Provocative hypothalamopituitary axis tests in severe head injury: correlations with severity and prognosis

Abstract

Objective: To evaluate the effect of severe head injury on both the secretion of basal pituitary hormones and the response to exogenous synthetic hypothalamic releasing factors administration.

Design: Prospective, clinical study.

Setting: General intensive care unit in a university teaching hospital, Italy.

Patients: Comatose, head-injured patients (n = 22), all intubated and mechanically ventilated, invasively monitored without previous endocrinologic problems and substitutive therapies.

Interventions: Routine neuroemergency procedures; administration of exogenous, synthetic hypothalamic releasing hormones.

Measurements and main results: Determinations of basal concentrations of growth hormone (GH), prolactin (PRL), thyroid-stimulating hormone (TSH), triiodothyronine, and thyroxine were performed daily in the first week and on days 15 and 16 after the trauma. Plasma insulin-like growth factor-I and cortisol were also determined on days 2, 7, and 15. We carried out a thyrotropin-releasing hormone (TRH) test for the evaluation of the PRL, TSH, and GH responses on days 1 and 16 after the trauma and a growth hormone-releasing hormone (GHRH) test for the evaluation of GH and PRL responses on days 2, 7, and 15 after the trauma. Outcome was evaluated at 6 mos with the GOS. Triiodothyronine showed low values, even if in the normal range; thyroxine remained in the normal range. Significant increases in insulin-like growth factor-I concentrations were observed on both days 7 and 15 compared with day 2 (p = .024 and p = .034, respectively). The GH response to GHRH was significantly greater on days 7 and 15 than in the very acute phase (p< .01 comparing days 7 and 15 vs. day 2). We found a higher GH response to GHRH on day 7 in group 1 vs. group 2 (as both peak and area under the curve, p = .018 and p = .015, respectively). No difference in GH response was detected on days 2 and 15. A "paradoxical" response of GH to TRH was observed on the day after the head trauma (basal vs. peak, p = .002) but not on day 16. The GH peak response to TRH was greater on day 1 in those patients with an unfavorable course (group 1 vs. group 2, p < .05). The TSH response to TRH was not significantly correlated to the severity of trauma, but it was significantly (p < .04) higher in group 1 than in group 2. Finally, a "paradoxical" PRL response to GHRH administration was present on day 2 (basal vs. peak, p=.0003), day 7 (basal vs. peak, p = .01), and on day 15 after the trauma (basal vs. peak, p = .04).

Conclusions: Some of the responses to provocative tests have been identified as "paradoxical" and seem to have a great importance in the definition of prognosis in severe head-injured patients, specifically the GH response to TRH and the PRL response to GHRH that are significantly correlated with outcome.