Hypothalamic–pituitary insufficiency may have diverse causes. The aim of this study was to determine the incidence of hypothalamic–pituitary insufficiency in patients with previous infectious diseases of the central nervous system (CNS) of different etiologies and mild-to-moderate clinical course.
Patient series. Basal and stimulated (insulin tolerance test) pituitary function testing was performed in 19 patients with previous neuroborreliosis, encephalitis, or meningitis following an interval of between 10 and 56 months (mean 26.1±13.1 months) after the acute event.
Four patients (21%; two males, two females) showed an isolated corticotropic insufficiency (peak cortisol <181.25 μg/l during the insulin tolerance test). Two patients (11%, males) showed borderline gonadotropic insufficiency (basal testosterone between 2.4 and 3.0 μg/l). No patient had somatotropic or thyrotropic insufficiency or evidence for diabetes insipidus; all had prolactin concentrations within the reference range.
Hypothalamic–pituitary dysfunction and especially isolated corticotropic insufficiency may develop in a relevant proportion of patients after infectious diseases of the CNS.
Hypothalamic–pituitary insufficiency may have diverse causes: the most common ones being pituitary adenoma, craniopharyngeoma, Sheehan's syndrome, lymphocytic hypophysitis, irradiation, or surgery (1). Despite numerous case reports, the incidence of hypothalamic–pituitary dysfunction following traumatic brain injury (TBI) and subarachnoid hemorrhage (SAH) has been underestimated for decades. In 1986, Edwards & Clark reported 6 own and 47 reviewed cases of post-traumatic hypopituitarism (2). Furthermore, in the year 2000, Benvenga et al. reviewed 314 cases, including 15 own patients (3). The endocrine dysfunctions reached from panhypopituitarism to various partial deficiencies and were sometimes transient. Hyperprolactinemia was reported in the majority of cases; diabetes insipidus occurred in up to 50% of the patients. A delay in diagnosis of months to years was common (2, 3). Since then, several studies evaluating the rate and risk factors of hypothalamic–pituitary dysfunction after TBI and SAH have been conducted (4, 5, 6, 7, 8, 9, 10).
Infectious diseases of the central nervous system (CNS) may also affect the hypothalamus and/or the pituitary, although this has not been reported very often and not yet been studied systematically. In most of the early case reports, the deficient hormone secretion has been caused by tuberculous meningitis (11, 12, 13). There are, however, reports to show that other infectious agents may also cause hypothalamic–pituitary dysfunction (14, 15, 16, 17, 18, 19). Therefore, the aim of the present investigation was to determine the incidence of hypothalamic–pituitary insufficiency in patients with previous infectious diseases of the CNS of different etiologies and mild-to-moderate clinical course. To address this issue, patients underwent basal and stimulated (insulin tolerance test) pituitary function testing following an interval of at least 6 months after the acute event.
Subjects and methods
Patients and study visits
The study was conducted according to the Declaration of Helsinki, and the study protocol was reviewed and approved by the local ethics committee. Medical records of patients admitted to the Neurological Department of the University Hospital of Marburg/Germany during the last 5 years for in-hospital treatment of infectious diseases of the CNS (meningitis, meningoencephalitis, encephalitis, neuroborreliosis) were screened. Patients of both sexes, aged 20–70 years, with an interval of at least 6 months between acute disease and medical record screening qualified for inclusion in this study. Exclusion criteria were any hormone intake, including oral contraceptives and topical/inhalative glucocorticoids; intake of neuroleptics; diabetes mellitus or other endocrine disorders; malignant disease; abuse of alcohol or drugs; or any other severe disease (especially coronary heart disease and a history of convulsion). Furthermore, severely disabled patients who were not able to give personal informed consent for participation in the study were excluded.
On the basis of 79 records screened, 5 had to be excluded, and another 8 patients could not be contacted any more because of unknown address. Of the remaining 66 patients, 38 refused to participate for personal reasons, and another 7 withdrew their consent after the first study visit before pituitary function testing.
During the first study visit, written informed consent was obtained from each subject. A physical examination, a routine laboratory testing, and an analysis of body composition using a body composition analyser (Akren-RJL BIA 101/S) were performed.
