Pituitary adenomas in young patients: when should we consider a genetic predisposition?

Pituitary adenomas (PAs) are increasingly being recognized in practice, with recent data indicating a prevalence of clinically-relevant PA of up to one per 1000 individuals [1]. Less than 10% of pituitary adenomas are, however, diagnosed before the age of 20 years [1,2]. Most are hormonally active, with prolactinomas being by far the most frequently encountered phenotype, followed by corticotropinomas, especially before puberty, and somatotropinomas. Nonsecreting PAs occur in a minority of adolescents and young adults with PA [2–4]. Nonetheless, PA in this age range represent an important clinical issue. Most are macroadenomas, with attendant mass effects due to their extrasellar extension, in particular headache and visual loss [2–4]. They are frequently endowed with specific, age-related endocrine morbidities owing to the impact of pituitary dysfunction on somatic growth, pubertal development and fertility. These effects may be life-long and can have a significant psychosocial impact. The need for multiple therapeutic modalities (including recurrent surgery and radiotherapy), and the risks for neurological sequalae and pituitary dysfunction, increase with tumor size and aggressiveness. Therefore, achieving an early diagnosis of PA provides an advantage to achieving a cure and a normal adult life. From this perspective, awareness of a genetic predisposition to PA is essential. Recent discoveries regarding the pathogenesis and epidemiology of PA may contribute to this effort and modify the clinical approach to young patients with PA.

The pathogenesis of PA is multifactorial, generally involving a number of tumor-promoting somatic events and extracellular signals inducing cell cycle dysregulation [5]. Accordingly, PAs are typically sporadic and develop in adult life, with an incidence peaking between 20 and 40 years of age for prolactinomas and between 40 and 60 years of age for other phenotypes. To date, only a handful of genetic factors have been recognized but genetic contribution is expected to be greater in young patients, who may have inherited one or more tumor susceptibility genes, explaining the early onset of the disease [4–6]. A genetic background is likely (indeed, irrespective of the patient’s age) if PAs occur in the presence of multiple endocrine neoplasia type 1 (MEN1), which is associated with a germline mutation in the MEN1 gene in nearly 80% of cases. In addition, germline mutation in the CDKN1B gene that encodes p27^Kip1 can lead to a related MEN1-like syndrome [7]. This condition however is exceptionally rare [8]. PA can also occur in the presence of complex diseases such as the Carney complex (CNC), due to a mutation in the PKRAR1A gene in approximately 60% of cases, or the McCune Albright syndrome (MAS), which is due to a post-zygotic activating mutation of the Gsα subunit gene. In CNC and MAS, somato–lactotroph hyperplasia may occur in parallel with, or as part of the formation of PA [4,6].

In recent years, evidence of PA occurring in a familial setting outside of the genetic conditions cited earlier has progressively accumulated in the literature. Traditionally it was believed that only somatotropinomas could occur in a familial setting, and even these were encountered infrequently. In 1999, 15 families were identified in Liège (Belgium) with PA with various phenotypes (not only somatotropinomas); these families had no extrapituitary disease and hence we designated this new entity as familial isolated pituitary adenoma (FIPA) [9]. During the following years, we expanded the cohort to other centers across Europe and elsewhere [10–12] and by 2006 we reported 138 patients in 64 FIPA kindreds [13]. FIPA kindreds can present either as homogeneous families with a single pituitary tumor phenotype among affected members, or as heterogeneous kindreds, in which different PA phenotypes can manifest within the same family. Our work has revealed that familial occurrence of PA was previously underestimated, with FIPA kindreds accounting for approximately 2% of patients with PA. Since no environmental factor has yet been recognized to induce PA in humans, this has strongly suggested the role of hereditary factors. In 2006, the aryl hydrocarbon receptor interacting protein (AIP) gene was identified as a new PA predisposition gene. Inactivating germline mutations of AIP were reported in large Finnish pedigrees with somatotropinomas and/or prolactinomas [14]. The AIP gene is a new pituitary tumor suppressor gene, located in 11q13, whose causative role in pituitary pathogenesis was further supported by the identification of germline AIP mutations (AIP^mut) in 15% of FIPA kindreds, including up to 50% of those families with homogeneous somatotropinomas [15]. By contrast, germline AIP mutations have rarely been encountered in unselected patients with sporadic PA (<1%) [16]. One exception of great interest is that AIP^mut may appear more frequently in young patients with growth hormone (GH)- or prolactin-secreting macroadenomas [17–19]. According to preliminary data from international series, distinct changes in the germline AIP sequence can be observed in up to 16% of young patients with pituitary macroadenomas [20]. At this time, germline mutations in the CDKN1B gene do not seem to play an important role in patients with familial PA [8,19,21].

