Exp Clin Endocrinol Diabetes 2024; 132(12): 678-696
DOI: 10.1055/a-2337-2265
Review

An Update on the Genetic Drivers of Corticotroph Tumorigenesis

Laura C. Hernández-Ramírez
1   Red de Apoyo a la Investigación, Coordinación de la Investigación Científica, Universidad Nacional Autónoma de México e Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán, Mexico City, Mexico
,
Luis Gustavo Perez-Rivas
2   Medizinische Klinik und Poliklinik IV, LMU Klinikum, LMU München, Munich 80336, Germany
,
Marily Theodoropoulou
2   Medizinische Klinik und Poliklinik IV, LMU Klinikum, LMU München, Munich 80336, Germany
,
Márta Korbonits
3   Centre for Endocrinology, Barts and The London School of Medicine, Queen Mary University of London, Charterhouse Square, London, UK
› Institutsangaben

Funding Information Sociedad Mexicana de Nutrición y Endocrinología — Financial support for research projects Dirección General de Asuntos del Personal Académico, Universidad Nacional Autónoma de México — http://dx.doi.org/10.13039/501100006087; PAPIIT projects TA200322 and TA200524 Universidad Nacional Autónoma de México — http://dx.doi.org/10.13039/501100005739; Coordinación de la Investigación Científica Society for Endocrinology — http://dx.doi.org/10.13039/501100000382; Equipment grant
 

Abstract

The genetic landscape of corticotroph tumours of the pituitary gland has dramatically changed over the last 10 years. Somatic changes in the USP8 gene account for the most common genetic defect in corticotrophinomas, especially in females, while variants in TP53 or ATRX are associated with a subset of aggressive tumours. Germline defects have also been identified in patients with Cushing’s disease: some are well-established (MEN1, CDKN1B, DICER1), while others are rare and could represent coincidences. In this review, we summarise the current knowledge on the genetic drivers of corticotroph tumorigenesis, their molecular consequences, and their impact on the clinical presentation and prognosis.


Introduction

Corticotrophinomas, also known as corticotroph tumours, are infrequent and usually benign pituitary neuroendocrine tumours (PitNETs). Derived from TPIT-expressing progenitors, most corticotrophinomas secrete adrenocorticotropic hormone (ACTH) excessively and lead to Cushing’s disease (CD), although some are clinically silent [1]. The untreated CD is associated with higher morbidity and mortality, and therefore, timely diagnosis and efficient treatment are critical [2] [3] [4] [5]; diagnosis is delayed by an average of 3.2 years [6]. Furthermore, the chronic complications of hypercortisolaemia result in an impaired quality of life that might persist even after successful treatment [7] [8]. The differential diagnosis of Cushing’s syndrome entails a complex series of biochemical and imaging studies [9], although CD is the most common cause of endogenous cortisol excess. Transsphenoidal surgery (TSS) is the first line of treatment and is successful in most cases, but therapeutic options are limited – and often ineffective – in case of postoperative disease persistence or recurrence [10] [11].

A great interest in the genetic defects driving corticotroph tumorigenesis has arisen in the last few years, particularly after the discovery that a large proportion of sporadic corticotrophinomas are driven by somatic variants in a single gene, USP8 [12] [13]. Pathogenic variants in other genes are less common and can be associated with specific phenotypes and clinical courses. Here, we review the current knowledge on the genetic drivers of corticotroph tumorigenesis ([Fig. 1]), their molecular consequences, and their impact on the clinical presentation and prognosis.

Zoom
Fig. 1 Summary of the genetic causes of corticotroph tumours, according to clinical presentation. Altered genes harbour variants in germline DNA unless marked as somatic. CS, Cowden syndrome; FIPA, familial isolated pituitary adenoma; CNC, Carney complex; CAH, congenital adrenal hyperplasia. *Case reports only.

Somatic genetic variants associated with sporadic corticotroph tumours

Next-generation sequencing (NGS) techniques have allowed the identification of somatic pathogenic variants in a group of genes ([Table 1]). About 40–70% of sporadic cases of CD are due to recurrent somatic variants in hotspot regions of USP8 (15q21.2) or USP48 (1p36.12) [12] [13] [14] [15] [16]. Variants in BRAF (7q34) and NR3C1 (5q31.3) have been described in a few cases, and their actual prevalence is still unknown. Co-occurrence of USP8, USP48, BRAF and NC3R1 variants has never been reported. Additionally, variants in TP53 (17p13.1), ATRX (Xq21.1), and DAXX (6p21.32) are rare in benign adenomas but are overrepresented in corticotrophinomas with aggressive features [16] [17] [18]. Furthermore, other somatic defects, with or without concomitant variants in more prominent genes (such as USP8, USP48, ATRX, NR3C1 and TP53), have also been described; however, their relevance needs to be clarified in further studies (GNAS [19] [20], FAT1, HCFC1, FBL, ANKRD27, STAG2, PKHD1, MXRA5, MAST4, AKAP6 and CHD2 [21]). Somatic MEN1 variants have rarely been identified in corticotroph tumours.

Table 1 Genes involved in the pathogenesis of corticotrophinomas: prevalence and clinical relevance.

Gene

Defects in corticotrophinoma*

Genotyped cases (total screened and prevalence range, if available)

Clinical features

USP8

Somatic hotspot
(Leu713-Thr723) GOF variants; germline hotspot GOF variant in one case

~600 cases (screened: ~2000, prevalence: 0–65%) [12] [13] [16] [18] [22] [23] [24] [25] [27] [28] [29] [30] [31] [33] [34] [37] [44] [49] [50] [51]

Sex:
 Higher OR for female [12] [13] [14] [16] [22] [23] [27] [28] [29] [43] [48]
Age at diagnosis:
 Older paediatric patients [23] [33] and younger adults [16] [24] [28] [34]
Tumour size:
 Smaller (or higher OR for microadenomas)
[12] [16] [22]
 Larger (or higher OR for macroadenomas) [31] [308]
Invasion:
 Lower OR [12] [18]
 Higher OR [51]
Remission:
 Higher OR [26] [27] [27]
Recurrence:
 Higher OR [23] [24] [31]
Hormone levels:
 Lower (ACTH, UFC, LDDST) [13] [22] [34] [49]
 Higher (ACTH, UFC, LDDST) [12] [48]
Germline defect: CD, developmental delay, dysmorphic features, ichthyosiform hyperkeratosis, chronic lung and kidney disease, cardiomyopathy, hyperinsulinaemia, and GH deficiency [38]

USP48

Somatic hotspot
(Met415) GOF variants

Fifty-six cases (screened: 577, prevalence: 0–23%) [15] [16] [31] [33] [34] [50]

