Copyright (c) Society of Critical Care Medicine and Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited

Critical Care Medicine www.ccmjournal.org 1

Objectives: To assess the clinical outcomes, predictors, and prevalence of anterior pituitary disorders following traumatic brain injury.Data Sources: We searched Medline, Embase, Cochrane Regis-try, BIOSIS, and Trip Database up to February 2012 and con-sulted bibliographies of narrative reviews and selected articles.Study Selection: We included cohort, case-control, cross-sectional studies and randomized trials enrolling at least five adults with blunt traumatic brain injury in whom at least one anterior pituitary axis was assessed. We excluded case series and studies in which other neu-rological conditions were indistinguishable from traumatic brain injury.Data Extraction: Two independent reviewers selected citations, extracted data, and assessed the risk of bias using a standard-ized form. Data Synthesis: We performed meta-analyses using random effect models and assessed heterogeneity using the I2 index.Results: We included 66 studies (5,386 patients) evaluating prevalence, 14 evaluating clinical outcomes, and 27 evaluating predictors. Thirty studies were at low risk of bias. Anterior pitu-itary disorders were associated with a trend toward increased ICU mortality (risk ratio, 1.79; 95% CI, 0.99–3.21; four studies) and no difference in Glasgow Outcome Scale score (mean differ-ence, –0.45; 95% CI, –1.10 to 0.20; three studies). Age (mean difference, 3.19; 95% CI, 0.31–6.08; 19 studies), traumatic brain injury severity (risk ratio, 2.15; 95% CI, 1.20–3.86 for patients with severe vs nonsevere traumatic brain injury; seven studies), and skull fractures (risk ratio, 1.73; 95% CI, 1.03–2.91; six stud-ies) predicted anterior pituitary disorders. Over the long term, 31.6% (95% CI, 23.6–40.1%; 27 studies) of patients had at least one anterior pituitary disorder. We observed significant heteroge-neity that was not solely explained by the risk of bias or traumatic brain injury severity.Conclusions: Approximately one third of traumatic brain injury patients have persistent anterior pituitary disorder. Older age, traumatic brain injury severity, and skull fractures predict anterior pituitary disorders, which in turn may be associated with higher ICU mortality. Further high-quality studies are warranted to better

1Centre de Recherche du CHU de Québec, Santé des Populations et Pratiques Optimales en Santé, Université Laval, Québec, QC, Canada.

2Division of Critical Care, Department of Anesthesiology, Université Laval, Québec, QC, Canada.

3Department of Medicine, Université Laval, Québec, QC, Canada.4Department of Family Medicine and Emergency Medicine, Université Laval, Québec, QC, Canada.

5Centre de recherche du CSSS Alphonse-Desjardins (CHAU de Lévis), Lévis, QC, Canada.

6Centre de Recherche Clinique Étienne-Le Bel, Université de Sherbrooke, Sherbrooke, QC, Canada.

7Department of Social and Preventive Medicine, Université Laval, Québec, QC, Canada.

8Hôpital du Sacré-Coeur de Montréal Research Center, Department of Medicine, Université de Montréal, Montréal, QC, Canada.

9Hôpital du Sacré-Coeur de Montréal Research Center, Department of Critical Care Medicine, Université de Montréal, Montréal, QC, Canada.

10 Centre de Recherche du CHU de Québec, Endocrinology and Nephrol-ogy, Université Laval, Québec, QC, Canada.

11Department of Medicine, McMaster University, Hamilton, ON, Canada.12 Department of Clinical Epidemiology and Biostatistics, McMaster Uni-

versity, Hamilton, ON, Canada.Supplemental digital content is available for this article. Direct URL cita-tions appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s website (http://journals.lww.com/ccmjournal).This study was supported by the Consortium pour le développement de la recherche en traumatologie and the Fonds de la recherche du Québec-Santé. Drs. Lauzier, Turgeon, Archambault, and Lamontagne are recipients of a Research Career Award from the Fonds de la recherche du Québec-Santé. Ms. Boutin holds a Doctoral Research Award. Dr. Archambault served as a board member for the Association des médecins d’urgence du Québec, is employed by the Universite ́ Laval, received grant support from the Fonds de recherche en Sante ́ Que ́bec (career award), and lec-tured for Queen’s University and Centre de sante ́ et services sociaux de Saguenay. Dr. Moore holds a New Investigator Award from the Canadian Institutes of Health Research. Dr. Cook is a Research Chair of the Cana-dian Institutes of Health Research. The remaining authors have disclosed that they do not have any potential conflicts of interest.For information regarding this article, E-mail: francois.lauzier@med.ulaval.ca

Crit Care Med

Copyright © 2013 by the Society of Critical Care Medicine and Lippincott Williams & WilkinsDOI: 10.1097CCM.0000000000000046

Clinical Outcomes, Predictors, and Prevalence of Anterior Pituitary Disorders Following Traumatic Brain Injury: A Systematic Review

François Lauzier, MD, MSc1,2,3; Alexis F. Turgeon, MD, MSc1,2; Amélie Boutin, MSc1; Michèle Shemilt, BSc1; Isabelle Côté, MD1,3; Olivier Lachance1; Patrick M. Archambault, MD, MSc1,2,4,5; François Lamontagne, MD, MSc6; Lynne Moore, PhD1,7; Francis Bernard, MD8,9; Claudia Gagnon, MD3,10; Deborah Cook, MD, MSc11,12

