Search

PURPOSE The combination of oxaliplatin and fluoropyrimidine for 6 months is one of the standard options for adjuvant therapy for high-risk stage II and III colorectal cancers (CRCs). The optimal duration of oxaliplatin to diminish neurotoxicity without compromising efficacy needs to be clarified. PATIENTS AND METHODS This open-label, randomized, phase III, noninferiority trial randomly assigned patients with high-risk stage II and III CRC to 3 and 6 months of oxaliplatin with 6 months of fluoropyrimidine groups (3- and 6-month arms, respectively). The primary end point was disease-free survival (DFS), and the noninferiority margin was a hazard ratio (HR) of 1.25. RESULTS In total, 1,788 patients were randomly assigned to the 6-month (n = 895) and 3-month (n = 893) arms, and 83.6% in the 6-month arm and 85.7% in the 3-month arm completed the treatment. The neuropathy rates with any grade were higher in the 6-month arm than in the 3-month arm (69.5% v 58.3%; P < .0001). The 3-year DFS rates were 83.7% and 84.7% in the 6-month and 3-month arms, respectively, with an HR of 0.953 (95% CI, 0.769 to 1.180; test for noninferiority, P = .0065) within the noninferiority margin. Among patients with stage III CRC treated by capecitabine plus oxaliplatin, the 3-year DFS of the 3-month arm was noninferior as compared with that of the 6-month arm with an HR of 0.713 (95% CI, 0.530 to 0.959; P = .0009). However, among patients with high-risk stage II and stage III CRC treated by infusional fluorouracil, leucovorin, and oxaliplatin, the noninferiority of the 3-month arm compared with the 6-month arm was not proven. CONCLUSION This study suggests that adding 3 months of oxaliplatin to 6 months of capecitabine could be considered an alternative adjuvant treatment for stage III CRC (ClinicalTrials.gov identifier: NCT01092481 ).

189 Background: ctDNA is an attractive alternative to tissue for its easy accessibility and real time representation of systemic tumor profile. We explored the utility of ctDNA profiling during the course of mCRC treatment. Methods: Serial blood samples were obtained from mCRC patients before and during first-line palliative chemotherapy at fixed intervals (after every four cycles) until confirmed disease progression. ctDNA was sequenced using targeted next-generation sequencing (NGS) platform with 106 genes. Changes of ctDNA profile and treatment outcome were comprehensively analyzed. Results: A total of 272 samples from 62 patients were analyzed. In the pre-treatment blood samples, 56 (90.3%) of patients had detectable ctDNA mutations including single nucleotide variants, short insertions/deletions and copy number changes (median 4.5 mutations/ patient, range 0 - 133). In 31 (50.0%) patients who had tissue NGS panel results performed in the clinic, overall concordance between mutations from ctDNA and tissue was 86.5%. In three patients, ctDNA mutational profiles were found to be completely different from tissue profiles. At further investigation, these patients were found to have a separate primary cancer in their colon. At the time of the first follow-up, most (98.0%) patients showed decrease of ctDNA from baseline, represented by average variant allele frequency (VAF) changes of all ctDNA mutations found. Clearance of ctDNA was achieved in 40 (78.4%) patients and was associated with longer progression-free survival (median PFS 11.8 moths in ctDNA clearance (+) vs. 4.7 months in ctDNA clearance (-), p < 0.001). The ctDNA clearance at the same time point was able to further discriminate the patients in same category by RECIST 1.1. Serial follow-up monitoring revealed three patterns of ctDNA changes at the time of clinical progressive disease (PD): 1) re-emergence or re-increase of baseline ctDNA mutations, 2) emergence of new resistance mutations, 3) radiologic PD without evidence of ctDNA progression. In the patients with detectable ctDNA at PD, the ctDNA progression preceded radiologic progression in 25 (58.1%) patients by median of 3.3 months. The patients in clinical PD without ctDNA progression showed different patterns of metastasis having mainly extrahepatic spread, while 77.8% of the patients with ctDNA progression had their progression confirmed in liver metastasis. Diverse resistant mutations and gene amplifications in PD patients were discovered by ctDNA sequencing. For seven (16.3%) of the PD patients, the newly identified mutations could be potential candidates of targeted therapy or clinical trial. Conclusions: ctDNA profile provided additional information to conventional evaluation methods and reflected dynamic changes. ctDNA monitoring may improve precise treatment decision-making for individual patients.

3592 Background: GC1118 is a novel anti-EGFR monoclonal antibody which has a unique binding epitope and superior ligand inhibition potential. It showed promising antitumor activity as a single agent in the phase I study. This study aimed to determine the maximum tolerated dose (MTD) and recommended phase 2 dose (RP2D) of GC1118 in combination with irinotecan or FOLFIRI in metastatic solid tumors and evaluate the efficacy of GC1118 plus FOLFIRI as a second line therapy for RAS and BRAF wild-type metastatic colorectal cancer. Methods: Phase 1b part was designed to evaluate weekly GC1118 (starting from 3 mg/kg) in combination with biweekly irinotecan (180mg/m2) or FOLFIRI (irinotecan 180mg/m2, leucovorin 400mg/m2, 5-FU 400 mg/m2 bolus, and 5-FU 2400mg/m2 over 46hrs) in a 3+3 design. In the phase 2a part, the RP2D of GC1118 is administered in combination with FOLFIRI in a Simon’s two stage design with objective response rate (ORR) as the primary endpoint. Results: 13 pts were enrolled in phase 1b and received 3mg/kg of GC1118 with irinotecan (N = 6) or FOLFIRI (N = 7). DLT occurred in 2 pts (G4 neutropenia, G2 rash) in irinotecan arm and 1pt (G3 neutropenia) in FOLFIRI arm with 3mg/kg of GC1118 and it was determined as MTD and RP2D. Adverse events (AE) of grade ≥ 3 included neutropenia (61.5 %), skin rash (15.4 %) and diarrhea (15.4%). Dose reductions due to GC1118-related AE were required in 6 (46.2%) patients. Among 10 response-evaluable pts in phase 1b, best overall response was PR in 3 and SD in 6, and median PFS was 12 months. In stage 1 of phase 2a (N = 9), 4 PR and 5 SD were observed (ORR 44.4%, 95% CI 13.7 – 78.8). We moved to stage 2, and are currently enrolling additional 20 pts. AE of grade ≥ 3 included neutropenia (66.7%), skin rash (22.2%) and diarrhea (11.1%). Updated data of the phase 2a part will be presented at the meeting. Conclusions: The MTD and RP2D of weekly GC1118 in combination with irinotecan or FOLFIRI was 3mg/kg. Preliminary results of GC1118 and FOLFIRI as a 2nd line treatment in mCRC suggests promising antitumor activity and acceptable safety profile. Clinical trial information: NCT03454620 .

