Audit of Pituitary Dysfunction after Traumatic Brain Injury: Caution in Interpretation of Glucagon Stimulation Test in Diagnosis of GH and ACTH Deficiency.
Tenorio Jimenez C 1,2, Niemi M 1, Malik A 1, Papadopoulou D 1, Ham T 2, Baxter D 2, Sharp DJ 2 & Goldstone AP 1,3
1 Imperial Centre for Endocrinology, Imperial College Healthcare NHS Trust; 2 Computational, Cognitive and Clinical Neuroimaging Laboratory; 3 MRC Clinical Sciences Centre, Imperial College London, Hammersmith Hospital, London, UK.
Introduction. Traumatic brain injury (TBI) is the leading cause of death and disability in young adults. Pituitary dysfunction after TBI has received increasing recognition in recent years because of its potential contribution to morbidity. Differences in patient selection, severity of TBI, time since injury, diagnostic endocrine tests and diagnostic criteria may explain the variation in the prevalence of pituitary dysfunction between studies.
Objectives. Audit of (i) the prevalence of pituitary dysfunction, and (ii) the validity of the glucagon stimulation test (GST) as a diagnostic tool for GH and ACTH deficiency after TBI.
Patients and Methods. 179 patients (137 males), mean ± SD age 41.0 ± 15.7yr (range 17.6-88.1) attended initial assessment in the TBI outpatient clinic at Charing Cross Hospital between July 2009 and Aug 2011. Median time since TBI was 0.27yr (range 0.05-46.5, 26% >1yr). 34% had an intracranial bleed and 9% needed craniotomy. 20% had an absolute contraindication (age, epilepsy, cardiovascular disease) for insulin tolerance testing (ITT) and 42% also had a relative contraindication (craniotomy, bleed). Patients had basal pituitary function tests and 135 (75%) had a GST for diagnosis of GH and ACTH deficiency. Previous audit has suggested maximum specificity (70%) and sensitivity (64%) for ACTH deficiency to be a GST peak cortisol <350 nmol/L. Where clinically appropriate, patients who failed the GST had a GHRH-Arginine test for confirmation of GH deficiency, a Metyrapone suppression test (MST) for confirmation of ACTH deficiency, and a water deprivation test (WDT) for confirmation of diabetes insipidus.
Results. There were no patients with TSH deficiency (out of 174) or diabetes insipidus (11/174 had normal WDT), but one patient had syndrome of inappropriate ADH. 2/169 had asymptomatic hyperprolactinaemia not due to other causes and 3/170 had gonadotrophin deficiency (1 on opiates). 30/135 had a GST peak cortisol <350 nmol/L (22.2%), but only 1/19 (5.2%) had confirmation of ACTH deficiency by failing a MST. None of the 14 patients with GST peak cortisol >350 nmol/L failed a MST. 35/135 had a GST peak GH <3 mcg/L (25.9%) and 55/135 peak GH <5 mcg/L (40.7%), but only 3/12 (25%) had GH deficiency confirmed by failing a GHRH-Arginine test. Patients with confirmed GH deficiency had worse ratings of Energy/Fatigue, Social Well Being, Social Functioning, General Health and Health Change in the SF36 quality of life (QoL) questionnaire, and worse ratings of Emotional Reaction and Social Isolation in the Nottingham Health Profile questionnaire. GH deficient patients have been commenced on GH replacement under NICE guidelines with symptomatic improvement.
Conclusions. The GST had a high false positive rate in diagnosing both GH and ACTH deficiency in patients after TBI. This is likely related to the low prevalence of true pituitary dysfunction in this group of patients. In our cohort, clinically significant pituitary problems after TBI was only confirmed in 10/174 patients (5.7%). Including the predicted results of those awaiting confirmatory diagnostic tests, we would estimate the maximum prevalence of pituitary dysfunction to be ~10% (18/174). However detection of pituitary dysfunction after TBI remains clinically important, while GH replacement is therapeutically beneficial given its significant neurocognitive and psychological consequences. Despite the high prevalence of contraindications to an ITT and associated resource implications, caution must be made not to over-diagnose GH and particularly ACTH deficiency after TBI by relying on a GST alone. Second line confirmatory tests must be performed for both axes since it is not yet possible to predict those patients with GH and ACTH deficiency through other means.