JPED Journal

Introduction

Hypopituitarism is defined as a “diminution or cessation of secretion of one or more hormones from the anterior pituitary gland” [1]. Hypopituitarism can be congenital or acquired [2]. Causes of acquired hypopituitarism are diverse and include tumors, autoimmune disease, infiltrative disease, chemotherapy, radiation exposure, trauma, and infection [2]. In childhood, isolated and idiopathic growth hormone deficiency is the commonest form of pituitary hormone deficiency [2,3]. 

Early diagnosis of hypopituitarism in children is crucial. The pituitary gland produces hormones involved in growth, metabolism, reproduction, and homeostasis [2]. Symptomatology of hypopituitarism depends on which hormones are insufficiently secreted and the severity of deficiency [2]. Hypopituitarism may lead to urgent conditions such as hypoglycemia or adrenal crisis. Disorders in cognitive development, growth failure, excessive weight gain and delayed puberty are examples of long-term morbidity. Laboratory results should be carefully interpretated as end-organ hormone deficiencies may hide central depletion [2]. Magnetic resonance imaging is indicated when hypopituitarism is suspected [2].

Infectious causes of hypopituitarism in children are uncommon and have principally been reported in literature as case series and case reports. Infectious meningitis, encephalitis, and hypophysitis can lead to pituitary hormone deficiency [3]. Mycobacterium tuberculosis is a known causative microorganism [3]. Pituitary abscess is rare but potentially life-threatening [4]. Cases of hypopituitarism due to fungal infections of the sellar region are sporadic [4]. The occurrence of infectious diseases as a cause of pediatric hypopituitarism has not been systematically reviewed yet.

The aim of this systematic review is to give a quantitative overview of the infectious causes of hypopituitarism in children. We will also systematically discuss clinical signs and symptoms as well as the extent of endocrinological recovery after infection treatment. Informing clinicians about infection as a cause of hypopituitarism is important for adequate diagnosis. Putting into perspective the infectious causes of hypopituitarism may outline which infections should be included in the differential diagnosis when hypopituitarism is diagnosed in a child.

Materials and Methods

We searched for literature in Embase, Medline (PubMed) and Web of Science Core Collection whereby we used both MeSH terms and free text. The aim of the search was to identify all reported pediatric cases of hypopituitarism caused by infection. The complete search strategy can be found in Supplementary Table 1. Through database searching we found 6241 articles which after deduplication resulted in 3644 articles that were subsequently screened by title and abstract (Figure 1). Another 3 articles were identified through snowballing. A PRISMA flow diagram is shown in Figure1.

We applied the following exclusion criteria: congenital hypopituitarism or infection, iatrogenic causes of infection or hypopituitarism developed post-operatively, conference abstracts and proceedings, reviews without additional case report or case series, articles of which the full text was either not available or only available in languages different from English or French, articles without sufficient methodology or case description, and irrelevant studies. To be included in this systematic review, the studies had to discuss hypopituitarism due to infection in patients aged 0 to less than 18 years old. In these studies, pediatric cases had to be reported in absolute numbers and separately from adult cases to make data extraction possible. Similarly, infectious causes had to be reported separately from other causes of hypopituitarism. Articles of which the full text was not available were excluded if title and abstract did not meet the inclusion criteria. However, where title and abstract met the inclusion criteria, full texts were requested from KU Leuven Library and assessed if available. Based on the aforementioned criteria, 3414 records were excluded. The full texts of the 230 remaining studies were then assessed for eligibility. Another 190 studies were excluded, leaving 43 primary studies for inclusion in this systematic review.

We critically evaluated each study on internal validity with specific attention given to the used hormonal tests and interpretation of their results. Case series of which we did not only extract clinical case information but also used the author’s data-analysis were evaluated by using the quality assessment tools developed by NHLBI [5]. We collected following data: patient characteristics, specific pituitary deficiencies and symptoms and clinical signs at the time of hormonal evaluation, underlying infectious diseases, hormonal replacement therapies (HRT) and extent of recovery of pituitary function during follow-up. These data were listed in result-tables and discussed. 

Figure 1: PRISMA flow diagram of the literature search. Figure adapted fromPage et al. [6].

Results

In this systematic review, we present 43 cases of children with various infectious diseases that caused hypopituitarism. A summarizing overview is given in Table 1.

