Central Diabetes Insipidus in Children and Adolescents: Twenty-Six Year Experience from a Single Centre

Abstract

Introduction. Paediatric cohorts of central diabetes insipidus (CDI) have shown varying prevalence for different causes of CDI. The objective of this study was to determine the causes of CDI and long-term outcome in children and adolescents from a Tertiary Paediatric Endocrinology unit. Methods. The clinic database was searched to identify patients with CDI managed between 1993 and 2019. Relevant clinical information was collected from patient records. Results. A total of 138 CDI patients, median age 6 years (range <1–18) at presentation, were identified. Principal CDI aetiologies were craniopharyngioma (n = 44), acute central nervous system (CNS) insult (n = 33), germinoma (n = 15), postneurosurgery (indication other than craniopharyngioma and germinoma, n = 20), midline CNS malformation (n = 14), Langerhans cell histiocytosis (n = 5), and familial (n = 2). Idiopathic CDI in this cohort was infrequent (n = 5). Patients with CNS malformations/infections presented with CDI at a younger age compared to patients with CNS tumours $\left(p<0.0001\right)$ . Five patients, initially presenting as idiopathic CDI, were subsequently diagnosed with germinoma after a median interval of 3.3 years. All patients with CDI related to craniopharyngioma and nearly all (87%) patients with CDI related to germinoma had concomitant GH, ACTH, and TSH deficiency. The majority of patients who manifested CDI due to acute CNS insult either deceased (30%) or had transient CDI (33.3%). Conclusion. Surgery for craniopharyngioma was the most common underlying aetiology of CDI with ubiquitous occurrence of panhypopituitarism in these patients. Manifestation of CDI in patients with acute CNS insult carries poor prognosis. We affirm that neuroimaging assessment in idiopathic CDI should be continued beyond 3 years from diagnosis as a significant number of patients exhibited progression of infundibular thickening 3 years post-CDI diagnosis.