711 Background: Modified CapeIRI (irinotecan 200 mg/m2 on day 1, capecitabine 1600 mg/m2 on days 1–14 every 3 weeks) with or without bevacizumab (± BV) has shown non-inferiority of overall survival compared with FOLFIRI ± BV based on the phase III study, AXEPT, as second-line chemotherapy for patients with mCRC. In this exploratory analysis of the AXEPT trial, we evaluated the impact of the prior chemotherapy with two different fluoropyrimidine backbones (fluorouracil and leucovorin vs oral fluoropyrimidine) on the efficacy of CapeIRI and FOLFIRI. Methods: Patients were randomized to receive standard FOLFIRI ± bevacizumab or modified CapeIRI ± bevacizumab after failure to fluoropyrimidine-based chemotherapy in the AXEPT study. Prior fluoropyrimidine backbones were categorized into oral fluoropyrimidine-based (eg, capecitabine or S-1) regimen (oral 5-FU group) and fluorouracil and leucovorin-based regimen (infusional 5-FU group). Assessed endpoints included overall survival, progression-free survival, response rate, and safety. Results: Prior fluoropyrimidine backbone was available for 642 patients among all 650 randomized patients (oral 5-FU group in 291, and infusional 5-FU group in 351). Median overall survival was 17.0 and 16.7 months for FOLFIRI ± BV and CapeIRI ± BV in the prior oral 5-FU group, and 14.9 and 16.7 months for FOLFIRI ± BV and CapeIRI ± BV in the prior infusional 5-FU group. Median progression-free survival was 7.9 and 8.6 months for FOLFIRI ± BV and CapeIRI ± BV in the prior oral 5-FU group, and 6.8 and 8.3 months for FOLFIRI ± BV and CapeIRI ± BV in the prior infusional 5-FU group. Conclusions: There was no significant difference in the efficacy of CapeIRI or FOLFIRI regardless of prior fluoropyrimidine backbones. Therefore, CapeIRI ± BV could also be effective for patients after failure of oral fluoropyrimidine-based chemotherapy (eg, CapeOX ± BV). Clinical trial information: NCT01996306. [Table: see text]

Purpose The primary purpose of our study was to evaluate whether maintenance chemotherapy with paclitaxel/gemcitabine (PG) was superior to observation in improving progression-free survival (PFS) in patients with metastatic breast cancer (MBC) who achieved disease control with an initial six cycles of PG as their first-line treatment. Patients and Methods The study was a prospective, randomized, multicenter, phase III trial. Patients MBC with who achieved disease control after six cycles of PG chemotherapy were randomly assigned to maintenance chemotherapy or observation until progression. Results Of 324 patients from 10 centers enrolled, 231 patients with MBC exhibited disease control (complete response + partial response + stable disease) with first-line PG and were randomly assigned to maintenance chemotherapy (n = 116) or observation (n = 115). The median age was 48 years (range, 28 to 76 years), median follow-up was 33 months, and median number of chemotherapy cycles in the maintenance group after random assignment was six. The median PFS time after random assignment was longer in the maintenance group than in the observation group (7.5 v 3.8 months, respectively; P = .026). The median overall survival (OS) time was longer in the maintenance group than in the observation group (32.3 v 23.5 months, respectively; P = .047). The rate of grade 3 or higher neutropenia after random assignment was higher in the maintenance group than in the observation group (61% v 0.9%, respectively; P < .001). Conclusion In patients with MBC who achieved disease control with an initial six cycles of PG chemotherapy, maintenance PG chemotherapy resulted in better PFS and OS compared with observation.

Purpose This study was undertaken to investigate the effects of epidermal growth factor receptor (EGFR) mutation and its downstream signaling on response and survival in non–small-cell lung cancer (NSCLC) patients treated with gefitinib. Patients and Methods For 90 consecutive NSCLC patients who had received gefitinib, EGFR mutation was analyzed by DNA sequencing of exons 18, 19, 21, and 23 in the EGFR tyrosine kinase domain. Expressions of phosphorylated (p) -Akt and p-Erk were determined via immunohistochemistry. Response rate, time to progression (TTP), and overall survival were compared between each group according to EGFR mutation, as well as p-Akt and p-Erk expression. Results Seventeen patients (18.9%; 95% CI, 10.8 to 27.0) harbored EGFR mutations. These mutations include deletions in exon 19 in seven patients, L858R in six patients, G719A in three patients, and a novel A859T in one patient. Response rate in patients with EGFR mutation was 64.7% (11 of 17 patients; 95% CI, 42.0 to 87.4), in contrast to 13.7% (10 of 73 patients; 95% CI, 5.8 to 21.6) in patients without mutation (P < .001). Moreover, these 17 patients with EGFR mutation had significantly prolonged TTP (21.7 v 1.8 months; P < .001) and overall survival (30.5 v 6.6 months; P < .001) compared with the remaining 73 patients without mutation. Although no significant correlation was detected between EGFR mutation and expressions of p-Akt or p-Erk, p-Akt overexpression was associated with prolonged TTP in patients with EGFR mutation. Conclusion Our data further support the importance of EGFR mutation with regard to gefitinib sensitivity. In addition to its predictive role, EGFR mutation confers significant survival benefits on NSCLC patients treated with gefitinib.

460 Background: Reports regarding sarcopenia as a prognostic factor in advanced biliary tract cancer (BTC) are rare. Furthermore, no study has investigated the dynamics of body weight with body muscle mass as a prognostic factor in advanced BTC patients undergoing palliative chemotherapy. Hence, we investigated whether sarcopenia affects survival in patients with BTC, with a co-analysis of body weight loss and body mass index (BMI). Methods: We consecutively enrolled patients with advanced BTC who received palliative chemotherapy between 2003 and 2013. Total muscle cross-sectional area (cm2) at the L3 level assessed by computed tomography was analyzed. We defined sarcopenia as a skeletal muscle index (SMI) < 48.5 cm2/m2 (men) and < 39.5 cm2/m2(women) using ROC curves. Results: The proportion of patients with sarcopenia upon diagnosis was 52.4% and 42.2% for men and women, respectively. By multivariate analysis, sarcopenia at diagnosis and decreased SMI during chemotherapy ( P = 0.008 and P < 0.001, respectively) were poor prognostic factors for overall survival (OS). Subgroup analysis revealed that sarcopenic patients who were overweight or obese (BMI ≥ 25 kg/m2) showed worse OS ( P < 0.001). Additionally, patients with both decreased BMI and SMI during chemotherapy had worse OS ( P < 0.001). Furthermore, patients with decreased SMI had shorter survival regardless of change in BMI. However, for patients with SMI maintained during chemotherapy, decreased BMI had no effect on survival ( P = 0.576). Conclusions: Sarcopenia, sarcopenic obesity and muscle depletion during palliative chemotherapy are meaningful prognostic factors in advanced BTC. Considering muscle depletion with weight change could help to more accurately predict prognosis of patients with BTC.

454 Background: Time to progression (TTP) is suggested as a reliable endpoint compared to progression free survival (PFS) in the clinical trials of hepatocellular carcinoma (HCC). However, the correlation between TTP and overall survival (OS) has never been studied. Methods: We searched Pubmed and Embase data to obtain data source. Eligible studies were randomized controlled phase III trials which evaluated the efficacy of systemic chemotherapy or molecular targeted therapy in advanced HCC. The association of treatment effects as shown by the hazard ratio (HR) of TTP and OS in each trial was assessed by Spearman rank correlation coefficient ( rs) and linear regression analysis. The association between median TTP and OS was also investigated. Results: Nine studies with a total of 18 treatment arms and 6318 patients were included. Incremental benefit from the study treatment in TTP from each trial was correlated with incremental benefit in OS. The rs value and R2 value between log (HRTTP) and log (HROS) was 0.73 (95% CI 0.12 – 0.94, p = 0.024) and 0.57. The minimum TTP effect to predict a treatment effect on OS was 0.63. Median TTP was associated with median OS. The rs value between TTP and OS was 0.73 (95% confidence interval (CI) 0.40 – 0.89, p < 0.001) and the corresponding R2 was 0.42. Conclusions: Our study results suggest that TTP could be used as a surrogate marker for OS in the clinical trials of advanced HCC. However, the results suggest modest correlation between treatment effects on TTP and OS. Along with individual-level analysis more evidences are needed to confirm our finding.

