Challenges in diagnosis and treatment of KCNJ11-MODY

Background

Monogenic diabetes (MD) is a rare type of diabetes resulting from a pathogenic variant in a single gene. It is associated with various clinical conditions, such as neonatal diabetes mellitus (NDM), maturity-onset diabetes of the young (MODY) and various diabetes-associated syndromes. MODY accounts for 1–5% of diabetes cases and is often misclassified as another type of diabetes, resulting in an average delay in diagnosis of ten years (1).

Pathogenic variants of the KCNJ11 gene are associated with NDM, MODY, type 2 DM and hyperinsulinaemic hypoglycaemia, all of which are the most frequent cause of NDM (2). KCNJ11-MODY is a very uncommon subtype of MODY, with only a few reported cases worldwide. It is caused by a pathogenic variant of the KCNJ11 gene (1). Sulfonylureas are typically the treatment of choice, leading to insulin release in a non-ATP-dependent manner (1). The exact prevalence, long-term response to therapy and prognosis of this subtype of MODY are not yet known.

Here, we describe the management of a case of de novo KCNJ11-MODY, with a focus on the transition from insulin to sulfonylureas. In addition, this case contributes to supporting the evidence of the effectiveness of sulfonylureas in the management of KCNJ11-MODY and also provides data on continuous glucose monitoring during the patient’s follow-up period. The literature currently lacks data concerning the treatment of KCNJ11-MODY.

Case presentation

A 24-year-old Caucasian woman was referred to the Endocrinology Department for newly diagnosed DM. On her first appointment, the patient complained of fatigue and polydipsia for several months, although she did not report that she was experiencing any other symptoms, such as polyuria, weight changes or polyphagia. A physical examination was deemed irrelevant, and her body mass index was 22.8 kg/m² (weight 60.5 kg, height 163 cm). Her weight remained stable throughout her entire adult life. No acanthosis nigricans or other skin abnormalities were observed. The patient’s blood pressure was 122/77 mmHg, and her pulse was 88/minute. At diagnosis, blood tests showed a haemoglobin A1c level of 93 mmol/mol (10.7%), fasting glucose of 15.5 mmol/L (278 mg/dL) and negative anti-glutamic acid decarboxylase and anti-insulin antibodies. Two months prior to the patient’s first visit, she had been prescribed metformin 700 mg twice daily and sitagliptin 25 mg and empagliflozin 25 mg daily, which had led to significantly improved glycaemic control. Fasting blood glucose levels ranged from 7.1 mmol/L (128 mg/dL) to 9.3 mmol/L (168 mg/dL), with levels before lunch ranging from 8.7 mmol/L (156 mg/dL) to 10.2 mmol/L (183 mg/dL) and levels before dinner between 9.3 mmol/L (167 mg/dL) and 13.9 mmol/L (250 mg/dL). At this point, the patient’s haemoglobin A1c was 68 mmol/mol (8.4%) and she did not report any changes in weight.

The patient’s birth weight was normal (3.180 kg; 30.2nd percentile), and her neonatal period and early childhood were unremarkable. However, medical records from primary care showed the onset of pre-diabetes from the age of 13 (haemoglobin A1c: 44 mmol/mol (6.2%)). Family history was relevant, as three grandparents had diabetes: her paternal grandmother was diagnosed at 37 years of age, her paternal grandfather at 65 years of age, and her maternal grandfather after the age of 50. None of her grandparents required insulin or were overweight when diabetes was diagnosed. There were no other cases of diabetes in the family (Fig. 1).

View Full Size

Genogram displaying family member with diabetes. The presence of diabetes is indicated by yellow boxes or circles, with an arrow pointing to the index patient. The classified type of diabetes, age at onset of diabetes, medication taken and the actual haemoglobin A1c (HbA1c) are also indicated. In persons without diabetes, the age corresponds to actual age with minor changes made for anonymity. Male subjects are indicated by boxes, while female subjects are indicated by circles. DM: diabetes mellitus; MODY: maturity-onset diabetes of the young.

Citation: Endocrinology, Diabetes & Metabolism Case Reports 2024, 4; 10.1530/EDM-24-0048

• Download Figure
• Download figure as PowerPoint slide

Type 1 DM was the most probable diagnosis, which led to the start of treatment with insulin (glargine U300 10 units daily) and the suspension of oral hypoglycaemic drugs. Concurrently, the patient was subject to continuous glucose monitoring, using the FreeStyle Libre 2 System®.

Investigation

Postprandial C-peptide performed during insulin therapy showed a normal level (0.91 mmol/L (2.75 ng/mL)), together with normoglycaemia (7.9 mmol/L (143 mg/dL)). Other antibodies (anti-zinc transporter 8 and anti-islet antigen 2) were not measured due to unavailability in our laboratory.

