDROME (METS) AND RISK OF COLORECTAL CANCER (CRC), A SYSTEMATIC REVIEW AND META-ANALYSIS Introduction: Metabolic syndrome (MetS) is defined as a cluster of risk factors for cardiovascular disease and type II diabetes which occur together and include hyperglycaemia, hypertension, raised triglyceride levels, low HDL levels and central obesity. The The aim of the study was to evaluate the association between MetS and colorectal cancer. A meta-analysis was performed on the studies that investigated the association between MetS and CRC. iv | poster discussions Methods: Medline/PubMed and Embase were searched for all articles published in English language related to MetS and CRC using the PRISMA guidelines. Articles published between 1965 and 2011 were reviewed to identify studies that reported the association of MetS, or syndromes that were indicative of the present day definition of MetS and CRC. There were 445 articles identified from the literature. A total of 374 abstracts were reviewed (by MA-W and AS) and 71 papers were examined. Results: The studies included 739,231 participants, which included 395,572 males and 337,756 females. This involved 11,769 cases of CRC. Significant associations between MetS and CRC were observed. Based on 14 studies, the overall OR for CRC incidence or mortality in MetS in both genders is 1.51, with a 95% C.I of 1.32 to 1.73 ( p < 0.00001. The incidence and mortality for males and females were OR 1.35 (95% C.I 1.19 to 1.53) and OR 1.47 (95% C.I 1.17 to 1.85) respectively. Eleven studies adjusted for confounding factor. Where studies which did not adjust for confounders were excluded, the OR was similar and retained significance at 1.40 (95% C.I 1.24 to 1.57) in both males and females. The OR for CRC morality in MetS was 1.90 (95% CI 1.00 to 3.60). The main finding from this meta-analysis is that MetS is significantly associated with increased risk of CRC (OR 1.51, 95% CI 1.32 to 1.73). This indicated that CRC risk may be greater among females with MetS (OR 1.47, 95% C.I 1.17 to 1.85) than males (OR 1.35, 95% C.I. 1.19 to 1.53), however as the confidence intervals overlap this gender difference is not significant. Most previous studies have stated the association between MetS and CRC risk is greater in males. Conclusion: This meta-analysis suggests a significant effect of metabolic syndrome on colorectal cancer, and emphasizes the importance of preventing and controlling MetS components; CRC is a disease to some extent preventable with changes in lifestyle and dietary habits. Malignancy itself may be the factor causing effects on lipids, so further confounding the influence of MetS, which has both systemic and hormonal effects in cancer pathogenesis. The exact mechanism how MetS relates to CRC pathophysiology is yet to be clarified and we suspect that MetS is more likely to precede the diagnosis of CRC. Therefore, understanding the pathophysiology of MetS would enable the development of therapeutic and preventative strategies in relation to the CRC, which could be made at a personalised level through, identify individuals more at risk. METABOLIC SYNDROME, ALCOHOL INTAKE AND COLORECTAL CANCER: DATA FROM A GREEK CASE CONTROL STUDY Introduction: Several studies support the involvement of Metabolic syndrome in colorectal cancer epidemiology. Alcohol intake seems to have a detrimental effect, despite the fact that moderate alcohol intake has been well documented to exacerbate a protective effect in cardiovascular diseases. The aim of the present study was to evaluate all possible associations of the aforementioned factors in a Mediterranean population. Methods: Between December 2009 and December 2010, 250 consecutive patients with a first diagnosis of CRC (defined by biopsy or histology) were selected from the admission list of Saint Savvas Cancer hospital and Alexandra General hospital. For the same period, 250 subjects (controls) without any clinical symptoms, signs or suspicious of any type of cancer in their medical history, were voluntarily selected from the general population and matched to the patients by age group and sex. The number of the enrolled subjects (n = 500) was decided through power analysis. Adherence to the Mediterranean diet was accessed using MedDietScore; a dietary index, validated for cardiometabolic diseases, that incorporates inherent characteristics of this traditional patter (theoretical range 0-55). Multi-adjusted logistic regression analysis controlling for age, sex, body mass index, physical activity status, smoking habits and family history, was conducted. Metabolic syndrome was defined using the NCEP ATP III criteria. Results: Fully adjusted multiple logistic regression revealed a protective association between moderate alcohol intake (defined as 12-35 g of ethanol per day) (0.38, 95% CI 0.22, 0.65). A 1-point increase in the MedDietScore was associated with 11% lower likelihood of having colorectal cancer (0.89, 95%CI 0.84, 0.94). Metabolic syndrome was positively associated with colorectal cancer (1.66, 95%CI 1.02, 2.69), as well as with alcohol intake ( p < 0.001). Conclusion: Metabolic syndrome is positively, whereas moderate alcohol intake is negatively associated with CRC. However, adherence to a healthy dietary pattern like the Mediterranean has a protective association with CRC independently of the present of other detrimental factors, so it should be promoted as a feasible and economical target for CRC prevention. METHYLATION BASED BIOMARKERS FOR NON-INVASIVE EARLY DETECTION OF COLORECTAL CANCER Introduction: Colorectal cancer is a common disease with a world-wide annual incidence of 1.2 million. The mortality rate is high, especially for patients diagnosed with an advanced tumor stage. We recently identified DNA methylation based biomarkers for early detection of colorectal cancer (and unpublished). In the beginning of 2012, these markers were licensed to the British biotech company Oxford Gene Technology for development into a non-invasive test suitable for colorectal cancer screening ( 659_ogt_signs_licensing_deal_for_colorectal_ cancer_biomarkers). Two tests based on cancer specific aberrant DNA methylation are currently commercially available, including analyses of VIM and SEPT9 in stool and blood samples, respectively. The aim of the present study was to compare the performance of the in house biomarker panel with previously published epigenetic biomarkers using a large series of tissue samples. Methods: Quantitative methylation-specific polymerase chain reaction was used to analyze the promoter methylation status of ITGA4, NTRK2, OSMR, TUBG2, SEPT9, and VIM in test and validation series comprising altogether 169 primary colorectal cancers, 105 matching normal mucosa samples, 104 adenomas, and 107 normal samples from colorectal cancer-free individuals. Receiver operating characteristic (ROC) curves were used to evaluate the overall performance of the biomarkers in tissue samples and the results were compared with the performance of the in house biomarker panel. Results: Across the test and validation series, ITGA4, SEPT9, OSMR, VIM, NTRK2, and TUBG2 were methylated in 91%, 82%, 79%, 67%, 7%, and 1% of the colorectal carcinomas, and in 79%, 88%, 78%, 54%, 3%, and 0% of the adenomas. The promoter methylation was highly cancer-specific and only 1%, 3%, 2%, 