Diabetes management in maternally inherited diabetes and deafness (MIDD): A review and a proposed treatment algorithm

Abstract

Maternally inherited diabetes and deafness (MIDD) is a mitochondrial disorder usually caused by the variant m.3243A>G in the MT‐TL1 gene. We have proposed that diabetes in MIDD arises from a combination of insulin resistance and impaired β‐cell function that is more likely to occur in the presence of high skeletal muscle heteroplasmy and moderate β‐cell heteroplasmy for m.3243A>G and to be driven by oxidative stress as a major pathophysiologic mechanism. There are no randomised trials specifically on the management of MIDD. An approach to MIDD informed by its pathophysiology could optimise management and be a framework for future trials. In this narrative review, we discuss the effects of existing anti‐hyperglycaemic medications on oxidative stress, mitochondrial function, and cardiorenal protection. We also review the published case reports and series on the management of diabetes associated with MIDD, and the safety of insulin and non‐insulin medications in MIDD. We found that glucagon‐like peptide‐1 receptor agonists and sodium‐glucose cotransporter‐2 inhibitors have favourable properties in addressing oxidative stress and mitochondrial function. They also harbour cardiorenal protection properties independent from their effect on glucose control that are fully relevant in MIDD, making them ideal candidates as first‐line agents for the management of MIDD. Accordingly, we share our perspective on a disease‐specific algorithm for the management of MIDD diabetes that could delay or prevent the development of cardiovascular and renal complications associated with this mitochondrial disease.

1. INTRODUCTION

Maternally inherited diabetes and deafness (MIDD) is a monogenic mitochondrial disorder caused by a pathogenic variant of the MT‐TL1 gene encoding for a leucine transfer RNA (tRNA^Leu(UUR)). Although one variant, m.3243A>G, accounts for more than 80% of the clinical cases, more than 22 pathogenic variants are known to cause MIDD. ¹ Review of published data led us to propose that MIDD diabetes is more likely to occur in the presence of high muscle heteroplasmy (promoting insulin resistance) and moderate β‐cell heteroplasmy (impairing β‐cell function) for m.3243A>G. ² MIDD is associated with increased oxidative stress and aberrant activation of the mammalian target of rapamycin (mTOR)C1 pathway. The MIDD phenotype is also influenced by mitochondrial metabolic retrograde signalling, a pathway that involves nuclear factor erythroid 2‐related factor 2 (Nrf2) and by other incompletely defined genetic and environmental factors (reviewed in Chanoine et al. ² ). Abnormalities in calcium handling have been reported; however, evidence is inconsistent as to whether they causally contribute to MIDD or reflect downstream effects of other processes such as oxidative stress. ³ , ⁴ , ⁵ Pathogenic variants of the MT‐TL1 gene can also be associated with Mitochondrial Encephalomyopathy, Lactic Acidosis and Stroke‐like Episodes (MELAS). This more severe clinical picture reflects, at least in part, higher cellular heteroplasmy for the pathogenic variant, is usually diagnosed at a younger age, and can also cause mitochondrial diabetes. ⁶

There are no randomised trials on the management of MIDD, the most common form of mitochondrial diabetes. ⁷ We are aware of only two sets of recommendations for the overall management of mitochondrial diabetes, including MIDD (Table 1).

TABLE 1.

Existing recommendations for the pharmacological management of mitochondrial diabetes.

References	Summary of recommendations (adapted from the original publications)
Ng a et al. 8	To be individualised: If weight loss is desirable First line: DPP‐4 inhibitor Second line: add an SGLT2 inhibitor OR a sulfonylurea. Third line: Consider prescribing both agents if HbA1c remains unsatisfactory with one agent or switching to insulin If weight loss is not desirable First line: GLP‐1R agonist OR SGLT2 inhibitor Second line: consider prescribing both agents if HbA1c remains unsatisfactory with one agent Third line: switch to insulin with or without GLP‐1R agonists if HbA1c remains unsatisfactory Comments Metformin is best avoided because of the enhanced risk of lactic acidosis
Yeung et al. 9	Evidence of insulin deficiency: initiate insulin therapy (with nutrition and physical activity) Evidence of insulin resistance: SGLT2 inhibitors and/or GLP‐1R agonists based on individual characteristics. Consider discontinuing metformin if already prescribed

Abbreviations: DPP‐4, dipeptidyl peptidase‐4; GLP‐1R, glucagon‐like peptide‐1 receptor; SGLT2, sodium/glucose cotransporter 2.

^a This algorithm is slightly different from the recommendation in the earlier Newcastle Mitochondrial Disease guidelines by the same group. ¹⁰

In this article, we share our perspective on a disease‐specific approach to MIDD that is grounded in our pathophysiological modelling of the disease. To develop this algorithm, we appraise the effects of existing anti‐hyperglycaemic agents on redox signalling pathways, oxidative stress and mitochondrial function, as these are important perturbations in MIDD and could therefore inform the efficacy of therapy on β‐cell function and insulin resistance. We also prioritise the cardiorenal protection of novel anti‐hyperglycaemic agents, highlighting data that the mechanisms of these benefits involve amelioration of mitochondrial function and oxidative stress and therefore have relevance in MIDD. We integrate this with the reported experience and use of these agents in patients with MIDD from the literature and focus on safety concerns and side effects most relevant to mitochondrial diseases. Because MIDD can affect women of childbearing age, we also reflect on the safety of the medications during pregnancy and breastfeeding. Drawing on these findings, we synthesise the data into an algorithm for the diagnosis, assessment and management of MIDD. However, it is important to note that the effects on blood glucose control, cardiorenal complications and oxidative stress have primarily been investigated in type 1 (T1D) and type 2 (T2D) diabetes, necessitating caution in extrapolating the findings to MIDD. Furthermore, although oxidative stress and mitochondrial dysfunction are major features in MIDD, there is limited clinical evidence to illustrate how targeted alleviation of these aspects can alter its phenotype or natural history. Therefore, our proposed approach is a foundation that requires validation by appropriate clinical studies.

