Algorithmic approach to the diagnosis of mitochondrial disorders: integrating clinical, biochemical, and genomic data

Introduction

Mitochondrial disorders (MDs) represent one of the most prevalent groups of inherited metabolic diseases, with an estimated incidence of approximately 1 in 5,000 live births and a carrier frequency close to 1 in 200 individuals. These disorders arise from defects in mitochondrial oxidative phosphorylation, affecting the cell’s ability to generate energy efficiently. Because mitochondria are essential for the function of energy-dependent tissues, MDs can involve multiple organ systems, including the central nervous system, skeletal muscle, heart, liver, and endocrine glands [1].

The clinical spectrum is highly variable, ranging from isolated organ involvement to severe multisystemic presentations with an early onset and a progressive course. This remarkable heterogeneity reflects the dual genetic origin of mitochondrial function, with disease-causing variants found both in mitochondrial DNA (mtDNA) and in nuclear genes encoding mitochondrial proteins. As a result, the diagnostic process remains particularly challenging, often requiring the integration of clinical, biochemical, and molecular data [2-3].

In this context, a multidisciplinary and algorithmic diagnostic approach can optimize patient evaluation and improve the identification of underlying molecular defects. The present study summarizes our experience with 240 patients clinically suspected of mitochondrial disease, focusing on the integration of clinical, biochemical, and genomic data into a structured diagnostic framework. This approach aims to illustrate a practical model for the evaluation of mitochondrial disorders and to emphasize the utility of an algorithm-based investigation in guiding genetic testing and diagnostic interpretation.

Materials and methods

Between March 2021 and October 2024, a total of 240 patients clinically suspected of mitochondrial disease were evaluated at the Institute of Mother and Child, the national referral center for rare diseases in the Republic of Moldova. Patient inclusion was based on the modified Nijmegen Mitochondrial Disease Scoring System (NMDS) [4], a validated tool designed to estimate the likelihood of mitochondrial dysfunction. The NMDS integrates clinical domains (neuromuscular, central nervous system, and multisystem involvement) with biochemical and neuroimaging findings, as well as muscle biopsy results, to generate a cumulative score reflecting the probability of mitochondrial disease. While the original scoring system also includes histopathological data from muscle biopsy, this component was not available for evaluation in our cohort. According to the NMDS criteria, a score of 1 indicates that mitochondrial disease is unlikely, scores of 2–4 suggest a possible disorder, values between 5 and 7 correspond to a probable disease, and scores ≥8 are consistent with a definite mitochondrial disease diagnosis. In this study, only patients with a total NMDS score ≥3 were included in the analysis. Written informed consent was secured from participants or their legal representatives following ethical guidelines.

Based on subsequent clinical and molecular evaluations, patients were classified into three groups: Group 1 – individuals with confirmed or likely mitochondrial involvement (n = 37); Group 2 – patients diagnosed with other genetic disorders (n = 44); and Group 3 – those remaining without a definitive genetic diagnosis after testing (n = 159).

All participants underwent a detailed clinical assessment focusing on neurological, neuromuscular, and multisystem manifestations, as well as family history and maternally inherited traits. Laboratory evaluation included an initial panel of routine tests: complete blood count, liver and kidney function tests, serum electrolytes, blood gases, lactate, ammonia, and creatine kinase. To further investigate potential inborn metabolic disorders, secondary screening was performed when indicated, including plasma amino acid profiling, acylcarnitine analysis, and urinary organic acid quantification. Neurophysiological assessments, such as audiometry, electromyography, and electroencephalography, as well as brain imaging by MRI or CT, and electrocardiography were conducted according to clinical need.

Genetic investigations followed a structured, stepwise approach aimed at detecting both mitochondrial and alternative genetic etiologies. All 240 patients with an NMDS ≥3 underwent initial screening using High-Resolution Melting qPCR (qPCR-HRM) targeting seven common pathogenic mitochondrial DNA mutations (m.3243A>G, m.8344A>G, m.8993T>G/C, m.13513G>A, m.3460G>A, m.11778G>A, m.14484T>C). In cases where mutations were detected, heteroplasmy was assessed by PCR-RFLP, with densitometric analysis of electrophoresis gels performed using ImageJ to provide a semi-quantitative evaluation. Selected patients with higher Nijmegen scores (≥6) underwent targeted Sanger sequencing of key mitochondrial coding regions, including genes encoding respiratory chain complexes and mitochondrial RNAs, and all identified variants were systematically interpreted and classified in accordance with American College of Medical Genetics and Genomics (ACMG) guidelines. POLG hotspot sequencing was performed for individuals with phenotypes suggestive of polymerase γ-related disorders.

