Clinical and Genetic Spectrum of VPS13A Disease Across India: A Retrospective Multicentre Genetically Confirmed Cohort

Introduction

Neuroacanthocytosis syndromes are rare inherited neurodegenerative disorders with movement disorders, acanthocytosis, and multisystem involvement [1]. Contemporary classification favors molecular terminology because acanthocytosis may be absent or technique-dependent [2, 3]. VPS13A disease, traditionally termed chorea-acanthocytosis, is an autosomal recessive disorder with a broad phenotype including chorea, dystonia, tics, orofacial and lingual involvement, feeding dystonia, dysarthria, dysphagia, psychiatric and cognitive symptoms, seizures, neuropathy and myopathy [3]. Early disease may mimic McLeod neuroacanthocytosis syndrome, Huntington disease, Tourette syndrome, Wilson disease, neurodegeneration with brain iron accumulation (NBIA), epilepsy, psychiatric illness, functional tic-like disorder, or atypical parkinsonism, especially when chorea is not dominant [2].

VPS13A disease has been reported worldwide and is not considered restricted to a particular ethnic or geographic population, although reliable population-specific prevalence estimates are unavailable. Recurrent variants and founder effects have nevertheless been identified in selected populations. In a Japanese series of 39 molecularly confirmed patients, an exon 60–61 deletion and c.4411C>T (p.Arg1471Ter) together represented more than half of the identified mutations and were concentrated in western Japan [4]. A 37-kb deletion involving exons 70–73 has been established as a founder variant in French Canadian families, while c.2343delA was identified in nine patients from three apparently unrelated Jewish families originating from Tunisia [5, 6].

Genetically confirmed reports from India remain limited and largely comprise individual families or small regional series, including four patients from three families of different Indian ethnic backgrounds [7, 8, 9, 10]. We therefore conducted a multicentre study to characterize the geographic and molecular spectrum of VPS13A disease across India and to examine patterns of delay to tertiary evaluation according to the initial presenting symptom.

Materials and Methods

We conducted a multicentre retrospective observational cohort of genetically confirmed VPS13A-related chorea-acanthocytosis across eight tertiary movement disorder centres in India (Supplementary Figure 1). Eligible patients were identified from clinical databases and investigator records, and data were collected between January and May 2026. All genetically confirmed cases available at participating centres were included. None of the patients included in this cohort had been previously reported in a published case report or case series. No sample-size calculation was performed. Ethical approval was obtained as per local norms (IEC #273/16.05.2023, RP-10/23, OP-20/12.01.2024). Written informed consent was obtained for publication of videos. This study is reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline for observational studies (Supplementary Table 1) [11].

Data were collected using a standardized proforma. Laboratory, electrophysiological, and imaging data were extracted from records and were not centrally standardized. Rubber-person gait was defined clinically as an unusual gait characterized by excessive truncal movements, intermittent lurching, and sudden buckling of the knees or trunk [12, 13, 14]. Imaging abnormalities were extracted from contemporaneous formal radiology reports and, where scans were available. Caudate atrophy was defined visually as reduction in the volume of the caudate head and/or body, with loss of normal caudate convexity and associated enlargement of the frontal horns of the lateral ventricles on axial or coronal images. There was no central blinded review or quantitative volumetric analysis. Acanthocyte results were abstracted from routine clinical laboratory reports performed by light microscopy. Electron microscopy was not performed, and preparation techniques and positivity thresholds were not standardized across centres. Genetic testing method, zygosity, variant location, and reported American College of Medical Genetics and Genomics (ACMG) classification were recorded. Variant localization was categorized as exonic/coding, intronic/splice-site or splice-region, mixed, or unclassifiable.

Continuous and categorical variables were summarized descriptively. Missing, unknown, or not-done values were excluded from variable-specific denominators. Diagnostic-delay comparisons were descriptive and exploratory. Selection and information bias were addressed by including all eligible genetically confirmed cases, using a standardized proforma, and reporting variable-specific denominators.

