Abstract
Patients with multiple endocrine neoplasia 1 (MEN1) syndrome have a germline mutation in the MEN1 gene. Loss of the wild-type allele can initiate endocrine tumorigenesis. Microscopic and macroscopic pituitary, parathyroid, and pancreatic tumors (referred to as the 3 P's) show loss of the wild-type MEN1 allele up to 100%. In contrast, the duodenal gastrinoma pathogenesis in MEN1 syndrome follows a hyperplasia-to-neoplasia sequence. Gastrinomas have loss of heterozygosity of the MEN1 locus in <50%, and invariably coincide with linear, diffuse, or micronodular gastrin-cell hyperplasia. The factor initiating the gastrin-cell hyperplasia-to-neoplasia sequence is unknown. In this perspective, we argue that hypercalcemia may promote the gastrin-cell hyperplasia-to-neoplasia sequence through the calcium sensing receptor. Hypercalcemia is present in almost all patients with MEN1 syndrome due to parathyroid adenomas. We propose a parathyroid-gut axis, which could well explain why patients with MEN1 syndrome are regularly cured of duodenal gastrinoma after parathyroid surgery, and might cause MEN1 syndrome phenocopies in MEN1-mutation negative individuals with parathyroid adenomas. This perspective on the pathogenesis of the gastrin-cell hyperplasia and neoplasia sequence sheds new light on tumorigenic mechanisms in neuroendocrine tumors and might open up novel areas of gastrinoma research. It may also shift focus in the treatment of MEN1 syndrome-related gastrinoma to biochemical prevention.
Background
The tissue-specific variation of cancer risk can largely be explained by the normal proliferative activity of cells. Every time a cell divides, there is a small risk of acquiring somatic DNA mutations, which might in turn initiate tumorigenesis. The difference in cancer prevalence arising from colorectal mucosa—with very frequent cell divisions—and endocrine cells of pancreatic islets of Langerhans—with rare cell divisions—is considerable (1). However, in certain instances, the odds are stacked against the endocrine cells. Patients with multiple endocrine neoplasia 1 (MEN1) syndrome have a germline inactivating MEN1 gene mutation. Menin, the MEN1 protein product, acts as a tumor suppressor in endocrine cells (2), and inactivation of both alleles can initiate tumorigenesis. Following Knudson's two-hit hypothesis, loss of heterozygosity (LOH) at the MEN1 locus on chromosome 11q13 or somatic mutation of the MEN1 wild-type allele is sufficient to initiate tumorigenesis, drastically increasing the risk of endocrine cancer, also in the pancreatic islets. Of note, the tumor suppressive effect of Menin is well-established in certain MEN1-associated endocrine organs, such as the parathyroid glands and the endocrine pancreas, but its role as tumor suppressor is more controversial in other cell types. For example, Menin may have a dual oncogene activating/tumor suppressive role in breast cancer and prostate cancer (3–6).
Indeed, parathyroid adenomas and pituitary tumors develop in more than 90% and 40% of patients with MEN1 syndrome, respectively (7). Duodenopancreatic neuroendocrine tumors (NET) are seen in virtually all patients with MEN1 syndrome (8). Endocrine tumors arising in patients with MEN1 syndrome often produce hormones causing clinical symptoms. Most patients have hypercalcemia due to excessive parathyroid hormone (PTH) secretion. Some patients may develop hypoglycemic events due to pancreatic insulinoma, or Zollinger Ellison Syndrome (ZES), that is, increased gastric acid secretion due to duodenal gastrinomas.
In accordance with the Knudson theory, patients with MEN1 syndrome harbor loss of the wild-type MEN1 allele by LOH in 100% of parathyroid adenomas (9), 100% of pituitary tumors (10, 11), 83% to 90% of pancreatic NETs and 100% of pancreatic microadenomas (neuroendocrine proliferation <5 mm; refs. 9, 12–14), whereas rare cases without LOH have somatic loss of function mutations of the wild-type allele (10). Remarkably, duodenal gastrinomas in MEN1 syndrome show a very different picture, with LOH of the MEN1 locus in only 21% to 46% of tumors (9, 12, 15). It was also noted that duodenal gastrinomas in MEN1 syndrome were invariably surrounded by areas of linear, nodular, and diffuse enteroendocrine gastrin-cell (G-cell) hyperplasia, none of which had loss of the MEN1 locus at chromosome 11q13 (15). This is intriguing, because it strongly suggests that loss of the wild-type MEN1 allele is not the initiating driver of the G-cell proliferations and hyperplasia, and not a prerequisite for the multifocal neuroendocrine tumors in the duodenum. As a result, a duodenal G-cell hyperplasia to neoplasia sequence was proposed (16). However, if MEN1 does not stack the odds against G-cells, what factor does (17, 18)?
Hypothesis
The nature of endocrine cells is to secrete enough hormones to meet demand. Overstimulation therefore commonly drives endocrine cell proliferation. As proliferation/hyperplasia is associated with an increased risk of mutations, this results in increased tumor formation (1). Indeed, this is the case in the rare glucagon cell hyperplasia and neoplasia (GCHN) syndrome (19, 20), affecting the pancreatic islets. In GCHN, a defective glucagon receptor results in reduced ureagenesis and gluconeogenesis related amino acid uptake in the liver. The resulting hyperaminoacidemia—the stimulating factor—promotes alpha-cell proliferation and hyperplasia, a feedback loop aiming to increase glucagon levels and restore the liver-alpha-cell axis (21). These proliferations increase the chance of acquiring mutations, among others in MEN1, eventually leading to development of multiple pancreatic α-cell tumors (20). Could a similar mechanism play a role in G-cell hyperplasia to neoplasia, especially in the setting of increased vulnerability due to germline heterozygous MEN1 mutations?
