Methylmalonic acidemia (MMA) is the most common autosomal recessive organic acidemia in China, with an overall incidence of approximately 1/50,000–1/100,000 in the general population, and higher incidence in regions with high consanguineous marriage rates (1). It arises from metabolic deficiencies in methylmalonyl-CoA mutase (MCM) or its coenzyme cobalamin (vitamin B₁₂) (1). Based on the age of onset, MMA is classified as early-onset (≤1 year) or late-onset (>1 year). Early-onset MMA is more prevalent, accounting for about 70–80% of childhood-onset cases, characterized by progressive deterioration and high mortality (neonatal-onset cases have an 80% mortality rate), whereas late-onset MMA is rare, with milder clinical manifestations but a risk of misdiagnosis as other neurological disorders (2, 3).

Pathogenetically, MMA is categorized by enzyme defects into MCM deficiency (Mut type, subdivided into Mut0 and Mut- based on enzyme activity) and vitamin B12 metabolic disorder (cbl type, including cblA, cblB, cblC, etc.) (4). Reported MMA-related pathogenic genes include MMACHC, MUT, MMAA, MMAB, MCEE, MMADHC, and others associated with atypical or rare MMA-complicated disorders (e.g., HCFC1, ACSF3) (4–6). Among these, MMACHC mutations are the most common cause of cblC-type MMA in Chinese populations, accounting for over 60% of genetically confirmed cases, while MUT mutations are the main cause of Mut-type MMA (5).

In terms of diagnostic technologies, recent advances have significantly improved MMA detection. Newborn screening using tandem mass spectrometry (MS) enables early identification of asymptomatic cases, with studies showing that screened patients have better long-term prognoses due to timely intervention (7, 8). Next-generation sequencing (NGS), including whole-exome sequencing (WES), has become a key tool for genetic confirmation, especially for cases with ambiguous biochemical results, as it can detect pathogenic variants in MMA-related genes with high accuracy (9). However, despite these advances, MMA remains frequently misdiagnosed or underdiagnosed due to its non-specific clinical manifestations—early-onset cases may present with seizures and developmental delay (easily confused with epilepsy or cerebral palsy), while late-onset cases may show mental and behavioral abnormalities (often misdiagnosed as immune-mediated encephalitis) (10, 11).

Against this background, this study summarized and discussed the clinical features, treatment, clinical outcome, and genetic characteristics of 9 children with MMA admitted to our hospital over nearly a decade. The novelty of this study lies in two aspects: first, it focuses on childhood-onset MMA (covering both early and late onset) in a single-center Chinese cohort, providing real-world clinical data that supplements existing epidemiological and clinical studies on MMA in China; second, it identifies a potential prognostic association between the MMACHC c.658-660del (p. Lys220del*) variant and poor outcomes, which has not been fully validated in previous small-sample studies. This work aims to help clinicians identify diagnostic clues, make early accurate diagnoses, and ultimately improve patient prognosis—with important implications for reducing medical costs and alleviating social healthcare burdens.

Methylmalonic acid is a metabolite of methylmalonyl-CoA in the metabolic pathways of isoleucine, valine, methionine, threonine, cholesterol, and odd-chain fatty acids. Under normal conditions, it is converted to succinate by methylmalonyl-CoA mutase (MCM) and vitamin B12, participating in the tricarboxylic acid cycle. Defects in MCM or vitamin B12 metabolism lead to abnormal accumulation of methylmalonic acid, propionic acid, and methylcitrate, causing multi-organ damage (nervous system, liver, kidneys, bone marrow) (4). Due to non-specific clinical manifestations, MMA diagnosis relies on biochemical (blood/urine MS) and genetic analyses (4).