For patients ultimately diagnosed with non-mitochondrial genetic conditions, most molecular data were obtained from next-generation sequencing (NGS) performed in accredited external laboratories, including whole-exome sequencing (WES) or targeted gene panels, tailored to the clinical phenotype. These analyses were conducted using standardized pipelines and quality control measures, ensuring reliable detection of pathogenic variants, likely pathogenic variants, and variants of uncertain significance (VUS). A smaller subset underwent whole-genome sequencing when clinically indicated. All identified variants were interpreted according to ACMG criteria, with reference to international databases such as ClinVar, OMIM, PanelApp, and dbSNP.

To integrate the diverse clinical, biochemical, instrumental, and genetic data into a coherent diagnostic framework, a phenotype-driven, tiered laboratory diagnostic algorithm for patients suspected of mitochondrial disorders was developed (Figure 1). The algorithm is structured to guide systematic decision-making by progressing sequentially from initial clinical evaluation, through targeted biochemical and instrumental assessments, to stepwise genetic testing. Its design emphasizes the stratification of patients based on Nijmegen score thresholds, clinical phenotype, and laboratory findings, allowing for the identification of primary mitochondrial disorders, guidance of appropriate molecular investigations, and recognition of alternative genetic etiologies. The algorithm thus serves as a practical framework to optimize diagnostic yield and support clinical interpretation.

Statistical analyses were conducted using SPSS (IBM SPSS Statistics, Version 27.0). Demographic, clinical, biochemical, and genetic variables were summarized using descriptive statistics, with continuous data presented as mean ± standard deviation (SD) or median with interquartile range (IQR), as appropriate, and categorical variables as absolute frequencies and percentages. Comparisons of categorical variables were performed using the Chi-square or Fisher’s exact test, while differences in continuous variables among groups were evaluated using the non-parametric Kruskal-Wallis H test, followed by post hoc pairwise comparisons with Bonferroni adjustment where applicable. A p-value < 0.05 was considered statistically significant.

Results

The study cohort included 240 patients, comprising 100 males (41.7%) and 140 females (58.3%). The mean age at study enrollment was 4.02 ± 4.37 years, whereas the mean age at onset of clinical manifestations was 1.14 ± 2.74 years, reflecting that the majority of patients exhibited symptomatology during early childhood. A positive family history suggestive of inherited neurometabolic conditions was observed in 22.9% of cases. Additionally, 28.7% of participants presented with perinatal complications, encompassing prematurity, intrauterine growth restriction, and neonatal hypoxic events.

During the follow-up period, 12 patients (5%) died, with a mean age at death of 2.52 ± 4.47 years. Mortality was disproportionately elevated in individuals with mitochondrial involvement (Group 1: 13.5%) relative to patients with alternative genetic disorders (Group 2: 6.8%) or those remaining without a definitive diagnosis (Group 3: 2.5%; p = 0.018). Furthermore, Group 1 patients exhibited an earlier onset of clinical manifestations (mean 0.98 ± 0.76 years) and demonstrated a more aggressive trajectory of multisystem involvement, underscoring the heightened disease burden in this population.

Clinical characteristics.

Neuromuscular impairment represented the hallmark of the studied cohort, being identified in 94.2% of patients, with muscular weakness (77.1%) and developmental delay (88.7%) emerging as the most prevalent features. Multisystemic involvement was documented in 86.7% of individuals, underscoring the systemic and progressive nature of these disorders. Endocrine and growth disturbances were noted in 69.2% of cases, followed by ophthalmologic manifestations (42.1%), gastrointestinal symptoms (28.7%), hematological abnormalities (27.1%), cardiovascular involvement (19.2%), renal impairment (6.3%), and auditory deficits (5%).

