No abstract

e15127 Background: In CRC, utility of immunotherapy (IT) remains limited to persons with microsatellite instability–high (MSI-H) status. Immunotyping of CRC patients is critical towards further establishing IT's potential. Both incidence and survival rates for CRC in the United States vary between races due to multiple factors including genetic heterogeneity. In this study, we have profiled tumor samples in a racially diverse population for the expression of four different members of the B7 family of immune check point regulators. Methods: Tissue microarray (TMA) was generated from 208 CRC patients. Formalin fixed paraffin embedded (FFPE) tissues were utilized for immunohistochemistry (IHC) post- antigen retrieval. Antibodies specific for IHC staining were used for B7-H3 (D9M2L), PD-L1(E1L3N), B7-H4/B7x(D1M81) and HHLA2(566.1). Each specimen was scored for percent staining of tumor tissue (4 quartile groups) and for intensity (1-4x), and then an overall score (1-16) was calculated. The clinical outcome of interest was overall survival (OS), measured as time from diagnosis of metastatic cancer to death. Analysis for differences in OS was by log rank test, while differences between mean staining was by t test. Race was designated as non-Hispanic white (NHW, n = 41), non-Hispanic black (NHB, n = 84), Hispanic (n = 75), and others (8). Results: B7-H3 protein expression showed strong cytoplasmic distribution. NHB patients had a mean lower expression than NHW patients (0.19 vs 0.41, p = 0.02). Correspondingly, NHB patients had a worse OS than NHW patients (606 vs 759 days). No discernible differences were found in Hispanic patients. PD-L1 showed membranous distribution with 17% expression without significant difference among patients of different racial origin. HHLA2 was more widely expressed with about 35% staining but without statistical significance. B7-H4/B7x failed to show any expression. Conclusions: Expression of B7H3 varies among patients with differential racial backgrounds. It has the potential to be a prognostic biomarker and may be a reason for worse outcome among NHB patients in our population. Other B7 immune checkpoint markers failed to show clinical relevance.

e16798 Background: There are no accepted guidelines for testing individuals at elevated risk for developing pancreatic duct adenocarcinoma (PC). We initiated a prospective screening and surveillance program for individuals at elevated risk for PC. Methods: Eligibility for the Pancreatic Cancer Early Detection Protocol (PCEDP) was based on germ line status and/or family history of PC, provided that the imparted risk was either five times that of the general population or 7.5% lifetime risk for developing PC. Testing was continued alternating between Endoscopic Ultrasound (EUS) and Magnetic Resonance Imaging (MRI) of the abdomen. Objectives were was to analyze the number, type, and location of pancreatic conditions found and their associations with genetic or family history; and to evaluate the outcomes and/or complications that may have resulted from our testing. Results: From April 2014 through October 2019 we received 238 queries, out of which 75 individuals (31%) enrolled in the PCEDP. Eligibility was based upon individual’s germ line only (45%), family history only (32%), and both (23%). Germ line mutations were observed in 34 (BRCA2), 9 (BRCA1), 4 (ATM), 3 (PALB2), and 3 (CDKN2A) individuals. Median age at consent was 57, 60% were female,and88%, 4%, 3%, and 1% self-identified as Caucasian, African American, Hispanic, and Asian, respectively. 133 EUS procedures and 83 MRIs have been performed. No serious adverse events occurred. Standard Insurance approved and paid for the vast majority of tests. Four individuals withdrew (5%) and three (4%) were lost to follow up. Ten individuals (13%) were found to have abnormal findings in the pancreas and therefore met an endpoint of the study, including seven anechoic cysts and three suspected intraductal papillary mucinous neoplasms (IPMN). All individuals with endpoints were recommended to continue surveillance with EUS. Eight of the ten endpoints were found on baseline EUS, one one from baseline MRI, and one was found on the 3rd EUS. One of the individuals with a 2.5cm IPMN seen on baseline EUS underwent a subsequent distal pancreatectomy, with pathology revealing high grade dysplasia. Conclusions: Screening and surveillance for PC using EUS alternating with MRI was feasible and well tolerated in our population of individuals with an elevated risk. Baseline EUS was successful in detecting 10/75 = 13% of enrollees with some abnormal pancreatic finding, including one requiring intervention with a high grade pre-malignant IPMN.

e15089 Background: Early-onset colorectal cancer (EO-CRC) incidence is increasing disproportionately among minorities compared to Non-Hispanic Whites (NHW). EO-CRC have aggressive features such as higher grade and advanced stages. The appropriate age to start screening colonoscopy (SC) in NHW and minorities remains controversial; varying between 45 and 50 years old. We aim to compare EO-CRC clinico-pathological characteristics and survival rates by race groups. Methods: Patients with colorectal adenocarcinoma (CRC) with available race and stage as per AJCC 6th edition were identified using the SEER registry (1973-2010). EO-CRC was defined as CRC before age 50 years. Clinico-pathological features, overall survival (OS) by Kaplan Meier curves and mortality predictors by multivariate analysis were evaluated by race groups. Results: 180 605 patients with CRC were identified; 10.2% had EO-CRC. Mean age of diagnosis was 42.7 years and EO-CRC frequency was higher in minorities (Hispanics (H):16.7%, Non-Hispanic Black (NHB):12.7% and Asian (A): 12.8%) compared to NHW (8.7%). EO-CRC in NHB was predominantly seen in females. The rectum was the most common location for all races. Two-thirds of tumors were located between the sigmoid and anal regions in all races except NHB that had higher frequencies of right-sided tumors. Compared to other races, NHB had worse OS at all stages and tumor locations. NHB was associated with 72% increased risk of death by multivariate analysis. Conclusions: Our data suggest that EO-CRC frequency, pathological features and OS differ by race group; hence SC guidelines should be tailored accordingly. SC would be considered early; especially in minorities. Complete colonoscopy should be considered for NHB given higher rates of right-sided tumors and worse OS; while sigmoidoscopy may be adequate for others up to age 50, given higher rates of tumors located in the sigmoid to anal region. [Table: see text]