Summary

Mahvash disease is an exceedingly rare genetic disorder of glucagon signaling characterized by hyperglucagonemia, hyperaminoacidemia, and pancreatic α-cell hyperplasia. Although there is no known definitive treatment, octreotide has been used to decrease systemic glucagon levels. We describe a woman who presented to our medical center after three episodes of small-volume hematemesis. She was found to have hyperglucagonemia and pancreatic hypertrophy with genetically confirmed Mahvash disease and also had evidence of portal hypertension (recurrent portosystemic encephalopathy and variceal hemorrhage) in the absence of cirrhosis. These findings established a diagnosis of portosinusoidal vascular disease, a presinusoidal type of portal hypertension previously known as noncirrhotic portal hypertension. Liver transplantation was followed by normalization of serum glucagon and ammonia levels, reversal of pancreatic hypertrophy, and resolution of recurrent encephalopathy and bleeding varices.

Under normal physiologic conditions, glucagon is secreted by pancreatic α-cells in response to hypoglycemia and hyperaminoacidemia. This process establishes a catabolic state by stimulating gluconeogenesis and glycogen degradation while also inhibiting glycolysis and further glycogen synthesis, ultimately increasing blood glucose.1 Mahvash disease is an exceedingly rare disease characterized by pathogenic variants in the glucagon receptor gene (GCGR) that render hepatocytes unresponsive to glucagon. The loss of this feedback mechanism results in reactive pancreatic α-cell hyperplasia, which often leads to pancreatic neuroendocrine tumors and hyperaminoacidemia.2 Mahvash disease has traditionally been characterized by symptoms driven by pancreatic stimulation. However, here we describe a woman with Mahvash disease who was found to have portosinusoidal vascular disease (previously termed noncirrhotic portal hypertension) with variceal hemorrhage and portosystemic shunting leading to recurrent portosystemic encephalopathy. These findings indicate that portosinusoidal vascular disease may be an underlying feature of Mahvash disease. Liver transplantation resolved the patient’s hyperglucagonemia by replacing defective hepatic glucagon receptors and instilling glucagon sensitivity. Replacing the liver, the site of increased portal-vein resistance, also resolved her intractable recurrent portosystemic encephalopathy and variceal hemorrhage.

Case Report

A 27-year-old woman without a substantial contributory medical history presented to our medical center after three episodes of small-volume hematemesis. She was found to have mild elevations in aminotransferase levels, with imaging showing chronic pancreatitis and possible hepatic cirrhosis without splenomegaly, as suggested by nodularity along the right hepatic lobe. She had no notable history of alcohol use. Findings on esophagogastroduodenoscopy were significant for small esophageal varices and portal hypertensive gastropathy. Endoscopic ultrasonography with needle biopsy of the pancreas was performed, preliminarily revealing a hyperplastic pancreas with multiple subcentimeter cysts and calcifications but no masses. Nadolol was prescribed as secondary prophylaxis to prevent recurrent variceal hemorrhage, and the patient was discharged home.

Final pathological examination of the pancreatic tissue revealed a well-differentiated neuroendocrine tumor, categorized as grade 1 according to the World Health Organization tumor classification criteria (Figure 1A). Positron-emission tomography with gallium-68–dotatate showed diffusely increased uptake throughout the pancreas (Figure 2A), and subsequent transjugular liver biopsy revealed cholestasis with patchy perisinusoidal fibrosis and sinusoidal dilation. There was no evidence of cirrhosis. Serum glucagon levels were considerably elevated (>25,000 pg per milliliter), leading to a presumptive diagnosis of glucagonoma and treatment with monthly injections of octreotide.

FIGURE 1

Pathological Findings.

FIGURE 2

Radiographic Evaluation throughout the Clinical Course.

TABLE 1

Laboratory Data.

In the 12 months after the first presentation, the patient came to our hospital four more times. One of these presentations led to admission to the intensive care unit for profound lethargy in the context of hyperammonemia, ammonia levels ranging from 146 μmol per liter to 354 μmol per liter, and glucagon levels of more than 8000 pg per milliliter (Table 1). At each presentation, the patient was treated with hemodialysis, which resulted in resolution of her symptoms and reduction in (but not normalization of) the ammonia level, but no measurable change in glucagon levels. At the time of one admission, the presence of non-tense ascites was noted in the context of global hypervolemia, and paracentesis revealed a high serum ascites albumin gradient of 1.8 g per deciliter (fluid albumin, 0.4 g per deciliter; serum albumin, 2.2 g per deciliter) and fluid protein level of 1 g per deciliter. There was concern for a paradoxical worsening of hyperammonemia with administration of octreotide, so octreotide therapy was discontinued. The patient was discharged with a plan for as-needed hemodialysis and a protein-restricted diet to control recalcitrant hyperammonemia.