3%, 0%, and 0% of the normal mucosa samples were methylated for the same genes. Resulting areas under the ROC curves for separating colorectal carcinomas from normal mucosa samples were 0.95, 0.93, 0.92, and 0.80 for ITGA4, SEPT9, OSMR, and VIM, respectively. In comparison, the in house biomarker panel achieved an area under the ROC curve of 0.97. Conclusion: In a direct comparison performed by a single lab, using the same protocol and the same series of tissue samples, the performance of our recently identified in house biomarker panel was shown to be better than previously identified DNA methylation based biomarkers, including VIM and SEPT9. This underscores the value of combining a manageable number of individually promising biomarkers into a panel, which can achieve more robustness in addition to a high sensitivity and specificity. From this comparison we conclude that the in house biomarker panel is highly suitable for developing into a non-invasive test for early detection of colorectal cancer. COMPARISON OF SCREENING METHODS WITHIN YHE CZECH NATIONAL COLORECTAL CANCER SCREENING PROGRAMME Introduction: Organized colorectal cancer screening programme in the Czech Republic was introduced in year 2000. Since 2009, three screening methods have been used: guaiac and immunochemical fecal occult blood tests (gFOBT and FIT resp.) and primary screening colonoscopy. To asymptomatic individuals aged 50 54, annual gFOBT or FIT is offered, followed by screening colonoscopy, if positive. In age of 55, there is a choice of either gFOBT/FIT biannually or primary screening colonoscopy in 10 years interval. Methods: Accuracy of screening tests - guaiac FOBT and immunochemical FOBT has been analyzed and compared to primary screening colonoscopy. Results: Since the beginning of on-line individual data management in year 2006 there were 77,695 screening colonoscopies performed. Adenomas were diagnosed in 24,558 patients (31.6%) and carcinomas in 3,296 patients (4.2%). Since year 2010, the differentiation between gFOBT and FIT has been recorded. The standardized ratio of positive predictive values (FIT/gFOBT) did not show significant difference between tests (0.97 [0.92 1.03] for adenomas and 0.99 [0.85 1.16] for cancers). Positive likelihood ratio of occult blood tests is 1.6 for adenoma and 4.8 for carcinoma comparing post-test prevalence with PSC prevalence. Conclusion: Testing by FIT provides similar positive predictive value to gFOBT. Pre/ post test comparison of prevalence confirms lower accuracy of gFOBT/FIT for adenomas and underlines the value of primary screening colonoscopy as a cancer prevention test. INFLUENCE OF BRAF V600E AND DIFFERENT KRAS MUTATIONS ON OUTCOME OF METASTATIC COLORECTAL CANCER PATIENTS TREATED WITH FIRST- LINE THERAPY Martina Rebersek1, Marko Boc2, Zvezdana Hlebanja1, Neva Volk3, Jernej Benedik2, Maja Ebert1, Tanja Mesti1 and Janja Ocvirk4 1Institute of Oncology, Ljubljana, 2Institute of Oncology, Ljubljana, 3Institute of Oncology, Ljubljana, 4Institute of Oncology, Ljubljana Introduction: KRAS mutation status in codons 12 and 13 is recognized as a predictive factor for resistance to anti- EGFR monoclonal antibodies. Recent retrospective studies suggested that patients ( pts) with different KRAS mutations (KRAS mts) respond different to specific systemic therapy and also pts with KRAS mt in codon 13 might respond to anti- EGFR monoclonal antibodies but with lower response rates than pts with KRAS wild-type (KRAS wt) tumors. Methods: Objective response (OR), time to progression (TTP), overall survival (OS) and BRAF status in metastatic colorectal cancer (mCRC) pts treated with first- line therapy according to KRAS status, were retrospectively analyzed. Results: In total 176 mCRC pts who received first- line therapy between May 2009 and October 2010 were included in retrospective analysis. The median age was 62 years (27- 86), M/F 61/39 %. The most common received therapies were XELIRI/ bevacizumab and XELOX/cetuximab in 29.5 % and in 22.1 % of pts respectively. KRAS mts were found in 63 pts (35.8 %), mt KRAS in codon 12 in 53 pts (84%) and mt KRAS in codon 13 in 10 pts (16%). BRAF V600E mt was detected in 13 of 176 pts (7.4%). OR in wt KRAS/wt BRAF pts was 54% (CR 14.7 %, PR 39.3 %) and in wt KRAS/mt BRAF pts was 38.5% (CR 15.4 %, PR 23.1 %), difference (diff ) was not statistically significant ( p= 0.378). Median OS of wt KRAS/wt BRAF pts and wt KRAS/mt BRAF pts was 107.4 months (mo) (95% CI: 82- 132.9 mo) and 45 mo (95% CI: 28.4- 61.5 mo) respectively, diff was statistically significant ( p= 0.042). TTP in wt KRAS/ wt BRAF pts and wt KRAS/mt BRAF pts was 16 mo (95% CI: 10.721.2 mo) and 12 mo (95% CI: 4.0 -15.0 mo) respectively, diff was not statistically significant ( p= 0.558). OR in codon 12 mt pts was 47% (CR 20.7%, PR 26.3%) and in codon 13 mt pts was 33% (CR 11%, PR 22%),dif was not statistically significant ( p= 0.08). TTP in mt codon 12 pts and mt codon 13 pts was 13.5 mo (95% CI: 9.018.0 months) and 9.3 mo (95% CI: 5.1- 13.5 mo) respectively, diff was not statistically significant ( p= 0.106). Conclusion: Mt BRAF pts have statistically significantly worse prognosis than wt BRAF pts, with lower response rates and progress earlier during systemic treatment. Codon 13 mt KRAS pts have lower response rate and progress earlier than KRAS codon 12 mt pts. Definitive role of mt BRAF and different KRAS mts are needed to be defined to select pts, who will benefit from specific combination of systemic therapy. KSR1 GENE POLYMORPHISM IN MCRC PATIENTS TREATED WITH FIRST-LINE FOLFIRI AND BEVACIZUMAB Introduction: Kinase suppressor of Ras (KSR) 1 is a scaffolding protein regulating the Raf/Mek/ERK cascade. Preclinical data showed that KSR1 could increase estrogen receptor transcriptional activity after exposure to estrogens. Recent retrospective analyses suggested a possible role for KSR1 rs2241906 C/T polymorphism in predicting the outcome of patients with metastatic colorectal cancer (mCRC). Methods: MCRC patients receiving first-line FOLFIRI + bevacizumab prospectively enrolled in a translational research program were selected on the basis of KRAS and BRAF mutational status availability. KSR1 rs2241906 polymorphism was analyzed on DNA extracted from peripheral blood by means of PCR and direct sequencing. Taking into account the connection between estrogen pathway and KSR1, subgroup analyses according to gender were pre-planned. Results: 287 patients were included. Main patients characteristics were the following: M/F = 175/112; median age = 62 (range 26-79); ECOG-PS 0/1-2 = 240/47; synchronous/metachronous disease = 213/74; Khne score (low/intermediate/high/ data missing) = 122/129/22/14; KRAS-BRAF mutational status (wt-wt/mut-wt/ wt-mut)= 123/146/18. In the overall KRAS-BRAF wt population, KSR1 polymorphism did not affect the outcome. The analysis performed according to gender showed that in the KRAS-BRAF wt population females with KSR1 rs2241906 T/- achieved a significantly better PFS (median 15.9 mos) in comparison to C/C variant carriers (median 8.8 mos) (HR = 0.44 [95%CI 0.21-0.91], p = 0.010); meanwhile males with T/- showed a worse PFS in comparison to those carrying the C/C variant (HR = 1.92 [95%CI 1.05-3.53], p = 0.021). These results were significant also in a multivariate model and a significant