Search

613 Background: Although growing number of studies have explored the role of vitamin D3 on various cancer types, no study has been conducted in the prospective cohort of advanced BTC. This analysis aimed to evaluate the predictive or prognostic implication of serum vitamin D3 in pts with advanced BTC treated with second-line chemotherapy. Methods: The analysis is the preplanned subgroup analysis of the NIFTY trial, a multicenter, open-label, randomized, phase 2b study which compared the second-line liposomal irinotecan plus 5-FU/leucovorin and 5-FU/leucovorin in pts with advanced BTC. Serum vitamin D3 (25-hydroxyvitamin D [25(OH)D]) levels were measured at baseline (i.e., C1D1) in a total of 176 pts out of 178 pts in the ITT population and were included in the current analysis. Correlation between OS and serum vitamin D3 as a continuous variable was analyzed for linear and non-linear associations. Results: The median age was 64 years (range, 37–84) and 75 patients (42.6%) were women. The baseline serum vitamin D3 levels ranged from 1.39 ng/ml to 140.02 ng/ml in the overall pts. There was no significant difference in the mean vitamin D3 concentration according to the sex (male vs. female: 24.1 vs. 26.7; p=0.43), body mass index (BMI) (3 and OS (HR=1.06 per 10 increase; 95% CI, 0.97–1.15; p=0.22). In addition, there was no non-linear association between vitamin D3 and OS (p=0.52 for the linear model and p=0.62 for the model with restricted cubic splines). In the subgroup analysis, higher vitamin D3 levels were associated with poorer OS (HR=1.16 per 10 increase; 95% CI, 1.04–1.30; p=0.0072) in female pts. Conclusions: Serum vitamin D3 levels did not show significant relationship with survival in pts with advanced BTC treated with second-line chemotherapy. High vitamin D3 levels might be associated with poorer survival in female pts. Further larger studies are needed.

737 Background: There is no reliable biomarker to guide treatment for patient with borderline resectable pancreatic ductal adenocarcinoma (PDAC) treated with neoadjuvant modified (m)FOLFIRINOX. We investigated the role of plasma circulating tumor DNA (ctDNA) sequencing to discover potential biomarkers for patients treated with neoadjuvant mFOLFIRINOX in our phase 2 clinical trial (ClinicalTrials.gov identifier: NCT02749136). Methods: Among 44 patients enrolled in the trial, patients with plasma ctDNA analysis performed at baseline or post-operative period were included in the analysis. Plasma cell-free DNA isolation and sequencing of ctDNA were performed using the Guardant 360 assay. Association of clinical outcomes with germline or somatic mutations of DNA damage repair (DDR) genes was analyzed. Somatic alterations other than DDR genes were also analyzed to find potential biomarkers of neoadjuvant mFOLFIRINOX. Progression free-survival (PFS) was defined as time from the initiation of mFOLFIRINOX to disease progression or any cause of death. Overall survival (OS) was defined as the time from the initiation of mFOLFIRINOX to any cause of death. Results: Among 44 patients, 28 patients had ctDNA sequencing data qualified for the analysis and were included in this study (15 patients with baseline data only, 3 patients with post-operative data only, and 10 patients with both baseline and post-operative data). Median age was 59 (range, 41-73) years and 15 patients (53.6%) were male. Among 25 patients with baseline plasma ctDNA data, 11 patients (44%) had somatic or germline alteration of DDR genes detected at baseline including ATM, BRCA1, BRCA2 and MLH1, and showed significantly better PFS than those without such DDR gene alterations (median 26.6 vs. 13.5 months, log-rank p = 0.004). Although there was a trend for better OS in patients with DDR gene alterations, this was not statistically significant (log-rank p = 0.29). The most observed baseline somatic alteration was TP53 (n = 7, 28%), followed by KRAS (n = 6, 24%). Patients with somatic KRAS mutation at baseline had significantly worse OS (median 8.5 months vs. not applicable, log-rank p = 0.003) than those without. Among 13 patients with post-operative plasma ctDNA data, 8 patients (61.5%) had detectable somatic alterations. Conclusions: Detection of germline or somatic mutations of DDR genes from plasma ctDNA at baseline was associated with better survival outcomes of patients with borderline resectable PDAC treated with neoadjuvant mFOLFIRINOX. Detection of DDR gene alterations from plasma ctDNA may have a potential role as biomarker for prognosis in patients with borderline resectable PDAC treated with mFOLFIRINOX.