Copyright (c) Society of Critical Care Medicine and Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited

Lauzier et al

2 www.ccmjournal.org

define the burden of anterior pituitary disorders and to identify high-risk patients. (Crit Care Med 2013; XX:00–00)Key Words: clinical outcomes; pituitary disorders; predictors; systematic review; traumatic brain injury

Pituitary disorders, defined as an isolated or combined hormone deficiency of the neurohypophysis and adeno-hypophysis, are a frequently overlooked complication of

traumatic brain injury. A prevalence of up to 30% has been reported (1, 2). Many symptoms of traumatic brain injury survivors such as fatigue, concentration difficulties, depres-sive symptoms, and low exercise capacity are nonspecific (3, 4) and overlap with symptoms of patients with pituitary disorders from other causes (5, 6). Those symptoms have a negative impact on quality of life, function, and work reinser-tion of traumatic brain injury survivors and could be attrib-utable to hormonal deficits. However, the association between pituitary disorders and disability remains uncertain. Several experts recently outlined the importance of defining the bur-den of pituitary disorders in traumatic brain injury patients and identifying high-risk patients for which a systematic screening strategy would be cost-effective (7–13). We there-fore performed a systematic review to comprehensively assess the clinical outcomes and the predictors of anterior pituitary disorders and to evaluate the prevalence of anterior pituitary deficits at different times following injury.

METHODS

Eligibility CriteriaWe performed a systematic review of prospective and retro-spective cohort studies, case-control studies, cross-sectional studies, and randomized clinical trials enrolling at least five adults (≥ 18 years old) with blunt traumatic brain injury in whom at least one anterior pituitary axis was assessed during ICU stay or subsequently. We excluded case studies and case series, studies of mixed population with no possible distinc-tion between patients with other acute neurological conditions like aneurysmal subarachnoid hemorrhage, studies related to previously published data, and studies in which individual hormonal deficits could not be determined (e.g., no normal values, reporting of mean or median of hormonal levels).

Systematic Review ObjectivesThe primary objective was to determine the clinical outcomes related to anterior pituitary disorders. We sought to determine their association with ICU, hospital and long-term mortality, ICU and hospital length of stay, duration of mechanical venti-lation, need for vasopressors, neurological prognosis based on the Glasgow Outcome Scale (14) or extended Glasgow Out-come Scale (15) (on which a lower score represents a poorer neurological outcome), and quality of life and functional sta-tus as measured by validated questionnaires such as the SF-36 (16) and the Functional Independence Measure (17). Our first

secondary objective was to evaluate potential predictors of anterior pituitary disorders: age, sex, severity of brain injury according to admission Glasgow Coma Scale (GCS) (18), brain damage severity according to CT imaging at admission, skull fractures, diffuse axonal injury, and secondary brain injury (hypoxemia, hypotension, and intracranial hypertension). Our other secondary objective was to evaluate the prevalence of anterior pituitary deficits at different times following injury: less than 3 months, 3–12 months, and greater than 12 months after injury.

Search Strategy and Study SelectionWe systematically searched Medline, Embase, Cochrane Cen-tral Register of Controlled Trials, BIOSIS, and Trip Database from their inception up to February 13, 2012. No restric-tion was applied regarding language and type of publication. We generated keywords and index terms related to pituitary disorders and merged it to a traumatic brain injury–specific strategy (19). To increase search sensitivity, we exploded each term. All terms were tailored to the thesaurus of each data-base. Two authors independently applied the selection criteria. Disagreements were resolved by consensus. The strategy used for Medline is presented in Appendix 1 (Supplemental Digital Content 1, http://links.lww.com/CCM/A782). The Clinical-Trials.gov registry, bibliographies of selected articles, relevant narrative reviews, and clinical practice guidelines (20) were also consulted.

Data AbstractionTwo investigators independently performed data extraction using a standardized, pilot-tested case report form. Any dis-agreement was settled by consensus. Data related to study characteristics, patient baseline characteristics based on admis-sion GCS and radiological findings, pituitary function assess-ment, secondary insults, and clinical outcomes were collected. We defined hypopituitarism as the deficit of at least one ante-rior pituitary hormone, namely, secondary adrenal failure, secondary hypothyroidism, secondary hypogonadism, and growth hormone deficit. A study was considered to have evalu-ated hypopituitarism when all anterior pituitary axes were assessed. Central diabetes insipidus was not considered. For each pituitary disorder, we considered the definitions used by the authors of included studies. When available, we used indi-vidual patient data to evaluate clinical outcomes, predictors, and prevalence of anterior pituitary disorders.

