Hypopituitarism after subarachnoid haemorrhage, do we know enough?

We searched Pubmed and Embase for articles published from 1995 to 2014 and used a combination of the term "subarachnoid haemorrhage" with "pituitary function", "hypopituitarism", "growth hormone", "thyroid function", "growth hormone", "cortisol", "gonadotropin", "testosterone function" and "diabetes insipidus". We also searched the reference lists of the articles identified by our search strategy. Two authors (LK, FK) screened titles and abstracts of all references listed in the search results independently. Of the remaining titles, full-text articles were retrieved and again screened for eligibility by both authors independently. In case of disagreement, consensus was sought through discussion. A third author (DD) was available if consensus could not be reached.

A full-text article was included in this review if it met all of the following criteria: 1) the study population consisted of patients with SAH caused by a ruptured aneurysm or, of a subgroup of patients with aneurysmal SAH; 2) the primary aim of the study was to investigate the incidence of endocrine dysfunction after SAH; 3) outcome was described in terms of levels of one or a combination of the following: ACTH, GH, TSH, cortisol, FSH, LH or testosterone, with a clear description of the assays; 4) time to laboratory investigation was at least 3 months after SAH; and 5) the study concerned a series of patients.

The two reviewers independently judged all studies by inception cohort, description of source population, description of inclusion criteria, follow-up more than 3 months, description loss to follow-up, standardized or valid measurements and data presentation of most important outcome measures. Items were scored as positive, negative or inconclusive (Table 1). A completed PRISMA checklist for quality assessment has been added as an Additional file 1. Data presented in de studies were then collected. Information on patient and study characteristics, inclusion and exclusion criteria, laboratory and outcome measurements were gathered from the selected articles.

Initially, 194 citations (abstracts) were found. Of these citations, 62 articles did not report relevant endocrine outcome. Eighty-eight studies concerned other diseases than aneurysmal SAH. Eighteen case reports were also excluded. Twenty-six full-text articles were collected of which 5 articles reported only early phase endocrine dysfunction [22]-[26], 3 were review articles [27]-[29], 1 article reported combined data of SAH with traumatic brain injury [30] and 1 large study was excluded because it concerned an internet based data collection study [31]. Finally sixteen studies fulfilled the inclusion criteria and were eligible for the current review (Figure 1).

Study population size ranged from 10-93 patients. Six studies were cross-sectional or retrospective cohort studies [17],[32]-[37]. Ten studies were conducted prospectively [37]-[45]. Interval between SAH and neuroendocrine assessment in studies reporting neuroendocrine outcome ranged from 3 months to 10 years.

