ACAD9 Deficiency

Isolated deficiency of respiratory chain complex I due to loss of function of an assembly factor with acyl-CoA dehydrogenase activity, presenting mainly with hypertrophic cardiomyopathy, lactic acidosis and myopathy and responding to riboflavin in a proportion of patients

Literature and regulatory information updated: 31 October 2026.

0. Initial Reference Table

1. Introduction

1.1. Definition

ACAD9 deficiency is a primary mitochondrial disease with autosomal recessive transmission caused by biallelic variants in the ACAD9 gene, which encodes a member of the acyl-CoA dehydrogenase family with a second function, that of assembly factor of complex I (NADH:ubiquinone oxidoreductase) of the respiratory chain [2, 19]. Online Mendelian Inheritance in Man (OMIM) defines it as a multisystem disease of usually infantile onset, with acute metabolic acidosis, hypertrophic cardiomyopathy (HCM) and muscle weakness, associated with deficient complex I activity in muscle, liver and fibroblasts, and classifies it as nuclear type 20 of mitochondrial complex I deficiency (MC1DN20) [2]. Orphanet describes it as a rare disease with neurological dysfunction, liver failure and cardiomyopathy due to complex I deficiency and places it under the entity "isolated complex I deficiency" [1, 8].

The clinical spectrum ranges from HCM with fatal lactic acidosis in the neonatal period, or already detectable in the fetus, to exercise intolerance with mild ventricular hypertrophy in adults, by way of forms with optic neuropathy, slowly progressive encephalomyopathy or episodes of liver dysfunction triggered by infections [10, 20–22]. Severity is divided, in practice, by the age at presentation: most patients present in the first year of life, a group in which half do not survive beyond 2 years, whereas more than 90% of those presenting after the first year are alive at 10 years [10]. The therapeutic peculiarity is the response to riboflavin, the precursor of flavin adenine dinucleotide (FAD) and flavin mononucleotide (FMN), which was associated with lower mortality in early-onset patients in a retrospective cohort [10, 19].

1.2. Historical overview

ACAD9 was cloned in 2002 by Zhang and colleagues, by random sequencing of a complementary DNA library from dendritic cells, as a protein of 621 amino acids with a mitochondrial leader sequence and the conserved motifs of the acyl-CoA dehydrogenase family; it oxidised stearoyl-CoA (C18:0) and palmitoyl-CoA (C16:0), but not medium- or short-chain substrates [6]. Ensenauer and colleagues, in 2005, showed the mitochondrial localisation, the two-step processing with removal of the first 37 amino acids and the maximal activity with long-chain unsaturated acyl-CoAs (C16:1, C18:1, C18:2 and C22:6), and Oey and colleagues, in 2006, identified it as the long-chain dehydrogenase of the embryonic and fetal brain [6].

In 2007, He and colleagues described "ACAD9 deficiency" as a new disorder of mitochondrial β-oxidation, in three patients: a previously healthy 14-year-old boy who died of a Reye-like episode with cerebellar infarction after a mild viral illness and aspirin ingestion; a 10-year-old girl with recurrent episodes of acute liver dysfunction and hypoglycaemia from 4 months of age; and a 4.5-year-old girl who died of cardiomyopathy, whose brother had died of the same cause at 21 months [22]. All had mild chronic neurological dysfunction and a marked deficiency of the ACAD9 protein; in the first patient a 4 base pair insertion 44 nucleotides from the initiation codon was found on one allele [2, 22].

In 2010, two independent studies moved the disease into the field of respiratory chain disorders. Nouws and colleagues showed that ACAD9 binds the assembly factors NDUFAF1 and ECSIT, that it is specifically required for complex I assembly and that its variants cause complex I deficiency without disturbance of long-chain β-oxidation; VLCAD (very long-chain acyl-CoA dehydrogenase), from which ACAD9 derives, does not participate in assembly [23]. Haack and colleagues, by exome sequencing of a single patient with severe isolated complex I deficiency and filtering of the mitochondrial proteins, identified compound heterozygous ACAD9 variants, demonstrated pathogenicity by correction of the deficiency in fibroblasts with the normal protein and found two further cases among 120 index cases with complex I deficiency [24]. One of the siblings in Haack's family, treated with riboflavin from birth, had normal psychomotor development at 5 years [2].

In 2011, Gerards and colleagues described, in a consanguineous family with fatigability, exercise intolerance and lactic acidosis from childhood, the homozygous variant c.1594C>T p.(Arg532Trp) and documented improvement in exercise tolerance and in complex I activity with high-dose riboflavin [25]. There followed the series of Collet (2016), who found ACAD9 variants in 8 of 20 children with cardiac hypertrophy and isolated complex I deficiency, of Dewulf (2016), with nine patients from three families, and of Schiff (2015), who related severity to residual dehydrogenase activity [26–28]. In 2018, Repp and colleagues published the largest cohort, with 70 patients from 50 families, and proposed a therapeutic trial of riboflavin in all patients [10]. In 2023, McGregor and colleagues described the structural basis of the ECSIT–ACAD9 interaction [29].

1.3. Nomenclature

Three designations with the same content coexist: "ACAD9 deficiency" (Orphanet, clinical literature), "mitochondrial complex I deficiency due to ACAD9 deficiency" (GeneReviews, GeneGuidance) and "mitochondrial complex I deficiency, nuclear type 20" (OMIM) [1, 2, 9]. The initial classification as a β-oxidation defect, which led to the inclusion of ACAD9 among the acyl-CoA dehydrogenases in the chapters on fatty acid oxidation, persists in the reference textbooks, which treat it in the chapter on β-oxidation defects and, at the same time, in the table of complex I assembly factors [5, 19]. The clinical and biochemical behaviour is that of a complex I deficiency: the acylcarnitine profiles are usually normal or uninformative [3, 30].

2. Molecular Genetics and Pathophysiology

2.1. The gene and the protein

The ACAD9 gene lies at 3q21.3, has 18 exons and encodes a protein of 621 amino acids (68.8 kDa) with a mitochondrial leader sequence; the mature form results from removal of the first 37 amino acids and begins at alanine 38 [3, 6]. The MANE Select reference transcript is NM_014049.5 and the protein NP_054768.2 [7]. Expression is high in the liver and neurons and minimal in skin fibroblasts, which has consequences for biochemical diagnosis in cell culture [28]. ACAD9 has 47% amino acid identity and 65% similarity with VLCAD; both function as homodimers associated with the inner mitochondrial membrane and catalyse the first step of β-oxidation of long-chain fatty acids [10].

2.2. Two functions: complex I assembly and acyl-CoA dehydrogenation

Mammalian complex I is made up of 44 subunits, seven encoded by mitochondrial DNA (mtDNA) and the remainder by the nuclear genome, and requires at least 19 specific assembly factors [5, 10]. It receives electrons from the NADH generated by the oxidation of pyruvate and fatty acids and is the main entry point of the respiratory chain [5]. ACAD9 forms, with NDUFAF1 and ECSIT, the mitochondrial complex I assembly (MCIA) complex; binding to ECSIT causes a conformational change that converts ACAD9 from a β-oxidation enzyme into an assembly factor, and the assembly function is independent of the dehydrogenase activity [3, 23]. The structure of the ECSIT–ACAD9 complex showed that ECSIT alters the FAD-binding loop of ACAD9, which leads to release of the cofactor and redirection of the protein to complex I assembly; an adjacent α helix distinguishes this capacity from VLCAD, and phosphorylation of ECSIT decreases its association with ACAD9 [29].

The dehydrogenase activity is not irrelevant. In human embryonic kidney (HEK293) cells, which express abundant ACAD9, knockout of the gene impaired both long-chain fatty acid oxidation and complex I, and both were restored by the normal protein [28]. ACAD9 forms the acylcarnitine C14:1 that accumulates in VLCAD deficiency [19]. Despite the overlap of substrates, ACAD9 and VLCAD do not compensate for each other in patients deficient in one of them, and the regulation and tissue distribution of the two enzymes are independent [22].

2.3. From the assembly defect to the disease

Loss of ACAD9 function reduces the amount of assembled complex I and of respiratory supercomplexes; in patients' fibroblasts, the ACAD9 protein was decreased in 12 of 14 lines, complex I-dependent respiration in 13 of 17 lines and the supercomplexes in all the lines studied [10]. The lack of correlation between residual protein and complex I activity, with lines almost devoid of ACAD9 and abundant supercomplexes, indicates that small amounts of functional protein suffice for assembly [10]. The metabolic consequence is that of any defect of oxidative phosphorylation: reduced ATP production in the tissues with the highest energy consumption, accumulation of NADH and a shift to anaerobic glycolysis, about 20 times less efficient, with production of lactate, pyruvate and alanine [5]. In human systemic disease, complete loss of oxidative phosphorylation is never observed and residual activity varies between tissues [5].

The animal model confirms the centrality of the heart and muscle. Complete inactivation of Acad9 in the mouse is embryonically lethal; cardiac inactivation abolishes complex I and the supercomplexes in the myocardium, renders ECSIT and VLCAD undetectable and causes, by the 14th day of life, cardiomyopathy with thickened walls and reduced ejection fraction, which suggests intrauterine onset; inactivation in skeletal muscle causes exercise intolerance and raised basal and post-exercise lactate [2]. The autopsy of a newborn with fatal lactic acidosis showed mitochondrial hyperplasia in cardiomyocytes, the diaphragm, the liver and the renal tubules [31].

The mechanism proposed for the response to riboflavin is an increase in mitochondrial FAD, which would favour cofactor binding, folding and stability of ACAD9 and, by this route, complex I assembly [3, 10]. In fibroblasts, riboflavin increased complex I activity in 9 of 15 lines (14% to 109%) and bezafibrate, an agonist of peroxisome proliferator-activated receptor alpha (PPAR-α), in 12 of 17 lines, with increased supercomplex formation in 4 of 5 [10].