Risk of Bias AssessmentWe evaluated the risk of bias using four key elements of the Guidelines for Assessing Quality and Potential Biases in Prog-nostic studies (21), the revised version of the Quality Assessment of Diagnostic Accuracy Studies tool (22), and the Strengthening the Reporting of Observational studies in Epidemiology state-ment (23). Studies were considered not to be at low risk of bias if inclusion and exclusion criteria were not reported, if study participants were recruited using volunteer sampling strategies (e.g., mail or telephone invitation to participate, flyers to recruit

Copyright (c) Society of Critical Care Medicine and Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited

Review Article

Critical Care Medicine www.ccmjournal.org 3

volunteers), if diagnostic criteria were not described, and if less than 90% of included patients underwent proper diagnostic testing (24). The adequacy of diagnostic testing was determined using consensus definitions (25) (Table 1). Each assessment of an anterior pituitary axis at a specific time frame was evalu-ated for the risk of bias. Therefore, a study evaluating several axes could be considered at low risk of bias for one axis and not being at low risk of bias for another.

Statistical AnalysisWe qualitatively described the results of included studies. For meta-analytical purposes, we pooled risk ratios assessing a spe-cific clinical outcome or predictor across studies, regardless of the pituitary axis being evaluated. When the association between anterior pituitary disorders and specific clinical outcome or pre-dictor was reported for more than one axis, we preferentially used the data related to hypopituitarism or, alternatively, to the most defective axis. When the integrity of an axis was evalu-ated on several occasions, we used the time interval farthest from trauma. Relative risk measures were pooled using Mantel-Haenszel approach and mean differences using inverse variances. To assess point prevalence, we used the Freeman Tukey-type arcsine square root transformation to stabilize the variances of the raw proportions (26, 27). We used Dersimonian-Laird ran-dom effect models. We used Revman v5 (The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark) for meta-analyses of clinical outcomes and predictors and R (R Development Core Team; R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria, 2011) for meta-analyses of prevalence (28). Point estimates are reported with 95% CIs, using I2 values to determine the degree of heterogeneity across studies (29).

Sensitivity AnalysesTo explain heterogeneity, we conducted sensitivity analyses reflecting predefined hypotheses based on study population

for prevalence (studies including more than 10% of mild trau-matic brain injury vs not) and on the risk of bias for clinical outcomes, predictors, and prevalence.

Assessment of Publication BiasPotential publication bias was evaluated using a visual inspection of funnel plots for the primary and secondary objectives when more than 10 studies reported the outcome of interest (30).

RESULTS

Systematic SearchWe retrieved 13,559 articles (Fig. 1). Thirteen thousand one hundred eighty-seven articles were excluded based on title and abstract screening. Of these, 373 publications were assessed for eligibility; 166 did not meet all inclusion criteria, 117 met at least one exclusion criteria, and 17 were duplicate publications. Sixty-six studies were included.

Description of Included StudiesMost studies were published in English (58 studies), while two studies were published in French, two in Spanish, two in Mandarin, one in Portuguese, and one in Persian. We included 22 prospective cohort studies, 2 retrospective stud-ies, 2 case-control studies, 1 randomized trial, and 39 cross-sectional studies. A total of 5,386 patients were evaluated (8–825 patients per study). The mean GCS was 8.0 (95% CI, 7.2–8.8; 20 studies), the mean age was 38.7 years (95% CI, 37.0–40.4; 50 studies), and most were men (74.1%; 95% CI, 71.6–76.6; 51 studies). Forty-eight studies (3,374 patients) included less than 10% mild traumatic brain injury patients. Thirty-seven studies reported inclusion and exclusion crite-ria (2,883 patients), 16 used a nonvoluntary sampling strat-egy (1,231 patients), 63 reported diagnostic criteria (4,296 patients), and 58 used proper diagnostic testing in more than 90% of patients (4,028 patients). Thirty studies (2,292

TABLE 1. Diagnostic Tests Considered as Appropriate for Pituitary Assessment in Different Time Frames

Time of Assessment Growth Hormone DeficitSecondary

Adrenal FailureSecondary

HypopituitarismSecondary

Hypogonadism

Short term (< 3 mo) Insulin-like growth factor-1 ACTH and cortisol TSH and FT4 Testo/E2 + FSH/LH

Insulin tolerance test CRH stimulation TRH stimulation

Glucagon stimulation

Mid term (3–12 mo) Insulin tolerance test ACTH stimulation TSH and FT4 GnRH stimulation

Glucagon stimulation Insulin tolerance test TRH stimulation Testo/E2 + FSH/LH

GHRH-arg stimulation CRH stimulation

Long term (> 12 mo) Insulin tolerance test ACTH stimulation TSH and FT4 GnRH stimulation

Glucagon stimulation Insulin tolerance test TRH stimulation Testo/E2 + FSH/LH

GHRH-arg stimulation CRH stimulation

ACTH = adrenocorticotropic hormone, TSH = thyroid-stimulating hormone, FT4 = free L-thyroxine, Testo = testosterone, E2 = estradiol, FSH = follicle-stimulating hormone, LH = luteinizing hormone, CRH = corticotrophin-releasing hormone, TRH = thyroid-releasing hormone, GnRH = gonadotropin-releasing hormone, GHRH = growth hormone–releasing hormone, Arg = arginine.

Copyright (c) Society of Critical Care Medicine and Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited

Lauzier et al

4 www.ccmjournal.org

patients) were considered at low risk of bias. Included stud-ies reporting clinical outcomes and predictors are outlined in Appendix 2 (Supplemental Digital Content 1, http://links.lww.com/CCM/A783). Studies reporting only prevalence are reported in Appendix 3 (Supplemental Digital Content 1, http://links.lww.com/CCM/A784).

