Wolfram syndrome and diabetes mellitus in Aotearoa, New Zealand: Phenotype and response to GLP‐1 receptor agonist therapy

Abstract

Aim

To describe the clinical characteristics of patients with Wolfram syndrome (WFS) and diabetes mellitus (DM) in Aotearoa, New Zealand. Review of response to therapy in those treated with glucagon‐like peptide‐1 receptor agonists (GLP1RA).

Methods

This retrospective cohort study describes 7 patients with WFS1 genetic variants (1 patient with WFS‐like syndrome), and DM from 5 New Zealand families (age range 5–33 years, 4 females, 3 males). All are receiving insulin. In 5 patients receiving GLP1RA therapy, we described their pre‐ and post‐treatment biometric parameters, glycaemic control and visual acuity.

Results

Genetic testing identified compound heterozygous variants in the WFS1 gene (inherited from each parent) in five patients. Another two were heterozygous carriers of a single WFS1 missense variant, one of which was associated with uniparental disomy. Among patients receiving GLP1RA therapy, reductions were seen in HbA1c (mean 11.6 mmol/mol) and total daily insulin dose (mean 0.25 units/kg/day).

Conclusions

We report genotypic and phenotypic variability in association with clinical features, including age of onset and severity. GLP1RA therapy was associated with improvements in diabetic control. Longer‐term follow‐up is required to monitor for sustained benefit and for progression or improvement in other WFS clinical features.

What‘s new?

1. What is already known?

• Wolfram syndrome (WFS) is a genetic, neurodegenerative condition associated with early onset diabetes mellitus.

• In vitro studies, animal models, and individual case reports have demonstrated benefits of Glucagon‐like peptide‐1 receptor agonists (GLP1RA) in improving glycaemic control and potentially in stabilising neuro‐ophthalmic disease.

2. What this study has found?

• Improvements in HbA1c and insulin total daily dose requirements were seen in WFS patients treated with GLP1RA.

3. What are the implications of the study?

• Treatment of people with WFS with GLP1RA may be associated with improved glycaemic control and reduced insulin requirements.

1. INTRODUCTION

Wolfram Syndrome (WFS) is a rare multi‐system neurogenerative genetic condition, with an estimated incidence between 1 in 160,000–805,000. ¹ , ² The condition is also known as DIDMOAD (Diabetes Insipidus, Diabetes Mellitus, Optic Atrophy, Deafness) due to its constellation of cardinal clinical features. Diagnosis requires either two major criteria (Insulin‐dependent diabetes before age 16; bilateral optic atrophy before age 16) or one major criterion and two minor criteria. Minor criteria include: insulin‐dependent diabetes mellitus or optic atrophy after the age of 16, Diabetes insipidus, sensorineural hearing loss, neurological manifestations, abnormalities of the urinary tract and family history of WFS. ³ Other clinical manifestations may include juvenile cataracts, intellectual impairment, psychiatric disorders, bowel dysfunction and hormonal or fertility impairment. ¹ , ² , ⁴ Symptoms emerge at different ages and progress at variable rates. Phenotypic presentations are diverse, which may reflect genetic heterogeneity. ⁵ The classical form of WFS is associated with recessive, loss‐of‐function variants in the Wolfram syndrome 1 (WFS1) gene. Less frequently, dominant variants have been reported in the WFS1 gene in those with non‐syndromic phenotypes, including low‐frequency sensorineural hearing loss, optic atrophy, congenital cataract, and diabetes ⁵ (Wolfram‐like syndrome). Another subtype of Wolfram syndrome affects a smaller proportion of patients with recessive variants in the CIDS2 gene responsible for Wolfram syndrome 2 (WFS2), who have increased rates of gastrointestinal bleeding and peptic ulceration in addition to the usual phenotype of DIDMOAD. ⁴ , ⁶

WFS1 encodes Wolframin, an endoplasmic reticulum (ER) transmembrane glycoprotein expressed in pancreatic β‐cells, as well as many other cells of the nervous system including retinal ganglion, optic nerve glial cells and the inner ear. ⁷ , ⁸ Wolframin is required for ER function, including membrane trafficking, protein folding, secretion and maintaining calcium homeostasis. ⁷ The rapid insulin production required in response to fluctuating blood glucose levels makes pancreatic beta‐cells susceptible to ER stress. Lack of functional CISD2 protein is thought to disrupt the influx of calcium from the ER to the mitochondria.

