Advances in understanding hypopituitarism

The understanding of hypopituitarism has increased over the last three years. This review provides an overview of the most important recent findings. Most of the recent research in hypopituitarism has focused on genetics. New diagnostic techniques like next-generation sequencing have led to the description of different genetic mutations causative for congenital dysfunction of the pituitary gland while new molecular mechanisms underlying pituitary ontogenesis have also been described. Furthermore, hypopituitarism may occur because of an impairment of the distinctive vascularization of the pituitary gland, especially by disruption of the long vessel connection between the hypothalamus and the pituitary. Controversial findings have been published on post-traumatic hypopituitarism. Moreover, autoimmunity has been discussed in recent years as a possible reason for hypopituitarism. With the use of new drugs such as ipilimumab, hypopituitarism as a side effect of pharmaceuticals has come into focus. Besides new findings on the pathomechanism of hypopituitarism, there are new diagnostic tools in development, such as new growth hormone stimulants that are currently being tested in clinical trials. Moreover, cortisol measurement in scalp hair is a promising tool for monitoring cortisol levels over time.


Introduction
Hypopituitarism is defined as a diminished function of the pituitary gland. First described in 1914 by Simmonds, it is also known as Simmonds' disease 1 . There are two main reasons for the hypofunction of the pituitary gland: it can result from pituitary dysfunction per se or from hypothalamic damage. In both cases, the production of pituitary hormones is diminished. When a single pituitary hormone is affected, this is called isolated pituitary deficiency. When two or more pituitary hormones are affected, this is referred to as multiple pituitary hormone deficiency. Panhypopituitarism is a state of reduction of all pituitary hormones.
Hypopituitarism may result from different causes and can be congenital or acquired. The most common causes are listed in Table 1. Depending on the specific anatomic lesion, patients present with different endocrine abnormalities. These either can be detected during newborn screenings such as congenital hypothyroidism or may present later in life (for example, with growth retardation or absence of puberty). As hypopituitarism may be associated with malformations of the central nervous system, such as malformation of the optic nerve, patients may present with different grades of sensory impairments. Acquired hypopituitarism can occur in all age groups and needs careful consideration of treatment indication depending on the nature and degree of the affected hormone axis.

