Abstract

Graphical Abstract

Open in a separate window

Graphical Abstract

New Findings About the MEN1–Encoded Protein Menin and its Functional Characterization

Menin, the protein product of the MEN1 gene, consists of 610 amino acids (menin isoform 2, NCBI reference sequence: NM_130799.2). There is a very rare minor isoform of menin (menin isoform 1, NCBI reference sequence: NM_000244.3), from a potential alternative splice site 15 bp downstream of exon-2 that inserts 5 amino acid residues (after amino acid 148). In the gene and protein databases, the norm is to designate the longest transcript as the primary isoform; therefore, sometimes menin is described as a 615 amino acid protein.

All studies of menin and its mutants have been conducted with the 610 amino acid isoform because the rare 615 amino acid isoform has not been observed in any cell types. The 67 kDa menin is widely expressed, and at the C-terminus contains 2 nuclear localization signals (NLSs) (NLS1: 479–497 and NLS2: 588–608) and 1 accessory NLS (aNLS: 546-572) (9, 14). Therefore, menin is detected in the nucleus as shown from experiments using green fluorescent protein–tagged menin, immunofluorescence, and Western blot analysis of subcellular fractions (9, 14). Posttranslational modifications of menin include phosphorylation at Ser394, Thr397, Thr399, Ser487, Ser543, and Ser583, SUMOylation, and palmitoylation that may enhance or suppress menin’s action as a tumor suppressor in the nucleus or its potential association with the cell membrane (15-17). The functional contribution of these posttranslational modifications has not been studied in MEN1-associated endocrine cell types or in a clinical context.

The 3-dimensional (3D) crystal structure of human menin (Protein Data Bank No. 3U84) has been successfully deciphered after the deletion of a single internal loop region that was predicted to be unstructured (amino acid residues 460-519) (18). The 3D structure of menin resembles a “curved left hand,” with a pocket formed by the “thumb” and the “palm.” The structure consists of 4 domains: a long β-hairpin N-terminal domain, a transglutaminase-like domain (“thumb”), a helical domain that contains 3 tetratricopeptide motifs (“palm”), followed by a C-terminal domain (“fingers”) (Fig. 2). The pocket or cavity formed by the “thumb” and the “palm” has been shown to facilitate protein-protein interactions (18, 19).

Open in a separate window

Figure 2.

Three-dimensional crystal structure of human menin alone or together with interacting partners. A, Structure of menin showing a pocket/cavity for protein-protein interaction (Protein Data Bank; PDB ID 3U84). The different domains of menin are color coded: “N-terminus” in pale green (1-101), “thumb” in green-cyan (102-230), “palm” in olive green (231-386), and “fingers” in dark green (387-end). B, Structure of menin interacting with the menin-binding motif (MBM) of JUND (amino acid 27-47, in purple) (PDB ID 3U86). C, Structure of menin interacting with the MBM of MLL1 (amino acid 6-13, in gold) (PDB ID 3U85). D, The ternary complex of menin with interacting regions in LEDGF (amino acid 347-435, in red) and MLL1 (MBM-LEGDF binding motif [LBM], amino acid 6-153, in gold) (PDB ID 3U88). To facilitate crystallization, the following regions were deleted (and not present in the structures shown in this figure): an unstructured loop (amino acid 460-519) in menin, a short fragment (amino acid 40-45) in the JUND-MBM and 2 loop regions (amino acids 16-22 and 36-102) in the MLL1-MBM-LBM. Structural images were generated by using PyMOL (Schrodinger Inc; https://pymol.org).

The tumor-suppressor role of menin in MEN1 and the tissue restricted pattern of MEN1-associated tumors has been replicated in mouse models. Germline homozygous knockout of Men1 (Men1^–/–) is embryonic lethal at E11.5 to E14.5, and germline heterozygous knockout of Men1 (Men1^+/–) generates viable mice that develop (at age > 12-15 months) hormone-hypersecreting tumors in the pancreatic islets (mainly insulinoma), anterior pituitary (mainly prolactinoma), and the parathyroid glands (mainly hyperplasia) (20-24). Consistent with the tumor suppressor role of menin in human MEN1 tumors, a second hit to the nontargeted copy of Men1 resulting in LOH is essential for tumor formation in Men1^+/– mice. The pancreatic islets of Men1^+/– mice show a pretumor stage of hyperplasia and dysplasia prior to LOH at the Men1 locus (25). Investigating the molecular aspects involved in these pretumor events can be helpful to understand tumor initiation and progression from tissue-specific menin haploinsufficiency and menin loss.

Tissue-specificity of the tumor suppressor role of menin has been shown in 2 conditional mouse models. First, mice with conditional loss of menin in the liver (Men1^f/f;ALB-Cre) do not develop tumors in the liver (26). Second, mice with conditional loss of menin in the whole pancreas (Men1^f/f;PDX1-Cre) develop tumors that originate from the β cells of endocrine pancreas (insulinoma), and not from any cells of exocrine pancreas (27). Interestingly, mice with conditional loss of menin in the glucagon-secreting pancreatic islet α cells (Men1^f/f;GLU-Cre) do not develop glucagonomas, instead they predominantly develop β-cell tumors (insulinomas) (28, 29). It is possible that after menin loss, α cells may transdifferentiate into β cells, or paracrine signals from menin-null α cells induce β-cell proliferation (28, 29). As expected, parathyroid-specific Men1-knockout mice (Men1^f/f;PTH-Cre) develop parathyroid hyperplasia and hypercalcemia, and pancreatic islet β-cell–specific Men1-knockout mice (Men1^f/f;RIP-Cre) develop insulinomas (25, 30-32). The Men1^f/f;RIP-Cre mice also develop prolactinomas due to the leaky expression of the RIP-Cre transgene in pituitary lactotroph cells. Similar to human MEN1, the prolactinomas in mice (Men1^+/– or Men1^f/f;RIP-Cre) are frequently observed in females. However, the reason for the sex bias of prolactinomas remains unknown. Although the conditional Men1-knockout mice do not depend on a spontaneous second hit for LOH of Men1, tumors develop 8 to 10 months after embryonic menin loss, and the basis for the delay in tumor formation is not known.

The functional characterization of menin has encountered various challenges due to the lack of any similarity to known proteins, lack of obvious functional motifs/domains, lack of normal or menin-null endocrine cell lines, and lack of ex vivo models of MEN1 tumors or their counterpart normal tissues (organoids or patient-derived xenograft [PDX]). Insights into how menin performs its tumor suppressor activity have been gained from the identification of interacting partners of menin in cell types unrelated to MEN1-associated tissues, followed by validation of some relevant targets in translational studies (33). Even though the sequence of menin does not reveal any functional attributes, direct or indirect interactions with more than 50 different proteins of known function have helped to provide clues about its role in various processes and pathways: cell adhesion, cell cycle progression, cell division, cell motility, cell signaling, cytoskeletal structure, DNA repair, genomic stability, and transcriptional regulation (34). Highly enriched among the interacting partners of menin are transcription factors and epigenetic regulators. Menin serves as a multifunctional protein through prominent functional contributions in transcriptional regulation as a coactivator or corepressor.

Epigenetic Regulation in the Tumorigenesis of MEN1–Associated Endocrine Tumors

DNA is wrapped around a histone octamer with 2 copies each of the 4 core histone proteins (H2A, H2B, H3, and H4), to form a nucleosome, which is the basic unit of chromatin. Epigenetic modifications to DNA and histone proteins can impart a closed or open chromatin structure for controlling access to the transcriptional machinery and to control other processes such as DNA replication and repair. Various epigenetic factors form multiprotein complexes, which may also include long noncoding RNAs, to function as enzymes or cofactors for “writing,” “reading,” or “erasing” the modifications on DNA and histones. DNA modifications include methylation, hydroxymethylation, and further oxidation. Posttranslational modifications of histones known as histone “marks” include methylation, acetylation, phosphorylation, ubiquitination, and other modifications. The precise regulation of these epigenetic modifications and their control mechanisms is essential to prevent abnormal cell proliferation and function that can lead to neoplasia and other conditions. Because epigenetic modifications can be written, read, and erased, they offer a therapeutic opportunity to restore aberrant epigenetic changes to the normal state with drugs that can block or enhance the enzymatic activity or critical interactions of epigenetic regulators.