interaction of gender with PFS according to KSR1 allelic variants was demonstrated ( p = 0.004). Conclusion: This prospectively conceived pharmacogenetic study confirms the role of KSR1 polymorphisms in affecting the outcome of mCRC KRAS-BRAF wt patients. In particular, these results indirectly indicate that estrogenic stimulation could affect the proliferation of KRAS-BRAF wt CRC cells through an interaction with KSR1. Preclinical investigations to further explore this preliminary hypothesis are ongoing. PHASE II TRIAL OF TARGET-GUIDED PERSONALIZED CHEMOTHERAPY IN FIRST LINE METASTATIC COLORECTAL Antonio Cubillo1, Jess Rodriguez-Pascual1, Fernando Lpez-Ros2, Carlos Plaza Lopez3, Elena Garca4, Rafael lvarez1, De Emilio Vicente5, Yolanda Quijano1, Ovidio Hernando1, Carmen Rubio6 and Manuel Hidalgo7 1Centro Integral Oncolgico Clara Campal, Madrid, 2Therapeutics Targets Laboratory, Madrid, 3Therapeutics Targets Laboratory, Madrid, 4Centro nacional de Investigaciones Oncolgicas, Madrid, 5Centro integral Oncolgico Clara Campal, Madrid, 6Centro Integral Oncolgico Clara Campal, Madrid, 7Centro Nacional de Investigaciones Oncolgicas, Madrid Introduction: Chemotherapy (Ch) options for patients ( pts) with colorectal cancer (CRC) have increased in the last years. However, there are no validated prospective molecular markers in CRC to select which agents are better to treat any individual case. The aim of this study was to determine the efficacy in terms of progression free survival (PFS) of a biomarker panel to guide treatment selection in this setting. Methods: Treatment naive, ECOG 0-1, metastatic CRC pts were accrued. Pts were prospectively analyzed with a predefined set of 10 molecular targets, including: KRAS, BRAF, and PI3K mutations and Topoisomerase-1(Top-1), ERCC-1, Thymidylate synthase (TS) and Thymidine phosphorylase (TP) expression by immunohistochemistry (IHC) performed in a tumor biopsy. Ch combination schema plus Cetuximab (C) or Bevacizumab (B) at standard doses was customized based on: Topo-1 +: Irinotecan (I). Topo-1- and ERCC-1 -: Oxaliplatin(O). Topo-1- and ERCC1 +: investigator option (I or O). TS -: Fluoropyrimidines (FP).TS +: No FP. TP -: 5-FU, TP +: Capecitabine. Maintenance C or B treatment was allowed. Primary outcome measure was PFS. Results: 74 pts were accrued and all of them received biomarker guide treatment. All of them began personalized. Interim analysis on 61 pts (38 males, median age 65) showed Topo-1 + in 33 pts (54%), ERCC-1- in 36 (59%) TS + in 44 (73%), TP in 61 (100%), K-ras native in 34 (55%), BRaf mutated in 2 (3.2%), with a median follow up time of 9.1 months (m). Median PFS (95% CI) is 8.6 (6.2-10.9) m, with a 41.3% (27.4-55.2) 12 m PFS. Overall clinical benefit (Response + Stabilizations) was 74.5% (65.6-83.4).Toxicities Grade 3 included 18% neutropenia, 4.9% asthenia and 3.3% anemia. 12 pts (23%) received loco-regional treatment (surgery or radiosurgery). Median Overall survival has not been reached. Conclusion: Target-Guided Personalized Ch in first line CRC pts is feasible and results in promising PFS with low toxicity. Update of results and more detailed data will be presented. PHASE I/II STUDY OF TIVANTINIB (ARQ 197) IRINOTECAN, AND CETUXIMAB IN PATIENTS WITH KRAS WILDTYPE, PREVIOUSLY TREATED, METASTATIC COLORECTAL CANCER Introduction: Combination therapy with cetuximab, an epidermal growth factor receptor (EGFR) inhibitor, plus irinotecan is a standard treatment for previously treated, KRAS wildtype, metastatic colorectal cancer (mCRC); however, tumors generally develop resistance to cetuximab. The MET receptor tyrosine kinase has been implicated in tumor cell migration, invasion, proliferation, and angiogenesis, iv | poster discussions and is associated with resistance to EGFR inhibitors. Tivantinib is an oral, selective MET inhibitor currently in phase II and phase III trials in patients with non-small cell lung cancer and hepatocellular carcinoma. The current phase I/II study assesses the activity of tivantinib combined with irinotecan plus cetuximab in previously treated patients with mCRC. Methods: Patients were required to have KRAS wildtype mCRC, ECOG performance status 1, and 1 previous lines of chemotherapy. Patient cohorts (n = 3) received escalating doses of tivantinib (120, 240, or 360 mg) twice daily (BID) and biweekly irinotecan (180 mg/m2) and cetuximab (500 mg/m2) in 28-day cycles. Patient plasma and tumor tissue samples were collected for pharmacokinetic and biomarker analyses. MET protein expression in tumor tissue was assessed via immunohistochemistry (IHC), with MET-positive samples defined as MET expression 2+ in 50% of tumor cells. Plasma biomarker levels were determined by ELISA. Radiographic response was assessed every 8 weeks by RECIST version 1.1. Results: Nine patients (median age, 53 years; range, 30-76 years) with ECOG performance status 0 (n = 4) or 1 (n = 5) received treatment. Median number of previous therapies was 2 (range, 1-4), including 1 patient with previous cetuximab treatment (a protocol violation). No dose-limiting toxicities were observed during the first treatment cycle. Tivantinib 360 mg BID was the recommended phase II dose (RP2D). Median treatment duration was 7.3 months (range, 1.6-15.1 months), and all 9 patients have discontinued treatment. Reasons for discontinuation included disease progression (n = 5), adverse events (AEs; n = 2), withdrawal of consent (n = 1), and investigator decision (n = 1). Reported grade 3 and 4 AEs were grade 4 neutropenia (n = 1), grade 3 fatigue (n = 3), and grade 3 anemia, leukopenia, diarrhea, small intestine obstruction, vomiting, mucositis, pneumonia, syncope, acneiform rash, and skin fissure (n = 1 each). The objective response rate was 44%. Best response was complete response (n = 1), partial response (n = 3), or stable disease (n = 4). Archival tumor samples were available for 7 patients, 4 of which were MET-positive by IHC. In the 4 MET-positive patients best response was partial response (n = 2) and stable disease (n = 1). Mean plasma levels of MET and hepatocyte growth factor were unchanged and mean plasma VEGF levels decreased 40% in 5 patients with available samples. Conclusion: The combination of tivantinib and irinotecan plus cetuximab was well tolerated and had antitumor activity in heavily pretreated patients with mCRC. The RP2D for tivantinib is 360 mg BID. The randomized, placebo-controlled phase II trial has completed enrollment (n = 122) and final results will be presented at a later date. Correlative studies are ongoing. BENEFIT OF PANITUMUMAB (PMAB) AS MONOTHERAPY IN PATIENTS WITH WT KRAS MCRC: A NUMBER-NEEDED-TO-TREAT (NNT) ANALYSIS PD-0019 Table 1 Introduction: In a phase III study (408), pmab monotherapy was shown to significantly improve progression-free survival (PFS) in patients with WT KRAS mCRC. The number needed to treat (NNT) is a widely used method of expressing the benefit from a particular intervention. The NNT is time-specific and quantifies the benefit of a treatment by describing how many patients need to be treated to prevent one undesirable event (e.g., disease progression) or achieve one beneficial outcome (e.g., tumour response). Here, we report the NNTs for PFS and objective response rate (ORR) in patients with WT KRAS tumours receiving pmab as monotherapy compared