Genetic causes of congenital hypopituitarism
Despite recent advances, genetic causes are still not identified in the majority of patients with congenital hypopituitarism. Affected patients can be subdivided into two groups: one with concomitant association with extra-pituitary abnormalities (syndromic diseases) and the other presenting with hypopituitarism alone. Congenital hypopituitarism may be isolated with a loss of function of a single pituitary axis or presenting with dysfunction of two or more pituitary axes. However, initial diagnosis of isolated growth hormone deficiency (GHD) may frequently develop into multiple pituitary hormone deficiencies as described in a long-term observational study with a mean follow-up of more than 15 years 2 . Analysis of the KIMS database (Pfizer International Metabolic Database) has shown that hypopituitarism in adults seems to be a dynamic condition where new hormone deficiencies may occur even years after initial diagnosis of isolated pituitary hormone deficiency 3 .
The diagnosis of congenital causes for hypopituitarism can be challenging. Sequencing technologies may provide the underlying genetic cause. Sanger sequencing can be used if a single gene mutation is suspected because of the patient's phenotype. More often, a multi-panel test seems to be more appropriate if no obvious clinical phenotype or magnetic resonance imaging finding is present. More recently, developed techniques such as next-generation sequencing and whole exome sequencing allow for the analysis of several relevant genes in parallel or coding regions of all genes, respectively. It has to be kept in mind that only 10% of cases with hypopituitarism can be explained by genetic alterations known so far 4 . Epigenetic alterations, mosaicism, and oligonucleotide repeat expansions as well as alterations outside the coding regions cannot be detected by the above-mentioned sequencing techniques. However, owing to these improved diagnostic tools, there have been several causative genetic alterations identified in the last few decades, ranging from chromosomal abnormalities to single gene mutations 4,5 . The most recent publications will be summarized here briefly.
In general, alterations in transcription factors implicated in pituitary organogenesis can be pituitary-specific, such as mutations in POU1F1, or can affect different tissues, as is the case in OTX2 mutations where pituitary deficiency is usually associated with severe eye abnormalities. In a recently published case report, an OTX2 mutation was identified in a neonate presenting with adrenocorticotropic hormone deficiency, bilateral microphthalmia, and agenesis of the left internal carotid artery, expanding the spectrum of pathophysiology associated with OTX2 mutations 6 .
In patients with hypogonadotropic hypogonadism (in combination with anosmia, known as Kallmann syndrome), plenty of genetic alterations affecting appropriate GnRH action have been described. Mutation in the genes FEZF1 7 , SOX10 8 , and FGF17, IL17RD, SPRY4, DUSP6, and FLRT3 9 have been identified. Whether these single mutations alone are sufficient to cause disease remains unclear. For clinicians, a panel-testing approach in patients with hypogonadotropic hypogonadism/Kallmann syndrome is recommended 5 . In patients with isolated GHD, genetic alterations hindering the release of functioning growth hormone have been identified. Mutations in the pituitary-specific transcription factor POU1F1 usually lead to an altered somatotropic, lactotropic, and thyreotropic axis. Recently, a heterozygous missense mutation in POU1F1 was identified as a causal trigger in a family with isolated GHD 10 . Moreover, a homozygous point mutation in the promotor region of the human growth hormone gene (GH1) was firstly described as causative for isolated GHD in siblings with normal pituitary imaging 11 . These two studies demonstrate the possibility of familial heredity that is either autosomal dominant or autosomal recessive.
The Italian Study Group on Genetics of multiple pituitary hormone deficiencies recently published a systematic review investigating the mutation frequency in the most common altered genes, PROP1, POU1F1, HESX1, LHX3, and LHX4 12 . The calculated prevalence of defects in these genes was quite low in the Italian study cohort (2.9% in sporadic and 12.5% in familial cases), and a worldwide prevalence of 12.4% has been estimated (from 11.2% in sporadic up to 63% in familial cases) identifying PROP1 as the most frequent mutated gene. Genetic alterations of PROP1 are usually associated with non-syndromic pituitary deficiencies. Nevertheless, dysfunction of this transcription factor can cause a variety of different phenotypes probably due to residual activity of altered protein synthesis 13 . Mutations in the non-pituitary-specific transcription factor HESX1 are commonly associated with syndromic clinical features and have been linked to septo-optic dysplasia. However, dysfunction of HESX1 may also present with a wide range of phenotypes 14 , making targeted genetic diagnostics in patients with multiple pituitary hormone deficiencies even more challenging.
It must also be kept in mind that the detection of a mutant gene does not mean dysfunction of the encoded product per se. Highperforming genetic sequence analysis enables the detection of a variety of genetic mutations and variants of unknown significance, as demonstrated in an analysis of the transcription factor GLI2 15 . GLI2 is a large polymorphic gene, and a relatively high frequency of GLI2 mutations and variants were identified in patients with congenital GHD without further brain abnormalities. In mice, a link between GLI2 mutations and prenatal ethanol exposure has been demonstrated, thereby implicating environmental factors in the etiology of GLI2-associated hypopituitarism 15 . Furthermore, variability in penetrance and not-yet-identified gene-gene interactions underline the need for functional studies in patients with hypopituitarism.
In patients with congenital holoprosencephaly, a midline defect with variable phenotypes, several genetic alterations in the sonic hedgehog signaling pathway have been identified. Pituitary stalk interruption syndrome and isolated pituitary hypoplasia have recently been proposed as mild phenotypic variants of holoprosencephaly, as several genetic mutations in holoprosencephaly-associated genes have been proven in both entities. Affected genes comprise TGIF1 and SHH, both relevant in sonic hedgehog signaling 16 . In familial clustering of the disease, a homozygous mutation of GPR161, a gene coding for a G-protein-coupled receptor, was recently identified 17 . Moreover, detection of a novel mutation in the CDON gene relevant in sonic hedgehog signaling 18 reveals the broad spectrum of implicating variants in pituitary dysgenesis and dysfunction.

