TRAUMATIC BRAIN INJURY (TBI) was recognized as a

Transcription

1 463 Traumatic Brain Injury as a Relevant Cause of Growth Hormone Deficiency in Adults: A KIMS-Based Study Felipe F. Casanueva, MD, PhD, Alfonso Leal, MD, Maria Koltowska-Häggström, MD, Peter Jonsson, Miklós I. Góth, MD ABSTRACT. Casanueva FF, Leal A, Koltowska-Häggström M, Jonsson P, Góth MI. Traumatic brain injury as a relevant cause of growth hormone deficiency in adults: a KIMS-based study. Arch Phys Med Rehabil 2005;86: Objectives: To characterize further the clinical manifestations and the efficacy of growth hormone (GH) replacement therapy in patients with adult-onset growth hormone deficiency (GHD) reported in the KIMS (Pfizer s international metabolic database) as caused by traumatic brain injury (TBI) and to compare them with nonirradiated patients whose GHD was due to a nonfunctioning pituitary adenoma (NFPA). Design: Observational study. Setting: Subjects selected from the KIMS database. Participants: Fifty-one patients with GHD resulting from TBI and 688 patients with GHD resulting from NFPA. Both groups were selected from the KIMS and had adult-onset GHD with GH replacement therapy only after KIMS entry and before and after KIMS entry. Interventions: Not applicable. Main Outcome Measures: Age, body mass index, age at disease onset, age at disease diagnosis, age at KIMS entry, final height, GH peak at testing, GH replacement dose, routine biochemical analysis, clinical manifestations of disease, and quality of life measurements. Results: Patients with TBI were significantly younger at study entry and were younger both at pituitary disease onset and at GHD diagnosis, but they showed a significant delay in treatment. When comparing patients not treated with GH before entering in the KIMS, patients with TBI were significantly shorter ( cm) than those with NFPA ( cm) in final height. TBI patients had lower GH reserves than NFPA patients, and although the latter group experienced more positive changes, both groups benefited from GH replacement therapy. Conclusions: Patients with GHD due to TBI showed a significant reduction in height and a reduction in pituitary GH From the Department of Medicine, Endocrine Section, School of Medicine and Complejo Hospitalario Universitario de Santiago, University of Santiago de Compostela, Santiago de Compostela, Spain (Casanueva); Division of Endocrinology, Hospital Virgen del Rocio, Seville, Spain (Leal); KIGS/KIMS Outcomes Research, Pharmacia AB, Stockholm, Sweden (Koltowska-Häggström, Jonsson); and Department of Medicine, National Medical Center, Budapest, Hungary (Góth). Supported by the Fondo de Investigación Sanitaria and the Instituto de Salud Carlos III, Ministerio de Sanidad y Consumo, Red de Grupos RGTO (grant no. G03/028), Red de Centros RCMN (grant no. C03/08), Secretaría Xeral de Investigación e Desenvolvemento (grant no. PGIDIT02BTF91801PR), Xunta de Galicia, the Ministerio Español de Ciencia y Tecnología, the Hungarian Scientific Research Fund (grant no. T035216), and the Hungarian Medical Research Council (grant no. 022/2003). No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the author(s) or on any organization with which the author(s) is/are associated. Reprint requests to Felipe F. Casanueva, PO Box 563, Santiago de Compostela E-15780, Spain, endocrine@usc.es /05/ $30.00/0 doi: /j.apmr reserve and were diagnosed and treated with inappropriate delay. Key Words: Brain injuries; Growth hormone deficiency dwarfism; Head trauma; Rehabilitation by American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation TRAUMATIC BRAIN INJURY (TBI) was recognized as a cause of neuroendocrine dysfunction more than 50 years ago. 1-4 More recently, several case reports and original articles have documented that head trauma of any cause may induce a variety of hormonal disabilities and their combinations Although diabetes insipidus and hyperprolactinemia are associated in most textbooks with sequelae of TBI, their presence is actually infrequent and is occasionally transitory. 12 It is not uncommon for the diagnosis of pituitary dysfunction after TBI to be made only several years after the traumatic event, and in some cases the diagnosis has been made more than 35 years after the original trauma. 