1098 Background: Combining immune checkpoint inhibitors with chemotherapy has become a promising therapeutic strategy in metastatic breast cancer. Eribulin is a potent microtubule inhibitor and modulates the immune microenvironment of tumor cells. Therefore, combining eribulin to nivolumab may synergize antitumor efficacy in metastatic breast cancer. Methods: Adult patients with histologically confirmed recurrent/metastatic HER2- breast cancer were enrolled prospectively from 10 academic hospitals in Korea (ClinicalTrials.gov Identifier: NCT04061863). Key eligibility criteria included prior treatment with taxanes and/or anthracyclines, ≥1 measurable disease, and ≤2 prior cytotoxic chemotherapies in the metastatic setting. Patients received nivolumab 200 mg intravenously (IV) on day 1 plus eribulin 1.4 mg/m2 IV on day 1 and 8 of every 3 weeks until disease progression or intolerable toxicity. The dose level was determined from safety profile of three patients in run-in phase. The primary endpoint was investigator-assessed progression-free survival (PFS) rate at 6 months. Secondary endpoints included investigator-assessed objective response rate (ORR) per RECIST v1.1, disease control rate (DCR), overall survival (OS), and toxicity profile of the combination treatment. The association between PD-L1 expression by SP142 Ab and efficacy was analyzed. Results: From August 2019 to June 2021, 90 patients (HR+HER2- 45 pts/TNBC 45 pts), with a median age of 51 (range 31–71), were enrolled in the study. With a median study follow-up time of 16.3 months, 68 (75.6%) patients experienced progressive disease. PFS rate at 6-months was 49.6% and 24.1% in patients with HR+HER2- and TNBC group, respectively. Median PFS was 5.6 months (95% CI: 4.3-6.8) and 3.0 months (95% CI: 1.3-4.7) for HR+HER2- and TNBC group, respectively. ORRs were 53.3% (CR:0, PR: 24) for HR+HER2- and 21.8% (CR1, PR: 12) for TNBC. Patients with PD-L1+ tumors (PD-L1 expression ≥ 1% on TC or IC) had similar ORR compared to PD-L1- tumors (ORR 50% vs. 53.8% in HR+HER2-, 30.8% vs. 29.0% in TNBC). The most common grade 3/4 adverse event was neutropenia (15/90, 16.7%), and the most common immune-related adverse events were grade 1/2 hypothyroidism (19/90, 21.1%) and grade 1/2 pruritus (16/90, 17.8%). Five patients had discontinued study treatment due to immune-related adverse events (3 pneumonitis, 1 hepatitis, 1 skin rash). Conclusions: In this parallel phase II clinical trial, the addition of nivolumab to eribulin showed promising efficacy and tolerable safety profile in previously treated HER2- MBC. Further survival and exploratory analyses to find predictive markers will be followed. Clinical trial information: NCT04061863. [Table: see text]

