Gastrointestinal Bleeding Is Common in Children With Familial Dysautonomia: A Case-Control Study (1980-2017): Presidential Poster Award

Abstract

Introduction: Familial dysautonomia (FD) is a rare hereditary disorder caused by mutations in the IKBKAP gene. The disease is characterized by sensory and afferent autonomic abnormalities. Clinical observations suggest that GI bleeds are not infrequent in patients with FD, but their actual frequency is unknown. We aimed to define the incidence of GI bleeding in patients with FD and its associated risk factors. Methods: Cases and age/sex matched controls were retrospectively identified using the NYU FD registry, which includes over 600 patients with genetically confirmed FD. Cases were defined as patients with FD with at least one episode of GI bleeding between January 1, 1980 and September 25, 2017. For each case we selected one to two age-matched controls with FD and no history of GI bleeding. Data on medication use, gastrostomy (G) tube placement, fundoplication, and ambulatory laboratory values obtained within one year before the GI bleed (or matched time period) were compared using univariate logistic regression models. Retinal fundus tortuosity scores (a noninvasive surrogate marker for the microcirculation) were collected from the most recent clinical visit. Continuous variables were evaluated with t-test and categorical variables by Chi squared test. Multivariate analysis was performed. Results: Thirty-nine FD patients with Gl bleeding and 52 controls were included. 62 GI bleeds occurred in 2,144 observed years (28.0 per 1000 person year) with a median age of bleeding at 11 years old (Figure 1). Fifty-three cases (85.5%) were upper GI bleeds from gastric and duodenal ulcers (Figure 2). Bleeds in patients with FD were more likely to occur in those with a G tube (p<0.01) or with a fundoplication (p<0.01) (Table 1). There were no differences in NSAID or SSRI use, age at fundoplication, or age of G tube placement. For a subset of the FD cohort with available ophthalmic fundus imaging (n=61), cases with GI bleeding had no difference in average arteriolar tortuosity score. Conclusion: The incidence of GI bleeding in FD is considerably higher than in the general pediatric population, which has been reported as 0.2 per 1000 person year. Rather than medications, G-tube and fundoplication appear to be the most important risk factors for gastrointestinal bleeding in patients with FD. Moreover, these procedures did not appear to be direct complications of these procedures. The potential role of other factors leading to mucosal injury remain important targets for study and potential interventions.1054_A Demographics. Bleeds in FD patients were more likely to occur in those with a G tube (p<0.01) and in those with a fundoplication (p<0.01). Statistical tests were not performed to compare medications given that zero cases had history of NSAID or SSRI use.1054_B Distribution of gastrointestinal bleed by age frequency. A total of 62 GI bleeds occurred with a median age of bleeding at 13.09 years old.1054_C Location of gastric ulcers. Fifty-three cases (85.5%) were upper GI bleeds from gastric and duodenal ulcers.