13 The hormonal deficit contributes to the chronic disability of TBI and its cognitive, physical, and social sequelae and to the reduction of quality of life (QOL) in these patients In any case, it is widely accepted that head trauma may result in hypopituitarism, but the prevalence and disease characteristics are unknown. All pituitary hormones, either isolated or in combination, may be affected by TBI, with growth hormone (GH) and gonadotropins being the most fragile after traumatic injury, even after mild traumas that do not require hospitalization and that patients are unable to remember. 11 Because the subtle clinical manifestations of head injury, such as decrease in strength, loss of working capacity, and impaired QOL, are reminiscent of the clinical manifestations of growth hormone deficiency 17,18 (GHD), studies have recently been undertaken to determine the prevalence of neuroendocrine dysfunction and particularly of GHD in adults with past brain injury. 19,20 Interest in this topic was spurred by the report that after specific selective testing, GHD may be observed in up to 15% of patients with previous TBI, even in the absence of clinical manifestations. 21 Because the prevalence and clinical characterization of GHD in patients who have suffered a TBI in the past are not well known and prospective studies will need a considerable number of years, our study took a diverse approach, analyzing the patients surveyed in the KIMS (Pfizer s international metabolic database). Initiated in 1994, the KIMS is a pharmacoepidemiologic survey for monitoring the safety and outcomes of GH replacement therapy and to improve understanding of GHD in adults. To date, the KIMS contains data on 8500 patients with GHD from 28 countries worldwide, and practically all causes of GHD are well represented in the database. The aim of our work was 2-fold: (1) to further characterize the clinical and biochemical manifestations of GHD due to TBI and (2) to evaluate the efficacy of GH replacement therapy in these patients.

2 464 BRAIN INJURY AND GH DEFICIENCY, Casanueva Table 1: Pituitary Hormone Deficiencies Other Than GH in the Total Cohort and in Naive Patients Total Cohort Group Total Status ACTH TSH LH/FSH ADH TBI 51 (100) Deficient 37 (72) 39 (76) 45 (88) 10 (19) Substituted 35 (68) 38 (74) 12 (23) * NFPA 686 (100) Deficient 522 (75) 508 (73) 577 (84) 161 (23) Substituted 492 (71) 474 (69) 100 (14) * Naive Patients Group Total Status ACTH TSH LH/FSH ADH TBI 25 (100) Deficient 22 (88) 23 (92) 25 (100) 5 (20) Substituted 21 (84) 22 (88) 6 (24) * NFPA 415 (100) Deficient 333 (80) 329 (79) 373 (84) 103 (24) Substituted 319 (76) 309 (74) 71 (17) * NOTE. Values are n (%). *ADH deficiencies were transitory or were substituted some time ago. METHODS From the pharmacoepidemiologic survey of adults with GHD contained in the KIMS, cases of patients with adult-onset GHD caused by TBI were identified and studied. To allow meaningful comparisons, it was decided to use patients with adult GHD due to nonfunctioning pituitary adenoma (NFPA) as controls. Anthropometric, biochemical, and clinical variables were analyzed within the groups and were compared in a stepwise analysis. For the data analysis, the KIMS terminology was used. In each patient, 4 relevant episodes were considered. The first was the initial event in this case, the head trauma episode or the beginning of NFPA development. The data for such an initial event are either not known, as in NFPA, or are not recorded or remembered, as in TBI. The second episode, pituitary disease onset, refers to the first date on which the clinical manifestations of hypopituitarism (signs, symptoms) were collected. It comes from a retrospective analysis of the clinical history performed by the physician and occurs years after the initial event. Third, the GHD diagnosis, is the date on which the biochemical tests unambiguously demonstrate a GHD. Fourth, the database or KIMS entry, is the date of entry in the registry and when replacement therapy with GH starts. Patients never treated with GH before KIMS entry were called naive, and those treated both before and after KIMS entry were called nonnaive. In