Hypopituitarism due to impaired hypophyseal blood supply
The anterior pituitary is one of the best-vascularized mammalian organs. Its high vessel density persists throughout the lifespan as the proper function of the pituitary is dependent on the rapid transportation of hypothalamic or peripheral factors from the bloodstream to the pituitary cells and the fast release of hypophyseal hormones into the blood 19,20 . The intrapituitary vessel density is reduced in -deficient mice, a model of pituitary hormone deficiency. However, as in these mice the development of hormone-producing cells is also impaired, it is not clear whether disturbed hormone transportation due to the reduced vessel density contributes to hypopituitarism in these mice 21 .
In general, blood supply-dependent pituitary insufficiencies are rare 22 . Mechanical ruptures of the long portal blood vessels, which connect the pituitary with the hypothalamus, have been suggested to contribute to pituitary insufficiency in patients with traumatic brain injury (see the next section). In patients with congenital pituitary stalk interruption syndrome, it is still not known to which extent an impaired portal blood vessel system contributes to the hypopituitarism seen in these patients 23,24 . Pituitary insufficiencies due to impaired hypophyseal blood supply are in most cases caused by compression of the long portal blood vessels or by pituitary apoplexy. Very often, these events are associated with or caused by pituitary tumors or local hemorrhages 22,25,26 . Depending on the degree of blood flow impairment, a partial pituitary insufficiency can be observed in affected patients. This partial insufficiency can be reversible after normalizing the hypophyseal blood flow (for example, by removing the compressing structure) 25,27 . Severe apoplexy leading to intrapituitary hypoxia and necrosis of hormoneproducing cells is a life-threatening condition and needs rapid and proper treatment 22,26 .
Reduced blood supply of the pituitary gland is attributed to Sheehan's syndrome, a common cause of hypopituitarism in postpartum women. The enlargement of the pituitary gland during pregnancy in combination with massive post-partum hemorrhage leads to a significant reduction of arterial blood flow and causes pituitary necrosis 28 . The prevalence of Sheehan's syndrome might have decreased in recent years in some European countries and this is probably due to better obstetric care 28 . In all other parts of the world, including the former Eastern European countries, the Middle East, the Mediterranean, and Africa, Sheehan's syndrome is still one of the most common causes of hypopituitarism. Signs and symptoms are often unspecific, and this may cause a delay in or impair correct diagnosis. A diagnostic delay of about 20 years was described and recently confirmed in Turkey 29,30 . Recently, an increased expression of the coagulation factor V gene was reported in patients with Sheehan's syndrome 31 . However, gene polymorphisms in coagulation factors II and V and in tumor necrosis factor-alpha seemed not to be attributed to the pathogenesis of Sheehan's syndrome 32 . In general, the role of coagulation factors in Sheehan's syndrome has to be further elucidated.