At the first clinical re-evaluation, the patient had increased to 12 units daily of glargine U300, which had resulted in adequate fasting glucose levels; however, she manifested postprandial glycaemic excursions, particularly at dinner (Fig. 2A). Treatment with ultra-rapid insulin (aspart) then commenced before dinner, consisting of two units with a glucose goal of 5.0–7.8 mmol/L (90–140 mg/dL) and an insulin sensitivity factor (ISF) of 3.3 mmol/L (60 mg/dL). Later on, three units of aspart before breakfast were added (with the same glucose goal of 5.0–7.8 mmol/L (90–140 mg/dL) and ISF of 3.3 mmol/L (60 mg/dL)) (Fig. 2B).

View Full Size

Glucose patterns in September 2022 (A) and in January 2023 (B). (A) Time in range was 75%, time above range 25%, glucose management indicator 53 mmol/mol, and glucose variability (defined as percent coefficient of variation) 23.8%. (B) Time in range was 78%, time above range 22%, glucose management indicator 53 mmol/mol, and glucose variability (defined as percent coefficient of variation) 22.4%. Graphics were obtained from Libreview®.

Citation: Endocrinology, Diabetes & Metabolism Case Reports 2024, 4; 10.1530/EDM-24-0048

• Download Figure
• Download figure as PowerPoint slide

Despite the low probability of MODY (0.7%) as calculated by the Exeter MODY probability calculator (accessed at https://www.diabetesgenes.org/exeter-diabetes-app/ModyCalculator) (3), a next-generation sequencing panel for MODY was requested. A heterozygous pathogenic variant c.679G>A, p.(Glu227Lys) was identified in the KCNJ11 gene (Kir6.2) [NM_000525.4], which supported the diagnosis of KCNJ11-MODY.

Treatment

An oral sulfonylurea, gliclazide, was then embarked upon, and basal-bolus insulin was progressively discontinued. Initially, gliclazide 30 mg (modified release) was prescribed in the morning, and rapid insulin was administered to correct hyperglycaemia (>10 mmol/L (180 mg/dL)). Meanwhile, the glucose and ISF goals were maintained. During this transition, the patient did not experience hypoglycaemia and excellent glycaemic control was achieved without the need for insulin aspart. In an attempt to discontinue insulin, gliclazide was increased to 60 mg a day after two weeks and the dose of glargine was reduced by half (to 6 units). However, on this day, the patient unfortunately experienced several hypoglycaemic episodes (minimum 2.0 mmol/L (52 mg/dL)). As a result, glargine was stopped and the patient maintained excellent glycaemic control with gliclazide 60 mg (0.99 mg/kg/day) alone, with hardly any hypoglycaemic events.

Outcome and follow-up

Four months after transition to gliclazide, excellent glycaemic control was achieved (time in range of 100%, with a mean glucose of 6.1 mmol/L (109 mg/dL), a glucose management indicator of 41 mmol/mol (5.9%) and glucose variability of 14.2% (defined as a percentage coefficient of variation)) (Fig. 3), which translated into a remarkable enhancement in the quality of life. One year after transition, the patient maintained excellent glycaemic control (haemoglobin A1c: 34 mmol/mol (5.3%)) and no microvascular complications have been diagnosed. The patient was referred to be booked in for a genetic appointment for genetic counselling, with the aim to identify other possible carriers of the pathogenic variant in her family, although it was confirmed that her parents did not have the pathogenic variant.

View Full Size

Glucose patterns in June 2023 (four months after gliclazide beginning). The graphic was obtained from Libreview®. Time in range was 100%, glucose management indicator 41 mmol/mol, and glucose variability (defined as percent coefficient of variation) 14.2%.

Citation: Endocrinology, Diabetes & Metabolism Case Reports 2024, 4; 10.1530/EDM-24-0048

• Download Figure
• Download figure as PowerPoint slide

Discussion

This case exemplifies the applicability of precision medicine in managing patients with DM. Through an exhaustive anamnesis, consultation of primary care medical records and the revelation of unexpected blood test results (normal C-peptide and absence of pancreatic autoimmunity), the medical team suspected a monogenic type of diabetes. The confirmation of a molecular diagnosis guided the correct approach, contributing to a substantial improvement in the patient’s quality of life.

The first KCNJ11-MODY case was reported by Bonnefond and coworkers in 2012, who identified the pathogenic variant c.679G>A, p.(Glu227Lys) in the KCNJ11 gene. Four of 12 family members had not been diagnosed with diabetes at the time of the genetic test, one suffered from impaired fasting glucose, and three had a normoglycaemic status (aged 11, 14 and 39) (4). This suggests that pathogenic variants in the KCNJ11 gene can be associated with different diabetes phenotypes and levels of penetrance, such as the most frequent subtypes of MD (1).