Knowledge regarding the clinical characteristics of AIP^mut PA has been accumulating rapidly over the past 3 years [6,22]. Preliminary data from an international study on 92 patients with AIP mutations and pituitary adenomas [23] indicate that 80% had somatotropinomas, with a mean age at diagnosis of less than 25 years, which may explain an intriguingly high rate of gigantism (>25%) in these patients. More than 90% of AIP^mut tumors were macroadenomas and most were invasive, making surgical treatment unable to control GH/IGF-1 hypersecretion in approximately two thirds of patients. Prolactinomas accounted for 12% of cases, most of them being large and invasive, with a mean age at diagnosis of approximately 20 years of age. In young sporadic patients with AIP mutations, the mean maximal diameter exceeded 25 mm [20]. Disease aggressiveness in patients with AIP^mut often translates into the use of multiple therapeutic modalities to achieve control with attendant difficulties in adequately preserving or restoring normal pituitary functions [23]. In addition, AIP may be involved in somatotropinoma response to somatostatin analogs and pharmacological resistance appears to be frequent in AIP^mut patients [23].

The potential aggressiveness of PA in young patients with AIP mutations strongly supports genetic screening starting in early childhood in kindreds carrying an AIP mutation. Recent experience with an AIP^R304X kindred affected by hard-to-treat large somatotropinomas, in whom an intrasellar macroadenoma in a 9-year-old boy with incipient gigantism was successfully resected, is encouraging and further supports this view [24]. Since up to 30% of patients with identified AIP mutations may present as supposedly ‘sporadic’ cases [23], screening for AIP mutations could be extended to a subset of non-familial PA patients. In fact, supposed sporadic presentation of familial disease may be explained by a low penetrance of the disease, poor information on the family history or the presence of environmental or genetic modifiers. Current knowledge strongly supports young age as the most important criteria for patient selection for screening in this setting, whereas additional criteria concerning tumor phenotype and behavior need to be defined further. In fact, the loss of immunostaining for AIP in paraffin-embedded sections of operated PA was first proposed as a prescreening tool [25], but further experience reporting detectable AIP immunostaining in most AIP^mut PA [24,26] and frequent AIP downregulation in aggressive somatotropinomas and in prolactinomas [24] do not support this view. Data from systematic AIP sequencing in somatotropinomas indicated young age as the only independent predictive factor for the presence of AIP mutations [24]. At the moment, young patients with somatotropinoma or large prolactinomas represent the best candidates for germline AIP sequencing.

Attempting to translate current knowledge into practical guidelines, we first strongly recommend that familial history should be considered systematically in young patients presenting with PA. A detailed familial interview should look for previous cases of PA among relatives as well as for clinical elements suggesting a familial history of MEN1 or CNC. Second, extrapituitary manifestations arguing for a complex disease, such as MEN1, CNC or MAS, should be looked for –in this perspective, cutaneous examination may often be informative. Familial and individual data will then allow one to select patients who should benefit from genetic screening for MEN1, or rarely for PKRAR1A or Gsα mutations. In a true FIPA context, MEN1 is unlikely and AIP screening should be considered first, with a 50% rate of germline AIP mutations in kindreds with homogeneous somatotropinomas. Although the probability of identifying a germline AIP mutation in other FIPA kindreds is lower, it may have familial implications, and systematic identification of AIP-negative FIPA kindreds is of benefit to help identify new candidate genes in the future. Finally, current experience in sporadic PA strongly supports AIP sequencing in the presence of early-onset somatotropinomas or large prolactinomas, with familial screening being important whenever an AIP mutation or suspicious variant is identified.

Financial & competing interests disclosure

The work described in this publication was supported in part by grants from the Fonds d’Investissement pour la Recherche Scientifique (FIRS) from the CHU de Liège, Belgium and the Italian Ministry for University and Research (MIUR), Neuromed Institute, IRCCS and the ‘Carlo Ferri’ Foundation for Prevention in Oncology, Italy. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.