Epigenetic events in MEN1–associated tumors

Various epigenetic changes have been reported in MEN1-associated tumors (Fig. 3). Evidence for a role of epigenetic regulation in the tumors of MEN1 is supported by the interaction of menin with histone-modifying proteins, particularly with histone lysine methyltransferases (KMTs) MLL1/KMT2A and MLL2/KMT2B in protein complexes that are responsible for writing the histone mark H3K4me3 (40, 41). Histone methyltransferases methylate lysine or arginine residues in the chain of amino acids that protrude from the nucleosome (histone tail). Histone H3 can be methylated on lysine residues at positions 4, 9, 27, 36, and 79 with 1, 2, or 3 methyl groups. Activation or silencing of gene expression is regulated by the level of methylation or demethylation of specific histone H3 lysine residues. H3K4me3 is a mark of actively transcribed genes, and H3K9me3 and H3K27me3 are associated with transcriptional silencing. Specific lysine demethylases (KDMs) can erase monomethylation, demethylation, or trimethylation of H3K4 or H3K9, such as lysine-specific demethylase 1 (LSD1/KDM1A), lysine-specific demethylase 2 (LSD2/KDM1B), and Jumonji AT-rich interacting domain 1A (JARID1A/KDM5A/RBP2).

Open in a separate window

Figure 3.

Schematic representation of epigenetic regulation in multiple endocrine neoplasia type 1 (MEN1)-associated tumors. Epigenetic modification in normal cells is shown on the left. Aberrant epigenetic change observed or predicted after menin loss in tumors is shown in the middle. Consequence of the aberrant epigenetic change is shown on the right. Green and red indicates the nature of the specific histone or DNA epigenetic modification, active or repressive mark of gene expression, respectively. Open black oval with a slanted line indicates loss of that histone mark. Alternative lengthening of telomeres (ALT) is activated in tumors and it is absent in normal cells (indicated by the blue open oval with a slanted line). Not shown is miR-24–mediated epigenetic regulation.

Epigenetic regulation in MEN1-associated islet tumors from H3K4me3 has been explored in mice with targeted β-cell–specific menin loss (Men1^f/f;RIP-Cre). Genome-wide distribution of the gene activation mark H3K4me3 and its counterpart recessive mark H3K27me3 has been examined in the pancreatic islets of 2-month-old Men1^f/f;RIP-Cre and control RIP-Cre mice (75). Immunohistochemistry with anti-H3K4me3 showed no significant change in the global/overall level of H3K4me3 in menin-null islets compared with control islets. In menin-null islets, loss of H3K4me3 correlated with gain of H3K27me3 within a specific set of genes, and the expression of such genes was significantly decreased compared to control islets, particularly the gene encoding insulin-like growth factor 2 messenger RNA binding protein 2 (Igf2bp2). Interestingly, the altered epigenetic marks (loss of H3K4me3 and gain of H3K27me3) and lower expression of Igf2bp2 could be reversed in menin-null islets with simultaneous deletion of the H3K4me3 demethylase Rbp2 (Men1^f/f;Kdm5a^f/f;RIP-Cre) (75). Immunohistochemistry with anti-H3K4me3 showed no significant change in the global/overall level of H3K4me3 in menin-Rbp2-null islets compared with control islets. The mice with β-cell–specific combined loss of menin and Rbp2 showed a decreased rate of islet tumor formation and prolonged survival (62). Therefore, specific epigenetic changes occurring as a consequence of menin loss could be reversed (by removing the Rbp2 histone demethylase), and the restoration of the epigenetic changes to the basal normal state could also reduce tumor formation.

The epigenetic regulation of MEN1-associated islet tumor cell proliferation has been explored for the interaction of menin with an arginine methyltransferase PRMT5 that deposits a repressive histone mark, H4R3me2s (36). In MEFs, menin together with PRMT5 was shown to repress the expression of the Gas1 gene, which is involved in the activation of the Hh signaling pathway. GAS1 is required for binding of the Sonic Hedgehog ligand to its receptor PTCH1 for the stimulation of Hh signaling. Therefore, in tumors with menin loss, GAS1 repression would be released to activate Hh signaling. Treatment of 8-month-old Men1^f/f;RIP-Cre mice with the Hh inhibitor GDC-0449 for 4 weeks reduced the proliferation of islet β cells by approximately 60%. The effect on tumor size and overall survival was not investigated. This study showed that blocking aberrant signaling due to the loss of a menin-dependent epigenetic mark could inhibit cell proliferation.

Another histone modification that has been studied in MEN1-associated islet tumors is histone acetylation that is associated with active transcription. Acetylation marks at lysine residues in histone tails are deposited by histone acetyl transferases (HATs) and read by the bromodomain contained in the bromodomain and extraterminal (BET) proteins. JQ1 is a small-molecule inhibitor of bromodomain interactions with acetylated histones. Treatment of 30-week-old Men1^f/f;RIP-Cre mice with twice-weekly injections of JQ1 for 1 month reduced the proliferation rate of islet β cells in the tumors by 49% to 55% and significantly increased apoptosis (76). The effect on tumor size and overall survival was not assessed in this study. Although the specific epigenetic changes in histone acetylation have not been investigated in MEN1-associated tumors, this study highlights the potential of targeting histone acetyl marks.

A few studies have investigated DNA methylation in MEN1-associated tumors, which is an epigenetic modification of CpG sites, particularly in gene promoter regions. DNA hypermethylation usually coincides with gene silencing. Also, H3K4me3 has been shown to protect CpG islands from DNA methylation to regulate gene transcription (77). DNA methylation can be blocked by directly inhibiting DNA methyltransferases (DNMTs) that establish, propagate, and maintain the stability of the DNA methylation mark. DNA hypermethylation was detected in a subset of MEN1 tumors. Global DNA hypermethylation was detected in parathyroid tumors and nonfunctioning pancreatic neuroendocrine tumors (pNETs) from MEN1 patients (78, 79). Also, promoter hypermethylation was observed as a frequent event in MEN1-associated advanced pNETs (80). Using a somatic gene transfer system in RIP-TVA mice, expression of DNMT1 increased β-cell proliferation, suggesting that DNMT1 could be targeted to inhibit DNA hypermethylation and β-cell proliferation (81).

MEN1-associated pNETs have been assessed for telomere length. Telomeres are specialized chromatin structures that protect chromosome ends. Alternative lengthening of telomeres (ALT) is a telomerase-independent process that is activated in cancer cells to prevent telomere shortening that accompanies normal proliferation of somatic cells. The correlation between ALT and prognosis is variable in different cancer types. Mutations in chromatin remodeling genes death domain-associated protein (DAXX) and α-thalassemia/mental retardation X-linked (ATRX) correlate with ALT activation in sporadic pNETs (82-85). In a study of nonfunctioning pNETs, 48% of sporadic and 25% of MEN1-associated pNETs were ALT positive, and ALT was associated with disease relapse (85).

Noncoding RNA (ncRNA)-mediated gene silencing is another epigenetic mechanism that has been investigated in MEN1-associated tumors. There are 2 types of ncRNAs: the short ncRNAs (fewer than 30 nucleotides) and the long ncRNAs (greater than 200 nucleotides). MicroRNAs are short ncRNAs that regulate gene expression at the transcriptional and posttranscriptional level. In the hyperplastic islets of 8-week-old Men1^f/f;RIP-Cre mice, and human parathyroid tumors, miR-24 (and its immature form miR-24-1) has been shown to target menin because increased miR-24 level correlated with decreased menin level (86-88). This mechanism of silencing menin may also contribute to the second somatic “hit” of MEN1 inactivation in MEN1-associated tumors that do not show LOH at the MEN1 locus (86).

Novel Insights Into the Genetics of MEN1

The current germline or somatic MEN1 genetic testing consists of DNA sequence analysis to screen coding exons and splice junctions for mutations, and multiplex ligation-dependent probe amplification for deletion/duplication (del/dup) analysis to screen for larger alterations. Since its discovery, the availability of genetic testing of the MEN1 gene has become an essential part in the diagnosis and management of MEN1. Genetic screening in MEN1 is informative to confirm the clinical diagnosis, for carrier ascertainment, and for early monitoring for tumors. Also, in families with clinical and genetic MEN1, relatives with a negative MEN1 genetic test can be excluded from the burden of lifelong tumor monitoring.

Clinical practice consensus guidelines developed by a panel of experts including physicians, surgeons, geneticists, and other specialists from international centers outlined recommendations for genetic testing in MEN1 (3). The current guidelines recommend that genetic counseling must be available to patients before and after genetic testing. In terms of who should be tested, the guidelines state that the test should be offered to: 1) an index case with clinical MEN1 (presenting with 2 or more MEN1-associated endocrine tumors), 2) asymptomatic first-degree relatives of an individual with genetic MEN1 (known MEN1-mutation carrier) as early as before age 5 years, 3) symptomatic first-degree relatives of an individual with genetic MEN1 who are presenting with at least one MEN1-associated tumor, and 4) patients with multigland parathyroid disease or parathyroid adenomas before age 30 years, and gastrinoma or multiple pancreatic islet tumors at any age.