When stratified by diagnostic category, patients with mitochondrial involvement (Group 1) exhibited a distinct and more severe clinical profile, marked by a significantly higher prevalence of severe neuromuscular dysfunction (40.5% vs. 15.9% in Group 2 and 12.5% in Group 3; p < 0.001), developmental regression (27.0% vs. 9.0% and 4.4%; p < 0.001), ophthalmic manifestations including ophthalmoplegia (64.8% vs. 40.9% and 37.1%; p = 0.009), and cardiovascular involvement (35.2% vs. 20.4% and 15.1%; p = 0.020).

Seizures represented one of the most prevalent neurological manifestations among the genetically confirmed groups, occurring in 70.3% of patients with mitochondrial involvement and 72.7% of those with alternative genetic disorders, compared to 47.7% in undiagnosed individuals (p = 0.002). By contrast, neurodevelopmental and behavioral impairments were significantly more common in patients with non-mitochondrial genetic conditions (88.6%) and in undiagnosed cases (83.0%) relative to the mitochondrial group (67.5%; p = 0.038). Dysmorphic features were also observed more frequently in non-mitochondrial (63.6%) and undiagnosed patients (53.4%) than in individuals with mitochondrial involvement (40.5%), although this difference did not achieve statistical significance (p = 0.116).

Biochemical findings.

Comprehensive biochemical profiling revealed consistent deviations from reference ranges across multiple metabolic parameters. Elevated serum lactate levels were recorded in 44.6% of patients, with a significantly higher prevalence among individuals with mitochondrial involvement (Group 1: 91.9%) compared with those harboring non-mitochondrial genetic disorders (Group 2: 36.4%) and undiagnosed patients (Group 3: 35.8%; p < 0.001). Analysis of continuous lactate concentrations using the Kruskal-Wallis H test confirmed significant differences across groups (H = 32.667, df = 2, p < 0.001). Post-hoc pairwise comparisons with Bonferroni adjustment indicated significantly higher lactate levels in Group 1 compared with both Group 2 and Group 3 (p < 0.001), while no significant difference was observed between Groups 2 and 3.

Plasma ammonia elevation was identified in 17.5% of the overall cohort, occurring most frequently among patients with mitochondrial disease. Increased creatine kinase (CK) and lactate dehydrogenase (LDH) activities were observed in 19.1% and 32.5% of cases, respectively, while transaminase abnormalities (ALT and AST) were noted in 20.0%. Electrolyte imbalances, primarily involving alterations in sodium, potassium, and calcium concentrations, were identified in 21.3% of patients.

Notably, hyperalaninemia, evaluated in a subset of cases, was detected in 7.5% of the overall cohort, with a marked predominance among patients with mitochondrial involvement (32.4%) compared with those in the non-mitochondrial (9.1%) and undiagnosed groups (1.3%; p < 0.001), highlighting its potential relevance as a discriminative biochemical indicator of mitochondrial dysfunction. Perturbations in acylcarnitine profiles were detected in 12.1% of tested patients, while urinary organic acid abnormalities were identified in 13.8%, reflecting subtle secondary metabolic alterations within the cohort.

Instrumental investigations.

Neurophysiological and imaging assessments provided further insight into the multisystem involvement of the studied cohort. Electroencephalographic modifications were identified in 45.0% of patients, whereas electromyographic abnormalities were observed in 16.7% of evaluated individuals. Neuroimaging assessments demonstrated structural or signal alterations in 49.6% of patients, with a significantly higher prevalence among those with mitochondrial involvement (75.0%) compared with non-mitochondrial (68.0%) and undiagnosed patients (38.0%; p < 0.001).

Within the mitochondrial group, the most frequently encountered neuroimaging findings were cerebral and cerebellar atrophy (40.5% vs. 15.9% in Group 2 and 14.4% in Group 3; p = 0.001), basal ganglia abnormalities (16.2% vs. 2.3% and 1.2%; p < 0.001), as well as distinctive features consistent with Leigh syndrome, which were exclusively observed in this group.

Cardiac evaluation by electrocardiography (ECG) revealed significant abnormalities in 14.5% of the total cohort, with a markedly higher prevalence in patients with mitochondrial involvement (29.7%) relative to non-mitochondrial (15.9%) and undiagnosed individuals (10.7%; p = 0.012), underscoring the systemic and multi-organ nature of disease in this population.

Consistently, the distribution of Nijmegen Mitochondrial Disease Scores, which integrate clinical, biochemical, and instrumental criteria, reflected this diagnostic pattern – high scores were observed in 27.0% of mitochondrial, 6.8% of non-mitochondrial, and 1.3% of undiagnosed cases (p < 0.001), as illustrated in Figure 2.

Genetic insights.

Initial molecular screening using the laboratory diagnostic algorithm for patients suspected of mitochondrial disorders via qPCR-HRM identified a subset of patients harboring established pathogenic mtDNA point mutations, including m.3243A>G (n = 3), m.8993T>G (n = 2), m.3460G>A (n = 2), and m.11778G>A (n = 1). Subsequently, 82 patients with Nijmegen Mitochondrial Disease Scores ≥6 underwent targeted Sanger sequencing of the mitochondrial coding regions as part of this stepwise diagnostic algorithm. This approach revealed an additional 24 individuals carrying variants categorized as pathogenic, likely pathogenic, or of uncertain significance, indicative of mitochondrial involvement. Most identified variants were located within genes encoding subunits of the respiratory chain, predominantly affecting Complex I (34.7%), followed by Complex V (26.0%) and mitochondrial RNA genes (21.7%), while alterations involving Complexes III and IV were comparatively less frequent. The spatial distribution of all identified variants across the mitochondrial genome is depicted in Figure 3, illustrating their clustering within key respiratory chain complexes.

Notably, pathogenic and likely pathogenic variants were enriched in genes encoding NADH dehydrogenase subunits (ND1–ND6), supporting their established role in mitochondrial dysfunction. Moreover, individuals carrying variants within Complex V of the mitochondrial respiratory chain demonstrated significantly elevated Nijmegen scores (χ² = 8.83, p = 0.032), suggesting a stronger clinical expression consistent with mitochondrial impairment.

Building upon the laboratory diagnostic algorithm for patients suspected of mitochondrial disorders, those individuals who underwent targeted Sanger sequencing of the mitochondrial genome but yielded no identifiable mtDNA abnormalities were further evaluated. Among this subset, six patients were deemed eligible for focused POLG testing via Sanger sequencing with capillary electrophoresis, which identified two positive cases. Additionally, three patients with mitochondrial involvement were identified through NGS as carrying nuclear variants affecting TWNK, DGUOK, and ETHE1, bringing the total number of individuals with confirmed mitochondrial pathology to 37, including those detected via qPCR-HRM, Sanger sequencing, and POLG analysis.

In contrast to patients with confirmed mitochondrial involvement, the genetic evaluation of Group 2 encompassed 44 individuals exhibiting non-mitochondrial disorders, originating from 42 unrelated families, with molecular data derived from high-throughput sequencing approaches, including WES and targeted gene panels. Across this subgroup, 49 molecular variants were identified, including 33 classified as pathogenic or likely pathogenic and 16 as variants of uncertain significance. Patients were distributed among five main diagnostic categories: Syndromic and Sensory Disorders (n = 15, 34.1%; e.g., TSEN54, SOX11, SMARCAL1, etc.), Inborn Errors of Metabolism (n = 12, 27.3%; e.g., ALDH7A1, PSAT1, G6PD, etc.), Channelopathies (n = 8, 18.1%; e.g., CACNA1A, SCN2A, SCN8A, etc.), Neuromuscular Disorders (n = 4, 9.1%; e.g., DARS2, GRIK2, TCAP, etc.), and Chromosomal Microdeletion Syndromes (n = 5, 11.4%; e.g., 4p16.3, 7q11.23, 9p13.1p12, etc.). Autosomal inheritance predominated, with 20 cases (43.5%) exhibiting recessive patterns and 18 (39.1%) dominant, whereas X-linked transmission was comparatively infrequent. Missense variants represented the majority of alterations (63.3%), followed by splice site, nonsense, and frameshift changes, with chromosomal microdeletions identified in 11.4% of patients. No individual gene was implicated in more than 4.5% of cases, highlighting the pronounced genetic heterogeneity characterizing this cohort.

Discussion

The diagnostic assessment of mitochondrial disorders (MDs) remains one of the most intricate challenges in clinical genetics, primarily due to their marked phenotypic and genotypic heterogeneity and the substantial overlap with other metabolic or neuromuscular conditions. In the present study, we applied an algorithmic, phenotype-driven strategy that integrated clinical, biochemical, and molecular data to enhance diagnostic precision in patients with suspected mitochondrial pathology. Central to this workflow was the use of the modified Nijmegen Mitochondrial Disease Scoring System (NMDS) as an initial stratification tool. When applying a threshold of NMDS ≥6 to guide molecular testing, the score exhibited high sensitivity (97.3%, 95% CI: 85.8-99.9%) and negative predictive value (99.2%, 95% CI: 95.9-99.98%), ensuring that clinically relevant cases were effectively captured, while maintaining moderate specificity (64.5%, 95% CI: 57.5-71.1%) and a positive predictive value of 33.3% (95% CI: 24.6-43.1%). These performance metrics emphasize the NMDS’s strength as a sensitive pre-screening instrument that minimizes false negatives and optimizes case prioritization for molecular analysis, thereby demonstrating its practical utility in diagnostic workflows.

The clinical spectrum observed across our cohort reflected the characteristic multisystemic nature of mitochondrial disorders. Neuromuscular and neurodevelopmental features predominated, aligning with the energy dependency of these tissues. Importantly, individuals within the mitochondrial group exhibited earlier disease onset, a higher frequency of severe neuromuscular dysfunction, and increased mortality, consistent with previous reports emphasizing the prognostic value of early multisystem involvement [5, 6]. These findings validate the integration of NMDS clinical domains as an effective preliminary filter to prioritize cases for molecular evaluation.

From a biochemical perspective, our findings underscored the diagnostic value of serum lactate and alanine as key discriminative indicators of mitochondrial dysfunction. Complementarily, instrumental investigations such as brain MRI and EEG further substantiated the multisystemic nature of the disease, revealing characteristic neuroimaging patterns and electrophysiological abnormalities consistent with mitochondrial involvement.

The genetic component of the diagnostic algorithm proved pivotal in establishing definitive molecular diagnoses and delineating the spectrum of mitochondrial involvement across the study group. Through the initial qPCR-HRM screening of the entire cohort, pathogenic variants were identified in 8 patients (3.3%), enabling the rapid detection of recurrent mtDNA mutations. Subsequent targeted sequencing of 82 individuals with higher Nijmegen scores or suggestive biochemical profiles yielded an additional 24 positive cases (29.3%), while POLG gene analysis, performed in five clinically indicative cases, identified two further molecularly confirmed patients (40%). The complementary application of next-generation sequencing in previously unresolved individuals revealed three additional diagnoses, culminating in a total of 37 patients (15.4%) with confirmed mitochondrial involvement. This diagnostic yield is comparable to previous studies that applied combined nuclear and mitochondrial genome sequencing. For example, Grigalionienė et al. (2023) reported 13.3% mitochondrial cases and an overall 21.7% diagnostic rate among 83 patients, using Sanger sequencing of the mitochondrial genome followed by WES in unresolved cases [7]. Similarly, Tolomeo et al. (2021) identified 25% positive cases in 111 patients through complete mtDNA sequencing and WES, emphasizing the value of integrated approaches [8]. In a larger study of 319 families, Schon et al. (2021) achieved a 31% diagnostic rate using WES after excluding common mtDNA and POLG variants [9]. Likewise, van der Ven et al. (2021), who analyzed both nuclear and mitochondrial genomes in 491 neuropediatric cases, found 6% mitochondrial and 45% non-mitochondrial diagnoses, demonstrating the clinical overlap between these groups [10]. In broader cohorts, Rouzier et al. (2024) reported 16% mitochondrial and ~20% overall diagnoses among more than 2000 patients using mainly targeted panels (~400 genes) [11], while Liu et al. (2025) achieved 38% mtDNA-positive cases and ~30% with nuclear gene variants affecting mitochondrial function in a Chinese pediatric cohort, using whole-exome and whole-mtDNA sequencing [12].

These findings collectively highlight that the use of Next Generation Sequencing substantially improves the diagnostic yield, enabling the identification of both mitochondrial and alternative genetic causes. Consistent with these observations, in our study we obtained an overall diagnostic rate of approximately 34% among 240 patients clinically suspected of mitochondrial disease, by integrating nuclear and mitochondrial genome analyses.