Size:
 Smaller [16]
Invasion:
 Higher OR [50]

NR3C1

Disperse

Nine cases (screened: 116, prevalence: 0–6.1%) [13] [15] [16] [33] [35] [56] [57]

Unknown

BRAF

Somatic hotspot (Val600Glu) GOF variant

Sixteen cases (screened: 517, prevalence: 0–16.5%) [13] [15] [16] [33] [35] [56] [57]

Unknown

TP53

Somatic LOF variants

Twenty-one cases (screened: 170, prevalence: 0–33%) [26] [35] [64] [65] [66] [67] [68] [69]

Higher number of interventions, higher OR for invasion, and shorter overall survival [18]

ATRX

Somatic LOF variants

Ten cases (screened: 41, prevalence: unknown) [16] [17] [26]

Aggressive and metastatic tumours [17]

DAXX

Somatic LOF variants

Two cases (screened: 18, prevalence: unknown) [16]

Unknown

MEN1

Germline or somatic LOF variants

At least 34 cases with germline variants [33] [75] [78] [79] [92] [93] [94] [95] [98] [99] [100] [101] [102] [103] [104] (2.9% out of 245 paediatric cases in one cohort [33]); somatic variant in one case [213]

MEN1 (might debut as isolated paediatric CD)

CDKN1B

Germline LOF variants

Five cases [125] (1.2% out of 245 paediatric cases in one cohort [33])

MEN4, isolated paediatric CD

SDHA

Germline LOF variant

One case (1.2% out of 245 paediatric cases in one cohort [29])

Isolated paediatric CD, possibly: 3PAs

PRKAR1A

Germline LOF variants

Three cases [156] [157] [158] (0.4% out of 245 paediatric cases in one cohort [33])

CNC

RET

Germline GOF variants

Two cases [142]

3PAs, MEN2B

DICER1

Germline LOF variants

Nineteen cases [181] [203] [204] [205]

DICER1 syndrome, PitB

MSH2

Germline LOF variants

Two cases [133] [212]

LS, aggressive corticotrophinoma, ACTH-producing metastatic PitNET

MLH1

Germline LOF variant

One case [213]

LS, aggressive corticotrophinoma

TSC2**

Germline or somatic LOF variants

Four cases, three with germline and one with somatic variants (1.6% out of 245 paediatric cases in one cohort [29])

TSC, isolated paediatric CD (somatic variant)

CABLES1**

Germline LOF variants

Five cases (1.2% out of 245 paediatric cases in one cohort [29])

Young-onset CD, large corticotrophinoma

*Only pathogenic and likely pathogenic variants were accounted for. ** Requires further evidence of causality. 3PAs, phaeochromocytoma, paraganglioma and pituitary neuroendocrine tumour; CD, Cushing’s disease; CNC, Carney complex; GOF, gain-of-function; LDDST, low-dose dexamethasone suppression test; LOF, loss-of-function; LS, Lynch syndrome; MEN1, multiple endocrine neoplasia type 1; MEN4, multiple endocrine neoplasia type 4; OR, odds ratio; PitB, pituitary blastoma.

USP8

USP8 codes for a protein with deubiquitinase activity (UniProt P40818) that regulates membrane trafficking and protein turnover through deubiquitination of its substrate proteins. About 40% of sporadic CD tumours carry USP8 variants. The prevalence of USP8 variants in sporadic corticotrophinomas varies between 0–65% depending on the published series, with an average of ~40% [12] [13] [14] [16] [18] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34]. Among ~600 positive CD cases,>99% of the variants are located within a hotspot region coding for the amino acids 713–723 [12] [13] [23] [25] [29] [31] [35], which includes the 14–3–3 binding motif Arg715-Ser716-Tyr717-Ser718-Ser719-Pro720 [36]. Most variants target residues Ser718, Pro720, or both, the most common events being the substitutions p.Ser718Pro, p.Ser719Pro, p.Pro720Arg and p.Pro720Gln, as well as the in-frame deletion p.Ser718del [12] [13] [27] [29] [37]. Variants are present in heterozygosity and only in the tumour DNA, with the exception of a single female patient who carried the USP8 p.Ser719Pro variant as a de novo heterozygous germline defect [38]. This individual developed paediatric CD in the context of a complex syndrome, including developmental delay, dysmorphic features, ichthyosiform hyperkeratosis, chronic lung and kidney disease, dilated cardiomyopathy, hyperinsulinaemia and partial growth hormone (GH) deficiency.

Interaction between USP8 and 14–3–3 is essential for the fine control of USP8 activity, as 14-3-3 retains USP8 in an inactive state in the cytosol [36]. Disruption of that interaction leads to a sustained DUB activity and prevents degradation of its cargo proteins [39]. Hotspot variants also promote cleavage and translocation of USP8 into the nucleus [12] [13]. In addition, a somatic variant upstream of the hotspot (p.Gly664Arg) has been reported in one case. In vitro experiments showed that it reduces USP8/14-3-3 interaction and results in sustained activation of USP8, similar to the hotspot mutants [31]. In corticotroph tumours, the best-characterized mechanism of action of USP8 involves EGFR signalling [13]. EGFR ubiquitination after its activation promotes EGFR endocytosis and lysosomal degradation [39]. USP8 variants reduce EGFR ubiquitination, increase the pool of EGFR in the plasma membrane and enable sustained ERK1/2 signalling, which in turn induces POMC transcription and ACTH secretion [13].

USP8 variants are very specific to corticotrophinomas, as they are absent in ectopic ACTH-producing tumours and other pituitary entities [12] [13] [14] [40] [41] [42]. They are also much more frequent in functional corticotroph tumours than in silent tumours [26] [43] [44]. Some studies suggest that USP8 variants define a specific molecular subgroup of corticotrophinoma, with differential expression of proteins relevant for the corticotroph physiology (cyclin-dependent kinase inhibitor 1B, CDKN1B; cyclin-dependent kinase 5 and Abl enzyme substrate 1, CABLES1; GR, glucocorticoid receptors; somatostatin receptor 5, SST5), proliferation, cell-to-cell junction, and epithelial-to-mesenchymal transition [24] [25] [28] [29] [35] [44] [45] [46]. Tumours with USP8 variants seem to partially retain the original corticotroph physiology, as they exhibit larger suppression of pro-opiomelanocortin (POMC) by dexamethasone test in vitro [29] and better response to desmopressin stimulation test in vivo than the wild type [34]. In vitro experiments have shown that they also respond better to the somatostatin receptor ligand pasireotide [37] [47], although the translation into clinical practice needs further evaluation [44].