547 Background: For patients who underwent curative-intent upfront surgery, adjuvant chemotherapy is the current standard of care. The previous, randomized phase 3 ESPAC-4 study showed significantly improved overall survival (OS) with GEMCAP compared to GEM. However, this study was conducted in European countries and its implication in Asian patients has not been explored yet. We conducted a retrospective analysis to evaluate the efficacy and safety of GEMCAP compared to GEM regimen. Methods: Between January 2017 and December 2020, a total of 292 patients who received adjuvant GEM or GEMCAP after curative-intent surgery in Asan Medical Center, Seoul, Korea, were included in this retrospective analysis. Results: Adjuvant GEM and GEMCAP were administered in 161 patients (55.1%) and 131 patients (44.8 %), respectively. Compared the GEMCAP group, age of patients were significantly older in the GEM group (median 66 vs 63 yo, p = 0.025); otherwise, there was no significant difference in baseline characteristics between two groups. With the median follow-up duration of 39.4 months (95% CI 36.9 - 45.0 months) in GEM group and 39.4 months (95% CI 34.7-41.6 months) in GEMCAP group, the median OS was 36.8 months (95% CI 29.7-43.5 months) and 46.1 months (95% CI 31.5 months – not reached) in the GEM group and GEMCAP group, respectively (unadjusted HR 0.72, 95% CI 0.51-1.02, p = 0.065). The median recurrence-free survival was 14.3 months (95% CI, 12.9-17.7 months) and 17.0 months (95% CI, 13.3-28.8 months) in the GEM group and GEMCAP group, respectively (p = 0.52). In the GEMCAP group, hand-foot skin reaction (any grade, 15.27 % vs 0.62 %, p < 0.001), neutropenia (78.6% vs 67.7%, p=0.037) and thrombocytopenia (30.53% vs 20.5%, p=0.035) were more common in the GEMCAP group compared to the GEM group. In multivariate analysis, adjuvant GEMCAP was significantly associated with better OS compared to adjuvant GEM (adjusted HR 0.64, 95% CI, 0.44-0.91, p = 0.014). Otherwise, moderate or poor histologic grade, lymph node positive, positive resection margin, and elevated CA 19-9 levels (> median) were significantly associated with poorer OS. Conclusions: In this retrospective analysis for Korean patients, adjuvant GEMCAP showed consistent clinical outcomes shown in the ESPAC-4 trial. As mFOLFIRINOX is the new standard of care for medically fit patients with resected pancreatic adenocarcinoma, further evaluation of optimal adjuvant chemotherapy in daily practice is warranted.

4006 Background: There is no globally established second-line therapy after progression on GemCis in BTC. Although ABC-06 trial showed the clinical benefit of mFOLFOX compared to active symptom control, further investigation is needed. Methods: NIFTY is an investigator-initiated, multicenter, open-label, randomized, phase 2b study. Pts with > 19 years, ECOG PS 0/1, histologically confirmed metastatic BTC, and disease progression on first-line GemCis were eligible. Pts were randomized 1:1 to nal-IRI (70 mg/m2, 90 min) plus 5-FU (2400 mg/m2, 46 hours)/LV (400 mg/m2, 30 min), every 2 weeks or 5-FU/LV, every 2 weeks until disease progression per investigator review or intolerable toxicities (stratification: primary tumor site, prior surgery and institution). Tumor response was evaluated per RECIST v1.1, every 6 weeks (fixed schedule). Primary endpoint is progression-free survival (PFS) per blinded independent central review (BICR). Secondary endpoints were PFS per investigator review, overall survival (OS) overall response rates (ORR), and safety. This study was designed to improve median PFS from 2 months (P0) to 3.3 months (P1; HR 0.6) with 2-sided alpha of 0.05, power of 80% and follow-up loss rates of 10%; a total of 174 pts were required. Results: A total of 178 patients were enrolled between SEP 2018 and FEB 2020; with exclusion of 4 pts who did not receive any study treatment, 174 pts (88 for nal-IRI plus 5-FU/LV group and 86 for 5-FU/LV group) were included in the Full Analysis Set. Median age was 64 yrs (range 37-84); 99/75 pts were male/female; 74/47/53 pts had intrahepatic/extrahepatic/gallbladder cancers. Pts characteristics were well balanced between two arms. With median follow-up duration of 6.1 mo (IQR 3.5-11.2), median PFS per BICR in nal-IRI plus 5-FU/LV group and 5-FU/LV group was 7.1 mo (95% CI, 3.6-8.8) and 1.4 mo (1.2-1.5), respectively (HR=0.56 [0.39-0.81], p=0.0019); median PFS per investigator review was 3.9 mo (2.7-5.2) and 1.6 mo (1.3-2.2), respectively (HR=0.48 [0.34-0.69], p