with best supportive care. Methods: The median duration of pmab treatment in 408 study was 2.8 months in the active treatment arm, and cross-over to pmab was allowed in the control arm after disease progression. Therefore, we calculated the number of patients needed to be treated with pmab monotherapy: to (A) prevent one tumour progression according to RECIST or death at 3 and 6 months; and (B) achieve either complete or partial tumour response according to RECIST. Hazard ratios (HR) were generated. Results: NNTs, HRs and median PFS are summarized in the table. Conclusion: In this analysis, NNTs for PFS and ORR were low reflecting a high clinical activity from pmab monotherapy in patients with WT KRAS mCRC. Stratification by skin toxicity grade 2 or ECOG status 0-1 further lowered the NNTs for PFS at both time points. PHARMACOKINETIC ANALYSIS OF PERCUTANEOUS HEPATIC PERFUSION (PHP) OF MELPHALAN IN PATIENTS WITH HEPATIC METASTASES FROM MELANOMA Erin Gardner1, William Figg2, Hughes Marybeth2 and James Pingpank, Jr.2 1Analytical Pharmacology Consultant, Vienna, Virginia, 2National Cancer Institute, Bethesda, Maryland Introduction: PHP (Delcath Systems Inc., New York, NY) is a non-invasive regional therapy that isolates the liver using a series of catheters, allowing infusion of high doses of chemotherapy directly into the hepatic artery. Systemic toxicity is minimized by extracorporeal hemofiltration of effluent hepatic venous blood, which is then returned to the systemic circulation. A randomized phase III study was conducted to compare PHP of high-dose melphalan with best alternative care in patients with ocular or cutaneous melanoma metastatic to the liver. A pharmacokinetic analysis, including an evaluation of filter extraction efficiency, was performed in a subset of patients from this study. Methods: Melphalan was perfused into the liver over 30 minutes via PHP at a maximum dose of 3.0 mg/kg of ideal body weight (maximum 220 mg/treatment). Extracorporeal hemofiltration was performed during and for 30 minutes after melphalan infusion. PHP was performed with patients under general anesthesia. Blood samples (7 mL) were collected during the first cycle of PHP melphalan simultaneously from the periphery (arterial line) and extracorporeal circuit (both pre- and post-filter). Sample collections were made at: baseline ( peripheral only); 15 minutes after the start of the infusion; immediately post-infusion; 5, 10, 15, and 30 minutes post-infusion. Plasma concentrations of melphalan were determined by PD-0020 Figure: Individual AUC by site, alter 3 0 mg/Kg (ideal body weight) melphalan Horizontal bars represent mean and 95% Cls high-pressure liquid chromatography with ultraviolet detection. Data were analyzed using a non-compartmental approach with WinNonlin v5.2 (Pharsight Corporation, Mountain View, CA). Concentration-time profiles were constructed for each sampling location (i.e. three profiles/patient). Pharmacokinetic parameters were: maximum plasma concentration (Cmax); area under the concentration-time curve from time zero to final sample (AUClast) calculated using the linear trapezoidal method; and filter efficiency [defined as: ( pre-filter AUClast) minus ( post-filter AUClast) divided by ( pre-filter AUClast)]. Results: Plasma samples were available from 48 patients, of which 40 were evaluable. Mean absolute melphalan dose was 191 mg (range, 137220 mg) and duration of perfusion was 30 minutes (range, 1652 minutes). Mean prefilter, postfilter and systemic Cmax values were 8726 ng/mL, 2330 ng/mL and 1429 ng/mL, respectively. Mean prefilter AUClast was 264,652 min*ng/mL, mean postfilter AUClast was 74,146 min*ng/mL and mean systemic AUClast was 50,777 min*ng/mL; individual patient data for AUClast are shown in the Figure. Mean filter extraction efficiency was 71.2% 10.4%. Filter efficiency did not change significantly with absolute melphalan dose ( p = 0.86, Spearman) or theoretical rate of perfusion ( p = 0.064, Spearman). Conclusion: The mean filter extraction efficiency of the PHP filtration system is 71%, and results appear to be consistent across patients (narrow 95% CI intervals). These findings indicate that the filter removes most of the melphalan administered via PHP. HAND-FOOT-SYNDROME (HFS) IS A CLINICAL PREDICTOR OF SURVIVAL IN CAPECITABINE TREATED PATIENTS (PTS) WITH COLORECTAL CANCER Introduction: HFS is a typical adverse event in capecitabine treated patients ( pts). Post hoc analyses of the AIO KRK-0104 trial showed that HFS is associated with improved prognosis in capecitabine recipients in stage IV colorectal cancer. Here we evaluated the potential association of HFS and survival in a pooled analysis of pts with metastatic colorectal cancer and locally advanced rectal cancer (Mannheim rectal cancer trial). Methods: Pts receiving capecitabine within the AIO KRK-0104 and Mannheim rectal cancer trials were analyzed. HFS was graded according to NCI-CTCAE. Occurrence and kinetics of occurrence per cycle were investigated defining early HFS as developing of any HFS during the fist two cycles. Groups (HFS versus non-HFS) were compared using Fishers exact test. Survival was compared using Kaplan-Meier estimation and logrank analysis. Overall survival (OS) and progression-free (PFS) or disease-free survival (DFS) were pooled. Results: Of 374 pts a total of 29.3% developed HFS. The frequency of early HFS was 70.9%. Baseline characteristics were comparable between the HFS and the non-HFS groups. Multivariate analysis for age, gender, ECOG status and UICC stage on the occurrence of HFS was done but none of these factors had influence. The percentage of all-grade haematological toxicities did not differ between both groups. We observed in the HFS group a significantly higher rate of all grade (but not grade 3-4) diarrhoea, stomatitis/mucositis and fatigue ( p < 0.01 resp.). Moreover, pts in the HFS group had improved PFS/DFS (29.0 vs. 11.4 months; p = 0.015, HR 0.69) and OS (75.8 vs. 41.0 months; p = 0.001, HR = 0.56). No difference in PFS/DFS and OS could be stated between the early HFS and late HFS pts groups. Conclusion: HFS as a sign of individual capecitabine metabolism is a clinical predictor of survival in colorectal cancer patients. Pts developing HFS had a higher probability of developing any-grade gastrointestinal toxicity and fatigue while no correlation with haematological toxicity was noticed. AFLIBERCEPT VERSUS PLACEBO IN COMBINATION WITH FOLFIRI IN PREVIOUSLY TREATED METASTATIC COLORECTAL CANCER: MEAN OVERALL SURVIVAL FOR SUBGROUPS FROM VELOUR Introduction: The VELOUR phase III trial evaluated the efficacy and safety of the novel fusion protein aflibercept (VEGF Trap) in combination with FOLFIRI in metastatic colorectal cancer (mCRC) patients previously treated with oxaliplatin. Median overall survival (OS) showed significant improvement with 13.50 months in the aflibercept arm vs. 12.06 months in the placebo arm ( p = 0.0032; HR = 0.82 [95.34% CI: 0.71 to 0.94]). Treatments were allocated through a 1:1 randomization stratified by ECOG PS (0, 1 or 2) and prior bevacizumab use. The purpose of this analysis was to estimate mean OS in ITT population and stratified subgroups of VELOUR for populating cost-effectiveness analysis. Methods: Because some patients were alive at the time of the final analysis, mean