Tuberculosis of the central nervous system

We identified tuberculosis (TBC) of the central nervous system as the most common cause of acquired hypopituitarism due to infection. We included a total of 17 cases of TBC followed by hypopituitarism during childhood [7-22].These cases are summarized in Table 1 and Supplementary Table 2, with description of specific pituitary deficiencies and symptoms and clinical signs at the time of hormonal evaluation. We also describe HRT and extent of recovery of pituitary function during follow-up.

TBC of the central nervous system principally occurred in the form of isolated tuberculous meningitis (TBM) (8 cases) [714]. Tubercular lesions additionally consisted of sellar and/ or suprasellar tuberculomas (7 cases) [15-20]. We also report a case of tubercular abscess, located in the suprasellar region [21]. One diagnosis of TBC had not formally been established [22]. In this case, antitubercular treatment of 12 months was initiated in a context of a positive intradermal tuberculin test, abnormal MRI-imaging of the hypophysis, no evidence of other systemic granulomatous disease, and negative tumor markers in blood and cerebrospinal fluid. A pituitary biopsy was not judged necessary because of the benign course of the disease under antitubercular treatment.

Assessment of pituitary activity with hormonal tests led to the identification of various hormone depletions (Table 1 and Supplementary Table 2). Anterior pituitary dysfunction was associated with diabetes insipidus in most patients (10 cases) [8-10,12,13,15,16,19,21,22]. One child suffered from panhypopituitarism [21]. In 2 studies, authors suspected hypopituitarism to be secondary to hypothalamic dysfunction [12,14].In 3 other studies, hypothalamic dysfunction was suspected to be the cause of the child’s weight excess [7,16,20]. Arguments that supported the hypothesis of hypothalamic insufficiency were the suprasellar location of tubercular active lesions or sequelae and/or the results of hormonal tests using pituitary releasing hormones. For instance, in one case report, a significant increase in levels of adrenocorticotropic hormone (ACTH) in response to exogenous corticotropin releasing hormone suggested a deficiency of the latter hormone [14].

In cases of isolated TBM, without coexistent tuberculoma or abscess of the hypophysis, diagnosis of hypopituitarism was made months to several years after treatment of TBC. However, in all children who had a sellar and/or suprasellar tuberculoma, hormonal deficits were present at the time of diagnosis of tuberculoma. Four children with tuberculoma presented with non-endocrinological features such as headache, vomiting, and visual disturbances (decreased visual acuity and/or bitemporal hemianopsia) [16,18,19]. One of these children had fever [19]. It is important to notice that two patients had acute manifestations of central hypoadrenalism: a 3-year-old boy who developed hypoglycemia and a 15-year-old boy who suffered from shock secondary to mild diarrhea [11,12]. They both had a history of TBM.

In many articles, authors did not report if HRT was initiated, which hindered data analysis [8,9,11,14,17,18]. Moreover, 7 studies did not report any patient follow-up [9-11,13,14,17,19]. In other instances, follow-up often revealed persistent requirement of HRT (4 cases) [7,12,15,22]. In one boy, persistent requirement of HRT at 3 years of follow-up was associated with the discovery of 2 suprasellar tuberculomas [15]. Three children died and for one of them this was due to severe hypopituitarism with hypothermia and bradycardia, which are consequences of profound hypothyroidism [16]. In another child, death was suspected to be secondary to cessation of corticosteroids [21]. The third cause of death was not specified [8]. Finally, in no more than two studies, complete recovery of pituitary function was obtained [18,20].

Pituitary abscess

We identified 12 cases of pituitary abscess associated with hypopituitarism in childhood [23-33]. These cases are summarized in Table 1 and Supplementary Table 3.

At initial hormonal evaluation, anterior pituitary axes were variably deficient (Table 1 and Supplementary Table 3). Concomitant diabetes insipidus was diagnosed in 9 patients [23-27,29,30,33]. Four patients presented with panhypopituitarism [24,26,30]. A total of 3 abscesses developed in a pre-existing pituitary lesion [24,31,32]. These pituitary disorders were Rathke’s cleft cysts in two children and adenoma in one child.

Most studies did not identify any risk factors for development of a pituitary abscess. Sinusitis was found to be a possible source of infection in 2 cases [25,33]. A 15-year-old girl was hospitalized for pneumonia one year prior to diagnosis of pituitary abscess [29]. Her pneumonia may have led to infection of the pituitary gland via hematogenous dissemination.

Bacteriological investigation was positive in 4 cases [25,29,31,32].

