Pituitary Adenoma 1, Multiple Types

A number sign (#) is used with this entry because of evidence that multiple types of pituitary adenoma (PITA1) are caused by heterozygous mutation in the aryl hydrocarbon receptor-interacting protein gene (AIP; 605555) on chromosome 11q13.

Description

Mutations in the AIP gene have been found predominantly in growth hormone (GH)-secreting adenomas, but have also been found in adrenocorticotropic hormone (ACTH)-secreting, thyroid hormone (TSH)-secreting, and prolactin (PRL)-secreting pituitary tumors.

Pituitary adenomas are benign monoclonal neoplasms of the anterior pituitary gland, accounting for approximately 15% of intracranial tumors. Growth hormone (139250)-secreting adenomas, also known as somatotropinomas, which clinically result in acromegaly, comprise about 20% of all pituitary tumors and are the second most common hormone-secreting pituitary tumor after prolactin (176760)-secreting tumors, which account for 40 to 45% of pituitary tumors. ACTH-secreting tumors, which result in Cushing disease, and thyrotropin (TSHB; 188540)-secreting tumors are much less common. Nonsecreting pituitary tumors, which account for about 33%, can cause symptoms due to local compressive effects of tumor growth (Vierimaa et al., 2006; Georgitsi et al., 2007; Horvath and Stratakis, 2008).

Acromegaly is characterized by coarse facial features, protruding jaw, and enlarged extremities (Vierimaa et al., 2006). Familial isolated somatotropinoma (FIS) is defined as the occurrence of at least 2 cases of acromegaly or gigantism in a family that does not exhibit features of other endocrine syndromes. FIS patients tend to have onset about 4 to 10 years earlier than patients with sporadic disease (Gadelha et al., 1999; Horvath and Stratakis, 2008).

Cushing disease is characterized by central obesity, moon facies, diabetes, 'buffalo hump,' hypertension, fatigue, easy bruising, depression, and reproductive disorders. Cushing disease is associated with increased morbidity and mortality, mainly due to cardiovascular or cerebrovascular disease and infections (summary by Perez-Rivas et al., 2015).

Familial isolated pituitary adenoma (FIPA) and pituitary adenoma predisposition (PAP) are terms referring to families in which 2 or more individuals develop pituitary tumors. Within a family, tumor types can be heterogeneous, with members of the same family having GH-secreting, prolactin-secreting, ACTH-secreting, or nonsecreting adenomas; in contrast, some families are homogeneous with regard to tumor type. Familial isolated somatotropinoma refers specifically to GH-secreting tumors and is usually associated with an acromegaly phenotype. Thus, FIS is a subset of FIPA or PAP (Toledo et al., 2007).

Schlechte (2003) discussed prolactinoma in general terms as a clinical, diagnostic, and therapeutic problem.

Genetic Heterogeneity of Pituitary Adenomas

Also see pituitary adenoma-2 (PITA2; 300943), caused by mutation in the GPR101 gene (300393); pituitary adenoma-3 (PITA3; 617686), caused by somatic activating mutations in the GNAS1 gene (139320); pituitary adenoma-4 (PITA4; 219090), caused by somatic mutation in the USP8 gene (603158); and pituitary adenoma-5 (PITA5; 617540), caused by mutation in the CDH23 gene (605516).

Patients with the chromosome Xq26.3 microduplication syndrome (300942) have growth hormone-secreting adenomas.

Familial acromegaly can also occur in association with multiple endocrine neoplasia type I (MEN1; 131100), Carney complex (CNC1; 160980), and the McCune-Albright syndrome (174800).