416 Background: Systemic inflammation is known to promote carcinogenesis in biliary tract cancer (BTC). Neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR) are indicative of systemic inflammation. We evaluated the clinical significance of systemic inflammation measured by NLR and PLR in patients with advanced BTC. Additionally, we also co-analyzed the dynamics of NLR and PLR during chemotherapy. Methods: We reviewed 554 patients with advanced BTC receiving palliative chemotherapy. NLR and PLR were obtained before initiation of palliative chemotherapy. Changes in NLR, PLR were obtained by subtracting the initial value from the value obtained after progression of chemotherapy. Results: Higher systemic inflammation status also had relation with a primary tumor site ( p = 0.044) and higher levels of CEA ( p = 0.041). The ROC cut-off values of NLR and PLR for predicting overall survival (OS) were 3.8 and 121, respectively. Patients with a high NLR or PLR had worse OS independently in multivariate analysis (7.10 vs. 10.23 months, p < 0.001; 8.43 vs. 11.87 months, p = 0.001, respectively). High NLR with increased NLR after chemotherapy is associated with worse OS and progression-free survival (PFS) ( p = 0.014, p= 0.020 respectively). Results are similar for PLR. Conclusions: Systemic inflammation represented by NLR and PLR, predicts the OS of patients with advanced BTC who are receiving palliative chemotherapy. In addition, considering NLR/PLR with a dynamic change of NLR/PLR during chemotherapy might help to predict a more accurate prognosis.

681 Background: Several studies have shown that capecitabine plus irinotecan (XELIRI) has promising efficacy and safety in patients with metastatic colorectal cancer. AXEPT is a non-inferiority, phase III comparison of XELIRI with or without bevacizumab vs 5-fluorouracil/folinic acid plus irinotecan (FOLFIRI) with or without bevacizumab as second-line therapy in patients with metastatic colorectal cancer. Methods: We undertook an open-label, non-inferiority, randomized phase III trial in South Korea, China and Japan. Patients were randomized 1:1 to XELIRI (irinotecan 200 mg/m2 on day 1 plus capecitabine 800 mg/m2 twice daily for 2 weeks in a 3-week cycle) with or without bevacizumab 7.5 mg/kg or FOLFIRI with or without bevacizumab. The primary endpoint was overall survival. Results: From December 2013 to August 2015, 650 patients were enrolled and 625 (311 in FOLFIRI and 314 in XELIRI) were included to safety analysis at the cutoff dates on August 31, 2016. Patient baseline characteristics including KRAS status and UGT1A1 gene polymorphism were similar between the two treatment arms. Prior chemotherapy with oxaliplatin was given to 97.4% in both arms. The overall incidence of grade 3-4 toxicity was 73% in FOLFIRI arm and 53.2% in XELIRI arm. Whereas FOLFIRI was associated with more grade 3-4 neutropenia (43.7% vs 16.2%), leucopenia (13.5% vs 7.3%) and all grades anemia (80.4% vs 69.4%) than XELIRI, XELIRI was associated with more all grades hand-foot syndrome (34.1% vs 14.8%) and grade 3-4 diarrhea (7.0% vs 3.2%). There was no significant difference in febrile neutropenia (4.2% in FOLFIRI and 2.9% in XELIRI). The addition of bevacizumab did not alter safety profiles between XELIRI and FOLFIRI. There was a treatment-related death in FOLFIRI arm (0.3%). Conclusions: Both FOLFIRI and XELIRI were safe and well tolerated, though there were differences in the rates and toxicity profiles at which adverse events occur. Clinical trial information: NCT01996306. UMIN000012263.

10067Background: Patients’ proper understanding and consent to participate are crucial for successful conduction of clinical trials (CT). We have conducted a survey, to assess the current patients’...

3502Background: Atezolizumab (atezo; MPDL3280A) is an engineered antibody that inhibits binding of PD-L1 to its receptors, PD-1 and B7.1. Atezo has demonstrated monotherapy activity in a multitude of human tumor types. However, response rates in microsatellite stable (MSS) CRCs have been lower than in other indications. In preclinical models, targeted inhibition of MEK leads to upregulation of MHC I on tumor cells, induces intratumoral T-cell infiltration and enhances anti-PDL1 activity. We therefore conducted a Phase Ib study combining cobi (MEK inhibitor) and atezo in patients (pts) with advanced solid tumors. Methods: Cobi was escalated from 20 to 60 mg daily (21 days on/7 days off) and combined with atezo 800 mg IV q2w. Tumor-specific expansion cohorts, including KRAS-mutant CRC, and serial biopsy cohorts in solid tumors were opened upon determination of the MTD. Safety, tolerability and confirmed ORR by RECIST v1.1 were evaluated. Results: As of October 12, 2015, 23 CRC (22 KRAS mutant, 1 WT) pts wer...

2521Background: GC1118 is a novel anti-EGFR monoclonal antibody and has unique binding epitopes and affinity. We conduct a first-in-human phase 1 study to investigate the maximum tolerated dose (MT...

4075Background: While overall survival (OS) is the most important survival endpoint in oncology clinical trials, it can be affected by sequential treatment after progression or by crossover from co...

e23268Background: We aim to investigate the correlation of ataxia-telangiectasia-mutated (ATM) protein expression with clinicopathological features and prognosis in patients with breast cancer. Met...

56 Background: Anti-angiogenic strategy in gastric cancer (GC) has been highlighted again due to the recent success of ramucirumab and apatinib. Therefore, the comprehensive network of VEGF, soluble VEGF receptor-2 (sVEGFR2) and cytokines and other angiogenic factors (CAF) in GC and their prognostic impact would be of importance, although they have been poorly understood. We aimed to find out the CAF signature associated with VEGF and sVEGFR2, and to explore their prognostic implication in GC. Methods: We measured pretreatment serum levels of 52 CAFs, including VEGF and sVEGFR2, using multiplex bead immunoassays and ELISA, in 70 patients who were diagnosed with GC in Seoul National University Hospital, and treated with palliative chemotherapy. Linear regression analysis for correlating CAFs with VEGF and sVEGFR2, and survival analysis by log rank test and Cox regression analysis were performed. Results: The VEGF signature was shown to be associated with seven CAFs (interluekin [IL]-7, IL-12p70, IL-2Ra, IL-10, stem cell factor, Fibroblast growth factor-basic, IL-3). The sVEGFR2 signature was associated with IL-4 and platelet-derived growth factor beta, but VEGF and sVEGFR2 showed no association with each other. Patients with high VEGF had a tendency to have worse overall survival (OS) than those with low VEGF (11.2 months versus 16.7 months; P = 0.061). However, among patients with high-sVEGFR2, OS was not different according to VEGF (12.1 months, high-VEGF versus 15.1 months, low-VEGF; P = 0.546). Interestingly, the poor prognostic impact of high-VEGF was far significant in patients with low-sVEGFR2 (10.9 months versus 16.8 months; P = 0.036). With this perspective, VEGF/sVEGFR2 ratio was significantly correlated with worse OS in univariate as well as multivariate analysis (HR 1.78 [95% CI 1.08-2.94], P= 0.024). Conclusions: Based on the comprehensive network analysis of CAF, VEGF and sVEGFR2 had distinct CAF signatures in GC. Consideration of both VEGF and sVEGFR2 confers more accurate prognostic implication compared with VEGF alone in GC. Regarding the angiogenic aspect, VEGF/sVEGFR2 ratio is significantly correlated with survival outcome in GC.