the first initial analysis, inclusion criteria for both groups were (1) patients were followed up in the KIMS after diagnosis of GHD, according to international criteria 22 ; (2) GHD occurred exclusively with adult onset; (3) patients were not irradiated; and (4) patients had been treated, at least after inclusion in the KIMS, with recombinant human GH (recombinant somatropin [Genotropin]), to allow for the analysis of the response to the replacement therapy. Thus, 739 patients of both sexes were studied: 51 were patients with TBI (34 men, 17 women), aged years, and 688 patients (438 men, 250 women) with NFPA, aged years, served as controls. The demographic analysis was performed in this initial total cohort. To analyze the effect of previous treatment with GH on final height, each of the 2 groups of the total cohort were further analyzed after being divided into 2 groups: those with no GH treatment at all before starting GH treatment after enrollment in the KIMS (naive) and patients with continuous GH treatment before and after entry into KIMS (nonnaive). In a third analysis, TBI and NFPA groups were strictly matched for gender and age at pituitary disease onset and by age at KIMS database entry. A further matching condition was that patients from both groups must have never been treated with GH before entering the database (naive). Thus, a total of 140 patients were studied: 15 patients with TBI, aged years, and 125 patients diagnosed with NFPA, aged years, who were controls. In this matched cohort, clinical and biochemical variables were studied both before and after 1 year of treatment with GH replacement therapy. The data are presented as mean standard error (SE) unless otherwise stated. Differences between groups were sought using the Wilcoxon rank-sum test. Values of P less than.05 were considered significant. RESULTS In the analysis of the unselected population (total cohort), in both groups that is, TBI (n 51) and NFPA (n 688) the sex distribution was similar (men with TBI, 64.7%; men with NFPA, 63.8%), and there were no statistical differences in height, weight, or body mass index (BMI). The number of pituitary deficiencies in addition to GHD was also similar in the 2 groups, with a tendency of most patients with TBI to have 3 additional deficiencies (table 1). In both groups, the most frequent additional deficiencies were luteinizing hormone (LH) and follicle stimulating hormone (FSH), followed by adrenocorticotrophic hormone (ACTH), thyroid-stimulating hormone (TSH), and antidiuretic hormone (ADH). Deficiencies were replaced according to standard criteria after diagnosis. The mean GH peak at diagnosis was g/l in the patients with TBI, which was not significantly different from that of the NFPA subjects ( g/l). The tests used for GHD diagnosis were the insulin tolerance test, followed in frequency by the arginine test, the glucagon test, and other tests. The insulinlike growth factor I (IGF-I) standard deviation (SD) score, the Quality of Life Assessment of Growth Hormone Deficiency in Adults (QOL-AGHDA) score, and the medical histories of both groups were similar. Clinical variables reflected in the database of patients with adult GHD, such as intermittent claudication, coronary heart disease, stroke, arthrosis, diabetes mellitus, convulsive disease, neoplasia, heart rate, and triglyceride levels, were similar in patients with TBI and in the patients with NFPA (data not shown). Some of the most relevant clinical variables are shown in table 2. Glucose, glycosylated hemoglobin (Hb A 1c ), and systolic blood pressure

3 BRAIN INJURY AND GH DEFICIENCY, Casanueva 465 Table 2: Clinical and Biochemical Variables From Patients With GHD due to TBI and NFPA Variable TBI Mean SE Between Groups NFPA Mean SE Age (y) Age at pituitary disease onset (y) Time from pituitary disease to KIMS (y) * Time from GHD diagnosis to KIMS (y) NS BMI (kg/m 2 ) NS Waist (cm) NS Diastolic blood pressure (mmhg) NS Systolic blood pressure (mmhg) Heart rate (beat/minute) NS BIA body fat (kg) NS BIA lean body mass (kg) NS Glucose (mg/dl) * Hb A 1c (%) Height (cm) NS Cholesterol (mg/dl) NS LDL cholesterol (mg/dl) NS HDL cholesterol (mg/dl) NS NOTE. To convert cholesterol values to