In this study, we chose WES as the genetic testing method instead of whole-genome sequencing (WGS). Compared with WGS, WES has limitations such as inability to cover large fragment gene deletions/duplications, deep introns, or gene regulatory regions. However, MMA is a known monogenic genetic disorder, and its pathogenic mechanism mainly involves mutations in enzyme-coding genes in the exon regions (9). Additionally, WES has advantages in cost-effectiveness, data interpretation accuracy, and clear recommendations in clinical guidelines (9), which ensures diagnostic efficiency while minimizing over-testing and waste of medical resources.

Clinical severity of MMA varies with age at onset: earlier onset correlates with more severe disease (11). Mut0 type, typically severe and neonatal-onset, presents with acute encephalopathy (vomiting, hypotonia, acidosis, hyperammonemia, coma, seizures) and high mortality (15). Mut-, cblA, cblB, and cblH types are milder, with onset ranging from neonates to adults. Childhood-onset cases are normal at birth, often manifesting within 1 year, with metabolic crises triggered by stress (infection, hunger, vaccines, high-protein diets) - these crises cause intellectual and motor developmental delay and regression, with potential involvement of blood, liver, kidney, skin, and peripheral nerves (15).

In our study, 7 early-onset MMA cases (onset <1 year) presented with recurrent seizures, developmental delay, abnormal metabolic indicators (lactate, glucose, liver enzymes, homocysteine), EEG abnormalities (background slowing, epileptiform discharges), and cranial MRI findings (subdural effusion, meningeal thickening, white matter signals, etc.). These manifestations overlap with common pediatric neurological diseases such as idiopathic epilepsy and hypoxic–ischemic encephalopathy. However, the presence of elevated homocysteine and lactic acid (consistent with metabolic dysfunction) should prompt clinicians to suspect genetic metabolic etiologies, and blood/urine tandem mass spectrometry should be performed promptly. Notably, 1 early-onset case was asymptomatic and detected via neonatal screening—this highlights the value of expanded newborn screening, as it can identify cases before symptom onset and enable early intervention (7, 8).

Late-onset MMA is less common but prone to misdiagnosis. Our two late-onset cases included a 5-year-old with subacute mental and behavioral abnormalities and consciousness disturbances, and an 8-year-old with infection-triggered dizziness and vomiting. The 5-year-old case was initially suspected of having immune-mediated encephalitis (due to subacute onset and mental symptoms), but normal cerebrospinal fluid immune antibodies and elevated homocysteine/lactic acid ruled out this diagnosis. Previous studies have shown that late-onset MMA with mental and behavioral abnormalities should be differentiated from immune-mediated encephalitis: the latter has acute/subacute onset, similar neurological symptoms, but normal metabolic indicators, characteristic MRI signals (cortical and basal ganglia involvement), and positive cerebrospinal fluid/peripheral blood immune antibodies (11). For unclear diagnoses, careful differentiation via blood biochemistry (metabolic indicators) and imaging is essential to avoid excessive testing and reduce medical/family burdens.

MMA treatment includes acute and long-term management (10). Acute treatment involves fluid resuscitation, correcting acidosis/hypoglycemia/electrolyte imbalance, limiting protein intake, and administering L-carnitine and vitamin B12; refractory hyperammonemia and acidosis may require dialysis (10, 16). Long-term treatment differs by vitamin B12 responsiveness: responsive cases (e.g., most cbl types) receive daily 1 mg intramuscular vitamin B12 (hydroxycobalamin preferred), with betaine (100–500 mg/kg/d) for hyperhomocysteinemia, supplemented by L-carnitine (50–100 mg/kg/d), folic acid (5–10 mg/d), and vitamin B6 (10–30 mg/d). Non-responsive cases (e.g., Mut0 type) require a special formula (free of isoleucine, valine, threonine, methionine) with limited natural protein (1.0–2.5 g/kg/d; breast milk preferred for infants), long-term L-carnitine (50–200 mg/kg/d), and monitoring/supplementation of micronutrients due to protein restriction (17, 18). Symptomatic treatment includes antiepileptics (e.g., levetiracetam, as used in our seizure cases) and liver protectants (for organ damage) (19).