No treatments are currently available to prevent or delay WFS or WFS‐like disorders. Previous work in rodent models of Wfs‐1‐deficient mice has demonstrated that GLP1RA can preserve glucose tolerance, reduce neuroinflammation, and prevent vision loss. ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ In human pre‐clinical models using pluripotent stem cell‐derived beta‐cells from individuals with Wolfram syndrome or cells silenced for WFS1 via small interfering RNA, Exenatide and Dulaglutide improved beta‐cell function and prevented apoptosis. Furthermore, in iPS cell‐derived neural precursors and cerebellar neurons from people with WFS, Exenatide improved mitochondrial function, decreased oxidative stress, and prevented apoptosis. ¹⁷

Clinical use of GLP1RA in people with WFS is limited. Amongst eight patients in four previously published case reports and a single case series, six patients had improvements in glycaemic control. ⁹ , ¹⁸ , ¹⁹ , ²⁰ Variable reductions in fasting glucose and reductions of insulin total daily dose by 20%–80% were reported. HbA1c was reduced by 1–15 mmol/mol, and increases in glucose time in range of 1%–18% were described ⁷ , ⁹ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ although not seen in all cases. ²⁰ None of these studies reported any change in weight or extra‐pancreatic features, other than one case series showing stability of optical coherence tomography (OCT) findings in 4 individuals treated with Liraglutide for 15–36 months. ²² Our study aimed to determine the glycaemic response to GLP1RA treatment in a national cohort of patients with genetically confirmed WFS (or WFS‐1 like syndrome) and diabetes mellitus in Aotearoa, New Zealand.

2. METHODS

For data collection, direct contact was made with paediatric and adult endocrinologists across New Zealand to obtain a list of patients under their care with known WFS1 variants and a diagnosis of diabetes mellitus (DM). All known patients with WFS1 variants and DM under the care of an endocrinologist in New Zealand were included in the study population. The study protocol adhered to the tenets of the Declaration of Helsinki, and informed consent for inclusion in the study was obtained from their primary clinician, with parental written consent provided on behalf of the children involved. Review of clinical characteristics, ophthalmological assessments and biochemical data (including insulin dosing, rates of hypoglycaemia, and reported adverse effects) was completed through review of medical records, and where applicable via the patients' ophthalmologists. Commencement of GLP1RA therapy (including type and maximal dose) for individual patients was at the discretion of the treating clinician. All patients in the cohort who received GLP1RA therapy were included in the GLP1RA data response analysis, except for one patient who had just commenced treatment at the time of data collection. We reviewed surrogate measures of glycaemic control, including insulin dose requirements and HbA1c, as these data were consistently available for comparison pre‐ and post‐GLP1RA treatment, unlike other parameters now commonly measured, such as glucose time in range.

2.1. Molecular investigations

Multiple diagnostic genetic testing laboratories performed testing according to the patient's location. Table 1 summarises the genetic testing locations and results. Next‐Generation Sequencing (NGS) covered exons and flanking regions of WFS1, with confirmation of identified pathogenic variants and novel variants by Sanger sequencing. Variants were annotated and interpreted according to the guidelines developed by the American College of Medical Genetics and Genomics (ACMG) and the Association for Molecular Pathology (AMP). For patients 2 and 4, cascade testing of the known familial variant was performed.

TABLE 1.

Genetic Analysis.