In our patient’s family, neither of the parents had been diagnosed with diabetes. Two hypotheses were accordingly postulated: i) the identified pathogenic variant occurred de novo and ii) it had variable penetrance. The paternal grandmother’s early diabetes at 37 suggests inheritance. However, the pathogenic variant was not identified in the parents, and therefore, this case of KCNJ11-MODY involves a de novo pathogenic variant. These results highlight the importance of suspecting this type of diabetes, even in the absence of a significant family history. Epidemiological studies have demonstrated that de novo pathogenic variants in MODY genes associated with the most frequent subtypes of MD can occur in 13.8–16.3% of cases (5, 6). Nevertheless, the prevalence in KCNJ11-MODY is currently unknown.

The MODY probability calculator showed good discrimination performance between MODY, type 1 diabetes (c-statistic 0.95; sensitivity 91%; specificity 94%) and type 2 diabetes (c-statistic 0.98; sensitivity 72%; specificity 91%) (7). Despite its overall performance, this case revealed certain limitations in its application. For example, the calculator may not work in cases without an affected parent, owing to variable penetrance and the occurrence of de novo pathogenic variants; in addition, it relies solely on whether the parents had DM, regardless of family history. Notably, this case shows that the MODY probability calculator overlooks other discriminatory biomarkers, such as C-peptide, pancreatic autoantibodies and extra-pancreatic features (excluding renal cysts) (8). To date, this calculator has only been validated in individuals aged 35 years or younger and never for patients aged over 35 at diagnosis (9). In addition, only the most frequent types of MD (GCK-MODY, HNF1A-MODY and HNF4A-MODY) (7) have been validated. Despite the calculator’s results, suspicion of MD in this case was supported by evidence of pre-diabetes in adolescence, normal C-peptide levels and the absence of pancreatic autoimmunity. Furthermore, with genetic tests becoming more accessible, the identification of a pathogenic variant will significantly impact disease management and quality of life.

The KCNJ11 gene encodes the inward rectifier potassium channel (Kir6.2) subunit of the pancreatic beta-cell ATP-sensitive potassium channel (K_ATP channel). The K_ATP channel is composed of four Kir6.2 subunits and four sulfonylurea receptor 1 (SUR-1) subunits, arranged as a hetero-octamer. When glucose is metabolized, ATP is generated and binds to the Kir6.2 subunit of the K_ATP channel, causing it to close. This leads to membrane depolarization, activation of voltage-gated calcium channels, calcium influx into the cell and exocytosis of insulin granules. The pathogenic variant in the KCNJ11 gene impairs K_ATP channel function and consequently insulin secretion (2). When sulfonylureas bind to the SUR-1 of the K_ATP channel, closing it through an ATP-independent mechanism, this promotes insulin secretion (2). Therefore, sulfonylureas can activate the impaired K_ATP channel in KCNJ11-MODY.

A multicentre study conducted by Bowman et al. and the Neonatal Diabetes International Collaborative Group showed the long-term efficacy and safety of sulfonylureas in the treatment of permanent NDM caused by KCNJ11 pathogenic variants. At a median follow-up at 10.3 (IQR: 9.2–10.9) years, the median haemoglobin A1c was 6.4 (5.9–7.3)%, and no severe hypoglycaemia or weight gain was recorded (3). Although pathogenic variants may vary, these data are encouraging for the long-term use of sulfonylureas in managing KCNJ11-MODY without any loss of efficacy, as seen in type 2 DM. In the majority of reported KCNJ11-MODY cases, sulfonylureas are the drug of choice and glycaemic control is good. In this case, excellent glycaemic control was achieved with a dose of 0.99 mg/kg/day of gliclazide. This dose was lower than the final equivalent dose of gliclazide in other clinical reports of KCNJ11-MODY (1.2 and 6.0 mg/kg/day) (2), or in the studies of both Pearson et al. (2.7 mg/kg/day) (10) and Bowman et al. (1.38 mg/kg/day) (3).

Declaration of interest

The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the work.

Funding

This work did not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector.

Patient consent

Written informed consent for the publication of their clinical details was obtained from the patient. Ethical approval for this publication was obtained from the Institutional Review Board (or Ethics Committee).

Author contributions

JG wrote this case report and was involved in the patient care. DR and JP were involved in the patient care and revised the draft. HF and SR helped in the writing of this case report and revised the draft. SBS, PF and JQ revised the draft.

Patient’s perspective

I discovered that I had diabetes in February 2022, at the age of 23, and was immediately referred to an endocrinology appointment, where the medical team started managing my case. I underwent a genetic test, which identified that my type of diabetes was KCNJ11-MODY. Thanks to this diagnosis, I was able to stop taking insulin and now only take one pill once a day.