Mo1270 THE ROLE OF EUS-GUIDED TISSUE ACQUISITION FOR GENOMIC ANALYSIS OF PANCREATIC ADENOCARCINOMA: IMPLICATIONS FOR PERSONALIZED MEDICINE

Genomic sequencing of pancreatic adenocarcinoma has emerged as a promising tool for personalized medicine, providing valuable prognostic information and potentially tailored therapeutic regimens. Endoscopic ultrasound (EUS) is now considered the primary modality for diagnosing and locally staging pancreatic cancer. However, it is not clear if EUS-guided tissue acquisition (EUS-TA) is reliably able to provide samples for genomic analysis. To determine the ability of EUS-TA to obtain sufficient tissue from pancreatic adenocarcinoma for genomic analysis and to compare the yield of fine needle aspiration (FNA) to fine needle biopsy (FNB). Patients at an academic tertiary care center who underwent either EUS-FNA (EUSN-3; Cook Medical) or FNB (SharkCore; Medtronic) with a 22-gauge needle and a final diagnosis of pancreatic adenocarcinoma from March 1, 2016 to November 15, 2017 were included in this retrospective study. Demographic and clinical data were extracted from the medical record and endoscopy database. Samples with a minimum of 10% tumor cellularity were determined to be sufficient for genomic analysis. Genomic testing was performed on the rinse material if the tissue was considered sufficient and the results were requested for clinical management. Genomic analysis entailed targeted next-generation sequencing with a panel for 47 commonly mutated genes in solid tumors. Wilcoxon rank-sum and Fisher’s exact tests were used to compare continuous and categorical variables, respectively, with an alpha = 0.05 level regarded to be statistically significant. The cohort consisted of 65 patients with pancreatic adenocarcinoma diagnosed by EUS-TA. Baseline characteristics included: median age 66.7 years (IQR 60.5, 73.7), 52% male sex, and 69% Caucasian race. Family history for pancreatic cancer was noted in 23% of patients, 42% had stage IV disease, and the median CA 19-9 at diagnosis was 326 U/mL (IQR 70, 2524). EUS-TA samples were sufficient for genomic analysis in 31/65 (48%). The odds of sufficient sampling for genetic testing were found to be 4.92-fold higher with FNB as compared to FNA (95% CI 1.28 – 18.52, p = 0.0183). Tumor location, tumor size, number of passes, and method of suction were not significantly associated with sufficient tissue to perform genomic analysis (Table 1). Of those samples undergoing genomic testing, the median number of mutations per tumor was 3 (IQR 2, 5). KRAS and TP53 mutations were the most frequently noted in 78% of tumors. Genomic sequencing of pancreatic adenocarcinoma using EUS-TA rinse material is feasible in a significant proportion of patients. EUS-FNB is associated with significantly increased odds of sufficient sampling for genomic testing as compared to EUS-FNA.Tabled 1Table 1: Possible Predictors of Adequate Sampling of Pancreatic Adenocarcinoma with EUS-TA for Genomic TestingFactorInsufficient for Genomic Testing (N = 34)Sufficient for Genomic Testing (N = 31)p-valueSampling Method0.029*FNA31 (60%)21 (40%)FNB3 (23%)10 (77%)Tumor Location1.00Head/Neck21 (54%)18 (46%)Body8 (50%)8 (50%)Tail5 (50%)5 (50%)Tumor Size (cm), median (IQR)3.3 (2.5, 4)3 (2.3, 3.5)0.17Number of Passes, median (IQR)4 (3, 5)3 (3, 4)0.36Method of Suction (FNA only)0.76Standard23 (62%)14 (38%)Wet8 (53%)7 (47%) Open table in a new tab