Rostomyan et al. (2015) performed a retrospective analysis of 208 patients with pituitary gigantism due to pituitary adenoma or hyperplasia. Most patients (78.4%) were male, and the median onset of rapid growth was 13 years of age for boys and 11 years for girls. Of the 143 patients who consented to genetic testing, 29% had AIP mutations, and microduplication at Xq26.3 (XLAG; 300942) was present in 2 familial isolated pituitary adenoma kindreds and in 10 sporadic patients. Rostomyan et al. (2015) noted that no genetic etiology was identified in more than 50% of the cases, and that the genetically unexplained cases showed more aggressive disease in terms of invasion, hormone levels, and lower control rates.

Clinical Features

Levin et al. (1974) reported 2 brothers with acromegaly confirmed by elevated serum GH levels and the finding of pituitary tumors. Both also had acanthosis nigricans.

Jones et al. (1984) reported an uncle and nephew with acromegaly. The authors considered MEN type I to be unlikely because of the absence of other endocrine disease at an advanced age.

Abbassioun et al. (1986) and McCarthy et al. (1990) also reported familial acromegaly.

Pestell et al. (1989) described a family in which 5 members over 3 generations had isolated functional pituitary adenomas. Four patients had acromegaly and 1 had galactorrhea from prolactin excess. Affected individuals were related as uncle and nephew or uncle and niece or as second cousins; no parent-child transmission was observed and there was no consanguinity. Pestell et al. (1989) proposed autosomal dominant inheritance with reduced penetrance. The authors considered the disorder in this family to be distinct from MEN1.

Links et al. (1993) reported a father and son with acromegaly associated with pituitary adenoma. The adenoma from the son was also found to secrete thyroid-stimulating hormone and prolactin. The father was deceased at the time of the report.

Berezin and Karasik (1995) studied 4 families in each of which more than one member were found to have prolactinoma. They concluded that there is a familial tendency to prolactinoma independent of its association with MEN1.

Gadelha et al. (1999) reported 2 unrelated families with isolated acromegaly/gigantism. In one family, 3 of 4 sibs were affected, with ages at diagnosis of 19, 21, and 23 years. In the other family, 5 of 13 sibs were diagnosed as affected at 13, 15, 17, 17, and 24 years of age. There was no history of consanguinity in either family, and the medical histories and laboratory results excluded MEN1 and the Carney complex.

Verloes et al. (1999) reported 3 unrelated families in which 2 members each had acromegaly not associated with other clinical features of MEN1. Two of the 6 patients also had galactorrhea due to prolactin secretion. Age at onset was usually in the twenties. After a review of similar families that had been published, Verloes et al. (1999) concluded that the disorder was a unique entity and showed autosomal dominant inheritance with reduced penetrance.

Jorge et al. (2001) reported a Brazilian family with acromegaly due to pituitary adenomas. The proband was a 24-year-old woman who presented with headaches, galactorrhea, menstrual irregularities, and progressive enlargement of hands and feet. Physical examination revealed evident acromegalic facial and acral features. Serum GH and prolactin were increased. The proband's brother presented at age 29 years with a 10-year history of progressive enlargement of hands, feet, and mandible. Serum growth hormone and insulin-like growth factor-1 (IGF1; 147440) were increased, but prolactin was normal. Other endocrine values were normal in both patients, excluding endocrine syndromes. Their father had acromegalic features confirmed by family pictures; he had died of an unrelated cause at the age of 40 years without endocrine evaluation. Molecular analysis of the sibs excluded germline mutations in the MEN1, GNAS1, GNAI2 (139360), and GHRHR (139191) genes.

Vierimaa et al. (2006) described a large kindred in northern Finland in which multiple individuals had pituitary adenomas, secreting either prolactin (5) or growth hormone (4); 2 individuals had a mixed tumor secreting both hormones. There were 3 clear cases of acromegaly or gigantism. Genealogy could be traced to the 1700s. Vierimaa et al. (2006) postulated that the phenotype represented a hereditary predisposition to pituitary adenomas (PAP) with very low penetrance. A second family had 2 individuals in 2 generations with somatotropinomas. Compared to patients with sporadic pituitary tumors, those with PAP had a significantly younger age at time of diagnosis (24.7 vs 43.6 years, P = 0.0003), but there were no differences in tumor size or sex distribution. Six of the 15 patients diagnosed under 35 years of age (40%) in the population-based series had PAP.