Sequencing and del/dup analysis can identify heterozygous MEN1 germline mutations in 70% to 90% of families with typical features of MEN1. A 2015 review of published germline mutations identified 576 unique mutations, and in 2019 the Universal Mutation Database of MEN1 mutations (UMD-MEN1 database) reported an additional 181 unique germline mutations (92, 93). These 757 unique MEN1 germline mutations cover the entire coding region with no hot spots.

The obviously pathogenic category of germline mutations (69%) predicts premature truncation of menin from nonsense mutations (14%), frame-shift mutations (42%), splice site mutations (10.5%), and large deletions (2.5%) (Fig. 4) (92). Missense mutations (25.5%) and in-frame insertion or deletion (indel) of one or more amino acids (5.5%) do not predict obvious inactivation of menin, and whether they are benign or pathogenic needs further investigation (92). Across multiple studies and among family members, no clear genotype-phenotype correlation has emerged from an analysis of mutation types or their location with the clinical manifestations of MEN1 (94). Similarly, somatic MEN1 mutations in sporadic tumors have not revealed any hot spots or a clear genotype-phenotype correlation with specific tumor types.

Open in a separate window

Figure 4.

Percentage distribution of the different types of MEN1 mutations. Obvious inactivation of menin is predicted by nonsense, frameshift, and splice mutations, and large deletions that constitute 69% of the mutations. Missense and in-frame insertion or deletion (indel) mutations are sliced out of the pie chart to indicate the potential of variants of unknown significance among these 2 types of mutations.

Germline MEN1 mutations are not found in 10% to 30% of patients who develop clinical features consistent with MEN1. These MEN1-mutation–negative patients may carry a germline MEN1 mutation in regions that are not interrogated by current genetic testing methods (such as untranslated, intronic, and regulatory regions), or the tumors may show somatic mosaicism (postzygotic MEN1 mutation), or they carry germline mutations in other genes (such as CDKN1B [114]), or the clinical manifestation of multiple tumors is a sporadic coincidence with no underlying germline mutation. Candidate gene analysis, whole-genome sequencing (WGS) or whole-exome sequencing (WES) approaches have been applied to decipher the germline genetic defects in MEN1-mutation–negative cases.

Advances in molecular genetic studies and their applications to genetic diagnosis of MEN1

One of the benefits of MEN1 genetic testing is to confirm the diagnosis of clinical MEN1. However, MEN1 genetic testing is negative in patients with clinical MEN1 who present with incomplete disease manifestations that represent phenocopies, or MEN1-like disease characterized by tumor in as few as 1 of the 3 main MEN1-associated endocrine tissues. The identification of susceptibility genes for endocrine tumor syndromes that have at least one overlapping feature with MEN1 has helped to extend the genetic diagnosis of MEN1 mutation-negative cases to include those additional genes in the genetic testing approach (Fig. 5). Among the 10% to 30% of germline MEN1 mutation–negative cases with or without a family history of clinical MEN1, a few may rarely test positive for germline mutations in genes for MEN1-like conditions (CDKN1B or other CDKI genes, CDC73, CASR, GNA11, AP2S1, GCM2, and AIP) (34, 98, 99). Therefore, the genetic predisposition in clinical MEN1-like cases should be further evaluated by genomic approaches to identify the responsible mutations and genes.

Open in a separate window

Figure 5.

Schematic diagram for a suggested approach to germline genetic screening in MEN1 and MEN1-like disease. MEN1: a patient with 2 or more MEN1-associated endocrine tumors. MEN1-like: a patient with as few as any 1 of the 3 main MEN1-associated endocrine tumors. Clinical MEN1: patients with MEN1 or MEN1-like disease features. Genetic MEN1: germline MEN1 mutation-positive. WES, whole-exome sequencing; WGS, whole-genome sequencing; +ve, genetic test is positive; -ve, genetic test is negative, and +ve*, genetic test is positive as per American College of Medical Genetics and Genomics and Association for Molecular Pathology guidelines.

One approach to find the potentially missing MEN1 mutations in clinical MEN1 cases is to screen the noncoding regions of MEN1 (promoter, introns, and untranslated) that are not part of the current genetic testing methods. In one study, targeted next-generation sequencing of the entire 7.2-kilobase genomic region of MEN1 was performed, and no mutation was detected in 16 of 76 patients. Also, none of the 76 cases had a point mutation or short indel mutation in the noncoding regions of MEN1, indicating that such mutations may be very rare (100).

There is only one study that has performed WGS of constitutional and tumor DNA samples from patients who were mutation negative in prior MEN1 genetic testing (101). Among the 6 patients analyzed, surprisingly, pathogenic MEN1 germline heterozygous mutations were identified in 3 (2 splice-site variants c.1186-2A > G and p.Arg223Arg (CGG > CGC), and a missense variant p.Pro12Leu) that was missed in the prior routine genetic testing. One patient showed a pathogenic germline heterozygous missense mutation in CASR (p.Ile555Val), and one patient had a germline heterozygous deletion on chromosome 1q that included CDC73. In the same study, WGS of tumor DNA samples from 6 other mutation-negative patients did not detect any somatic variants in recurrent genes that may act as tumor suppressors (101). Therefore, the results of this WGS analysis raises the possibility of missing a germline mutation in prior routine genetic testing, perhaps due to older sequencing or variant classification methods. Thus it may be useful to repeat the MEN1 genetic testing of individuals who present with clinical MEN1 for whom finding an MEN1 mutation is highly likely (tumors of parathyroid, pituitary, and endocrine pancreas; or tumors of parathyroid and endocrine pancreas).

Collaborative efforts and other novel approaches can be considered to identify mutations in MEN1 or other genes in the 10% to 30% of mutation-negative cases of clinical MEN1. Blood transcriptome sequencing has been used for the identification of rare-disease genes, which is RNA sequencing analysis of RNA isolated from whole blood samples to detect any evidence of altered transcription as a consequence of DNA variants (102, 103). If the transcript is expressed in blood cells (eg, MEN1), the effect on splicing and expression level can be detected from missense, synonymous, and loss-of-function mutations within the coding exons. Loss-of-function mutations can lead to lower transcript levels through nonsense-mediated decay. Also, blood transcriptome analysis can identify the decreased expression of one allele due to a variant in the regulatory region. When combined with WGS, the corresponding DNA variants in genes of interest can be identified that are responsible for the altered transcription. One of the limitations of this approach is the inability to identify causal genes due to lack of expression in blood cells. Also, this approach may not successfully identify disease susceptibility genes if causal variants do not affect splicing or expression of the gene.

Clinical Course of Genetically Positive and Negative MEN1 Patients

MEN1 is mainly characterized by the occurrence of parathyroid tumors, DP-NETs, pituitary adenomas, and adrenal tumors (3, 104-107). A diagnosis of MEN1 can be made based on clinical, familial, or genetic criteria. Thus, for a clinical diagnosis patients should have 2 or more MEN1-associated tumors; for a familial diagnosis patients should have 1 MEN1-associated tumor plus a first-degree relative with MEN1; and for a genetic diagnosis a germline MEN1 mutation needs to be identified (105, 106). Across all these diagnostic criteria, up to 90% of patients will be found to have a MEN1 mutation. To date, however, no genotype-phenotype correlations have been reported, and even within the same family the tumor types and age of tumor onset can differ significantly (105). Therefore, long-term radiological and biochemical screening should be performed and appropriate treatment undertaken, as it has been reported that early diagnosis of tumors with appropriate interventions can significantly improve patient survival (3, 105, 108, 109). It is important to note, however, that MEN1 mutations can also give rise to familial isolated primary hyperparathyroidism (FIHP) (110). FIHP is characterized by the occurrence of hereditary PHPT without occurrence of other MEN1-associated tumors during a follow-up period of more than 10.4 years, and the development of PHPT at age 51.4 ± 14 years (110). Therefore, a diagnosis of FIHP should also be considered in patients who present with PHPT at an advanced age and show no further MEN1 manifestations after a 10-year follow-up.