Clinical Outcomes of Anterior Pituitary DisordersFourteen studies included at least one clinical outcome. ICU mortality was reported in four studies (416 patients) (31–34)

(Fig. 2). Anterior pituitary dis-orders were associated with a nonsignificant increase in ICU mortality (risk ratio, 1.79; 95% CI, 0.99–3.21). Only two stud-ies with low risk of bias reported mortality (32, 34), precluding sensitivity analysis. Three stud-ies (204 patients) evaluated the neurological prognosis with the Glasgow Outcome Scale score reported on a continu-ous scale (33, 35, 36). Anterior pituitary disorders were not associated with unfavorable neurological outcome (mean difference, –0.45; 95% CI, –1.10 to 0.20). Only one study evaluated neurological progno-sis using the extended Glasgow Outcome Scale and found no association with anterior pituitary disorders (37). Two studies (88 patients) suggested impaired functional status on the Functional Independence Measure scale with anterior pituitary disorders at reha-bilitation admission (38, 39), whereas two studies (117 patients) suggested impaired improvement of Functional Independence Measure scores during rehabilitation (38, 40). Two studies (120 patients) showed impaired quality of life using the SF-36 questionnaire (37, 41). The limited number of studies for these outcomes precluded pooling the results.

Predictors of Anterior Pituitary DisordersTwenty-seven studies reported at least one potential predic-tor. Among baseline charac-

teristics, increased age was associated with anterior pituitary disorders (mean difference, 3.19; 95% CI, 0.31–6.08; 19 stud-ies; 1,057 patients) (35–53), whereas sex was not (risk ratio for male gender, 1.02; 95% CI, 0.80–1.30; 15 studies; 870 patients) (32, 34–39, 41, 43, 44, 46, 49, 51, 53, 54) (Appendix 4, Supple-mental Digital Content 1, http://links.lww.com/CCM/A785). None of these associations were significant when considering only studies with low risk of bias (data not shown). Injury severity was associated with an increased risk of anterior pitu-itary disorders (risk ratio, 1.91; 95% CI, 1.17–3.13 for patients

Figure 1. Flow diagram of included studies.

Copyright (c) Society of Critical Care Medicine and Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited

Review Article

Critical Care Medicine www.ccmjournal.org 5

with nonmild vs mild traumatic brain injury; six studies; 322 patients and risk ratio, 2.15; 95% CI, 1.20–3.86 for patients with severe vs nonsevere traumatic brain injury; seven studies; 425 patients) (35, 38, 41, 44, 49, 54–56) (Fig. 3). Among CT imaging findings, skull fractures were associated with a greater risk of developing anterior pituitary disorders (relative risk ratio, 1.73; 95% CI, 1.03–2.91; six studies; 357 patients) (33, 35, 37, 41, 54, 57), whereas brain edema at admission CT was not (relative risk ratio, 2.24; 95% CI, 0.69–7.23; five studies; 336 patients) (33, 37, 41, 54, 58) (Fig. 3). We found no study at low risk of bias that appropriately assessed the association between GCS, CT imaging findings, and the risk of developing anterior pituitary disorders. Only two studies quantitatively reported the effect of secondary brain injuries (hypoxemia and hypotension) on anterior pituitary disorders and found no association (33, 37). One study observed a nonsignificant association between anterior pituitary disorders and diffuse axonal injury (40). Two studies evaluated MRI during the recovery phase and suggested that patients with pituitary dis-orders were more likely to have MRI pituitary abnormalities such as empty sella (33, 46).

Prevalence of Anterior Pituitary DisordersSixty-six studies reported the prevalence of anterior pituitary disorders. Estimates are presented in Figure 4 and reported according to the time of hormonal assessment after the injury (< 3 mo, 3–12 mo, and > 12 mo after traumatic brain injury). In studies evaluating all pituitary axes, 31.6% (95% CI, 23.6–40.1%; 27 studies) patients had long-term hypopituitarism, as defined by at least one anterior pituitary disorder (Fig. 4A). We observed significant statistical heterogeneity among all assess-ment time frames and for each anterior pituitary axis (I2 from 77.7% to 96.5%). We performed sensitivity analyses of studies including less than 10% of patients with mild traumatic brain injury and of studies with low risk of bias and observed impor-tant residual heterogeneity (I2, 52.0–98.0%).

Assessment of Publication BiasVisual inspection of the funnel plots of studies evaluating clini-cal outcomes was limited by the paucity of studies retrieved. Inspection of funnel plots suggests a lack of studies with small sample size evaluating the association between anterior pituitary disorders, brain injury severity, and sex.

DISCUSSIONIn this systematic review, we found that approximately one third of patients with blunt traumatic brain injury develop at least one anterior pituitary disorder following their injury. There was significant heterogeneity among studies reporting prevalence that were not solely explained by methodological quality and traumatic brain injury severity. We found that anterior pituitary disorders in traumatic brain injury were associated with a nonsignificant increased risk of death in the ICU, but not with unfavorable neurological outcome. We also found that increased age, traumatic brain injury severity, and skull fractures were associated with a greater risk of developing anterior pituitary disorders, whereas sex and brain edema at admission CT scan were not.