After an overnight fast, basal concentrations of growth hormone (GH), insulin-like growth factor-I (IGF-I), adrenocorticotropin (ACTH), cortisol, luteinizing hormone (LH), follicle-stimulating hormone (FSH), testosterone (in males), estradiol and progesterone (in females), thyrotrophin (TSH), free tri-iodothyronine (fT3), free thyroxine (fT4), and prolactin (PRL) were determined and an insulin tolerance test (ITT) was performed (Table 1). All subjects received 0.10–0.15 IU/kg regular i.v. insulin (Insuman Rapid; Sanofi-Aventis, Deutschland GmbH, Frankfurt, Germany). Additional insulin was given if blood glucose concentrations below 40 mg/dl were not reached within 15 min. Samples for the determination of glucose, GH, ACTH, and cortisol were drawn. In addition, bedside determinations of glucose levels were done throughout the test.
Insulin tolerance test: time of cortisol, growth hormone (GH), and adrenocorticotropin (ACTH) determinations.
|Regular insulin i.v. bolus (0.10–0.15 IU/kg body weight) given at 0 min|
|Hormone||−30 min||0 min||15 min||30 min||60 min||90 min||120 min|
No adverse events occurred during the ITT. Adequate hypoglycemia with clinical symptoms was achieved in all but two patients. One of these patients showed elevated basal glucose concentrations before the test because of formerly unknown diabetes mellitus. One patient showed normal basal glucose concentrations but did not reach adequate hypoglycemia. The ITT was repeated on a second occasion with an increased insulin dose, but again no adequate hypoglycemia was achieved. Diabetes mellitus was suspected and an oral glucose tolerance test proposed, but the patient denied further investigations. These two patients were excluded from the analysis.
Eventually, data of 19 patients (13 males, 6 females) were included in the final analysis. They were 22–65 years old (mean 38.7±11.7 years). Time interval since diagnosis ranged from 10 to 56 months (mean 26.1±13.1 months). Diagnosis was proved by lumbar puncture. In all patients, an elevation of protein concentrations and pleocytosis of cerebrospinal fluid indicated a disturbance in the blood–brain barrier. A direct examination of bacteria and antibody determinations (herpes simplex, varicella, Epstein–Barr, neuroborreliosis, tick-borne encephalitis) in the cerebrospinal fluid and serum were performed. Of all these patients, 4 suffered from neuroborreliosis, 2 from encephalitis (2 tick-borne encephalitis), and 13 from meningitis (1 herpes simplex, 1 varicella, 1 enterovirus, 10 of unknown origin). Imaging (cranial computed tomography or magnetic resonance imaging) was performed in patient numbers 2, 4–8, 10, 14, 15, and 19. The degree of disease was mild (no neurological deficits) or moderate (neurological deficits, no impairment of consciousness or other severe complication, no treatment at an intensive care unit) in all cases. All patients were treated with acyclovir and different antibiotics, leading to improvement of symptoms. Patient characteristics are given in Table 2.
Patient characteristics: age (years), sex (male, female), time from disease to testing (months), infectious disease, neurological deficits during acute CNS infection, and actual neurological and self-experienced deficits (corticotropic insufficient patients are marked in gray shades).
|Age||Sex||Time||Infectious disease||Neurological deficits||Actual symptoms|
|1||63||M||16||Neuroborreliosis||Vertigo||Fatigue, paraesthesia (legs)|
|2||37||M||49||Meningitis, unknown etiology||No||Cephalgia|
|3||42||M||18||Meningitis, unknown etiology||No||Fatigue|
|4||29||F||19||Meningitis, unknown etiology||Hemihypesthesia and hemiparesis||No|
|5||36||M||12||Herpes simplex meningitis||No||Lack of concentration|
|6||34||F||45||Neuroborreliosis||Cramps, generalized paresis||Myalgia, cramps, generalized paresis, cold intolerance, lack of concentration, reduced vision|
|10||36||M||30||Meningitis, unknown etiology||No||No|
|11||24||M||17||Tick-borne encephalitis||Reduced vision||No|
|12||65||M||31||Neuroborreliosis||Dysbasia||Fatigue, dysbasia, RLS, paraesthesia, sleep disturbance|
|14||50||F||22||Meningitis, unknown etiology||No||Fatigue|
|15||38||M||22||Meningitis, unknown etiology||No||Fatigue, forgetfulness|
|16||43||F||20||Meningitis, unknown etiology||No||Mood disturbance|
|17||38||M||28||Meningitis, unknown etiology||No||No|
|18||30||F||22||Meningitis, unknown etiology||No||Fatigue|
|19||48||M||10||Meningitis, unknown etiology||No||Cephalgia|
Statistical analysis was performed using SPSS 12.0 for Windows. Groups were compared by one-factorial ANOVA, and correlations established according to Pearson. P<0.05 was considered significant. Data are given as mean±s.d.