Despite elevated serum glucagon levels, hyperammonemia, and pathological evidence of a pancreatic neuroendocrine tumor, the patient did not have the classic characteristics of glucagonoma syndrome, which include weight loss, necrolytic migratory erythema, diabetes, and mucosal abnormalities. We therefore suspected she had Mahvash disease, a rare genetic disorder of hepatic glucagon receptors that is associated with pancreatic enlargement and pancreatic neuroendocrine tumors.

The patient underwent genetic testing to confirm the diagnosis of Mahvash disease, and germline sequencing of GCGR revealed two variants (NM_000160.5:c.247del and NM_000160.5:c.463del) that predict truncation of the GCGR protein (p.Trp83GlyfsTer61, p.Ala155ProfsTer32). We submitted these variants into the ClinVar database (accession numbers pending). Parental sequencing confirmed that these variants were in trans (i.e., both copies of the GCGR gene carried a truncating variant). Both are null variants in GCGR, and loss of function is an established cause of Mahvash disease. Both variants are absent in the gnomAD v3.1 database of more than 130,000 control alleles, and the patient’s phenotype and biochemistry are highly specific for Mahvash disease, which is a single genetic entity.3 These findings render a classification of the variants as pathogenic according to criteria established by the American College of Medical Genetics and Genomics and the Association for Molecular Pathology criteria.4

Octreotide is used in the treatment of Mahvash disease, but because the use of octreotide in this patient led to recurrent hospitalizations, alternative treatments were considered. Unlike previously described persons with Mahvash disease,5 this patient showed evidence of portal hypertension without cirrhosis — meeting criteria for portosinusoidal vascular disease — accompanied by intractable recurrent portosystemic encephalopathy. Therefore, she was placed on a waiting list for liver transplantation, with Model for End-stage Liver Disease (MELD) exception points because of her genetic diagnosis. (MELD scores determine a patient’s ranking on the waiting list for a liver transplant.) The decision to pursue transplantation was based on previous evidence of the efficacy of liver transplantation in other metabolic disorders.6

While awaiting transplant, the patient was hospitalized multiple times owing to sequelae of portal hypertension, including episodes of hyperammonemic encephalopathy and hematemesis due to gastric varices. On June 11, 2022 (28 months after the initial presentation), she underwent orthotopic liver transplantation (with a liver from a donor after brain death; the donor and the recipient were positive for Epstein–Barr virus, and the donor was negative and the recipient was positive for cytomegalovirus). Examination of the liver explant revealed nonspecific findings, including moderate steatosis and mild fibrosis (stage F1 on a scale ranging from F0 to F4, with higher scores indicating worse fibrosis), without evidence of cirrhosis (Figure 1B). Glucagon levels decreased dramatically after transplantation: 1576 pg per milliliter 2 days after transplantation, 401 at 30 days after transplantation, and 75 (normal) at 270 days after transplantation. Ammonia levels, which had previously remained high despite hemodialysis, also decreased after transplantation and have since remained at normal levels. Within 1 week after transplantation, the pancreas had decreased in size by approximately 30% in both lateral and anterior–posterior dimensions. Follow-up magnetic resonance imaging showed interval decreases in the extent of diffuse pancreatic enlargement (Figure 2B and 2C), indicating a reduction in the pancreatic hyperplasia. Since undergoing liver transplantation, the patient has been free of encephalopathy, and repeat endoscopies performed on days 73 and 290 after transplantation showed no evidence of esophageal or gastric varices.

Discussion

Mahvash disease is an autosomal recessive condition that has been described fewer than a dozen times in the literature. Although reported cases are scarce, the prevalence of Mahvash disease, as estimated on the basis of genomic studies, is as high as four cases per million persons.7 Among the known cases, a number of overlapping features have been reported, including abdominal pain, hyperglucagonemia without glucagonoma syndrome, pancreaticomegaly due to α-cell hyperplasia, hyperaminoacidemia, and the development of pancreatic neuroendocrine tumors.5,7 Although the patient described here has the typical signs of Mahvash disease, the additional findings of portal hypertension and portosystemic encephalopathy make her presentation unique even among rare company.

In Mahvash disease, hepatocytes are rendered unresponsive to glucagon, which results in disruption of the feedback loop between the liver and the pancreas that helps regulate glucose and amino-acid metabolism. This disruption stimulates pancreatic α-cell hyperproliferation and causes pancreaticomegaly, overproduction of glucagon, and ultimately development of pancreatic neuroendocrine tumors.8,9 As a result, the complications of the disease are primarily pancreatic in nature, and the focus of management of the disease has been the pancreas. Although pancreatectomy has been considered, there remain no firm criteria to indicate the conditions under which to pursue this treatment. In addition, somatostatin analogues have been used to suppress hyperglucagonemia, but the efficacy of this approach is not well known.5,10,11 The benefits of sunitinib, everolimus, and peptide-receptor radionuclide therapy in the treatment of patients with Mahvash-related pancreatic neuroendocrine tumors have not yet been established, although their benefits have been shown in the treatment of other syndromes associated with pancreatic neuroendocrine tumors.