Patient	Laboratory	Gene	Mutational Analysis	Segregation	Pathogenicity	Previously reported	dbSNP number (if available)
1	Molecular Vision Laboratory	WFS1	NM_006005.3:c.1515C>G (p.Cys505Trp)	Maternal	LP*	Same patient reported by Hull et al. 2024 23	N/A
		WFS1	NM_006005.3:c.1699_1704dupCTCTTT, (p.Leu567_Phe568dup)	Paternal	LP*
2	Canterbury Health Laboratories	WFS1	NM_006005.3:c.1515C>G (p.Cys505Trp)	Maternal	LP	Same patient reported Hull et al. 2024 23
		WFS1	NM_006005.3:c.1699_1704dup, (p.Leu567_Phe568dup)	Paternal	LP		N/A
3	Exeter Laboratory	WFS1	NM_006005.3:c.1254_1255del (p.Phe419fs Profs* 123)	Maternal	P*	Novel	N/A
		WFS1	NM_006005.3:c.1260dup (p.Ile421fs* 122)	Paternal	P*	Novel	N/A
4	Canterbury Health Laboratories	WFS1	NM_006005.3:c.1254_1255del (p.Phe419fs)	Maternal	P	Same patient reported by Hull et al. 2024 23	N/A
		WFS1	NM_006005.3:c.1260dup (p.Ile421Hisfs* 122)	Paternal	P	Same patient reported by Hull et al. 2024 23	N/A
5	Molecular Vision Laboratory	WFS1	NM_006005.3:c.1944G>A (p.Trp648Ter)	ND	P	ClinVar data base	rs104893879
		WFS1	NM_006005.3:c.2254G>T (p.Glu752Ter)	ND	P	ClinVar data base	rs210239579
6	Centogene Rare Disease Company	WFS1	Autosomal Dominant NM_006005.3c.937C>T (p.His313Tyr)	ND	P	ClinVar database	rs886044563
7	Exeter Laboratory	WFS1	NM_006005.3:c1433G>A, (p.Trp478Ter) Location GRCh37 (hg19) Chr4:g.6302955 Single mutation plus maternal uniparental disomy	Maternal	P	ClinVar database	rs377726402

Note: Laboratory: MVL: Molecular Vision Laboratory, Hillsboro, Oregon, United States CHL: Canterbury Health Laboratories, New Zealand. Centogene: Centogene (Rare Disease Company), Rostock, Germany. Exeter: Exeter Laboratory (Royal Devon and Exeter, NHS). dbSNP Database of Single Nucleotide Polymorphisms. ND Not done N/A Not Available.

Abbreviations: LP, Likely Pathogenic; P, Pathogenic, VUS, Uncertain Significance.

^*Novel variant at the time of reporting.

3. RESULTS

3.1. Patient characteristics

A total of 7 patients with DM and genetically confirmed WFS (n = 6) or Wolfram‐like syndrome (n = 1) from 5 ethnically diverse families, including Samoan, Asian, MELAA, European, and NZ European, were characterised. In all 6 patients with Wolfram syndrome, but not in the patient with Wolfram‐like syndrome, DM was the presenting symptom. The age of DM diagnosis ranged from 2 to 15 years. Five patients with WFS1 variants had optic atrophy, and four had hearing loss (Table 2). Genotype analysis revealed two variants in the WFS1 gene in all genetically confirmed cases, except one patient with an autosomal dominant variant and another with a single variant and maternal uniparental disomy (Table 1). Eight different genetic mutations were identified, including four novel variants. In patients presenting with DM, the mean time to WFS diagnosis after DM onset was 7.7 years (range 0–22 years). GLP1RA treatment started between 2 and 9 years after diagnosis of DM. In response to GLP1RA, there was a mean reduction in HbA1c and total daily insulin dose, but no significant change in BMI (Tables 3, 4). Changes in visual acuity are summarised in Table 5.

TABLE 2.

Demographic Characteristics and Phenotype of patients with WFS.

Patient	Current Age	Sex	Ethnicity	Age of DM Diagnosis (years)	Age of WFS Diagnosis (years)	Age of GLP1RA commencement (years)	Optic Atrophy	Diabetes Insipidus	Hearing Loss	Urological Disorder	Psychological disorder	Neurological disorder
1	20	F	Pacific Peoples	8*	16	17	Y	N	Y	N	S	N
2	33	M	Pacific Peoples	9*	31	33	Y	N	N	Y	S	Y
3	13	M	MELAA	3*	8	11	Y	S	N	N	Y	Y
4	5	M	MELAA	2*	2	4	Not yet assessed	N	N	N	N	N
5	14	F	European	4*	11	12	Y	N	Y	Y	Y	S
6	22	F	Asian	15*	16	N/A	Y	N	Y	N	N	N
7	10	F	NZ European	5*	6	8	N	N	Y	N	N	Y

Abbreviations: F, Female; M, Male; MELAA, Middle Eastern/Latin American/ African; N, No; N/A, Not applicable; S, suspected (not yet confirmed); Y, Yes.