Assays and pituitary insufficiency
GH and IGF-I were determined using a solid-phase chemiluminescence immunometric assay (Immulite 2500; EURO/DPC, Llanberis, UK), and progesterone using a chemiluminescence assay (DXI 800; Beckmann Coulter, Krefeld, Germany). All other endocrine parameters were determined using the ADVIA Centaur System (Bayer HealthCare); cortisol, estradiol, testosterone, fT3, and fT4 using a competitive immunoassay; and LH, FSH, PRL and TSH using a sandwich immunoassay (direct chemiluminescence technology). Due to technical reasons, the method of ACTH determination had to be changed during the study. ACTH was determined using an immunoluminometric assay (LUMItest ACTH; Brahms, Hennigsdorf, Germany) in patient numbers 1, 3, 4, 6, 8, 11, 12, 16, and 18, and a solid-phase chemiluminescence immunometric assay (IMMULITE 2500; EURO/DPC) in patient numbers 2, 5, 7, 9, 10, 13, 14, 15, 17, and 19.
Corticotropic insufficiency was defined as peak cortisol concentration during the ITT <181.25 μg/l (500 nmol/l) (10). Severe somatotropic deficiency was defined as peak GH concentration during the ITT <3 ng/ml (0.14 pmol/l), and partial deficiency as peak GH concentration between 3 and 5 ng/ml (0.23 pmol/l) (20). Gonadotropic deficiency was defined as testosterone concentration <2.4 μg/l (8.4 nmol/l) without LH elevation (>9.2 U/l) in males, referring to the reference range given by the manufacturer of the assay, and the presence of menstrual disturbances in females. As a higher cut-off for males has also been proposed (21), a borderline gonadotropic deficiency was defined as testosterone concentration between 2.4 and 3.0 μg/l (10.5 nmol/l) in males. Thyrotropic deficiency was defined as fT3 concentration <3.1 pmol/l (2.0 pg/ml) and/or fT4 concentration <7.5 pmol/l (0.59 ng/dl) without TSH elevation (>5.6 mU/l). PRL concentrations were considered normal between 2.1 and 17.7 μg/l (93.2 and 785.9 pmol/l) in men and between 2.8 and 25.0 μg/l (124.3 and 1109.8 pmol/l) in women. Posterior pituitary insufficiency was considered in case of reported polydipsia and polyuria and electrolyte disturbances.
Four patients (21%; two males, two females) showed a peak of cortisol <181.25 μg/l during the ITT. In the patients with a corticotropic deficiency, peak cortisol ranged from 115 to 172 μg/l (mean 141.0±28.3 μg/l), and in the non-deficient patients from 183 to 240 μg/l (mean 210.5±19.7 μg/l), the difference being highly significant (P=0.000). The corticotropic insufficiencies were not combined with any further pituitary deficiency (Table 3).
Hormone concentrations: basal and stimulated cortisol (μg/l), growth hormone (GH; ng/ml), adrenocorticotropin (ACTH; pg/ml), basal thyrotrophin (TSH; mU/l), free tri-iodothyronine (fT3) and free thyroxine (fT4; pmol/l), luteinizing hormone (LH) and follicle-stimulating hormone (FSH; U/l), testosterone (μg/l), estradiol (ng/l), progesterone (μg/l), prolactin (PRL; μg/l), and insulin-like growth factor-I (IGF-I; ng/ml).