USP8 variants are much more common in female patients (range 27–68%) than in males (range 3.5–38%) [12] [13] [14] [16] [22] [23] [27] [28] [29] [43] [48]. It is unclear whether their prevalence is similar in paediatric (range 0%–31%) [23] [30] [33] and adult cases (range 12%–5%) [12] [18] [22] [24] [25] [27] [28] [29] [31] [34] [44] [49] [50], although they seem to be more frequent in teenagers diagnosed at older age [23] [33] and adults at younger age [16] [24] [28] [34]. The association of USP8 mutational status with hormone levels is inconsistent and, in some cases, contradictory. Some authors reported associations with lower plasma ACTH [13] [22] [34], higher ACTH/tumour size ratio [12], lower [49] or higher preoperative 24-hour urinary-free cortisol (UFC) [24], or higher postoperative UFC [43]. Nevertheless, most studies could not show significant differences between wild-type tumours and those with USP8 variants. Several publications suggest that tumours with USP8 variants tend to have a more benign phenotype, such as smaller size [12] [16] [22], lower invasion rates into the cavernous sinus [12] [18] or higher chances of achieving postoperative remission [26] [27]. Other studies have found evidence for a higher risk of invasion [51] or higher risk of recurrence after remission [23] [24] [31] and have reported USP8 variants in difficult-to-manage cases, including few patients with aggressive tumours and pituitary carcinomas [32] [48].


USP48

Variants in USP48, which encodes another deubiquitinase (UniProt Q86UV5), were discovered by NGS in USP8 wild-type corticotroph tumours [15] [16]. USP48 variants have been described in 1.5–23% of cases, all USP8 wild type [15] [16] [31] [33] [34] [50]. Similar to USP8, USP48 variants are more frequent in female patients [15] [16]. One study found tumours with USP48 variants to be significantly smaller [16], while another reported a higher rate of cavernous sinus invasion compared to wild-type tumours [50]. All somatic variants identified affect a single amino acid (Met415) and confer higher deubiquitinase activity to the mutant USP48 protein. A potential mechanism of action in corticotroph tumour cells may be through the stabilizing effect of USP48 on GLI1, a downstream transcription factor effector of the hedgehog pathway [52], which contributes to corticotroph pathophysiology via its crosstalk with CRH signalling [53] [54]. In a corticotroph tumour cell model, overexpression of a USP48 mutant enhanced CRH-induced (but not basal) POMC promoter activity in a GLI1-dependent manner, suggesting that it may promote corticotroph tumorigenesis by amplifying the trophic action of CRH [16].


NR3C1

Variants in the NR3C1 (nuclear receptor subfamily 3 group C member 1) gene that encodes for the glucocorticoid receptor (UniProt P04150) were initially described as rare events [55]. NGS studies showed somatic NR3C1 variants in ~6% of cases, but no clinical correlations have been established [13] [15] [16] [33] [35] [56] [57]. NR3C1 loss-of-function (LOF) in the corticotroph cells should impair the response to the negative adrenal feedback, although the contribution of this specific genetic defect to corticotroph tumorigenesis is unclear, given its low prevalence [56]. In addition, at the germline level, this genetic defect causes generalized glucocorticoid resistance [58]. Interestingly, germline LOF NR3C1 variants have been documented in two cases of CD, including a young man with generalized glucocorticoid resistance and a dominant-negative de novo variant, as well as a paediatric female patient carrying a germline likely pathogenic variant and no additional clinical features [33] [59].


BRAF

The oncogenic BRAF variant p.Val600Glu was reported in 16% of USP8 wild-type corticotrophinomas in one cohort [15], but not in other Asian, American or European cohorts [16] [31] [32] [33] [34] [50]. This hotspot variant is a frequent gain-of-function (GOF) defect in cancer that results in a constitutively active RAS-RAF-MEK-ERK pathway, and it has been exploited as a therapeutic target in various neoplasms [60] [61] [62] [63]. The evidence so far indicates that the BRAF p.Val600Glu variant is extremely rare or absent in the majority of adult and paediatric CD patient cohorts.


TP53

Variants in the TP53 tumour suppressor gene were initially considered rare events described only in isolated cases of aggressive and metastatic corticotrophinomas [64] [65] [66] [67] [68] [69]. NGS on selected cohorts of USP8 wild-type, aggressive and/or metastatic corticotroph tumours revealed somatic TP53 variants in up to 33% of cases [16] [17]. A large multicentre study on functional corticotrophinomas reported TP53 variants in 14% of all functional macroadenomas and in 24% of invasive cases [18]. TP53 variants correlated with higher Knosp grades and parasellar invasion, more therapeutic procedures and increased disease-specific mortality. In a case of progressive corticotroph tumour growth after bilateral adrenalectomy (Nelson syndrome), a TP53 variant was detected in the specimen obtained after radiation, but not in the pre-radiated surgical specimens, and it was attributed to the mutagenic action of radiation [67]. However, it should be noted that TP53 variants have also been detected in specimens obtained prior to radiation [18]. The majority of cases report TP53 variants already at first surgery and in both primary tumour and metastases, suggesting an early event in pituitary tumour progression [18] [69]. In addition, a single case of CD due to a microadenoma in a female paediatric patient carrying a germline likely pathogenic TP53 variant with a Li-Fraumeni-like family history has recently been described [33].


ATRX and DAXX

ATRX (alpha thalassemia/mental retardation syndrome X-linked, UniProt P46100) and DAXX (death-domain-associated protein, UniProt Q9UER7) are involved in chromatin remodelling and alternative lengthening of telomeres [70]. Inactivating ATRX and DAXX variants are commonly found in pancreatic neuroendocrine neoplasms (PanNENs) and are mutually exclusive [71]. ATRX and DAXX bind to TP53 and in several cancers, ATRX/DAXX variants can be found concomitantly with TP53 variants. ATRX variants were reported in aggressive and metastatic corticotrophinoma [17] [26] [69] [72]. Their prevalence in metastatic corticotroph tumours is estimated at ~30% [17]. In almost half of the reported cases, ATRX variants were accompanied by TP53 (or TP53 plus PTEN/NF1/2) variants. DAXX variants were found in two USP8 wild type aggressive corticotroph tumours, which also had TP53 variants [16]. In some reports, they were described in both primary tumour and metastasis, while in others, only in metastasis [17] [69] [72]. Strong ATRX immunopositivity was found in the majority of corticotroph tumours, and a lack of nuclear ATRX immunostaining was observed in several cases of corticotroph tumours carrying ATRX variants [17] [72] [73].