TPS780 Background: Life-prolonging systemic therapy such as chemotherapy (FOLFOX, XELOX, FOLFIRI, 5-FU/LV) with / without molecular targeted agents (bevacizumab, afilibercept, cetuximab, panitumumab) are important treatment for metastatic colorectal cancer (mCRC). On the other hand, it is required to establish an evidence of convenient therapy including oral anticancer agent which is expected lower risk of safety. XELIRI was already examined by various doses in this decade, and result of AIO 0604 was regarded as one of the most appropriate regimen. The BIX phase II study showed the tolerability and promising efficacy of the AIO regimen for Japanese mCRC patients (Hamamoto Y, et al. Oncologist 2014). Methods: Asian XELIRI Project (AXEPT) is an East Asian collaborated, open label, randomized phase III clinical trial designed to demonstrate the non-inferiority of XELIRI (+bevacizumab) to the standard of care FOLFIRI (+bevacizumab) as the second-line chemotherapy in patients with mCRC. The primary endpoint is median overall survival. The secondary endpoints are progression-free survival, time to treatment failure, overall response rate, disease control rate, relative dose intensity, safety, correlation between UGT1A1 polymorphisms (*28, *6) and safety. Eligible patients were 20 years of age and older, had histologically proven CRC, ECOG performance status 0-2, adequate organ function, progression or difficult to continue after first line regimen. Patients were randomized 1:1 to standard FOLFIRI (+bevacizuamb 5mg/kg day1), repeated every 14 days (Group A) or XELIRI, Irinotecan 200mg/m2 day1, and capecitabine 1600mg/m2 day 1-14 (+bevacizumab 7.5mg/kg day1), repeated every 21 days (Group B). A Total of 464 events were needed to show non-inferiority with a two-sided α of 0·05 and a power of 80%, a target sample size of 600 patients was required. (The 95% CI upper limit of the hazard ratio was pre-specified as less than 1.3.). Enrollment has started in December 2013. As of August 2015, 98 centers in the Japan, South Korea and China are participating in this trial. Clinical trial information: NCT01996306. UMIN000012263. Clinical trial information: NCT01996306.

411 Background: Host antitumor immunity measured by neutrophil-lymphocyte ratio (NLR) is reported to be associated with prognosis of cancer patients. Standard uptake value (SUV)by18F-FDG PET represents the metabolic activity of tumor itself. The correlation of host immunity and tumor metabolic activity has not been studied. We investigate the association of these two factors and evaluate their roles in prediction of overall survival (OS) in metastatic pancreatic cancer (MPC) patients who receive palliative chemotherapy. Methods: We reviewed 396 MPC patients receivingpalliative chemotherapy. NLR was obtained before initiation of chemotherapy and after 1st cycle of chemotherapy. In 118 patients who were evaluated with 18F-FDG PET before initiation of chemotherapy (PET cohort), maximum SUV (SUVmax) was collected. Cut-offs for each variable were determined by ROC curve. Results: In multivariate analysis, higher NLR was associated with worse OS ( < 2.5, 9.0 m; 2.5-4.4, 7.2 m; ≥ 4.5, 3.9 m; p< 0.001). Reduction of NLR after 1st cycle of chemotherapywas associated with betterOS (8.0 m vs 6.1 m; HR 0.653; p< 0.001). We made the risk scoring model considering both NLR (score 0, NLR < 2.5; score 1, 2.5 ≤ NLR < 4.5; score 2, NLR ≥ 4.5) and ΔNLR(score 0:ΔNLR < 0; score 1: ΔNLR ≥ 0), which identified 4 risk groups with different prognosis (group with risk score 0 vs 1 vs 2 vs 3: OS 9.7 vs 7.9 vs 5.7 vs 2.6 months; HR 1 vs 1.329 vs 2.137 vs 7.915, respectively; P < 0.001). In PET cohort (118 patients), NLR and SUVmax were independent prognostic factors for OS. The risk model considering both NLR (score 0, NLR < 2.5; score 1, 2.5 ≤ NLR < 4.5; score 2, NLR ≥ 4.5) and SUVmax (score 0:SUVmax < 4.5; score 1: SUVmax ≥ 4.5), which define 4 risk groups with different OS. (group with risk score 0 vs 1 vs 2 vs 3: OS 11.8 vs 9.8 vs 7.2 vs 4.6 months; HR 1 vs 1.536 vs 2.958 vs 5.336, respectively; P < 0.001). Conclusions: NLR and SUVmax as simple parameters of host antitumor immunity and tumor metabolic activity, respectively, have significant impacts on the survival of metastatic pancreatic cancer patients independently. Consideration of both aspects allows more precise prediction of patients’ prognosis.

3576 Background: The purpose of this randomized phase III trial was to evaluate the addition of synthetic 3-hydroxy-3methyglutaryl coenzyme A (HMG-CoA) reductase inhibitor, simvastatin to XELIRI/FO...

11094 Background: Programmed death-ligand 1 (PD-L1) expression in cancer cells or tumor microenvironment is under investigation as a candidate biomarker in the new drug development of immuno-oncolo...

4050 Background: Trastuzumab-based chemotherapy is standard treatment for HER2-positive advanced gastric cancer (AGC). Although increased HER2 gene amplification by fluorescent in situ hybridizatio...

415 Background: Ampulla of Vater cancer (AoV ca) is rare tumor, and its adjuvant treatment after curative resection has not been well established. The purpose of this study is to see the clinical outcomes of curatively resected AoV ca patients and to figure out the role of adjuvant treatment. Methods: We reviewed 227 AoV ca patients who underwent curative resection at the Seoul National University Hospital between 1997 and 2012. Clinical factors, pathologic findings, adjuvant treatment pattern, disease-free survival (DFS) and overall survival (OS) were analyzed. Results: Median age was 61.5 year old (range, 33.8-88.2) and 125 patients were male. T1/T2 stage was found in 63.9% of patients and T3/4 stage was in 36.1%. Seventy-seven patients (33.9%) had node positive disease (N+) and seventy-three patients (32.1%) had well-differentiated histology. OS of all patients was 90.9 months (95% CI: 58.4-128.09). 5-Year OS- and DFS- rate were 58.2% and 62.5%, respectively. On multivariate analysis, adverse prognostic factors for OS included poorly differentiated histology (HR 1.35, 95% CI: 1.01-1.81, p=0.045), elevated CEA (HR 2.55, 95% CI: 1.29-5.04, p=0.007), elevated CA 19-9(HR 1.86, 95% CI: 1.08-3.21, p=0.027), and stage (HR 2.34, 95% CI: 1.43-3.82, p=0.001). A total of 104 patients (46.3%) received adjuvant treatment (concurrent chemoradiotherapy (CCRT) 32, CCRT with maintenance chemotherapy 58, chemotherapy alone 9, radiotherapy alone 5). Most of the chemotherapies were 5-FU-based ones. Among patients with T1/T2 & N (-) (n=111), only 25 patients (22.5%) received adjuvant treatment. In contrast, among patients with T3/T4 or N (+) (n=116), 79 patients (68.1%) received adjuvant treatment. 5Y-OS rate was 74.1% in patients with T1/T2 and N (-) and that of patients with T3/T4 or N (+) was 43.0% (p

83 Background: Weight loss during chemotherapy is a significant prognostic factor for poor survival in patients advanced gastric cancer (AGC). However, in most studies, weight loss was measured at the end of chemotherapy, limiting its clinical use. In this study, we evaluated whether weight loss during the first month of chemotherapy could predict survival outcomes in patients with AGC. Methods: We analyzed 719 patients with metastatic or recurrent AGC who were receiving palliative chemotherapy. We calculated initial body mass index (BMIi), percent weight loss after one month of chemotherapy (ΔW1m), percent weight loss after last administration of chemotherapy (ΔWend), and average weight loss per month during chemotherapy (ΔW/m). We correlated these data with overall survival (OS) by receiver operating characteristic (ROC) curves and Kaplan-Meier curves, and performed a subgroup analysis using Cox regression. Results: The probabilities of longer OS had stronger correlations with ΔW/m and ΔW1m than with ΔWend or BMIi. The optimal cutoff values of ΔW/m and ΔW1m for predicting shorter survival were 1% and 3%, respectively. A significant positive correlation between ΔW1m and ΔW/m (r2 = 0.591, p < 0.001) was observed. Patients with ΔW1m more than 3% significantly younger, had worse performance status, more diffuse-type Lauren classification, more HER2-negative pathology, a higher number of involved organs, and more peritoneal seeding at initial presentation. OS of patients with ΔW1m more than 3% were significantly shorter than patients with less weight loss (ΔW1m ≥3%: 9.7