2. PHARMACOLOGICAL AGENTS

2.1. Metformin

Metformin is commonly the first‐line oral anti‐hyperglycaemic agent in the treatment of T2D. In T1D, it may also improve glucose control, insulin requirements and body weight. ¹¹ It reduces glucose primarily by decreasing hepatic gluconeogenesis, a process that necessitates the mitochondria and adenosine triphosphate (ATP) consumption. The exact mechanism remains incompletely understood and involves inhibition of the mTORC1 pathway through adenosine monophosphate‐activated protein kinase (AMPK) dependent and possibly independent mechanisms. ¹² , ¹³ Additional mechanisms may include altered gut motility, ¹⁴ augmentation of glucagon‐like peptide‐1 (GLP‐1) response, ¹⁴ improved insulin sensitivity and increased intestinal glucose efflux. ¹⁵ Although not approved for this indication, clinical studies suggest that metformin, in contrast to insulin and sulfonylureas, may offer cardiovascular ¹⁶ and renal protection ¹⁷ independently from glucose‐lowering.

2.1.1. Effects of metformin on mitochondrial function and oxidative stress

Metformin has several effects on mitochondrial homeostasis and oxidative stress. Through AMPK‐dependent pathways, clinically relevant doses of metformin stimulate complex 1 of the oxidative phosphorylation chain with augmented mitochondrial respiratory chain activity and fatty acid oxidation in the liver of high‐fat diet‐fed mice. ¹² Data from other animal studies support that metformin ameliorates reactive oxidative species (ROS) levels and mitochondrial oxidative stress, including in the pancreas, through activation of AMPK, sirtuins (SIRT1 and SIRT3) and Nrf2. ¹⁸ Through these pathways, metformin also enhances mitophagy (clearance of damaged mitochondria) and mitochondrial biogenesis. ¹⁹ A trial of metformin versus lifestyle changes in patients with newly diagnosed T2D demonstrated superior reduction in serum biomarkers of oxidative stress in the metformin group. ²⁰ Cardiovascular and renal protective effects are also associated with a decrease in markers of oxidative stress. ¹⁶ , ²¹

2.1.2. Safety of metformin in MIDD

Controversy exists regarding the risk of metformin and lactic acidosis in patients with mitochondrial diabetes, as lactate could theoretically accumulate in the setting of mitochondrial dysfunction superimposed with metformin‐induced inhibition of glycerophosphate dehydrogenase and complex 1. ²² However, complex 1 inhibition likely occurs only with supratherapeutic doses and not with clinically relevant doses. ¹² Clinical evidence of this risk is limited to case reports and small case series (Table S1). Other groups, however, have reported positive experiences with the efficacy and safety of metformin use in some patients with MIDD. ²³ , ²⁴ A Cochrane review did not identify any case of fatal or nonfatal lactic acidosis in 70 490 patient‐years of metformin use. ²⁵ Although patients with mitochondrial diabetes were not specifically identified, it is likely that patients with undiagnosed MIDD were included in these studies. Moreover, an expert consensus unanimously agreed that metformin could be safely used in the management of primary mitochondrial diseases. ²⁶ Therefore, our perspective is that metformin could be an option for patients with MIDD with careful monitoring and risk assessment. The risk of lactic acidosis is increased in the setting of liver and renal failure (particularly when the glomerular filtration rate [GFR] is <30 mL/min/1.73 m²), which should be contraindications to its use. Caution may also be needed in patients with a history of stroke‐like episodes (MELAS) as a potential risk of new neurologic events has been raised by some case reports, although clinical evidence remains limited (Table S1). Otherwise, common side effects include abdominal discomfort, bloating and diarrhoea. Metformin has a low risk of hypoglycaemia.

2.2. Sulfonylureas

Sulfonylureas are insulin secretagogues that bind to the SUR1 receptor on pancreatic β‐cells and increase insulin release. ²⁷ They decrease glycated hemoglobin (HbA1c) by 1 to 1.5% in patients with T2D. Another type of sulfonylurea receptor, SUR2, is found in a variety of tissues, including the heart and the kidney. However, extra‐pancreatic actions of sulfonylureas at clinically relevant pharmacological doses have not been demonstrated. ²⁷ The effect of sulfonylureas on cardiovascular disease (CVD) in patients with T2D is inconsistent, possibly reflecting differences among agents and bias in published studies. ²⁸

2.2.1. Effects of sulfonylureas on mitochondrial function and oxidative stress

Correction of hyperglycaemia by sulfonylureas results from increased insulin secretion, a process that generates oxidative stress and metabolic load in the mitochondria. In cell culture models, treatment of β‐cells with glibenclamide (glyburide) or glimepiride activates mTOR and increases ROS production and rates of apoptosis. ²⁹ , ³⁰ Overactivation of the mTOR pathway by glibenclamide may also interfere with mitophagy, impairing clearance of damaged mitochondria. Interestingly, oxidative stress is not exacerbated by gliclazide, which has free radical scavenging properties due to its unique azabicyclic ring. ³⁰ Data from human islet and clinical studies also suggest that gliclazide better attenuates markers of oxidative stress compared to other sulfonylureas. ³¹ Low dose gliclazide can also augment the production of glucose‐dependent insulinotropic polypeptide (GIP) and GLP‐1, two hormones that are protective against oxidative stress and mitochondrial function. ³² Sulfonylureas may provide less durable glucose‐lowering and β‐cell preservation compared to insulin, ³³ raising questions about whether they accelerate β‐cell burnout and mitochondrial dysfunction in part due to oxidative and endoplasmic reticulum (ER) stress. Gliclazide, however, is associated with lower secondary failure rates and time to initiation of insulin compared to older sulfonylureas such as glibenclamide. ³⁴

2.2.2. Safety of sulfonylureas in MIDD

There are several reports of experience with sulfonylureas in MIDD. ⁷ The main side effects are hypoglycaemia and weight gain. Generally, the risk of hypoglycaemia can be lower with gliclazide. ³⁵ Sulfonylureas are metabolised by the liver and excreted in urine (and bile) and caution must be taken in case of chronic kidney disease (CKD). ²⁷

2.3. Thiazolidinediones

Thiazolidinediones (TZDs) are insulin sensitisers that activate the nuclear peroxisome proliferator‐activated receptor γ (PPAR‐γ). In newly diagnosed patients with T2D, pioglitazone and metformin monotherapy caused a similar decrease in HbA1c (1.3% and 1.5%, respectively). ³⁶ In patients with T1D and features of insulin resistance, adjunctive treatment with TZDs may also reduce insulin requirements. ³⁷ Clinical studies demonstrate beneficial effects of TZDs in decreasing the progression from prediabetes to diabetes, improving insulin sensitivity and preserving β‐cell secretory function. ³⁸ The effects on the β‐cell may be achieved through both indirect (decrease in lipotoxicity and glucotoxicity) and direct effects.