Results

VPS13A variants spectrum

Detailed variant information was available for 17 of 18 patients. Genetic testing was performed by whole-exome sequencing in 12/18 (66.7%), targeted gene panel in 3/18 (16.7%), clinical exome sequencing in 2/18 (11.1%), and single-gene study in 1/18 (5.6%) patients; variants were homozygous in 9/18 (50.0%) and compound heterozygous in 8/18 (44.4%) patients, with zygosity not available for 1/18 (5.6%) (Supplementary Figure 4). The molecular spectrum was heterogeneous, with variants distributed across VPS13A and comprising frameshift, nonsense, canonical splice-site or splice-region, in-frame deletion, missense, and multiexon deletion variants (Figure 1 and Table 2). Several variants were absent from the population and clinical databases reviewed and appeared to be previously unreported. Two variants were identified independently in more than one unrelated patient: the canonical splice-site variant c.144+1G>A in Patients 3 and 15 and the frameshift variant c.3904delG in Patients 7 and 8. The remaining identified variants were each observed in a single patient.

Figure 1

Table 2

VPS13A variant spectrum, testing method, zygosity, localization, and ACMG classification in genetically confirmed VPS13A disease.

PATIENT ID	GENETIC TEST	ZYGOSITY	REFSEQ	VPS13A VARIANT(S)	PROTEIN CHANGE	VARIANT CLASS/LOCALIZATION	ACMG/SOURCE CLASSIFICATION	GNOMAD
Patient 1	Whole-exome sequencing	Homozygous	NM_033305.3	Exon 8 c.584del	p.Leu195TyrfsTer12	Exonic/coding; frameshift	Likely pathogenic (PVS1, PM2)	Novel
Patient 2	Whole-exome sequencing	Compound heterozygous	NM_033305.3	c.1595 + 4_1595 + 7del/ c.4629_4630 + 8del	Not available/ p.Glu1543AspfsTer12	Exon-Intron junction 17 deletion/ Exon 38-intron 39 splice junction deletion	VUS (Splice AI 0.99)/ Likely pathogenic (Splice AI score 1.00)	Novel
Patient 3	Targeted gene panel	Homozygous	NM_033305.3	Intron 2 c.144+1G>A	Not available	Intronic/splice-site	Likely pathogenic (Splice AI score 0.99)	Novel
Patient 4	Whole-exome sequencing	Homozygous	NM_033305.3	Intron 26 c.2824+1G>T	Not available	Intronic/splice-site	Likely pathogenic (Splice AI score: 1.00)	Novel
Patient 5	Whole-exome sequencing	Homozygous	NM_033305.3	Exon 33 c.3802C>T	p.Arg1268Ter	Exonic/coding; nonsense	Pathogenic (PVS1, PM2)	Minor allele frequency 0.004% (gnomAD) Not reported in SAS gnomAD Reported in ClinVar 2184590
Patient 6	Whole-exome sequencing	Homozygous	NM_033305.3	Exon 45 c.5881C>T	p.Arg1961Ter	Exonic/coding; nonsense	Pathogenic (PVS1, PM2, PM3)	Minor allele frequency 0.006% (gnomAD) Minor allele frequency (MAF): 1.1 × 10–5 in SAS gnomAD Reported in ClinVar 1458994
Patient 7	Whole-exome sequencing	Homozygous	NM_033305.3	Exon 34 c.3904delG	p.Glu1302 LysfsTer13	Exonic/coding; frameshift	Pathogenic (PVS1, PM2, PP4)	Minor allele frequency 1.24e-6 (gnomAD) Not reported in SAS gnomAD Absent in ClinVar
Patient 8	Targeted gene panel	Homozygous	NM_033305.3	Exon 34 c.3904delG	p.Glu1302SerfsTer11	Exonic/coding; frameshift	Pathogenic (PVS1, PM2, PP4)	Minor allele frequency 1.24e-6 (gnomAD) Not reported in SAS gnomAD Absent in ClinVar
Patient 9	Whole-exome sequencing	Compound heterozygous	NM_033305.3	Exon 34 c.3908G>A/Intron 59 c.8212-2A>T	p.Trp1303Ter/splice effect	Mixed exonic + intronic/splice-region	Pathogenic (PVS1, PM2,PM3)/Pathogenic (PVS1, PM2, PM3)	Novel ClinVar 2064491/Novel
Patient 10	Whole-exome sequencing	Compound heterozygous	NM_033305.3	Intron 40 c.4957-1G>T/exon 54 c.7423dup	Splice effect/not recorded	Mixed exonic + intronic/splice-region	Likely pathogenic (PVS1, PM2) (Splice AI score 0.98)	Novel
Patient 11	Whole-exome sequencing	Compound heterozygous	NM_033305.3	Exon 7 c.521C>A/exon 71 c.9403delC	p.Ser174Ter/p.Arg3135Glufs*2	Exonic/coding; nonsense + frameshift	Likely pathogenic (PVS1, PM2)	Novel
Patient 12	Clinical exome sequencing	Compound heterozygous	NM_033305.3	Intron 60 c.8325+1G>A/intron 65 c.8908-1G>C	Not available	Intronic/splice-site	Likely pathogenic (PVS1, PM2, PP5) (Splice AI score 1.00)/Likely pathogenic (PVS1, PM2) (Splice AI score 0.84)	Minor allele frequency 5.77e-6 in gnomAD ClinVar 1347984 (classified variably as likely pathogenic and VUS)/Novel
Patient 13	Targeted gene panel	Not available	NM_033305.3	Not available	Not available	Not available	Pathogenic	–
Patient 14	Clinical exome sequencing	Homozygous	NM_033305.3	Exon 46 c.5996_5998del	p.Arg1999del	Exonic/coding; in-frame deletion	Pathogenic (PVS1, PM2, PP4)	Novel
Patient 15	Whole-exome sequencing	Compound heterozygous	NM_033305.3	Intron 2 c.144+1G>A/Intron 53 c.7419 + 2dup	Not available	Intronic/splice-site	Likely pathogenic (Splice AI score 0.99)/ Pathogenic (Splice AI score 0.88)	Novel
Patient 16	Whole-exome sequencing	Compound heterozygous	NM_033305.3	Exon 43 c.5514dup/exons 6–25 deletion c.[385 + 1_386-1]_[2667 + 1_2668-1]del	p.Asp1839ArgfsTer32/structural deletion	Exonic/coding + structural deletion	Pathogenic (PVS1, PM2, PM3)/Pathogenic (PVS1, PM2, PM3)	Novel
Patient 17	Whole-exome sequencing	Homozygous	NM_033305.3	Exon 26 c.2820C>G	p.Tyr940Ter	Exonic/coding; nonsense	Pathogenic (PVS1, PM2, PP4)	Novel
Patient 18	Single-gene study	Compound heterozygous	NM_033305.3	Exonic c.5247dup/c.9275G>A	p.Arg1750Thrfs*20/ p.Arg3092His	Exonic/coding	Pathogenic (PVS1, PM2, PM3)/VUS (PM2, PM3, PP3)	Novel/Minor allele frequency (0.0004%) in gnomAD Absent in SAS gnomAD Reported in ClinVar 3603267