PROP1
Acquired hypopituitarism due to brain trauma Acquired hypopituitarism can be observed in several neurological diseases. It was first described in a patient after traumatic brain injury leading to post-traumatic hypopituitarism 33 . There are two distinct pathomechanisms causing post-traumatic hypopituitarism 34 . The pituitary gland can be affected directly at the moment of the trauma (for example, by blast injuries, skull fractures with involvement of the sella turcica, or pituitary stalk disruptions). Not only do direct damages of the pituitary gland lead to hypopituitarism but also secondary effects of the trauma such as compression of the pituitary gland by brain swelling, impairment of the vascularization through swelling or vasospasms, and consecutive cell death of pituitary cells result in post-traumatic hypopituitarism.
Post-traumatic hypopituitarism may have causes other than just a single event. Repetitive concussions in contact sports such as football 35 or boxing 36 may also result in post-traumatic hypopituitarism.
As described above, the perfusion of the pituitary gland is crucial for the survival of pituitary cells. From an anatomical perspective, acidophil growth hormone-producing cells are located mainly at the periphery of the pituitary gland, thus making them sensitive to perfusion deficits and hypoxia. This fact contributes to the finding that somatotropic insufficiency is the most frequent hormonal deficit after brain injury 37-39 , especially in chronic patients 40,41 .
Recently, this finding has been controversially discussed. In line with recent guidelines, GHD has to be defined by stimulation tests such as the insulin tolerance test or the growth hormone-releasing hormone (GHRH)-arginine test 42 . In a Danish National Study, only a low incidence of GHD of less than 1% after brain injury was reported when patients received a confirmatory stimulation test 43 . Hence, the high prevalence and especially the relevance of posttraumatic GHD have been questioned 44 .
This controversial discussion has extended to patients with subarachnoid hemorrhage (SAH). The first reports showed high prevalence rates of hypopituitarism after SAH of 47.5% 38,45 , especially of GHD with a prevalence of 20 to 25% in chronic patients 46 . Recent findings suggest that these high prevalence rates might be overestimated 47,48 . The inclusion of different patient cohorts with distinct characteristics such as different severity grade of TBI/SAH and the use of different diagnostic tests might contribute to these controversial findings 49 . This outlines that test stringency, especially for GHD, is very important in these patient cohorts. Additional diagnostic tools would be helpful. Recently, circulating microR-NAs were identified in patients with post-traumatic hypopituitarism, suggesting that the microRNAs miR-126-3p and miR-3610 may act as potential biomarkers for hypopituitarism after TBI 50 .
The diagnosis of hypopituitarism may be difficult in patients after brain injury, since signs and symptoms of pituitary dysfunction may overlap with chronic sequelae of TBI and SAH. However, it was shown that pituitary dysfunction had no effect on the commonly reported fatigue following SAH 51 . Moreover, endocrine disturbances such as hypogonadism and hypothyroidism may occur as physiological stress responses to critical illness in patients after very severe brain injury, even long after the trauma 52 . Hence, pathological laboratory findings do not automatically implicate the presence of post-traumatic hypopituitarism 53 .
But not only TBI and SAH may lead to hypopituitarism. Pituitary dysfunction also occurs after stroke with an estimated prevalence rate of about 19%. Thus, hypopituitarism seems to be less frequent in stroke patients than in patients after TBI or SAH 54 . Despite the impact of stroke on morbidity and mortality, research on hypopituitarism after stroke has been very limited in recent years 55 . Moreover, hypopituitarism may rarely occur after infectious diseases like encephalitis or meningitis 56,57 and should be considered in these patients presenting with an unusual course.
In addition, cranial radiation is the most common cause of iatrogenic hypopituitarism 58 . The radiation of brain or neck tumors may affect the pituitary gland, thus leading to hypopituitarism 59 . In children with acute lymphoblastic leukemia, GHD seems to be the most common hormonal deficit after cranial radiation 60 . In adults, an increasing prevalence of hypopituitarism from 12.5% after 2 years up to 90% 15 years after radiation has been reported 61 . As hypopituitarism due to cranial irradiation will increase as a consequence of more cancer patients receiving this treatment, establishing well-functioning collaboration between endocrinologists and oncologists is essential for adequate followup for these patients.
The role of autoantibodies in the pathophysiology of secondary hypopituitarism has been discussed in more detail in recent years. Damage of the pituitary integrity might expose pituitary antigens, leading to autoimmune-mediated damage of the pituitary gland in the presence of pituitary antibodies. However, the impact of pituitary antibodies on the pathophysiology of secondary hypopituitarism after brain injury remains unclear. There are several findings indicating that secondary hypopituitarism is associated with the presence of autoantibodies. These findings seem to be stable even years after the event 62 . However, it has to be kept in mind that there is no standardization or even proper validation of autoantibody assessment and such findings have not been confirmed by other research groups so far 63 .

Auto-antibodies of the pituitary gland and the hypothalamus
Autoimmunity has recently been discussed as a further pathomechanism of hypopituitarism. Two different types of antibodies play a role in hypopituitarism: antipituitary antibodies and antihypothalamus antibodies.
Although antipituitary antibodies were first described in 1965 in a patient with Sheehan's syndrome 64 , it is still not clear what their exact target is 63 . Alpha-enolase, gamma-enolase, the pituitary gland-specific factors 1 and 2, and corticotroph-specific transcription factor (TPIT) have been discussed as possible targets of antipituitary antibodies 65 . The measurement of antipituitary antibodies is complicated by different methodological problems, and the clinical use of antipituitary antibodies and antihypothalamus antibodies is very limited so far because of the lack of standardized and validated methods. In autoimmune hypophysitis, also known as lymphocytic hypophysitis, the measurement of antipituitary antibodies can be considered, but sensitivity and specificity are low 65 .
Ricciuti et al. recently showed that antipituitary antibodies are more common in patients with pituitary diseases (24%) when compared with healthy controls (5%), but a discrimination of pituitary diseases has not been possible 66 .
De Bellis et al., in a longitudinal study, recently demonstrated that children with isolated GHD and high antipituitary antibody titers were not likely to show remission 67 . Furthermore, antipituitary antibody-positive patients were at risk for delayed puberty due to hypogonadotropic hypogonadism. The presence of antipituitary antibodies may precede hypogonadism 68 .
Less is known about antihypothalamus antibodies. These antibodies seem to target corticotropin-releasing hormone (CRH)secreting cells in the hypothalamus 69 . The presence of antihypothalamus antibodies was described in amateur boxers with hypopituitarism 36 as well as in patients with idiopathic hypopituitarism 70 . In general, the literature on antihypothalamus antibodies is very scarce. A standardization and proper validation of methods are necessary for a better understanding of these results. This would also enable wider use of these autoantibody assessments.