Syndromes of endocrine neoplasia

Multiple endocrine neoplasia type 1

This entity (MIM #131100) consists of the association of various endocrine and nonendocrine neoplasms, mainly primary hyperparathyroidism (PHPT, 75–100% of patients), gastrointestinal and PanNENs (41–75%), and PitNETs (30–65%) [74] [75] [76] [77] [78] [79] [80]. Less constant associations include adrenocortical tumours, phaeochromocytomas, bronchopulmonary and thymic NENs, lipomas, angiofibromas, collagenomas, meningiomas, leiomyomas, hibernomas ependymomas, and breast cancer [81]. Half of cases display autosomal dominant (AD) inheritance, of which 70–93% are caused by LOF germline and rarely mosaic, heterozygous MEN1 variants (11q13.1) [79] [80] [82] [83] [84] [85] [86] [87] [88]. The same genetic defect underlies only one-third of sporadic cases, and de novo variants are rare [79] [84] [85] [89]. MEN1 (UniProt O00255–2) is a 610 amino acid predominantly nuclear scaffold protein that regulates gene transcription, genome stability, and cell proliferation, thus acting as a classic tumour suppressor in multiple tissues [90] [91].

A pituitary tumour is the first manifestation of MEN1 in 12–32% of patients, with prolactinomas accounting for two-thirds of these lesions [75] [76] [77] [79] [92] [93] [94] [95]. According to most studies, MEN1 LOF-associated PitNETs present at younger ages, are more often plurihormonal and behave more aggressively than non-MEN1 tumours [92] [93] [95] [96] [97]. Only ̴4% of MEN1-associated pituitary tumours are corticotrophinomas, yet MEN1 LOF is the most common germline genetic cause of CD (2.9% of cases in a paediatric cohort), with at least 34 genotyped cases reported in the literature [33] [75] [78] [79] [92] [93] [94] [95] [98] [99] [100] [101] [102] [103] [104]. Similar to humans, Men1 knockout mice develop pituitary tumours with clear predominance of prolactin and prolactin/GH-expressing tumours [105]. In one study, however, 3% of heterozygous Men1 exon 1–2 deletion mice developed ACTH-expressing pituitary tumours, accounting for 10% of all PitNETs. In contrast, that histotype was not observed in eight previously reported knockout lines [106].

The documented MEN1-associated CD cases include an equal proportion of males and females (n=17 each) [33] [92] [93] [94] [95] [98] [99] [100] [101] [102] [103] [104]. Out of the cases with available data, 61.8% (n=21) were caused by microadenomas, 17.6% (n=6) were due to macroadenomas, and multiple tumours were found in 11.8% (n=4). Of the four patients with multiple tumours, three developed microprolactinomas, and one had a nonfunctioning gonadotroph macroadenoma [99] [101] [103]. A single case of CD caused by a plurihormonal macroadenoma expressing ACTH, prolactin (PRL), GH, and luteinizing hormone (LH) has been published [102]. In most other cases with available data, tumours were monohormonal, displayed no particular imaging or histopathological features and responded well to surgery. CD was diagnosed in childhood or adolescence in 34.4% of cases (n=11) and was the first manifestation of MEN1 in 35.3% (n=12) [33] [92] [93] [98] [100] [103]. Cases presenting during adulthood (n=8 with available data) were diagnosed at age 40.9±10.1 years [93] [94] [99] [101] [102]. Because simplex MEN1 cases have been identified among patients with paediatric CD, genetic testing could be helpful for risk identification, early diagnosis of other disease components, and genetic counselling. Hypercortisolaemia in the setting of MEN1 is usually due to a corticotroph tumour but is caused by primary adrenal disease in one-fifth of cases and is rarely secondary to ACTH secretion from other NENs [101].


Multiple endocrine neoplasia type 4

Caused by germline LOF germline heterozygous CDKN1B (12p13.1) variants, this syndrome (MIM #610755) explains ̴2% of MEN1 cases negative for MEN1 defects [107] [108] [109]. Manifestations highly vary among CDKN1B variant carriers, but AD inheritance with incomplete penetrance is well-proven [110]. The most common component of multiple endocrine neoplasia type 4 (MEN4) is PHPT, although renal angiomyolipomas, adrenal nonfunctioning tumours, uterine fibroids, gastrinomas, gastric carcinomas, gastrointestinal and PanNENs, neuroendocrine cervical carcinomas, bronchial neuroendocrine tumours (NEN), and papillary thyroid carcinoma have also been described [107] [108] [109] [111] [112] [113] [114] [115] [116]. CDKN1B (UniProt P46527) is a negative regulator of multiple cyclin-dependent kinase/cyclin complexes that inhibits the progression from the G1 to the S phase of the cell cycle [117] [118].

Low CDKN1B expression is a common feature of human corticotroph tumours, and Cdkn1b knockout mice develop fully penetrant ACTH-secreting hyperplasia or tumours of the pituitary pars intermedia, but no other PitNETs [119] [120] [121] [122] [123] [124]. In contrast, corticotroph tumours account for only 36.8% (n=7) out of the 19 pituitary tumours associated with germline CDKN1B variants in the literature (reviewed in [125]). Presentation was sporadic in 85.7% (n=6) of these cases, and an equal proportion of cases was explained by tumours<10 mm. Onset in adulthood and additional MEN4-associated tumours were documented in 28.6% of cases (n=2), while 71.4% (n=5) presented with isolated paediatric CD [108] [110] [126]. CDKN1B variants associated with these cases were classified as pathogenic or likely pathogenic, except for two missense defects that were variants of uncertain significance (VUS). Functional analyses rendered mixed results for both VUS; one of them was a previously known germline change in familial isolated pituitary adenoma (FIPA) and familial cancer predisposition, and somatic defect in other neoplasms [127] [128] [129].


Phaeochromocytoma and paraganglioma syndromes

An infrequent phenotype recently termed as “three P association” (3PAs) consists of the association of a pituitary tumour with a phaeochromocytoma or paraganglioma (PPGL) in a single patient or in different members of the same family [130] [131]. Germline defects in the PPGL-associated genes SDHA, SDHB, SDHC, SDHD, and SDHAF2 (namely SDHx genes), explain ̴40% of the genetically tested 3PAs, as well as rare cases of isolated pituitary tumours [33] [132] [133]. These genes encode the four subunits and the assembly factor 2 of succinate dehydrogenase (SDH), a mitochondrial enzyme that participates in the Krebs cycle and in oxidative phosphorylation [134]. SDH dysfunction leads to tumorigenesis via pseudohypoxia, reactive oxygen species, and defective apoptosis and DNA methylation [134] [135] [136] [137].