[i] *ACMG = American College of Medical Genetics and Genomics; ClinVar = clinical variant archive; gnomAD = Genome Aggregation Database; RefSeq = Reference Sequence; SAS = South Asian; VUS = variant of uncertain significance. ACMG evidence codes are shown as reported in clinical genetic reports.

Discussion

India remains an underexplored setting for VPS13A disease, with published evidence largely confined to individual reports and small familial or regional series [7, 8, 9, 10]. This multicentre cohort broadens its documented geographic distribution across India and adds substantially to the limited molecular data available from South Asian populations. However, the observed distribution represents tertiary-referral ascertainment rather than regional prevalence. Geographic clustering or under-representation may reflect differences in clinical recognition, access to molecular testing, referral pathways, and publication practices. Because ancestry and haplotypes were not systematically examined, the present study cannot establish founder effects or population-specific enrichment within India.

The principal strength of this cohort lies in its molecular data. Twenty-three distinct VPS13A variants were identified across 17 genotyped patients, and some appeared to be previously unreported, substantially extending the known allelic spectrum of VPS13A disease beyond the handful of variants previously described from India. This scale of novel-variant discovery, together with two variants recurring independently in unrelated patients empahsizes the value of systematic multicentre molecular characterisation in a population that remains markedly underrepresented in existing VPS13A variant databases. The relatively high representation of splice-altering variants and the identification of a multiexon deletion support comprehensive molecular testing that includes splice-junction coverage and copy-number analysis [15].For non-canonical splice-region variants and variants of uncertain significance, segregation and RNA studies may demonstrate aberrant splicing, while erythrocyte chorein Western blotting can provide complementary functional evidence of VPS13A dysfunction [16, 17, 18]. However, normal chorein expression does not exclude VPS13A disease, because some pathogenic missense variants may impair protein function without causing protein misfolding or degradation, while small deletions may produce a truncated protein with electrophoretic mobility similar to that of the wild-type protein [13].

