Diabetes insipidus associated with Langerhans cell histiocytosis: Is it reversible?

Fourteen of 58 (24%) children with Langerhans cell hisiocytosis (LCH) currently attending the Hospital for Sick Children (London) developed thirst and polyuria during the course of their disease. Three had single-system disease confined to bone, and 11 had multisystem disease. The median age at presentation of LCH was 2 years 0 months, and polyuria/polydipsia developed at a median age of 3 years 9 months (range 1 month before diagnosis of LCH to 4 years after diagnosis). Each child had a water deprivation test with measurement of urinary arginine vasopressin (AVP) to document diabetes insipidus. The doses of 1-desamino-8-D arginine vasopressin (DDAVP) required to control symptoms were compared at diagnosis and at a mean follow-up of 7 years 8 months. Local and systemic treatment was recorded. Ten of 14 children were shown to have "complete" diabetes insipidus, whilest the other four had "partial" diabetes insipidus. Seven children were treated with irradiation. with or without systemic chemotherapy, six with systemic chemotherapy only, and one with DDAVP replacement only. No child, including two with partial diabetes insipidus irradiated within 4 weeks of the onset of symptoms, lost symptoms of polyuria/polydypsia and none was able to discontinue DDAVP replacement. One child treated with Etoposide showed a temporary rise in urinary AVP level to within the normal range but still needed DDAVP to control her symptoms. The mean doses of DDAVP at onset of diabetes insipidus and at follow-up were 9.3 micrograms and 18 micrograms daily, respectively. We conclude that the most appropriate treatment for reversing diabetes insipidus complicating Langerhans cell histiocytosis is yet to be determined. Precise documentation of posterior pituitary dysfunction, including measurement of urinary AVP levels, is essential if the effects of new forms of treatment are to be assessed accurately.