2.4. Inheritance, penetrance and variant spectrum

Transmission is autosomal recessive in all the families described [1, 2]. The sources consulted do not describe dominant transmission or symptomatic heterozygous carriers. In the cohort of 70 patients, 52 different variants were found (34 already known and 18 novel): 42 missense, 1 frameshift, 1 nonsense, 7 splice and 1 affecting the initiation codon; no patient had two predicted loss-of-function variants, which, together with the embryonic lethality of the Acad9-null mouse, suggests that complete absence of ACAD9 is incompatible with life [10]. The variants are distributed throughout the gene, with no identified founder variants and no correlation with specific regions or domains [10].

The protein nomenclature of publications predating normalisation should be checked against the MANE transcript before being transcribed into a report; the case of c.380G>A illustrates the discrepancy between the published and the predicted amino acid [7, 25]. The conversions in the table, including those of the variants in trans cited within it (c.130T>A, c.941T>C, c.1636G>A, c.1030-1G>T, c.869G>A and c.453+2T>A), were obtained in VariantValidator (GRCh38) and the classifications in ClinVar, both consulted in October 2026 [7, 32]. The patient of Li and colleagues, with a frameshift and a missense variant, does not contradict the absence of patients with two predicted loss-of-function variants [10, 21].

2.5. Genotype–phenotype correlations

Clinical severity does not correlate with complex I activity measured in fibroblasts, but with the residual dehydrogenase activity of ACAD9: Schiff and colleagues tested 16 variants from 24 patients in a prokaryotic expression system and found a significant inverse correlation between residual activity and the severity of the phenotype [28]. The variants with little impact on dehydrogenase activity cluster in the C-terminal domain and the inactivating ones in the catalytic domain [6]. The authors concluded that treatment must counter both complex I dysfunction and that of β-oxidation [28]. The intrafamilial variability described by Dewulf (siblings stable in their mid-20s on riboflavin, alongside infants with early death in other families) and the heterogeneous response to riboflavin in lines with the same variant make it impossible to predict the course from the genotype alone [10, 27, 30]. Concordance between the fibroblast response to riboflavin and the clinical response was observed in 7 of 8 cell–patient pairs [10].

2.6. The class: isolated complex I deficiencies of nuclear cause

Isolated complex I deficiency is the most frequent enzyme defect among the disorders of oxidative phosphorylation, with extreme genetic heterogeneity, a majority of nuclear causes and no obvious genotype–phenotype correlations [4]. The phenotypes include macrocephaly with progressive leukodystrophy, non-specific encephalopathy, HCM, myopathy, liver disease, Leigh syndrome, Leber hereditary optic neuropathy (LHON) and some forms of Parkinson's disease [4]. OMIM numbers 39 nuclear types (MC1DN1 to MC1DN39); ACAD9 is type 20. Table 10.4 of Saudubray groups the genes into structural subunits, assembly factors and cofactors [4, 5].

Within the class, ACAD9 deficiency stands out for its cardiac and muscular predominance with relative sparing of the brain and for the response to riboflavin. In the French series of children with cardiac hypertrophy and isolated complex I deficiency, ACAD9 was the most frequent gene, with biallelic variants in 8 of 20 (40%) [26]. In the fibroblast lines tested with palmitic acid, the defects of NDUFS7 and NDUFAF5 responded and those of ACAD9, NDUFA12 and NDUFV2 did not [38].

3. Clinical Spectrum and Manifestations by System

3.1. The reference cohort and its limitations

The quantitative description of the disease rests on the international cohort of Repp and colleagues (2018): 70 patients (41 female) from 50 families, 29 of them previously unpublished, gathered through a collaboration of centres in Europe, China and the United States, with clinical data collected by a 93-question online survey completed by the responsible physician [10]. In 55 patients, complex I deficiency was demonstrated in skeletal muscle (44), cardiac muscle (7) and/or fibroblasts (26) [10]. The denominators vary from item to item, because not all data were available for all patients, and the information is retrospective; the cohort mixes patients treated and not treated with riboflavin, and therefore does not describe a pure natural history [10].

3.2. Age at presentation and two prognostic groups

Fifty of the 70 patients presented in the first year of life and 20 later; survival in the first group was significantly worse (p = 6.49×10⁻⁵), with 50% of deaths by 2 years of age, whereas more than 90% of the patients with later presentation were alive at 10 years [10]. At the time of publication, 37 patients were alive with a median age of 14 years (24 days to 44 years) and the median age at death was 3 months (1 day to 44 years) [10]. Van Hove and colleagues, citing the same literature, report 70% mortality by 2 years with onset before 1 year and 10% with later onset [37]. The difference between the 50% and the 70% results from different readings of the curves and from the mixture of treated and untreated patients; Repp's data, published with the Kaplan-Meier curve, are the primary source.

3.3. Prenatal and neonatal period

In the cohort of 70 patients, the prenatal findings included intrauterine growth restriction (6 of 60; 10%), oligohydramnios (4 of 60; 7%), fetal heart rhythm abnormalities (2 of 60; 3%), fetal cardiomegaly (1 of 60; 2%) and reduced fetal movements (1 of 60; 2%); the three patients with fetal cardiac findings died on days 1, 2 and 280 of life [10]. In the neonatal period, lactic acidosis was recorded in 21 of 63 (33%), cardiomyopathy in 15 of 63 (24%), respiratory failure with mechanical ventilation in 6 of 54 (11%), rhythm abnormalities in 4 of 54 (7%) and severe hepatic or renal dysfunction in 2 of 54 (4%) each [10]. Familial recurrences of prenatal HCM with intrauterine growth restriction and confirmed ACAD9 defects have been published, as have fetal presentations with cardiomegaly and death shortly after birth [36, 39]. Dewulf and colleagues observed a patent ductus arteriosus in 5 of 9 patients from two families [27].

3.4. Heart

Cardiomyopathy is the dominant manifestation: it was present at presentation in 44 of 66 patients (67%) and during the course in 56 of 66 (85%), usually of the hypertrophic type [3, 10]. In primary mitochondrial diseases, HCM is typically concentric and without left ventricular outflow tract obstruction, systolic dysfunction is more frequent than in sarcomeric HCM and dilated cardiomyopathy often arises as the evolution of a previous HCM [40]. The cardiac spectrum of ACAD9 deficiency ranges from left ventricular hypertrophy evident only on the electrocardiogram or mild on echocardiography to severe hypertrophic, dilated or combined cardiomyopathy [27, 30]. Heart failure and cardiovascular collapse are frequent causes of death [37]. The electrocardiogram (ECG) is abnormal in 68% of mitochondrial cardiomyopathies, with T-wave inversion in V6 as the most frequent abnormality, and atrioventricular and bundle branch block, ventricular pre-excitation, a long QTc and ventricular ectopy may occur [40].

3.5. Skeletal muscle and exercise tolerance

Muscle weakness was present at presentation in 21 of 48 patients (44%) and during the course in 37 of 49 (75%); exercise intolerance in 21 of 49 (43%) and 34 of 47 (72%), respectively [10]. In the later-onset forms, exercise intolerance with lactic acidosis may be the main complaint from childhood, as in the consanguineous family of Gerards, with subsarcolemmal mitochondrial proliferation and severe complex I deficiency in muscle [25]. Muscle biopsy may show ragged red fibres and mitochondrial proliferation [12, 34]. Survivors of early cardiac forms, including transplant recipients, later developed muscle weakness and exercise intolerance [26].

3.6. Nervous system

The central nervous system is spared in most patients, unlike the usual picture in complex I deficiencies. In the cohort of 70 patients, severe intellectual disability was reported in 1 of 51 (2%) and mild in 14 of 48 (29%); severe developmental delay in 4 of 52 (8%) and mild in 23 of 51 (45%); all patients with severe delay or intellectual disability had had early onset [10]. Neither optic atrophy nor pigmentary retinopathy was recorded in any of the 70 patients [10]. On the available magnetic resonance imaging (MRI), there were basal ganglia abnormalities in 4 of 24 (17%), leukoencephalopathy in 5 of 21 (24%), global cerebral atrophy in 2 of 20 (10%) and isolated cerebellar atrophy in 1 of 20 (5%); spectroscopy showed a lactate peak in 2 of 16 (13%) [10]. Outside this cohort, predominantly neurological phenotypes have been described: an 11-month-old infant with microcephaly, dystonia and lactic acidosis and normal complex I activity in muscle [12]; a child with optic and peripheral neuropathy without cardiac involvement, with visual improvement on riboflavin [20]; and a 27-year-old man with global delay, progressive hearing loss, ataxia, dysarthria, intellectual disability, cerebellar atrophy and optic atrophy [21]. Leigh syndrome, macrocephaly, seizures and Reye-like encephalopathic episodes form part of the described spectrum [3, 12, 22]. In the survivors of Collet's series, cognitive impairment and seizures appeared late, even when absent before heart transplantation [26].

3.7. Liver, kidney and other organs

Hepatic crises are part of the original description: recurrent episodes of acute liver dysfunction and hypoglycaemia triggered by minor illnesses, and a fatal Reye-like episode with diffuse microvesicular steatosis [2, 22]. Saudubray includes ACAD9 variants among the causes of liver failure in infancy, with crises that are mostly recurrent and triggered by fever and infections [41]. In the cohort of 70 patients, hypoglycaemia, rhabdomyolysis and liver failure, characteristic of VLCAD deficiency, were infrequent [10]. Collet and colleagues recorded proximal tubulopathy, renal failure, secondary ovarian failure and optic atrophy in the survivors [26].

3.8. Functional capacity

Among the patients with data, 69% had age-appropriate behaviour (27 of 39), 63% attended or had completed mainstream education (26 of 41), 81% sat unaided (34 of 42), 80% walked unaided (33 of 41), 80% ate and drank unaided (33 of 41), 73% managed personal hygiene without help (30 of 41) and 79% communicated with words (34 of 43); more than 70% carried out activities of daily living at the level of their peers [10].

