New-Onset Diabetes Presenting With Hyperosmolar Hyperglycemic State in a Lean Adolescent With Atypical Rett Syndrome Using Antipsychotics Free

A 17-year-old medically complex Hispanic female presented with new-onset diabetes. She had a history of features of RTT (negative for MeCP2 mutation screening), autism spectrum disorder, dystonia, seizure disorder, dysphagia/gastrostomy-tube dependence, neurocognitive deficits, and mood disorder/aggression requiring multiple neuropsychiatric medications.

She presented with a 1-week history of abrupt-onset fatigue, polyuria, and lethargy and a 5-lb weight loss. Laboratory testing confirmed HHS, and she tested negative for four islet autoantibodies. Her initial A1C was 11.5%, which could indicate she had elevated blood glucose levels over the preceding 3 months versus the degree of hyperglycemia in the preceding days/weeks increased her A1C rapidly to this level. Other pertinent biochemical test results are shown in Table 1. MODY testing was not pursued given its likely low probability.

On clinical examination, she was lean (BMI 17.9 kg/m², 10th percentile) without acanthosis nigricans and was noted to have marked altered mental status and poor peripheral perfusion. She was admitted to the intensive care unit and treated with intravenous fluid resuscitation, followed by hypertonic saline for concerns of cerebral edema, given her rapid deterioration in mental status. In addition, she was placed on intravenous insulin infusion at 0.05 units/kg/hour until the HHS resolved within 72 hours and gastrostomy tube feeds were resumed. As hyperglycemia persisted, she was discharged home with an intensive subcutaneous insulin regimen consisting of insulin glargine 10 units daily and insulin lispro three times daily with meals based on an insulin-to-carbohydrate ratio of 1:10 (1 unit insulin for each 10 g carbohydrate), a correction factor of 70 mg/dL, and a target blood glucose of 100 mg/dL to optimize glycemic control.

Her family history was largely noncontributory except for family members with type 2 diabetes on the paternal side. She was on multiple medications, including oxcarbazepine, clonazepam, benztropine, baclofen, dextroamphetamine/amphetamine, risperidone, sertraline, polyethylene glycol, and levonorgestrel. In particular, she was noted to have been taking risperidone for nearly 10 years to manage behavioral aggression associated with RTT.

The psychiatry team was consulted during her hospital admission given concerns for risperidone-induced HHS. However, her mother declined discontinuation of risperidone until she could be seen by her primary psychiatrist. A month after her hospital discharge, she was switched from risperidone to aripiprazole after consulting with her psychiatrist.

At her diabetes follow-up 2 months post-diagnosis, her A1C was 6.5% on a basal/bolus insulin regimen, with a total daily insulin dose of 0.72 units/kg/day. At her diabetes follow-up 6 months post-diagnosis, her A1C was 6.7%, and a repeat C-peptide test was 2.57 ng/mL, which suggested likely preserved β-cell function. Of note, the corresponding glucose with this C-peptide was not available at the time.

She had been clinically diagnosed with atypical RTT at 8 years of age because of a loss of purposeful hand skills and spoken communication and the presence of stereotypical repetitive hand movements. Mutations in Rett-related disorders, including MECP2, FOXG1, and CDKL5, were not identified. Exome sequencing (ES) done when she was 17 years of age identified a heterozygous c.1798C>T (p.R600W) pathogenic variant in the ZMYND11, which is known to cause autosomal-dominant intellectual disability (OMIM: 616083) and is characterized by developmental delay/intellectual disability (DD/ID), epilepsy, microcephaly, neuropsychiatric behavioral problems, feeding difficulties, and dysmorphia. Moreover, this variant was previously reported in several patients with DD/ID plus additional problems, confirming its pathogenicity in our patient (12,13). ES also identified a heterozygous c.127T>C (p.Y43H) variant of unknown clinical significance in IGF1R. Of note, defects in IGF1R are suggestive of potential resistance to insulin-like growth factor-1 (IGF-1) (OMIM: 270450) and characterized by intrauterine growth retardation, postnatal growth, increased plasma IGF-1 levels, and diabetes.

1. Is there an association between RTT and nonautoimmune diabetes?

2. What are the possible roles of antipsychotics in diabetes development?

3. Is there an association between RTT and HHS?

4. Is there an association between antipsychotic use and HHS?

5. What could be done differently to prevent the development of diabetes and/or presentation with HHS in patients using antipsychotics?

Here, we describe a novel case of new-onset diabetes with negative islet autoantibodies in a lean child with features of RTT presenting with hyperglycemic emergency. Although the patient had a family history of type 2 diabetes, her thin clinical phenotype and the absence of acanthosis nigricans suggested otherwise. Furthermore, her C-peptide levels at the time of presentation were lower than expected for type 2 diabetes, although one might argue that this could have been the result of transient β-cell dysfunction secondary to glucotoxicity. At the same time, she was unlikely to have type 1 diabetes, considering her nonketotic and nonacidotic presentation, negative islet autoantibodies, and C-peptide levels higher than in typical cases of type 1 diabetes. Thus, diabetes type in this case could be classified as “specific types of diabetes due to other causes” (i.e., neuropsychiatric medication–induced diabetes).