At least 25 cases of PitNETs associated with germline pathogenic or likely pathogenic SDHx variants have been documented [33] [138]. The affected gene was SDHB in 12, SDHD in six, SDHA in five, and SDHC in two cases. Sixteen tumours were prolactinomas, four were somatotrophinomas, two were nonfunctioning pituitary tumours, one was a metastatic PitNET of gonadotroph origin, one was a corticotrophinoma, and the tumour type was not reported for one case. The presentation was 3PAs in 20 cases (80%) and isolated in the rest. Loss-of-heterozygosity (LOH) was detected in seven out of eleven cases where it was analysed, and two macroprolactinomas due to an SDHB defect occurred in a single family [130]. SDHx-driven pituitary tumours display a particular histopathological pattern characterized by cytoplasmic vacuoles of variable size [130] [139]. Concordantly, Sdhb +/- mice develop hyperplasia of GH and prolactin-producing cells with nuclear inclusions or pseudoinclusions [131]. The nature of these features remains unclear but could be explained by the engulfment of SDHx-associated dysfunctional mitochondria [140].

Four cases of CD presenting with a 3PAs phenotype have been published. One patient carried a pathogenic RET variant, two had negative genetic tests (for SDHx in one case and SDHx, MEN1, and VHL in the other), and one was not tested but had a family history compatible with MEN2A [131] [141] [142] [143]. More recently, a case of isolated paediatric CD (age 8 years) caused by a pathogenic germline SDHA frameshift variant was reported. The tumour was a 7 mm corticotrophinoma with reduced SDHA, SDHB and SDHD immunostaining and LOH at the variant locus. The variant was of maternal origin, but there were no other cases of pituitary tumours or PPGLs in the family [33]. In addition, three germline SDHA VUS were found in three different patients with isolated paediatric CD. One germline SDHD VUS was found in two other paediatric CD patients; one of them had a family history compatible with FIPA, but the cosegregation was not proven [33] [131]. Functional studies are required to assess the pathogenic potential of these VUS.


Carney complex (CNC)

CNC is a rare syndrome with endocrine and cardiocutaneous manifestations [144]. LOF PRKAR1A (coding for protein kinase A, PKA, regulatory subunit-1-alpha 17q24.2, MIM #160980) variants are identified in three-quarters of cases, either as de novo defects or inherited in an AD manner [145] [146]. While one of the most prominent and frequent manifestations of the disease is adrenal Cushing’s syndrome, CNC has only recently been associated with CD [147] [148]. In the pituitary, the cAMP/PKA pathway is involved in cell differentiation and in response to secretagogues and trophic factors in most cell types [149], but alterations most often affect the somatotroph lineage. Single micro or macroadenomas or multiple microadenomas of somatotroph, lactotroph or mammosomatotroph origin, often with surrounding hyperplastic areas, with or without LOH, are characteristic of CNC [150] [151] [152]. PRKAR1A variants have not been detected as somatic changes in sporadic PitNETs [145] [153] [154] [155].

Only three cases of CD in patients with CNC have been reported in the literature, and thus, LOF PRKAR1A variants remain a very rare cause of CD. The three individuals had apparently sporadic presentation but carried germline frameshift PRKAR1A variants and developed additional manifestations of CNC, including PPNAD in one case [156] [157] [158]. The de novo origin of the variant was confirmed in one individual [158]. In one female patient, CD was highly suspected at age three years, but a corticotroph tumour was not proven [156]. The other two patients were males diagnosed with CD at ages 17 and 31 years due to<10 mm corticotrophinomas with LOH at the variant locus [157] [158]. The scarcity of reported cases, however, could be misleading since the diagnosis of CD in this context is challenging, given the high penetrance of PPNAD [159]. Together with McCune-Albright syndrome, MEN1, and, very rarely, MEN type 2 (MEN2, a possible coincidence), CNC could be a cause for “double Cushing’s” [157].


Multiple endocrine neoplasia type 2

GOF variants in RET (10q11.2), coding for a transmembrane receptor tyrosine kinase, are the underlying cause of all MEN2 phenotypes: MEN2A, MEN2B and familial medullary cancer (95–98% of cases) both as inherited or de novo forms [160] [161] [162] [163] [164]. Although pituitary tumours are not considered a component of MEN2, the RET protein, as a dependence receptor, plays an important role in the maintenance of somatotroph cell numbers in normal circumstances [165]. The occurrence of pituitary tumours in germline RET variant carriers has only been documented in four cases in the literature [142] [166] [167] [168], which may represent coincidences. Corticotroph tumours accounted for two of these cases and in both individuals, CD preceded typical MEN2 manifestations. The first patient was a 68-year-old male with MEN2A due to a pathogenic RET variant who also developed a phaeochromocytoma, PHPT and MTC [142]. He had a biochemical diagnosis of CD and achieved remission after TSS, but a corticotroph tumour was neither observed by magnetic resonance imaging nor histologically confirmed. A second surgery was performed 15 years later due to recurrence, again with no histological confirmation [142]. The second patient was a male who carried a pathogenic RET variant and was diagnosed with CD at age 13 years. On examination, coarse facial features, mucosal neuromas, and a marfanoid habitus were noticed, and he ultimately developed MTC. Of note, a corticotrophinoma was not observed by imaging studies and was only confirmed after a third TSS, after which he finally achieved remission [168]. Patients with MEN2 phenotypes without genetic testing or negative for RET variants have also been reported [132]. Other causes of hypercortisolaemia are rare among MEN2 patients, although ectopic ACTH secretion from an MTC may occur [169].



Familial cancer predisposition

DICER1 syndrome

The DICER1 syndrome, DICER1 tumour predisposition syndrome, DICER1 pleuropulmonary blastoma (PPB) familial tumour predisposition syndrome, or PPB familial tumour and dysplasia syndrome (MIM #601200) is an AD entity including MEN and nonendocrine cancer [170]. The characteristic features of the syndrome include dysembryonic tumours that are extremely rare outside this setting, such as PPB, cystic nephroma, ovarian sex cord-stromal tumour, nasal chondromesenchymal hamartoma, ciliary body and cerebral medulloepithelioma, anaplastic kidney sarcoma, pineoblastoma, embryonal rhabdomyosarcoma, and pituitary blastoma (PitB) [170]. Wilms tumour, juvenile hamartomatous intestinal polyps, and pulmonary cysts can also occur, but the most common manifestation of DICER syndrome is thyroid disease, in the form of differentiated thyroid carcinoma, or childhood-onset multinodular goitre. Some of the manifestations have a characteristic age of onset [171] [172].

