4104 Background: Gut microbiota established crucial roles in host metabolism and several diseases, particularly cancers. Distinct bacterial profiles from intestine are found to be a potential factor in carcinogenesis, while some of those are associated with detrimental treatment to the responsiveness in various types of cancer. Association of gut microbiota and treatment outcome has been thoroughly investigated in intrahepatic cholangiocarcinoma (ICCA). The present study aimed to compare gut microbiota profiles between chemotherapy responder and non-responder in ICCA patients. Methods: Unresectable or metastatic ICCA patients were recruited in the study. The criteria of all patients were naïve to chemotherapy. All patients received first line combination of cisplatin 25 mg/m2 and gemcitabine 1000 mg/m2 at day1 and day 8 in every 21-day cycle and were given up to 8 cycles. The primary endpoint was to evaluate the association between gut microbiota and the objective response rate. Bacterial genomic DNA samples were extracted from the stool using a commercial genomic DNA isolation kit (Qiagen, Hilden, Germany) and then underwent next-generation sequencing. Data analyses were conducted by using QIIME2 and Analysis of Compositions of Microbiomes with Bias Correction (ANCOM-BC). Results: Seventeen ICCA patients were recruited. The objective response rate was 23.5% comparable to historical study. Chemotherapy responder group was defined as partial response patients (N = 4) and chemotherapy non-responder group included stable of disease and progressive disease (N = 13). Baseline characteristics were similar including age, sex, ECOG performance status, numbers of metastatic organ and CA 19-9 in two groups. Median Age were 63 years old (60-70) and 61 (45-73), respectively. Alpha-diversity and beta-diversity were not different between two groups. In chemotherapy non-responder group, some taxa were dominantly observed including Ruminococcaceae, Oribacterium, Oxalobacter, Peptostreptococcus, Aggregabacter, which had been previously reported increasing in ICCA patients. Interestingly, Ruminococcaceae, which was previously documented to be correlated with vascular invasion, was significantly increased abundance in non-responder group. Moreover, the median progression-free survival (PFS) was improved in chemotherapy responder group compared to non-responder group 11.8 and 6.6 month, respectively. Large sample sizes study to confirm this study result should be developed. Conclusions: Distinct bacteria from gut microbiota such as Ruminococcaceae has inversely associated with chemotherapy response in ICCA patients receiving first line cisplatin and gemcitabine combination, suggesting the possibility of the poorer PFS.