382 Background: Liposomal irinotecan (nal-IRI) plus fluorouracil/leucovorin (5-FU/LV) has demonstrated its clinical benefit in patients with metastatic pancreatic adenocarcinoma (mPAC) who progressed on prior gemcitabine-based chemotherapy. However, its role in patients who previously treated with conventional irinotecan has not been investigated. We analyzed clinical outcomes of nal-IRI + 5-FU/LV in mPAC patients after progression on conventional irinotecan-containing chemotherapy. Methods: In this multicenter retrospective analysis, a total of 35 patients with mPAC who received nal-IRI + 5-FU/LV after progression on irinotecan-containing regimen, between January 2017 and March 2020, were included. The ratio of time-to-progression (TTP) with nal-IRI + 5-FU/LV to TTP with conventional irinotecan (TTPr) was correlated with duration and cumulative dose of prior conventional irinotecan. Results: The median age was 58 years (range, 35-73) and 16 patients (46%) were male. All patients received prior irinotecan as the component of FOLFIRINOX. The median duration of prior irinotecan was 4.6 months (range, 0.5-16.8) and median cumulative dose of prior irinotecan was 1230 mg (range, 150-4650). Objective response rate of nal-IRI + 5-FU/LV was 2.9% (1 partial response) and stable disease was achieved in 31.4% (n = 11). With median follow-up duration of 9.2 months [95% CI, 7.8-10.5], the median PFS and OS were 2.0 months [95% CI, 1.4-2.6] and 4.4 months [95% CI, 3.6-5.7], respectively. 6-month PFS rate was 16.3% and OS rate was 37.5%. The median TTPr was 0.41 (range 0.07-2.07) and this showed negative correlation between cumulative dose of prior irinotecan (R = -0.37, p = 0.041). There was a tendency for the negative correlation between TTPr and duration of prior irinotecan (R = -0.35, p = 0.062). Most common grade 3-4 toxicities were neutropenia (20%) and fatigue (8.6%). Conclusions: Nal-IRI + 5-FU/LV showed only modest efficacy for mPAC patients who progressed on conventional irinotecan-containing chemotherapy. Cumulative dose of prior conventional irinotecan may be correlated with the efficacy of nal-IRI + 5-FU/LV.

720 Background: Borderline resectable pancreatic cancer (BRPC) and locally advanced unresectable pancreatic cancer (LAUPC) are heterogeneous disease entity with various prognosis. Based on the phase III PRODIGE trial, (m)FOLFIRINOX has been widely used for the management of patients with BRPC and LAUPC. Considering the lack of large phase 3 trial of (m)FOLFIRINOX for BRPC and LAUPC, real-life evidences of (m)FOLFIRINOX are needed. Methods: In this retrospective analysis, 199 patients who received at least one dose of (m)FOLFIRINOX between February 2013 and January 2017 were included. Endpoints of this study were objective response rates (ORR), surgical resection rate, progression-free survival (PFS) and overall survival (OS). Results: Median age was 60 years (range, 33-79) and 62.3% of patients were male. Pancreas head (n=112, 56.3%) was the most common primary tumor site, followed by body (n=42, 21.1%) and multifocal (n=34, 17.1%). By an independent radiology review, patients were classified to BRPC (n=75, 37.7%) and LAUPC (n=124, 62.3%). With median 40.3 months (95% CI, 36.7-43.8) of follow-up duration in surviving patients, ORR was 26.6% (n=53), median PFS and OS were 10.6 months (95% CI, 9.5-11.7) and 17.1 months (95% CI, 13.2-20.9), respectively. There was no difference in PFS and OS between BRPC and LAUPC (median PFS, 11.1 months [95% CI, 8.8-13.5] vs. 10.1 months [95% CI, 8.4-11.8], p=0.47); (median OS, 18.4 months [95% CI, 16.1-20.8] vs. 17.1 months [95% CI, 13.2-20.9], p=0.50). Curative-intent surgery (R0 and R1) was done in 63 patients (33.2%, 49 for R0 and 14 for R1) after treatment with (m)FOLFIRINOX. Resection rates were 58.2% in BRPC patients and 19.4% in LAUPC patients (p