^*Antibody negative for GAD/IA2/ZnT8.

TABLE 3.

Baseline Clinical characteristics and Glycaemic control indicators (prior to GLP1RA treatment).

Baseline Characteristics before commencement of GLP1RA Treatment
Patient	Use of GLP1RA (Yes/No)	Age (years)	Last Hba1c (mmol/l)	Insulin Total Insulin dose (units/day)	Random C‐peptide (pmol/L)	Weight (kg)	Height (cm)	BMI (kg/m2)	BP (mmHg)
1	Yes	17	85 (9.9%)	115	651	106.6	163	39.6	133/85
2	Recently commenced Liraglutide 1.8 mg Sc daily – no follow up data available.	33	80 (9.5%)	85–100	60	96	172	32.2	120/75
3	Yes	11	83 (9.7%)	32	53	36.9	145.7	17.3	N/A
4	Yes	4	62 (7.8%)	14	N/A	17.6	107.7	15.4	N/A
5	Yes	12	49 (6.6%)	20	117	31.9	144.7	15.2	N/A
6	No	22	65 (8.1%)	30–40	N/A	61	150.7	26.8	102/66
7	Yes	8	55 (7.2%)	10–22	196	28.8	131	16.8	106/54

TABLE 4.

Clinical characteristics and Glycaemic control indicators post GLP1RA therapy, in 5 patients receiving GLP1RA.

Characteristics Post GLP1RA treatment
Patient	GLP1RA Drug/dose	GLP1RA Duration	Age (years)	Last Hba1c (mmol/l)	Insulin Total Insulin dose (units/day)	Random C‐peptide (pmol/L)	Weight (kg)	Height (cm)	BMI (kg/m2)	BP (mmHg)
1	Dulaglutide 1.5 mg s/c per week	20 months	20	53 (7%)	85–90	N/A	105	N/A	39.5	114/80
3	Liraglutide 1.8 mg s/c daily	16 months	13	78 (9.3%)	30–45	N/A	37.5	N/A	15.5	N/A
4	Liraglutide 0.42 mg sc daily	9.5 months	5	50 (6.7%)	2	N/A	18.4	112.7	14.5	96/52
5	Liraglutide 0.9 mg S/c daily	7 months	14	39 (5.7%)	21–25	N/A	36.8	146.7	16.2	97/63
7	Dulaglutide 1.5 mg s/c per week	48 months	10	56 (7.3%)	3–10	N/A	33.2	139.6	17	125/79

TABLE 5.

Change in Visual acuity (VA) over time.

Patient	Ophthalmic diagnosis	Year	Right Eye	Left Eye	Comment
1	Optic Atrophy (WFS1) Mild diabetic retinopathy, early macular disease.	2022	6/36	6/60	Pre GLP1RA
		2024	6/60	6/60	Post GLP1RA
2	Optic Atrophy (WFS1)	2016	6/60	6/60	Pre GLP1RA
	Absence of diabetic retinopathy	2025	6/120	6/120	GLP1RA recently started. Post GLP1RA not yet available. The severity of eye disease at baseline will limit assessment for progression.
3	Optic Atrophy (WFS1)	2019	6/12	6/12	Pre GLP1RA
		2024	6/30	6/24	Post GLP1RA
4	No formal VA measures on record, clinically no ophthalmic concerns
5	Optic Atrophy (WFS1)	2022	6/45	6/45	Pre GLP1RA
		Not yet repeated.
6	Optic Atrophy	2019 2024	6/7.5 6/9.5	6/9.5 6/9.5–2	Pre GLP1RA (not started)
7	Optic atrophy (WFS1)	2021 2024	6/6 6/6	6/6 6/6	Pre GLP1RA Post GLP1RA

Detailed phenotype characteristics are presented below.

