Acute neuro-endocrine profile and prediction of outcome after severe brain injury

Abstract

Object: The aim of the study was to evaluate the early changes in pituitary hormone levels after severe traumatic brain injury (sTBI) and compare hormone levels to basic neuro-intensive care data, a systematic scoring of the CT-findings and to evaluate whether hormone changes are related to outcome.

Methods: Prospective study, including consecutive patients, 15-70 years, with sTBI, Glasgow Coma Scale (GCS) score ≤ 8, initial cerebral perfusion pressure > 10 mm Hg, and arrival to our level one trauma university hospital within 24 hours after head trauma (n = 48). Serum samples were collected in the morning (08-10 am) day 1 and day 4 after sTBI for analysis of cortisol, growth hormone (GH), prolactin, insulin-like growth factor 1 (IGF-1), thyroid-stimulating hormone (TSH), free triiodothyronine (fT3), free thyroxine (fT4), follicular stimulating hormone (FSH), luteinizing hormone (LH), testosterone and sex hormone-binding globulin (SHBG) (men). Serum for cortisol and GH was also obtained in the evening (17-19 pm) at day 1 and day 4. The first CT of the brain was classified according to Marshall. Independent staff evaluated outcome at 3 months using GOS-E.

Results: Profound changes were found for most pituitary-dependent hormones in the acute phase after sTBI, i.e. low levels of thyroid hormones, strong suppression of the pituitary-gonadal axis and increased levels of prolactin. The main findings of this study were: 1) A large proportion (54% day 1 and 70% day 4) of the patients showed morning s-cortisol levels below the proposed cut-off levels for critical illness related corticosteroid insufficiency (CIRCI), i.e. <276 nmol/L (=10 ug/dL), 2) Low s-cortisol was not associated with higher mortality or worse outcome at 3 months, 3) There was a significant association between early (day 1) and strong suppression of the pituitary-gonadal axis and improved survival and favorable functional outcome 3 months after sTBI, 4) Significantly lower levels of fT3 and TSH at day 4 in patients with a poor outcome at 3 months. 5) A higher Marshall CT score was associated with higher day 1 LH/FSH- and lower day 4 TSH levels 6) In general no significant correlation between GCS, ICP or CPP and hormone levels were detected. Only ICPmax and LH day 1 in men was significantly correlated.

Conclusion: Profound dynamic changes in hormone levels are found in the acute phase of sTBI. This is consistent with previous findings in different groups of critically ill patients, most of which are likely to be attributed to physiological adaptation to acute illness. Low cortisol levels were a common finding, and not associated with unfavorable outcome. A retained ability to a dynamic hormonal response, i.e. fast and strong suppression of the pituitary-gonadal axis (day 1) and ability to restore activity in the pituitary-thyroid axis (day 4) was associated with less severe injury according to CT-findings and favorable outcome.

Figure 1

Proportions of patients (n = 45) presenting hormone values above or below laboratory reference interval day 1 and day 4 after sTBI. Reference intervals are given in Table 1.

Figure 2

a) Serum fT3 and TSH levels day 1 and day 4 in survivors and non-survivors 3 months after injury. Wilcoxon sign rank test between groups and paired Student´s t-test between day 1 and day 2 results. Values are means ± sem. b) Serum fT3 and TSH levels day 1 and day 4 in patients with favorable (GOS 4-5) and unfavorable (GOS 1-3) outcome 3 months after injury. Wilcoxon sign rank test between groups. Values are means ± sem.

Figure 3

a) Serum LH, FSH, testosterone and fc-testosterone levels day 1 and day 4 in male survivors and non-survivors 3 months after injury. Note that the value for fc-testosterone is divided by 10. Wilcoxon sign rank test between groups. Values are means ± sem. b) Serum LH, FSH, testosterone and f fc-testosterone levels day 1 and day 4 in males with favorable (GOS 4-5) and unfavorable outcome 3 months after injury. Note that the value for fc-testosterone is divided by 10. Wilcoxon sign rank test between groups. Values are means ± sem.

Figure 4

a) Probability of death at 3 months related to serum LH levels day 1 after sTBI. Probability results are from ROC analysis. b) Probability of unfavorable outcome at 3 months related to serum LH levels day 1 after sTBI. Probability results are from ROC analysis.