689 Background: FOLFIRINOX is one of standard 1st-line regimens for patients (pts) with advanced PDAC. However, there is no globally established 2nd-line regimen after failure of FOLFIRINOX. Although gemcitabine-based regimens are recommended by multiple guidelines and widely used in daily practice, further analysis is needed to reveal the magnitude of clinical benefit with these regimens. Nab-P+Gem is another standard 1st-line regimen for PDAC, but there are limited data as 2nd-line therapy in PDAC. Therefore, we conducted multicenter retrospective analysis of 2nd-line nab-P+Gem after progression on FOLFIRINOX in pts with advanced PDAC. Methods: Between Feb 2016 and Feb 2019, a total of 103 pts with histologically documented PDAC who received nab-P+GEM after progression on 1st-line FOLFIRINOX were identified among 5 referral cancer centers in South Korea. Results: Median age was 60 years and 50 pts (49%) were male. All but one pts had ECOG performance status of 0-1 at the time of nab-P+GEM. At the time of nab-P+GEM, 25 (24%) and 78 (76%) patients had locally advanced and metastatic disease, respectively. Median overall survival (OS) and progression-free survival (PFS) with nab-P+GEM was 9.8 months (95% CI: 8.9-10.6) and 4.6 months (95% CI: 3.7-5.5), respectively. Among pts with measurable disease (n = 95), partial response and stable disease were achieved in 8 (8%) and 56 (54%), respectively. Median OS from the start of 1st-line FOLFIRINOX was 20.9 months (95% CI: 15.2-26.6). Most common adverse event of all grade was anemia (77%), followed by neutropenia (60%), fatigue (52%), thrombocytopenia (45%), and peripheral neuropathy (30%). Most common grade 3-4 adverse events were neutropenia (36%), anemia (9%), and peripheral neuropathy (8%). Conclusions: In medically fit pts with advanced PDAC who failed on 1st-line FOLFIRINOX, nab-P+GEM was effective and well tolerated as 2nd-line therapy.

4078 Background: Although many studies using whole-exome sequencing or targeted sequencing have reported the molecular profile of BTC, its clinical implications remains unclear. In this study, we assessed a predictive role of DDR gene mutations in advanced BTC patients treated with platinum-containing regimen. Methods: Eighty-eight patients with pathologically-confirmed BTC who received first-line gemcitabine-cisplatin combination (n = 69) or fluoropyrimidine-oxaliplatin combination (n = 19) were included in this analysis. Targeted exome sequencing was performed using Foundation Medicine T7 assay or in-house OncoPanel AMC. Germline or somatic mutations in ATM, ATR, BAP1, BARD1, BRCA1, BRCA2, BRIP1, CHEK2, FAM175A, GEN1, MLH1, MSH2, MSH6, MRE11A, NBN, PALB2, PMS2, RAD50, RAD51, RAD51C, RAD51D, and XRCC2 were classified as DDR gene mutations. Data regarding baseline characteristics and treatment outcomes were retrospectively obtained from medical records. Results: The median age was 62 years (range, 25-78), with male comprising 64.8% (n = 57). By primary tumor site, 21 patients with GBC (23.9%), 44 with ICC (50.0%) and 23 with ECC (26.1%) were included. Most patients received palliative chemotherapy for their initially metastatic (50.0%) or recurred (44.3%) disease; the rest 5.7% had locally advanced disease. The median PFS and OS of overall patients were 7.1 and 16.1 months, respectively with median follow-up duration of 20.2 months. DDR gene mutations were found in 63.5% of patients. BRCA2 (18.2%) was most frequently mutated, followed by ATM (13.6%), and ATR (8.0%). DDR gene mutations were significantly associated with prolonged PFS (presence vs. absence; median, 6.9 vs. 5.7 months; P = 0.013) and OS (median, 21.0 vs. 13.3 months, P = 0.009). The impact of DDR gene mutations remained significant in multivariate analyses for PFS that included other prognostic factors (hazard ratio, 0.51; P = 0.009), but not for OS. Conclusions: The presence of DDR gene mutations might be a promising predictive biomarker for response to platinum-based chemotherapies in advanced BTC. Future investigation using novel agents targeting DDR gene alteration in BTC are warranted.