The mechanism underlying the trend toward increased risk of ICU mortality with anterior pituitary disorders remains unclear. In the acute phase of traumatic brain injury, several hormones regulated by the pituitary gland may play an impor-tant role. For instance, untreated corticosteroid deficiency may contribute to hemodynamic instability and results in second-ary brain insult. Other hormones such as estrogen and proges-terone have neuroprotective properties in animal models (59, 60) and may decrease cerebral edema, modulate the inflamma-tory cascade, and reduce apoptosis (60). In our review, stud-ies assessing ICU mortality were underpowered to adjust for important covariates. We cannot exclude that anterior pitu-itary disorders could be markers of illness severity rather than an independent predictor for mortality.

Figure 2. Mortality and Glasgow Outcome Scale (GOS) according to the occurrence of any pituitary disorder at any time of assessment. A risk ratio > 1.0 represents an increased risk of mortality in patients with pituitary disorders. A negative mean difference represents a lower GOS score (i.e., a worse neurological prognosis) in patients with pituitary disorders. M-H = Mantel-Haenszel.

Copyright (c) Society of Critical Care Medicine and Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited

Lauzier et al

6 www.ccmjournal.org

The lack of association between anterior pituitary disorders and neurological recovery may be attributable to multiple fac-tors. First, there may be no association. Second, we retrieved a limited number of studies evaluating neurological outcomes, therefore limiting statistical power. Third, most studies used non-optimal diagnostic testing. The prevalence of anterior pituitary disorders varied greatly depending on the diagnostic criteria and

the diagnostic test being used (2). Therefore, improper identifica-tion of patients with anterior pituitary disorders might have led to biased assessments of the association between anterior pitu-itary disorders and neurological outcomes. Finally, the Glasgow Outcome Scale (14) is relatively insensitive to change (14, 61) and might have failed to capture subtle but clinically relevant changes in neurological outcome related to anterior pituitary disorders.

Figure 3. Association between traumatic brain injury severity, skull fractures, and brain edema with any pituitary disorder at any time of assessment. A risk ratio > 1.0 represents an increased risk of pituitary disorders in patients with nonmild traumatic brain injury (TBI), severe TBI, skull fracture, and brain edema on CT, respectively. M-H = Mantel-Haenszel.

Copyright (c) Society of Critical Care Medicine and Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited

Review Article

Critical Care Medicine www.ccmjournal.org 7

Figure 4. Prevalence of pituitary disorders. Hypopituitarism refers to the presence of at least one anterior pituitary hormone. Only studies evaluating all anterior pituitary axes were considered for this analysis. A, Hypopituitarism (at least one defective axis). B, Growth hormone deficit. C, Secondary adrenal failure. D, Secondary hypothyroidism. E, Secondary hypogonadism. TBI = traumatic brain injury.

Copyright (c) Society of Critical Care Medicine and Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited

Lauzier et al

8 www.ccmjournal.org

We observed an association between age and anterior pitu-itary disorders. Age is a well-established prognostic criterion in traumatic brain injury (62, 63). The propensity for ante-rior pituitary disorders may relate to the decreased neuroplas-ticity observed with aging (64, 65), which to some extent can also impair recovery of hypothalamic-pituitary function (66). We also observed an association between skull fractures and anterior pituitary disorders. Skull fractures may be a surrogate for anatomic lesions of the hypothalamus and pituitary gland (67–69) although the absence of structural abnormalities on imaging does not preclude pituitary dysfunction (70). The asso-ciation between injury severity and anterior pituitary disorders may, in part, be related to an increased risk of secondary brain injury. The anterior pituitary is mainly perfused by the pitu-itary portal veins running through the diaphragm of the sella turcica along the pituitary stalk (71), making it vulnerable to brain edema and intracranial hypertension which occurs more frequently in patients with severe traumatic brain injury (72).

The prevalence of anterior pituitary disorders at different time frames varied highly across studies. This heterogeneity was not solely explained by studies with an unclear or high risk of bias or studies including a significant proportion of patients with mild traumatic brain injury. Other factors such as differ-ences in study population, the timing of hormonal assessment, and the reference tests used may partly explain the heterogene-ity observed.

Our findings are limited by the paucity of studies evaluat-ing clinical outcomes and predictors of anterior pituitary dis-orders. Most studies were not at low risk of bias. Except for the association between brain injury severity and anterior pitu-itary disorders, we observed substantial statistical heterogene-ity that was not solely explained by the methodological quality of included studies. Authors may have reported clinical out-come selectively or descriptively, which makes results unsuit-able for meta-analyses and introduce a reporting bias (30). Most studies did not use multivariable analyses, probably due to small sample sizes, therefore precluding a comprehensive evaluation of the predictors and clinically relevant outcomes after adjustment for other important covariates. Accordingly, it was impossible to evaluate the impact of individual anterior pituitary axis failure on clinical outcomes.

