Screening for hypopituitarism in 509 patients with traumatic brain injury or subarachnoid hemorrhage

We performed a screening on patients with traumatic brain injury (TBI) or subarachnoid hemorrhage (SAH) to determine the prevalence of post-traumatic hypopituitarism in neurorehabilitation in a cross-sectional, observational single-center study. In addition, the therapeutic consequences of our screening were analyzed retrospectively. From February 2006 to August 2009, patients between 18 and 65 years (n=509) with the diagnosis of TBI (n=340) or SAH (n=169) were screened within two weeks of admittance to neurorehabilitation as clinical routine. Blood was drawn to determine fasting cortisol, free thyroxine (fT4), prolactin, testosterone or estradiol, and insulin-like growth factor I (IGF-I). Patients with abnormalities in the screening or clinical signs of hypopituitarism received further stimulation tests: growth hormone releasing hormone -L-arginine-test and adrenocorticotrophic hormone (ACTH)-test (n=36); ACTH-test alone (n=26); or insulin tolerance test (n=56). In our screening of 509 patients, 28.5% showed lowered values in at least one hormone of the hypothalamus-pituitary axis and 4.5% in two or more axes. The most common disturbance was a decrease of testosterone in 40.7% of all men (in the following 13/131 men were given substitution therapy). Low fT4 was detected in 5.9% (n=3 were given substitution therapy). Low IGF-I was detected in 5.8%, low cortisol in 1.4%, and low prolactin in 0.2%; none were given substitution therapy. Further stimulation tests revealed growth hormone deficiency in 20.7% (n=19/92) and hypocortisolism in 23.7% (n=28/118). Laboratory values possibly indicating hypopituitarism (33%) were common but did not always implicate post-traumatic hypopituitarism. Laboratory values possibly indicating hypopituitarism were common in our screening but most patients were clinically not diagnosed as pituitary insufficient and did not receive hormone replacement therapy. A routine screening of all patients in neurorehabilitation without considering the time since injury, the severity of illness and therapeutic consequences seems not useful.