2.3.1. Effects of TZDs on mitochondrial function and oxidative stress

In vitro studies have demonstrated a protective effect of TZDs on mitochondrial ROS levels and oxidative stress in the β‐cell. This effect at least partly involves PPAR‐γ and an increase in the expression of catalase, an antioxidant enzyme. ³⁹ In cultured β‐cells exposed to hyperglycaemia, pioglitazone upregulates antioxidant defenses, with attenuation of the mTORC1 axis. ⁴⁰ TZDs can additionally promote peroxisome proliferator‐activated receptor‐gamma coactivator 1‐α activity and mitochondrial biogenesis. ⁴¹ They also improve markers of ER stress in islets in diabetic rodent models, with possible promotion of β‐cell proliferation. ⁴² TZD treatment in obese insulin‐resistant rat models has been shown to improve the redox balance in peripheral tissues and reduce the accumulation of lipid radical formation and associated by‐products. ⁴³ Pioglitazone also modulates oxidative stress and inflammatory signalling in endothelial and vascular tissue, contributing to positive effects on atherosclerosis. ⁴⁴

2.3.2. Safety of thiazolidinediones in MIDD

To our knowledge, there is only one report of TZD use in MIDD, ⁴⁵ in which pioglitazone improved glycaemic control, attenuated markers of insulin resistance, and was well tolerated. In general, TZDs do not commonly cause hypoglycaemia. ⁴⁶ They are associated with weight gain, fluid retention, haemodilution and heart failure. ³⁸ Although TZDs can worsen congestive heart failure, pioglitazone has been shown to lower the risk of atherosclerotic cardiovascular events. ⁴⁶

2.4. Glucagon‐like peptide‐1 receptor agonists

Incretins are peptides that stimulate insulin release in a glucose‐dependent manner. The two main incretins are GIP and GLP‐1. The action of GLP‐1 is mediated through stimulation of the glucagon‐like peptide‐1 receptor (GLP‐1R) that is present in several tissues, including the endocrine pancreas (glucose‐dependent insulin secretion from pancreatic β‐cells and inhibition of glucagon release from pancreatic α‐cells), stomach (delayed gastric emptying), muscle, fat, heart, kidney and hypothalamus (increased satiety). ⁴⁷

GLP‐1R agonists cause a decrease in HbA1c of 1%–2% and induce weight loss. ⁴⁸ GLP‐1R agonists also improve insulin sensitivity independently from their weight‐lowering effect ⁴⁹ and confer potent cardiovascular (including atherosclerotic disease and congestive heart failure) and renal protection. ⁵⁰

Dual GLP‐1 and GIP receptor agonist agents, such as tirzepatide, have recently become available and exert more potent glucose and weight reduction compared to GLP‐1R agonists. ⁵¹ The addition of GIP receptor agonism further potentiates insulin secretion, satiety, insulin sensitivity, and may blunt GLP‐1‐triggered nausea. ⁵¹ Dual agonists may also offer superior cardiovascular protection and therefore be particularly suitable for MIDD patients, although head‐to‐head comparison trials remain to be completed. ⁵²

2.4.1. Effects of GLP‐1R agonists on mitochondrial function and oxidative stress

In vitro and in vivo studies have demonstrated that GLP‐1R agonists regulate genes and metabolic pathways involved in mitochondrial function and morphology, apoptosis, inflammation, oxidative stress and ER stress. ⁵³ , ⁵⁴ , ⁵⁵ They increase antioxidant capacity not only in the pancreas but also in the heart, brain, liver and kidney through pathways including AMPK, protein kinase A (PKA), exchange protein kinase activated by cyclic AMP 2 (Epac2), protein kinase B, SIRT1 and Nrf2, and attenuation of aberrant mTOR hyperactivity. ⁵⁶ , ⁵⁷ This also corresponds to increased mitophagy and improvement in mitochondrial membrane potential, mass and morphological characteristics in various tissues. ⁵³ , ⁵⁸ Treatment with GLP‐1R agonists has also been shown to accentuate mitochondrial recovery and adaptation to injury or stress in pancreatic β‐cell models. ⁵⁹

A meta‐analysis of 40 clinical trials performed in patients with T2D showed that GLP‐1R agonists significantly decreased serum markers of oxidative stress. ⁶⁰ Ikonomidis et al. ⁶¹ also demonstrated that GLP‐1R agonists exert a superior improvement of neurohormonal markers (GDF‐14 and mOTS‐C) of oxidative stress and mitochondrial activation in patients with T2DM and high cardiovascular risk in comparison to treatment with insulin. Moreover, the renoprotective effects of GLP‐1R agonists even in the absence of diabetes have highlighted the improvements in inflammation, oxidative stress and perfusion in the kidney as mechanisms which are likely causal. ⁶²

Limited data from rodent models suggest enhanced improvements in oxidative stress and antioxidant activity in the kidney and heart with tirzepatide compared to GLP‐1R‐only agonists. ⁶³ , ⁶⁴ There are currently no human or clinical data to support these findings.