Other Features

Lopez-Velasco et al. (1997) found that hypertension was present in approximately 43% of patients with active acromegaly and in 28% of patients in whom acromegaly was cured. Studies of other cardiac parameters, including functional cardiac indexes and echocardiography, showed that hypertension was independently related to cardiac morphology and to systolic and diastolic function. Acromegaly was related to an increase in left ventricular mass, stroke volume, cardiac output, and isovolumic relaxation time, which were independent of the presence of hypertension. In the 5 patients in whom active acromegaly was successfully treated, left ventricular mass and left ventricular posterior wall thickness were reduced 1 year later. Lopez-Velasco et al. (1997) concluded that asymptomatic morphologic and functional cardiac abnormalities present in acromegalic patients are independently related to acromegaly and hypertension, suggesting the existence of a specific acromegalic myocardiopathy that might be aggravated by the coexistence of hypertension.

Among 25 patients with uncomplicated acromegaly and 25 controls, Colao et al. (2002) found similar resting blood pressure, whereas heart rate at rest and systolic blood pressure at peak exercise were higher in the patients. The left ventricular mass index was higher in acromegalic patients than in controls; 7 patients had left ventricular hypertrophy. Diastolic function was similar in the 2 groups. The ejection fraction at rest, but not at peak exercise, was significantly increased in the patients compared with controls; as a consequence, the exercise-induced changes in the ejection fraction were lower in patients than controls. At common carotid ultrasonography, young patients with acromegaly had increased diastolic peak velocity and increased intima media thickness. The authors concluded that short-term GH excess, despite causing enhanced cardiac performance at rest, reduces cardiac performance on effort and impairs vascular morphology. These deleterious effects of early-onset acromegaly were ameliorated by suppressing GH/IGF1 levels for 6 months.

Parkinson et al. (2001) found that women with active acromegaly had serum IGF1 values 82 ng/ml less than males (P less than 0.02) for a given serum GH value. In females receiving oral estrogen, mean serum IGF1 for a given GH value was 130 ng/ml lower than in males (P = 0.01), but only 60 ng/ml lower than in the remaining 45 females (NS; P = 0.2). The authors concluded that there is a gender difference in the relationship between serum GH and IGF1 in patients with active acromegaly consistent with relative GH resistance observed in normal and GH-deficient females, which may, in part, be mediated by estrogen.

Mapping

Thakker et al. (1993) found loss of heterozygosity (LOH) for chromosome 11q13 in 4 somatotrophinomas derived from non-MEN1 patients with acromegaly.

Gadelha et al. (1999) found loss of heterozygosity of chromosome 11q13 in all pituitary adenomas isolated from affected members of 2 unrelated families with acromegaly. None of the patients had germline mutations in the MEN1 gene, and a somatic mutation was not identified in tumor tissue from 1 patient. Gadelha et al. (1999) concluded that LOH in these affected family members was independent of MEN1 changes and due to another tumor suppressor gene in the 11q13 region.

By linkage analysis of 2 unrelated families with familial isolated somatotropinomas, Gadelha et al. (2000) found linkage to an 8.6-cM region on chromosome 11q13.1-13.3 (maximum 2-point lod scores of 3.0 or more between FGF3 (164950) and D11S1335). Stratakis and Kirschner (2000) recalculated the lod scores for 11q using the germline alleles reported by Gadelha et al. (2000); this analysis yielded 2-point lod scores that were strongly positive, but not conclusive. Stratakis and Kirschner (2000) concluded that LOH at 11q13 was likely to be a tertiary hit at the tumor tissue level.