OS must be estimated by extrapolating the trial Kaplan-Meier curve using a survival function. Five standard parametric distributions were tested: exponential, Weibull, lognormal, log-logistic and Gompertz. Akaikes Information Criteria (AIC), Bayesian Information Criteria (BIC) and graphical methods were used to evaluate the goodness of fit of the distributions. Survival functions were fit to each treatment arm separately or by combining the two arms and using treatment as covariate to control for variation. Mean OS was estimated for the ITT population and stratification subgroups accordingly. The subgroup of patients with PS = 2 is not presented as it represents only 27 patients (2.2% of the 1226 ITT patients). Results: The log-logistic function provided the best fit of the VELOUR data for the ITT population and for all subgroups except in the placebo arm for the PS = 0 subgroup and for the prior bevacizumab subgroup where the lognormal and Weibull distributions respectively fit best. Using the best fits, mean survival was greater in the aflibercept arm compared to placebo arm in all subgroups, by an amount ranging from 1.4 to 9.8 months. Conclusion: These additional analyses of mean OS in the ITT population and stratification subgroups support the primary VELOUR analyses demonstrating a survival benefit across patient subgroups of Aflibercept/FOLFIRI over Placebo/ FOLFIRI. These results have important implications in terms of informing consistent clinical and economic benefit of aflibercept across general, as well as clinically relevant patient subpopulations in second-line mCRC. Study NCT00561470 was funded by sanofi, in partnership with Regeneron. BENEFIT OF ADDING PANITUMUMAB (PMAB) TO 1ST/ 2ND-LINE CHEMOTHERAPY IN PATIENTS WITH KRAS WT MCRC: A NUMBER-NEEDED-TO-TREAT (NNT) ANALYSIS Introduction: Data from two randomized, multi-center studies have shown pmab plus chemotherapy (FOLFOX4 in 1st-line therapy [PRIME study] and FOLFIRI in iv | poster discussions PD-0022 Table 1 2nd-line therapy [181 study]) significantly improved progression-free survival (PFS) in patients with WT KRAS mCRC. NNT is a widely used method of expressing the benefit from a particular intervention. The NNT is time-specific and quantifies the benefit of a treatment by describing how many patients need to be treated to prevent one undesirable event (eg, disease progression) or achieve one beneficial outcome (eg, tumor response). Here, we retrospectively calculate the NNT for PFS and objective response rate (ORR) in patients with WT KRAS mCRC receiving pmab plus chemotherapy compared with chemotherapy alone for each study. Methods: Using data from a prespecified final analysis scheduled to occur 30 months after the last patient was enrolled in PRIME and 181, we calculated the number of patients needed to be treated with pmab plus chemotherapy vs chemotherapy alone to: (A) to prevent one tumour progression according to RECIST or death at 6 and 12 months; and (B) achieve either complete or partial tumour response according to RECIST. Hazard ratios were generated. Results: Median pmab duration was 6.4 months in PRIME, and 5.5 months in study181. NNTs are summarized in the table. Conclusion: In this analysis, NNTs for PFS and, particularly, ORR are low reflecting the high therapeutic benefit of the addition of pmab to chemotherapy in patients with WT KRAS mCRC. Stratification by skin toxicity grade 2 or ECOG status 0-1 further lowered the NNT for PFS at both time points. PHASE I/II STUDY OF FOLFIRI PLUS THE MEK1/2 INHIBITOR PIMASERTIB (MSC1936369B) AS SECOND-LINE TREATMENT FOR KRAS MUTATED METASTATIC COLORECTAL CANCER Introduction: Pimasertib is a highly selective inhibitor of the MEK1/2 kinases of the MAPK pathway. It demonstrates potent antitumor activity in cell lines and xenograft models with activating (mainly BRAF and KRAS) mutations. A two-part study, comprising a safety run-in part followed by a randomized phase II part, was designed to investigate FOLFIRI plus pimasertib as second-line treatment for patients with KRAS mutated (mt) metastatic colorectal cancer (mCRC). The results of the safety run-in part, conducted primarily to determine the maximum tolerated dose (MTD) and the recommended phase II dose (RP2D), are reported here. Methods: Patients with KRAS mt mCRC progressing on first-line oxaliplatin plus fluoropyrimidine bevacizumab were eligible. A 3 + 3 design, with fixed dose-escalation based on dose-limiting toxicities (DLTs) occurring during the first cycle (28 days) of treatment, was used to investigate FOLFIRI plus daily oral pimasertib (dosed 5 days on 2 days off, starting dose 45 mg/day). The decision to proceed to the next dose level was made by a safety monitoring committee, comprising investigators and representatives of the sponsor. Results: Sixteen atients (median age 64 years [range 38-81] and ECOG performance status 0-1) received FOLFIRI plus pimasertib 45 mg/day (n = 10) or 60 mg/day (n = 6) (safety population). Two patients were not evaluable for the dose escalation and DLT analysis (n = 14). No DLTs were observed in the first 3 patients treated at 45 mg/day, allowing dose escalation to 60 mg/day. On pimasertib 60 mg/day, 2/5 patients had DLTs, both grade 3 mucositis/stomatitis. This led to a per-protocol expansion of the 45 mg/day cohort, during which 1/6 patients had a DLT (grade 3 hyponatremia). The median number of pimasertib cycles initiated was 3 (1-11). Over both doses, the most common treatment-emergent adverse events (TEAEs) were asthenia (44%), diarrhea (44%), mucosal inflammation (38%), ocular events (38%, mainly grade 1/2 macular degeneration) and nausea, rash and vomiting (31% each). Seven patients had at least one grade 3 treatment-related TEAE: diarrhea (2 patients), mucosal inflammation (3 patients) and neutropenia (3 patients). Eight patients had 1 serious TEAE: 4 on pimasertib 45 mg/day and 4 on pimasertib 60 mg/day. Five patients had 1 TEAE leading to permanent treatment discontinuation, 3 on pimasertib 45 mg/day and 2 on 60 mg/day. Over both doses, 9 patients had on-treatment electrocardiogram changes compared with baseline readings. Reasons for stopping treatment included AE (3 patients), disease progression (6 patients) and consent withdrawal (3 patients). As of 23 November 2011, 3 patients remained on treatment with FOLFIRI plus pimasertib. Pimasertib dosed once-daily concomitantly with FOLFIRI exhibited a pharmacokinetic profile comparable to pimasertib monotherapy. Conclusion: In combination with FOLFIRI, dose escalation of pimasertib from 45 mg/day to 60 mg/day was limited by toxicity. Due to the fact that 45 mg/day is not considered to be an active dose in this setting, progression to the phase II part of the study was not recommended by the sponsor. EVALUATION OF CODON 12 AND 13 KRAS MUTATIONS AS BIOMARKERS OF RESPONSE TO PANITUMUMAB IN PATIENTS WITH METASTATIC COLORECTAL CANCER Introduction: Panitumumab, a fully human monoclonal antibody directed against the epidermal growth factor receptor (EGFR), has demonstrated significant improvement in progression-free survival (PFS) in patients with wild-type (WT) KRAS metastatic colorectal cancer (mCRC) in three phase 3 trials: the first-line study 20050203 ( panitumumab + FOLFOX4), the second-line study 20050181 ( panitumumab + FOLFIRI), and the