Dimopoulou et al. retrospectively analysed 30 patients between one and two years after SAH but did not use a stimulation test for the evaluation of growth hormone function [17]. Aimaretti et al. conducted a prospective follow-up study of 40 patients after SAH derived from multiple Italian centres. The patients were all conscious and measured 3 months after discharge from the ICU. GHRH + arginine test was used to measure growth hormone function [37]. Aimaretti et al. prospectively studied 32 patients in Italian hospitals, and performed basal hormonal tests and GHRH + arginine test as a dynamic test to establish GHD between 3 and 12 months after SAH [32]. Brandt et al. selected 10 patients with fatigue after SAH and measured corticotrophin, growth hormone and thyrotrophic function using insulin tolerance test (ITT) and TSH-Thyroid releasing hormone stimulation tests12 month after SAH. In 30% of the patients ITT was not performed [33]. Kreitschman-Andermahr et al. retrospectively studied 40 SAH patients from a cohort of 274 patients after excluding patients with liver disease, coronary heart disease, convulsions, DM, depression, severe confusional state or vegetative state after discharge. ITT and THRH-LHRH were used as dynamic tests for assessment of ACTH, TSH and GH function 12 to 72 months after SAH [35]. Kreitschman-Andermahr et al. retrospectively measured basal hormones in 45 patients 3 to 24 moths after SAH. Only 14 patients had dynamic tests [36]. Jovanovic et al. retrospectively evaluated endocrine function in 93 patients, between one and ten years after SAH, however stimulation tests were not used [34]. Tanriverdi et al. prospectively analysed 22 patients one year after SAH using basal and dynamic tests for ACTH and GHD [39]. Karaca et al. did a follow-up study, three years after SAH of 20 patients investigated by Tanriverdi et al. in the abovementioned study using basal hormonal tests and glucagon stimulation test [45]. They found 4 cases of GHD three years after SAH of whom three did not have GHD one year after SAH. Dutta et al. evaluated endocrine function in 60 SAH patients with anterior communicating artery (A-com) and middle cerebral artery (MCA) aneurysms using only basal hormonal tests. Part of the study was retrospective, analysing patients one year after SAH and partly prospectively analysing patients 6 months after SAH [46]. Kronvall et al. prospectively analysed 45 patients in the acute phase and 3 to 6 months after SAH, using basal hormonal test and GHRH-arg test for GHD. They did not use a dynamic test to establish ACTH deficiency [44], Khursheed et al. prospectively analyzed 73 patients nine months after SAH for TSH and gonadotropin deficiency and not the other anterior pituitary hormones [43]. Blijdorp et al. prospectively analysed 84 patients and reported preliminary data of 43 patients using basal hormonal tests, synacten test when ACTH deficiency was suspected and a ghrelin test in the early phase after SAH and confirmatory GHRH-arg test after six months [42]. The most prominent methodological shortcomings were the incomplete reports of patient selection [17],[32],[33],[35],[37],[46], selection bias [17],[33]-[35],[42],[46] and inadequate laboratory testing [17],[33],[35],[36],[38],[43],[44]. Some studies did not use dynamic tests to determine growth hormone or corticotrophin deficiency [17],[34],[43],[44],[46]. In some reports, description of the statistical analysis and results were incomplete [36] or even absent [33].

In total, 671 patients were examined for hypopituitarism after SAH (Table 2). The proportion of patients with endocrine dysfunction varied from 0 - 55%. Growth hormone deficiency occurred in 0 to 29%, adrenocorticotropic deficiency in 0 to 40%, gonadotropin (luteinizing hormone, follicle stimulating hormone and testosterone) deficiency in 0 to 40% and thyroid stimulating hormone deficiency in 0 to 20% of patients in different studies. The largest prospective study by Klose et al. evaluated 62 patients for an average of 14 months (range 11-26 months) after SAH. Although they found some evidence of hypopituitarism after initial testing, they were not able to confirm hypopituitarism by confirmatory laboratory tests. They found no evidence of hypopituitarism in the long term after SAH [40]. In another well designed study, two different confirmatory tests were used to establish GHD adjusting the outcome for body-mass index [41]. After 12 months, 12% of the patients had pituitary dysfunction (PD).

Inconsistent results for predicting hypopituitarism after SAH were reported. In a series of 93 patients with SAH, cerebral vasospasm and hydrocephalus were identified as risk factors for pituitary dysfunction [34]. Kreitschmann-Andermahr et al. found a significant association of female sex and the presence of corticotrophin deficiency [35]. Kronvall et al. reported that younger age was significantly associated with pituitary dysfunction at follow-up [44] but Tanriverdi et al. found higher age to be associated with growth hormone deficiency in the acute phase after SAH [39]. These findings were not confirmed by other studies [17],[32],[40],[47].

In a series of 26 patients, 14 (54%) had neuropsychological deficits, but only 1 patient suffered neuroendocrine dysfunction at six months after SAH [38]. A study of 40 patients evaluated the effect of neuroendocrine dysfunction on quality of life and psychiatric symptoms. Low basal cortisol level was associated with low quality of life scores and high depression scores [48]. Severe GH deficiency was associated with low scores on the energy subscale of Nottingham Health Profile (NHP) questionnaire. Gardner et al. did not find an association between PD and quality of life after SAH, measured using the quality of life in adult GHD assessment (QOL-AGHDA) [41].