mmol/l, divide by The statistical comparisons are between groups. Abbreviations: BIA, bioimpedence analysis; HDL, high-density lipoprotein; LDL, low-density lipoprotein; NS, not significant. *P.05. P.01. P.001. (SBP) were significantly lower in patients with TBI than in those with NFPA. Patients with head trauma were younger at entry into the KIMS (TBI, y; NFPA, y; P.001), younger at pituitary disease onset (TBI, y; NFPA, y; P.001), and younger at GHD diagnosis (TBI, y; NFPA, y; P.001). However, TBI patients showed longer intervals between diagnosis and treatment, reflected in the time from pituitary disease onset to KIMS entry for the TBI patients ( y) and the NFPA subjects ( y) (P.02). Patients with TBI did not take a significantly longer time from GHD diagnosis to initiation of GH treatment ( y) than did patients with NFPA ( y) (fig 1). To understand whether the type of GH replacement therapy affected some biologic parameters, both groups were divided into those without treatment until entry into the database (naive) and those with continuous treatment from GHD diagnosis to database entry and thereafter (nonnaive). In the naive group (TBI, n 25; NFPA, n 415), TBI patients, compared with NFPA patients, were shorter at database entry ( cm vs cm, respectively; P.05) (fig 2) and had lower IGF-I SD scores ( vs , respectively; P.05). Patients with TBI were also significantly younger and had lower SBP than patients with NFPA (data not shown). However, in the nonnaive patients, there were no significant differences in height ( cm vs cm) or in IGF-I SD scores ( vs ) between the patients with TBI and NFPA, respectively (fig 2). Also in this group, TBI patients were younger and had significantly lower Hb A 1c values, but no differences in blood pressure were observed (data not shown). In these nonnaive groups, GH replacement before KIMS entry was undertaken for years in patients with TBI and for years in those with NFPA. When the cohort was restricted to those patients without a previous treatment with GH before entering into the KIMS database (naive) and when the 2 groups were matched for sex, age at pituitary disease onset, and age at database entry, a total of 15 patients with TBI and 125 with NFPA were studied. Patients of these matched cohorts were analyzed both before and 1 year after treatment with recombinant GH. The clinical and biochemical changes in both groups after 1 year of GH replacement therapy are shown in table 3 as a delta change. Within the TBI group, after 1 year of GH replacement therapy, a significant clinical improvement was observed in the IGF-I SD scores; in the NFPA group an improvement was observed in the waist, body fat mass (reduction) and lean body mass, QOL-AGHDA scores, and total and low-density lipoprotein cholesterol (reduction), and an increase was observed in blood glucose. Although improvement of some variables was more evident and significant in patients with NFPA, both groups benefited from the 1-year GH replacement therapy. Interestingly, patients with TBI required a higher maintenance dose of GH (.51.24mg/d) than patients with NFPA (.39.02mg/d) (delta increments.30.06mg/d and.14.02mg/d, respectively; P.05) to sustain a similar normal age- and sex-related IGF-I level. Obviously, the height differences persisted: patients with TBI were cm tall and those with NFPA were cm tall (P.05) (fig 2). To understand whether the height differences observed throughout the study were due to genetic or ethnic reasons, an analysis of patients countries or regions of origin was performed. Patients with the diagnoses of TBI and NFPA came from similar origins, which made us discount an ethnic explanation for the differences observed (data not shown). In this matched cohort, no differences were observed in the number or in the rate of replacement therapy of pituitary hormone deficiencies other than GH (table 1). DISCUSSION Our work has systematically observed that adult patients with GHD due to TBI are significantly shorter and were treated later than controls that is, patients with GHD due to NFPA. To summarize our main findings, data can be grouped into (1) time until diagnosis, (2) GH reserve, (3) height, and (4) effectiveness of GH replacement.

