ESPE Abstracts

Introduction: Congenital hyperinsulinism (CHI) is the most common cause of hypoglycaemia in infancy caused by dysregulated insulin secretion. The management of severe hypoglycaemia often requires the administration of high dextrose-containing fluids through a central venous catheter (CVC). However, CVCs carry the risk of complications including thrombosis. We sought to determine the incidence of CVC-associated thrombosis in patients with CHI and examine associated risk factors.

Methods: Patients with CHI admitted at a specialist treatment centre requiring CVC placement from 2014 to 2017 (n=26) were retrospectively reviewed for the incidence of thrombosis and potential risk factors. Non-parametric tests were used for differences between groups with and without thrombosis.

Results: 5 of 26 (19%) patients with CHI requiring CVC placement for the management of hypoglycaemia developed thrombosis in the CVC, confirmed by ultrasound scanning, giving an incidence of 4.9 thromboses per 1000 CVC days. Thrombosis was identified at a median (range) time of 12 (2–17) days after CVC insertion at age 13 (6–139) days. Mutations in ABCC8 were detected in two patients who were treated surgically; one was paternally-inherited (focal CHI), and the other was homozygous (diffuse CHI). The remaining three patients with no known mutations were managed with diazoxide. Males were more frequent in those with thrombosis (100% vs 71%, P=0.29), but there was no difference in the frequency of mutations (40 vs 38%, P=0.94), and the incidence of focal CHI (20 vs 24%, P=0.85) compared to those without thrombosis. Similarly, there was no difference in the maximum concentration (%) of dextrose [20 (12.5–50) vs 20 (10–50), P=0.70] or glucagon infusion (mcg/kg per hour) [15 (3–20) vs 10 (4–25), P=0.54] between the groups with and without thrombosis. The duration (days per patient) of high concentration dextrose (>15%) through CVCs was also similar [12.5 (7–27) vs 15 (2–107), P=0.73], while the duration of CVC placement (days per patient) was marginally less in the group with thrombosis [13 (2–41) v 22 (2–213), P=0.17], indicating earlier CVC withdrawal following thrombosis.

Conclusion: Thrombosis was identified in a clinically relevant proportion of patients with CHI and CVC (19%). No definite association was identified with putative extrinsic risk factors. Intrinsic factors specific to CHI, such as hyperinsulinism, may increase the risk of thombosis development in CHI patients. Our data supports the use of prophylactic anticoagulant therapy in severe CHI.