Patient 1 is a 20‐year‐old Samoan female diagnosed with WFS at age 16 years. She was initially diagnosed with antibody‐negative Type 1 DM (T1DM) at age 8 and started insulin therapy. She had a strong family history of diabetes, with 5 paternal generations affected by T2DM. Patient 1 has one sibling also affected by WFS with DM (patient 2), and another older sibling affected by DM diagnosed in childhood, but with negative genetic testing for WFS. At age 15, patient 1 was diagnosed with optic atrophy, which led to genetic testing and her WFS diagnosis. At the time of testing, both genetic variants identified in patient 1 were novel (Table 1). She has optic atrophy, unilateral sensorineural hearing loss, anxiety, hair loss and micro‐albuminuria (Table 2). In the 10 years prior to GLP1RA commencement, she had poor glycaemic control, with HbA1c measurements between 85 and 110 mmol/mol (9.9%–12.2%). She commenced GLP1RA therapy (Dulaglutide) at age 17 years, and after 27 months, her HbA1c significantly improved to 53 mmol/mol (7%). Her insulin requirements reduced, and microalbuminuria resolved. There was no change in her BMI which remained similar to 39.6 kg/m2. Her visual acuity has continued to decline, but hearing loss remained stable.

Patient 2 is the 33‐year‐old male sibling of patient 1. He was diagnosed with presumed T1DM at age 9. He had a significant delay in the diagnosis of WFS, which was confirmed at age 31 with cascade genetic testing. His phenotypic features include progressive optic neuropathy with severe visual impairment, neurogenic bladder, erectile dysfunction, transient hand numbness, and mood disturbance. His genetic tests showed the same gene variants identified in patient 1. Patient 2 has chronic poor glycaemic control with HbA1c readings of 75–99 mmol/L (9hile on a basal/bolus insulin regime–11.2%) while on a basal/bolus insulin regimen (Lantus Glargine/Novorapid Aspart). At the time of data collection, patient 2 just commenced GLP1RA treatment, 24 years after diagnosis of DM, and had no changes to his HbA1c, necessitating a shift to automated insulin delivery.

Patient 3 is a 13‐year‐old Latin American male diagnosed with WFS at age 8. He was initially diagnosed with T1DM at age 3 years. His clinical features include optic atrophy, anxiety, developmental delay, attention deficit hyperactivity disorder (ADHD) and features of autism. Patient 3 was also born prematurely. Patient 3 started on a GLP‐1RA, Liraglutide, at age 11 when his HbA1c was 71 mmol/mol (8.6%), in addition to insulin via a Tandem T‐Slim pump. He had a period of temporary cessation of Liraglutide due to gastrointestinal symptoms. However, he recommenced the GLP1RA when his symptoms were ongoing and thus seemingly unrelated. Prior to interruption in GLP1RA therapy, his HbA1c reduced to 55 mmol/mol (7.2%), after 16 months of treatment. After stopping GLP1RA therapy, his HbA1c increased to 68 mmol/mol (8.4%). There has been some decline in his visual parameters on recent ophthalmic assessments, with progressive thinning of the retinal nerve fibre layer. Initial workup for diabetes insipidus has been inconclusive and remains under surveillance.

Patient 4 is the 5‐year‐old brother of patient 3, who was diagnosed with DM and WFS at age 2, after presenting with hyperglycaemia and genetic testing demonstrating the same variants as his brother. To date, he has not had a formal ophthalmic examination and has no other clinical features of WFS. Although he lacks a second major diagnostic criterion for WFS (i.e. confirmed optic atrophy), WFS is assumed on the basis of DM (major criteria), family history and confirmed WFS1 genetic variants (minor criteria). ³ , ²⁴ He initiated Liraglutide at age 4 when his HbA1c was 62 mmol/mol (7.8%), in addition to his usual basal/bolus Insulin therapy. After 9.5 months of continuous GLP1RA therapy, his HbA1c reduced to 50 mmol/mol (6.7%), and his insulin requirements reduced to low‐dose basal insulin only.