4 466 BRAIN INJURY AND GH DEFICIENCY, Casanueva Considering that in both groups of patients, subjects with GHD of childhood onset were excluded, the differences between time at pituitary disease onset and at the start of treatment were remarkable. Patients with TBI were younger at database entry, younger at pituitary disease onset, and younger at GHD diagnosis than patients with NFPA; however, they were treated later than NFPA patients in terms of time elapsed between pituitary disease onset and time of treatment initiation (KIMS entry). Interestingly, once the GHD diagnosis was established by biochemical testing, the time to treatment initiation was roughly similar, which suggests that the delay of TBI patient treatment resides in the late detection of the signs and symptoms of pituitary disease. The earlier onset of pituitary disease after the TBI episode fits well with data from the literature that indicate that (1) head trauma causing hypopituitarism is due mostly to road collisions; (2) it is a medical problem most common in young men, with the male-to-female ratio of 5:1; (3) it occurs between the ages of 11 and 29; and (4) the head trauma can be minor but still induce significant impairment in pituitary function. 11 The second observation is also a common finding in other studies; in TBI-mediated hypopituitarism, the onset of the pituitary disease manifestations occurred several years earlier than GHD diagnosis and treatment. 5,12,23 However, there was also a considerable time from pituitary disease onset to GHD diagnosis in patients with NFPA, probably because of the low rate of tumor growth and the fact that the slower the growth rate, the more insidious the clinical manifestations of a brain tumor. Still, the data indicated that delay in GHD diagnosis was greater in patients with TBI than in those with NFPA. Considering the low number of patients with TBI in the whole database cohort compared with the astonishing number of car accidents worldwide, it is more than possible that a considerable number of patients with TBI-mediated GHD are in fact undiagnosed. The delayed diagnosis and the low rate of diagnosis seem to suggest a general unawareness of the incidence of hypopituitarism after brain injury and that patients with TBI experience an inappropriate delay in diagnosis and in starting treatment. The finding that in previously untreated patients with GHD, those with TBI were more than 4cm shorter than controls was new and unexpected information. A careful analysis was undertaken to ascertain any bias in the interpretation of data or in the sorting of subjects. A bias based on ethnic origin was Fig 1. Patients from the total cohort with GHD due to TBI or NFPA. (A) Boxplot (showing median, upper and lower quartile, and minimum and maximum levels) of age, age at pituitary disease onset, and age at the diagnosis of GHD. (B) Boxplot of years from pituitary disease onset until GH treatment (KIMS entry) and from GHD diagnosis to GH treatment. *P<.05. Fig 2. Mean height SE in patients with GHD due to TBI compared with those with GHD due to NFPA. The total cohort was originally divided into those with GH treatment before and after entry into the KIMS database (nonnaive) (left columns) and those with GH treatment only after entry in the database (naive) (middle columns). In the right columns, naive patients from the matched cohort (matched by age, sex, age at pituitary disease onset, age at GH replacement initiation) after 1 year with GH replacement therapy. *P<.05.

5 BRAIN INJURY AND GH DEFICIENCY, Casanueva 467 Table 3: Changes With Respect to Basal (Delta Values) of Clinical and Biochemical Variables From Patients With GHD due to TBI and to NFPA, After 1 Year of GH Replacement Therapy TBI NFPA Variable Mean SE Within Group Between Groups Mean SE Within Group Dose (mg/d) * Height (cm) NS * NS BMI (kg/m 2 ) NS NS NS Waist (cm) NS NS * BIA body fat mass (kg) NS NS * BIA body lean mass (kg) NS NS Systolic blood pressure (mmhg) NS NS NS Diastolic blood pressure (mmhg) NS NS NS Hb A 1c (%) NS NS NS QOL-AGHDA score NS NS IGF-I, SD score NS Cholesterol (mg/dl) NS NS * HDL cholesterol (mg/dl) NS NS LDL cholesterol (mg/dl) NS NS * Glucose (mg/dl) NS NS NOTE. The statistical analyses are within group and between groups. *P.05. P.01. P.001. discarded, because a systematic study of the origin of cases showed that the number of patients with TBI or with NFPA