|Cortisol||Cortisol peak||GH||GH peak||ACTH||ACTH peak||TSH||fT3||fT4||LH||FSH||Testo.||Estr.||Prog.||PRL||IGF-I|
In the patients with corticotropic deficiency, basal cortisol ranged from 81 to 144 μg/l (mean 103.3±29.3 μg/l), basal ACTH from 7.5 to 14.1 pg/ml (mean 11.0±3.5 pg/ml), and peak ACTH during the ITT from 14.1 to 72.8 pg/ml (mean 35.3±25.9 pg/ml). In the non-deficient patients, basal cortisol ranged from 53 to 186 μg/l (mean 121.6±36.5 μg/l), basal ACTH from 4.6 to 40.5 pg/ml (mean 15.7±10.3 pg/ml), and peak ACTH during the ITT from 4.6 to 245.9 pg/ml (mean 70.0±74.7 pg/ml). The differences between both groups were not significant (P=0.369, 0.391, and 0.382). Peak cortisol concentrations were not correlated with basal cortisol, basal ACTH, or peak ACTH concentrations (P=0.315, 0.412, and 0.080).
In patients reporting self-experienced fatigue, peak cortisol concentrations were lower than those in patients not reporting fatigue (165.4±38.9 vs 213.6±18.4 μg/l; P=0.002).
No patient showed a peak GH <5.0 ng/ml during the ITT. Only one patient showed a response close to a partial somatotropic insufficiency (peak GH 5.6 μg/l). This particular patient showed a sufficient cortisol response during the ITT and normal IGF-I concentrations.
Basal GH ranged from 0.05 to 17.9 ng/ml (mean 2.63±4.55 ng/ml), peak GH from 5.6 to 36.9 ng/ml (mean 17.34±9.98 ng/ml), and basal IGF-I from 86.3 to 211 ng/ml (mean 150.6±34.9 ng/ml). Fifteen patients had IGF-I concentrations within the age-related reference range, three patients between −2 and −1 s.d., and one patient below −2 s.d. (22).
No patient showed fT3 or fT4 concentrations below the reference range. Basal fT3 ranged from 3.8 to 6.1 pmol/l (mean 4.7±0.52 pmol/l), basal fT4 from 9.3 to 13.0 pmol/l (mean 11.3±1.2 pmol/l), and basal TSH from 0.23 to 3.4 mU/l (mean 1.13±0.73 mU/l).
No male patient showed testosterone concentrations below 2.4 μg/l; all women had a regular menstrual cycle. Two male patients showed a borderline gonadotropic insufficiency with testosterone concentrations between 2.4 and 3.0 μg/l and low LH concentrations. One of these patients reported loss of erection and libido, and the other did not show any clinical evidence for a gonadotropic insufficiency.
Basal testosterone ranged from 2.7 to 5.9 μg/l (mean 4.0±1.1 μg/l), basal estradiol from 20.0 to 185.0 ng/l (mean 87.2±61.9 ng/l), basal progesterone from 0.55 to 6.20 μg/l (mean 2.1±2.1 μg/l), basal sex hormone binding globulin (SHBG) from 13.0 to 87.0 nmol/l (mean 37.1±19.5 nmol/l), basal LH from 1.7 to 46.0 U/l (mean 6.2±10.0 U/l), and basal FSH from 1.8 to 32.0 U/l (mean 6.9±6.8 U/l).
All patients showed PRL concentrations within the reference range. Basal PRL ranged from 2.4 to 16.7 μg/l (mean 7.1±2.9 μg/l).
Posterior pituitary insufficiency
No patient showed evidence for diabetes insipidus.
Body impedance analysis
Body mass index (BMI) ranged from 16.6 to 34.7 kg/m2 (mean 25.5±5.1 kg/m2), fat mass from 5.7 to 42.2 kg (mean 17.4±9.9 kg), fat-free mass from 34.0 to 84.0 kg (mean 63.3±12.8 kg), and body water from 24.9 to 61.5 kg (mean 46.3±9.4 kg).
BMI and body composition were not different between patients with a corticotropic insufficiency and non-deficient patients (23.8±5.3 vs 25.9±5.2 kg/m2, 13.7±8.1 vs 18.3±10.4 kg, 58.8±12.8 vs 64.5±13.0 kg and 43.1±9.3 vs 47.2±9.5 kg; P=0.473, 0.417, 0.445, and 0.448).