Our systematic review has several strengths. Compared to previous systematic reviews (1, 2), we used a comprehen-sive search strategy with no restriction of language or year of publication. We minimized publication bias by searching rel-evant narrative reviews and the Biosis database, which includes records of several scientific conferences. Our search strategy, combined with the addition of more than 22 studies published since the last systematic review (2), substantially increased the body of literature summarized. We independently performed the study eligibility evaluation, data abstraction, and assess-ment of the risk of bias in duplicate. We assessed the risk of bias of included studies by evaluating several key methodologi-cal aspects (appropriate description of inclusion and exclusion criteria, adequate sampling strategy, proper diagnostic testing, and minimal lost to follow-up). We also used the Freeman

Tukey-type transformation and random effects model to assess pooled prevalence. This approach provides a more accurate estimate of pooled prevalence and accounts for heterogeneity across studies.

CONCLUSIONSIn summary, approximately one third of traumatic brain injury patients will develop at least one anterior pituitary disorder. Older age, brain injury severity, and skull fractures are asso-ciated with an increased risk of anterior pituitary disorders, which in turn may be associated with higher ICU mortality. A limited number of studies at low risk of bias have focused on this topic. Large multicenter cohort studies using standard reference tests, adjusting for important covariates, and limit-ing loss to follow-up are warranted to better define the burden of pituitary disorders in traumatic brain injury patients and to identify high-risk patients for whom a systematic screening strategy would be cost-effective.

ACKNOWLEDGMENTSWe thank Drs. Lucy Lu and Saif Shahrzad who helped translat-ing foreign language articles and Dr. Bobby Senay who helped for data abstraction. We finally thank Xavier Neveu and Stéph-anie Camden for their review of statistical analysis.

REFERENCES 1. Schneider HJ, Kreitschmann-Andermahr I, Ghigo E, et al:

Hypothalamopituitary dysfunction following traumatic brain injury and aneurysmal subarachnoid hemorrhage: A systematic review. JAMA 2007; 298:1429–1438

2. Kokshoorn NE, Wassenaar MJ, Biermasz NR, et al: Hypopituitarism following traumatic brain injury: Prevalence is affected by the use of different dynamic tests and different normal values. Eur J Endocrinol 2010; 162:11–18

3. Hellawell DJ, Taylor RT, Pentland B: Cognitive and psychosocial out-come following moderate or severe traumatic brain injury. Brain Inj 1999; 13:489–504

4. Dikmen SS, Machamer JE, Powell JM, et al: Outcome 3 to 5 years after moderate to severe traumatic brain injury. Arch Phys Med Rehabil 2003; 84:1449–1457

5. Ho KK; 2007 GH Deficiency Consensus Workshop Participants: Consensus guidelines for the diagnosis and treatment of adults with GH deficiency II: A statement of the GH Research Society in associa-tion with the European Society for Pediatric Endocrinology, Lawson Wilkins Society, European Society of Endocrinology, Japan Endocrine Society, and Endocrine Society of Australia. Eur J Endocrinol 2007; 157:695–700

6. Rosén T, Wirén L, Wilhelmsen L, et al: Decreased psychological well-being in adult patients with growth hormone deficiency. Clin Endocrinol (Oxf) 1994; 40:111–116

7. Ciancia S: [Pituitary insufficiency after traumatic brain injury: Consequences? Screening?]. Ann Fr Anesth Reanim 2012; 31:e117–e124

8. Dusick JR, Wang C, Cohan P, et al: Pathophysiology of hypopituita-rism in the setting of brain injury. Pituitary 2012; 15:2–9

9. van der Eerden AW, Twickler MT, Sweep FC, et al: Should anterior pituitary function be tested during follow-up of all patients presenting at the emergency department because of traumatic brain injury? Eur J Endocrinol 2010; 162:19–28

10. Liew A, Thompson CJ: Is hypopituitarism predictable after traumatic brain injury? Nat Clin Pract Endocrinol Metab 2008; 4:126–127

Copyright (c) Society of Critical Care Medicine and Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited

Review Article

Critical Care Medicine www.ccmjournal.org 9

11. Aimaretti G, Ghigo E: Should every patient with traumatic brain injury be referred to an endocrinologist? Nat Clin Pract Endocrinol Metab 2007; 3:318–319

12. Klose M, Feldt-Rasmussen U: Does the type and severity of brain injury predict hypothalamo-pituitary dysfunction? Does post-traumatic hypopituitarism predict worse outcome? Pituitary 2008; 11:255–261

13. Dusick JR, Wang C, Cohan P, et al: Pathophysiology of hypopituita-rism in the setting of brain injury. Pituitary 2012; 15:2–9

14. Jennett B, Bond M: Assessment of outcome after severe brain dam-age. Lancet 1975; 1:480–484

15. Jennett B, Snoek J, Bond MR, et al: Disability after severe head injury: Observations on the use of the Glasgow Outcome Scale. J Neurol Neurosurg Psychiatry 1981; 44:285–293

16. Guilfoyle MR, Seeley HM, Corteen E, et al: Assessing quality of life after traumatic brain injury: Examination of the short form 36 health survey. J Neurotrauma 2010; 27:2173–2181

17. Granger C, Hamilton B, Keith R, et al: Advances in functional assess-ment for medical rehabilitation. Top Geriatr Rehabil 1986; 1:59–74

18. Teasdale G, Jennett B: Assessment of coma and impaired conscious-ness. A practical scale. Lancet 1974; 2:81–84

19. Mercier E, Boutin A, Lauzier F, et al: Predictive value of S-100β protein for prognosis in patients with moderate and severe traumatic brain injury: Systematic review and meta-analysis. BMJ 2013; 346:f1757