Beyond diagnostic yield, the structured, multi-tiered algorithm employed in this study offers several pragmatic advantages that reinforce its applicability in routine clinical practice. By combining sequential biochemical and molecular steps calibrated to NMDS stratification, the workflow enables a rational allocation of laboratory resources, mitigating both the overuse of comprehensive genomic sequencing and the risk of underdiagnosing clinically meaningful cases. This tiered architecture ensures that high-yield genetic tests are preferentially directed toward individuals with the strongest clinical and biochemical signals, thereby improving cost-effectiveness, shortening diagnostic timelines, and reducing diagnostic odysseys commonly experienced by patients with heterogeneous neurometabolic presentations.

Despite its strengths, this study presents several limitations that warrant consideration. The unavailability of muscle biopsy data prevented histopathological confirmation in diagnostically equivocal cases, particularly where morphological or enzymatic analyses could have provided decisive clarification of mitochondrial involvement. Furthermore, although mitochondrial DNA was comprehensively interrogated, nuclear genomic analysis was limited to selected targets or externally performed exome sequencing, thereby restricting the ability to detect the full spectrum of pathogenic variants in nuclear-encoded mitochondrial genes. These limitations are particularly significant given that NGS is now recognized as a cornerstone of modern diagnostic workflows for mitochondrial disease, markedly enhancing diagnostic yield and enabling the differentiation between primary mitochondrial defects and phenocopies [13-15]. The absence of systematic, broad-scale nuclear genomic testing in all patients likely resulted in an underestimation of the true number of genetically confirmed cases. Future work incorporating uniform whole-exome or whole-genome sequencing, supported by functional validation, will be critical for maximizing diagnostic accuracy and aligning with current evidence-based standards.

Conclusions

This study highlights the diagnostic significance and practical utility of a systematically structured, algorithmic framework for the comprehensive evaluation of patients with clinical suspicion of mitochondrial disease. By integrating multidimensional data sources – including clinical, biochemical, and molecular parameters – the proposed diagnostic pathway provides a coherent and harmonized approach, capable of delineating complex phenotypic spectra and effectively distinguishing mitochondrial from non-mitochondrial genetic etiologies.

Among the 240 individuals with Nijmegen Mitochondrial Disease Scores ≥3, molecular analyses confirmed mitochondrial involvement in 37 patients (15.4%) and identified alternative genetic conditions in 44 individuals (18.3%), whereas 159 cases (66.3%) remained unresolved following the applied molecular testing strategy. This distribution reflects the pronounced heterogeneity of clinical presentations and corresponds to a combined diagnostic rate of 33.7% across the cohort.

Collectively, this structured, algorithm-driven workflow demonstrated high efficacy in optimizing case prioritization, refining diagnostic categorization, and synthesizing multidisciplinary data into a unified evaluative framework. These findings underscore the critical value of algorithmic approaches in enhancing diagnostic precision, standardizing assessment, and supporting reproducible decision-making in heterogeneous rare disease populations.

Competing interests

None declared.

Authors’ contribution

DS, NU, and VS conceived the study and contributed to the study design. DS, DB, NU, and VS performed data collection and were involved in the acquisition of clinical and laboratory data. DS and DB carried out data processing and preliminary analyses. DS performed the genetic analyses, including qPCR-HRM screening and Sanger sequencing. AN and CD carried out the biochemical investigations, including urinary organic acid analysis and acylcarnitine profiling. All authors critically reviewed the work, provided important intellectual input, and approved the final version of the manuscript.

Informed consent for publication

Obtained.

Ethics approval

The study was approved by the Research Ethics Committee of Nicolae Testemițanu State University of Medicine and Pharmacy (Decision no. 3 of 09.09.2020).

Acknowledgements and funding

This study was conducted with support from the institutional research project “Diagnosis and Monitoring of Genetic Diseases in the Prevention of Maternal and Child Health Disorders (140102, DiMoGEN)” and the Romanian bilateral project MEMOMAR, funded by the Ministry of Education and Research, CCCDI–UEFISCDI, grant number PN-IV-PCB-RO-MD-2024-0539 within PNCDI IV. Additional support was provided by the Ministry of Health of the Republic of Moldova.