panitumumab monotherapy study 20020408. Mutations in codons 12 and 13 of KRAS are recognized biomarkers that predict lack of response to anti-EGFR antibody therapies. This retrospective analysis assessed the prognostic and predictive impact of individual codon 12 and 13 KRAS mutations on PFS, overall survival (OS), and response rate across three-lines of therapy. Methods: Patients were randomized 1:1 to receive FOLFOX4, FOLFIRI, or best supportive care with or without panitumumab 6.0 mg/kg every two weeks. The primary endpoint in studies 20050203 and 20020408 was PFS; OS was a secondary endpoint. In study 20050181, the co-primary endpoints were PFS and OS. Mutant (MT) KRAS codon 12 and 13 status was ascertained with the Therascreen K-RAS Mutation Kit (Qiagen) that identifies the seven most common KRAS mutations in codons 12 and 13 (KRAS G12A, G12C, G12D, G12R, G12S, G12V, G13D). PD-0023 Table 1 Results: KRAS ascertainment rates were 93% (study 20050203), 91% (study 20050181), and 92% (study 20050408). A total of 1052 patients with MT KRAS codon 12 and 13 alleles were included in the analysis: 440 patients from 20050203, 486 patients from 20050181, and 126 patients from 20020408. The frequency of MT KRAS alleles was conserved between studies and baseline demographic and clinical features were generally balanced. There were no consistent results that any single MT KRAS allele, compared with the other MT KRAS alleles or the entire MT KRAS group, differentially affected PFS or OS in patients treated in the control or panitumumab-containing arms. Only two individual MT KRAS alleles were significantly associated with outcomes by interaction testing: in the panitumumab + FOLFOX4 arm of study 20050203 G12V was favorably and G13D was unfavorably associated with OS. Response rates between MT KRAS allele groups were similar in the first-line 20050203 and second-line 20050181 studies. In the panitumumab monotherapy setting (20020408), no patients with MT KRAS codon 12 or 13 mCRC tumors responded. Finally, pooled analysis of all 3 trials suggested that only the G12A KRAS allele was significantly associated with a negative treatment effect on OS. Conclusion: The results from this retrospective analysis of three phase 3 trials indicate that patients with MT KRAS codon 12 or 13 mCRC tumors are unlikely to benefit from panitumumab therapy. Therefore, only patients with WT KRAS tumors should receive panitumumab therapy. QUALITY OF LIFE IN PATIENTS WITH ADVANCED GASTRIC CANCER ENROLLED IN THE INTERNATIONAL, PHASE 3 GRANITE-1 STUDY Jaffer Ajani1, Jin Li2, Taroh Satoh3, Tomohiro Nishina4, A. Alarcn-Rozas5, Junji Furuse6, Wei Liu7, Min-Hee Ryu8, Was Mansoor9, Todd Roma10, Heind Smith10, Jocelynn Booth10, Michaela Sedova11, Shriya Bhushan10, Tarek Sahmoud10, Syed Rizvi12 and Yung-Jue Bang13 Introduction: Advanced gastric cancer is an aggressive, difficult-to-treat disease associated with poor outcomes. Treatment options for patients who progress after first- or second-line chemotherapy are limited, particularly as many patients have poor performance status (PS) and are not eligible for additional chemotherapy. Advanced gastric cancer treatment options that improve outcomes without significantly decreasing quality of life (QoL) are needed. Methods: In the randomized, double-blind, phase 3 GRANITE-1 study, patients aged 18 years with confirmed advanced gastric cancer, disease progression after 1 or 2 lines of previous systemic chemotherapy, and an Eastern Cooperative Oncology Group (ECOG) PS 2 were randomized 2:1 to oral everolimus 10 mg/ day plus best supportive care (BSC) or placebo plus BSC. Randomization was stratified by region (Asia [China, Japan, Korea, Taiwan, Thailand, and Hong Kong] vs rest of the world) and by the number of previous lines of systemic chemotherapy (1 vs 2). Study drug was continued until disease progression or unacceptable toxicity. ECOG PS was assessed at baseline, every 2 weeks for the first 4 weeks, and every 4 weeks thereafter until the end of treatment. The European Organization for the Research and Treatment of Cancer (EORTC) QLQ-C30 QoL questionnaire and gastric-specific STO22 module were completed at the start of treatment and every 4 weeks thereafter until disease progression. The primary endpoint was overall survival (OS). Secondary endpoints included the time to definitive deterioration of ECOG PS, defined as a decrease of 1 category and no subsequent improvement, and the time to a definitive 5% deterioration in the global health status/QoL and physical, social, and emotional functioning scales of the QLQ-C30 questionnaire. Results: A total of 656 patients (median age, 62.0 years; 73.6% male; 91.9% with ECOG PS 1) were enrolled in the study and received everolimus (n = 439) or placebo (n = 217). Median OS was 5.39 months with everolimus and 4.34 with placebo (HR for OS, 0.90; 95% CI, 0.75-1.08; P = 0.1244). The median time to definitive deterioration in ECOG PS by 1 category did not differ between treatment groups (2.30 months with everolimus vs 2.23 months with placebo; HR, 0.96; 95% CI, 0.76-1.20; P = 0.6925). A trend for slightly longer time to definitive 5% deterioration in global QoL was observed with everolimus vs placebo (median, 1.51 months vs 1.45 months; HR for definitive 5% deterioration from baseline, 0.84; 95% CI, 0.69-1.03; P = 0.0936). Trends for slightly longer time to definitive 5% deterioration in the social (HR, 0.87; 95% CI, 0.70-1.09; P = 0.2108) and emotional (HR, 0.83; 95% CI, 0.67-1.02; P = 0.0735) functioning scales were also observed; no difference between treatment groups was detected for physical functioning (HR, 0.95; 95% CI, 0.78-1.15; P = 0.5711). Over time, everolimus recipients had higher mean scores for the emotional functioning subscale of the QLQ-C30 and lower mean scores (i.e., lower level of symptoms) for the pain and reflux subscales of the STO22 module. Conclusion: Everolimus did not cause significant deterioration in ECOG PS or QoL in patients with previously treated advanced gastric cancer. iv | poster discussions FEASIBILITY OF PERIOPERATIVE CHEMOTHERAPY WITH INFUSIONAL 5-FU, LEUCOVORIN AND OXALIPLATIN +/- DOCETAXEL, IN ELDERLY PATIENTS WITH LOCALLY ADVANCED ESOPHAGOGASTRIC CANCER Introduction: The aim of this study was to determine feasibility and efficacy of perioperative chemotherapy in elderly, potentially operable esophagogastric cancer patients. Methods: Patients aged 65 or older with locally advanced esophagogastric adenocarcinoma were randomized to perioperative chemotherapy consisting of four preoperative and four postoperative cycles of infusional 5-FU, leucovorin, and oxaliplatin (FLO) with or without docetaxel 50mg/m2 (FLOT), every 2 weeks. Results: 44 patients with a median age of 70 years were randomized to either FLO (22) or FLOT (22) chemotherapy. In the FLO and FLOT group, 20 and 18 patients completed four cycles of preoperative chemotherapy and 17 (77.3%) and 15 (68.2%) patients proceeded to surgery, with 88.2% and 93.3% R0 resections, respectively. FLOT was associated with significantly more treatment-related NCI-CTC grade 3/4 adverse events, including neutropenia (P < 0.0001), leukopenia (P < 0.0001), stomatitis (P = 0.02) and nausea (P = 0.02) and a higher rate of patients experiencing a 10-points deterioration of EORTC QoL global health status scores (FLOT, 54%; FLO 18%) at 8 weeks. No perioperative mortality was observed, however postoperative morbidity was nearly doubled with the FLOT regimen (60% versus 35.3% for FLO), mostly due to wound infections in the FLOT group (20%). 