20. Ghigo E, Masel B, Aimaretti G, et al: Consensus guidelines on screening for hypopituitarism following traumatic brain injury. Brain Inj 2005; 19:711–724

21. Hayden JA, Côté P, Bombardier C: Evaluation of the quality of prognosis studies in systematic reviews. Ann Intern Med 2006; 144:427–437

22. Whiting PF, Rutjes AW, Westwood ME, et al; QUADAS-2 Group: QUADAS-2: A revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med 2011; 155:529–536

23. von Elm E, Altman DG, Egger M, et al; STROBE Initiative: The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: Guidelines for reporting observational studies. Lancet 2007; 370:1453–1457

How much is too much? Eur J Epidemiol 2004; 19:751–760 25. Schneider HJ, Aimaretti G, Kreitschmann-Andermahr I, et al:

Hypopituitarism. Lancet 2007; 369:1461–1470 26. Stuart A, Ord JK: Kendall’s Advanced Theory of Statistics. Sixth

Edition. London, England, Arnold Publishers, 1994 27. Jaynes ET: Confidence intervals vs. Bayesian intervals. In: Foundations

of Probability Theory, Statistical Inference, and Statistical Theories of Science. Harper WL, Hooker CA (Eds). Dordrecht, The Netherlands, Reidel, 1976, pp 175–257

28. R Development Core Team: R: A Language and Environment for Statistical Computingfor Statistical Computing, 2012

29. Higgins JP, Thompson SG, Deeks JJ, et al: Measuring inconsistency in meta-analyses. BMJ 2003; 327:557–560

30. Sterne JA, Sutton AJ, Ioannidis JP, et al: Recommendations for exam-ining and interpreting funnel plot asymmetry in meta-analyses of ran-domised controlled trials. BMJ 2011; 343:d4002

31. Matsuura H, Nakazawa S, Wakabayashi I: Thyrotropin-releasing hor-mone provocative release of prolactin and thyrotropin in acute head injury. Neurosurgery 1985; 16:791–795

32. Llompart-Pou JA, Raurich JM, Pérez-Bárcena J, et al: Acute hypo-thalamic-pituitary-adrenal response in traumatic brain injury with and without extracerebral trauma. Neurocrit Care 2008; 9:230–236

33. Krahulik D, Zapletalova J, Frysak Z, et al: Dysfunction of hypothalamic-hypophysial axis after traumatic brain injury in adults. J Neurosurg 2010; 113:581–584

34. Dimopoulou I, Tsagarakis S, Theodorakopoulou M, et al: Endocrine abnormalities in critical care patients with moderate-to-severe head trauma: Incidence, pattern and predisposing factors. Intensive Care Med 2004; 30:1051–1057

35. Bondanelli M, De Marinis L, Ambrosio MR, et al: Occurrence of pituitary dysfunction following traumatic brain injury. J Neurotrauma 2004; 21:685–696

36. Jeong JH, Kim YZ, Cho YW, et al: Negative effect of hypopituita-rism following brain trauma in patients with diffuse axonal injury. J Neurosurg 2010; 113:532–538

37. Bavisetty S, Bavisetty S, McArthur DL, et al: Chronic hypopituitarism after traumatic brain injury: Risk assessment and relationship to out-come. Neurosurgery 2008; 62:1080–1093; discussion 1093–1094

38. Bondanelli M, Ambrosio MR, Cavazzini L, et al: Anterior pituitary func-tion may predict functional and cognitive outcome in patients with traumatic brain injury undergoing rehabilitation. J Neurotrauma 2007; 24:1687–1697

39. Bondanelli M, Ambrosio MR, Margutti A, et al: Evidence for integrity of the growth hormone/insulin-like growth factor-1 axis in patients with severe head trauma during rehabilitation. Metabolism 2002; 51:1363–1369

40. Park KD, Kim DY, Lee JK, et al: Anterior pituitary dysfunction in moder-ate-to-severe chronic traumatic brain injury patients and the influence on functional outcome. Brain Inj 2010; 24:1330–1335

41. Wachter D, Gündling K, Oertel MF, et al: Pituitary insufficiency after traumatic brain injury. J Clin Neurosci 2009; 16:202–208

42. Agha A, Phillips J, O’Kelly P, et al: The natural history of post-traumatic hypopituitarism: Implications for assessment and treatment. Am J Med 2005; 118:1416

43. Agha A, Rogers B, Sherlock M, et al: Anterior pituitary dysfunction in survivors of traumatic brain injury. J Clin Endocrinol Metab 2004; 89:4929–4936

44. Klose M, Juul A, Poulsgaard L, et al: Prevalence and predictive fac-tors of post-traumatic hypopituitarism. Clin Endocrinol (Oxf) 2007; 67:193–201

45. Lieberman SA, Oberoi AL, Gilkison CR, et al: Prevalence of neuroen-docrine dysfunction in patients recovering from traumatic brain injury. J Clin Endocrinol Metab 2001; 86:2752–2756