included in this study was similar in the different countries. Because the loss in centimeters is small and height in the normative population is quite variable, the absence of similar findings in other published studies is not surprising: only after comparing a large number of subjects does this finding become evident. However, the age at pituitary disease onset cannot fully explain the height loss. In the third part of our study, GHD patients of both TBI and NFPA were strictly matched by sex, age, and age at disease onset, but the height differences were still significant. The reduction in height in TBI patients suggests 2 things: (1) that GHD was established at an earlier age that is, when growth was not finished yet, and (2) that most TBI patients received GH replacement too late in their lives for it to be effective. The third observation from our study was the direct and inferential evidence indicating that patients with GHD due to TBI had less GH reserve than patients with NFPA. Inferential information is that TBI patients have lower blood glucose and Hb A 1c than NFPA patients. As direct evidence, they were significantly shorter, had lower IGF-I values, and needed higher GH doses of replacement therapy to maintain the target IGF-I levels. Furthermore, based on the database, the GH peak obtained at diagnosis tended to be lower in the TBI group. The above considerations suggest that the GHD in patients with TBI may be more pronounced than in other causes of GHD for example, NFPA although the pathologic basis of this finding is not clear. It is of particular interest that the chronic neurobehavioral problems and the severe reduction in QOL that affect many patients with brain injury are similar to those described for adult subjects with GHD. 24 However, when such symptoms are seen, brain concussion is usually cited as the cause, and the endocrinopathy is often not considered. This should be addressed, because hypopituitarism-related neurobehavioral problems normally improve with hormonal replacement therapy. 10,20,25,26 Another observation was that after 1 year of replacement therapy with GH, the rate of improvement in several biochemical variables was observed in both patients with TBI and those with NFPA, although more evident in the latter. This suggests that early diagnosis and treatment is unambiguously beneficial for patients with TBI and GHD. We are well aware of the limitations of our study. Because our database was set up specifically for patients with GHD who exhibited the signs or symptoms of GHD beforehand, it is not possible to state the prevalence of GHD due to TBI in the general population. In addition, a cross-sectional study presents its own inherent limitations, more so when a small number of subjects is studied. However, the stringent selection of the initial cohort and the rigorous matching of both groups in the subsequent analysis may add power to the results presented, because it eliminates several confounding variables. Although hypopituitarism secondary to TBI was described more than 50 years ago, it was initially reported anecdotally, and a systematic study was lacking. The topic has recently received greater attention since, thanks to more sophisticated methods of diagnosis, it became evident that a considerable number of patients experience a definitive GHD after TBI, even after trauma of mild intensity. 11,21 This work endorses the view that a large number of post-tbi patients have GHD but are undiagnosed, despite the full clinical manifestation of hormonal deficiencies. 5,11 No definitive anatomic explanation has been provided for the GHD that follows head trauma. In fact, pituitary lesions, such as anterior lobe infarcts, posterior lobe hemorrhages, and partial or complete transection of the stalk, have been reported. 27,28 In addition, hypothalamic lesions were reported in 50% of head trauma cases, including microhemorrhages and ischemic lesions, which may sometimes be bilateral. 29,30 In any case, it is worth noting that there is a relatively low number of cases with TBI reported as the cause of GHD in most clinical units, even with an expanding number of TBIs due mostly to automobile collisions. 31 This fact supports the view that in the last few decades, diagnosis of GHD has been overlooked in a large proportion of patients with TBI. CONCLUSIONS Adult patients with GHD due to TBI show a significant reduction in height and pituitary GH reserve, experience an