Patient 5 is a 14‐year‐old European female, diagnosed with T1DM at age 4. Although patient 5 has no known family history of diabetes, she was screened for monogenic diabetes and found to have two WFS1 gene variants at age 11. Her clinical features include optic atrophy, behavioural issues, anxiety, high‐frequency hearing loss, and enuresis. She started Liraglutide at age 12 years, in addition to her basal‐bolus insulin (Lantus Glargine/Novorapid Aspart). Her diabetes was well controlled at baseline, with an HbA1c of 49 mmol/mol (6.6%). After 20 months of GLP1RA therapy, she has maintained excellent glycaemic control with an HbA1c of 46 mmol/mol (6.4%). Her insulin TDD doses are largely unchanged, and she has had no progressive hearing loss. There has been some decline in her visual parameters, including colour vision and thinning of retinal nerve fibres. She is also treated with Sodium Valproate for potential neuroprotective effects in WFS. ⁴

Patient 6 is a 22‐year‐old Chinese woman with sensorineural hearing loss diagnosed in her first 2 years of life. Following the subsequent development of DM at age 15, optic atrophy and premature ovarian insufficiency, she underwent WFS genetic testing at age 16. Her genetic tests were consistent with an autosomal dominant WFS1‐like disorder due to a single pathogenic variant in the WFS1 gene. Her diabetes is currently managed with basal/bolus insulin (Lantus Glargine/Novorapid Aspart). She has not started GLP1RA therapy. She has had some decline in her vision and thickness of retinal nerve fibres on serial ophthalmic examinations (Table 5).

Patient 7 is a 10‐year‐old NZ European girl, diagnosed with antibody‐negative DM at age 5 and WFS at age 6. She has no affected family members. Her clinical features include bilateral sensorineural hearing loss, hypotonia, and mild optic disc pallor. Patient 7 commenced Dulaglutide at age 8, when her HbA1c was 55 mmol/mol (7.2%), in addition to her basal/bolus insulin therapy. Her HbA1c remained close to target (<53 mmol/mol, <7%) after 48 months of GLP1RA therapy, and she had a 59% reduction in her insulin TDD. She reported a reduction in hypoglycaemic episodes, and her neuro‐ophthalmic assessments were stable.

4. DISCUSSION

This national cohort of seven patients with WFS1 variants and DM from five pedigrees includes glycaemic responses from five individuals treated with GLP1RA. The mean duration of GLP1RA therapy in these patients was 20.1 months at the time of reporting. (Table 5). Excluding patient 2, who was initiated on a GLP1RA after a 24‐year delay from DM diagnosis without glycaemic impact, the other 5 patients, initiated on GLP1RA (between 2 and 9 years after DM diagnosis), had a mean reduction in HbA1c of 11.6 mmol/mol (1.1%) (Table 6). Only 1/5 GLP1RA‐treated patients had an HbA1c at target (≤53 mmol/mol, 7%) at baseline. Of the remainder, 2/4 (50%) achieved their target HbA1c after GLP1RA treatment, and all 4/4 (100%) achieved an HbA1c of <58 mmol/mol (7.5%). Among the GLP1RA‐treated patients, the average reduction in insulin total daily dose (TDD) was 7.9 units/day, or 0.25 units/kg/day (ranging from −27.5 to +6.5 units/day, or − 0.7 to +0.16 units/kg/day) (Table 6). There was little impact of GLP1RA on the BMI of all 5 treated patients.

TABLE 6.

Summary of all patients with WFS1 variants and DM, response to GLP1RA.

	Mean	Range
Time between DM diagnosis and WFS diagnosis* (years)	7.7	0–22
Time between DM diagnosis and GLP1RA commencement (years)	6	2–9
Duration of time on GLP1RA (months)	20.1	7–48
Change in HbA1c (mmol/mol, %)	−11.6 (1.1%)	−32 to +1 (−5.1 to +2.2%)
Change in total daily insulin dose (units/day)	−7.9	−27.5 to +6.5
Change in insulin total daily insulin dose (units/kg/day)	−0.25	−0.7 to +0.16

^*Excluding patient 6 with Autosomal Dominant Wolfram‐like Syndrome.

Our findings suggest that GLP1RA therapy may be associated with improved glycaemic control in patients with WFS1 variants and DM. In previously published case reports of WFS patients treated with GLP1RA, reductions in insulin TDD of up to 70% have been reported. ⁸ , ²⁰ In our cohort, the mean reduction in insulin TDD was 28%, with one patient (patient 4) achieving an 86% reduction. It is possible, given this patient's very low residual insulin requirement (2 units/day), that he may ultimately cease insulin therapy. Other case reports of GLP1RA use in WFS suggest this treatment may increase c‐peptide levels and insulin responses; however, serial c‐peptide measurements were unavailable in our population. ¹⁵ A lack of HbA1c response in patient 2 (not included in the analysis) could reflect GLP1RA initiation after a long duration of established DM or only recent initiation with short treatment exposure.