Infectious diseases of the CNS may cause hypothalamic and/or pituitary dysfunction. The aim of this study was to determine the incidence of hypothalamic–pituitary insufficiency in patients with previous infectious diseases of the CNS of different etiologies and mild-to-moderate clinical course.
In most previous case reports, isolated posterior pituitary insufficiency following infectious diseases of the CNS has been described in children: in a retrospective analysis, severe CNS infection was seen in 8 out of 73 children with central diabetes insipidus. The infectious agents were group B streptococcus, Haemophilus influenzae, Streptococcus pneumoniae, and unknown virus (23). Central diabetes insipidus developed in five infants with congenital cytomegalovirus infection (24) and in a 5-year-old boy with encephalitis caused by Coxsackie virus B1 after resuscitation (25). Only few cases of isolated posterior pituitary insufficiency in adults following CNS infections have been reported, mainly affecting immunocompromised patients. Infectious agents were herpes simplex (26), S. pneumoniae (27), and cryptococci (28) in patients with AIDS, and herpes simplex in one female with ectopic ACTH-dependent Cushing's syndrome due to metastatic carcinoid. Autopsy revealed involvement of the hypothalamus with a viral destruction of vasopressin-producing neurons in this particular patient (29). Finally, an otherwise healthy woman developed diabetes insipidus 3 weeks after acute herpes simplex encephalitis (30).
Basal anterior pituitary hormone levels were normal (26, 30) or were not determined (23, 25, 27, 28) in these patients, and stimulation tests have been performed in only few (24). Reports of anterior pituitary dysfunction following infectious diseases of the CNS are rare, and the incidence has not yet been studied systematically. In almost all reported cases, the anterior pituitary insufficiency was caused by viruses (14, 15, 16, 17, 18), and only one bacterial meningoencephalitis associated with diabetes insipidus and suspected corticotropic insufficiency has been reported. A CT of the head revealed a contrast-enhanced suprasellar lesion in this particular patient (19). The endocrine deficiencies reached from complete panhypopituitarism including diabetes insipidus (14, 15, 16, 18) to various partial defects (14, 15, 17). Data of these patients are given in Table 4.
Reported cases of anterior pituitary insufficiency after infectious diseases of the CNS: age (years), sex (male, female), time from disease to diagnosis (years), infectious agent, neurological and endocrine deficits.
|Age||Sex||Time||Infectious agent||Neurological deficits||Endocrine deficits||Reference|
|1||43||F||0||Unknown virus||Retrobulbar neuritis||Gonadotropic, thyrotropic and somatotropic insufficiency, diabetes insipidus||Hagg et al. (14)|
|2||53||M||0||Coxsackie B5||Paresis of the sixth nerve on both sides||Gonadotropic, thyrotropic and somatotropic insufficiency|
|3||23||M||3||Influenza A||No||Somatotropic and corticotropic insufficiency||Kupari et al. (15)|
|4||39||F||0||Unknown virus||Paresis of the right arm, face and third nerve, retrobulbar neuritis||Panhypopituitarism, diabetes insipidus|
|5||54||F||0||Herpes simplex||Convolutions and unresponsiveness||Corticotropic and somatotropic insufficiency, hypoprolactinemia|
|6||41||M||0||Unknown virus||Somnolence and disorientation||Panhypopituitarism, hyperpro-lactinemia, diabetes insipidus||Lichenstein et al. (18)|
|7||53||F||0||Unknown bacteria||Tremor, lethargy, loss of cognitive functions and visual hallucinations||Corticotropic insufficiency, diabetes insipidus||Jew et al. (19)|
|8||76||F||0||Herpes simplex||Akinetic parkinsonian syndrome||Corticotropic, thyrotropic and gonadotropic insufficiency, hyperprolactinemia, diabetes insipidus||Ickenstein et al. (16)|
|9||49||M||27||Herpes simplex||Expressive aphasia||Thyrotropic and gonadotropic insufficiency||Vesely et al. (17)|
In our study, we found no evidence for posterior pituitary deficiencies; however, no dynamic testing (i.e., water deprivation test) has been performed and partial posterior pituitary insufficiencies might have been missed. On the contrary, there were a substantial number of isolated corticotropic deficiencies. With these findings, we are in concordance with Kreitschmann-Andermahr et al. (10), who found a high incidence of isolated corticotropic deficiencies in patients with previous SAH. Contrary to their data, in our present study, other anterior pituitary deficits despite two borderline gonadotropic deficiencies were not seen, especially no somatotropic deficiencies and no differences in body composition.
It is reasonable to assume that the incidence and pattern of hormonal deficiencies after an acute infectious (meningo-) encephalitis may vary with the type of causative agent, the localization of brain lesion, as well as with the severity of disease. The hormonal deficiencies may be transient (15, 19) or permanent (15, 16, 17, 23, 30).
Endocrine deficiencies have been reported following infectious diseases of the CNS of different etiologies. Three of the patients with corticotropic insufficiency in our study had suffered from meningitis of unknown origin; in one patient meningitis caused by enterovirus had been diagnosed. It has been suggested that focal neurological symptoms from the basal regions of the brain might be suggestive of an increased risk of hypothalamic–pituitary damage (14, 15), but heterogeneous neurological deficits have been reported in affected patients. Few patients (15) and also all patients with corticotropic deficiency in our study had presented with symptoms of general illness (cephalgia, myalgia, fever), but no neurological deficits at all. In only one patient (15), imaging of the head had been performed during acute illness, revealing no abnormalities of the pituitary gland or hypothalamus. In earlier case reports, endocrine investigations and elevated basal PRL levels (probably due to the loss of tonic inhibition of the pituitary lactotrophs by the hypothalamus) were suggestive of a hypothalamic rather than a pituitary lesion (14, 16, 18). On the contrary, in other patients, the pituitary gland seemed directly affected (15, 17). In our investigation, PRL concentrations were within the normal ranges in all patients. However, the presence of an isolated corticotropic deficiency has been taken as a hint for a hypothalamic rather than a pituitary damage (10). The site of the endocrine dysfunction and the reason for the predominance of isolated corticotropic insufficiency remain speculative.
Especially, self-reported persisting fatigue following an infectious CNS disease might be suggestive of the presence of a hypothalamic–pituitary damage. Of the seven patients reporting fatigue, four proved to have an isolated corticotropic insufficiency in our study. A hydrocortisone substitution therapy was started. One patient (no. 18) experienced massive subjective improvement of well-being and vitality. Two patients (numbers 3 and 15) seemed not to benefit from the hormone substitution. Both had a moderately diminished peak cortisol during the ITT. Another patient (no. 13) declined a substitution therapy. Although these early results do not allow any final conclusions, they point toward a possible use and necessity of hormone substitution therapy in some patients following infectious diseases of the CNS.
Finally, it has to be mentioned that our data might be biased for the following two reasons: during recruitment many patients denied to take part in our study as they felt completely healthy and favoring the inclusion of patients with self-experienced deficits might lead to an overestimation of the incidence of hypothalamic–pituitary insufficiencies. On the other hand, the acute disease of most patients in the published case reports were more severe than that in our patients, resulting in neurological deficits and even death. Only slightly or moderately ill patients with a good neurological outcome were included in our study, and the incidence of pituitary insufficiencies might be higher in patients with severe persistant neurological deficits.
In summary, it may be that hypothalamic–pituitary dysfunction and especially isolated corticotropic insufficiency may develop in a relevant proportion of patients after mild and moderate infectious diseases of the CNS, and that the clinical picture of some patients with endocrine insufficiency might be misinterpreted as an ordinary post-encephalitic syndrome. Furthermore, prospective studies investigating patients during and after infectious diseases of the CNS and the effects of hormone replacement therapy are warranted.
The authors thank Novo Nordisk, Mainz, Germany for providing the grants, which made this investigation possible.
AimarettiGAmbrosioMRDi SommaCFuscoACannavoSGasperiMScaroniCDe MarinisLBenvengaSdegli UbertiECLombardiGManteroFMartinoEGiordanoGGhigoE. Traumatic brain injury and subarachnoidal haemorrhage are conditions at high risk for hypopituitarism: screening study at 3 months after the brain injury. Clinical Endocrinology2004;61:320–326.
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. Journal of Clinical Endocrinology and Metabolism 1998 83 379–381.