In 10% to 30% of MEN1 patients who are diagnosed based on clinical criteria, MEN1 mutations are not identified (106, 111). These are referred to as phenocopies, and are clinically challenging because the manifestations and familial penetrance of the disease are not defined, and therefore the risk of tumor occurrence and subsequently the most appropriate screening protocols are debatable. Furthermore, analysis of a large MEN1 family cohort consisting of 152 members indicated that 10% of individuals in the family who were diagnosed as having MEN1 did not have the MEN1 mutation, thereby indicating that phenocopies can also occur in MEN1 patients with a family history of MEN1. Determining whether an individual is an MEN1 is of particular importance because it has been reported that MEN1-mutation–negative patients have less aggressive disease, with the median age of first tumor manifestation 13 years later, and median age of second manifestation 9 years later compared to MEN1-mutation–positive patients (5). Furthermore, no third manifestations were identified in mutation-negative patients compared to 76 patients in the mutation-positive group, and median survival was 14 years greater in the negative vs positive MEN1-mutation patient groups (age 87 vs 73 years) (106). The disease course for each MEN1-mutation–negative patient is however likely to differ, and subclassification of this group based on genetic analysis will help inform the most appropriate screening and treatment (112). To date, a number of genes have been reported as causing MEN1 phenocopies. These genes include the cyclin-dependent kinase inhibitor 1B (CDKN1B), which encodes the cell cycle regulating tumor suppressor protein p27^Kip1; rearranged during transfection (RET) proto-oncogene, which encodes a receptor tyrosine kinase (RTK) for members of the glial cell line–derived neurotrophic factor family of extracellular signaling molecules; CDC73, which encodes parafibromin, a subunit of the polymerase II-associated factor protein complex that associates with the RNA polymerase II subunit POLR2A and with a histone methyltransferase complex; CASR, which encodes a G-protein coupled receptor; and aryl hydrocarbon receptor-interacting protein (AIP), which encodes a tumor suppressor protein that interacts and colocalizes with the protein kinase A subunits R1-α (PRKAR1A) and C-α (PRKACA) (3, 99, 101, 113).

CDKN1B mutations have been observed in patients with MEN1-associated tumors. These patients are classified as having MEN4, and to date more than 25 cases have been reported (114-116). The most common tumors to arise in MEN4 patients are parathyroid tumors, followed by pituitary adenomas and pNETs, and then occasional adrenal tumors and nonendocrine tumors such as lipomas and meningiomas (114-116). Furthermore, details of the age of onset of these tumors are limited; however, PHPT has been reported in a 15-year-old individual, indicating that, similar to MEN1, tumors may develop in MEN4 patients during childhood or adolescence (117). Although the clinical manifestations appear similar to MEN1, some key differences have been observed; for example, there are currently no reported cases of prolactinomas in MEN4, and the prevalence of DP-NETs in MEN4 patients is only approximately 25%, compared to up to 70% in MEN1 patients (114). In addition to CDKN1B mutations, mutations in other CDK family members including CDKN2B encoding p15^INK4b, CDKN2C encoding p18^INK4c, and CDKN1A encoding p21^Cip1 have also been identified in MEN1 patients (118). The number of reported cases for these mutations is low, and therefore accurate predictions of tumor manifestation is difficult; however, parathyroid, pituitary, and adrenal tumors have been reported, as well as prostate and breast tumors in these patients (118). Moreover, it is predicted that patients with CDKN1B mutations account for only approximately 3% of MEN1-like individuals, with mutations in CDKN2B, CDKN2C, and CDKN1A estimated to account for up to 1%, 0.5%, and 0.5%, respectively (114, 118). Therefore, mutations in the CDK family may occur in only a small proportion of patients defined as being MEN1 phenocopies.

Analysis of patients and families diagnosed initially to have MEN1 based on clinical criteria has also highlighted additional genes that may represent MEN1 phenocopies, including RET, CDC73, CASR, and AIP (3, 99, 102, 113). RET mutations are usually associated with MEN2 (previously MEN2A) and MEN2B (previously MEN3). MEN2 is characterized by the occurrence of medullary thyroid carcinoma (MTC), pheochromocytoma and parathyroid tumors; while in MEN2 parathyroid tumors are rare, and the occurrence of MTC and pheochromocytoma is found in association with a marfanoid habitus, mucosal neuromas, medullated corneal fibers, and intestinal autonomic ganglion dysfunction leading to megacolon (3). However, a patient presenting with Cushing disease due to a corticotrophinoma at age 48 years, who later developed PHPT and was diagnosed clinically to have MEN1, but in whom a MEN1 mutation was not identified, was subsequently found to have MTC and pheochromocytoma at age 66 years, and an RET mutation (116 RET mutation) (119), consistent with a diagnosis of MEN2. MEN1 phenocopies associated with CDC73 mutations, a gene in which mutations usually cause hyperparathyroidism–jaw tumor (HPT-JT) syndrome characterized by parathyroid tumors, ossifying jaw fibromas, renal tumors, and uterine neoplasms (120), have been reported in 2 unrelated patients who had an initial diagnosis of MEN1, but on genetic analysis were shown to have a CDC73 mutation (99, 100); one had PHPT and a prolactinoma (100), and the other had acromegaly, PHPT, and a pNET (100). MEN1 phenocopies associated with CASR mutations, which usually give rise to familial hypocalciuric hypercalcemia type 1 and FIHP (121), have been reported in 2 unrelated patients; one had acromegaly and hypercalcemia possibly due to PHPT (99), and the other had a NET hepatic metastasis from an unknown primary tumor, and PHPT (102). Thus, these patients represent MEN1 phenocopies because they have developed MEN1 tumor manifestations on a background of hereditary disorders associated with parathyroid tumors. MEN1 phenocopy associated with an AIP mutation, which usually gives rise to familial isolated pituitary adenomas, has been reported in a patient with acromegaly and PHPT, thereby illustrating the occurrence of a MEN1 phenocopy in which parathyroid tumor development arose on a background of a hereditary disorder associated with pituitary adenomas (3, 113). These findings indicate that if a patient is diagnosed with MEN1 based on the combined occurrence of PHPT and a pituitary tumor, then testing for MEN1, RET, CDKN1B, CDKN2B, CDKN2C, CDKN1A, CDC73, CASR, or AIP mutations would be advisable to determine if the patient has MEN1, or could potentially have MEN2, MEN4, HPT-JT, FHH, FIHP, or familial isolated pituitary adenoma. In addition, the prevalence of both PHPT and pituitary adenomas is rapidly rising, with PHPT increasing from 76 to 233 per 100 000 women, and 30 to 85 per 100 000 men over the past 2 decades (122), and pituitary tumors identified in more than 25% of unselected autopsies and 20% of the population undergoing intracranial imaging (123). Therefore, the potential of patients to develop coincidental parathyroid and pituitary tumors, and thus meeting the MEN1 clinical criteria, is also increasing. This again highlights the need for detailed germline genetic testing as well as familial investigation of MEN1 patients (112).

Cohorts of patients who have MEN1-like syndrome but tested negative for MEN1, CDKN1A, CDKN1B, CDKN2B, CDKN2C, CDC73, CASR, RET, and AIP mutations have also been reported (100, 102). It is possible that these patients may have mutations in noncoding regions of the MEN1 gene that affect menin expression, for example, in promotor or enhance regions. Sequencing of cDNA may therefore be of benefit to identify, for example, splicing changes. It is, however, also probable that additional, yet unreported genes are involved in the pathogenesis of MEN1 phenocopies. To identify novel MEN-causing genes will likely require genetic analysis of large cohorts of patients (102). This is likely because novel genes will be occurring in less than 20% of clinically diagnosed MEN1 patients. Thus, in summary the clinical course of patients who are MEN1-mutation–negative differs from that of MEN1-mutation–positive patients. Currently, genetic testing for genes including CDKN1A, CDKN1B, CDKN2B, CDKN2C, CDC73, CASR, RET, and AIP may highlight MEN1 phenocopies; however, these still account for only approximately 5% to 10% of MEN1-mutation–negative cases. Therefore, future studies identifying novel genes will be important for determining the treatment and screening of MEN1-mutation–negative MEN1 patients.

Clinical Presentation of MEN1 in Children and Adolescents

Approximately 12% to 17% of MEN1 patients are diagnosed with the disease in the first 2 decades of life (124-126). Clinically evident disease appears uncommon before adolescence, with consensus guidelines currently recommending phenotype screening of confirmed MEN1 carriers commencing by age 5 years (3). Even if penetrance of MEN1 mutations is age dependent, clinical manifestations of MEN1 have occurred in some patients by age 5 years. Therefore, clinical guidelines suggest the performance of genetic testing in asymptomatic relatives of MEN1-mutated patients as soon as possible, certainly within the first decade of life. MEN1 germline-mutation analysis should be recommended in individuals presenting at an early age with a single, apparently sporadic MEN1-associated tumor (3).

Early manifestation of the classical MEN1 endocrine disorders in young patients can be the first manifestation of the syndrome and thus their recognition may help not only the close monitoring of patient’s treatment, but also direct the screening for other endocrinopathies.

The seminal paper on the role of a germline MEN1 mutation causing pituitary adenomas in children and adolescents described prolactin (PRL)-secreting tumors as the most frequent manifestation, with growth hormone (GH) excess being rare and possibly related to GH-releasing hormone–secreting pancreatic tumor (127). The age at diagnosis of MEN1 syndrome varies according to the clinical, familial, or genetic diagnosis. Clinically a functional tumor is diagnosed much earlier than a nonfunctional tumor (128). The familial diagnosis allows the recognition of gene carriers at birth with the possibility of following the natural history of the disease in a given patient. Future tests based on wide genome analysis will allow the identification of gene mutations even in sporadic asymptomatic carriers. Diagnosis in children has to do with the presymptomatic screening of at-risk patients, which allows for earlier detection and intervention, with a resultant decrease in mortality and morbidity associated with these tumors (129). Presymptomatic screening recommendations for MEN1 management have been based on the youngest age at which disease manifestations have been reported, and this cannot be considered a precise optimal timing.

