Stickler syndrome is characterized by ocular, auditory, skeletal, and orofacial abnormalities. We describe a family with autosomal recessive Stickler syndrome. The main clinical findings consisted of high myopia, vitreoretinal degeneration, retinal detachment, hearing loss, and short stature. Affected family members were found to have a homozygous loss-of-function mutation in COL9A2, c.843_c.846 + 4del8. A family with autosomal recessive Stickler syndrome was previously described and found to have a homozygous loss-of-function mutation in COL9A1. COL9A1, COL9A2, and COL9A3 code for collagen IX. All three collagen IX α chains, α1, α2, and α3, are needed for formation of functional collagen IX molecule. In dogs, two causative loci have been identified in autosomal recessive oculoskeletal dysplasia. This dysplasia resembles Stickler syndrome. Recently, homozygous loss-of-function mutations in COL9A2 and COL9A3 were found to co-segregate with the loci. Together the data from the present study and the previous studies suggest that loss-of-function mutations in any of the collagen IX genes can cause autosomal recessive Stickler syndrome.

244 Background: Pancreatic ductal adenocarcinoma (PC) is associated with multiple hereditary cancer syndromes. Genes implicated in hereditary PC include ATM, BRCA1, BRCA2, CDKN2A, EPCAM, MLH1, MSH2, MSH6, PALB2 and PMS2. The advent of multi-gene hereditary cancer panel testing streamlines diagnoses and medical management for clinicians and patients. Our objective was to assess the yield of pathogenic/likely pathogenic variants (PV/LPV) in individuals with PC undergoing panel testing as an initial test at GeneDx. Methods: We retrospectively reviewed panel test results of 605 individuals reporting a personal history of PC. Panel testing evaluated up to 32 genes associated with hereditary cancer. Individuals reporting neuroendocrine pathology or previous BRCA1/BRCA2 testing were excluded. Results: In this cohort, 61 PV/LPV were detected in 57 individuals in the following genes: ATM (17), BRCA2 (14), BRCA1 (5), CDKN2A (5), PALB2 (5), CHEK2 (4), MLH1 (2), MUTYH (2), PMS2 (2), BARD1 (1), FANCC (1), MSH2 (1), RAD51D (1) and TP53 (1), corresponding to a positive yield of 9.4% (57/605). Fifty-one of 61 PV/LPV were detected in genes associated with PC (84%) while 10 PV/LPV (16%) were identified in other genes including BARD1, CHEK2, FANCC, MUTYH, and RAD51D. The diagnostic yield among those reporting a family history of PC (33/294, 11.2%) was not statistically different from those without a reported family history (24/311, 7.7%). However, PV/LPV in ATM were detected more often in individuals reporting a family history of PC compared to those without a family history (4.1% vs. 1.6%, p=0.018). Conclusions: In total, 9.4% of patients with PC tested positive for PV/LPV in 14 different genes by panel testing. Although the majority of PV/LPV were identified in known PC genes, 16% of positive findings occurred in genes not typically associated with PC. ATM was most commonly implicated and more frequently reported in patients reporting family histories of PC. Assessing whether these genes are indeed causally related to PC and/or are possibly associated with other cancer types requires further investigation. Based on our results we conclude multi-gene panel testing may be considered as a first option for patients with PC regardless of their family history.