2.4.2. Safety of GLP‐1R agonists in MIDD

Several recent case reports have reported adequate tolerance and improved diabetes control in MIDD when insulin was either replaced by or used in conjunction with newer agents such as GLP‐1R agonists, and in some cases to eliminate the need for exogenous insulin. ⁹ , ⁶⁵ , ⁶⁶ , ⁶⁷ In general, GLP‐1R agonists can cause weight loss and potential gastrointestinal side effects (nausea, vomiting and diarrhoea). ⁶⁸ In MIDD, body mass is often decreased compared to the general population, and further weight loss may not be desirable. However, weight loss caused by GLP‐1R agonists is dose dependent, and significant effects on HbA1c with minimal weight loss by low dose GLP‐1R agonists have been demonstrated. ⁶⁹ Recent conflicting reports on the effects of GLP‐1R agonists on eye complications in older patients with primarily T2D have emerged and require further clarification. ⁷⁰ One of the major benefits of GLP‐1R agonists is a low risk of hypoglycaemia unless in combination with insulin secretagogues or exogenous insulin.

Despite its more potent clinical efficacy, tirzepatide has a similar safety profile to GLP‐1R agonists without a higher incidence of gastrointestinal adverse effects. ⁷¹ There is one report of tirzepatide (dual GLP‐1/GIP receptor agonist) use in a patient with mitochondrial diabetes caused by a m.3271T>C mutation with good efficacy and tolerability. ⁷²

2.5. Dipeptidyl peptidase 4 inhibitors

Dipeptidyl peptidase 4 (DPP‐4) inactivates GLP‐1 and GIP. DPP‐4 inhibitors are oral agents that prevent the degradation of GLP‐1 and increase its concentration and duration of action. Like GLP‐1R agonists, they stimulate insulin secretion in a glucose‐dependent manner, inhibit glucagon secretion, and do not cause hypoglycaemia. In contrast to GLP‐1R agonists, DPP‐4 inhibitors promote physiological concentrations of GLP‐1, do not improve insulin sensitivity, do not cause nausea or vomiting, and are weight neutral. ⁷³ They lower HbA1c by 0.5%–0.8%. ⁷³ DPP‐4 inhibitors have no demonstrated cardioprotective effect. ⁷³ Modest beneficial effects on the kidney include a decrease in albuminuria but no significant effect on estimated GFR (eGFR). ⁷⁴

2.5.1. Effects of DPP‐4 inhibitors on mitochondrial function and oxidative stress

In vitro and in vivo animal studies have suggested a beneficial effect of DPP‐4 inhibitors on mitochondrial oxidative stress and function, which is largely driven by the consequences of increased GLP‐1 activity as described earlier. ⁷⁵ , ⁷⁶ Additionally, inhibition of DPP‐4 was also found to inhibit apoptosis in human islets from organ donors with and without T2D, mediated through increased GLP‐1, and overall β‐cell mass may increase as well. ⁷⁷ , ⁷⁸ However, in patients with T2D, DPP‐4 inhibitors had no additional effect on oxidative stress markers when added to existing sodium‐glucose cotransporter‐2 (SGLT2) inhibitor or GLP‐1R agonist treatment. ⁷⁹ In a randomised trial comparing sitagliptin and glimepiride, there was a small but significant increase in biological antioxidant potential (assessed via the amount of serum trivalent iron that is deoxidised) in the sitagliptin but not in the glimepiride group, and no changes in reactive oxygen metabolite‐derived compounds. ⁸⁰

2.5.2. Safety of DPP‐4 inhibitors in MIDD

Their safety profile is generally good, ⁷³ and there are case reports demonstrating their safety and tolerance in MIDD. ⁹ , ⁸¹ DPP‐4 inhibitors have a low risk of hypoglycaemia. They may, however, be associated with cholelithiasis and pancreatitis. DPP‐4 inhibitors are excreted in the kidney, except for linagliptin, which is excreted via the biliary pathway and can be used in end‐stage renal disease (ESRD).

2.6. Sodium‐glucose cotransporter‐2 inhibitors

SGLT2 inhibitors are oral agents that block the reuptake of sodium and glucose in the proximal tubule of the kidney, thereby promoting natriuresis and glycosuria. SGLT2 inhibitors lower HbA1c by an average of 0.5%–0.8%. ⁸² Improvements in insulin sensitivity and lower insulin secretion have been observed in patients with T2D but are thought to be secondary to the decrease in glucotoxicity and shift from glucose to free fatty acids as a source of energy. ⁸³ , ⁸⁴ In T1D, SGLT2 inhibitors can reduce HbA1c, insulin requirements and body weight modestly, albeit with a higher risk of diabetic ketoacidosis (DKA). ⁸⁵

Independently of their effect on diabetes control, SGLT2 inhibitors improve cardiovascular health irrespective of the presence of diabetes, ⁸⁶ and are routinely used now in the management of heart failure even for patients with diabetes. ⁸⁶ The benefit of SGLT2 inhibition on the progression of chronic kidney disease in patients with and without diabetes has also been established. ⁸⁶ However, its ability to lower glucose diminishes with GFR and is minimal when the GFR is <30 mL/min/1.73 m². ⁸⁷

2.6.1. Effects of SGLT2 inhibitors on mitochondrial function and oxidative stress

SGLT2 inhibitors decrease oxidative stress and affect mitochondrial function in a variety of tissues, including heart, kidney and pancreas. ⁸⁸ As reviewed by Packer, ⁸⁸ these specific effects on oxidative stress and mitochondrial function appear to be critical mechanisms for their cardiovascular benefit. SGLT2 inhibitors can modulate cellular metabolism, notably inducing an increase in nutrient deprivation signals in cells (such as AMPK and sirtuins) over energy surplus signals (such as mTORC1), leading to a decrease in the generation of ROS and ER stress. These effects may in part be mediated indirectly by urinary caloric loss, ketogenesis and increased metabolism of lipids as substrates. Although SGLT2 channels are not present on β‐cells, treatment with SGLT2 inhibitors improves markers of β‐cell function, survival, oxidative stress and proliferation, likely also indirectly by reducing glucotoxicity and metabolic workload (‘β‐cell rest’). ⁸⁴ , ⁸⁹ These benefits are most pronounced when treatment is initiated early in the course of diabetes.