Using haplotyping and allelotyping techniques to evaluate 8 families with FIS and 15 sporadic somatotropinomas, Soares et al. (2005) narrowed the candidate locus to a 2.21-Mb region on chromosome 11q13.3. LOH at this region was found in all families and in 40% of sporadic tumors. Three potential candidate genes in this region were sequenced, but no mutations were found.

By whole-genome single-nucleotide polymorphism (SNP) genotyping of a large Finnish family with pituitary adenoma predisposition, Vierimaa et al. (2006) found linkage to chromosome 11q12-q13. The results yielded a lod score of 7.1 when combined with a second affected family that shared the linked haplotype. No mutations were identified in the MEN1 gene, which maps to this region.

Molecular Genetics

In affected individuals from a large Finnish family with pituitary adenoma predisposition, Vierimaa et al. (2006) identified a heterozygous germline mutation in the AIP gene (Q14X; 605555.0001). Five individuals had prolactinomas, 4 had somatotropinomas, and 2 had a mixed tumor comprising both cells. Further screening identified this mutation in 6 of 45 patients from a population-based cohort with acromegaly. Affected Italian sibs were found to have an R304X mutation (605555.0003). Loss of heterozygosity at the AIP locus was detected in all 8 pituitary tumors analyzed, including somatotropinomas, prolactinomas, and mixed-type tumors.

In the Brazilian sibs with acromegaly and GH-secreting pituitary adenomas reported by Jorge et al. (2001), Toledo et al. (2007) identified a heterozygous mutation in the AIP gene (605555.0007). A 41-year-old brother with the mutation was clinically unaffected, but was found on imaging to have a small, apparently nonsecreting pituitary nodule. A 3-year-old boy with the mutation was also unaffected, but was younger than the average age at symptom onset.

In 9 of 460 patients from Europe and the U.S. with pituitary adenomas, Georgitsi et al. (2007) identified 9 different germline mutations in the AIP gene (see, e.g., 605555.0004-605555.0006). Eight patients had GH-secreting tumors and acromegaly, and 1, a 26-year-old Polish patient, had Cushing syndrome due to an ACTH-secreting tumor (see 605555.0008). Age at diagnosis ranged from 8 to 41 years.

Daly et al. (2007) studied the frequency of AIP gene mutations in a large cohort of patients with familial isolated pituitary adenoma from 9 different countries. Seventy-three FIPA families were identified, with 156 patients with pituitary adenomas; the FIPA cohort was evenly divided between families with homogeneous and heterogeneous tumor expression. Eleven FIPA families had 10 germline AIP mutations; 9 of the mutations were novel. Tumors were significantly larger (p = 0.0005) and diagnosed at a younger age (p = 0.0006) in AIP mutation-positive versus mutation-negative subjects. Although somatotropinomas predominated among FIPA families with AIP mutations, mixed GH/prolactin-secreting tumors, prolactinomas, and nonsecreting adenomas were also found. Approximately 85% of the FIPA cohort and 50% of those with familial somatotropinomas were negative for AIP mutations.

Barlier et al. (2007) did not identify mutations in the AIP gene in 107 European patients with sporadic pituitary adenomas, including prolactinomas (49), somatotropinomas (26), ACTH-secreting tumors (2), TSH-secreting tumors (1), and nonfunctioning tumors (29). One additional patient with a somatotropinoma was found to have a germline mutation in the AIP gene (R22X; 605555.0009). Barlier et al. (2007) concluded that germline AIP mutations are infrequent in patients with sporadic pituitary adenomas.

Igreja et al. (2010) analyzed the AIP gene in 38 families with FIPA, in which at least 2 family members had pituitary adenoma without features of MEN1 (131100) or Carney complex (see 160980), and identified mutations in 11 of the families, including 3 with large deletions. The authors reviewed the clinical characteristics of these 38 families and 26 previously reported families (Leontiou et al., 2008), confirming that patients with AIP mutations had a lower mean age at diagnosis. Igreja et al. (2010) noted that overall, AIP mutations were implicated in 20 (31%) of the 64 families in their FIPA cohort.

Kamenicky et al. (2015) screened their entire cohort of 263 patients with acromegaly or gigantism for germline mutations in AIP and identified mutations in 8 patients with somatotropinomas (3.0%), 6 (75%) of whom had gigantism. None of the 263 patients carried germline mutations in both GPR101 and AIP.

Pathogenesis

Shimon and Melmed (1997) reviewed the multiple molecular events known at the time that result in pituitary adenomas. These events include early chromosomal mutations (11q13, 13q14 LOH) and possibly expression of pituitary-specific protooncogenes and/or growth factors including GNAS1, CREB (see CREB1; 123810 and CREB2; 123811), HST (see FGF4; 164980) and TGF-alpha (see TGFA; 190170). Subsequent permissive factors allowing clonal expansion of the transformed cell include hypothalamic hormone receptor signals, paracrine growth factor signals, and disordered cell cycle regulation.

In 6 of 14 sparsely granulated human somatotroph adenomas, Asa et al. (2007) identified somatic mutation of codon 49 (H49L or H49R) of the growth hormone receptor gene (GHR; 600946) within an extracellular cysteine-rich immunoglobulin-like loop. In vitro functional studies with mutant rabbit Ghr showed that codon 49 mutations impaired receptor processing, activation, and binding of GH. Mutant Ghr was retained within cytoplasmic granules in the endoplasmic reticulum, and there was relative resistance of mutant Ghr to activation of intracellular signaling by GH. Thus, mutant Ghr showed ineffective sensing of ambient GH and lacked negative feedback on GH production and growth, suggesting another pathogenetic mechanism for a subgroup of pituitary somatotroph adenomas. Asa et al. (2007) noted that the findings were significant for treatment, in that the disruption of GH autoregulation by a GHR mutation in sparsely granulated adenomas renders GHR antagonism a more appropriate therapeutic option than GH antagonism, since the former would be less likely to be associated with treatment-induced tumor activation.

In studies of acromegalics with abnormally high levels of growth hormone, Boguszewski et al. (1997) evaluated the proportion of circulating non-22-kD isoforms of GH and found the proportion was fairly constant in different samples from the same patient, regardless of the GH level. A wide variation of values was observed among acromegalics, both before (14-51%) and after surgery (8-62%). The proportion of non-22-kD GH isoforms was increased in untreated patients, compared with controls (26.6 vs 17.4%; P less than 0.01), and the values correlated significantly to tumor size, mean 24-hour GH concentration, serum PRL, and extracellular water. They concluded that acromegalics have an increased proportion of circulating non-22-kD GH isoforms. Although values are fairly constant in different samples from an individual, a large spectrum can be observed among patients. This variability suggested to Boguszewski et al. (1997) that different pituitary adenomas secrete GH isoforms in variable amounts. Their observation that a higher proportion of non-22-kD GH isoforms is present in patients not truly cured after surgery suggested to the authors that the evaluation of non-22-kD GH isoforms can be useful in the follow-up of acromegalic patients.

History

Chahal et al. (2011) identified the arg304-to-ter mutation in the AIP gene (605555.0003) in DNA extracted from the teeth of an Irish patient with gigantism who lived from 1761 to 1783. This patient's skeleton had been examined by Harvey Cushing, who identified an enlarged pituitary fossa and ascribed his gigantism to a pituitary adenoma. Chahal et al. (2011) also identified this mutation in 4 contemporary northern Irish families who presented with gigantism, acromegaly, or prolactinoma and had the same haplotype. Using coalescent theory, Chahal et al. (2011) inferred that these persons shared a common ancestor who lived about 57 to 66 generations earlier. The skeleton of the patient (Charles Byrne, known as 'The Irish Giant'; Bergland, 1965) is exhibited in the Hunterian Museum of the Royal College of Surgeons in London, near the skeleton of Caroline Crachami (see 210730).