Dynamic endocrine testing and magnetic resonance imaging in the long term follow-up of childhood Langerhans cell histiocytosis

Children treated for Langerhans cell histiocytosis (LCH) are at risk for short and long term endocrine sequelae, but biological predictors of specific deficits are not well defined. We evaluated the frequency and progression of LCH-related endocrine deficits during long term follow-up and assessed the ability of dynamic endocrine testing to identify patients at risk for late anterior or posterior pituitary hormone dysfunction. The 17 patients (5 males and 12 females) were followed a median of 10 yr after diagnosis of single system (n = 6) or multisystem (n = 11) disease. Study evaluations, performed a median of 4.1 yr after the diagnosis, comprised pituitary hormone responses to the appropriate challenge, 7-h water deprivation test, 3% hypertonic saline infusion, and magnetic resonance imaging (MRI). The six patients with GH deficiency at the time of evaluation had a significantly lower GH response to GHRH than the other patients [median peak, 7.3 us. 21.5 mu g/L (P = 0.03); median area under the curve, 4.7 vs. 13.5 mu g/L (P = 0.03)]; levels in the latter group did not differ significantly from those in 20 age- and sex-matched controls with constitutional or familial short stature. Two patients who had GH responses to GHRH of 20.6 and 23 ng/mL at 2.8 and 9.5 yr of age developed GH deficiency at 6.5 and 11.2 yr of age, respectively. The TSH response to TRH was less than 10 mU/L in three patients, two of whom later developed central hypothyroidism. ACTH and cortisol responses to CRF, and PRL responses to TRH were normal in all cases, and LH and FSH responses to GnRH were compatible with pubertal stage; Abnormalities in arginine vasopressin responses to water deprivation or hypertonic saline infusion were seen only in four patients who had preexisting diabetes insipidus (DI); one patient who later developed DI had normal findings. On standard MRI, posterior pituitary hyperintensity was absent only in the patients with DI. Pituitary stalk thickening was seen in seven patients, including three who did not have DI and had normal arginine vasopressin responses. Delayed posterior and anterior enhancement on dynamic MRI was present in two patients, both of whom later developed central hypothyroidism. Patients with single system disease had a lower 5-yr probability of LCH reactivation (41% vs. 83% for those with multisystem disease; P = 0.21) and a significantly lower risk of endocrine dysfunction (P = 0.007). In this series, dynamic evaluation of pituitary function was not a useful predictor of late endocrine sequelae, with the possible exception of the progressively decreasing TSH response to TRH. Similarly, a standard MRI was not predictive, although dynamic imaging may be informative regarding evolving pituitary hormone deficiency.