Adverse Fertility, Pregnancy Outcomes, and Impact of Parental MEN1

Women affected by MEN1 typically show classical endocrinopathies during their reproductive years. Some of these endocrine manifestations are known to potentially affect reproductive health. In a report describing a family genotyped by linkage analysis with 2 of the siblings found to be homozygotes, homozygosity did not result in a more severe phenotype, while resulting in unexplained infertility possibly at the time of conception (148). Surprisingly, until very recently there was little published research regarding the impact of MEN1 on pregnancy and a corresponding lack of data to guide antenatal management, except for a few case reports. There are, however, case reports that outline the management of PHPT and sporadic pituitary tumors and DP-NETs in pregnancy in MEN1. Data concerning the impact of MEN1 on fertility and pregnancy is based on the accumulated experience with more common sporadic, single-organ dysfunction, such as isolated PHPT, pituitary tumors, and NETs.

Genetic and Clinical Signatures of Functioning and Nonfunctioning Duodenopancreatic Neuroendocrine Tumors in MEN1 and Their Impacts on Diagnosis and Therapy of These Neoplasia

In MEN1 patients, DP-NETs are highly prevalent and are today the major cause of premature MEN1-related death because of metastasized disease (124, 159, 160). As early as in the first decade of life, a few MEN1 patients have been reported to develop DP-NETs (125). At age 50 years, approximately 50% of patients suffer or have suffered from a DP-NET, rising to a prevalence of almost 90% at older than 80 years (108). Within MEN1 families, there are some indications that tumors arise at an earlier age in patients within successive generations (161). During the lives of patients, multiple functioning as well as nonfunctioning tumors occur throughout the pancreas, which further complicates the management for individual patients (162). Therefore, MEN1 carriers are intensively screened presymptomatically from a young age to enable timely interventions for preventing morbidity and metastasized disease (3). Insights into the clinical and genetic signatures of the different types of DP-NETs and the impact on the management of patients is very important for an adequate management of patients. Because of the complexity, and for well informed and up-to-date decision making tailored to individual patient needs, MEN1 care within an experienced multidisciplinary team dedicated to collaborative research is of the utmost importance (163).

Functioning (hormone-producing) DP-NETs are often diagnosed because of elevation of plasma biochemical markers and the endocrine syndrome that is caused by the hormone produced by the NET. The management of functioning DP-NETs entails both the treatment of the functional syndrome caused by hormonal hypersecretion as well as mitigating the oncologic risk of distant metastases, while minimizing treatment-related morbidity and mortality. This is particularly challenging in gastrinoma, the most frequently occurring functioning DP-NET in MEN1. Gastrinomas occur in approximately 30% of patients with MEN1 (110, 164-168). These tumors produce gastrin, which, if unopposed, leads to acid hypersecretion of the stomach with subsequent severe peptic ulceration and gastrointestinal bleeding, the so-called Zollinger-Ellison syndrome (169). Gastrinomas in MEN1 are often small and in most cases located in the submucosa of the duodenum (170, 171). Proton pump inhibitors are effective for the treatment of peptic ulcer disease and therefore suggested for the treatment of the majority of patients with gastrinomas (3, 172). The prognosis of MEN1-related gastrinomas is hard to interpret because there is a wide variation among studies with regard to diagnosis and treatment (173-178). In a French cohort study, gastrinomas were associated with an increased risk of distant metastases irrespective of tumor size. In this study metastasized gastrinomas were not significantly associated with patient survival (160). However, in a subgroup of MEN1 patients, gastrinomas appeared to have an aggressive course of disease with distant metastases and early death. In this study aggressive tumor behavior was associated with larger tumor size and higher gastrin levels (173). In a recent population-based study the life expectancy of MEN1 patients with a gastrinoma was shown to be reduced. Also, in this study fasting serum gastrin levels were associated with overall survival and could therefore provide a justification for selecting those patients who might benefit from surgery (179). However, since surgery for gastrinomas is often extensive, not always curative, and is associated with morbidity, controversies exist about indication and timing (3, 172, 180).

Interestingly, past Helicobacter pylori exposure was associated with increased prevalence and severity of hypergastrinemia in MEN1 patients (181). Based on these findings, routine H pylori serotyping is suggested in all MEN1 patients, with eradication therapy for those who demonstrate evidence of active infection.

The most frequently occurring functioning pNETs in patients with MEN1 are insulinomas, with an incidence of 10% to 15% (33, 108, 182). Insulinomas lead to symptomatic and potentially life-threatening hypoglycemia. Patients are often young and localization of the insulinoma in the presence of multiple other pNETs is difficult. Therefore, deciding on the type and extent of surgery is complex (183). Although studied only in 6 MEN1 patients with evidence of an insulinoma, the recently developed ⁶⁸Ga-exendin-4 positron emission tomography–computed tomography (PET-CT) scan appears to help guide selective pancreas surgery (184). This functional imaging technique makes use of the glucagon-like peptide-1 (GLP-1) receptor combined with exendin-4, a synthetic GLP-1 analogue. Although different surgical strategies are followed, resection of a MEN1-related insulinoma is often effective but depends on the individual patient’s characteristics (185). The same dilemma applies to the even more seldom occurring functioning pNETs such as glucagonoma, VIPoma, and GHRH-oma, which also lead to symptoms because of overproduction of hormones and can have a poor prognosis (186).

Nonfunctioning pNETs (NF-pNETs) are the most commonly occurring pNETs in patients with MEN1 (133, 187, 188). Therefore, the clinical practice guidelines advise screening for new tumors and monitoring of already existing tumors by plasma biochemical tumor markers combined with radiological examinations or EUS (3). Based on the recent literature, the annual use of tumor markers can no longer be recommended for diagnosing NF-pNET (189-191). The preferred radiological examination is the magnetic resonance imaging scan, not only because of the lower risk of cumulative ionizing radiation exposure but also because of the better sensitivity compared with CT scans (189). EUS is the most sensitive imaging modality; however, it is operator dependent and invasive. Functional imaging using ⁶⁸Ga-DOTA PET-CT seems to be most useful for the detection of metastasis of prevalent NF-pNET (191). Up to now, there has been no agreement on the optimal follow-up schedule and the timing of surgical intervention, which is the only curative option. Given the paucity of high-quality evidence, appraisal of the existing literature still leads to different opinions on the timing and extent of interventions for surgical interventions (192). Among the multidisciplinary teams taking care for patients with MEN1, the management of NF-pNETs is one of the greatest dilemmas. On the one hand, surgery for NF-pNET often leads to major short- and long-term complications that should be taken into account in the decision making before proceeding to surgery (193). On the other hand, up to now, surgery has been the only curative option for NF-pNET. In one small nonrandomized and nonblinded study, the somatostatin analogue (SSA) lanreotide appeared to improve the progression-free survival (PFS) of patients with nonmetastatic NF-pNET of less than 2 cm (194). However, before this can be translated to clinical practice, more robust evidence is needed.

In MEN1-related NF-pNETs, well-informed decision making toward an optimal follow-up and treatment plan requires insight into the natural course and prognostic factors (163). NF-pNET was scarce and often at the risk of bias (195). This emphasizes the caution that is needed before applying these scientific findings to clinical practice.

In the follow-up of patients, the size of pNETs is often used for deciding on the follow-up schedule and when to proceed to surgery. From the available studies it can be concluded that a larger tumor size is associated with a higher chance for metastases and a worse survival (195). In addition, from the studies comparing the effect of surgical resection with a watchful-waiting strategy, it can be concluded that patients with tumors smaller than 2 cm have an overall low absolute risk for metastases and death (196-198). Therefore, a watchful-waiting strategy for those patients seems to be oncologically safe, with surgery not lowering the risk for metastases enough when outweighing the risk for complications.

In daily clinical practice tumor growth over time is often used as a predictor for the course of the disease in individual patients. In the aforementioned review, 2 studies assessing the growth of NF-pNETs were identified (195, 199, 200). Overall, from the current available literature it can be concluded that NF-pNETs smaller than 2 cm have a generally stable course, and a watch-and-wait strategy appears to be safe in those patients (195). However, metastasis of NF-pNETs smaller than 2 cm do seldom occur and reliable predictors for the course of disease in individual patients are urgently needed to enable decision making for individual patients (163).