3.9. Natural history and treated disease

The available observations do not rigorously separate natural history from treated disease. The data closest to natural history are those of 17 patients with onset in the first year not treated with riboflavin, of whom 16 died [10]. The reports of long survival include patients treated with riboflavin from childhood: two siblings stable in their mid-20s with mild left ventricular hypertrophy and a woman who reached 26 years with a normal pregnancy [27, 34]. The late-onset transplant recipients of Repp's cohort were alive at 15 and 35 years [10].

4. Clinical Recognition and Warning Signs

4.1. When to suspect it

The most frequent scenario is that of an infant with HCM, often biventricular, accompanied by lactic acidosis, hypotonia or poor weight gain, with or without heart failure [10, 30]. The 2023 European Society of Cardiology (ESC) guideline on cardiomyopathies considers it mandatory to exclude inborn errors of metabolism, including Pompe disease, β-oxidation defects and mitochondrial diseases, in infants with biventricular HCM, frequently with heart failure and systolic dysfunction, and at least one metabolic warning sign: muscular hypotonia, raised creatine kinase (CK) or transaminases, consanguinity or matrilineal inheritance [15]. The same guideline points to lactic acidosis, myoglobinuria and leukopenia as suggestive of mitochondrial disease, and to learning difficulty, sensorineural deafness, visual impairment and muscle weakness as extracardiac signs of mitochondrial disease in patients with HCM [15]. In mitochondrial diseases in general, suspicion increases when two or more organs are involved and when there is worsening with infections, vaccinations or exertion [40].

In the older child, the adolescent and the adult, ACAD9 deficiency should be considered in the face of exercise intolerance with lactic acidosis, with or without left ventricular hypertrophy, and in mitochondrial myopathy with complex I deficiency [14, 25, 35]. The neurological phenotypes (microcephaly with dystonia, optic and peripheral neuropathy, slowly progressive encephalomyopathy) are rare, but the absence of cardiomyopathy does not exclude the diagnosis [10, 12, 20, 21]. In the approach to metabolic myopathies in adults, Saudubray points to exercise intolerance (cramps or exertional fatigue) and recurrent rhabdomyolysis as the hallmarks, and distinguishes mitochondrial diseases, with early fatigue or myalgia disproportionate to the weakness, from β-oxidation defects, with rhabdomyolysis triggered by prolonged exercise or fasting [42]. He and colleagues recommended considering ACAD9 deficiency when a long-chain β-oxidation defect is suspected but not confirmed, particularly after a Reye-like episode [2, 22].

4.2. Warning and emergency signs

4.3. Newborn screening

ACAD9 deficiency is not part of biochemical newborn screening panels, because the acylcarnitine profiles are usually normal or uninformative [3, 30]. C14:1-acylcarnitine, the marker of VLCAD deficiency in screening, is formed by ACAD9, but does not discriminate ACAD9 deficiency [19, 43]. Genomic newborn screening is under investigation: in the Australian BabyScreen+ study, genome sequencing from dried blood spots in 1000 newborns, targeting 605 genes of early-onset, severe and treatable diseases, identified 16 (1.6%) with a high-probability result, only one of them detected by conventional screening, with a mean time to result of 13 days [44]. The Belgian Baby Detect genomic newborn screening study has completed recruitment [45]. The sources consulted do not allow it to be stated that ACAD9 is part of the panels of these studies.

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Criteria

There are no formal clinical criteria for ACAD9 deficiency. The diagnosis is established by the identification of biallelic pathogenic or likely pathogenic variants in ACAD9, in a patient with a compatible phenotype [3, 10]. No direct biochemical assay of ACAD9 is available in clinical practice [12]. Demonstration of complex I deficiency supports the diagnosis, but does not identify the responsible gene among the dozens of subunits and assembly factors, and normal activity does not exclude it [3, 12].

5.2. Initial biochemical assessment

The GeneReviews differential diagnosis table makes ACAD9 deficiency less likely when urinary 2-methyl-2,3-dihydroxybutyrate is present or when the lactate/pyruvate ratio is normal [9].

Interpretation of lactate requires attention to the sampling conditions. In respiratory chain defects, hyperlactataemia is permanent and highly variable, often maximal in the postprandial state, and is accompanied by a simultaneous rise in the lactate/pyruvate ratio (normal around 10:1, above 20 in these defects) and in the 3-hydroxybutyrate/acetoacetate ratio, which reflect the cytoplasmic and mitochondrial redox states; a normal or low lactate/pyruvate ratio points instead to pyruvate dehydrogenase deficiency [47]. Lactaturia with raised plasma alanine suggests genuine hyperlactataemia, because the renal reabsorption threshold for lactate is around 4 mmol/L, and lactate in the cerebrospinal fluid is more specific than in plasma [40, 47]. Many patients with respiratory chain defects have no hyperlactataemia, and some nuclear variants that alter the respiratory chain, most of them described in complex I genes, do not produce it [46, 47]. The fasting test is uninformative in these defects, a glucose load may raise lactate and the exercise test has neither sufficient sensitivity nor specificity to diagnose a mitochondrial myopathy [47].

5.3. Cardiac assessment

Baseline cardiac assessment includes a 12-lead ECG and echocardiography with left ventricular mass, systolic and diastolic function, wall thicknesses, and a search for outflow tract obstruction and valvular abnormalities; HCM is defined by unexplained left ventricular wall thickness above 15 mm or a Z-score above +2 [14, 40]. Cardiac MRI with late gadolinium enhancement quantifies function and fibrosis and is indicated when echocardiography is inconclusive or before invasive treatments [14, 40]. A 24-hour Holter recording is indicated if there are palpitations, paroxysmal events or cardiomyopathy [14]. Global left ventricular hypokinesia and pre-excitation are findings described in mitochondrial cardiomyopathies [15]. Endomyocardial biopsy with electron microscopy is an option when a mitochondrial or storage cardiomyopathy is suspected and the result could change management, weighing the risk [15].

5.4. Genetic testing

Next-generation sequencing with full coverage of the mitochondrial disease genes, or exome sequencing, is preferable to the study of individual genes, because different genes produce the same phenotype; if a panel is negative, the exome follows [5, 11]. Trio analysis facilitates variant interpretation; Clinical Cardiogenetics advocates early genetic testing, particularly in the acute setting, and the ESC considers a multidisciplinary approach with a geneticist and metabolic disease specialists mandatory in infantile-onset disease [15, 40]. ACAD9 deficiency was the first complex I deficiency solved by exome sequencing, and trio sequencing was what established the diagnosis in the cases with normal enzymology [12, 24]. Genomics is increasingly used in practice, with frequent variants of uncertain significance (VUS) [5]. ClinGen has no published actionability assessment for ACAD9, which does not amount to the absence of effective intervention [3].

The report should state the gene, the transcript and its version (NM_014049.5), the HGVS c. and p. nomenclature, the zygosity, the classification according to the criteria of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP) and the phase of the variants, confirmed by testing the parents when the patient is a compound heterozygote [30, 40]. Molecular testing in blood has limitations: deep intronic variants, regulatory variants (the promoter insertion described by He and colleagues) and structural changes may escape the exome; in He's original series, only one of the four alleles studied was characterised [6, 22]. A negative result in a patient with a suggestive phenotype does not exclude the disease and justifies genome sequencing, RNA studies or functional studies [5, 11]. In the face of a VUS, demonstration that it is in trans with a pathogenic variant (parental testing) and segregation in affected siblings help reclassification [40].

5.5. Biopsy and functional studies

Muscle biopsy is no longer mandatory when genetic diagnosis is available, but retains value in the critically ill patient, in the functional confirmation of novel variants and when genetics is inconclusive [5, 40]. The 2015 Mitochondrial Medicine Society (MMS) guideline prefers open biopsy of the vastus lateralis, with COX, SDH, NADH-TR and combined SDH/COX histochemistry, electron microscopy and enzymology of complexes I to IV in frozen tissue or fresh mitochondria, normalised to citrate synthase or complex II [11]. In ACAD9 deficiency, complex I deficiency has been demonstrated mainly in skeletal muscle, but it may be absent: Aintablian's infant had mitochondrial proliferation and normal complex I activity in muscle, and two fibroblast lines with homozygous variants had no complex I deficiency [10, 12]. Fibroblasts express little ACAD9 and assays in this tissue may give false negatives [11, 28]. In research studies, fibroblasts allow measurement of complex I-dependent respiration, the amount of ACAD9 protein by western blot, supercomplex assembly by native gel electrophoresis (BN-PAGE) and the response to riboflavin or bezafibrate [10].

5.6. Practical algorithm

5.7. Prenatal and preimplantation diagnosis

When the two familial variants are known, prenatal diagnosis in chorionic villi or amniocytes and preimplantation genetic diagnosis are technically straightforward and follow the principles of Mendelian diseases [5, 49]. Biochemical prenatal diagnosis (complex I activity in fetal cells) has given erroneous results and does not replace molecular testing [46]. In the face of fetal HCM with growth restriction without a previous family diagnosis, inclusion of ACAD9 in prenatal investigations was proposed by the authors who described the familial recurrence [39].

6. Differential Diagnosis

6.1. Comparative table

6.2. Narrative on the main differentials

The immediate differential in the infant with HCM and lactic acidosis is that of the metabolic cardiomyopathies. The ESC considers it mandatory, in the face of biventricular HCM with metabolic warning signs, to exclude Pompe disease, β-oxidation defects and mitochondrial diseases [15]. Pompe disease is distinguished by acid α-glucosidase activity, by the often extreme hypertrophy and by the response to enzyme replacement therapy, which reverses the hypertrophy in infants [15]. The long-chain β-oxidation defects, with the exception of CPT I deficiency, present with HCM and arrhythmias in infancy and are distinguished by the acylcarnitine profile [40, 41]. The relationship between ACAD9 and VLCAD deserves particular attention: the two proteins are homologous, ACAD9 forms C14:1-acylcarnitine and both cause cardiomyopathy, but the hypoglycaemia, rhabdomyolysis and liver failure of VLCAD deficiency are rare in ACAD9 deficiency, and the two enzymes do not compensate for each other [10, 19, 22]. A C14:1 above 1 µmol/L correlates well with confirmed VLCAD deficiency [43].