Our study underscores the importance of an underlying hepatic pathophysiology in the context of a traditional focus on the pancreatic complications of Mahvash disease. Blocking glucagon signaling results in the downregulation of genes involved in the metabolism of amino acids, leading to hyperaminoacidemia.12,13 We hypothesized that an impaired ability of amino acids to enter the liver and undergo ureagenesis results in the secondary impairment of the urea cycle. This leads to an increased nitrogen load and hyperammonemia — to a level beyond that caused by the primary effect of portosystemic shunting — because the extrahepatic metabolism of high levels of circulating glutamine, mediated by glutaminase, generates glutamic acid and ammonia. There was no evidence of a primary urea-cycle defect, since the ratios among ureagenic amino acids (ornithine, citrulline, and arginine) were normal (Table 1) and the pattern of this patient’s amino-acid profile is similar to that of a pediatric patient with biallelic GCGR mutations.14

Although hyperaminoacidemia is established in mouse models of Mahvash disease, hyperammonemia has not been described previously in persons with Mahvash disease.13,15 Our study provides evidence that hyperammonemia (related to portosystemic shunting with or without derangements in amino acid metabolism) is a feature of Mahvash disease that is accompanied by severely decreased GCGR function. There is a general genotype–phenotype correlation between residual GCGR action and the severity of the clinical presentation of Mahvash disease.5 Patients who retain partial GCGR function tend to present at an older age with less aggressive pancreatic neuroendocrine tumors.16 Patients with lesser GCGR function tend to present at a younger age and with pancreatitis. The patient we describe here probably has no GCGR function, which would explain her uniquely severe phenotype.

In Gcgr-null mice (i.e., mice with complete loss of function of the GCGR), portal pressures have not been measured and abnormal liver histologic findings have not been reported. In fact, the phenotypes of both germline and liver-specific Gcgr knockout mice are similar, with no signs of portal hypertension being described in either type.9 Early in our patient’s course of disease, liver biopsy revealed only nonspecific findings: mild perisinusoidal fibrosis and sinusoidal dilation without cirrhosis. Pathological examination of the explant liver revealed moderate steatosis — thought to be related to the patient’s protein-restricted diet — and an increase in fibrosis, which was mild (stage F1 fibrosis limited to perisinusoidal, portal, and periportal regions) and not associated with obliterative venopathy, incomplete septal fibrosis, or cirrhosis. In the context of clinical evidence of portal hypertension, both initial biopsy and explant findings establish the diagnosis of portosinusoidal vascular disease, which represents a new entity associated with the condition.17 Therefore, it is possible that Gcgr knockout mice — and patients with Mahvash disease — may have as-yet unrecognized or subclinical portal hypertension with or without altered liver parenchyma, a finding that could direct future research.

Owing to our patient’s considerable symptom burden, multiple presentations, and need for intensive care, she underwent evaluation for liver transplantation. Although to our knowledge this procedure had not been previously performed in a patient with Mahvash disease, there were two proposed benefits of transplantation. First, there was a theoretical basis for replacing her native liver with a donor organ that has functional glucagon receptors, permitting normal glucagon signaling. Second, transplantation could potentially reverse the complications of portal hypertension, including persistent hyperammonemia and variceal hemorrhage.

Pancreatectomy was also considered as a second procedure, to take place after liver transplantation. However, our hypothesis that the transplant would result in reversal of the pancreatic hypertrophy (as a result of the reinstatement of natural glucagon signaling) proved to be true: serial imaging in the post-transplantation setting has confirmed a reduction in pancreas size (Figure 2B and 2C). Pancreatectomy was not pursued. At the time of this writing, pancreatic measurements are not normalized. In addition, although the pancreatic masses have persisted approximately 1 year after transplantation, the gradual decrease in size of the massive pancreas suggests that new pancreatic neuroendocrine tumors are unlikely to arise. Whether the existing pancreatic neuroendocrine tumors will ultimately disappear — as has been seen in animal models of Mahvash disease after reconstitution of glucagon receptor function — remains to be determined and necessitates ongoing surveillance.18

A patient with biallelic truncating variants in GCGR resulting in Mahvash disease had unique clinical aspects, including portosinusoidal vascular disease and acute decompensations after the administration of octreotide, characterized by hyperammonemic encephalopathy. These aspects have not been previously noted in reports of Mahvash disease, and although there is no definitive evidence that their cause is syndromic, there is a biologic basis to conclude that they are related to the patient’s underlying disease. It appears that liver transplantation resolved the sequelae of portal hypertension, normalized the levels of both ammonia and glucagon, and partially reversed pancreatic hypertrophy. Further investigation is necessary to elucidate the variable presentations of Mahvash disease and more firmly establish liver transplantation as a potential treatment.