Several studies were undertaken in MEN1 patients assessing the prognostic value of tumor tissue–based markers as assessed by the pathological examination of surgically resected NF-pNETs (80, 85, 201, 202). From research in non-MEN1–related NETs, it is clear that tumors with a higher World Health Organization grade according to the mitotic index and higher Ki67 labeling index are associated with worse patient survival (203). In only one study, tumor grade was assessed as a predictor for survival of MEN1 patients who were operated because of a NF-pNET (201). A higher tumor grade as assessed by mitotic index, in contrast to Ki 67 labeling index, was associated with a higher risk of liver metastasis. These patients generally underwent surgery for tumors larger than 2 cm.

Epigenetic changes contribute to tumor development. This mechanism is potentially reversible and might therefore be a new therapeutic target. Promotor hypermethylation is reported in sporadically occurring pNETs (204). In MEN1-related pNET, the association between promotor hypermethylation expressed as cumulative methylation index and clinical outcome was also studied. Patients with distant metastasis of an NF-pNET appeared to have a higher cumulative methylation index (80).

Cejas et al showed that within the group of NF-pNETs, a distinction can be made between α-cell tumors expressing the transcription factor ARX and β-cell tumors expressing the transcription factor PDX1 (85). In this study, both transcription factors were assessed by immunohistochemistry. The occurrence of distant metastasis after surgical resection of MEN1-related NF-pNETs that were often larger than 2 cm was strongly associated with the expression of ARX and not PDX1. Interestingly, distant metastasis almost exclusively occurred in cases with ALT within this subgroup of tumors expressing ARX positivity.

Of the currently available markers, the WHO grade of tumors can be used to assess the risk for metastasis. Because grade 2 tumors are at a higher risk for metastasis, after resection of these tumors patients should be carefully followed. In addition, tissue-based prognostic factors such as the transcription factors ARX and PDX1 can be assessed by immunohistochemistry. Furthermore, ALT can be an important addition as a biomarker for predicting the risk for metastasis in individual patients with tumors expressing ARX. New biomarkers such as multianalyte circulating transcriptomas, tissue-based molecular factors, and image-based markers are needed for predicting the course of tumors in individual MEN1 patients (195). In the coming decade the goals for MEN1-related DP-NETs should be to attain the ability to predict aggressive behavior early on in the disease course and develop new therapies to prevent metastatic disease. Where patients with sporadic pNET often present with metastatic disease and research is aimed at devising novel treatment options for advanced disease and prediction of treatment response, in MEN1 there is a window of opportunity to prevent metastatic disease. Since MEN1 is a rare condition and high-quality research is urgently needed, patients should be treated in centers of expertise, in which multidisciplinary dedicated care is combined with collaborative research to search for new insights into more individualized care for patients (163).

Novelties in the Surgical Approaches in MEN1 Endocrine Tumors

Surgical management of MEN1 is complex and controversial, given the multifocal and multiglandular nature of the disease and the high risk of tumor recurrence even after surgical intervention. Establishing the diagnosis of MEN1 before making surgical decisions and referring affected individuals to a surgeon with experience in treating MEN1 can be critical in preventing unnecessary operations or inappropriate surgical approaches.

Treatment for duodenopancreatic neuroendocrine tumors

The timing and extent of surgery for DP-NETs are controversial and depend on many factors, including the severity of symptoms, extent of disease, functional component, location and necessity of simple enucleation, subtotal or total pancreatectomy, and pancreaticoduodenectomy (PD; Whipple procedure). Pancreatoduodenectomy has been associated with higher cure rates and improved overall survival; it also has higher rates of postoperative complications and long-term morbidity (208). The risks and benefits should be carefully considered, and surgical decisions should be made on a case-by-case basis. With regard to open or minimally invasive (laparoscopic or robotic) approaches, minimally invasive pancreatectomy appears to be safe and associated with a shorter length of stay and fewer complications in selected patients (209).

Minimally invasive surgery techniques have evolved remarkably over the past few decades in the field of surgical oncology, including minimally invasive pancreatectomy. Because of the relative simplicity of distal pancreatectomy, minimally invasive distal pancreatectomy has been widely accepted and increasingly performed with reported safety (210). Large retrospective analyses as well as recent systematic reviews have shown that minimally invasive distal pancreatectomy resulted in less blood loss and shorter hospital stays compared with open distal pancreatectomy, and there were no significant differences in incidence rates of short-term complications, including postoperative pancreatic fistula, or in mortality rates or complete gross tumor resection rates between the 2 techniques (211, 212). Alfieri and colleagues reported a large Italian multicenter comparative study comparing laparoscopic and robotic distal pancreatectomy for pNETs. The study included a total of 181 patients (96 robotic and 85 laparoscopic distal pancreatectomies), and the authors reported that both approaches are safe and efficacious for pNET treatment, with a similar conversion rate, postoperative morbidity, and pancreatic fistula rate between groups (213). The robotic approach, reported to have a higher spleen-preservation rate and less blood loss compared with laparoscopic PD, is a technically demanding surgical procedure associated with high mortality and morbidity rates. Techniques for minimally invasive PD have continued to evolve (214, 215), most significantly with the emergence of the robotic surgery platform (216).

The LEOPARD-2 trial was conducted in the Netherlands. This multicenter national study was terminated prematurely after accrual of 99 patients owing to safety concerns, with reported 10% 90-day mortality in the laparoscopic group compared with 2% in the open group (217). The results of the LEOPARD-2 trial clearly demonstrate the safety concerns of laparoscopic PD. Augmented surgical dexterity, particularly the wide range of instrument articulation provided by the robotic surgery platform, may improve the safety and generalizability of robotic PD, and several retrospective cohort studies have reported promising results (218). The accumulating reports support the use of robotic minimally invasive PD (Fig. 6); however, prospective randomized, controlled trials are warranted to determine the safety and noninferior oncologic outcomes of robotic PD compared with open PD for patients with pNETs.

Open in a separate window

Figure 6.

Intraoperative view after robotic pancreaticoduodenectomy of a pancreatic neuroendocrine tumor (pNET) in the head of the pancreas before reconstruction. CHA, common hepatic artery; IVC, inferior vena cava; SMA, superior mesenteric artery; SMV, superior mesenteric vein.

Recent advances in interventional gastroenterology have allowed for new treatments as emerging adjuncts to standard care in patients with pNETs. Multiple applications of different techniques have been demonstrated to locally treat pNETs. The studies are small and none have had large cohorts of MEN1 patients, but several deserve mention. Local therapies include ethanol ablation using 1 to 2 or more cycles of ethanol lavage at various concentrations for local pNET treatment (219, 220). In 2015, Park et al (221) published results from a series of 11 patients with 14 lesions. Ten patients had nonfunctional NETs and 2 patients had symptomatic insulinomas (1 with 3 lesions). Endoscopic ethanol ablation was performed with 98% ethanol over multiple sessions, and patients were followed for 1 year. Postprocedure complete resolution was seen in 8 of 13 lesions (61.5%); both insulinoma patients were asymptomatic. Three patients had mild pancreatitis and required stent placement for PD stricture. All 3 patients with acute pancreatitis had more than 2-mL ethanol lavage in a single procedure. EUS-guided brachytherapy, photodynamic therapy (222), laser ablation therapy (223), CyberKnife frameless radiosurgery, regarded as image-guided radiotherapy, have also been described as emerging technologies.

Radiofrequency ablation (RFA) is perhaps the most common endoscopic treatment considered safe. The RFA technique emits thermal energy resulting in coagulative necrosis of the surrounding tissue. Recent studies have shown promising results with RFA in patients with unresectable pancreatic cancer in the open, laparoscopic, or percutaneous setting. EUS-guided RFA (EUS-RFA) allows real-time imaging of pancreatic tumors and may result in safe tissue ablation. In 2015, Armellini et al (224) demonstrated the safety and feasibility of EUS-RFA in a patient with pNET who refused surgery. Post ablation, the patient remained asymptomatic and the CT after 1 month showed complete radiological ablation—suggesting RFA can be a potential alternative to surgery in select cases. Lakhtakia and Seo (225) and Waung et al (226) reported that EUS-RFA used in 4 cases of insulinoma resulted in complete clinical resolution in all patients (complete morphological resolution in 2 patients) following EUS-RFA with 12 and 10 months of follow-up, respectively.

Benefits of affecting the systemic immunomodulatory response have also been studied with RFA of pancreas lesions. The initial results assessing a systemic response from RFA have proved feasibility for locally advanced pancreatic tumors (not specifically pNETs) with regard to adding the benefit of evidence of immunomodulation. One study group observed a general activation of an adaptive response decrease of immunosuppression. The plausibility of this occurring with pNETs needs further study (227).