4082 Background: For patients with advanced BTC, standard chemotherapy has limited benefit and no molecular targeted agents have been approved. Pembrolizumab is an anti PD-1 immune checkpoint inhibitor which has shown modest activity for advanced BTC patients in prior single-arm phase I/II studies. Considering the heterogeneity of BTC, more data are needed to evaluate the clinical outcomes of pembrolizumab in unresectable or metastatic BTC. Methods: In this prospective cohort study, 39 patients with PD-L1 positive BTC who received pembrolizumab in Asan Medical Center, Seoul, Korea were included (ClinicalTrials.gov identifier, NCT03695952). PD-L1 expression was assessed using immunohistochemistry and PD-L1 positive tumors were defined as the expression of PD-L1 in ≥ 1% of tumor cells. Pembrolizumab was given at a fixed dose of 200 mg intravenously, every 3 weeks. Results: The median age was 61 years old (range, 41-76) and 22 (56.4%) patients were male. Intrahepatic cholangiocarcinoma (CCA) was the most common type (n = 18, 46.2%), followed by gallbladder cancer (n = 12, 30.8%) and extrahepatic CCA (n = 9, 23.1%). Most of the patients had distant metastasis (n = 37, 94.9%). Pembrolizumab was administered as 2nd-, 3rd- and 4th or greater line chemotherapy in 18 (46.2%), 16 (41.0%) and 5 (12.8%) patients, respectively, and median 2 cycles (range 1-10) of pembrolizumab were given. In 36 patients whose response was assessable, partial response (PR) and stable disease were achieved in 4 (11.1%) and 13 (36.1%), respectively. In 19 (52.8%) patients, progressive disease was the best response. In patients with PR, the median time to response was 2.1 months (95% confidence interval (CI), 0.4 – 3.9). With a median follow-up duration of 4.4 months (95% CI, 2.4 – 6.4), median progression-free survival and overall survival was 1.5 months (95% CI, 0.4 – 2.6) and 4.3 months (95% CI, 2.6 – 6.1), respectively. No grade 3/4 adverse events (AEs) were reported and grade 1/2 fatigue (n = 4, 10.3%) was the most common AE. Conclusions: In PD-L1 positive BTC, pembrolizumab showed modest efficacy with 11.1% of response rates although our patients were heavily pretreated. Considering the limited therapeutic options and poor survival for these patients, further evaluation of immunotherapy including biomarker analysis is needed.

312 Background: Although neoadjuvant chemotherapy (NACT) has been widely investigated, the magnitude of the clinical benefit and the potential risk of NACT followed by surgery compared with upfront surgery remains unclear for patients with locally advanced pancreatic cancer (LAPC). Methods: This retrospective, prospective cohort-based analysis included 135 patients who underwent NACT followed by surgery and 359 patients who received upfront surgery for LAPC between October 2005 and April 2017. Disease-free survival (DFS) and overall survival (OS) from surgery were compared between the two groups. Results: There were no significant differences in gender (male, 53% vs 56%) and age (median 60 vs 61 years) between the NACT followed by surgery group and upfront surgery group. As NACT, gemcitabine-based regimens and FOLFIRINOX were used in 69 (51%) and 66 (49%) patients, respectively. The NACT followed by surgery group showed significantly less advanced T stage (T3–4, 93% vs 99%, p = 0.001) and N stage (N+, 49% vs 71%, p < 0.001) than the upfront surgery group. NACT followed by surgery was significantly associated with better OS (median, 25.4 [18.6–32.2] vs 17.1 [15.5–18.7] months, p = 0.001) and DFS (median, 9.0 [95% CI, 6.8–11.2] vs 7.1 [6.4–7.8] months, p = 0.005) than upfront surgery. These results were consistent in the multivariate analysis for OS (adjusted hazard ratio [aHR], 0.73 [95% CI, 0.56–0.96], p = 0.02) and DFS (aHR, 0.72 [95% CI, 0.56–0.93], p = 0.01). There was no difference in length of hospital stay (median 13 vs 17 days, p = 0.14) for surgery between the two groups, and the NACT followed by surgery group showed a significantly lower incidence of postoperative complication than the upfront surgery group (38% vs 27%, p = 0.03). Conclusions: The present study revealed that NACT followed by surgery may provide survival benefit compared with upfront surgery in LAPC without causing significant safety issues.

e21150Background: Despite osimertinib revealed marked antitumor activity in EGFR T790M mutant lung adenocarcinoma, second tumor biopsy is inevitably required for detection of T790M mutation. We inv...