Following microorganisms were detected: Staphylococcus aureus (2 cases), Staphylococcus epidermidis (1 case), Streptococcus viridans (1 case), Streptococcus spp. (not further specified) (1 case), and Corynebacterium argentoratense (1 case). In 5 instances, cultures remained negative [24,27,28,30,33]. Only 2 of these patients received antibiotic therapy before samples were taken for culture [28,30]. For one of these 5 patients, the authors did not mention any antibiotic therapy at all besides surgery [24]. Individual bacteriological data were not available in 2 articles [23,26].

Duration of symptoms prior to presentation was variable. Symptoms were most often present for several weeks or months, with a maximum duration of 2 years [23,25-30,32]. Pituitary dysfunction was symptomatic in the majority of children (10 cases) [23-27,29,30,32,33]. Headache was the commonest symptom not related to endocrine disturbances and was present in 9 children [23,25,26-29,31-33]. Remarkable is that only one patient had fever [33]. Biological markers of inflammation, such as leukocytosis with an elevated count of neutrophils, elevated C-reactive protein and/or erythrocyte sedimentation rate were positive in 4 cases [23,27-29].

Three children had acute symptoms, such as altered consciousness, emesis, neck pain, and visual disturbances [28,29,33], suggestive for additional meningitis, encephalitis, and/or intracranial hypertension. An 11-year-old boy had bitemporal hemianopsia, decreased visual acuity, and sudden-onset headache caused by pituitary adenoma complicated by apoplexy and abscess [31].

HRT was started in all patients, except for the child presenting with isolated hypoprolactinemia [31]. In 5 patients, all hormone deficiencies were substituted [23,27,28,32,33]. Information regarding HRT was incomplete in 2 studies [24,30]. Hypocortisolemia and diabetes insipidus were always corrected by HRT [23-30,32,33].

During follow-up, recovery of all affected pituitary axes was observed in one child [25]. Most authors reported persistent requirement of HRT after treatment of pituitary abscess, sometimes with partial recovery of pituitary function [24,26,28,29,32,33]. The follow-up period varied considerably between articles and often lasted several months, up to a maximum duration of 17 years [24-26,28,29,31-33]. Three studies did not report follow-up [23,27,30].

Other case studies

The remaining 14 cases are detailed in Table 1 and Supplementary Table 4 and involved children with active or a history of infection different from TBC and pituitary abscess [34-46]. Group B streptococcal meningitis was the most frequent infection in this group, affecting 4 children [35,36,38,46]. Hypopituitarism was suggested to be a sequela of infection in 4 patients, as it was diagnosed at least 1 year after the infectious disease: poliomyelitis (1 case), group B streptococcal meningitis (2 cases), and viral meningitis (1 case) [34,36,38,39]. In the other 10 patients, hypopituitarism occurred during active infection or within the months after [35,37,40-46]. Various microorganisms were involved (see Table 1 and Supplementary Table 4). Among these patients, 7 exhibited features of severity such as hypotension, tachy- or bradycardia, hypotonia, altered consciousness, and/or seizures [35,40,43-46]. The pituitary-thyroid axis was found to be deficient in all 7 children. Well-documented thyroid-stimulating hormone (TSH) deficiency was present in 6 of them [40,43-46]. ACTH deficiency and/or low cortisol were also very frequent (6 cases) [35,40,43,44,46]. Six critical children suffered from meningitis, meningoencephalitis [35,40,45,46], or infectionassociated encephalopathy [44].

Furthermore, we identified 2 studies that investigated pituitary function in children (19 and 12 patients below the age of 18 years) who had a history of infectious meningitis of different etiologies [47,48]. The time between the acute event and the hormonal evaluation varied from 3 months to 4.3 years. No pituitary hormone deficiencies were detected. At last, we included a study that retrospectively examined growth in 22 boys with HIVinfection (without overt acquired immunodeficiency syndrome) secondary to repeated transfusions with clotting factor concentrate for treatment of hemophilia [49]. Three boys had a growth velocity below <2 standard deviations for age, for more than 2 years after the onset of HIV-infection. However, assessment of pituitary activity did not reveal abnormalities of growth hormone, cortisol, or thyroxine levels.

Table 1: Summarizing table of the infectious causes of hypopituitarism in children; HRT: hormonal replacement therapy; FU: followup; TBC: tuberculosis; TBM: tuberculous meningitis.