The beneficial effects of GLP1RA have been clearly established in T2DM for improving glycaemic control. There is also increasing evidence suggesting benefits for patients with T1DM. Meta‐analyses of patients with T1DM treated with GLP1RA therapy have shown reductions in body weight, blood pressure, HbA1c, and insulin requirements. ²⁵ , ²⁶ Two trials have indicated that initiation of GLP1RA in T1DM may help preserve beta cell function. ²⁵ , ²⁷ In patients with T1DM treated with GLP1RA, benefits were seen without an increase in severe hypoglycaemia, diabetic ketoacidosis, or severe adverse events. ²⁸ This has similarly been observed in case reports involving GLP1RA‐treated patients with WFS. In our population, increased rates of hypoglycaemia were not observed, and GLP1RAs were well tolerated. We did not observe clinically significant weight loss (mean reduction in BMI of 0.32 kg/m²), consistent with previous case reports of WFS patients receiving GLP1RA. ⁷ , ⁹ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ Whilst BMI reduction has been observed in patients with T1DM and T2DM treated with GLP1RA, ²⁹ in these populations, GLP1RA have generally been commenced in patients with elevated BMI. In our population, only one patient receiving GLP1RA had an elevated BMI; 4 patients had a low baseline BMI. Whilst improvements in glycaemic control have been associated with weight reduction in other DM patient populations, our findings may suggest that some improvements in glycaemic control with GLP1RA use in WFS patients may be weight‐independent.

GLP1RA therapy has been associated with neuroprotective benefits in WFS. Research supports a role for delaying the onset or progression of optic complications without clear evidence of improving established disease. ⁸ , ¹³ , ¹⁴ In our cohort, no improvements in ophthalmologic or neurological complications have been reported, and most patients have experienced some progression (Table 5). The ophthalmic manifestations of patients with WFS in Aotearoa have recently been described by Hull et al. ²³ As optic atrophy is slowly progressive, and our study's follow‐up period was relatively short, any altered trajectory of neuro‐ophthalmic disease progression may not have been detected.

The optimal timing for commencement of GLP1RA therapy in WFS has not been clearly established. Given the potential to reduce ER‐mediated cell stress and destruction, it is plausible that initiating GLP1RA therapy before the onset of organ dysfunction could delay the onset of clinical disease. All patients in our cohort had established DM at the time of WFS diagnosis, and in the patient with WFS‐like syndrome (patient 6), optic atrophy preceded DM diagnosis. Some animal models suggest that GLP1RA treatment may delay the onset of diabetes and protect against the development of vision loss. ⁸ If affected individuals are known, genetic testing of family members could enable empiric treatment of affected relatives in childhood, thereby enabling assessment of whether these complications could be prevented.

Limitations of this study include its retrospective, observational nature and the differing baseline severity of WFS complications and GLP1RA regimens in treated patients. The small cohort size and the requirement for individual patients to consent to non‐standard medical therapy may introduce selection bias. Variables, including adherence to other diabetes management regimens over time, differences in concomitant therapies and diabetes medications, and variability in the frequency and intensity of clinical follow‐up between patients, can confound analyses. Small population sizes limit the interpretation of statistical significance and the assessment of causality. Large randomised clinical trials with GLP1RA in WFS would be required to confirm improvement in glycaemic control and evaluate the impact on other manifestations of WFS. Given the rare nature of WFS, recruitment for such a study may prove logistically challenging. International rare disease databases may enable larger cohort studies or combined analyses.

5. CONCLUSION

We describe genetic and phenotypic variability in WFS manifestation, with overall good response in 5 of 6 patients with WFS1 variants and DM treated with GLP1RA (reduced HbA1c and total daily insulin dose).

FUNDING INFORMATION

No sponsorship or funding has been required for this study.

CONFLICT OF INTEREST STATEMENT

The authors have nothing to report.

L’Amie A, Vincent AL, Jefferies CA, et al. Wolfram syndrome and diabetes mellitus in Aotearoa, New Zealand: Phenotype and response to GLP‐1 receptor agonist therapy. Diabet Med. 2026;43:e70405. doi: 10.1111/dme.70405