Objectives: Diazoxide is first line treatment for hypoglycaemia due to hyperinsulinaemic hypoglycaemia (HH). However, the FDA has raised serious concerns regarding diazoxide-induced pulmonary hypertension (PH) in 2015. Although sporadic cases of PH have been reported, no HH cohort has been systematically characterised to understand severity and risk factors for diazoxide-induced PH.
Methods: To investigate the onset, progress and associated factors in PH, patients with HH who developed diazoxide-induced PH in four regional centres were retrospectively reviewed. PH diagnosis was based on clinical and/or echocardiography evidence. Child and treatment-related risk factors were analysed for association. The time intervals from diazoxide initiation to onset and resolution of PH were also recorded.
Results: Twelve cases were identified (5M:7F). HH was diagnosed at median (range) 12 (1,180) days, with diazoxide started 3 (1,76) days from diagnosis, reaching a highest dose of 8.0 (2.5,20) mg/kg per day. Only 3 (25%) patients had mutations in ABCC8/KCNJ11 establishing genetic causation. Total fluid intake was 170 (100,180) ml/kg per day prior to treatment. The majority developed PH within 2 weeks of diazoxide (12 (2,90) days), with 3 patients requiring intensive care ventilation (two requiring high frequency oscillation). Two-thirds of (8/12) patients had baseline echocardiography before initiation of diazoxide. Diazoxide dose reduction did not ameliorate PH but complete diazoxide withdrawal led to PH resolution at a variable time of 32.5 (3,985) days. In three patients, PH has yet to resolve after 6 months. Risk factors for the development of PH included low birthweight (and/or intrauterine growth restriction) in 6 (50%) and fluid intake exceeding 130 ml/kg per day in 10 (83%) patients. Eight (67%) patients also had congenital heart disease (CHD), predominantly secundum atrial septal defects and patent arterial ducts, likely related to the early transitional circulation. The presence or absence of CHD did not influence the time to develop PH (13 (2,90) v 11 (6,13) days, p=0.37) or time for PH resolution (28 (7,90) v 37 (3,985) days, P=0.99) respectively.
Conclusion: PH is a serious complication of diazoxide therapy in HH occurring at an unpredictable time from initiation of treatment. We recommend vigilance for PH in low birthweight infants with fluid intake exceeding 130 ml/kg per day. PH-specific echocardiography should be performed before diazoxide treatment to identify underlying CHD, followed by weekly monitoring for at least the first 2 weeks of treatment. If PH is identified, diazoxide should be discontinued to facilitate PH resolution.
27 - 29 Sep 2018
European Society for Paediatric Endocrinology