Clinical outcome of MMA depends on subtype, onset age, and treatment adherence (1, 3). Vitamin B12-responsive types (especially cblA) have favorable prognoses (70% healthy survival), while non-responsive types (especially Mut0) have poor outcomes (60% mortality, 40% severe developmental delay). Neonatal-onset cases have 80% mortality, while late-onset cases are milder (1, 3). In our study, 7 of 9 cases achieved favorable outcomes, which we attribute to early diagnosis (median time from onset to diagnosis: 24 days) and timely intervention (e.g., vitamin supplementation, antiepileptics, special formula). In contrast, the two poor outcomes were due to either delayed intervention (Case 7, diagnosed 1 year and 2 months after onset, leading to developmental delay) or treatment non-responsiveness (Case 1, died of refractory metabolic complications).

A key finding of our study is the potential association between the MMACHC c.658-660del (p. Lys220del*) variant and poor prognosis. Case 6 (who died) carried MMACHC compound heterozygous variants (c.609G > A/p. Trp203* and c.658-660del/p. Lys220del*), while Case 4 (with mild developmental delay) carried only the c.609G > A variant. These two cases had similar intervals from onset to diagnosis (20 days vs. 1 months) and comparable initial clinical symptoms (seizures, metabolic acidosis), but drastically different outcomes. Previous literature on the MMACHC c.658-660del variant is limited: a 2022 study by Hu et al. reported this variant in 3 of 244 Chinese MMA pedigrees, but did not analyze its prognostic significance (5). Our data suggest that this variant may be a predictor of poor outcome, but further validation in larger cohorts is needed.

Other genetic variants in our study are consistent with previous reports. The MUT c.323G > A (p. R108H) and c.982C > T (p. L328F) variants (Case 2) have been classified as LP in ClinVar, and are known to be associated with mild Mut-type MMA (3), which aligns with Case 2’s favorable outcome. The MMACHC c.217C > T (p. Arg73*) and c.365A > G (p. His122Arg) variants (Case 8) are common pathogenic variants in Chinese cblC-type MMA patients (5), and Case 8’s favorable clinical outcome is consistent with the generally favorable prognosis of late-onset cblC-type MMA (20).

In summary, MMA lacks specific clinical manifestations, with early-onset cases predominating and high misdiagnosis risk. Clinicians should suspect genetic metabolic causes in children <1 year with seizures and developmental delay, abnormal metabolic indicators (homocysteine, lactic acid, ammonia), or late-onset cases with mental and behavioral abnormalities. Prompt blood and urine mass spectrometry (for biochemical confirmation) and genetic testing (for subtype classification) are recommended to guide treatment. Advances in diagnostic technologies (e.g., newborn screening, WES) have enabled earlier intervention, which is critical for improving long-term outcomes. The MMACHC c.658-660del (p. Lys220del*) variant may be associated with a poor prognosis in MMA patients, but further studies are needed to validate this finding.

This study has several limitations that should be acknowledged: (1) Small sample size: Only 9 cases were included, which may limit the generalizability of our findings (especially the prognostic association of the MMACHC c.658-660del variant). Future multi-center studies with larger cohorts are needed to validate these results. (2) Single-center and retrospective design: The study was conducted at a single children’s hospital, and the retrospective nature may have led to missing data (e.g., incomplete genetic testing in 5 cases, lack of long-term follow-up in some cases). Prospective studies would provide more reliable data on clinical outcomes. (3) Incomplete genetic testing: Genetic results were unavailable for 5 cases, which prevented a comprehensive analysis of genotype–phenotype correlations. Additionally, WES may have missed large fragment deletions/duplications or intronic variants, which could be detected by WGS or copy number variation (CNV) analysis. (4) Limited follow-up duration: The median follow-up time was 3 months, and long-term outcomes (e.g., neurocognitive development in late childhood, renal function in adulthood) were not evaluated. Longer follow-up is needed to assess the long-term impact of MMA and its treatment.