However, several potential direct effects of SGLT2 inhibitors have been proposed. Studies examining SGLT2 inhibitor treatment on isolated cultured cells (of various types) have observed effects on nutrient signalling in cells even in the absence of changes in environmental glucose and ketone levels, nor requiring the presence of SGLT2 channels on the cell membrane, raising the possibility that SGLT2 inhibitors may exert direct effects on relevant energy sensor pathways like mTOR. ⁹⁰ Improvements in mitochondrial function in diverse cell types have also been demonstrated (including those that do not express SGLT2), with evidence of enhanced autophagic flux and mitophagy and increased mitochondrial biogenesis and mass. ⁹¹ This may be driven by changes in energetic signals that regulate autophagy and attenuation of inflammatory and pro‐fibrotic pathways involving Nuclear Factor kappa‐light‐chain‐enhancer of activated B cells and NLRP3. ⁹¹

In clinical trials of patients with T2D and high cardiovascular risk, SGLT2 inhibitors significantly improve markers of oxidative stress and mitochondrial activity (GDF‐14 and mOTS‐C) compared to treatment with insulin and potentially greater than treatment with GLP‐1R agonists. ⁶¹ The reduction in inflammation, mitochondrial oxidative stress and fibrosis, in addition to improvements in glomerular hyperfiltration, is also hypothesised to be key mechanisms in their renoprotective benefits. ⁸⁶ , ⁸⁸

2.6.2. Safety of SGLT2 inhibitors in MIDD

The efficacy and safety of SGLT2 inhibitors in MIDD have been described in recent case reports. ⁷ , ⁹ Overall, the common side effects of this class are volume depletion and genital mycotic infections. ⁹² SGLT2 inhibitors have a low risk of hypoglycaemia. ⁹² They are associated with increased rates of DKA, ⁹² including euglycaemic DKA, particularly in patients with insulin deficiency and/or during acute illness. In theory, patients with MIDD may be more vulnerable to this due to impaired insulin secretion, necessitating caution in patients with evidence of insulinopaenia. However, the use of SGLT2 inhibitors in patients with type 1 diabetes is increasing as several studies have demonstrated that the risk of DKA can be mitigated by careful sick day counselling to interrupt the use of the medication in high‐risk states. ⁹³ This suggests that the class can also be used in patients with MIDD while ensuring sick day counselling.

2.7. Insulin

Insulin is required for the management of autoimmune T1D. It is used in T2D when lifestyle and non‐insulin medications do not achieve glycaemic targets, as T2D progresses towards absolute or relative insulin deficiency, and during periods of metabolic decompensation or pregnancy. Exogenous insulin is available in a variety of formulations that have been studied in multiple trials. It may be delivered via subcutaneous injection or insulin pumps.

2.7.1. Effects of exogenous insulin on mitochondrial function and oxidative stress

The effect of exogenous insulin on β‐cell and mitochondrial oxidative stress in patients with MIDD diabetes is unclear. In humans with T2D, exogenous insulin preserves the function of the β‐cell for several months or years, likely through a decrease in glucotoxicity, ⁹⁴ and may improve overall oxidative stress (assessed through 24 h urinary 8‐isoprostaglandin F2a). ⁹⁵ This contrasts with in vitro data showing that supraphysiological concentrations of insulin promote oxidative stress and cell death‐inducing mechanisms in MIN‐6 cells. ⁹⁶

2.7.2. Safety of insulin in MIDD

Hypoglycaemia is the most common side effect and results from excess insulin administration. Due to its anabolic effect, administration of insulin can be associated with weight gain, which may not be relevant in patients with MIDD who often have difficulty gaining weight. One case series reported the development of leg oedema in four patients with MIDD diabetes receiving insulin, of which two cases resolved spontaneously and two resolved following treatment with Coenzyme Q10 (CoQ10) for several months. ⁹⁷

3. USE OF INSULIN AND NON‐INSULIN AGENTS IN MIDD

Despite clear differences between the phenotype of MIDD diabetes, T1D and T2D, ² MIDD diabetes is often mistaken for T1D or T2D and managed according to their respective guidelines. Prior to the approval of GLP‐1 receptor agonists (2005) and SGLT2 inhibitors (2013), insulin, sulfonylureas or metformin were most commonly used as a first management approach in MIDD. ⁷ , ⁹

A systematic review by Naylor et al. ⁷ examining published cohort studies and case reports/series on patients with mitochondrial diabetes caused by the m.3243A>G mutation identified a total of 242 patients. Of these, 167 (72%) used insulin. We identified another series of five patients with MIDD diabetes of 7–47 years duration treated with insulin. ⁹⁸ There is usually no structured description of outcomes and whether insulin treatment reflects insulin dependence or lack of availability of novel pharmacological agents is unclear. The average time between diagnosis of diabetes and insulin therapy is 11 years, and 63% start insulin within 6 years after diagnosis. ⁹⁹

4. SAFETY IN PREGNANCY AND LACTATION

MIDD affects patients of childbearing age and carries a high risk of obstetrical complications for the mother and foetus. ¹⁰⁰ Compared to patients with other mitochondrial diseases, patients with the m.3243A>G phenotypic variant are at particularly high risk for gestational diabetes mellitus (GDM) and complications. ¹⁰¹ Therefore, it is important to consider the role and safety of these medications in pregnancy and breastfeeding.

Exogenous insulin, including insulin analogues, does not cross the placenta and does not have adverse consequences on the breastfed neonate. ¹⁰² It is used to manage T1D in pregnancy and is also the first‐line pharmacotherapy for GDM and T2D in pregnancy. ¹⁰² There are data on metformin and glyburide use in pregnancy; however, they are not generally recommended over insulin for diabetes in pregnancy. ¹⁰² TZDs, GLP‐1R agonists, DPP‐4 inhibitors and SGLT2 inhibitors are contraindicated during pregnancy due to possible teratogenicity and/or paucity of safety data, ¹⁰² nor are they approved during breastfeeding. GLP‐1R agonists should be discontinued 3 months before pregnancy.