363 Background: Body composition has emerged as a prognostic factor in cancer patients. We investigated whether sarcopenia at the diagnosis and progressive loss of skeletal muscle were associated with survival in pancreatic cancer (PC) patients receiving palliative chemotherapy. Methods: We retrospectively reviewed PC patients receiving palliative chemotherapy between 2003 and 2010. Skeletal muscle cross-sectional area at L3 was measured by computed tomography. Sarcopenia was defined using the previously published sex-specific cutoff points for Korean people. Loss of skeletal muscle was classified by sex-specific cutoffs from ROC curve. Results: Among 484 patients, 260 (53.7%) patients were more than sixty years old and 295 (61.0%) patients were male. Overall, 187 (38.6%) patients were sarcopenic at the diagnosis (male, 2/m2; female, 2/m2). Decreased skeletal muscle index (SMI) during the chemotherapy, which was defined as a reduction by more than 0.21 for male and by more than 2.19 for female, was observed in 198 (77.3%) male patients and 61 (38.1%) female patients. Decreased body mass index (BMI) by more than 1 was observed in 149 patients (37.3%) without difference between genders. Median overall survival (OS) of whole patients was 8.4 months [95%CI 7.6-9.2]. In the multivariate analysis, sarcopenia (p=0.001), decreased SMI (p=0.003), and decreased BMI (p=0.001) was significantly poor prognostic factors for OS. When we analyzed four groups by SMI and BMI changes (maintained SMI and BMI, maintained SMI and decreased BMI, decreased SMI and maintained BMI, decreased SMI and BMI), the groups with maintained SMI had longer survival than the groups with decreased SMI regardless of BMI changes. Median OS was 11.5 months with maintained SMI and 8.1 months with decreased SMI (HR 0.534 and 1, p=0.004) in decreased BMI groups and 9.9 months with maintained SMI and 8.6 months with decreased SMI (HR 1 and 1.490, p=0.002) in maintained BMI groups, respectively. The analyses separately done by gender showed similar results. Conclusions: Decrease of SMI during chemotherapy was a significantly poor prognostic factor for survival regardless of BMI changes.

4061 Background: In advanced gastric cancer (AGC), HER2 is a validated therapeutic target. However, the metabolic landscape of AGC based on the HER2 status has not been reported. Furthermore, the p...

4062 Background: Trastuzumab-based chemotherapy is considered as standard treatment for patients with HER2-positive advanced gastric cancer (AGC). However, the biomarkers to predict the sensitivity...

351 Background: The optimal treatment strategy for locally advanced pancreatic cancer (LAPC), especially the role of chemoradiotherapy (CCRT), is still in debate. We compared the clinical outcomes of CCRT and palliative chemotherapy alone (CA) in patients with LAPC. Methods: We consecutively enrolled LAPC patients treated between 2003 and 2010. AJCC 7th edition was followed for the diagnostic criteria of LAPC. We retrospectively evaluated the clinical outcomes according to treatment groups (CCRT vs CA). Results: A total of 86 patients were enrolled. Median age was 60 years. ECOG PS was 0-1 in 77 (89.5%) and 2 in 9 (10.5%). Forty five patients (52.3%) were treated with CCRT and 41 patients (47.7%) with CA. Baseline characteristics were not significantly different between CCRT and CA group. In the CCRT group, gemcitabine (n=7, 15.6%), 5-FU (n=10, 22.2%), and capecitabine (n=28, 62.2%) were concurrently used with radiation. Radiation was delivered with 55.8Gy/ 31fraction. All of the CA group patients were treated with gemcitabine-based chemotherapy. Median progression free survival (PFS) and overall survival (OS) of whole patients were 6.9 months [95%CI 4.8-9.0] and 12.7 months [95%CI 11.6-14.3]. PFS and OS of CCRT versus CA was 8.9 months [95%CI 6.8-11.0] vs 3.7 months [95%CI 2.9-4.5] (p

252 Background: Metabolic landscape evaluated by 18F-FDG positron emission tomography (PET) in advanced biliary tract cancer (BTC) has not been studied. Furthermore, the clinical meanings of metabolic features have rarely been reported. We evaluated the metabolic features using 18F-FDG PET and its clinical significances in advanced BTC. Methods: We consecutively enrolled advanced BTC patientswho underwent PET prior to palliative chemotherapy between 2003 and 2013. We retrospectively evaluated the findings of PET such as SUVmax, lesion numbers, involved organ and pathologic findings and other clinical factors including response to chemotherapy, progression-free survival (PFS) and overall survival (OS). Results: A total of 106 patients were enrolled (intrahepatic cholangiocarcinoma (ICC):53, extrahepatic BTC :7, gallbladder cancer (GB Ca) :30, and ampulla of vater cancer (AoV Ca): 16 patients) ECOG PS 0-1 in 89 and PS 2 in11 patients. 52.8 % of patients received gemcitabine-based chemotherapy and 47.1% of patients received 5-FU-based chemotherapy. The SUVmax was median 7.8 (range 0-44). Fifty-four percent of patients showed higher SUVmax in metastatic lesion than primary site. The SUVmax was different according to primary origin (ICC: 9.10, extrahepatic BTC: 5.90, GB Ca : 9.10, AoV Ca :6.37, p=0.008) and histologic differentiation (well-differentiated: 4.95, moderately-differentiated :6.60, poorly-differentiated :14.50, p=0.004) Patients with a SUVmax of >7.5 had more poorly differentiated histology and more PET uptake-lesions (p < 0.05) than those with a SUVmax of 7.5. In multivariate analysis, SUVmax was also a significant prognostic factor for OS (p=0.012) and PFS (p=0.039). Conclusions: Metabolic landscapes of advanced BTC are different according to primary origin and histology. This metabolic feature such as SUVmax could be a potential prognostic factor for OS and PFS in advanced BTC.

4044 Background: A causal relationship between diabetes mellitus (DM) and pancreatic cancer (PC) is well established. However, in patients with advanced PC (APC) who receive palliative chemotherapy, the impact of DM on prognosis is unclear. Methods: Between 2003 and 2010, we enrolled consecutive patients with APC, all recipients of palliative chemotherapy, with the provision that DM disease status could be properly defined. Enrollees were stratified by diagnosis, in accordance with 2010 DM criteria (AHA/ADA). DM at least 2 years' duration prior to diagnosis of APC qualified as remote-onset DM (vs recent-onset). Clinical characteristics and outcomes were then analyzed. Results: Of the 349 enrollees with APC, 183 (52.4%) had DM. In patients with DM, 160 had DM at the time APC was diagnosed (remote onset, 87; recent onset, 73); and the remaining 23 developed DM during the course of APC treatment. Ultimately, 73.2% (134/183) of DM patients received antidiabetic medication (metformin (56), sulfonylurea (62), insulin (43)). By multivariate analysis, cancer extent (HR, 1.792; p < 0.001) and weight loss during chemotherapy (HR, 1.270; p = 0.08) were associated with diminished overall survival (OS), whereas a diagnosis of DM (HR, 0.788; p = 0.05) conferred a positive effect on OS ( OS 8.4 months in DM patients vs 7.5 months in non-DM patients, p = 0.04). Among DM patients, recent-onset DM trended toward prolonged OS, compared with remote-onset/subsequent DM (9.8 months vs. 7.9 months, respectively; HR, 0.789; p = 0.142). Neither antidiabetic medication in general nor sulfonylurea or insulin specifically affected OS. However, recipients of metformin survived longer than non-recipients (HR, 0.693; 95% CI, 0.492-0.977; p = 0.036). Relative to the APC cohort overall including non-DM patients, metformin conferred better survival as well (11.0 months vs. 7.8 months, HR 0.712, p = 0.067), given similar baseline clinical characteristics. Conclusions: In patients with APC receiving palliative chemotherapy, recent-onset DM (within 2 years of APC diagnosis) and metformin treatment are positive prognosticators, associated with prolonged OS.