Drug-induced hypopituitarism
The discussion of iatrogenic hypopituitarism has focused mainly on induction of autoimmune hypophysitis by immune checkpoint-inhibiting medication used in cancer treatment. A recently published review provides a good overview of the prevalence and pathophysiology of ipilimumab-induced autoimmune hypophysitis associated with pituitary dysfunction 71 . In brief, the incidence of hypophysitis is less than 10% with endocrine dysfunction occurring approximately 9 weeks after the initiation of therapy on average. Drug cessation and high-dose glucocorticoid administration have been proposed to manage these side effects. Continuation of the therapy may lead to hypopituitarism and can necessitate pituitary hormone substitution.

Recent developments in the diagnosis of hypopituitarism
In the case of suspected hypopituitarism, testing for the function of a single pituitary axis should be performed stepwise and according to the clinical presentation. Guidelines (for example, clinical practice guidelines of the Endocrine Society 72 ) provide advice for a reasonable screening and use of appropriate tests. However, there are some new developments in the diagnosis of isolated pituitary deficiencies, which will be highlighted here.
Analysis of human scalp hair has gained more attention in recent years. Originally used for consumption behavior in a forensic context, analysis of the content of steroids, mainly cortisol, has been proposed as a promising biomarker for individual cortisol exposure over a defined period of time 73 . The sampling procedure is easy and non-invasive, and samples can be stored at room temperature for years. Determination of cortisol content in human scalp hair was recently used for retrospective determination of disease onset in a patient with primary adrenal insufficiency 74 . Furthermore, measurement of hair cortisol content in patients with corticotroph insufficiency could be considered as a useful tool for therapy monitoring rather than as a diagnostic tool 73 .
The diagnosis of GHD requires confirmation using a growth hormone stimulation test. The insulin hypoglycemia test represents the gold standard, but there are several contraindications and the test procedure is physically demanding for patients. Body mass index-related cutoff values are used in the GHRH-arginine test 75 , and recent studies have emphasized new cutoff levels for the glucagon stimulation test, which seems to be an acceptable alternative to the insulin hypoglycemia test 76 . Furthermore, new growth hormone stimulants for oral application are currently being validated in clinical trials 77,78 .

Conclusion and future perspective
In patients with acquired pituitary insufficiency, understanding of the pathomechanisms has increased in recent years. The importance of a proper blood supply and the role of autoimmunity have been discussed. A more profound understanding of the underlying causes leading to secondary hypopituitarism may lead to recommendations for rational screening of hypopituitarism in more precisely defined risk populations and therefore may prevent under-or overdiagnosis.
Gene sequencing methods have markedly improved. New diagnostic tools led to the discovery of different genetic alterations in patients with congenital hypopituitarism. These mutations mainly affect genes important for pituitary ontogenesis. However, the diagnosis of congenital causes for hypopituitarism can be challenging and remains unsolved in the majority of patients. Depending on the phenotype and medical history of a single patient, an individual approach for stepwise genetic testing is essential. Increasing our understanding of molecular pathways relevant in pituitary function also includes expanding on current knowledge of pituitary cell differentiation and proliferation. Very recently, it was shown that pituitary cells seem to have the capability to regenerate, which provides a promising approach for therapeutic strategies for patients with permanent hypopituitarism 79 . Nevertheless, activation of pituitary stem cells and cell transplantation in the treatment of pituitary insufficiency is still a long way off 80,81 .
Author contributions MRS performed the literature search on genetics, drug-induced hypopituitarism, and diagnostic findings and prepared the first draft of the manuscript. AK wrote the sections on acquired hypopituitarism and autoimmunity and revised the first draft. UR contributed to the section on vascularization. GKS added his expertise by carefully reading and revising the manuscript, especially the sections on Sheehan's syndrome and radiation-induced hypopituitarism. All authors were involved in the revision of the draft manuscript and have agreed to the final content.

Competing interests
The authors declare that they have no competing interests.

Grant information
The author(s) declared that no grants were involved in supporting this work.