Around 70% of cases are due to heterozygous germline LOF DICER1 (14q32.13) variants appearing de novo in 10–20% of patients, while somatic mosaic variants have been detected in a further 10% [171] [172] [173] [174]. The latter defects are associated with an extremely rare phenotype known as GLOW syndrome (global developmental delay, lung cysts, overgrowth, and Wilms tumour) [174] [175] [176]. Germline DICER1 variants are most often accompanied by somatic second hits; somatic defects also occur in sporadic tumours [176] [177] [178] [179] [180] [181] [182] [183] [184] [185] [186] [187]. Germline variants are not clustered in hotspots, but mosaic and somatic defects are missense variants that are usually located within the RNase IIIb domain of the protein [170] [176] [184] [188] [189]. DICER1 participates in the processing of mature microRNAs (miRNAs) and small interfering RNAs (siRNAs); its inactivation leads to reduced expression of 5′-derived mature miRNAs, which greatly alters the expression of miRNA targets [187] [189] [190] [191].

PitB, a poorly differentiated ACTH-expressing anterior pituitary neoplasm with an oncofetal molecular signature, has a very low penetrance (<1%) [184] [192] [193]. Most PitBs are diagnosed in neonates or infants as CD or silent pituitary tumours, but two cases have been reported in childhood and young adulthood [194] [195] [196]. Nineteen out of the twenty PitBs so far genotyped were due to LOF DICER1 variants, but it is not clear if any of them were caused by somatic defects [172] [194] [195] [196]. In most cases, PitB was the first manifestation of the syndrome, and nine of these patients died during infancy or childhood [195] [196]. Because PitB is considered a pathognomonic lesion of the DICER1 syndrome, its diagnosis should prompt germline DICER1 screening [172]. These tumours overexpress PRAME, which has been exploited as a target for immunotherapy in other neoplasms [197]. Aside from PitB, a causal association of DICER1 variants with sporadic PitNETs has not been proven [198].


Lynch syndrome

Also known as hereditary non-polyposis colorectal cancer (CRC), this AD syndrome is defined by an increased risk for developing CRC, endometrium, ovary, gastric, small bowel, urinary tract, biliary tract, brain, skin, pancreas, and prostate cancer [199]. Germline heterozygous LOF defects in the DNA mismatch repair (MMR) genes MLH1 (3p22.2, MIM #609310), MSH2 (2p21-p16.3, MIM #120435), MSH6 (2p16.3, #614350), and PMS2 (7p22.1, MIM #614337), usually accompanied by somatic second hits, as well as deletions of the last exons of EPCAM (2p21, MIM #613244, causing MSH2 silencing) underlie this phenotype [200] [201] [202] [203] [204] [205] [206] [207] [208]. The MMR system excises base-base mismatches to increase the fidelity of DNA replication and is involved in the response to mechanisms of DNA damage; its failure leads to increased spontaneous mutagenesis and microsatellite instability [209] [210].

Eight cases of pituitary tumours presenting in carriers of germline MMR defects have been documented in the literature, affecting patients within the fourth and seventh decades of life [133] [211] [212] [213] [214] [215]. These cases include one macroprolactinoma, one aggressive prolactinoma, one aggressive nonfunctioning PitNET, one undifferentiated sellar carcinoma, and three ACTH-producing PitNETs. The tumour type was not specified in one case reported in a register-based study [211]. Four patients had MSH2 variants, one carried an MSH6 variant, another one had both genes affected, and individual cases were due to PMS2 and MLH1 variants. Out of the three ACTH-producing tumours, one was a metastatic PitNET and two were aggressive corticotrophinomas [133] [212] [213]. One of the latter tumours displayed LOH at the variant locus (MSH2), while somatic MEN1 and MSH6 variants were detected in the other case [133] [213]. Interestingly, in one of these patients the corticotroph tumour was the first manifestation of LS [133]. Pituitary tumours seem to be causally associated with MMR defects, and although they remain rare LS-associated neoplasms, their unusually aggressive potential in this context should be kept in mind.



Phakomatoses

Tuberous sclerosis complex (TSC)

The occurrence of hamartomatous lesions affecting brain, skin, heart, lungs, and kidneys, as well as neurological complications such as seizures and behavioural, psychiatric, intellectual, academic, neuropsychological, and psychosocial difficulties characterizes TSC, with a range of tumorous and cutaneous manifestations [216] [217]. TSC is classified as TSC1 (MIM #191100) or TSC2 (MIM #613254), depending on its causative genetic defect: LOF heterozygous variants in TSC1 (9q34.13) or TSC2 (16p13.3), respectively [218] [219]. An AD pattern of inheritance is present in one-third of patients, while the rest are simplex cases due to de novo variants [216]. Germline TSC1 or TSC2 variants are found in around 70–90% of all familial or simplex TSC patients, while germline or somatic mosaic defects cause a small proportion of cases [217] [220] [221] [222]. TSC1 and TSC2 are part of a protein complex that negatively regulates the serine/threonine kinase MTOR and its LOF leads to an overactive MTORC1. [223] [224] [225] [226] [227]. Aside from their role in TSC, somatic defects in these genes have been observed in NENs and urothelial, bladder, and renal cancer [227].

PitNETs have been reported only in eight patients with TSC, while, interestingly, in a rat model with a naturally occurring Tsc2 LOF variant, pituitary tumours are observed with a frequency of 58%, with LOH in one out of three tumours analysed [228]. The human cases include one silent gonadotrophinoma, two somatotrophinomas, one suspected but unconfirmed prolactinoma, and four corticotrophinomas [229] [230] [231] [232] [233] [234]. The CD cases included one young adult (macroadenoma) and three paediatric patients (microadenomas) [33] [232] [233]. Evidence of a pathogenic or likely pathogenic TSC2 variant was available for three cases, and LOH was documented in one out of two cases where it was investigated [33] [233]. Two individuals carrying likely pathogenic variants had no family or personal history of TSC at ages 10 and 15 years. In addition, a pathogenic variant was reported as a somatic change in an 8 mm corticotrophinoma from a paediatric individual with CD [33].


PTEN hamartoma tumour syndrome

The PTEN hamartoma tumour syndrome encompasses a spectrum of phenotypes, including Cowden syndrome, MIM #158350), Lhermitte-Duclos disease, Bannayan-Riley-Ruvalcaba syndrome, PTEN-related Proteus syndrome, and PTEN-related Proteus-like syndrome [235]. Cowden syndrome is a rare AD disorder characterized by dermal trichilemmomas and papillomatous papules, bowel, breast and thyroid hamartomatosis, macrocephaly, and high lifelong risk for breast, endometrial, renal colorectal, and non-medullary thyroid cancer [236].