2565 Background: HM781-36B, a novel pan-HER tyrosine kinase inhibitor, showed potent in vitro and in vivo antitumor activities for EGFR mutant models including T790M mutation. In the previous 2-weeks on, 1-week off phase I study, the maximum tolerated dose (MTD) was determined as 24 mg/day. Phase I study with a continuous daily dosing schedule was conducted to determine the recommended dose (RD), and to assess the effect of food on pharmacokinetics (PK) in patients (pts) with advanced solid tumors. Methods: Eligible pts had advanced malignancies refractory to standard therapies. Standard 3+3 scheme was used in the dose escalation study, and additional 8 patients were enrolled to test food effects. Results: 20 pts (median age: 55 years [range32-77], M:F=13:7, ECOG PS 0/1/2/3 8/12/0/0) were enrolled (5 NSCLC, 4 colon cancer, 3 stomach cancer, 2 breast cancer, 2 rectal cancer, 2 common bile duct cancer, 1 pancreatic cancer and 1 esophageal cancer); 12 in the dose escalation and 8 in the food effect study cohort. Twelve pts were heavily pretreated(≥4 regimens). Dose limiting toxicities (DLTs) were G3 anorexia in 1 pt at 18mg/day, G3 diarrhea and anorexia in 1 pt, and drug compliance last values, whereas Cmax values were lower in fed condition than in fasted condition. Conclusions: Continuous daily dosing schedule of HM781-36B is safe and well tolerated in advanced solid tumors. It exerts anticancer activity, without being influenced by food. Updated data will be presented at the meeting. Clinical trial information: NCT01455584.

551 Background: Although young age breast cancer represents poor prognosis, no definitive explanation could have been made for the phenomenon. A tumor proliferation marker Ki67 is known to be a marker for both prognosis and prediction for chemotherapy responsiveness, and its level varies widely depending on the breast cancer subtype. This study was aimed to analyze Ki67 in relationship with age in hormone receptor positive breast cancer patients. Methods: We retrospectively reviewed 9061 consecutive cases of hormone receptor positive invasive breast cancer from data base at Seoul National University Hospital (SNUH) (between 2000 and 2012), Samsung Medical Center (SMC) (between 2004 and 2010), and National Cancer Center (NCC) (between 2001 and 2010) in Korea. Patients with estrogen receptor (ER) or progesterone receptor (PR) positive tumors were included irrespective of HER2 amplification. A multicenter data of Ki67 level identified by immunohistochemistry (IHC) and age at diagnosis were analyzed. Patients who underwent neoadjuvant systemic therapy were excluded. Results: Total 6222 cases from SNUH, 976 from SMC and 1863 from NCC were included. The three datasets were analyzed separately due to variable IHC methods in each institute. Mean ages were 49.30 years (range 20-86), 47.75 years (range 22-81) and 45.31 years (range 25-59), and mean Ki67 levels were 4.66% (range 1-100), 22.98% (range 1-97) and 14.58% (range 1-90), at SNUH, SMC, NCC respectively. Ki67 level was inversely proportional with age at diagnosis in all three datasets, and the level was significantly higher for patients

3550 Background: There have been controversies in prognostic impact of mucinous histology in colorectal cancer (CRC) and its implication in pts treated with adjuvant FOLFOX is unclear. This study aimed at elucidating the molecular characteristics and prognostic implication of mucinous histology in pts treated with adjuvant FOLFOX. Methods: Stage II and III CRC pts who received adjuvant FOLFOX were analyzed. Pts were grouped according to the mucinous content: > 50%, mucinous adenocarcinoma (MAC); < 50%, adenocarcinoma with intermediated mucinous component (AIM); and without any mucinous component, nonmucinous adenocarcinoma (NMA). Clinicopathologic features, MSI status (N = 518), CpG island methylator phenotype (CIMP) (N = 322) BRAF mutation (N = 269) and disease-free survival (DFS) were compared. Results: Among a total of 521 pts, 27 (5.2%) had MAC, 41 (7.9%) AIM, and 453 (86.9%) NMA. MAC and AIM had higher frequency of proximal location and lower angiolymphatic invasion. MAC had higher proportion of T4 tumors. AIM had higher frequency of age ≥65 years and female. In terms of molecular characteristics, MAC and AIM showed similarly higher proportion of MSI-high and CIMP-high compared to NMA. BRAF mutation also showed similar trend. In contrast to the similarities between MAC and AIM, DFS was significantly different. MAC showed significantly worse DFS compared with AIM and NMA, whereas AIM and NMA showed similar DFS. Multivariate analysis revealed MAC as an independent negative prognostic factor of DFS (adjusted HR 7.96, 95% CI 3.76-16.8). Conclusions: AIM and MAC has distinct clinico-pathologic features and molecular characteristics compared with NMA. Only MAC but not AIM has an adverse prognostic impact on stage II or III CRC treated with adjuvant FOLFOX compared with NMA. [Table: see text]

182 Background: To evaluate the prognostic significance of phosphorylated Akt (p-Akt), phosphorylated mammalian target of rapamycin (p-mTOR), and total phosphatase and tensin homolog deleted on chromosome 10 (PTEN) expression in patients undergoing adjuvant chemoradiotherapy for proximal extrahepatic bile duct (EHBD) cancer. Methods: Sixty-three patients with proximal EHBD cancer who underwent curative resection followed by adjuvant chemoradiotherapy were enrolled into this study. Postoperative radiotherapy was delivered to tumor bed and regional lymph nodes up to a median of 40 Gy (range; 40-54). Fifty-nine patients also received fluoropyrimidine chemotherapy as a radiosensitizer. p-Akt, p-mTOR, and PTEN expression were assessed with immunohistochemical staining on the tissue microarray. Results: p-Akt, p-mTOR, and PTEN were expressed in 23 (36.5%), 17 (27.0%) and 25 patients (39.7%), respectively. p-Akt expression was associated with distant metastasis and overall survival, but not with loco-regional recurrence. The 5-yr distant metastasis-free and overall survival rates were 25.8% vs. 58.2% (p = 0.007), and 27.5% vs. 50.2% (p = 0.0167) in patients with negative and positive expression, respectively. On multivariate analysis, nodal involvement was the only significant prognosticator predicting inferior distant metastasis-free survival (p = 0.0105), while p-Akt expression had a borderline significance (p = 0.0541). As for overall survival, p-Akt expression was a marginally significant prognosticator (p= 0.0635), whereas other risk factors lost the statistical significance. Conclusions: p-Akt expression was a favorable prognostic factor for patients undergoing curative resection followed by adjuvant chemoradiotherapy for proximal EHBD cancer.

505 Background: Circumferential resection margin (CRM) and distal resection margin (DRM) have different impact on clinical outcomes after preoperative chemoradiotherapy (CRT) followed by surgery. Effect and adequate length of resection margin as well as impact of treatment response after preoperative CRT was evaluated. Methods: Total of 403 patients with locally advanced rectal cancer underwent preoperative CRT followed by total mesorectal excision between January 2004 and December 2010. After applying the criterion of margin less than 0.5 cm for CRM and/or less than 1 cm for DRM, 158 cases were included as a study cohort. All patients underwent conventionally fractionated radiation with dose over 50 Gy and concurrent chemotherapy with 5-FU or capecitabine. Median follow-up duration was 44.9 months. Results: The 5-year overall survival (OS), disease-free survival (DFS), locoregional relapse-free survival (LRFS), and distant metastasis-free survival (DMFS) were 83.3%, 75.6%, 86.3%, and 77.4% respectively. CRM of 1.5 mm and DRM of 7 mm were cutting points showing maximal difference using maximal chi-square method. In univariate analysis, the shorter CRM was significantly related with worse clinical outcomes, whereas DRM was not. In multivariate analysis, CRM of 1.5mm, ypN, and perineural invasion were prognosticators for OS, DFS, LRFS, and DMFS. CRM was not a significant prognostic factor for good responders, defined as patients with near total regression or T down-staging. However, poor responders demonstrated a significant difference according to the CRM status. Conclusions: Close CRM, defined as 1.5 mm, was a significant prognosticator, but the impact was different for treatment response. Postoperative treatment strategy may be individualized based on this finding. However, findings from this study needs to be validated with larger independent cohort. [Table: see text]