Alongside this molecular contribution, the cohort provides descriptive evidence that diagnostic delay differs systematically by presenting phenotype, a clinically actionable finding. Chorea may prompt earlier consideration of a hereditary neurodegenerative disorder, whereas patients presenting with seizures may initially remain within epilepsy-care pathways until additional movement or multisystem manifestations emerge [6, 19]. Psychiatric, neuropathic, or other non-choreic presentations may similarly lead patients through separate specialist pathways before the unifying disorder is recognized. These comparisons are descriptive and based on small numbers, but they identify a clinically relevant hypothesis for evaluation in larger cohorts.

The measured interval should not be interpreted as equivalent to the time to definitive molecular diagnosis or as an intrinsic property of VPS13A disease. It is influenced by geographical and financial access, referral practices, availability of genetic testing, and healthcare-seeking behaviour [20, 21]. Predominantly out-of-pocket genetic testing may further restrict ascertainment in India, although socioeconomic status and the source of funding for testing were not systematically recorded. The findings nevertheless suggest that awareness of VPS13A disease among epileptologists, psychiatrists, and neuromuscular specialists, not only movement-disorder specialists, may facilitate earlier referral of patients with evolving non-choreic presentations.

Despite clinically significant bulbar dysfunction, all patients continued oral feeding, including those for whom enteral feeding had been recommended. Non-acceptance appeared to reflect overlapping psychosocial and healthcare-access barriers, particularly perceived stigma, fear, and limited local support for tube-feeding care. However, nutritional adequacy, aspiration risk, formal swallowing assessments, reasons for recommending enteral feeding, and patient goals of care were not systematically recorded. These findings should therefore not be interpreted as demonstrating preserved swallowing safety or adequate nutrition among patients who continued oral intake.

Caudate-predominant atrophy was the most frequent imaging finding, whereas putaminal abnormalities and cortical and cerebellar atrophy occurred in smaller subsets. These findings are supportive but not specific and may overlap with Huntington disease and other hereditary choreas [22, 23]. Functional imaging studies have also demonstrated striatal metabolic, perfusion, and dopaminergic abnormalities in VPS13A disease, but these have not been validated as reliable discriminators from Huntington disease [24, 25, 26]. Neuroimaging should therefore be interpreted in the clinical context, with definitive distinction dependent on molecular testing.

Acanthocytes were not documented in four genetically confirmed patients despite testing. This finding should be interpreted cautiously because smear preparation, microscopy methods, positivity thresholds, and sample-handling procedures were not standardized across participating centres. Acanthocytosis is a supportive, technique-dependent finding, and the proportion of acanthocytes has not shown a consistent relationship with disease severity [27, 28]. Accordingly, the absence of documented acanthocytosis should not preclude molecular testing when VPS13A disease is clinically suspected. Together, these clinical, imaging, and molecular observations support a staged diagnostic approach in resource-variable settings, as summarized in Supplementary Figure 5.

This study is limited by its retrospective design, modest sample size, and non-standardized clinical, laboratory, and imaging assessments. Neuroimaging was interpreted qualitatively at individual centres without centralized review or predefined quantitative thresholds. Acanthocyte-testing methods could not be compared across centres, and validated longitudinal measures of movement, speech, swallowing, functional progression, and treatment response were unavailable. Recruitment from tertiary movement-disorder centres and the requirement for molecular confirmation introduce referral and ascertainment biases and may exclude patients without access to specialist care or genetic testing. Furthermore, socioeconomic, ancestry, and haplotype data were not systematically collected. Formal swallowing assessments, aspiration events, nutritional adequacy, concurrent oral and enteral feeding plans, and goals-of-care considerations were not systematically collected. Despite these limitations, this multicentre cohort broadens the geographic and molecular spectrum of VPS13A disease reported from India and identifies apparently novel and recurrent variants, while its descriptive delay patterns highlight non-choreic presentations, particularly seizures, as potential points for improving earlier recognition.

Additional File

The additional file for this article can be found as follows: