Dual phenotype in a patient with adult-onset hyperinsulinism and diabetes caused by an activating MAFA variant

Abstract

Context

The coexistence of diabetes and hyperinsulinemic hypoglycemia (HI) within a family is rare. Activating MAFA variants have been described in 3 families with this dual phenotype and are associated with sex-dependent clinical patterns.

Methods

We studied a family where members were affected by HI (n = 5), diabetes (n = 5), or both conditions (n = 1). Clinical, biochemical, and imaging data were collected. Genetic testing for variants in known monogenic diabetes and HI genes was performed in the proband, followed by cascade testing in available relatives. Pancreatic tissue was examined by immunohistochemistry and immunofluorescence.

Results

A heterozygous, p.(Ser64Phe) MAFA variant was identified in the male proband, who had presented with diabetes at 28 years and developed HI due to insulinomatosis 6 years later. The variant was confirmed in 5 relatives, 4 with HI and 1 with diabetes (including obligate heterozygotes). Among the 5 presenting with HI, 3 were female, whereas 4 of 6 presenting with diabetes were male, consistent with reported sex-dependent patterns.

Increased nuclear expression of the transcription factor MafA (MAFA) was detected by immunofluorescence in insulin-positive tumor regions compared to the adjacent islets, supporting a pathogenic role for the variant in beta-cell regulation.

Conclusion

We describe the fourth family with a pathogenic MAFA variant and the second individual with the dual phenotype of diabetes and HI. Our findings highlight challenges in clinical management of this condition and underscore the need to consider MAFA in cases of adult-onset HI or those with an atypical course of diabetes especially when there is family history.

The combination of diabetes mellitus and hyperinsulinemic hypoglycemia (HI) within an individual is paradoxical and extremely rare in clinical practice. This dual phenotype has been described in some rare monogenic conditions, such as ABCC8- and HNF4A-related congenital HI, where dominant loss-of-function variants initially cause HI in infancy that typically remits during childhood or adolescence, but can later manifest as diabetes in adulthood [1-4].

Another rare example of pathogenic variants associated with both HI and diabetes involves dominant activating variants in the MAFA gene [5]. To date, 2 such variants, p.(Ser64Phe) and p.(Thr57Arg), have been described in 3 families, where they cosegregated with HI or diabetes among affected family members [5-7]. Unlike ABCC8-HI and HNF4A-HI, MAFA-HI is an adult-onset condition. Notably, in one of the previously reported cases with MAFA, HI and diabetes occurred sequentially; the female patient was first diagnosed with gestational diabetes at the age 27, which progressed to impaired glucose tolerance between ages of 33 and 39. She then developed HI at age 55 [5].

In cases with MAFA variants, HI arises secondary to insulinomatosis, with a mean age of diagnosis of 39 years [5]. The HI phenotype is more frequent in females, whereas diabetes is more commonly observed in males [5]. This sex-dependent pattern has been recapitulated in a mouse model, where premature aging and accelerated senescence of male Mafa^S64F beta-cells contributed to diabetes development [8]. Moreover, in 4 of the reported individuals with a MAFA variant, congenital cataracts and/or congenital glaucoma were reported in addition to diabetes [5].

The MAFA gene encodes the V-Maf avian musculoaponeurotic fibrosarcoma oncogene homolog A protein (MaFA), which functions both as an oncogene and as a transcription factor in pancreatic beta-cells [9-11]. MAFA regulates the transcription of multiple genes involved in glucose-stimulated insulin secretion, including the insulin (INS) gene [12]. Both the p.(Ser64Phe) and p.(Thr57Arg) variants affect highly conserved residues within MAFA's N-terminal transactivation domain, impairing phosphorylation. This disruption increases protein stability and enhances its transactivation potential [5]. Insulinomatosis in individuals with MAFA activating variants is thought to result from elevated protein levels and enhanced activity, which drives the expression of genes involved in cell cycle regulation, including CCND2, a key regulator of beta-cell proliferation [13].

In the present study, we report the fourth family with a MAFA pathogenic variant causing adult-onset hyperinsulinism and diabetes, including one family member who exhibited both phenotypes.

Methods

Clinical features

The male proband presented at the age 28 years with symptoms suggestive of diabetes, including fatigue and dry mouth. He had a proportional body structure, with a height of 174 cm, a body weight of 70 kg, and a normal body mass index (BMI) of 23.1 kg/m². Laboratory tests confirmed fasting hyperglycemia (8.8 mmol/L) with elevated glycated hemoglobin (HbA1c; 7.6%; 60 mmol/mol) and negative antibodies to glutamate decarboxylase (Table 1). Other pancreatic autoantibodies were not tested. Thyroid hormone levels were normal. He was subsequently diagnosed with type 2 diabetes, prescribed 850 mg of metformin daily, and advised to follow a low-carbohydrate diet. One month later, and after metformin was discontinued, target glycemic control was achieved, with HbA1c levels ranging from 6.1% to 6.8% (43-51 mmol/mol) over the following 6 years. Imaging studies, including thyroid and abdominal ultrasound as well as brain magnetic resonance imaging (MRI) were unremarkable.

Table 1.

Phenotypic and genetic information of affected family members

Relationship to proband (individual ID)	Diabetes	Hyperinsulinism	Additional features	Genotype
Male proband (IV-III)	Diagnosed at age 28 years, BMI at diagnosis 23.1 kg/m2. Fasting blood glucose: 8.8 mmol/L, HbA1c: 7.6% (NR: 4.8-5.9%), C-peptide: 1.28 ng/mL (NR: 0.8-4.2) GAD antibody negative (<10 mIU/mL). Treated with metformin	Episodic hypoglycemia (glucose 2.5 mmol/L) at 34 that increased with frequency and severity. At 36 years, HI confirmed on fasting test (4 hours fasting) plasma glucose: 2.4 mmol/L, C-peptide: 3.63 ng/mL, insulin: 60.4 mU/mL (NR: 2.9-24), proinsulin: 43.1 pmol/L (NR: 0.5-6.1). Ongoing at 42 years of age.	None	p.(Ser64Phe)/N
Sister (IV-I)	None	None	None	N/N
Sister (IV-II)	None	None	None	N/N
Female second cousin (IV-IV)	None	Diagnosed at age 34 (plasma glucose: 2.5 mmol/L, C-peptide: 2.77 ng/mL, insulin: 14.5 mU/mL), coincided with weight gain.	Ovarian cancer	p.(Ser64Phe)/N
Female second cousin (IV-V)	None	None	None	N/N
Female second cousin (IV-VI)	None	None	None	N/N
Female second cousin (IV-VII)	Diagnosed at age 40, diet treated	None	None	Not tested
Mother (III-II)	None	Symptoms of hypoglycemia started at 46 years. Hospitalized at 65 years with severe hypoglycemia (plasma glucose: 1.9-2.2 mmol/L, C-peptide: 2.8 ng/mL [NR: 0.81-3.85 ng/mL]). Chronic pancreatitis.	None	Obligate heterozygote
Female first cousin once removed (III-IV)	None	Diagnosed at age 39, resection of 90% of pancreas. Relapse at 64	None	p.(Ser64Phe)/N
Male first cousin once removed (III-IX)	Diagnosed at age 60, diet treated	None	None	Not tested
Male first cousin once removed (III-XII)	None	Diagnosed at age 45 (according to relatives)	None	Not tested
Grandmother (II-II)	Diagnosed at age 65, diet treated	None	None	Obligate heterozygote
Great aunt (II-III)	None	None	Congenital cataracts and glaucoma	Not tested
Great uncle (II-IV)	None	Diagnosed at age 50 (according to relatives)	None	Obligate heterozygote
Great uncle (II-VI)	Diagnosed at age 60, medically treated	None	None	Not tested
Great uncle (II-XIII)	Age at diagnosis unknown. Insulin treated	None	None	Not tested

Individual numbers correspond to those in the pedigree shown in Fig. 2. Data are presented for individuals with available clinical or genetic information. N/N indicates that the p.(Ser64Phe) variant was tested for, but was not detected.

Abbreviations: BMI, body mass index; HbA1c, glycated hemoglobin; HI, hyperinsulinism; NR, normal range.

At 34 years of age, he experienced the first episodes of hypoglycemia (blood glucose <2.5 mmol/L) (Supplementary Figure 1) [14], which increased with frequency and severity over the following 2 years. The hypoglycemia was nonketotic. A 72-hour fasting test was stopped after 4 hours when HI was confirmed (plasma blood glucose; 2.4 mmol/L, insulin; 60.4 mU/mL) (Table 1). Abdominal ultrasound and computed tomography (CT) were unremarkable, whereas an MRI highlighted a possible 8-mm lesion in the pancreatic head. Subsequent ⁶⁸Ga–positron emission tomography (PET)/CT imaging (GEDiscovery PET/CT scanner, GE Medical System, Waukesha, WI, USA) demonstrated inconclusive contrast uptake in the same region.

At 37 years of age, the proband underwent enucleoresection of a suspected insulinoma located in the pancreatic head. Pathohistological examination confirmed an insulin-positive neuroendocrine tumor (5 × 4 mm), consistent with a small insulinoma, and foci of neuroendocrine dysplasia, measuring up to 300 μm in size. Hypoglycemia recurred within 1 month postsurgery, with similar frequency and severity. Treatment with short-acting octreotide and diazoxide was initiated but discontinued due to lack of efficacy. At 39 years of age, the proband was reexamined due to continued symptoms of hypoglycemia, in particular, in relation to physical activity. During a 72-hour fast, the patient had severe symptoms of hypoglycemia after 35 hours and concomitant hyperinsulinemic hypoglycemia (plasma glucose 2.6 mmol/L, insulin 12.4 mU/mL, C-peptide 1.94 ng/mL and proinsulin 53 pmol/L) [15]. Subsequent imaging including F-18-DOPA PET/CT, 68-Ga-DOTATOC PET/CT and endoscopic ultrasound examination (EUS) did not show any significant focal lesions or tumors. The patient has since been treated with long-acting octreotide (Sandostatin LAR) at 20 to 30 mg every 4 weeks; however, he has had moderate reduction in the frequency and severity of hypoglycemic episodes. Episodes with glucose levels below 3 mmol/L and severe symptoms, such as loss of consciousness or those requiring assistance for recovery, have generally not been observed. However, maintaining this level of control requires considerable effort. The patient relies on continuous glucose monitoring (CGM) with alarm functionality (Libre 2), carries a glucagon pen for the prevention of acute severe hypoglycemia, and limits physical activity, among other preventative measures. Due to periodic increases in frequency of hypoglycemic episodes, imaging with F-18-DOPA PET/CT and 68-Ga-DOTATOC PET/CT has been repeated at 40 years of age, and 68-Ga-DOTATOC PET/CT at 42 years of age, but still without evidence of focal lesions or tumors in the pancreas.

Family members

Family history revealed that 5 of the proband's maternal relatives had adult-onset hyperinsulinism, 5 had diabetes, and 1 individual had congenital cataracts and glaucoma (Table 1 and Fig. 1). One of the individuals with HI also had a history of early-onset ovarian cancer.

Figure 1.

Partial pedigree illustrating clinical phenotypes and genotype data. Squares indicate males and circles indicate females. Black-filled symbols represent individuals with adult-onset hyperinsulinism; gray-filled symbols represent individuals with diabetes; a black-and-white hatched symbol represents an individual with congenital cataracts and glaucoma; a gradient symbol (light gray to black) represents an individual with diabetes followed by hyperinsulinism. Genotypes are shown for individuals who underwent genetic testing, with N/N indicating no variant detected. An asterisk marks obligate heterozygotes, the proband is marked by an arrow.

Genetic testing

Targeted next-generation sequencing (tNGS) of 17 genes for monogenic diabetes and/or congenital HI (ABCC8, CACNA1D, GLUD1, GCK, GCP3, HADH, HK1, HNF1A, HNF4A, INSR, KCNJ11, KDM6A, KMT2D, MAFA, PMM2, SLC16A1, TRMT10A) was performed on a leukocyte DNA sample from the proband as previously described [16]. This included on- and off-target copy number variant analysis [17]. Upon identification of a pathogenic variant, co-segregation studies were performed by Sanger sequencing of samples from 6 available family members [5].

Phenotype-driven genetic testing had previously been performed in 2 family members by a commercial laboratory: the proband and second cousin (IV-IV, Fig. 1) underwent testing with the Invitae tNGS Multi-Cancer Panel (Invitae, San Francisco, CA, USA), which captures 74 genes, including MEN1, BRCA1, and BRCA2.

Haplotype analysis

Haplotype analysis was performed using whole-genome sequencing data from the proband along with the proband of each family with the p.(Ser64Phe) MAFA variant described by Iacovazzo et al [5]. Reads were aligned to BWA MEM v0.7.15 and processed with Picard v2.7.1, GATK v3.7. Variants were annotated with Alamut Batch Standalone v1.11 (Sophia Genetics, Lausanne, Switzerland). Shared haplotypes were defined as ≥150 consecutive identical variants across the MAFA region.

Immunohistochemistry of pancreatic tissue

Tissue specimens were analyzed by microscopy of hematoxylin-eosin (H&E) staining of formalin-fixed, paraffin-embedded 4-µm thick sections and immunohistochemical staining using the BenchMark Ultra immunostainer (Ventana Medical Systems, Tucson, AZ) with the OptiView-DAB detection kit (Ventana Medical Systems, Tucson, AZ), nuclear counterstaining was done with the BenchMark Ultra instrument using Hematoxylin II (Ventana Medical Systems, Tucson, AZ) and coverslipping using a Tissue-Tek Film coverslipper (Sakura, Alphen aan den Rijn, The Netherlands). An overview of the antibodies, dilutions, incubation times and epitope retrieval procedures are provided in Supplementary Table 1 [14].

Immunofluorescence of pancreatic tissue

Pancreatic samples from the proband were dewaxed and subjected to heat-induced epitope retrieval in Tris-EDTA buffer (10 mM Tris Base, 1 mM EDTA pH 9 for 20 minutes). After blocking, sections were sequentially probed with Rabbit anti-MAFA (Abcam Ab26405, RRID:AB_776146, 1:750 1 hour at room temperature), followed either by detection with the Dako REAL EnVision-HRP detection system (Agilent Technologies K500711-2) and mounted for the immunohistochemistry studies, or by detection with goat-anti-Rabbit 488 antibody (RRID:AB_2576217). Following washes, sections were probed with mouse monoclonal anti-glucagon antibody (Abcam, ab10988, RRID:AB_297642, 1:2000 for 1 hour), followed by detection with goat-anti-mouse Alexa Fluor 555 antibody (RRID:AB_2633276). Finally, sections were probed with guinea pig anti-insulin (Agilent, IR002, RRID: AB_2800361, 1/5 for 1 hour), followed by detection with goat-anti-guinea pig Alexa Fluor 647 antibody (RRID:AB_2535867) in the presence of DAPI (1 μg/mL) to identify cell nuclei. Sections were not stripped between antibody incubations. Sections were mounted and imaged on an Akoya Phenolmager HT scanner.

MAFA expression analysis

To evaluate MAFA expression across multiple human tissues, we obtained gene-level expression, bulk-tissue data from the version 8 release of the Genotype-Tissue Expression (GTEx) project. All 54 tissue types were ranked by median MAFA expression levels (expressed in transcripts per million [TPM]) [18].

Ethical considerations

The study complied with the Declaration of Helsinki, with informed written consent obtained from all individuals. The study was approved by the Wales Research Ethics Committee 5 (22/WA/0268), with participants recruited to the Genetic Beta Cell Research Bank (IRAS: 316050).

Results

Heterozygous MAFA variant in 6 individuals, including 3 obligate heterozygotes

Targeted sequencing analysis identified a previously reported heterozygous MAFA variant, c.191C>T, p.(Ser64Phe) (NM_201589) in the proband [5]. No additional pathogenic or likely pathogenic variants were detected in genes included on the MODY, HI, or multi-cancer targeted sequencing panels. The MAFA c.191C>T, p.(Ser64Phe) variant was also heterozygous in the proband's second cousin and the cousin's mother; both clinically diagnosed with HI. Detection of the variant in these individuals enabled classification of the proband's mother (III-II), maternal grandmother (II-II), and maternal great uncle (II-IV) as obligate heterozygotes (Fig. 1). Samples from the other 6 relatives with diabetes, HI, or congenital cataracts and glaucoma were not available for testing. The proband's 2 unaffected sisters and 2 unaffected second cousins did undergo genetic testing; none were heterozygous for the variant (Fig. 1). Haplotype analysis did not identify a shared region spanning MAFA between the 3 probands with this variant (data not shown), indicating that p.(Ser64Phe) is likely to be a recurrent variant arising independently on multiple alleles.

Clinical characteristics of 6 individuals with hyperinsulinism

The male proband initially presented with diabetes in adulthood, which quickly progressed to HI. Five of his family members had initially presented with HI, 2 were male and 3 were female (Table 1). Four of these (1 male, 3 females) were heterozygous for the MAFA variant (including the 2 obligate heterozygotes). In the remaining case, a sample was not available for testing. Overall, the median age at HI diagnosis for the proband and 5 family members was 42 years.

Insulinomatosis was confirmed in the proband following surgical resection of a small insulinoma (measuring 5 mm), which did not resolve the hypoglycemia and could not be controlled with diazoxide or octreotide. The proband's mother (III-II), an obligate heterozygote, had experienced symptoms of hypoglycemia beginning at the age of 46 years. She had been hospitalized at the age of 65 with severe hypoglycemia (plasma glucose: 1.9-2.2 mmol/L, C-peptide: 2.8 ng/mL, [normal range 0.81-3.85 ng/mL], 24-hour urinary cortisol: 55 mg/day [normal range 45-380 mg/day]). Abdominal ultrasound showed features of chronic pancreatitis. She died on the seventh day of hospitalization, with an autopsy report citing the cause of death as “a benign pituitary tumor with the formation of an empty sella turcica, secondary adrenal insufficiency and hypoglycemia syndrome.”

A female cousin (III-IV), heterozygous for the variant on genetic testing, was clinically and biochemically diagnosed with HI at 39 years. She underwent 3 pancreatic surgeries to control hypoglycemia. This culminated in a 90% pancreatectomy at the age of 52 years which resulted in her achieving normoglycemia for over 10 years. The HI recurred at 64 years. This woman's daughter (IV-IV) was also heterozygous for the MAFA variant on genetic testing and was diagnosed with HI (clinically, biochemically) at 34 years. She had undergone an oophorectomy at the age of 31 due to a mucinous adenocarcinoma of the ovary (tumor size 10 cm). Her tNGS Multi-Cancer Panel was negative

According to the investigated relatives, a great uncle (obligate heterozygote, II-IV) had been diagnosed with HI at 50 and a male cousin (genetics not confirmed, III-XII) had been diagnosed with HI at 45. The biochemistry results confirming this diagnosis and details of their treatment are not available (Table 1).

Clinical characteristics of 6 individuals with diabetes

Six individuals, including the proband, initially presented with diabetes; 4 were male and 2 were female. Genetic testing was not performed on any of the proband's family members with diabetes, although the maternal grandmother (II-II), who was diagnosed with diabetes at 65 years was an obligate heterozygote. The median age at diabetes onset in this group was 60 years. Of these, 3 were managed with diet, and 2 were treated with oral hypoglycemic agents or insulin.

Immunohistochemistry confirmed insulinomatosis in the proband

Microscopy of H&E and immunohistochemical stained sections showed multiple insulin-expressing pancreatic neuroendocrine lesions, most corresponding to neuroendocrine dysplasia (0.25-0.5 mm) (Fig. 2A-2D) and 1 lesion qualifying as a small pT1 insulinoma (5 mm in diameter) (Fig. 2E-2H).

Figure 2.

Histological and immunohistochemical features of MAFA mutation-positive insulinomatosis in the proband. A-D, Multiple small neuroendocrine dysplastic foci are shown. A, Overview image showing pancreatic tissue containing neuroendocrine dysplastic foci surrounding a pancreatic duct (H&E staining) (Scale bar: 250 µm). B, Strong synaptophysin expression in the neuroendocrine dysplastic foci (synaptophysin immunostaining) (Scale bar: 500 µm). C, Strong insulin expression (insulin immunostaining) (Scale bar: 50 µm). D, Negativity for glucagon (glucagon immunostaining) (Scale bar: 50 µm). E-H, A very small neuroendocrine tumor consistent with insulinoma is shown, measuring 5 mm. E, H&E staining (Scale bar: 1 mm). F, Strong insulin expression (insulin immunostaining) (Scale bar: 1 mm). G, Ki67 index is very low (below 1%) (Ki67 immunostaining) (Scale bar: 1 mm). H, Negativity for glucagon in the tumor (glucagon immunostaining) (Scale bar: 1 mm).

Aberrant nuclear MAFA expression in beta-cells detected by immunofluorescence

Pancreatic tissue from the proband stained for MAFA demonstrated weak nuclear expression within the adjacent normal islets, but notable strong nuclear MAFA staining in the tumor regions (Fig. 3). When the proband tissue was co-stained for insulin and glucagon, tumor regions exhibited strong nuclear MAFA and cytoplasmic insulin (Fig. 3). These results support the pathogenic role of the variant within the beta-cells.

Figure 3.

MAFA staining present in the nuclei of cells within areas of insulinomatosis in the proband. Imaging of staining of MAFA (magenta), insulin (cyan), glucagon (yellow) and 4,6-diamidino-2-phenylindole (DAPI) (gray) in 2 areas of pancreatic tissue resected from the proband, an area of insulinomatosis (top row) and normal tissue (bottom row). MAFA is present in the nuclei of the insulinomatosis at a much higher level than in the normal tissue. Scale bar, 100 µM.

Expression data confirms MAFA is well-expressed in pancreas but not ovarian tissue

Bulk pancreatic tissue showed the third-highest MAFA expression among all tissues analyzed (median TPM: 1.060) [18]. In contrast, ovarian tissue exhibited extremely low expression (median TPM: 0.03442), ranking as the third-lowest across tissues (Supplementary Figure 2) [14].

Discussion

We describe the fourth family with a pathogenic MAFA activating variant with diabetes and HI. Within this family, we report the second individual with a dual phenotype, with the proband initially presenting with diabetes in early adulthood followed by HI driven by insulinomatosis 6 years later [5]. This clinical trajectory contrasts with the more common dominant ABCC8- and HNF4A-related forms of HI, in which HI typically manifests in very early infancy and, in some cases, later progresses to diabetes in adulthood [3, 19, 20]. In contrast, our patient fits within the spectrum of adult-onset HI, a presentation most frequently associated with insulinomas, either in the context of MEN1 or non-MEN1-related disease. Our case aligns with the latter, as the proband had multiple pancreatic neuroendocrine dysplastic foci in addition to a small 5 mm insulinoma. These results are important given that very few genes have been implicated in adult-onset HI. While SLC16A1 promoter variants represent another monogenic cause of adult-onset hyperinsulinemic hypoglycemia, they are not associated with insulinoma formation, highlighting the heterogeneity of adult-onset HI [21].

In this family, 3 of 5 individuals who initially presented with HI were female, whereas 4 of 6 individuals who presented with diabetes were male. Although the sample size is too small for firm conclusions, this is consistent with previously reported sex-dependent patterns of the disease where HI is more common in females and diabetes in males [5, 8]. The genotype could, however, not be confirmed in 5 family members with diabetes, raising the possibility that some individuals are phenocopies which is possible given the high prevalence of type 2 diabetes in the general population.