6 468 BRAIN INJURY AND GH DEFICIENCY, Casanueva undue delay in both diagnosis and replacement therapy, and are probably frequently undiagnosed. Acknowledgment: We gratefully acknowledge the technical collaboration of Mary Lage. References 1. Escamilla RF, Lisser H. Simmonds disease. J Clin Endocrinol 1942;2: Porter RJ, Miller RA. Diabetes insipidus following closed head injury. J Neurol Neurosurg Psychiatry 1948;11: Witter M, Tasher R. Hypophysar-hypothalamische kramkheitsbilder mach stumptem schadeltrauma. Fortschr Neurol Psychiatr 1957;25: Cyran E. Hypophysenschadigung durch Schadelbasisfraktur. Deutch Med Wochenschrift 1918;44: Edwards OM, Clark JD. Post-traumatic hypopituitarism. Six cases and review of the literature. Medicine (Baltimore) 1986;65: Altman R, Pruzanski W. Post-traumatic hypopituitarism. Ann Intern Med 1961;55: Rudman D, Fleischer AS, Kutner MH, Raggio JF. Suprahypophyseal hypogonadism and hypothyroidism during prolonged coma and head trauma. J Clin Endocrinol Metab 1977;45: Fleischer AS, Rudman DR, Payne NS, Tindall GT. Hypothalamic hypothyroidism and hypogonadism in prolonged traumatic coma. J Neurosurg 1978;49: Castaner MF, Ayma J, Martinez MJ, Vilardell E. Diabetes insipida e hipopituitarismo postraumáticos. Consideraciones diagnósticas a propósito de tres casos. Med Clin 1981;78: Vance ML. Hypopituitarism. N Engl J Med 1994;330: Benvenga S, Campenni A, Ruggeri RM, Trimarchi F. Hypopituitarism secondary to head trauma. J Clin Endocrinol Metab 2000; 85: Eiholzer U, Zachman M, Gnehm HE, Prader A. Recovery from post-traumatic anterior pituitary insufficiency. Eur J Pediatr 1986; 145: Gunn IR, Beastall GH, Matthews DM, Bath JC. Post-traumatic hypothalamic-pituitary dysfunction presenting with biochemical features of primary hypothyroidism. Ann Clin Biochem 1991; 28(Pt 4): Mazaux JM, Richer E. Rehabilitation after traumatic brain injury in adults. Disabil Rehabil 1998;20: Colantonio A, Dawson DR, McLellan BA. Head injury in young adults, long-term outcome. Arch Phys Med Rehabil 1998;79: Morton MV, Wehman P. Psychosocial and emotional sequelae of individuals with traumatic brain injury: a literature review and recommendations. Brain Inj 1995;9: Lieberman SA, Hoffman A. Growth hormone deficiency in adults; characteristics and response to GH replacement. J Pediatr 1996; 128(5 Pt 2):S Carroll PV, Christ ER, Bengtsson BA, et al. Growth hormone deficiency in adulthood and the effects of growth hormone replacement: a review. Hormone Research Society Scientific Committee. J Clin Endocrinol Metab 1998;83: Springer J, Chollet A. A traumatic car crash. Lancet 2001;357: Kelly DF, Gaw IT, Cohan P, Berman N, Swerdloff R, Wang C. Hypopituitarism following traumatic brain injury and aneurismal subarachnoid hemorrhage: a preliminary report. J Neurosurg 2000;93: Lieberman SA, Oberoi AL, Gilkison CR, Masel BE, Urban RJ. Prevalence of neuroendocrine dysfunction in patients recovering from traumatic brain injury. J Clin Endocrinol Metab 2001;86: Consensus guidelines for the diagnosis and treatment of adults with growth hormone deficiency: summary statement of the Growth Hormone Research Society Workshop on Adult Growth Hormone Deficiency. J Clin Endocrinol Metab 1998;83: Benvenga S, Le Giudice F, Campenni A, Longo M, Trimarchi F. Post-traumatic selective hypogonadotropic hypogonadism. J Endocrinol Invest 1997;20: Levin HS, Gary HE, Eisenberg HM, et al. Neurobehavioral outcome 1 year after severe head injury. Experience of the Traumatic Coma Data Bank. J Neurosurg 1990;73: Lopez-Guzman A, Salvador J, Albero R, et al. Selective growth hormone deficiency of hypothalamic origin following severe head injury. Acta Paediatr 1992;81: McGauley GA, Cuneo RC, Salomon F, Sonksen PH. Psychological well-being before and after growth hormone treatment in adults with growth hormone deficiency. Horm Res 1990;33(Suppl 4): Ceballos R. Pituitary changes in head trauma (analysis of 102 consecutive cases of head injury). Ala J Med Sci 1966;3: Kornblum RN, Fisher RS. Pituitary lesions in craniocerebral injuries. Arch Pathol 1969;88: Crompton MR. Hypothalamic lesions following closed head injury. Brain 1971;94: Daniel PM, Treip CS. The pathology of the pituitary gland in head injury. Mod Trends Endocrinol 1961;2: Centers for Disease Control and Prevention. Traumatic brain injury in the United States: a report to Congress. Atlanta: 2003.