No controlled studies have examined the efficacy and safety of these agents specifically during diabetes for patients with MIDD, and published reports on their experience in this setting are limited. Given the absence of robust clinical and safety data for non‐insulin medications in this setting, insulin remains the mainstay for pharmacologic treatment of diabetes in pregnancy in MIDD as well. ¹⁰⁰ , ¹⁰³ , ¹⁰⁴ , ¹⁰⁵ Metformin is often avoided due to the concern for lactic acidosis. ¹⁰

5. PROPOSED GUIDELINES

5.1. Confirmation of MIDD diagnosis

The presence of diabetes in a patient with a family history of deafness and diabetes inherited in a matrilinear manner strongly suggests the diagnosis of MIDD, in particular in the presence of other MIDD‐associated symptoms. ²⁴ Diagnosis should be confirmed by genetic testing in the blood and in other easily obtained tissues/fluids such as urine, hair and saliva. If the results are inconclusive, or if an assessment of muscle heteroplasmy is required, a muscle biopsy can be performed. ¹⁰⁶ Muscle heteroplasmy can also be assessed indirectly through urine heteroplasmy ¹⁰⁷ and other markers such as VO2 max and basal serum lactate. ¹⁰⁸

5.2. Assessment of MIDD diabetes and its complications

Guidelines for the assessment of mitochondrial diseases are discussed extensively elsewhere. ¹⁰⁹ , ¹¹⁰ The 2025 American Diabetes Association guidelines define prediabetes as a HbA1c 5.7%–6.4% (39–47 mmol/mol) and diabetes as a HbA1c ≥6.5% (48 mmol/mol) but do not mention mitochondrial diabetes. ¹¹¹ In contrast, the 2017 Guidelines from the Mitochondrial Medicine Society ¹⁰⁹ and the Newcastle Mitochondrial Disease 2020 guidelines ¹⁰ recommend measuring HbA1c at baseline and thereafter annually in asymptomatic patients. Glucose metabolism (fasting blood glucose, HbA1c and C‐peptide), renal function (urea, creatinine and urinary albumin‐to‐creatinine ratio), lactate and cardiovascular function (blood pressure, electrocardiogram and cardiac ultrasound) should be assessed at baseline and annually (or as clinically indicated) in MIDD patients.

5.3. Lifestyle, diet and nutritional supplements

A general approach to optimise physical activity, nutrition and nutritional supplements has been reviewed elsewhere and should be discussed on a case‐by‐case basis. ¹⁰⁶ As a rule, a healthy diet should be followed while catabolism and prolonged fasting should be prevented. ¹⁰⁹ No specific diet has been demonstrated to be of benefit for the prevention or management of MIDD diabetes. The ketogenic diet has shown neurological benefits in some patients with MELAS ¹¹² and a high‐fat diet is known to downregulate genes necessary for oxidative phosphorylation and mitochondrial biogenesis. ¹¹³ However, to our knowledge, the effect of this approach on mitochondrial diabetes has not been investigated.

Regular physical activity is decreased in subjects with mitochondrial diseases ¹¹⁴ and is an integral part of the management of mitochondrial diseases. ¹¹⁵ In MIDD, insulin resistance may precede the development of β‐cell dysfunction. ² Regular exercise (mild to moderate, as tolerated) appears to be safe and to improve muscle heteroplasmy and oxidative phosphorylation. ¹¹⁶ , ¹¹⁷ , ¹¹⁸ The effects on insulin sensitivity in MIDD/MELAS, however, remain to be confirmed.

The role of nutritional supplements remains inconclusive. ¹¹⁰ CoQ10 may improve hearing loss progression, ¹¹⁹ lactate elevation after exercise, ¹¹⁹ and myocardial and muscular dysfunction, ¹²⁰ but evidence for CoQ10 and L‐carnitine on β‐cell function and glycaemia is limited and conflicting. ¹¹⁹ , ¹²¹ There is currently insufficient evidence for L‐arginine in reducing stroke‐like episodes in MELAS. ¹²² A small trial has suggested the benefit of taurine supplementation in reducing stroke‐like episodes, and further study is required. ¹²³ There are no trials that have examined the impact of L‐arginine or taurine on glycaemic control in MIDD.

5.4. Diabetes ketoacidosis

MIDD diabetes can first manifest as DKA. There are no specific guidelines for the management of MIDD DKA. A case of ‘switched metabolic acidosis’ was reported in a MELAS patient in whom lactic acidosis worsened while ketoacidosis improved. ¹²⁴ The authors suggested that a rapid intracellular glucose increase following insulin administration could not be metabolised efficiently by affected mitochondria, resulting in increased lactate production. This suggests that DKA management in MIDD may benefit from ongoing lactate measurements and slow correction of hyperglycaemia.

5.5. Pharmacological approach to MIDD diabetes

We propose an empirical management approach for MIDD diabetes with therapeutic agents that in addition to lowering glucose will also (1) target the hallmark metabolic perturbations of MIDD such as oxidative stress in the pancreas, kidney and heart; (2) potentially improve insulin resistance and β‐cell function; (3) protect cardiovascular and kidney function (Figure 1, Table 2). It is important to note the differences in the pathophysiology of cardiovascular and renal complications between MIDD and in T1D/T2D. In MIDD, it reflects primarily the underlying genetic mitochondrial defect (in the absence of diabetes), followed, as diabetes develops, by the effects of chronic hyperglycaemia (reviewed further in Chanoine et al. ² ). In both MIDD and T1D/T2D, oxidative stress and mitochondrial dysfunction play a major role in the pathophysiology of cardiovascular and renal complications in addition to diabetes. ² , ⁸⁸

FIGURE 1.