46. Schneider HJ, Sämann PG, Schneider M, et al: Pituitary imaging abnormalities in patients with and without hypopituitarism after trau-matic brain injury. J Endocrinol Invest 2007; 30:RC9–RC12

of traumatic brain injury (TBI) and its possible relation with cognitive dis-abilities and mental distress. J Endocrinol Invest 2004; 27:1048–1054

48. Schneider HJ, Schneider M, Saller B, et al: Prevalence of anterior pituitary insufficiency 3 and 12 months after traumatic brain injury. Eur J Endocrinol 2006; 154:259–265

49. Tanriverdi F, Ulutabanca H, Unluhizarci K, et al: Three years prospec-tive investigation of anterior pituitary function after traumatic brain injury: A pilot study. Clin Endocrinol (Oxf) 2008; 68:573–579

50. Zgaljardic DJ, Guttikonda S, Grady JJ, et al: Serum IGF-1 concentra-tions in a sample of patients with traumatic brain injury as a diagnostic marker of growth hormone secretory response to glucagon stimula-tion testing. Clin Endocrinol (Oxf) 2011; 74:365–369

51. Peñalver D, Pavón I, Olivar J, et al: Evaluación de la función hipofisaria tras episodio neurológico grave. Endocrinol Nutr 2005; 52:283–289

52. Carlson NE, Brenner LA, Wierman ME, et al: Hypogonadism on admission to acute rehabilitation is correlated with lower functional status at admission and discharge. Brain Inj 2009; 23:336–344

53. McCarthy CF, Wills MR, Keane PM, et al: The Su-4885 (Methopyrapone) response after head injury. J Clin Endocrinol Metab 1964; 24:121–124

54. Kelly DF, Gonzalo IT, Cohan P, et al: Hypopituitarism following trau-matic brain injury and aneurysmal subarachnoid hemorrhage: A pre-liminary report. J Neurosurg 2000; 93:743–752

55. Aimaretti G, Ambrosio MR, Di Somma C, et al: Residual pituitary function after brain injury-induced hypopituitarism: A prospective 12-month study. J Clin Endocrinol Metab 2005; 90:6085–6092

56. Klose M, Juul A, Struck J, et al: Acute and long-term pituitary insuf-ficiency in traumatic brain injury: A prospective single-centre study. Clin Endocrinol (Oxf) 2007; 67:598–606

57. Richard I, Rome J, Lemené B, et al: [Post-traumatic endocrine defi-cits: Analysis of a series of 93 severe traumatic brain injuries]. Ann Readapt Med Phys 2001; 44:19–25

Copyright (c) Society of Critical Care Medicine and Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited

Lauzier et al

10 www.ccmjournal.org

58. Agha A, Thornton E, O’Kelly P, et al: Posterior pituitary dysfunc-tion after traumatic brain injury. J Clin Endocrinol Metab 2004; 89:5987–5992

59. Wigginton JG, Pepe PE, Idris AH: Rationale for routine and immediate administration of intravenous estrogen for all critically ill and injured patients. Crit Care Med 2010; 38:S620–S629

60. Stein DG, Wright DW, Kellermann AL: Does progesterone have neu-roprotective properties? Ann Emerg Med 2008; 51:164–172

61. Hall K: Overview of functional assessment scales in brain injury reha-bilitation. NeuroRehabilitation 1992; 2:98–113

62. Perel P, Arango M, Clayton T, et al; MRC CRASH Trial Collaborators: Predicting outcome after traumatic brain injury: Practical prognostic models based on large cohort of international patients. BMJ 2008; 336:425–429

63. Mushkudiani NA, Engel DC, Steyerberg EW, et al: Prognostic value of demographic characteristics in traumatic brain injury: Results from the IMPACT study. J Neurotrauma 2007; 24:259–269

64. Kempermann G, Gast D, Gage FH: Neuroplasticity in old age: Sustained fivefold induction of hippocampal neurogenesis by long-term environmental enrichment. Ann Neurol 2002; 52:135–143

65. Timaru-Kast R, Luh C, Gotthardt P, et al: Influence of age on brain edema formation, secondary brain damage and inflammatory response after brain trauma in mice. PLoS One 2012; 7:e43829

66. Fu Q, Gremeaux L, Luque RM, et al: The adult pituitary shows stem/progenitor cell activation in response to injury and is capable of regeneration. Endocrinology 2012; 153:3224–3235

67. Crompton MR: Hypothalamic lesions following closed head injury. Brain 1971; 94:165–172

68. Kornblum RN, Fisher RS: Pituitary lesions in craniocerebral injuries. Arch Pathol 1969; 88:242–248

69. Ceballos R: Pituitary changes in head trauma (analysis of 102 con-secutive cases of head injury). Ala J Med Sci 1966; 3:185–198

70. Benvenga S, Campenní A, Ruggeri RM, et al: Clinical review 113: Hypopituitarism secondary to head trauma. J Clin Endocrinol Metab 2000; 85:1353–1361

71. Yuan XQ, Wade CE: Neuroendocrine abnormalities in patients with traumatic brain injury. Front Neuroendocrinol 1991; 12:209–230

72. Narayan RK, Kishore PR, Becker DP, et al: Intracranial pressure: To monitor or not to monitor? A review of our experience with severe head injury. J Neurosurg 1982; 56:650–659