1003 Background: Chemotherapy provides a survival benefit in patients with metastatic breast cancer (MBC), but the optimal duration of chemotherapy remains controversial. Primary purpose of the study was to evaluate whether the maintenance chemotherapy with gemcitabine/paclitaxel (GP), which is one of the two regimens which showed a survival gain from a randomized trial, is superior to observation in terms of progression free survival (PFS) in responding patients with MBC after 6 cycles of GP as first-line treatment. Methods: This study is a prospective, randomized, multi-center, phase III study. Patients who achieved response (CR+PR+SD) following 6 cycles of GP chemotherapy (gemcitabine 1250 mg/m2 on day 1 and 8 plus paclitaxel 175 mg/m2 on day 1 every 3 weeks) randomized to maintenance till progression or observation arm. The trial was conducted by the Korean Cancer Study Group (KCSG). Results: Among total 324 patients enrolled between 2007 and 2010 from 10 centers, 231 responding patients to were randomly assigned to maintenance chemotherapy (n=116) or observation (n=115). Median age was 49 (range 28-76). The numbers of hormone receptor (HR)+ve and HR-ve patients were 172 (74.5%) and 59 (25.5%), respectively. The median No. of chemotherapy cycles in maintenance group was 12 (range 6-32). During median 33 months of follow-up, median PFS was superior in maintenance than in observation (12.0 vs. 8.3 months, p=0.030). Patients < age 50 years (hazard ratio 0.50, p=0.001) and HR-ve patients (hazard ratio 0.52, p=0.019) received more benefit from maintenance chemotherapy in terms of PFS. Median OS was superior in maintenance than in observation (36.8 vs 28.0 months, p=0.047). Neurotoxicity (≥ grade 2) was more common in maintenance than in observation without statistical significance (41.7% vs 33.3%, p=0.210). Serial assessment of Quality of Life (QoL) did not show any significant difference between two groups. Conclusions: Maintenance GP chemotherapy for responding patients with MBC showed clinical benefit in terms of PFS and OS without impairment of QoL.

3624 Background: CpG island methylator phenotype (CIMP) is characterized by concurrent methylation of multiple CpG islands in tumor DNA, which can inactivate tumor suppressor genes or promote carcinogenesis. The prognostic impact of CIMP on treatment outcome of colon cancer patients receiving adjuvant 5-fluorouracil/leucovorin/oxaliplatin (FOLFOX) is unclear. We investigated CIMP markers in colorectal cancer patients treated with adjuvant FOLFOX. Methods: Sporadic colorectal cancer patients treated with curative resection followed by adjuvant FOLFOX were included. DNA was extracted from formalin-fixed paraffin-embedded surgical specimen. 8 CpG island loci (CACNA1G, CRABP1, IGF2, MLH1, NEUROG1, P16, RUNX3 and SOCS1) were examined using MethyLight analysis. Disease-free survival (DFS) was evaluated according to each methylation loci. Results: A total of 322 patients were included. Methylation at 1 or more loci was observed in 150 patients (46.6%) and 6 or more loci in 15 (4.7%). During a median follow-up duration of 39.7 months, 55 recurrences were observed. Three year DFS in the patient cohort was 84%. CRABP1 (23.9%) was the most frequently methylated loci, followed by p16 (22.7%) and NEUROG1 (20.8%). Patients having methylation at NEUROG1 (3 year DFS 78% in (+) vs. 86% in (-), p = 0.014) and p16 (3 year DFS 78% in (+) vs. 86% in (-), p = 0.12) had worse DFS, whereas methylation at MLH1 had better DFS (3 year DFS 100% in (+) vs. 86% in (-), p = 0.19). In a combined analysis, patients with MLH1(-)/NEUROG1(+)/p16(+) had worst treatment outcome compared to MLH1(-)/NEUROG1(+) or p16(+), MLH1(-)/ NEUROG1(-) /p16(-), and MLH1(+) (3 year DFS 62%, 82%, 87%, and 100%, respectively; p = 0.002). In multivariate analysis, NEUROG1(+)/p16(+) was associated with significantly higher recurrence compared with other patients (adjusted hazard ratio (HR) 2.15 (95% confidence interval (CI) 1.08 - 4.27, p = 0.029). Conclusions: Methylation status of NEUROG1, p16, and MLH1 is associated with recurrence following adjuvant FOLFOX in stage II/III colorectal cancer. Further validation and translational studies to improve treatment outcome in the subset of patients are warranted in the future.

3076 Background: HM781-36B is a pan-HER tyrosine kinase inhibitor, which showed a potent activity against the gefitinib- or erlotinib-resistant, EGFR L858R/T790M double mutant cells. A phase I study was conducted to determine the MTD, pharmacokinetics, and antitumor activity. Methods: Eligible pts had advanced malignancies refractory to standard therapies. Standard 3+3 scheme was used in the dose escalation part, and additional 12 pts were enrolled in the expansion cohort of molecular enrichment. Results: In dose-escalation part, 43 pts (median age: 55 yrs (range 25-82), M:F=25:18, ECOG PS 0/1/2/3: 23/17/2/1, median prior chemotherapy: 4) were treated. DLTs were G3 diarrheas in 5 pts, one at 12 mg, 16 mg, 24 mg, and two at 32 mg. The MTD was determined as 24mg. The most common drug-related adverse events were diarrhea, stomatitis, rash, pruritus, and anorexia. Among 41 evaluable pts, 4 pts achieved PR (1 unconfirmed, duration of response: 11.9 mo, 7.07 mo+, 4.5 mo+), and 19 pts had SD. Two of 4 PR pts were Her2-positive breast cancer pts. The median duration of treatment in pts with PR or SD was 3.87 (2.47- 15.17) months. In the dose range of 0.5 to 24 mg, it showed linear pharmacokinetics proportional to dose-escalation, relatively short half-life, and little accumulation. Additional 12 pts in the expansion cohort are under treatment at 24 mg (6 pts: EGFR-mutant NSCLC, 3 pts: Her2-positive gastric cancer, 2 pts: Her2-positive breast cancer, 1 pt: rectal cancer). Conclusions: HM781-36B was safe and well tolerable in advanced solid tumors. Preliminary evidence of anticancer activity has been observed. Updated data will be presented at the meeting.