Proposed algorithm for the diagnosis, assessment and management of maternally inherited diabetes and deafness (MIDD) diabetes. *Although m.3243A>G in the MT‐L1 gene accounts for more than 80% of the clinical cases, more than 22 pathogenic variants are known to cause MIDD. ¹ Genetic testing can be performed using blood, urine, hair and saliva. If the results are inconclusive, muscle heteroplasmy can be assessed directly by muscle biopsy or indirectly by urine heteroplasmy ¹⁰⁷ or markers such as VO2 max and basal serum lactate. ¹⁰⁸ **Lower doses may be necessary to avoid undesirable weight loss and/or gastrointestinal side effects. If available, consider intensification to a dual glucagon‐like peptide‐1 (GLP‐1)/glucose‐dependent insulinotropic polypeptide (GIP) receptor agonist. ***We included metformin as a first‐line therapy, with conditions. Similar to glucagon‐like peptide‐1 receptor agonist (GLP‐1RA) and sodium/glucose cotransporter 2 inhibitor (SGLT2i), it may decrease oxidative stress and insulin resistance; cardiovascular and renal benefits have been recently suggested. It is very affordable. Lactic acidosis, although rare, is a concern and we therefore suggest shared decision‐making with patients and careful monitoring for lactic acidosis. See the text for further rationale. DKA, diabetic ketoacidosis; DPP‐4i, dipeptidyl peptidase 4 inhibitors; ECG, electrocardiogram; FBG, fasting blood glucose; GFR, glomerular filtration rate; HHS, hyperglycaemic hyperosmolar syndrome.

TABLE 2.

Rationale for the recommendations of anti‐diabetic agents in patients with MIDD.

Drug	Rationale	Characteristics of the patients who would most likely benefit from the medicine	Characteristics of the patients who should be cautious or should avoid the medicine
GLP‐1R agonist (or dual GLP‐1/GIP receptor agonist)	Improves oxidative stress and mitochondrial function systemically (through direct mechanisms), cardiorenal protection, glycaemic lowering (HbA1c decrease 1.5%–2%, >2% with dual agonists)	Excess weight, CVD, heart failure, CKD, high insulin requirements	Low body weight, undernourished, low oral intake, significant baseline diarrhoea, constipation, pregnancy is planned within 3 months
SGLT2 inhibitor	Improves oxidative stress and mitochondrial function systemically (through indirect and possibly direct mechanisms), cardiorenal protection, moderate glycaemic lowering (HbA1c decrease 0.5%–0.8%)	CVD, heart failure, CKD (glucose‐lowering effect diminished when GFR <45, but can be started if GFR >20 for renoprotective benefits), high insulin requirements	Low baseline C‐peptide (consider addition of exogenous insulin given risk of DKA, emphasise sick‐day counselling such as the STOP‐DKA protocol), recurrent genitourinary infections, ESRD
Pioglitazone	Improves β‐cell oxidative stress (involving an increase in PPAR‐γ and catalase) and insulin resistance, glycaemic lowering (HbA1c decrease by 1%–1.5%), low cost	Low body weight, high insulin requirements and/or insulin resistance, atherosclerotic CVD	Heart failure, peripheral oedema, bladder cancer
DPP‐4 inhibitor	Modest glycaemic lowering (HbA1c decrease by 0.5%)	Mild hyperglycaemia	Avoid combination GLP‐1RA and DPP‐4i
Gliclazide	Potential benefits on β‐cell oxidative stress (as a free radical scavenger, not a sulfonylurea class effect), glycaemic lowering (HbA1c decrease by 1%–1.5%), low cost		Unpredictable blood sugars with frequent hypoglycaemia, use with caution if GFR <15
Metformin	May improve oxidative stress and mitochondrial homeostasis (through AMPK‐dependant and possibly independent mechanisms), glycaemic lowering (HbA1c decrease by 1%–1.5%), low cost	High insulin requirements and/or insulin resistance	Risk of lactic acidosis (use with shared decision‐making and monitor for lactic acidosis), reduce dose depending on GFR, avoid if GFR <30 or liver failure

Abbreviations: AMPK, AMP activated protein kinase; CVD, cardiovascular disease; DKA, diabetic ketoacidosis; DPP‐4, dipeptidyl peptidase 4; GFR, glomerular filtration rate; GIP, gastric inhibitory peptide; GLP‐1R, glucagon‐like peptide‐1 receptor; PPAR‐γ, proliferator‐activated receptor γ; SGLT2, sodium/glucose cotransporter 2.

The data summarised earlier suggest that both GLP‐1R agonists and SGLT2 inhibitors can be considered as first‐line agents for diabetes in MIDD (Table 3). The systemic cardiorenal benefits associated with these agents are mechanistically supported by their effects on ameliorating oxidative stress, mitochondrial function and inflammation. The choice between GLP‐1R agonists and SGLT2 inhibitors should be made based on an individual basis (contraindications, preexisting complications and weight) and adapted based on the response to treatment and the side effects. For instance, in MIDD, GLP‐1R agonists should be started at a low dose with appropriate titration to limit undesired weight loss or significant gastrointestinal dysmotility. GLP‐1R agonists and SGLT2 inhibitors have also been recently investigated in combination in T2D patients with possible additive effects on chronic complications. ¹²⁵ If accessible, then novel dual GLP‐1/GIP receptor agonists could also be considered in place of GLP‐1R agonists.

TABLE 3.

Comparison of the metabolic effects of glucagon‐like peptide‐1 receptor (GLP1‐R) agonists and sodium/glucose cotransporter 2 (SGLT2) inhibitors, based on data mainly from patients with type 2 diabetes.

	GLP‐1R agonists	SGLT2 inhibitors
Insulin response	Glucose‐stimulated insulin release	Increase renal glucose excretion with decrease insulin response
Insulin resistance	Decreases, independent from weight loss	Decreases, secondary to the decrease in plasma glucose and in glucotoxicity, and to a shift from glucose to free fatty acids as a source of energy
Antioxidant effects	Yes	Yes
Weight loss	Modest to severe, dose‐dependent	Modest
Cardiac complications	Lower risk of myocardial infarction	Lower risk of heart failure in patients with and without diabetes
Renal complications	Slows progression of diabetes kidney disease	Slows progression of chronic kidney disease in patients with and without diabetes
Contraindications	History of pancreatitis, pregnancy and lactation	Severe renal insufficiency (eGFR <30 mL/min/1.73 m2), pregnancy and lactation
Side effects	Weight loss, gastrointestinal side effects (dose‐related): nausea, vomiting, diarrhoea, injection site reactions, headache, pancreatitis and cholelithiasis	Urinary tract infections and genital mycotic infections, less common: euglycaemic DKA and hypotension

Abbreviations: DKA, diabetic ketoacidosis; eGFR, estimated glomerular filtration rate.