At least four reports of pituitary tumours in patients with CS exist in the literature [237] [238] [239] [240]. The causative defect was determined in only one case a 29-year-old female patient with a macroprolactinoma who carried a germline pathogenic PTEN variant [239]. Another patient developed a microprolactinoma and a paraganglioma, but genetic testing failed to identify a causative defect in PTEN, SDHB, SDHC, and SDHD [237]. The tumour type was not specified in one individual [238]. The most recent report was of an ACTH-producing metastatic pituitary tumour in a 52-year-old female patient with Cowden syndrome without clinical signs of CD [240]. She was concurrently diagnosed with meningiomas, and had a history of macrocephaly, multiple mucocutaneous lesions, multinodular goitre and a breast fibroadenoma, but no genetic test was conducted in this individual. The clinical association of corticotrophinomas with phakomatoses remains a rare occurrence and additional research is required to assess causality.



Familial isolated corticotrophinoma

The diagnosis of pituitary tumours in two or more members of the same family in the absence of other clinical features is termed FIPA (MIM #102200) and accounts for 2–4% of all PitNETs [241]. Three-quarters of FIPA-associated pituitary tumours are somatotrophinomas, prolactinomas, or mixed GH/prolactin-producing tumours that arise at a younger age compared with sporadic cases [242] [243] [244]. Corticotroph tumours occur in only 7.5% of FIPA families and kindreds with only this type of tumours are extremely rare (1.4%) [245]. The genetic cause of FIPA seems to imply multiple genes and remains unknown in most cases.

Germline LOF AIP (11q13.2) variants underlie 15–20% of all FIPA cases but 25–50% of pituitary tumours in families with exclusively somatotrophinomas, displaying AD inheritance and incomplete penetrance [243] [245] [246]. AIP defects also explain one-third of all cases of gigantism and 6–20% of young-onset pituitary tumours [245] [247] [248]. AIP is a co-chaperone protein involved in pathways relevant for pituitary tumorigenesis, such as cAMP/PKA and RET-dependent apoptosis [249] [250] [251] [252] [253]. Its role in neoplasia is complex and entails both tumour suppression and oncogenesis, depending on the cellular context [254] [255]. Most pituitary tumours associated with AIP defects are sparsely granulated macroadenomas with increased macrophage infiltration, presenting clinically as GH excess, with prolactinomas representing 10% of cases [243] [245] [256]. Five cases of CD associated with germline AIP variants (reference sequence NM_003977.4) have been published, but none of these cases supports the role of AIP in pure corticotroph tumorigenesis [245] [257] [258] [259] [260] [261]. All cases have apparently sporadic presentations. Three variants, p.Lys103Arg in one paediatric case [100], as well as p.Arg304Gln [257] (with allele frequency 0.1%, and four homozygotes in gnomAD v4.0) and p.Arg16His [262] (allele frequency 0.2% and five homozygotes in gnomAD v4.0), found in adults, are VUS or likely benign variants [245]. A fourth variant (p.Lys58Asn, classified as VUS) has been identified in a patient with an apparently multihormonal pituitary tumour secreting both prolactin and ACTH. This patient responded well to dopamine analogue treatment with normalised prolactin and cortisol levels and reduction in tumour size over 14.4 years of follow-up [261]. A fifth case (p.Leu251Argfs*52) had a similar clinical course (prolactin and ACTH staining, response to cabergoline) [258]. Furthermore, a rodent model of biallelic Aip deficiency did not develop corticotroph adenomas, while PIT1 lineage tumours (GH, PRL and thyroid stimulating hormone) were all present [263].

Tandem microduplications including GPR101 (Xq26.3), are the cause of X-linked acrogigantism (MIM #300942), an infrequent form of GH excess with onset in infancy or early childhood [264]. These copy number variants are detected constitutively, or as somatic mosaicism, and although most X-LAG cases occur sporadically, FIPA has been documented [265]. GPR101 defects have not been implicated in CD [266].

Germline heterozygous LOF CDH23 (10q22.1) variants are another reported underlying cause of FIPA (MIM #617540), accounting for one-third of FIPA families and 12% of sporadic pituitary tumours in one cohort [267]. This study reported four rare germline CDH23 VUS in an equal number of individuals with CD, but no experimental validation was performed. This genetic defect requires further confirmation.

Another molecular defect possibly contributing to FIPA was recently proposed, after finding germline missense PAM variants in three first-degree relatives with gigantism and in sporadic patients with functioning pituitary tumours [232]. This gene encodes an enzyme catalysing the C-terminal amidation of secreted peptides; LOF was demonstrated in vitro for multiple variants [268] [269]. Among these cases, two patients with isolated paediatric CD carried a frameshift and a 5′-UTR variant that displayed deleterious effects. A frameshift variant carrier family member displayed elevated midnight serum cortisol but no clinical features of CD. A second study found additional deleterious variants in sporadic patients with functioning PitNETs, including two corticotrophinomas causing cyclical CD [270]. PAM variants seem to result in pituitary hypersecretion, but their role in pituitary tumorigenesis requires further clarification.


Germline defects with no evidence of familial disease

CABLES1 variants

CABLES1 (18q11.2) encodes a negative regulator of cell cycle progression that facilitates the interaction of tyrosine kinases with their substrates, thereby modulating crucial phosphorylation cascades [271] [272] [273]. CABLES1 also stabilizes and prevents the degradation of cell-cycle regulators and interacts with TP53 and TP73 to trigger apoptosis [274]. Inactivation of the mouse orthologue Cables1 promotes tumour formation as well as cell proliferation and survival in vitro [ 275]. Although knockout mice are prone to developing colon and endometrial cancer, Cables1 might not be a cancer driver per se [276] [277]. In corticotroph cells, Cables1 is one of the main transactivated genes in response to glucocorticoids, making it a possible mediator of the regulatory adrenal-pituitary feedback loop [278].

The above-mentioned data, plus the observation that CABLES1 immunostaining was often reduced in human corticotrophinomas, prompted the search for CABLES1 genetic defects [278]. Five germline heterozygous missense CABLES1 variants have been identified in five sporadic corticotroph tumours [33] [279]. In vitro, these variants displayed either reduced capacity to block cell proliferation under dexamethasone treatment (four variants) or reduced half-life (one variant) [33] [279] [280]. Three variants were confirmed to be germline, one with proven inheritance, but there was no somatic LOH in any of these cases. No coexisting hotspot somatic drivers of corticotrophinomas were detected either. Interestingly, all patients (three adults and two children) developed aggressive, either functioning (three cases) or silent (two cases) corticotroph tumours. Further research is required to fully define this recently proposed association.