This patient will need to be followed over the long term to determine her ongoing insulin requirements. Insulin remained the primary treatment throughout her clinical course. As her glycemic control improves, weaning from insulin may be considered; however, this seems less likely to happen because of her ongoing use of an antipsychotic agent that is likely contributing to her hyperglycemia. Of note, metformin, an antidiabetic agent, was not prescribed for this patient because she did not have an elevated BMI and did not experience the rapid weight gain typically noted in patients with type 2 diabetes. Additionally, her diagnosis of diabetes, as discussed earlier, did not quite correlate with type 2 diabetes, as the physical exam was not notable for acanthosis nigricans.

Antipsychotics are known to be associated with the development of type 2 diabetes, obesity, and acute decompensation, including DKA (14). However, HHS with antipsychotics is extremely rare in both adults and children. Cerimele (15) described a case of a 39-year-old male with Asperger’s syndrome and metabolic syndrome, including hypertension and hyperlipidemia, presenting with HHS while taking risperidone. Roefaro and Mukherjee (16) reported another case of an adult male presenting with olanzapine-induced HHS.

Although the exact mechanism of HHS and/or diabetes development in patients taking antipsychotics is unclear, a plausible explanation could be impairment in insulin action. Specifically, drug-induced insulin resistance resulting from weight gain side effects and/or changes in body fat distribution or direct effect on target tissues has been noted in those using antipsychotics (14). Whicher et al. (17) reported that there are several hormones, neurotransmitter receptors, and neuropeptides orchestrating the antipsychotic-induced weight gain (17). Interestingly, leptin is significantly elevated in individuals taking antipsychotics. This drug-induced leptin resistance leads to increased appetite and weight gain. Moreover, 78% of these patients have improved glycemic control after removal or reduction in dosage of antipsychotic medication, as noted in our patient (3). Notably, our patient did not present with elevated BMI, arguing against the metabolic effect being secondary to abnormal weight gain. Alternatively, it is presumed that antipsychotic medications directly mediate β-cell dysfunction and subsequent insulin resistance, as seen in our patient. However, the contribution of ZMYND11 in making our patient prone to diabetes remains elusive.

Extensive literature review revealed that there are only four case reports of RTT and diabetes, as summarized in Table 2. This finding was mirrored by our observation that diabetes has rarely been identified in the large number of patients with RTT seen at our center. Furthermore, it has been reported that there is a possible biological association between diabetes and RTT that is mediated by MECP2, leading to impairment of insulin secretion or dysfunction at the insulin receptor level. However, we could not conclude this in our patient given the absence of genetic mutation, nor did we have functional mutation studies to further confirm. Although it is unknown whether these previously published cases with RTT and new-onset diabetes involved the use of atypical antipsychotics, it is crucial to note that neither of them presented with HHS. Thus, it is presumed that atypical antipsychotic use in our patient most likely contributed to the development of new-onset diabetes and HHS in her case.

RTT is a complex neurogenetic developmental disorder that is characterized by an initial period of developmental acquisition, followed by loss or regression of purposeful hand skills and spoken communication, an abnormal gait, and stereotypical hand movements. After the regression, a long period of plateau is followed by a later period of motor deterioration. Associated problems include growth failure, gastrointestinal problems, epilepsy, and autonomic dysfunction. Anxiety, autistic features, and behavior problems, including aggressive behaviors, may occur. These problems may require treatment with various medications. Of individuals with typical RTT, ≥95% are determined to have a mutation in the MECP2 gene. Rarely, other genetic mutations have been identified as being associated with the clinical features of RTT. The reported patient was found to have genetic mutations other than in the MECP2 gene. The relationship between her clinical features and the development of diabetes and HHS is uncertain.

Ours is the only case known to date describing a pediatric patient with features of RTT presenting with HHS and new-onset diabetes. Interestingly, our patient had an additional likely pathogenic variant in ZMYND11. To the best of our knowledge, there is no known association between ZMYND11 and diabetes. The possible contribution of variants of unknown significance in IGF1R cannot be excluded. It is possible that mutation in a receptor that is similar to the insulin receptor could be the source of insulin resistance in our patient. This possibility, combined with diabetogenic effects of her antipsychotic agent, as well as her RTT, may have triggered a severe hyperglycemic state, resulting in HHS.

In conclusion, our case suggests an association between the use of atypical antipsychotics and the development of diabetes and HHS in a patient without identifiable risk for insulin resistance or deficiency. In addition, although it may have been just coincidental to have diabetes develop in a child with atypical RTT, the few existing case reports raise suspicion for genetic predisposition.

Further studies are needed to determine the molecular underpinnings of these possible associations, such as the role of antipsychotic medications causing insulin resistance and type 2 diabetes in the absence of metabolic risk factors. Clinicians should be aware of the possible development of diabetes and hyperglycemic emergencies in children not otherwise at risk who are using antipsychotic medications.

• Multiple case reports suggest that neuropsychiatric medications are a trigger for HHS and new-onset diabetes. Thus, prescribers are highly encouraged to be cognizant of this metabolic risk and to screen for diabetes at timely intervals.

• The link between RTT and diabetes is extremely rare, and the exact pathogenesis remains to be elucidated.

• A balancing act is required to alleviate the burden of psychosis with antipsychotics while preventing associated metabolic abnormalities in RTT.

No potential conflicts of interest relevant to this article were reported.

M.V. extracted the data, reviewed the literature, and drafted the initial manuscript. D.P., A.H.-M., and D.G.G. reviewed and edited the manuscript. M.T. critically reviewed and revised the manuscript. All authors were actively involved in the discussion for this project. M.V. is the guarantor of this work and, as such, had full access to all the data in the case study and takes responsibility for the contents of this article.