G-cells are present in the gastric antrum and duodenum, and are stimulated to release gastrin by luminal peptones or amino acids, parasympathetic vagal nerve reaction to stomach distention, and—notably—extracellular increase of calcium levels (22, 23). Hypercalcemia, commonly seen in MEN1 syndrome due to parathyroid adenomas, might therefore play a key role in initiating G-cell hyperplasia.
G-cells have a calcium-sensing receptor (CaSR; ref. 22), which is also expressed in gastrinomas causing ZES (24). The ability of G-cells to react to calcium has long been exploited to support the diagnosis of ZES after intravenous calcium gluconate injection. Recently, it was shown in a murine model that CaSR in fact functions as a regulator of G-cell growth, and CaSR null mice have low numbers of G-cells (25). Moreover, in human gastric biopsies in the setting of absorptive hypercalciuria, characterized by increased CaSR sensitivity, G-cell hyperplasia was seen in all patients. In addition, nodular gastric ECL-cell hyperplasia and fundic gland polyps were observed in some cases, as often seen in ZES (26).
Almost all MEN1 syndrome patients develop parathyroid adenomas and experience hypercalcemia during their life. Similarly, duodenal G-cell hyperplasia is found in all patients with MEN1 (17). Furthermore, although the symptoms of ZES sometimes precede those of hyperparathyroidism, almost all patients with ZES have hyperparathyroidism (27). Studies have shown that following successful parathyroidectomy in MEN1 syndrome patients with ZES, 20% of patients were biochemically cured of ZES, without removing any duodenal or pancreatic NET (28). In line with this observation, chronic sporadic primary hyperparathyroidism may mimic MEN1 syndrome due to development of ZES or duodenal gastrinomas, thereby fulfilling the clinical criteria for MEN1 syndrome: gastrinomas were found to be common in MEN1 mutation-negative MEN1 syndrome probands, of which 90% had primary hyperparathyroidism (29). In contrast with MEN1 mutation-positive probands, recurrent primary hyperparathyroidism was seen less often (9% vs. 56%).
As calcium is well known to provoke gastrin secretion, parathyroid surgery can cure ZES, and CaSR plays a role in regulating the number of G-cells present in the mucosa, it is plausible that parathyroid disease plays a causal role in G-cell hyperplasia. Following the paradigm that the cancer risk of a particular cell type can in part be explained by the proliferative activity, this would imply that hyperparathyroidism may also increase risk of G-cell neoplasia (gastrinoma). In the setting of MEN1 syndrome, the heterozygous germline MEN1 defect makes G-cells even more prone to neoplastic progression, independent of their initial hormonal stimulus (Fig. 1). However, it is clear that a proportion of duodenal gastrinomas will remain (functionally) dependent of their growth stimulus, as evidenced by the cured patients with ZES after parathyroidectomy (28).
Clinical consequences and research proposals
Currently, indications for parathyroid surgery in MEN1 syndrome include hypercalciuria, nephrolithiasis, symptomatic hypercalcemia, and reduced bone density. Surgery on young patients with MEN1 syndrome with asymptomatic hyperparathyroidism is controversial in view of the risk of recurrence and more complicated resurgery. However, in the light of the possible G-cell hyperplasia-neoplasia sequence in the duodenum, early presymptomatic control of biochemical parameters would be an interesting strategy to reduce the chance of developing malignant gastrinoma. The calcimimetic agent cinacalcet has shown to reduce the hypercalcemia by increasing CaSR sensitivity for extracellular calcium (30). CaSR is also expressed on the parathyroid chief cells secreting PTH, and under normal circumstances, calcium negatively regulates PTH secretion. By allosteric binding, cinacalcet increases calcium sensitivity and reduces PTH secretion, in turn decreasing calcium levels. However, it remains to be studied if increasing CaSR sensitivity might also stimulate G-cell proliferation, comparable with absorptive hypercalciuria (26), and eventually increase the risk of G-cell tumors, especially in the setting of a germline MEN1 mutation. Of note, treatment with cinacalcet has indeed been demonstrated to increase gastrin concentrations and gastric acid secretion in healthy individuals (31).
Unfortunately, no in vivo gastrinoma tumor model is available. The lack of gastrinomas in many murine MEN1 syndrome models has been puzzling, but importantly, MEN1 mice regularly lacked parathyroid adenomas or hypercalcemia (32). Mice with a gastrointestinal epithelium-specific deletion of Men1 might be especially suited to study the effect of cinacalcet or primary hyperparathyroidism on duodenal gastrinoma development (33). Studying human G-cell hyperplasia on biopsies pre- and post-parathyroid surgery would be very challenging, especially as it is yet unclear what the effect of proton pump inhibitors, calcium supplementation, or cinacalcet might have on G-cell proliferation. Nevertheless, supported by extensive literature, we hypothesize that it is plausible that parathyroid adenomas enhance duodenal G-cell and gastrinoma activity. By promoting G-cell proliferation, hyperparathyroidism-induced hypercalcemia may be the missing link in the hyperplasia to neoplasia sequence of gastrinomas in MEN1 syndrome. The proposed parathyroid–gut axis might present novel gastrinoma treatment targets and opportunities. Early biochemical cure of parathyroid disease may decrease disease burden related to G-cell proliferations in patients with MEN1 syndrome.
Authors' Disclosures
No disclosures were reported.
Acknowledgments
This work was supported by a grant from Maag Lever Darm Stichting (no. CDG 14-020).