“Feedback tumours”

In patients with primary adrenal insufficiency, ACTH levels rise due to a lack of adrenal-pituitary negative feedback. This process may, on rare occasions, lead to symptomatic pituitary hyperplasia and even potentially corticotroph tumour formation, as it has been documented in patients with Addison’s disease, the most common cause of primary adrenal insufficiency [281] [282] [283] [284] [285] [286] [287] [288] [289] [290]. While in the past, Addison’s disease was often caused by tuberculosis or Waterhouse–Friderichsen syndrome (adrenal haemorrhage due to severe infection), today, this condition is usually due to destruction of the adrenal cortex by the adaptive immune system, which in most cases is mediated by autoantibodies against CYP21A2 [291]. A complex genetic background underlies this entity, and it is unknown if any of its predisposition loci is also involved in pituitary tumorigenesis. Corticotrophinomas may also occur in patients with monogenic causes of adrenal insufficiency, such as congenital adrenal hyperplasia (CAH) or X-linked adrenal hypoplasia congenita.

Congenital adrenal hyperplasia

This diagnosis includes multiple entities characterized by enzymatic deficiencies involving adrenal steroidogenesis. Over 95% of the cases are due to LOF CYP21A2 (6p21.33, MIM #201910) variants with autosomal recessive inheritance and rarely occurring de novo, while the rest of the cases are due to CYP11B1, HSD3B2, CYP17A1, CYP11A1, POR, and STAR defects [292]. The clinical presentation ranges from the classic salt-wasting and simple-virilising forms in patients with severe CYP21A2 deficiency to the non-classic presentation, characterized by milder degrees of hypocortisolaemia and hyperandrogenaemia. At least four cases of corticotrophinomas presenting in patients with CAH have been documented, all affecting women within the third and fourth decades of life [293] [294] [295]. One patient was homozygous for a CYP21A2 variant associated with non-classic CAH, another one was a compound heterozygous for variants associated with classic and non-classic CAH, and a third one carried a heterozygous variant associated with severe CYP21A2 deficiency. In these three cases, the diagnosis of CAH was established during the diagnostic workup for CD and all had pituitary microadenomas, with histological confirmation in two cases [294] [295]. The patient with no genetic test had a history of ambiguous genitalia and presented with compressive symptoms of a pituitary macroadenoma, confirmed as a corticotrophinoma [293]. The clinical presentation of CD in these patients was characterized by pronounced stigmata of hyperandrogenaemia and variable degrees of hypercortisolaemia, depending on the degree of CYP21A2 deficiency; interestingly, the patient with virilising CAH developed the largest tumour. Notwithstanding these data, it is likely that additional molecular disruptions have contributed to driving corticotroph tumorigenesis in these rare cases.


X-linked adrenal hypoplasia congenita (X-linked AHC)

X-linked AHC (MIM #300200) consists of adrenal hypoplasia with primary adrenal insufficiency associated with hypogonadotropic hypogonadism [296]. Two-thirds of patients develop acute infantile adrenal insufficiency, while in the rest, the adrenal insufficiency appears in childhood and, rarely, in early adulthood [297] [298] [299]. Central hypogonadism is diagnosed most often due to delayed or arrested puberty [300] [301] [302]. Most patients are males carrying hemizygous LOF variants in NR0B1 (Xp21.2); heterozygous females are rarely affected [303] [304] [305]. NR0B1 encodes an orphan nuclear receptor with a role in the development of the adrenal glands and the hypothalamic-pituitary-gonadal axis, mainly via repressing steroidogenesis [306]. A single case of a corticotroph tumour associated with a germline frameshift NR0B1 variant has been reported [307]. This male patient had pre-existing adrenal insufficiency, primary hypothyroidism, and hypogonadotropic hypogonadism. He was diagnosed with an invasive corticotrophinoma causing CD at age 33 years, for which three TSS and radiotherapy were required to achieve remission. Maternal inheritance of the genetic defect was proven, but no other affected family members were identified. The causative role of the primary adrenal insufficiency and, therefore, the theoretically heightened stimulatory effect on the central regulation of ACTH is unclear.



Conclusions

A large proportion of cases of CD are explained by a single somatic driver, while the remaining are due to a heterogeneous genetic background involving multiple germline and somatic defects. Most of our knowledge on the genetic basis of this complex entity derives from research conducted over the last decade. For USP8 variants, most of the available data point towards a relatively benign and treatment-responsive clinical phenotype. In contrast, TP53, ATRX, and likely Lynch syndrome-related gene defects translate into aggressive behaviour. Defining the impact of other individual genetic defects on the pathogenesis of corticotrophinomas remains a pending matter. For instance, it is unclear whether germline defects lead to specific clinical features in CD, other than younger disease onset and the risk of developing other neoplasms in some cases. Further research efforts should focus on developing evidence-based guidelines for genetic testing in CD, as well as on identifying and characterising potential pharmacological targets. For now, the risk of pathogenic/likely pathogenic germline variants is low in patients with corticotrophinomas if no other syndromic manifestation is present. Therefore, germline genetic testing is only suggested in children. Regarding somatic testing, in aggressive macroadenomas, ATRX immunostaining could be informative, but genetic assessment of somatic DNA of the tumour is not recommended outside of clinical research settings in a recent publication, Lin et al. identified TP53 variants or extensive LOH in 9/14 treatment-refractory corticotroph tumours, six of them carrying concomitant ATRX (four) or DAXX (two) alterations [309].



Conflict of Interest

The authors declare that they have no conflict of interest.


Correspondence

Laura C. Hernández-Ramírez, MD, PhD, Associate Researcher C.
Red de Apoyo a la Investigación
Coordinación de la Investigación Científica
Universidad Nacional Autónoma de México e Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán
15 Vasco de Quiroga
Radiation Oncology building, 2nd floor
Belisario Domínguez sección 16
14080 Tlalpan, CDMX
Mexico

Publikationsverlauf

Eingereicht: 22. März 2024
Eingereicht: 16. Mai 2024

Angenommen: 03. Juni 2024

Accepted Manuscript online:
03. Juni 2024

Artikel online veröffentlicht:
03. Juli 2024

© 2024. Thieme. All rights reserved.

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany


Zoom
Fig. 1 Summary of the genetic causes of corticotroph tumours, according to clinical presentation. Altered genes harbour variants in germline DNA unless marked as somatic. CS, Cowden syndrome; FIPA, familial isolated pituitary adenoma; CNC, Carney complex; CAH, congenital adrenal hyperplasia. *Case reports only.