Among the mitochondrial causes, ACAD9 deficiency is the most frequent cause of HCM with isolated complex I deficiency in childhood (8 of 20 in a French series) [26]. The other complex I defects mainly produce Leigh syndrome, leukoencephalopathy and neurological involvement, and cardiomyopathy in them is an accompanying feature [4, 5]. The cardiac and muscular predominance with preserved cognition points towards ACAD9, but does not replace genetic testing [10]. Barth syndrome (boys, neutropenia, 3-methylglutaconic aciduria) and Sengers syndrome (congenital cataracts) have their own extracardiac signs, and PRKAG2 cardiomyopathy in the fatal congenital form can mimic a respiratory chain disorder [40, 41]. Early-onset sarcomeric HCM, including double or compound variants, should be excluded even without a family history [15].

The response to riboflavin is not exclusive to ACAD9 deficiency. MADD due to ETFDH, the deficiencies of the riboflavin transporters (RFVT2 and RFVT3, formerly Brown-Vialetto-Van Laere and Fazio-Londe syndromes), FAD synthase deficiency (FLAD1) and that of the mitochondrial FAD transporter (SLC25A32) also respond, and Table 45 of Saudubray lists ACAD9 deficiency alongside these conditions among the indications for riboflavin [19, 50, 53]. The mild MADD-like acylcarnitine profile, present in about 60% of RFVT2/3 deficiencies, and low plasma flavins help to separate these groups [19].

In the predominantly neurological forms, the differential includes Leigh syndrome of any cause and ECHS1 deficiency; urinary 2-methyl-2,3-dihydroxybutyric acid favours ECHS1, and a normal lactate/pyruvate ratio makes ACAD9 deficiency less likely [9]. In Reye-like hepatic episodes with recurrent liver dysfunction, the β-oxidation defects (LCHAD, MCAD) are the main differential, and He and colleagues suggested considering ACAD9 deficiency when these are not confirmed [2]. In the forms with optic neuropathy, LHON and mtDNA diseases enter the differential [4, 20].

7. Treatments and Risk Management Strategies

7.1. General principles

There is no curative treatment and no randomised trial has been conducted in ACAD9 deficiency [3, 10]. Management rests on four components: riboflavin, started as early as possible; treatment of cardiomyopathy and heart failure according to the cardiology guidelines; prevention and early treatment of metabolic decompensation; and multisystem surveillance at a centre with experience in mitochondrial disease and cardiology [10, 14, 40]. The applicable recommendations are, for the most part, extrapolated from the general guidelines on primary mitochondrial disease (MMS 2015 and 2017, the 2022 Australian adaptation, the 2020 Delphi consensus on drugs) and from the 2023 ESC guideline on cardiomyopathies, because there is no guideline specific to this disease [11, 13–15, 54].

7.2. Riboflavin

Mechanism. Riboflavin (vitamin B2) is the precursor of FAD and FMN, cofactors of complex I and of numerous β-oxidation dehydrogenases; in ACAD9 deficiency, an increase in mitochondrial FAD should favour cofactor binding, folding and stability of the protein and, thereby, complex I assembly [10, 55].

Evidence. Most of the evidence comes from the retrospective cohort of 70 patients. Of the 67 with treatment data, 20 were untreated and in 15 the information was missing; among the 31 with a clinical effect reported by the physician, there was benefit in 20 (65%) and no effect in 11 (35%), without clinical deterioration or adverse effects [10]. Among the 39 patients with onset in the first year and treatment data, 7 of 22 treated with riboflavin died and 16 of 17 untreated patients died (p = 5.34×10⁻⁵ in the comparison of the Kaplan-Meier curves) [10]. In families 1 and 33, the first affected child died before 2 years of age without riboflavin and the siblings treated from diagnosis were alive at 10, 1.5 and 11 years [10]. In fibroblasts, riboflavin increased complex I activity in 9 of 15 lines, and the cellular response agreed with the clinical one in 7 of 8 pairs; in one patient, the cells responded and the patient did not [10]. The authors identified the limitations: the study was not designed as a trial, data on the start, dose and duration of riboflavin existed in only a minority and were not analysed, and the clustering of deaths at the end of the first year exposes the analysis to survival selection bias (the patients who live long enough to be diagnosed and treated are, from the outset, those with a less rapid course) [10]. Before this cohort, the evidence consisted of reports of myopathic patients improving on riboflavin, including the sibling treated from birth in Haack's family and Gerards' family [2, 25]. There are genotypes without a response, such as homozygous p.(Ala220Val), whose fibroblasts did not improve with riboflavin [30, 33]. Riboflavin improved visual acuity in a child with optic neuropathy [20].

Recommendations. The 2017 MMS guideline recommends considering riboflavin in the myopathy associated with ACAD9 deficiency (and the combination of coenzyme Q10 with riboflavin in ETFDH-related myopathy) [14]. The 2022 Australian adaptation recommends considering riboflavin 50–200 mg daily in the myopathy associated with ACAD9 deficiency [13]. Repp and colleagues propose a therapeutic trial with 20 mg/kg/day, up to a maximum of 200 mg/day, in all patients with this diagnosis, and consider it reasonable to start it in patients with a compatible phenotype while genetic testing is under way [10]. Saudubray, citing the same cohort, reports 100–400 mg/day orally, and its medicines table gives 100–400 mg/day divided into 2 to 3 doses for the indications of riboflavin, among them ACAD9 deficiency [5, 53].

The dose these sources share is 20 mg/kg/day in infants and children, not exceeding 200 mg/day, which coincides with the maximum of the Australian guideline [10, 13]. Daily doses of 100 to 400 mg, such as those of Saudubray, exceed the maximum of 200 mg proposed by the cohort and the Australian guideline [10, 13, 53]. Neither the minimum duration of the therapeutic trial nor formal response criteria have been established; the sources record clinical course, lactate, cardiac function and survival as outcomes [10, 34].

Safety. In the cohort of 70 patients, no adverse effects or clinical deterioration attributable to riboflavin were reported [10]. In RFVT2/3 deficiencies, doses of 10–80 mg/kg/day have been used with minimal side effects [19].

Regulatory status. Oral riboflavin is not a medicine authorised for this indication in any of the jurisdictions consulted, and its use in ACAD9 deficiency rests on guidelines and on the cohort, not on a label.

7.3. Other cofactors and supplements

In the cohort of 70 patients, several were taking coenzyme Q10, biotin and L-carnitine, with only anecdotal positive effects [10]. The general recommendations differ. The 2015 MMS guideline recommends offering coenzyme Q10 to most patients with mitochondrial disease and offering α-lipoic acid and riboflavin, giving L-carnitine when a deficiency is documented and introducing the supplements one at a time [11]. The 2022 Australian adaptation considers that there is insufficient evidence of benefit from supplements outside specific conditions (thiamine with a ketogenic diet in pyruvate dehydrogenase deficiency, coenzyme Q10 in primary coenzyme Q10 deficiency, riboflavin in ACAD9 deficiency) and recommends carnitine, 50–100 mg/kg/day in 2 to 3 doses, only when there is a deficiency [13]. Saudubray notes that vitamin "cocktails" have no evidence base to support or refute their use, although riboflavin in complex I deficiency has a rationale [5]. Clinical Cardiogenetics allows empirical supplementation with thiamine, riboflavin, biotin and coenzyme Q10 in critically ill patients with primary mitochondrial disease [40]. In Kotlukova's case, the infant received levocarnitine 100 mg/kg/day and coenzyme Q10, without riboflavin [30].

The sources consulted give no coenzyme Q10 dose for ACAD9 deficiency; the published doses refer to the condition in which the indication is established, primary coenzyme Q10 deficiency, and their application to ACAD9 deficiency is an extrapolation. The Australian adaptation gives 10–30 mg/kg/day in two doses, and Saudubray, in the biosynthesis defects, at least 30 mg/kg/day in children and up to 3 g/day in adults [5, 13]. In a study cited by the 2009 MMS guideline, the plasma concentration did not increase on moving from 2400 to 3000 mg/day in adults, which suggests saturation of absorption above 2400 mg/day; bioavailability depends on the formulation [55]. For L-carnitine in acute illness, the 2009 MMS guideline reports the intravenous route with at least 100 mg/kg/day and doses used up to 300 mg/kg/day, maintaining intravenously the previous oral dose when this is higher, alongside the oral maintenance dose of 50–100 mg/kg/day of the Australian guideline [13, 55].

7.4. Treatment of the cardiomyopathy

Treatment of mitochondrial cardiomyopathy follows the general guidelines for heart failure and HCM, with diuretics, angiotensin-converting enzyme (ACE) inhibitors and beta blockers [40]. In the cohort of 70 patients, the physicians reported a positive effect on heart failure with beta blockers in 14 of 44 patients (32%), ACE inhibitors in 6 of 40 (15%), calcium channel blockers in 1 of 37 (3%) and diuretics in 3 of 39 (8%); no patient received digitalis and in only one patient was worsening reported on a beta blocker [10]. The choice of beta blocker is a point of divergence: Clinical Cardiogenetics, on the basis of in vitro data on inhibition of mitochondrial function related to lipophilicity, proposes atenolol as first choice and advises against carvedilol, metoprolol and propranolol, whereas the 2020 Delphi consensus considers beta blockers safe in primary mitochondrial diseases, as are enalapril and amiodarone [40, 54]. The decision should be based on the cardiological indications, with the Delphi consensus as the more recent position.