623 Background: Ethnic diversity of genetic polymorphism can result in individual differences in the efficacy and toxicity of cancer chemotherapy. Methods: A total of 292 Korean patients (183 men and 109 women) from six hospitals in Korea were prospectively enrolled. Patients had resected stage III or high-risk stage II colon cancer and were treated with 12 cycles of adjuvant oxaliplatin plus leucovorin plus 5-fluorouracil (FOLFOX) chemotherapy. 20 germline polymorphisms in 10 genes (TS, MTHFR, ERCC1, XPD, XRCC1, ABCC2, AGXT, GSTP1, GSTT1 and GSTM1) were analyzed from peripheral blood. TS genotype in 5’UTR was classified as ‘high’ (2R/3G, 3C/3G, and 3G/3G) or ‘low’ (2R/2R, 2R/3C, and 3C/3C). Results: Most patients (86.3%) received 12 complete cycles of FOLFOX chemotherapy. In contrast to previous studies in Caucasians, neutropenia (grade 3–4, 60.5%) was frequently observed in our Korean patients, whereas only 16.4% experienced grade 2 or more sensory neuropathy. Neutropenia was more frequent in MTHFR 677TT [adjusted odds ratio (OR) 2.32, 95% confidence interval (CI) 1.19–4.55] and ERCC1 19007TT (adjusted OR 4.58, 95% CI 1.20–17.40) genotypes. Patients harboring XRCC1 23885GG had lower risk of neuropathy (adjusted hazard ratio 0.56, 95% CI 0.32-0.99) and longer time to the onset of grade 2-4 neuropathy. MTHFR 677TT and XRCC1 23885GG genotype was also more prevalent in Koreans compared to Caucasians, and the difference of genotypic frequency could partly explain the ethnically different toxicity profile. After median 49.4 months of follow-up, there were 58 (19.9%) relapses and 19 (6.5%) deaths. TS ‘low’ genotype [adjusted hazard ratio (OR) 1.83, 95% CI 1.003–3.34] was significantly related to shorter disease-free survival. Overall survival was not significantly different according to the polymorphisms. Conclusions: Polymorphisms in MTHFR, XRCC1 and TS are related to toxicities and disease-free survival in patients with colon cancer. These polymorphisms may explain the ethnic difference in toxicity profile following adjuvant FOLFOX chemotherapy.

269 Background: In advanced HCC, the role of salvage chemotherapy after failure of prior chemotherapy including sorafenib has not been fully studied. The aims of this study were to evaluate the efficacy of infusional 5-fluorouracil (5-FU), doxorubicin, and mitomycin-C (iFAM) as salvage regimen in advanced HCC and to identify the prognostic factors in a salvage setting. Methods: Eligibility included the following: 1) age > 18 years; 2) Diagnosed as HCC; 3) previously treated with palliative chemotherapy; 4) ECOG PS 0-2; 5) Child-Pough Class A or B; and 6) adequate BM and renal function. iFAM consisted of 5-FU 800 mg/m2 over 10 hours on days 1-5, doxorubicin 30 mg/m2 on day 1 and mitomycin-C 8 mg/m2 on day 1, every 4 weeks. Response evaluation was done every 8 weeks with RECIST criteria. Results: A total of thirty-five patients were enrolled. The median age was 50.2 years (range: 23.3-74.7). ECOG PS was; PS 0-1 (80%) and PS 2 (20%). The etiology of HCC was; HBV (82.8%), HCV (2.9%) and others (14.3%). Liver function was; Child Pough A (91.4%) and B (8.6%). Nine patients had macrovascular invasion and all had extrahepatic metastatic lesion. Prior chemotherapies were sorafenib or 5-FU based regimens. A total of 94 cycles of iFAM were delivered (median 2: range 1-7). There were 2 PR and 6 SD, resulting in RR 5.7% (95% CI, 1.6-18.6%) and DCR 22.9% (95% CI, 12.1-39%). The PFS and OS were 2.4 months and 8.6 months, respectively. The 6- and 12-month survival rates were 67.7% and 35.3%, respectively. Grade 3/4 hematologic toxicity was neutropenia (14.3%). Frequent non-hematological toxicities were; anorexia (45.7%), alopecia (40%), stomatitis (20%), all of which were grade 1/2. In multivariate analysis, PS and serum α-FP were significantly associated with OS. Patients were classified into 3 risk groups based on the number of prognostic factors (0 vs. 1 vs. 2). This classification was well-correlated to OS (16.1 months vs. 10.5 months vs. 3.8 months, respectively; p < 0.0001). Conclusions: iFAM is an effective and tolerable treatment option in advanced HCC and could be considered as salvage regimen. PS and α-FP are significant prognostic factors in a salvage setting.

626 Background: Systemic chemotherapy is still mainstay of treatment of metastatic colorectal cancer (mCRC), while the role of palliative resection is not convincing. We intended to find out the role of palliative resection in mCRC. Methods: A total of 1,015 patients were diagnosed with mCRC at Seoul National University Hospital between 2000 and 2009. Except 169 patients who received curative metastectomy of liver and/or lung, 847 patients were retrospectively analyzed. Results: Out of 847 patients, 556 (65.6%) had metastasis at the time of diagnosis and 291 (34.4%) had recurrence after surgery. The median age was 61 (range, 16-88) and 491 (58.0%) were male. The median number of metastatic site was 1 (range, 1-6) and 738 (87.1%) had limited metastasis (number of metastatic site ≤2). The liver was most frequently involved site (451, 53.2%). Surgery was done in 527 (62.2%). One hundred three patients received resection of both primary and metastatic sites (group 1), while 347 and 78 received resection of primary (group 2) and metastatic sties (group 3), respectively. R0 resection was done in 95 patients (G1: 53, G3: 42), while R1/2 was done 431 in patients (G1: 56, G2: 336, G3:39). Of 95 patients with R0 resection, 93 (97.8%) had limited metastasis (number of metastatic organ ≤2) in the peritoneum, lymph nodes, liver, or lung. The median overall survival (OS) was 19.0 months (95% CI, 17.8-20.1) and resected patients had prolonged median OS compared with patients never resected (21.3 vs 14.1 months, p < 0.001). In multivariate analysis, R0 resection was associated with superior OS compared to R2 resection (51.3 vs 18.7 months; HR, 3.0; 95% CI, 1.86 to 4.85, P Conclusions: Palliative resection with curative intent and chemotherapy confer long-term survival on subsets of mCRC with limited metastasis.

1088 Background: Lapatinib (L) is a dual tyrosine kinase inhibitor of EGFR and HER2. Lapatinib plus capecitabine (C) is an effective treatment option in trastuzumab (T)-refractory HER2-positive (+)...

7173 Background: EGFR mutation is significantly associated with objective response and prolonged survival in NSCLC patients treated with gefitinib. However, presence of mutant nonresponses and nonmutant responses mandates investigation of other molecular markers for more effective prediction of gefitinib sensitivity in NSCLC. It is unclear whether low CA repeat number in EGFR intron 1 has such a predictive role. Methods: Advanced NSCLC patients received gefitinib 250mg/day. EGFR mutation in exons 18 - 21 were identified by direct sequencing of PCR products of DNA extracted from archival paraffin embedded tissue. Number of CA repeat in intron 1 of EGFR was determined by GeneScan with tumoral DNA. Baseline characteristics, mutational status, CA repeat number and efficacy of gefitinib were analyzed in respect to each other. Results: To date, 73 patients were evaluable for EGFR mutation, CA repeat and gefitinib responsiveness. 14 patients (19.2%) harbored EGFR mutation (7 deletion in exon 19, 4 L858R, 1 L861Q, 1 G719A, and 1 insertion in exon 20). Most common CA repeat genotype was 20/20 repeat (31 patients) followed by 16/20 repeat (15 patients). Patients were classified as having either low CA repeat (sum of both allele ≤ 37 repeats) or high CA repeat (≥ 38 repeats). 34 patients (46.6%) had low repeat, whereas 39 patients (53.4%) had high repeat. Patients with EGFR mutation showed better objective response (response rate [RR] 57.1% vs. 10.2% in wild type [WT], p < 0.001), time-to-progression (TTP) (p = 0.031, median 5.1 vs. 1.9 months in WT), and overall survival (OS) (p = 0.051, median14.5 vs. 7.4 months in WT). In the whole study population, low CA repeat was not associated with RR (p = 0.38), TTP (p = 0.15), or OS (p = 0.51). However, in the 59 patients without an EGFR mutation, patients with low CA repeat tended to have better objective response (RR 17.2% [5/29] in low repeat vs. 3.3% [1/30] in high repeat, p = 0.10) and significantly better TTP (p = 0.019, median 2.2 months in low repeat vs. 1.2 months in high repeat). Conclusions: Our results suggest that low number of CA repeats in EGFR intron 1 may have possible role in prediction of gefitinib responsiveness when analyzed together with EGFR mutational status. [Table: see text]