Pharmacological Therapies in MEN1-associated Endocrine Tumors

MEN1 patients often have multiple, multifocal tumors that occur at a younger age and have a higher metastatic potential compared to the single sporadic counterpart tumors (104, 233). This means that surgery may not be possible, especially in the case of pNETs. In addition, with the increase in screening and subsequent earlier diagnosis, pharmacological treatments may also be administered to delay surgical intervention, or to control symptoms in functioning tumors (3). Selecting the optimal pharmacological therapy is however challenging because clinical trials are often not undertaken in MEN1 patients, and instead results are extrapolated from those undertaken in patients with a single endocrine tumor, or based on descriptive, small-cohort studies. Nevertheless, there are a number of pharmacological, or medical, treatments available for MEN1-associated tumors, that can be broadly divided into biotherapies and chemotherapies.

Biotherapies are agents that target specific tumor-associated receptors, or signaling pathways. For NETs, these predominately include agents that target somatostatin receptor (SSTR), mechanistic target of rapamycin (mTOR), and receptor tyrosine kinase (RTK) signaling (109). The first-line pharmacological treatment for MEN1-associated NETs is commonly a SSA (234). SSAs bind to SSTRs, a family of G protein–coupled receptors consisting of SSTR1-5, to activate their downstream signaling pathways (235). MEN1-associated NETs can express all 5 SSTRs; however, they most commonly express SSTR2 and SSTR5 (234, 236). Targeting SSTR signaling, which affects antisecretory and antiproliferative pathways, is effective in facilitating symptom control due to hormone hypersecretion and in reducing tumor burden (104, 234, 235). Three SSAs, octreotide, lanreotide, and pasireotide, have been used clinically for treating NETs. Octreotide and lanreotide, which primarily bind to SSTR2, have been demonstrated to have efficacy in pituitary and gastroenteropancreatic tumors NETs (237-242). Specific evidence for the efficacy of octreotide and lanreotide in MEN1 patients has also been demonstrated in a study of 5 patients with gastroenteropancreatic NETs associated with hypergastrinemia, in whom 3 months’ treatment reduced gastrin hypersecretion, and reduced the size of liver metastases (243); a retrospective study of 40 individuals with MEN1-associated DP-NETs, in whom 12 to 15 months’ octreotide treatment resulted in tumor response in 10%, and stable disease in 80% of patients (244); in a longitudinal, open-label study of 8 MEN1 patients with pNETs smaller than 2 cm, in whom octreotide treatment resulted in a decrease in GEP hormones and stable disease in approximately 80% of patients (245); and in a prospective observational study, in which lanreotide treatment in 23 MEN1 patients with pNETs smaller than 2 cm significantly improved PFS (203). Pasireotide, which binds to SSTRs 1–3 and 5, has been reported to be effective in treating pituitary adenomas and pNETs (246-249). Thus, pasireotide has been reported to ameliorate hypoglycemia in insulinoma patients, and to be effective in patients who are non-responsive to octreotide and lanreotide (246, 247, 250-252). However, clinical trials have indicated that the efficacy of pasireotide is not significantly superior to octreotide and lanreotide in controlling disease progression, and that it is associated with a higher frequency of hyperglycemia that requires medical intervention, thereby limiting its use (242, 246, 249). Furthermore, due to the lack of detailed clinical trials in MEN1 patients the evidence for the use of pasireotide for MEN1-associated tumors is currently unclear. SSAs have also been used in peptide receptor radionuclide therapy (PRRT), whereby they are labeled with a β-emitting radionuclide, for example ¹⁷⁷lutetium (253). One prospective randomized study indicated an increase in PFS in PRRT vs SSA alone in metastatic midgut NETs, and currently PRRT is recommended as a third-line treatment for metastatic midgut and pNETs (111). Small retrospective case series have also highlighted the benefit of PRRT for metastatic MEN1-associated tumors (254, 255).

In addition to SSTRs, other receptors can be targeted for NET treatment. For example, NETs are often highly vascularized and express RTKs including vascular endothelial growth factor receptor (VEGFR), insulin-like growth factor 1 receptor, and platelet-derived growth factor receptor (104, 109). The RTK inhibitors sunitinib and pazopanib have been reported to increase PFS from approximately 5.5 to 11.4 months in patients with pNETs (107, 256). However, the evidence of the efficacy of these inhibitors in MEN1 patients is inconclusive, as only one study specifically examined sunitinib treatment in MEN1 patients, and no improvement in PFS was observed in the 2 patients included in the study (257). In addition to SSTRs and RTKs, pituitary tumors can also express G protein–coupled dopaminergic type 2 receptors, and dopamine agonists (eg, cabergoline) are well known to lower PRL levels and decrease tumor size in prolactinomas in non-MEN1 and MEN1 patients (258). As well as receptors, molecules targeting downstream signaling pathways have also been shown to have efficacy in NETs, including inhibitors of mTOR signaling (83). Everolimus, an mTOR inhibitor approved for the treatment of advanced NETs, has been demonstrated to increase PFS of metastatic pancreatic and lung NET patients from approximately 4 to approximately 11 months (259, 260). Evidence for everolimus for the treatment of locally advanced or metastatic pNETs in MEN1 patients has also been demonstrated in a retrospective study of 6 patients in which PFS was higher in individuals with MEN1 mutations (33.1 months) compared to those with sporadic non-MEN1 disease (12.3 months) (257, 261).

Finally, surgical removal of the multiple parathyroid tumors that occur in patients with MEN1 is the definitive treatment (3). However, in patients in whom surgery has either failed or is contraindicated, drugs acting as CaSR agonists (calcimimetics, eg, cinacalcet) may be successfully used to treat the hypercalcemia and PHPT (262). Thus, in 8 patients up to 30 mg twice daily cinacalcet treatment has been reported to reduce median serum calcium by 0.35 mmol/L, and decrease serum parathyroid hormone by a median of 5.05 pmol/L, with no change in urinary calcium; none of the patients demonstrated recurrent stone formation (262).

Chemotherapies, including alkylating agents, antimicrotubule agents, topoisomerase inhibitors, antimetabolites, and cytotoxic antibodies, have been used to treat NETs. These are, however, reserved for pancreatic, thymic, and bronchopulmonary NET patients with metastatic disease, high tumor burden, or high proliferative index (3, 109). Commonly used agents include the alkylating agents streptozocin and temozolomide, the antimicrotubule agent docetaxel, the topoisomerase inhibitor doxorubicin, and the antimetabolites capecitabine and gemcitabine, either alone or in combination (3, 109). Thus, for high-grade, poorly differentiated pNETs, it has been reported that streptozocin-based chemotherapy is effective (109); however, the benefits for non-pNETs or for MEN1-associated NETs is unclear because of the lack of detailed clinical trials in these patients.

Over the last few decades the increased knowledge of menin function and the development of robust MEN1 tumor models has led to the preclinical evaluation of a number of novel pharmacological therapies specifically targeting MEN1-associated tumor pathways. This includes gene replacement therapies, epigenetic targeting therapies, Wnt pathway inhibitors, and novel mTOR and RTK inhibitors. Furthermore, these and existing pharmacological agents have also been evaluated for their efficacy in chemoprevention. These preclinical studies are discussed subsequently and are shown in Fig. 7.

Open in a separate window

Figure 7.

Emerging therapies for MEN1. Menin, encoded by the MEN1 gene, has roles in multiple pathways associated with cell proliferation. These can be targeted by emerging compounds, including receptor tyrosin kinase (RTKs) inhibitors, novel mechanistic target of rapamycin (mTOR) inhibitors, β-catenin antagonists, epigenetic modulators, and thrombospondin analogues. In addition, preclinical studies indicate MEN1 gene replacement may have efficacy in MEN1 patients, and somatostatin (SST) analogues may have chemopreventive efficacy.

Menin is a tumor suppressor, and loss of menin expression or function results in NET tumor formation. Therefore, MEN1 gene replacement therapy has the potential as a therapy for all MEN1-associated tumors. Thus, injection of a recombinant nonreplicating adenoviral serotype 5 vector containing Men1 cDNA under the control of a cytomegalovirus promoter into pituitary tumors of a conventional heterozygous Men1 knockout mouse (Men1^+/–) resulted in increased menin expression, and decreased proliferation without inducing an immune response (263). Furthermore, systemic delivery of a hybrid adeno-associated virus and phage vector that displays active octreotide on the viral surface that allows targeted delivery of a tumor necrosis factor transgene, significantly reduced the size of pNETs in a conditional Men1 pancreatic knockout mouse model (Men1^f/f;RIP-Cre) (264). Despite the promise of gene therapy representing a potentially curative treatment for MEN1-associated NETs, its translation into clinical trials may be complex because challenges around delivery, specificity, and off-target effects still need to be clarified.