357 Background: Gemcitabine plus cisplatin (GP) is the standard first line chemotherapy for patients with advanced biliary tract cancer (BTC). In the pivotal ABC-02 study, patients received up to 24 weeks (6-8 cycles) of three-weekly GP. In daily practice setting, however, patients without progression often receive GP more than 6-8 cycles. It is uncertain whether maintenance treatment has clinical benefit in patients without progression on GP. Methods: Advanced BTC patients treated with GP between April 2010 and February 2015 in Asan Medical Center, Seoul, Korea, were retrospectively analyzed. Among the patients who did not progressed and stopped GP after 6-8 cycles, patients were stratified according to the further treatment; those with or without further cycles of GP (maintenance group vs observation group). Primary endpoint was overall survival (OS). Results: Among 740 patients, 231 patients (31.2%) were eligible for this analysis; 111 for observation group, 120 for maintenance group. In observation group, 76 patients (68.5%) stopped GP due to completion of scheduled chemotherapy and 27 patients (24.3%) due to the patients’ request or toxicity. There were no statistically significant differences in baseline characteristics between two groups. Median OS from the initiation of GP was 20.5 months [95% CI 15.4-25.6] and 22.4 months [95% CI 17.0-27.8] in the observation and maintenance group, respectively (p = 0.32). Median progression-free survival (PFS) was 10.4 months [95% CI 7.0-13.8] and 13.2 months [95% CI 11.3-15.2], respectively (p = 0.22). These were consistent in the multivariate analyses for OS and PFS after the adjustment of prognostic factors. Conclusions: In our analysis, maintenance therapy of GP was not associated with improved survival outcomes. Considering the potential disadvantages such as cumulative toxicities, maintenance therapy may not be beneficial in patients who did not progressed on 6-8 cycles of GP.

354 Background: AG and FOLFIRINOX have been established as standard first-line treatment in mPC patients based on the improved efficacy compared to gemcitabine monotherapy. Because there was no head-to-head comparison between these regimens, however, there is lack of data which regimen is preferable in patients with mPC. Therefore, we performed retrospective analysis comparing the efficacy of AG and FOLFIRINOX in daily practice setting. Methods: We analyzed a total of 308 patients with confirmed mPC who received AG (n = 149) or FOLFIRINOX (n = 159) as first-line treatment between 2013 and 2016 at Asan Medical Center, Seoul, Korea. Primary endpoint was overall survival (OS) and secondary endpoints were progression-free survival (PFS) and overall response rates (ORR). Results: Median age was slightly older in AG group than FOLFIRINOX group (62 vs. 60 years, p = 0.02). Except this, there was no significant difference between the two groups in terms of baseline characteristics including sex (male, AG/FOLFIRNOX): 56%/67%, ECOG performance status (0-1): 97%/99%, pancreatic head location: 32%/40%, and baseline CA19-9 level ( > UNL): 77%/82%. ORR (28% vs. 31%) and disease control rate (74% vs. 74%) did not differ between two groups (p = 0.45, and p = 0.96, respectively). Although there was no significant difference in PFS between the two groups (AG: median 6.8 months [95% CI: 5.8-7.8] vs. FOLFIRINOX: 5.1 months [95% CI: 4.2-5.9]; p = 0.15), OS was significantly better in AG group compared to FOLFIRINOX group (median 12.3 months [95% CI: 11.0-13.7] vs. 9.7 months [95% CI, 8.1-11.4]; p = 0.001). Among the patients who showed progression, 81% (180/223) received second-line chemotherapy and there was no difference between the two groups (AG vs. FOLFIRINOX: 76% vs. 85%, p = 0.09). 5-FU monotherapy or combination with oxaliplatin were given in 97% (73/75) of patients in AG group and gemcitabine-based regimens (including 2 cases of AG) were given in 97% (102/105) of patients in FOLFIRINOX group. Conclusions: Both AG and FOLFIRINOX are feasible and active as first-line treatment for patients with mPC. In daily practice setting, AG showed comparable efficacy outcomes with FOLFIRINOX.