Devices follow the general guidelines. The 2017 MMS guideline recommends a low threshold for pacing, ablation of supraventricular tachycardias and of Wolff-Parkinson-White syndrome, and an implantable cardioverter-defibrillator (ICD) in patients at risk of sudden death, with left ventricular wall thickness above 30 mm or with sustained or non-sustained ventricular tachycardia [14]. The 2023 ESC guideline recommends estimating the risk of sudden death with the HCM-SCD Risk tools and, in children, HCM Risk-Kids, and using an ejection fraction below 50% in the shared decision on an ICD; in a meta-analysis of about 3000 patients cited by the guideline, late gadolinium enhancement was associated with a 2.32-fold higher risk of sudden death, aborted sudden death or appropriate ICD discharge and with 2.1-fold higher overall mortality [15]. The European calculator for adults has not been validated for those under 16 years of age or for metabolic or syndromic HCM; in children, an enlarged left atrium, a raised E/E′ ratio, reduced global longitudinal strain and extensive late enhancement were associated with sudden death [40].

7.5. Heart transplantation

The 2017 MMS guideline accepts heart transplantation in end-stage heart failure, weighing the comorbidities and the natural history of the mitochondrial disease concerned [14]. In ACAD9 deficiency, the age at onset appears to determine the outcome: in the cohort of 70 patients, the two transplant recipients with onset in the first year died (one 4 years after transplantation performed at 2 years, the other at 3 months, after an ICD and transplantation), and the two with later onset, transplanted at 9 and 18 years, were alive at 15 and 35 years with satisfactory cardiac function [10]. In Collet's series, 3 of 8 children were transplanted and 2 survived, but the survivors later developed cognitive impairment, seizures, muscle weakness, exercise intolerance, tubulopathy, renal failure, ovarian failure or optic atrophy, even when these were absent before transplantation; the authors recommend caution in the indication [26]. Van Hove summarises that the transplanted patients later developed neurological, muscular, renal and ophthalmological problems [37].

7.6. Patients without a response to riboflavin

For early-onset patients without a response to riboflavin, mortality is high and the published options amount to isolated cases [37, 58]. Kadoya and colleagues treated a child with progressive cardiomyopathy from 8 months of age and a left ventricular ejection fraction of 45.4% at 20 months with sodium pyruvate, a beta blocker and coenzyme Q10, without riboflavin, with sustained normalisation of cardiac function [58]. Van Hove and colleagues treated a 6-month-old infant without a response to riboflavin (20 mg/kg/day in three doses), with an ejection fraction of 18%, with calcium pyruvate 900 mg/kg/day, nicotinamide riboside 300 mg twice daily, D,L-3-hydroxybutyrate 900 mg/kg/day, coenzyme Q10 100 mg twice daily (26.7 mg/kg/day), carvedilol and bezafibrate increased progressively to 300 mg twice daily (86 mg/kg/day), eight times the adult dose; lactate and NT-proBNP fell, the ejection fraction stabilised near 45% and there was no hepatic or muscular toxicity, but the child died of heart failure with infection at 10.5 months [37]. The authors attributed the limited efficacy to the weak PPAR agonist activity of bezafibrate: the usual adult dose (about 10 mg/kg/day) gives peaks of 25 µM and the dose used gave 100 µM, below the concentrations of 300–400 µM effective in culture [37]. These approaches are experimental and bezafibrate is outside the approved indications; in a double-blind crossover trial in five patients with VLCAD deficiency, bezafibrate brought no improvement [10].

7.7. Diet, fasting and exercise

The basic recommendation is to avoid prolonged fasting and catabolism [11, 30]. As for a high-fat diet, the sources diverge. The 2022 Australian adaptation allows it to be considered in patients with complex I deficiency, after β-oxidation defects have been excluded, and Clinical Cardiogenetics reports the ketogenic diet as advised in complex I defects [13, 40]. The 2015 MMS guideline states that there is no evidence for adjusting the diet to the respiratory chain results, and the 2009 MMS guideline notes that data on the ketogenic diet in mitochondrial diseases with a secondary β-oxidation defect are limited [11, 55]. In the particular case of ACAD9, the fatty acid dehydrogenase component and the absence of a fibroblast response to palmitic acid argue against generalising a high-fat diet without individual assessment [28, 38].

Supervised, progressive aerobic and resistance exercise is considered safe and beneficial in mitochondrial myopathies, with mandatory cardiac screening before a programme is started; in patients with cardiomyopathy, the intensity should be prescribed by the cardiology team and exercise testing performed under supervision, weighing the risk of decompensation [11, 14].

7.8. Intercurrent illness, intensive care and the perioperative period

The MMS recommends a medical alert bracelet, an up-to-date summary document from the reference centre (symptoms, baseline cardiorespiratory function, medication, contraindicated treatments and an emergency plan) and prevention of catabolism with dextrose-containing fluids before, during and after procedures [11, 14]. In acute decompensation, assessment includes biochemistry, glucose, transaminases and lactate; treatment consists of intravenous dextrose, withdrawal of potentially toxic drugs and correction of the metabolic abnormalities, maintaining the supplements whenever possible; lipids may be used even with a secondary β-oxidation defect [11, 14]. In intensive care, echocardiography is recommended in the event of haemodynamic dysfunction, stress corticosteroids in hypotension refractory to vasopressors, monitoring of lactate and pyruvate on introduction of feeding and during ventilator weaning, and nutrition within the first 48 hours [14]. For anaesthesia, minimise fasting, add glucose to the fluids, exercise caution with volatile anaesthetics and with muscle relaxants in myopathic patients or those with respiratory depression, and limit propofol to short procedures; local anaesthetics are well tolerated and no established link with malignant hyperthermia exists [11]. The 2020 Delphi consensus considers volatile anaesthetics, propofol for induction, ketamine, fentanyl and non-depolarising neuromuscular blockers safe, with additional monitoring in the myopathic phenotype [54]. Deterioration with intercurrent illnesses and death in cardiogenic shock or metabolic crisis have been described in infants [30, 33].

In acute primary hyperlactataemia of the newborn and infant, Saudubray describes metabolic acidosis with an increased anion gap and dehydration, treated with supportive measures that usually suffice to bring lactate below the level that causes severe acidosis; when hyperlactataemia persists because of a glucose-rich infusion, it is corrected by lowering the intravenous glucose concentration to 5% or even 2.5% [59]. This indication complements that of the MMS, which recommends intravenous dextrose in decompensation: glucose prevents catabolism, but the rate should be adjusted to the lactate [11, 59]. Extracorporeal clearance is not indicated (the exception is TMEM70 deficiency). A trial of thiamine, riboflavin and biotin is recommended, as is treatment of secondary carnitine deficiency with L-carnitine [59]. Dichloroacetate, an inhibitor of pyruvate dehydrogenase kinase, at 50 mg/kg/day in one or two doses in the newborn and infant with primary hyperlactataemia (with no maximum dose defined in the source; the medicines table of the same book gives 25 mg/kg/day orally in two doses in primary lactic acidosis [53]), lowers lactate and corrects the acidosis in pyruvate dehydrogenase and respiratory chain defects, but has little effect on the clinical state beyond reducing tachypnoea, and chronic use causes reversible sensorimotor peripheral neuropathy; it is not an authorised medicine and its use is experimental [59]. In adults with encephalopathy without a diagnosis, a short trial of high-dose B vitamins (thiamine, riboflavin, pyridoxine, biotin, folate and cobalamin) is recommended [59].

7.9. Drugs and safety

The Delphi consensus concluded that the 46 drugs or groups studied are, in general, safe in primary mitochondrial diseases, recommending monitoring for adverse effects and of lactate in the first days for the drugs without clear evidence of mitochondrial toxicity in vivo; the level of evidence is that of expert opinion [54]. The 2017 MMS guideline recommends caution with statins, corticosteroids, metformin and antiretrovirals in mitochondrial myopathy [14].

7.10. Pregnancy, transition and other situations

The 2017 MMS guideline recommends close monitoring by the specialists, follow-up by a high-risk obstetric service (risk of preterm birth and pre-eclampsia), early and late oral glucose tolerance testing and preconception counselling [14]. Clinical Cardiogenetics adds the risk of cardiac deterioration and of magnesium sulphate toxicity [40]. In ACAD9 deficiency, a normal pregnancy has been described, with elective caesarean section and a healthy newborn girl, at 26 years of age, in a woman with myopathy, HCM and epilepsy on riboflavin since the age of 4 [34]. For air travel, patients with cardiomyopathy should consider monitoring of oxygen saturation and access to supplemental oxygen [14]. The sources consulted do not describe a specific transition programme; the 2017 MMS guideline recommends that care be coordinated by a mitochondrial disease centre [14].

7.11. Therapeutic studies

These studies do not allow comparisons with one another: they differ in population, design and outcome. The comparison between treated and untreated patients in the cohort is observational, without randomisation or control of the date treatment began, and is subject to survival selection bias; the results in fibroblasts do not necessarily translate into clinical benefit, as the negative bezafibrate trial in VLCAD deficiency showed [10]. The authors of the cohort called for a controlled trial of riboflavin, which had not been registered by the reference date [10, 45].

7.12. Developments of the past 6 to 12 months

A search of ClinicalTrials.gov on 31 October 2026 found no therapeutic trials directed at ACAD9 deficiency and no riboflavin trials in complex I deficiency; the only records with the term "ACAD9" are the Baby Detect genomic newborn screening study (completed) and a compassionate use programme of triheptanoin in disorders of energy metabolism, no longer available [45]. There are no applications for authorisation, opinions or regulatory authorisations for this condition at the FDA or the European Medicines Agency in the sources consulted [56]. The 2025–2026 literature is at the level of case reports: a late and mild phenotype with a novel variant (p.(Val546Met)) that reduces complex I activity by more than 50% in a cell model; slowly progressive encephalomyopathy in an adult; and fatal infantile HCM with homozygous p.(Ala220Val), with death before the genetic result [21, 30, 35]. At the preclinical level, the structure of the ECSIT–ACAD9 complex published in 2023 opened the possibility of strategies directed at the assembly function [29].

7.13. Divergences between sources

The clinically relevant divergences are four. The riboflavin dose: 20 mg/kg/day with a maximum of 200 mg/day (cohort) and 50–200 mg/day (Australian guideline) against 100–400 mg/day (Saudubray) [10, 13, 53]. General supplements: offered to most patients by the 2015 MMS guideline, without sufficient evidence according to the Australian adaptation [11, 13]. The high-fat diet in complex I deficiency: accepted by the Australian adaptation and by Clinical Cardiogenetics, without evidence according to the MMS [11, 13, 40]. The drugs to be avoided: precautionary lists of the 2015 MMS guideline and of Clinical Cardiogenetics against the conclusion of general safety of the 2020 Delphi consensus, which is the most recent [11, 40, 54].

8. Reproductive Options and Family Counselling

8.1. Recurrence risks

ACAD9 deficiency is transmitted in an autosomal recessive manner [2]. When both parents are heterozygous, the risk of each pregnancy being affected is 25%, that of being an unaffected carrier 50% and that of inheriting neither variant 25% [5, 30]. Consanguinity is frequent in the families described (Gerards' family; first cousins in Kotlukova's case) [25, 30]. Heterozygotes have, according to the sources consulted, no clinical manifestations. The offspring of an adult patient are affected only if the other parent is also a carrier; the woman described with a pregnancy at 26 years of age had a healthy daughter [34]. No founder variants have been described, so the carrier frequency in the general population cannot be estimated from a common allele [10].

8.2. Cascade testing and testing of siblings

Once the two variants have been identified, the parents should be tested to confirm the phase, and the siblings may be studied [40]. Testing of minor siblings has a clinical justification, because riboflavin started immediately after diagnosis was associated with the survival of younger siblings in families 1 and 33 of the cohort of 70 patients and because early cardiac surveillance is possible [10, 14]. In Kotlukova's case, carrier testing was proposed for the healthy sisters [30]. When the genetic cause is not identified, Clinical Cardiogenetics suggests clinical assessment of first-degree relatives, with lactate, amino acids, glucose and organic acids according to the proband and periodic cardiological assessment when cardiomyopathy is the main feature [40]. The sources consulted do not describe specific recommendations on carrier testing in healthy minors in this disease.

8.3. Prenatal and preimplantation diagnosis

With the familial variants known, prenatal diagnosis by chorionic villus sampling or amniocentesis and preimplantation genetic diagnosis are possible and follow the principles of Mendelian diseases [5, 49]. The 2023 ESC guideline recommends discussing these options in good time when counselling couples with a child affected by a hereditary cardiomyopathy with known variants, taking into account cultural, religious, legal and access-related aspects [15]. Prenatal diagnosis based on enzyme activity in fetal cells has given errors and should not replace molecular testing [46]. On ultrasound, fetal HCM or cardiomegaly with intrauterine growth restriction, in a family without a diagnosis, justify including ACAD9 in prenatal investigations [36, 39].

8.4. Content of counselling

Counselling should address the prognostic variability linked to the age at onset, the incomplete response to riboflavin (about two thirds with reported benefit), the relative cognitive and functional preservation of survivors, the limited significance of complex I activity in fibroblasts and the uncertainties about heart transplantation in early-onset patients [10, 26]. Genetic information retains value for the family even when it arrives too late for the proband, as in Kotlukova's case [30].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Frequency and what each figure measures

The prevalence of ACAD9 deficiency is unknown: Orphanet records the prevalence class "unknown" and 23 validated cases, and the parent entity, isolated complex I deficiency, also has no estimated prevalence [3, 8]. In 2017, Aintablian and colleagues counted fewer than 25 published cases [12]; in 2018, Repp's cohort gathered 70 patients from 50 families, 29 of them previously unpublished [10]. From the frequencies of deleterious alleles in gnomAD, Repp and colleagues estimated about 59 affected births per year in Europe and 689 worldwide; this is a genetic estimate, not an observed incidence, and depends on the criteria used to classify the alleles [10]. The proportion among disease causes indicates the relative importance: ACAD9 was responsible for 8 of 20 (40%) French cases of infantile cardiac hypertrophy with isolated complex I deficiency and is considered one of the commonest causes of respiratory chain deficiency [10, 26].

For the group of mitochondrial diseases, the prevalence estimates are 5–15 per 100,000 people with childhood onset and 10–15 per 100,000 with adult onset, and the incidence of respiratory chain dysfunction in children under 6 years of age in Sweden was 1 in 11,000 [40, 60]. Severe HCM with onset in infancy has an estimated frequency of 1 in 47,000 newborns, according to the authors of the 2026 case [30]. These figures refer to groups and cannot be converted into a prevalence of ACAD9 deficiency.

9.2. Geography

The cohort of 70 patients was gathered through a collaboration of centres in Germany, the United Kingdom, France, Italy, Belgium, the Netherlands, Austria, Switzerland, Denmark, Poland, China and the United States; no founder variants were identified [10]. The published isolated cases come from several continents, including Russia, Israel, Martinique (France) and China [21, 30, 34, 35]. In the sources consulted there are no data on ACAD9 deficiency in Portugal or Brazil; in a Portuguese cohort of 40 patients with the Leigh syndrome spectrum, ACAD9 appeared only in the gene panel studied, with no patients identified in the sources read [61]. The absence of data in these sources does not demonstrate an absence of cases.

9.3. Special populations

The groups requiring particular attention are the fetus and the newborn (presentations with growth restriction, cardiomegaly and fatal lactic acidosis), the pregnant woman with mitochondrial disease and the patient who is a candidate for heart transplantation [14, 26, 39]. Consanguineous families and those who have already lost a child to cardiomyopathy benefit from early molecular diagnosis and reproductive counselling [30].

9.4. Reference networks and patient associations

In Europe, patients fall within MetabERN (https://metab.ern-net.eu/), ERN GUARD-Heart for cardiomyopathy (https://guardheart.ern-net.eu/) and ERN EURO-NMD for myopathy (https://ern-euro-nmd.eu/) [16–18]. The Mitochondrial Medicine Society (https://www.mitosoc.org/) publishes the standards of care cited [62]. The patient associations with information on mitochondrial diseases include the United Mitochondrial Disease Foundation, in the United States (https://www.umdf.org/), The Lily Foundation, in the United Kingdom, which funded part of the cohort of 70 patients (https://www.thelilyfoundation.org.uk/), the Mito Foundation, in Australia (https://www.mito.org.au/), and International Mito Patients (https://www.mitopatients.org/), whose table of potentially harmful drugs was reviewed in the 2020 Delphi consensus [10, 54, 63–66].

10. Illustrative Clinical Case

Kotlukova and colleagues published in 2026 the case of a 4-month-old infant, the daughter of first cousins, born of a fourth pregnancy without complications, by elective term caesarean section, weighing 3170 g with Apgar scores of 8/8 [30]. The family had two healthy daughters aged 9 and 6 years and a third daughter who had died at 4 months with a similar picture, attributed to progressive cardiomyopathy; the pedigree is in Figure 1 of the article (https://pmc.ncbi.nlm.nih.gov/articles/PMC13431430/figure/F1/) [30]. Because of the family history, the child had an echocardiogram at 2 months, which was normal, and a repeat at 3 months, which showed for the first time signs of HCM, after which she was referred to a federal centre in Moscow [30].

On admission, the blood glucose was 3.8 mmol/L and normalised with a glucose infusion; lactate rose progressively from 4.1 to 6.0, 7.9 and 8.0 mmol/L (reference 0.5–1.6), with only partial and transient falls with fluids, and the base deficit ranged between −11.1 and +0.4, consistent with persistent metabolic acidosis [30]. Echocardiography showed symmetrical biventricular hypertrophy without obstruction (outflow tract gradient of 6 mmHg): interventricular septum 9.0 mm (Z-score 3.76), posterior wall 8.3 mm (Z-score 4.82), myocardial mass with a Z-score of 6.7, normal cavity dimensions and preserved systolic function (ejection fraction 70.4%, fractional shortening 38%) (Figure 2: https://pmc.ncbi.nlm.nih.gov/articles/PMC13431430/figure/F2/) [30]. The ECG showed sinus rhythm with signs of biventricular hypertrophy and isolated ectopic beats, and the 24-hour Holter recording a tendency to tachycardia (Figure 3: https://pmc.ncbi.nlm.nih.gov/articles/PMC13431430/figure/F3/); troponin I was normal and NT-proBNP was 3055 ng/mL [30]. There were minimal, non-specific dysmorphic features and suspected hearing loss; the acylcarnitines were normal and the urinary organic acids showed non-specific abnormalities [30].

Levocarnitine and spironolactone were started and, on discharge, it was recommended to avoid prolonged fasting, with spironolactone 2 mg/kg/day, coenzyme Q10 and levocarnitine 100 mg/kg/day, with immediate admission in the event of unexplained weakness or lethargy; riboflavin was not started because discharge took place before the genetic result [30]. The exome identified the variant NM_014049.5:c.659C>T p.(Ala220Val) in homozygosity, classified as likely pathogenic and deposited in ClinVar, with both parents confirmed as carriers by Sanger sequencing (Figure 4: https://pmc.ncbi.nlm.nih.gov/articles/PMC13431430/figure/F4/) [30, 32]. The child died at home, from a metabolic crisis and acute cardiovascular failure, before the result was known; the family was subsequently invited for genetic counselling and testing of the sisters (timeline in Figure 5: https://pmc.ncbi.nlm.nih.gov/articles/PMC13431430/figure/F5/) [30].

The case brings together the warning signs of the infant with HCM of metabolic cause: consanguinity, a sister who died of cardiomyopathy at the same period of life, biventricular HCM with lactic acidosis and normal acylcarnitines [15, 30]. The same variant had already been described, in homozygosity, in an infant with severe HCM and death at 6 months, whose fibroblasts did not improve with riboflavin [33]. The management in the report (carnitine and coenzyme Q10, without riboflavin until the genetic result) differs from the recommendation of Repp and colleagues to start riboflavin when the phenotype is compatible while genetic testing is under way, although, with this genotype, the response to riboflavin would have been uncertain [10, 33].

11. Prognosis and Long-Term Follow-Up

11.1. Prognosis

The strongest prognostic factor is the age at presentation: 50% of deaths by 2 years of age in patients with onset in the first year, against survival above 90% at 10 years in those with later onset [10]. The median age at death in the cohort of 70 patients was 3 months and the living patients had a median age of 14 years, up to 44 years [10]. In the group of childhood-onset mitochondrial diseases, the presence of cardiomyopathy was associated with 71% mortality against 26% without cardiomyopathy, survival at 16 years was 16% against 95%, and onset before 6 months multiplied the risk of death tenfold [40]. Riboflavin was associated with lower mortality in early-onset patients (7 of 22 deaths among those treated against 16 of 17 among those untreated), with the limitations of a retrospective study [10]. Severity correlates with the residual dehydrogenase activity of ACAD9, not with complex I activity [28]. Survivors have, for the most part, functional capacity close to that of their peers (more than 70% with equivalent activities of daily living) [10], but may later develop neurological, muscular, renal, ovarian or optic manifestations, including after heart transplantation [26]. Adult life, with pregnancy, has been documented on riboflavin [27, 34].

11.2. Surveillance

11.3. Long-term follow-up

Follow-up should be provided at a tertiary centre with experience in metabolic diseases and in the cardiology of mitochondrial diseases, with an emergency document updated at every visit [14, 40]. Annual nutritional assessment in children, growth monitoring and screening for endocrinopathies form part of the general standards of care [13, 14]. In transplant recipients, surveillance should extend to the organs that may be affected later [26].

12. Summary and Final Considerations

ACAD9 deficiency is a deficiency of respiratory chain complex I, with autosomal recessive transmission, caused by loss of function of a protein that is simultaneously a complex I assembly factor and a long-chain acyl-CoA dehydrogenase. Described in 2007 as a β-oxidation disorder and redefined in 2010 as a complex I deficiency, it is one of the most frequent causes of respiratory chain deficiency and the most frequent cause of hypertrophic cardiomyopathy with isolated complex I deficiency in childhood. The typical presentation is that of an infant with HCM, lactic acidosis and weakness, with cognition as a rule preserved, but the spectrum includes lethal fetal forms, exercise intolerance in adults, optic neuropathy and encephalomyopathy.

The prognosis depends on the age at onset: half the patients with onset in the first year die by 2 years of age, and more than 90% of those presenting later are alive at 10 years. Complex I activity may be normal and the acylcarnitine profiles are usually uninformative, so the diagnosis is molecular, by exome sequencing or a broad panel, ideally in a trio. Riboflavin, at 20 mg/kg/day up to 200 mg/day, was associated with clinical benefit in about two thirds of the treated patients and with lower mortality in those with early onset in a retrospective cohort, and should be started when the phenotype is compatible while genetic results are awaited. No oral riboflavin medicine is authorised in the United States or in Portugal for this indication, and there are no controlled trials.

Management combines riboflavin with cardiological treatment according to the guidelines, prevention of catabolism, critical review of medication in the light of the 2020 consensus and annual cardiac surveillance that may be spaced out in stable patients. Heart transplantation had good results in late-onset patients and poor results in early-onset ones, with a risk of subsequent extracardiac manifestations. Genetic counselling should include testing of the siblings, because of the advantage of early treatment, and the options of prenatal and preimplantation diagnosis.