Provenance and peer review

Not commissioned, externally peer-reviewed.

Authors’ ORCID IDs

Doina Secu – https://orcid.org/0000-0002-8571-0524.

Daniela Blaniță – https://orcid.org/0000-0001-7736-3406

Natalia Ușurelu – https://orcid.org/0000-0001-8685-3933

Alina Nicolescu – https://orcid.org/0000-0001-7022-8893

Călin Deleanu – https://orcid.org/0000-0001-8206-5227

Victoria Sacară – https://orcid.org/0000-0001-9200-0494

References

1. Ng YS, Turnbull DM. Mitochondrial disease: genetics and management.  J Neurol. 2016;263(1):179-91. doi: 10.1007/s00415-015-7884-3.
2. Graham BH. Diagnostic challenges of mitochondrial disorders: complexities of two genomes. Methods Mol Biol. 2012;837:35-46. doi: 10.1007/978-1-61779-504-6_3.
3. Forny P, Footitt E, Davison JE, Lam A, Woodward CE, Batzios S, et al. Diagnosing mitochondrial disorders remains challenging in the omics era. Neurol Genet. 2021;7(3):e597. doi: 10.1212/NXG.0000000000000597.
4. Morava E, van den Heuvel L, Hol F, de Vries MC, Hogeveen M, Rodenburg RJ, Smeitink JA. Mitochondrial disease criteria: diagnostic applications in children. Neurology. 2006;67(10):1823-6. doi: 10.1212/01.wnl.0000244435.27645.54.
5. Rahman S. Mitochondrial disease in children. J Intern Med. 2020;287(6):609-633. doi: 10.1111/joim.13054.
6. Mancuso M. Complex neurological and multisystem presentations in mitochondrial disease. Handb Clin Neurol. 2023;194:117-124. doi: 10.1016/B978-0-12-821751-1.00003-8.
7. Grigalionienė K, Burnytė B, Ambrozaitytė L, Utkus A. Wide diagnostic and genotypic spectrum in patients with suspected mitochondrial disease. Orphanet J Rare Dis. 2023;18(1):307. doi: 10.1186/s13023-023-02921-0.
8. Tolomeo D, Orsucci D, Nesti C, Baldacci J, Battini R, Bruno C, et al. The diagnostic approach to mitochondrial disorders in children in the era of next-generation sequencing: a 4-year cohort study. J Clin Med. 2021;10(15):3222. doi: 10.3390/jcm10153222.
9. Schon KR, Horvath R, Wei W, Calabrese C, Tucci A, Ibañez K, et al. Use of whole genome sequencing to determine genetic basis of suspected mitochondrial disorders: cohort study. BMJ. 2021;375:e066288. doi: 10.1136/bmj-2021-066288.
10. van der Ven AT, Johannsen J, Kortüm F, Wagner M, Tsiakas K, Bierhals T, et al. Prevalence and clinical prediction of mitochondrial disorders in a large neuropediatric cohort. Clin Genet. 2021;100(6):766-770. doi: 10.1111/cge.14061.
11. Rouzier C, Pion E, Chaussenot A, Bris C, Ait-El-Mkadem Saadi S, et al. Primary mitochondrial disorders and mimics: insights from a large French cohort. Ann Clin Transl Neurol. 2024;11(6):1478-1491. doi: 10.1002/acn3.52062.
12. Liu Z, Pan K, Wang M, Jin Y, Yang W, Chen K, et al. Novel pathogenic mtDNA variants in Chinese children with neurological mitochondrial disorders. Ann Clin Transl Neurol. 2025;12(3):586-601. doi: 10.1002/acn3.52315.
13. Wu TH, Peng J, LiYang, Chen YH, Lu XL, Huang JT, et al. Use of dual genomic sequencing to screen mitochondrial diseases in pediatrics: a retrospective analysis. Sci Rep. 2023:13(1):4193. doi: 10.1038/s41598-023-31134-5.
14. Sun G, Huang W, Wang L, Wu J, Zhao G, Ren H, et al. Molecular findings in patients for whole exome sequencing and mitochondrial genome assessment. Clin Chim Acta. 2024;561:119774. doi: 10.1016/j.cca.2024.119774.