217 Background: Androgen deprivation therapy (ADT) has been the main treatment option for patients with locally advanced or metastatic prostate carcinoma. However, the impact of continued ADT on clinical outcomes in castration-resistant prostate cancer (CRPC) patients treated with cytotoxic chemotherapy is controversial. We conducted a randomized phase II study to assess the impact of continued ADT in patients with CRPC receiving docetaxel plus prednisolone chemotherapy. Methods: Patients with CRPC were randomly assigned (in a 1:1 ratio) to receive docetaxel (75mg/m2 every 3 weeks) plus prednisolone (5mg twice daily on days 1-21) with ADT (leuprolide 11.25 mg long-acting depo every 12 weeks, ADT group) and docetaxel plus prednisolone without ADT (No-ADT group). For patients in the No-ADT group, ADT was resumed at the end of chemotherapy. The primary endpoint was time to PSA progression. Secondary objectives included PSA response rates, progression-free survival (PFS), and overall survival (OS). Suspension of leuprolide support prevented the attainment of our original goal of enrolling 80 men. Results: Between April 2011 and July 2014, 45 patients were enrolled in this study; 23 patients were randomly assigned to the ADT group and 22 to the No-ADT group. The median age was 68 years (range = 49-81) and median baseline serum PSA was 60.3 ng/mL (interquartile range = 14.7-157.3). Time to PSA progression was 6.5 months in ADT group and 4.3 months in No-ADT group, respectively (P = 0.673). PSA response rates were 55.2% and 44.8% in the ADT group and the No-ADT group, respectively (P = 0.463). The median PFS and OS were 5.3 months and 19.4 months for ADT patients and 4.3 months and 20.8 months for No-ADT patients, respectively (P = 0.608, P = 0.635). The testosterone level rose to more than the castrate level in 5 (22.7%) patients of the No-ADT group. Conclusions: Clinical outcomes were not significantly different when patients with CRPC received concurrent ADT, or were not so treated, when receiving docetaxel plus prednisolone chemotherapy. Clinical trial information: NCT01487902 Clinical trial information: NCT01487902.

121 Background: Adjuvant chemotherapy with S-1 has been proven effective for patients with curatively resected advanced gastric cancer with D2 lymph node dissection in Japan. We assessed the safety and feasibility of adjuvant S-1 chemotherapy in Korean patients with stage II, III or IV(M0) gastric cancer. Methods: A total of 305 patients with stage II, III or IV(M0) gastric cancer received adjuvant S-1 chemotherapy following curative gastrectomy with D2 lymph node dissection in Asan Medical Center between October 2007 and December 2009. Adjuvant chemotherapy with S-1 was started 3-6 weeks after surgery and it was administered orally twice daily at the dose of 40 mg/m2 for 4 weeks followed by 2 weeks of rest, every 6 weeks 8 times for 1 year. We retrospectively reviewed the medical records of the patients and evaluated the safety and feasibility of adjuvant S-1 chemotherapy in Korean patients. Results: Among the 305 patients, 248 (81.3%) and 198 (64.9%) patients completed 4 and 8 cycles of adjuvant chemotherapy, respectively. The most common reasons for discontinuation of treatment were adverse event (43.9%) and recurrence (26.2%). Among the 305 patients, 75 (24.6%) patients required dose reduction because of toxicities. The most common grade 3/4 toxicities were neutropenia (12.8%), diarrhea (5.3%), abdominal pain (3.8%), and anemia (3.3%). Multivariate analysis showed that total gastrectomy (H.R. 2.50; 95% C.I. 1.32-4.72, p=0.005) and female gender (H.R. 1.95; 95% C.I. 1.03-3.66, p=0.039) were independent risk factors for grade 3/4 hematologic toxicities, and old age (>65 years) (H.R. 2.92; 95% C.I. 1.50-5.69, p=0.002) was an independent risk factor for grade 3 non-hematologic toxicities. Conclusions: Adjuvant chemotherapy with S-1 for 1 year was safe and feasible in Korean patients. Old age, female gender, and total gastrectomy were independent risk factors for severe toxicities of adjuvant S-1 chemotherapy.