13. References

1. Orphanet. Acyl-CoA dehydrogenase 9 deficiency (ORPHA:99901). Available at: https://www.orpha.net/en/disease/detail/99901
2. OMIM. #611126 Mitochondrial complex I deficiency, nuclear type 20; MC1DN20. Baltimore: Johns Hopkins University. Available at: https://omim.org/entry/611126
3. GeneGuidance. ACAD9 — curation of the gene and of the condition (Mitochondrial complex I deficiency due to ACAD9 deficiency). Version dated 2026-09-26.
4. OMIM. #252010 Mitochondrial complex I deficiency, nuclear type 1; MC1DN1 (genetic heterogeneity of complex I deficiency). Available at: https://omim.org/entry/252010
5. Rahman S, Mayr JA. Disorders of oxidative phosphorylation. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 10, sections 10.1–10.5 (Tables 10.4 and 10.5). doi:10.1007/978-3-662-63123-2_10
6. OMIM. *611103 Acyl-CoA dehydrogenase family, member 9; ACAD9. Available at: https://omim.org/entry/611103
7. Freeman PJ, Hart RK, Gretton LJ, Brookes AJ, Dalgleish R. VariantValidator: accurate validation, mapping, and formatting of sequence variation descriptions. Hum Mutat. 2018;39(1):61-68. doi:10.1002/humu.23348. PMID: 28967166. Accessed (GRCh38, NM_014049.5) on 2026-10-31.
8. Orphanet. Isolated complex I deficiency (ORPHA:2609). Available at: https://www.orpha.net/en/disease/detail/2609
9. Ganetzky R, Kong C, Elsharkawi I. Mitochondrial short-chain enoyl-CoA hydratase deficiency. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 2019 [updated 2026 Mar 12]. Available at: https://www.ncbi.nlm.nih.gov/books/NBK542806/
10. Repp BM, Mastantuono E, Alston CL, Schiff M, Haack TB, Rötig A, et al. Clinical, biochemical and genetic spectrum of 70 patients with ACAD9 deficiency: is riboflavin supplementation effective? Orphanet J Rare Dis. 2018;13(1):120. doi:10.1186/s13023-018-0784-8. PMID: 30025539; PMC6053715.
11. Parikh S, Goldstein A, Koenig MK, Scaglia F, Enns GM, Saneto R, et al. Diagnosis and management of mitochondrial disease: a consensus statement from the Mitochondrial Medicine Society. Genet Med. 2015;17(9):689-701. doi:10.1038/gim.2014.177. PMID: 25503498; PMC5000852.
12. Aintablian HK, Narayanan V, Belnap N, Ramsey K, Grebe TA. An atypical presentation of ACAD9 deficiency: Diagnosis by whole exome sequencing broadens the phenotypic spectrum and alters treatment approach. Mol Genet Metab Rep. 2017;10:38-44. doi:10.1016/j.ymgmr.2016.12.005. PMID: 28070495; PMC5219625.
13. Sue CM, Balasubramaniam S, Bratkovic D, Bonifant C, Christodoulou J, Coman D, et al. Patient care standards for primary mitochondrial disease in Australia: an Australian adaptation of the Mitochondrial Medicine Society recommendations. Intern Med J. 2022;52(1):110-120. doi:10.1111/imj.15505. PMID: 34505344; PMC9299181.
14. Parikh S, Goldstein A, Karaa A, Koenig MK, Anselm I, Brunel-Guitton C, et al. Patient care standards for primary mitochondrial disease: a consensus statement from the Mitochondrial Medicine Society. Genet Med. 2017;19(12):1380-1397. doi:10.1038/gim.2017.107. PMID: 28749475; PMC7804217.
15. Arbelo E, Protonotarios A, Gimeno JR, Arbustini E, Barriales-Villa R, Basso C, et al. 2023 ESC Guidelines for the management of cardiomyopathies. Eur Heart J. 2023;44(37):3503-3626. doi:10.1093/eurheartj/ehad194. PMID: 37622657.
16. MetabERN — European Reference Network for Hereditary Metabolic Disorders. Available at: https://metab.ern-net.eu/
17. ERN GUARD-Heart — European Reference Network for rare, low prevalence and complex diseases of the heart. Available at: https://guardheart.ern-net.eu/
18. ERN EURO-NMD — European Reference Network for neuromuscular diseases. Available at: https://ern-euro-nmd.eu/
19. Morris AAM, Spiekerkoetter U. Disorders of mitochondrial fatty acid oxidation & riboflavin metabolism. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 12, sections 12.1 (ACAD9) and 12.2. doi:10.1007/978-3-662-63123-2_12
20. Gueguen N, Piarroux J, Sarzi E, Benkirane M, Manes G, Delettre C, et al. Optic neuropathy linked to ACAD9 pathogenic variants: A potentially riboflavin-responsive disorder? Mitochondrion. 2021;59:169-174. doi:10.1016/j.mito.2021.05.002. PMID: 34023438.
21. Li S, Li Y, Chen Y. Mitochondrial encephalomyopathy caused by a novel ACAD9 mutation: a case report. Front Hum Neurosci. 2026;20:1828023. doi:10.3389/fnhum.2026.1828023. PMID: 42370067; PMC13294475.
22. He M, Rutledge SL, Kelly DR, Palmer CA, Murdoch G, Majumder N, et al. A new genetic disorder in mitochondrial fatty acid beta-oxidation: ACAD9 deficiency. Am J Hum Genet. 2007;81(1):87-103. doi:10.1086/519219. PMID: 17564966; PMC1950923.
23. Nouws J, Nijtmans L, Houten SM, van den Brand M, Huynen M, Venselaar H, et al. Acyl-CoA dehydrogenase 9 is required for the biogenesis of oxidative phosphorylation complex I. Cell Metab. 2010;12(3):283-294. doi:10.1016/j.cmet.2010.08.002. PMID: 20816094.
24. Haack TB, Danhauser K, Haberberger B, Hoser J, Strecker V, Boehm D, et al. Exome sequencing identifies ACAD9 mutations as a cause of complex I deficiency. Nat Genet. 2010;42(12):1131-1134. doi:10.1038/ng.706. PMID: 21057504.
25. Gerards M, van den Bosch BJ, Danhauser K, Serre V, van Weeghel M, Wanders RJ, et al. Riboflavin-responsive oxidative phosphorylation complex I deficiency caused by defective ACAD9: new function for an old gene. Brain. 2011;134(Pt 1):210-219. doi:10.1093/brain/awq273. PMID: 20929961.
26. Collet M, Assouline Z, Bonnet D, Rio M, Iserin F, Sidi D, et al. High incidence and variable clinical outcome of cardiac hypertrophy due to ACAD9 mutations in childhood. Eur J Hum Genet. 2016;24(8):1112-1116. doi:10.1038/ejhg.2015.264. PMID: 26669660; PMC4970679.
27. Dewulf JP, Barrea C, Vincent MF, De Laet C, Van Coster R, Seneca S, et al. Evidence of a wide spectrum of cardiac involvement due to ACAD9 mutations: Report on nine patients. Mol Genet Metab. 2016;118(3):185-189. doi:10.1016/j.ymgme.2016.05.005. PMID: 27233227.
28. Schiff M, Haberberger B, Xia C, Mohsen AW, Goetzman ES, Wang Y, et al. Complex I assembly function and fatty acid oxidation enzyme activity of ACAD9 both contribute to disease severity in ACAD9 deficiency. Hum Mol Genet. 2015;24(11):3238-3247. doi:10.1093/hmg/ddv074. PMID: 25721401; PMC4424958.
29. McGregor L, Acajjaoui S, Desfosses A, Saïdi M, Bacia-Verloop M, Schwarz JJ, et al. The assembly of the Mitochondrial Complex I Assembly complex uncovers a redox pathway coordination. Nat Commun. 2023;14(1):8248. doi:10.1038/s41467-023-43865-0. PMID: 38086790; PMC10716376.
30. Kotlukova N, Kadykova A, Gorelova T, Dzik L, Islanov I, Zaklyazminskaya E. Case Report: Homozygous mutation in the ACAD9 gene revealed in a pediatric patient initially diagnosed with familial cardiac hypertrophy. Front Med (Lausanne). 2026;13:1824617. doi:10.3389/fmed.2026.1824617. PMID: 42549236; PMC13431430.
31. Leslie N, Wang X, Peng Y, Valencia CA, Khuchua Z, Hata J, et al. Neonatal multiorgan failure due to ACAD9 mutation and complex I deficiency with mitochondrial hyperplasia in liver, cardiac myocytes, skeletal muscle, and renal tubules. Hum Pathol. 2016;49:27-32. doi:10.1016/j.humpath.2015.09.039. PMID: 26826406.
32. National Center for Biotechnology Information. ClinVar: ACAD9 variants (VCV000030882, VCV000030884, VCV000858760, VCV004526405). Available at: https://www.ncbi.nlm.nih.gov/clinvar/?term=ACAD9%5Bgene%5D
33. Nouws J, Wibrand F, van den Brand M, Venselaar H, Duno M, Lund AM, et al. A Patient with Complex I Deficiency Caused by a Novel ACAD9 Mutation Not Responding to Riboflavin Treatment. JIMD Rep. 2014;12:37-45. doi:10.1007/8904_2013_242. PMID: 23996478; PMC3897792.
34. Jacobi-Polishook T, Yosha-Orpaz N, Sagi Y, Lev D, Lerman-Sagie T. Successful pregnancy in a patient with mitochondrial cardiomyopathy due to ACAD9 deficiency. JIMD Rep. 2020;56(1):9-13. doi:10.1002/jmd2.12157. PMID: 33204590; PMC7653261.
35. Giguet-Valard AG, Ait-El-Mkadem Saadi S, Duclos S, Lacombe D, Bellance R, Bellance N. A Late-Onset and Mild Phenotype of Mitochondrial Complex I Deficiency Due to a Novel Reported Variant Within the ACAD9 Gene. Int J Mol Sci. 2025;26(15):7128. doi:10.3390/ijms26157128. PMID: 40806260; PMC12346219.
36. Lagoutte-Renosi J, Ségalas-Milazzo I, Crahes M, Renosi F, Menu-Bouaouiche L, Torre S, et al. Lethal Neonatal Progression of Fetal Cardiomegaly Associated to ACAD9 Deficiency. JIMD Rep. 2016;28:1-10. doi:10.1007/8904_2015_499. PMID: 26475292; PMC5059192.
37. Van Hove JLK, Friederich MW, Hock DH, Stroud DA, Caruana NJ, Christians U, et al. ACAD9 treatment with bezafibrate and nicotinamide riboside temporarily stabilizes cardiomyopathy and lactic acidosis. Mitochondrion. 2024;78:101905. doi:10.1016/j.mito.2024.101905. PMID: 38797357; PMC11390326.
38. Theunissen TEJ, Gerards M, Hellebrekers DMEI, van Tienen FH, Kamps R, Sallevelt SCEH, et al. Selection and Characterization of Palmitic Acid Responsive Patients with an OXPHOS Complex I Defect. Front Mol Neurosci. 2017;10:336. doi:10.3389/fnmol.2017.00336. PMID: 29093663; PMC5651253.
39. Dubucs C, Aziza J, Sartor A, Heitz F, Sevely A, Sternberg D, et al. Severe Antenatal Hypertrophic Cardiomyopathy Secondary to ACAD9-Related Mitochondrial Complex I Deficiency. Mol Syndromol. 2023;14(2):101-108. doi:10.1159/000526022. PMID: 37064341; PMC10091013.
40. Blank AC, Breur JMPJ, Fuchs SA, Koop K, Baas AF. Mitochondrial cardiomyopathies. In: Baars HF, van der Smagt JJ, Doevendans PAFM, van Tintelen JP, editors. Clinical Cardiogenetics. 3rd ed. Cham: Springer; 2020. p. 179–91. doi:10.1007/978-3-030-45457-9_11
41. Saudubray JM, García-Cazorla Á. Clinical approach to inborn errors of metabolism in paediatrics. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 1, sections 1.4.8–1.4.9 (Tables 1.12 and 1.13). doi:10.1007/978-3-662-63123-2_1
42. Mochel F, Sedel F. Inborn errors of metabolism in adults: a diagnostic approach to neurological and psychiatric presentations. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 2, sections 2.2.5 and 2.3.10 (Table 2.12). doi:10.1007/978-3-662-63123-2_2
43. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. p. 304–9 (Very long-chain acyl-CoA dehydrogenase deficiency).
44. Lunke S, Downie L, Caruana J, Kugenthiran N, De Fazio P, Hollizeck S, et al. Feasibility, acceptability and clinical outcomes of the BabyScreen+ genomic newborn screening study. Nat Med. 2025;31(12):4236-4245. doi:10.1038/s41591-025-03986-z. PMID: 41068466; PMC12705431.
45. ClinicalTrials.gov. Search for "ACAD9" and "complex I deficiency riboflavin" (API v2). Available at: https://clinicaltrials.gov/
46. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. p. 353–63 (Introduction to lactic acidemias).
47. Touati G, Mochel F, Artuch R. Diagnostic procedures. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 3, sections 3.1.2, 3.2.1 and 3.2.4 (Table 3.4). doi:10.1007/978-3-662-63123-2_3
48. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 84–7 (Cardiomyopathy in children under 10 years).
49. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 474–5 (Leigh encephalopathy).
50. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. p. 331–40 (Multiple acyl-CoA dehydrogenase deficiency).
51. Linhart A. Metabolic cardiomyopathy. In: Baars HF, van der Smagt JJ, Doevendans PAFM, van Tintelen JP, editors. Clinical Cardiogenetics. 3rd ed. Cham: Springer; 2020. p. 151–66. doi:10.1007/978-3-030-45457-9_9
52. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 460–3 (Hypertrophic cardiomyopathy).
53. Morris AAM, Jones S. Medications used in the treatment of inborn errors of metabolism. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 45, table (riboflavin). doi:10.1007/978-3-662-63123-2_45
54. De Vries MC, Brown DA, Allen ME, Bindoff L, Gorman GS, Karaa A, et al. Safety of drug use in patients with a primary mitochondrial disease: An international Delphi-based consensus. J Inherit Metab Dis. 2020;43(4):800-818. doi:10.1002/jimd.12196. PMID: 32030781; PMC7383489.
55. Parikh S, Saneto R, Falk MJ, Anselm I, Cohen BH, Haas R, et al. A modern approach to the treatment of mitochondrial disease. Curr Treat Options Neurol. 2009;11(6):414-430. doi:10.1007/s11940-009-0046-0. PMID: 19891905; PMC3561461.
56. U.S. Food and Drug Administration. Drugs@FDA: search for riboflavin (openFDA drugsfda, 7 applications; none oral). Available at: https://api.fda.gov/drug/drugsfda.json?search=openfda.generic_name:%22RIBOFLAVIN%22
57. INFARMED. Infomed — search for "riboflavina" (Stimulex, revoked). Available at: https://extranet.infarmed.pt/INFOMED-fo/
58. Kadoya T, Sakakibara A, Kitayama K, Yamada Y, Higuchi S, Kawakita R, et al. Successful treatment of infantile-onset ACAD9-related cardiomyopathy with a combination of sodium pyruvate, beta-blocker, and coenzyme Q10. J Pediatr Endocrinol Metab. 2019;32(10):1181-1185. doi:10.1515/jpem-2019-0205. PMID: 31473688.
59. Schiff M, Mochel F, Dionisi-Vici C. Emergency treatments. In: Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 4, sections 4.2.1.3 and 4.2.5. doi:10.1007/978-3-662-63123-2_4
60. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 490–5 (Mitochondrial DNA diseases).
61. Baldo MS, Nogueira C, Pereira C, Janeiro P, Ferreira S, Lourenço CM, et al. Leigh Syndrome Spectrum: A Portuguese Population Cohort in an Evolutionary Genetic Era. Genes (Basel). 2023;14(8):1536. doi:10.3390/genes14081536. PMID: 37628588; PMC10454233.
62. Mitochondrial Medicine Society. Available at: https://www.mitosoc.org/
63. United Mitochondrial Disease Foundation. Available at: https://www.umdf.org/
64. The Lily Foundation. Available at: https://www.thelilyfoundation.org.uk/
65. Mito Foundation. Available at: https://www.mito.org.au/
66. International Mito Patients. Available at: https://www.mitopatients.org/