11 (64.7%) and 9 (60.0%) of patients in the FLO/FLOT group were able to undergo postoperative chemotherapy. Compared with the FLO group, the FLOT group had a better response rate (61.9% vs. 18.2%) and a trend towards an improved median disease free survival (21.1 vs. 12.0 months; p = 0.09). Conclusion: Age alone is not a contra-indication for the selection of patients for three-drug regimens in the multimodal treatment approach. FLOT seems to be more effective than FLO, but is associated with an increased toxicity and higher postoperative morbidity. Therefore, careful patient selection is warranted. PREDICTION OF INTRACTABLE ASCITES AND PROGNOSIS AFTER SURGERY IN GASTRIC CANCER PATIENTS WITH LIVER CIRRHOSIS Jenny Hong National Cancer Institute, National Institutes of Health, Bethesda, MD Introduction: To determine the rate of occurrence of intractable ascites in patients with gastric cancer and cirrhosis; to examine which factors of the MELD score and Child class are associated with the occurrence of intractable ascites and whether these factors may be used as markers to determine the extent of lymph node dissection for gastric cancer patients with liver cirrhosis; and to determine the effect of intractable ascites on prognosis. Methods: This retrospective analysis was completed by surveying 160 patients diagnosed with cirrhosis before or during surgery among 9100 patients who had undergone radical gastrectomy for gastric cancer between October 1994 and December 2005 at Samsung Medical Center. Results: Intractable ascites occurred in 9 of 160 patients. There was a significant difference in the extent of lymph node dissection (D1/D2; p = 0.04) and Child class ( p = 0.04) with respect to the occurrence of intractable ascites. In Child class B, intractable ascites occurred in none of the 12 patients undergoing D1 dissection and in 5 of 12 patients undergoing D2 ( p = 0.03). The three-year survival rate of all patients was 71%. The occurrence of intractable ascites significantly reduced the three-year survival rate (without 73% vs. with 33% ( p < 0.01). Cox regression analysis identified serum albumin as a statistically significant factor affecting survival ( p = 0.03). Conclusion: With respect to gastrectomy in patients with liver cirrhosis, D2 dissection may be performed on patients with Child class A. However, patients with Child class B who underwent D1 dissection had a decreased occurrence of intractable ascites with an improved prognosis. The MELD score is not useful in determining the extent of lymph node dissection. TWO DISTINCT SUBTYPES OF GASTRIC ADENOCARCINOMA ASSOCIATED WITH PROGNOSIS AND SENSITIVITY TO THERAPY Ju-Seog Lee1 and Sang Cheul Oh2 1UT MD Anderson Cancer Center, Houston, TX, 2Korean University, Seoul Introduction: Heterogeneity of gastric adenocarcinoma is reflected in the unpredictable outcomes when patients with similar stage of cancer are treated with empiric approaches. Thus, molecular subtypes and their associated biomarkers need to be established to optimize therapy of gastric adenocarcinoma. We aim to uncover subgroups of gastric adenocarcinoma that have distinct biological characteristics associated with clinical outcome and identify gene-expression signatures that best reflect the biological and clinical characteristics of each subgroup. Heterogeneity of gastric adenocarcinoma is reflected in the unpredictable outcomes when patients with similar stage of cancer are treated with empiric approaches. Thus, molecular subtypes and their associated biomarkers need to be established to optimize therapy of gastric adenocarcinoma. We aim to uncover subgroups of gastric adenocarcinoma that have distinct biological characteristics associated with clinical outcome and identify gene-expression signatures that best reflect the biological and clinical characteristics of each subgroup. Methods: We analyzed gene expression profiling data from 93 patients with gastric adenocarcinoma to uncover subtypes and identify a gene expression signature associated with prognosis and response to adjuvant chemotherapy. The association of the signature with prognosis was validated in four independent cohorts of 646 patients. Results: Gene expression profiling identified two subtypes of gastric adenocarcinoma, S and L, in exploration cohort (n = 93) and the prognosis of the S subtype was significantly worse than that of the L subtype (P =.04 for recurrence-free survival [RFS]). The 299-gene signature, associated with the S subtype, was a highly significant predictor of both overall survival (OS) and RFS in four independent test cohorts. In multivariate analysis, the signature was the most significant risk factor among clinical variables examined together (hazard ratio, 2.27; 95% confidence interval, 1.38-3.73; P =.001 for OS). The signature also predicted lack of benefit from adjuvant chemotherapy. Conclusion: This novel gene expression signature was highly associated with benefit from adjuvant chemotherapy and prognosticated outcome of gastric adenocarcinoma patients who underwent gastrectomy. 2-YEAR FOLLOW-UP OF A PHASE II STUDY ON CATUMAXOMAB AS PART OF A MULTIMODAL APPROACH IN PRIMARILY RESECTABLE GASTRIC CANCER Introduction: Perioperative chemotherapy (CT) has demonstrated as survival benefit in locally advanced gastric cancer (GC) in randomized trials. However, the overall cure rate is 30-40% and a significant number of patients are not able to receive the postoperative part of their CT regimen. In Europe, the trifunctional antibody catumaxomab is approved for the treatment of malignant ascites based on a pivotal trial which also included GC patients. A new multimodal approach combining neoadjuvant CT, followed by gastrectomy and intraperitoneal (i.p.) immunotherapy with catumaxomab was assessed in a single-arm multicenter phase II study. We here report 2-year follow-up data. Methods: GC pts (T2/T3/T4, N + /, M0) received 3 cycles of neoadjuvant fluoropyrimidine/platinum-based CT followed by en-bloc R0-gastrectomy. Catumaxomab was administered i.p. as intraoperative bolus (10 g) followed by 4 consecutive 3-hour infusions of 10-150 g. Primary safety endpoint was the rate of predefined postoperative complications observed during 30 days after surgery. Key efficacy endpoints included disease-free (DFS) and overall survival (OS). Results: The original study data presented at the WCGC in 2011 showed that the primary endpoint was met and the described application regimen is safe. At time of surgery, 27.8% of patients were stage I, 27.8% of patients were stage II, 22.3% of patients were stage III and 14.8% of patients were stage IV as assessed according to pTNM measures. At 24 months 39/54 (safety analysis set) patients were still alife, 14/ 54 were dead, (one patient lost to follow-up), 24/37 had no progression, only 11/37 patients relapsed (for 2 patients disease status was not recorded). At the 2 year cut off DFS was 56.4% (95% CI: 4169%), OS was 75% (95% CI: 6085%). Conclusion: Catumaxomab as part of a multimodal therapy in primarily resectable GC is a feasible option. The 2-year follow up efficacy results show promising data for DFS and OS in a cohort of locally advanced gastric cancer pts. PREOPERATIVE CHEMORADIATION FOR RESECTABLE ADENOCARCINOMA OF THE PANCREAS (ISRCTN 78805636): PATTERN OF RECURRENCE. Henriette Golcher1, Helmut Witzigmann2, Lukas Marti3, Jochen Lange3, Wolf Bechstein4, Christiane Bruns5, Henry Jungnickel2, Johann Hauss6, Stefan Schreiber7, Thomas Brunner8, Gerhard Grabenbauer9, Susanne Merkel10, Rainer Fietkau10 and Werner Hohenberger10 Introduction: Standard treatment for resectable pancreatic carcinoma is primary surgery, and most patients die of recurrent cancer. Chemoradiation is used to treat locally unresectable pancreatic cancer which is meant to stop local progress at least for a while. Neoadjuvant chemoradiation was introduced prior to surgery in locally resectable cancer in this trial. The local effect of chemoradiation and its influence on resection rates and recurrence patterns after tumor resection is examined now. Methods: Patients with histologically proven ductal adenocarcinoma of the pancreatic head encircling peripancreatic vessels for <180 were randomized into two groups: Group A - primary surgery, Group B - preoperative chemoradiation (55.4Gy; gemcitabine/cisplatin), followed by surgery 6 weeks later. In 2005 a protocol amendment suggested adjuvant chemotherapy (gemcitabine). All patients were followed up regularly by CT-scan. Results: Between 01.06.2003 and 31.12.2009, 73 patients were randomised in 8 centres (A n = 37, B n = 36). Due to slow recruitment the trial was closed early ( planned n = 254). By 09.03.2012 data of 68 patients (A/B n = 34) could be evaluated, 3 patients withdrew consent (A/B n = 2/1), 1 centre did not sent data (A/B n = 1). Tumor resection was performed in 24 patients (A) and in 20 patients (B) by intention-to-treat-analysis. Explorative laparotomy revealed locally unresectable tumor in 6 patients (A/B n = 4/2) and distant metastasis in 13 patients (A/B n = 5/8). By now 20 patients (A) and 16 patients (B) had recurrence of cancer: 3 patients (A) had residual tumor and/or residual filiae and died within 7 months, 1 patient (B) had residual tumor and died 5 months after randomization. 6 patients (A/B n = 5/1) had local recurrence first, whereas 15 patients (A/B n = 7/8) had distant metastases as primary recurrence localization. 9 patients (A/B n = 4/5) had local recurrence and distant metastases diagnosed simultaneously. The localization of cancer recurrence was not known in 2 patients (A/B n = 1/1). There was no significant difference in median time from randomization until tumor progression with 10 months in group A and 9 months in group B ( p = 0.4) after tumor resection. Even though median time to local recurrence was delayed after chemoradiation with 7 months in group A and 10 months in group B. Conclusion: In this trial the pattern of local and distant recurrence of pancreatic cancer after tumor resection was not significantly different with and without preoperative chemoradiation treatment. But this might be due to low patient numbers because of early stopping the trial for slow recruitment. iv | poster discussions

This is a preview of a remote PDF:

Panagiotis Heras, Vasiliki Niarou, Evaggelos Andrikopoulos, Aikaterini Spiliopoulou, Anastasios Braimakis, Shawn Young, Allen Cohn, Brian Pottorf, Howard Shapiro, Graham Sellers, Kelly Birdsey, Andrea Spaulding, Junichi Arita, Nobuyuki Takemura, Ryuji Yoshioka, Ono Yoshihiro, Fumitake Suzuki, Rintaro Koga, Takeshi Sano, Akio Saiura, Toshiharu Yamaguchi, Thomas Gruenberger, Friedrich Längle, Josef Thaler, Wolfgang Eisterer, Alfred Lenauer, Stefan Stremitzer, Judith Stift, Birgit Gruenberger, Valerie Bridoux, Karoui Mehdi, Lilian Schwarz, Francis Michot, Jean Jacques Tuech, Calvin Park, Michael Lin, Thomas Hugh, Weng Ng, Encarnación González-Flores, Julia Ruiz, Beatriz Gonzalez-Astorga, Verónica Conde, Javier Garcia Garcia, Aranzazu Gonzalez-Vicente, Jesus Soberino, Cynthia Gonzalez, Luque-Caro Raquel, Carmen Sanchez-Toro, Joaquina Martinez-Galan, Juan Ramón Delgado, Sebastian Stintzing, Alexander Grothe, Ralf Hoffmann, Volker Heinemann, Markus Rentsch, Alexander Muacevic, Christoph Trumm, Fotios Loupakis, Dongyun Yang, Shibao Feng, Chiara Cremolini, Wu Zhang, Carlotta Antoniotti, Christiane Langer, Herbert Hurwitz, Leonard Saltz, Alfredo Falcone, Heinz Joseph Lenz, Hugo Bernard, Feriel-Lylia Messekher, Véronique Lussier, Charles Ménard, Maryam Alfa-Wali,, Anand Sharma, Sadie Boniface, Paris Tekkis, Allan Hackshaw, Anthony Antoniou, Niki Kontou, Theodora Psaltopoulou, Nick Soupos, Evangelos Polychronopoulos, Dimitrios Xinopoulos, Athena Linos, Meletios Dimopoulos, Demosthenes Panagiotakos, Deeqa Ahmed, Trude Ågesen, Stine Danielsen, Michael Bretthauer, Espen Thiis-Evensen, Geir Hoff, Torleiv Rognum, Arild Nesbakken, Ragnhild Lothe, Guro Lind, Stepan Suchanek, Ondrej Majek, Gabriela Veprekova, Ladislav Dusek, Miroslav Zavoral, Martina Rebersek, Marko Boc, Zvezdana Hlebanja, Neva Volk, Jernej Benedik, Maja Ebert, Tanja Mesti, Janja Ocvirk, Marta Schirripa, Fotios Loupakis, Chiara Cremolini, Yang Dongyun, Melissa Janae Labonte, Wu Zhang, Lisa Salvatore, Takeru Wakatsuky, Carlotta Antoniotti, Giuseppe Aprile, Sara Lucchesi, Georgios Giamas, Stebbing Justin, Alfredo Falcone, Heinz Joseph Lenz, Antonio Cubillo, Jesús Rodriguez-Pascual, Fernando López-Ríos, Carlos Plaza Lopez, Elena García, Rafael Álvarez, De Emilio Vicente, Yolanda Quijano, Ovidio Hernando, Carmen Rubio, Manuel Hidalgo, Cathy Eng, Bessudo Alberto, Nashat Gabrail, Taina Lopez, Hamim Zahir, Reinhard von Roemeling, Johanna Bendell, Eric Van Cutsem, Alan Rong, Yves Humblet, Salvatore Siena, Stephan Braun, Roger Sidhu, Marc Peeters, Erin Gardner, William Figg, Hughes Marybeth, James Pingpank Jr., Sebastian Stintzing, Andreas Hochhaus, Ludwig Fischer von Weikersthal, Ruediger Laubender, Deniz Gencer, Iris Burkholder, Volker Heinemann, Ralf Hofheinz, Florence Joulain, Sheikh Iqbal, Françoise Diamand, Eric Van Cutsem, Martin Hoyle, Carmen Allegra, Michel Ducreux, Marc Peeters, Salvatore Siena, Jean-Yves Douillard, Cornelis Punt, Stephan Braun, Alan Rong, Roger Sidhu, Teresa Macarulla, Josep Tabernero, Andrés Cervantes, Susana Roselló, Eric Van Cutsem, Sabine Tejpar, Hans Prenen, Erika Martinelli, Teresa Troiani, Fank Campana, Bernard Laffranchi, Virginie Jego, Oliver von Richter, Fortunato Ciardiello, Marc Peeters, Jean-Yves Douillard, Eric Van Cutsem, Siena Salvatore, Kathy Zhang, Richard Williams, Jeffrey Wiezorek, Jaffer Ajani, Jin Li, Taroh Satoh, Tomohiro Nishina, A. Alarcón-Rozas, Junji Furuse, Wei Liu, Min-Hee Ryu, Was Mansoor, Todd Roma, Heind Smith, Jocelynn Booth, Michaela Sedova, Shriya Bhushan, Tarek Sahmoud, Syed Rizvi, Yung-Jue Bang, Sylvie Lorenzen, C. Pauligk, N. Homann, H. Schmalenberg, F. Tauchert, E. Jäger, S.-E. Al-Batran, Jenny Hong, Ju-Seog Lee, Sang Cheul Oh, Carsten Bokemeyer, Karsten Ridwelski, G. Ramadori, Djordje Atanackovic, Dirk Arnold, Ewald Wöll, Markus Buechler, Christian Krüger, Christoph Schuhmacher, Henriette Golcher, Helmut Witzigmann, Lukas Marti, Jochen Lange, Wolf Bechstein, Christiane Bruns, Henry Jungnickel, Johann Hauss, Stefan Schreiber, Thomas Brunner, Gerhard Grabenbauer, Susanne Merkel, Rainer Fietkau, Werner Hohenberger. Poster discussions, Annals of Oncology, 2012, iv19-iv30, DOI: 10.1093/annonc/mds152