We include metformin as an option for first‐line therapy, albeit with conditions. It has several promising properties addressing oxidative stress and insulin resistance, was the backbone for glycaemic control in trials of other agents, and may also have cardiorenal benefits (although it is not approved in these indications). Its low cost and easy access are other key advantages when affordability is an issue that limits the use of other agents. Lactic acidosis, although rare, is a competing concern in patients with MIDD based on limited clinical data (see Table S1). Our perspective is that metformin may still be offered through shared decision‐making with the patient on its benefits and risks and with close surveillance of lactate levels. This could be compared to the growing use of SGLT2 inhibitors with monitoring and education in patients with T1D despite the associated increased risk of DKA.

We suggest that the other agents could be considered as a second‐line approach. The TZD pioglitazone has positive effects on oxidative stress, insulin resistance, and potentially prolongs β‐cell function when started early in diabetes and could have a role in MIDD. Gliclazide may have antioxidant properties in addition to a lower risk of hypoglycaemia and a more durable treatment response compared to other sulfonylureas (e.g., glibenclamide/glyburide) and is a potential option as well. DPP‐4 inhibitors have modest glucose‐lowering efficacy without cardiorenal protection but are usually well‐tolerated with comparatively lower side effects than the other agents and are weight neutral. Prior experience of the use of these agents in patients with MIDD has been reported, providing limited evidence of their efficacy and tolerability (as outlined earlier).

Sick‐day counselling for medications, particularly SGLT2 inhibitors, sulfonylureas and metformin, should be reviewed with patients to reduce the risk of complications. As patients with MIDD could progress earlier to insulin deficiency and therefore have a higher risk of DKA while on SGLT2 inhibitors, protocols such as the STOP‐DKA resource should be reviewed with patients to reduce this risk. ⁹³

Like other types of diabetes, intensive insulin therapy should be initiated in patients presenting with metabolic decompensation, namely symptomatic hyperglycaemia, DKA or the hyperglycaemic hyperosmolar syndrome (HHS). Given that absolute insulin deficiency is unlikely near the time of diagnosis of MIDD, it is possible that β‐cell function and insulin response could recover as the glucotoxicity dissipates, and several reports have suggested that insulin therapy can be de‐escalated as non‐insulin agents (particularly SGLT2 inhibitors and GLP‐1R agonists) are introduced. ⁹ , ⁶⁵ , ⁶⁶ , ⁶⁷ Insulin should also be added to the regimen upon failure to achieve glycaemic targets despite maximally tolerated therapy with non‐insulin agents, or earlier if insulinopaenia is suspected based on C‐peptide levels. Furthermore, we suggest that insulin remains the treatment of choice in those with MIDD who have diabetes in pregnancy.

We suggest that the standard glycaemic target of an HbA1c <7% be pursued when possible, with consideration for higher targets individualised to the patient and depending on the risk of hypoglycaemia, frailty and care requirements and life expectancy.

6. CAN DEVELOPMENT OF MIDD DIABETES BE DELAYED OR PREVENTED?

Although the penetrance of diabetes is high (>85%), it is not a universal characteristic of MIDD, and the age of appearance varies widely (average age at diagnosis of 37 years, range 11–68 years). ¹²⁶ An important question is whether evaluation of insulin resistance, determination of heteroplasmy in muscle, clinical assessment or changes in laboratory markers such as Homeostatic Model Assessment for Insulin Resistance or HbA1c could predict progression towards diabetes. This could lead to early interventions such as lifestyle optimisation and/or pharmacological treatment before prediabetes or diabetes develop. For example, SGLT2 inhibitors for heart failure or chronic kidney disease are associated with a 26% reduction in the risk of diabetes in individuals even without baseline dysglycaemia, ¹²⁷ prompting consideration of whether similar interventions could be applicable to patients with MIDD. Future longitudinal studies should focus on the relationship between baseline (or evolution of) muscle heteroplasmy or changes in laboratory markers and progression towards MIDD diabetes.

7. CONCLUSION

MIDD is a monogenic cause of diabetes that is underrecognised and for which no evidence‐based therapeutic guidelines are available. Over the last 20 years, novel anti‐hyperglycaemic agents with exciting characteristics have been developed for the management of T2D. In addition to glucose‐lowering, antioxidant properties, mitochondrial regulation and protective effects on the heart, kidney and pancreatic β‐cells have been identified.

We have proposed a treatment algorithm that focuses on the specific characteristics of MIDD. As the pathophysiology of MIDD becomes better understood, ² further clinical studies are needed to confirm how oxidative stress affects its phenotype and to interrogate whether the clinical efficacy of anti‐diabetic therapies is modulated by their capacity to improve oxidative stress and mitochondrial function. Incorporation of patient experiences, values and perspectives towards treatment should also be a key goal of future studies. These efforts will strengthen the translation of the mechanistic understanding of MIDD to inform precision medicine and evidence‐based recommendations for patients with MIDD.

AUTHOR CONTRIBUTIONS

AC and J‐PC are the guarantors of this work. AC researched data, co‐wrote, reviewed and edited the manuscript. DMT contributed to the discussion and reviewed the manuscript. J‐PC designed the study framework, researched data, co‐wrote, reviewed and edited the manuscript.

CONFLICT OF INTEREST STATEMENT

Ahsen Chaudhry, David M. Thompson and Jean‐Pierre Chanoine report no conflict of interest relevant to this work.

Supporting information

Data S1. Supporting Information.

Chaudhry A, Thompson DM, Chanoine J‐P. Diabetes management in maternally inherited diabetes and deafness (MIDD): A review and a proposed treatment algorithm. Diabetes Obes Metab. 2026;28(2):826‐839. doi: 10.1111/dom.70240

DATA AVAILABILITY STATEMENT

Data sharing not applicable to this article as no datasets were generated or analysed during the current study.