Our immunofluorescence results further support the mechanistic role for MAFA in this disease, showing stronger nuclear MAFA expression in tumor regions with colocalized cytoplasmic insulin, whereas adjacent normal islets exhibited absent or weak nuclear MAFA. This pattern is consistent with the proposed molecular mechanism, in which MAFA acts as a transcription factor involved in a glucose-stimulated insulin secretion [13].

One individual within the family had young-onset ovarian cancer of unknown genetic cause despite comprehensive genetic testing. Interestingly in one reported family with a MAFA activating variant, 4 individuals developed solid organ tumors (age 53-75 years) [7]. While these occurrences are intriguing given the known role of MAFA as an oncogene, it seems unlikely that MAFA dysregulation is a major driver of neoplasia in our patient, as GTEx expression data does not support substantial baseline MAFA expression in the ovaries. These observations may therefore reflect coincidental occurrences rather than evidence of a direct role for MAFA in extra-pancreatic tumorigenesis. However, a potential oncogenic effect of MAFA overexpression or dysregulated expression in specific non-β-cell contexts cannot be excluded and warrants further study.

The clinical course of the HI in the proband and his cousin highlights the challenges in managing adult-onset HI, as surgical interventions, including enucleoresection or pancreatectomy, often provide only temporary remission. Pharmacological therapies, such as diazoxide or somatostatin receptor analogues, showed variable efficacy, emphasizing the need for ongoing monitoring and individualized management strategies.

Furthermore, routine visualization techniques had a limited diagnostic yield, in contrast to a previous report [5]. In the proband, Ga-68-DOTATOC did not detect insulinomatosis at initial presentation, and subsequent explorative EUS and repeated PET/CT imaging did not confirm any lesion in the setting of recurrent HI.

Our study has limitations, including the inability to obtain in depth clinical data and DNA from all affected family members which restricted clinical observations. This included an individual with congenital cataracts and glaucoma, which appears to be a rarer feature of this condition [5]. It is interesting to note that this individual had no reported history of diabetes or HI which differs to the 4 previously reported patients with a MAFA variant and congenital cataracts and/or congenital glaucoma, who all had diabetes [5].

MAFA was reported as a disease-causing gene in 2018 and the related syndrome was incorporated into the 2022 WHO Classification of Endocrine and Neuroendocrine Tumors under familial cancer syndromes [22]. Despite this, MAFA is not yet included in all commercial multi-gene panels for hereditary cancers. This highlights a gap in genetic testing that may delay recognition of MAFA-associated risks and underscores the need to update panels to reflect emerging gene-disease associations. Furthermore, because MAFA variants cause autosomal dominant diabetes and adult-onset hyperinsulinemic hypoglycemia, the gene should be considered for inclusion in genetic testing panels for these 2 disorders.

In conclusion, our study highlights the distinctive clinical profile of MAFA-related disease, characterized by adult-onset hyperinsulinism and diabetes that can occur independently or sequentially within the same individual. Awareness of this dual phenotype is essential for early recognition, accurate genetic counseling, and tailored clinical management of affected families.

Abbreviations

BMI

body mass index

CT

computerized tomography

EUS

endoscopic ultrasound examination

GTEx

Genotype-Tissue Expression

HbA1c

glycated hemoglobin

HI

hyperinsulinemic hypoglycemia

MRI

magnetic resonance imaging

PET

positron emission tomography

tNGS

targeted next-generation sequencing

Funding

This research was funded in whole, or in part, by the Wellcome Trust [223187/Z/21/Z]. For the purpose of open access, the author has applied a CC BY public copyright license to any Author accepted Manuscript version arising from this submission.

Disclosures

The authors declare that there are no competing interests.

Data availability

Restrictions apply to the availability of some or all data generated or analyzed during this study to preserve patient confidentiality or because they were used under license. The corresponding author will on request detail the restrictions and any conditions under which access to some data may be provided.

Availability of data and materials

The datasets generated and/or analyzed during the current study are not publicly available to protect patient privacy but are available from the corresponding author on reasonable request.