Thyrotrophin and prolactin release in prolonged critical illness: dynamics of spontaneous secretion and effects of growth hormone-secretagogues.

Abstract

Infusion of GH secretagogues appears to be a novel endocrine approach to reverse the catabolic state of critical illness, through amplification of the endogenously blunted GH secretion associated with a substantial IGF-I rise. Here we report the dynamic characteristics of spontaneous nightly TSH and PRL secretion during prolonged critical illness, together with the concomitant effects exerted by the administration of GH-secretagogues, GH-releasing hormone (GHRH) and GH-releasing peptide-2 (GHRP-2) in particular, on night-time TSH and PRL secretion. Twenty-six critically ill adults (mean +/- SEM age: 63 +/- 2 years) were studied during two consecutive nights (2100-0600 h). According to a weighed randomization, they received 1 of 4 combinations of infusions, within a randomized, cross-over design for each combination: placebo (one night) and GHRH (the next night) (n = 4); placebo and GHRP-2 (n = 10); GHRH and GHRP-2 (n = 6); GHRP-2 and GHRH + GHRP-2 (n = 6). Peptide infusions (duration 21 hours) were started after a bolus of 1 microgram/kg at 0900 h and infused (1 microgram/kg/h) until 0600 h. Serum concentrations of TSH and PRL were determined by IRMA every 20 minutes and T4, T3 and rT3 by RIA at 2100 h and 0600 h in each study night. Hormone secretion was quantified using deconvolution analysis. During prolonged critical illness, mean night-time serum concentrations of TSH (1.25 +/- 0.42 mlU/l) and PRL (9.4 +/- 0.9 micrograms/l) were low-normal. However, the proportion of TSH and PRL that was released in a pulsatile fashion was low (32 +/- 6% and 16 +/- 2.6%) and no nocturnal TSH or PRL surges were observed. The serum levels of T3 (0.64 +/- 0.06 nmol/l) were low and were positively related to the number of TSH bursts (R2 = 0.32; P = 0.03) and to the log of pulsatile TSH production (R2 = 0.34; P = 0.03). GHRP-2 infusion further reduced the proportion of TSH released in a pulsatile fashion to half that during placebo infusion (P = 0.02), without altering mean TSH levels. GHRH infusion increased mean TSH levels and pulsatile TSH production, 2-fold compared to placebo (P = 0.03) and 3-fold compared to GHRP-2 (P = 0.008). The addition of GHRP-2 to GHRH infusion abolished the stimulatory effect of GHRH on pulsatile TSH secretion. GHRP-2 infusion induced a small increase in mean PRL levels (21%; P = 0.02) and basal PRL secretion rate (49%; P = 0.02) compared to placebo, as did GHRH and GHRH + GHRP-2. The characterization of the specific pattern of anterior pituitary function during prolonged critical illness is herewith extended to the dynamics of TSH and PRL secretion: mean serum levels are low-normal, no noctumal surge is observed and the pulsatile fractions of TSH and PRL release are reduced, as was shown previously for GH. Low circulating thyroid hormone levels appear positively correlated with the reduced pulsatile TSH secretion, suggesting that they have, at least in part, a neuroendocrine origin. Finally, the opposite effects of different GH-secretagogues on TSH secretion further delineate particular linkages between the somatotrophic and thyrotrophic axes during critical illness.