One of the key functions of menin is the regulation of gene transcription through epigenetic mechanisms, including the modification of histones via methyltransferase and deacetylase complexes (described earlier). The use of epigenetic-targeting compounds may therefore have utility in treating MEN1-associated tumors. Preclinical studies have shown that an inhibitor, JQ1, of the BET protein family that binds to acetylated histone residues to promote gene transcription may have efficacy in MEN1-associated pNETs as it significantly reduced proliferation and increased apoptosis of pNETs in a conditional Men1 knockout mouse model (76). Furthermore, JQ1 has also been shown to reduce secretion of ACTH from the pituitary tumor cell line, AtT20 (265). In addition, another BET inhibitor, CPI203, was able to significantly reduce proliferation of a BON-1 xenograft model, further indicating the likely utility of BET inhibition for pNETs (266). The utility of histone deacetylases inhibitors (HDACis) has also been evaluated in preclinical NET models. The HDAC5 inhibitor LMK-235 significantly reduced proliferation and increased apoptosis of pNET cell lines (267), and etinostat, an HDAC1/3 inhibitor, could significantly inhibit the expression of master regulator proteins of metastatic gastroenteropancreatic NETs, as well as reduce tumor growth of the midgut H-STS cell line xenograft mouse model (268). In addition, the HDACi SAHA has been reported to significantly decrease proliferation, and increase apoptosis of a growth hormone and PRL-secreting GH3 rat pituitary tumor cell line, and to decrease cell viability and ACTH secretion from AtT20 and human-derived corticotroph tumor cells (269, 270). Similarly, the HDACi trichostatin also significantly decreased proliferation and inhibited ACTH production in AtT20 cells (271). A number of BET and HDACis are already in clinical trials, and thus these may offer a promising novel therapeutic approach for MEN1-associated NETs.

Menin has also been reported to promote β-catenin phosphorylation, resulting in inhibition of Wnt signaling; therefore, when menin function is lost β-catenin accumulates in the nucleus and promotes gene transcription (104). The β-catenin inhibitor PRI-724 has been shown to significantly decrease pNET cell line viability (272), and a preclinical study in conditional Men1 knockout mice indicated that the β-catenin antagonist PKF115-584 decreased the number and size of pNETs, as well as increased overall mouse survival (63). PRI-724 is already in clinical trials for other conditions (272), and therefore this may provide a novel pharmacological agent for assessment in MEN1-NETs.

RTK and mTOR inhibitors are already in clinical use, although they may have limited efficacy and patients often stop responding to these inhibitors. Therefore, a number of preclinical studies have investigated different approaches for targeting RTK and mTOR signaling pathways. This includes targeting angiogenic pathways with an anti-VEGF (the cytokine that binds to the RTK, VEGFR) monoclonal antibody. However, studies in a RipTag2 insulinoma mouse model indicated that although these compounds reduced tumor burden, they increased invasiveness and metastasis (273). This increased invasiveness may be overcome by targeting multiple RTK pathways simultaneously. For example, concurrent inhibition of VEGF and cMET using sunitibib and crizotinib, respectively, reduced invasion and metastasis in the RipTag2 pNET model (274). Furthermore, the VEGF, PDGF, and fibroblast growth factor targeting the small kinase inhibitor nindetanib has been reported to decrease tumor growth and prolong survival of Rip1Tag2 insulinomas without increasing local invasiveness or metastatic spread (275). In addition, inhibition of nitric oxide synthase may also prove a novel angiogenic-targeting pharmacological approach, as the nitric oxide synthase inhibitor L-NAME could increase constriction of tumor-supplying arterioles in pNETs of a conventional Men1 knockout mouse model (276). Also, functionally active angiogenic peptides, including for thrompospondin 1 have been reported to suppress angiogenesis and tumor growth in RipTag2 pNETs, and as menin interacts with SMAD3, which is downstream of thrompospondin 1 receptor signaling, this may also provide a menin-targeted therapy for MEN1 patients (104, 277). Finally, novel mTOR inhibitors are being investigated; for example, sapanisertib may provide a novel pharmacological agent for everolimus-resistant pNETs. Thus, sapanisertib caused tumor shrinkage of MEN1-mutant patient xenografts that were implanted in female athymic nude mice, and these included tumors that were nonresponsive to everolimus (278).

Pharmacological agents have efficacy in reducing proliferation and hormone secretion; however, they may also have utility for chemoprevention, which would be of particular importance in MEN1 patients, to prevent or delay tumor occurrence. SSAs may have a role in chemoprevention as paseriotide treatment of a conventional Men1 knockout mouse model and a Pdx1-Cre Men1 knockout model (Men1^f/f;Pdx1-Cre) significantly reduced pNET occurrence (279, 280). Furthermore, lanreotide treatment of a conventional Men1 knockout mouse model also significantly decreased the number of new pNETs, and pNET size compared to vehicle-treated mice (281). In addition, clinical trials indicate that patients treated with octreotide have stable disease (244, 245). Therefore, the administration of SSAs to familial or genetically diagnosed MEN1 patients may provide a novel approach for preventing or controlling tumor development.

Preclinical studies of novel agents targeting menin specific pathways that are perturbed on menin loss in tumors may provide a wealth of pharmacological approaches for the treatment of MEN1-associated tumors. These will, however, need to be carefully evaluated in prospective clinical trials specifically recruiting MEN1 patients.

Prognosis and Quality of Life

MEN1 is linked to a higher mortality when compared to a normal population or to nonaffected members in MEN1 families, with DP-NETs and thymic NETs representing the main cause (70%) of death (3, 282). Even if uncommon, other tumors such as adrenal carcinoma and parathyroid carcinoma are fatal (13, 283). A predisposition to breast cancer, mainly of the luminal type, has been reported in female MEN1 patients at a mean age of 45 years, earlier than in the general population, thereby prompting the recommendation for breast cancer screening in MEN1 patients around age 40 years (284).

In females with MEN1, the mortality is lower than in males with MEN1, probably due to the lower incidence of thymic NETs in females. Moreover, sporadic MEN1 cases show higher mortality than familial cases. All these observations affect the surveillance strategies for MEN1 patients.

The publication of the last guidance paper (3) has certainly contributed to the decrease in percentage of deaths related to MEN1 tumors, with a significant improvement in the management of MEN1 patients.

The oncological nature of MEN1, the multiplicity of tumors, the aggressiveness of some of the neoplasms, the lack of solid data on prognosis and of phenotype genotype correlations, increase the potential for a considerable influence of the disease in the health-related quality of life (HRQoL) of the affected patients. Recent studies support this hypothesis, showing a role of the disease and its management in the lower QoL scores in adults with MEN1 (285-288). Important variables that influence HRQoL in MEN1 patients are the possibility of being referred to a dedicated multidisciplinary center and being supported by a dedicated patient advocacy group.

Unfortunately, specific questionnaires measuring HRQoL for MEN1 or other hereditary cancers have not yet been developed, contributing to an underestimation or overestimation of specific traits of the syndrome

Conclusions and Future Prospects

Considerable progress has been made in the past decade to study the natural history, diagnosis, and management of MEN1, and in basic and preclinical research to understand the pathophysiology of MEN1-associated tumors. This review has outlined the major advances since the publication of the most recent MEN1 guidance paper (3). The combined efforts of basic researchers and clinicians have been instrumental to alleviate the morbidity and mortality of affected individuals. National registries and databases have been developed that can facilitate access to relatively large well-characterized patient populations for diagnostic and therapeutic developments. Patient associations and federations have been developed worldwide that can reinforce future screening or intervention programs. In parallel, centers for excellence are now being recognized and these represent quality contacts for reliable information and clinical care for any type of patients’ needs. Basic research on the characterization of menin and mouse models on Men1 loss has revealed protein interactions and pathways affected in tumors that can be targeted for potential therapeutic options.

However, many important questions are still unanswered, and appropriate and timely recommendations must be offered to the clinicians involved in the care of MEN1 patients.

MEN1 is a complex disorder predisposing to more than 20 benign and malignant endocrine and nonendocrine neoplasms. The multidisciplinary team involved in the clinical care of these patients should represent experts in endocrinology, radiology, gastroenterology, surgery, and genetics working in tight connection. This is an essential requirement for the best care of this difficult disease.

Periodic reassessment of the biomedical literature and raw genetic data by experts is encouraged to help to identify new genes or as yet unrecognized syndromes in individuals with specific syndromic features but who lack known genetic susceptibility mutations. Further work is needed to identify epigenetic or modifying factors that may explain even rare associations with uncommon tumors for MEN1. Biobanks of MEN1 biological materials should be established.

Preclinical and clinical evidences point to the efficacy of potential pharmacological treatments for MEN1 tumors. These promising data are a clear indication for the development of multicenter clinical trials nationally and internationally. These studies can benefit from the development of liquid biopsy assays that could predict specific therapeutic options for patients and efficacy of the therapy.

Additionally, the design of a dedicated HRQoL questionnaire will represent a necessary step to increase the analytical effectiveness of any given treatment in MEN1. Advocacy associations will be instrumental to carry on these studies.

Acknowledgments

Glossary

References