Vol.:(0123456789) Journal of Neurology (2026) 273:263 https://doi.org/10.1007/s00415-026-13794-1 NEUROLOGICAL UPDATE Non‑invasive biomarkers for diagnosis and monitoring of primary mitochondrial diseases Ignazio Giuseppe Arena1 · Shamini Saravanabavan1 · Rita Horvath1 · Jelle van den Ameele1,2  Received: 8 February 2026 / Revised: 26 March 2026 / Accepted: 31 March 2026 © The Author(s) 2026 Abstract Primary mitochondrial diseases (PMDs) represent a clinically and genetically heterogeneous group of disorders characterized by impaired oxidative phosphorylation and multisystem involvement, commonly affecting the nervous system. As therapeutic development accelerates, there is a growing need for robust biomarkers capable of supporting diagnosis, stratifying patient subgroups, monitoring disease progression, and providing sensitive pharmacodynamic readouts for clinical trials. This review summarizes recent advances in three major non-invasive biomarker domains relevant to PMDs: circulating serum and molecular biomarkers, functional and digital endpoints, and neuroimaging modalities. Circulating markers, such as FGF21, GDF15, NfL, and NAD⁺-related signatures, have each been proposed for diagnosis and to follow disease progression, while multi-omics approaches are paving the way toward integrated molecular phenotyping. Digital health technologies, including accelerometry and gait analytics, enable objective quantification of real-world functional impairment, although disease-specific validation remains an unmet need. Neuroimaging offers mechanistic insights through metabolic (MRS, CEST), perfusion (ASL), and molecular modalities (mitochondrial PET tracers). Cutting-edge tools, such as Multi-Spectral Optoacoustic Tomography (MSOT), Raman spectroscopy, and Near-Infrared Spectroscopy (NIRS), promise real-time or spatially resolved assessment of mitochondrial function. Together, these developments outline multidimensional biomarker approaches for PMDs, with the potential to directly measure target engagement and clinically meaningful phenotypes in future therapeutic trials. Future progress will depend on longitudinal validation, harmonized acquisition protocols, and the integration of multimodal platforms to support upcoming therapeutic trials and precision medicine strategies. Keywords  Mitochondrial disease · Biomarkers · Functional endpoints · Wearable devices · Digital health technologies · Phenotyping · Neuroimaging · Magnetic resonance imaging · Positron emission tomography · Precision medicine · Clinical trials Introduction Primary mitochondrial diseases (PMDs) comprise a clini- cally heterogeneous group of disorders caused by patho- genic variants in mitochondrial or nuclear genes, resulting in impaired oxidative phosphorylation (OXPHOS). With an estimated prevalence of approximately 1 in 4,300 individu- als [1], PMDs represent one of the most common inherited metabolic disorders. They may present with a wide range of neurological and multisystem manifestations, each posing significant challenges for diagnosis and management [2]. Despite major advances in genetic diagnosis [3] and the understanding of mitochondrial disease pathophysiology [4, 5], effective therapeutic options remain limited [6]. To date, only three drugs have received regulatory approval for the treatment of specific forms of PMDs: idebenone for Leber’s Hereditary Optic Neuropathy (LHON) [7] in Europe, and more recently, elamipretide for Barth Syndrome [8] and deoxy-pyrimidines for Thymidine kinase 2 (TK2)- related mtDNA depletion syndrome [9, 10] in the United States. However, most mitochondrial syndromes still lack Rita Horvath and Jelle van den Ameele have contributed equally. * Ignazio Giuseppe Arena ia454@cam.ac.uk * Jelle van den Ameele jv361@cam.ac.uk 1 Department of Clinical Neurosciences, University of Cambridge, Cambridge, UK 2 MRC Mitochondrial Biology Unit, University of Cambridge, Cambridge, UK http://crossmark.crossref.org/dialog/?doi=10.1007/s00415-026-13794-1&domain=pdf http://orcid.org/0000-0002-2744-0810 Journal of Neurology (2026) 273:263 263   Page 2 of 14 disease-modifying interventions [6, 10–13], and an increas- ing number of emerging therapeutic strategies underscore the urgent need for reliable and clinically meaningful bio- markers [14, 15]. Biomarkers, defined as measurable indicators of normal biological processes, pathogenic mechanisms, or responses to therapeutic interventions [16], hold particular promise in PMDs. Given the pronounced clinical heterogeneity and diagnostic complexity of these disorders, robust biomarkers with direct relevance for the underlying pathophysiology are essential to improve diagnostic accuracy and patient strati- fication. There is an urgent need for biomarkers to monitor disease progression, predicting acute neurological events such as stroke-like episodes or sudden visual loss, and objec- tively assessing treatment responses or target engagement in experimental medicine studies and clinical trials [17]. Recent years have witnessed progress in biomarker dis- covery, driven by targeted and untargeted omics approaches, advances in imaging, and in digital and functional assess- ments tailored to specific phenotypes [18]. However, despite their promise, most still require rigorous validation. The translation of candidate biomarkers into clinical practice or trials requires a progressive stepwise process of analytical and clinical validation followed by regulatory qualification within frameworks established by agencies, such as the European Medicines Agency (EMA), the U.S. Food and Drug Administration (FDA), or the UK Medicines and Healthcare products Regulatory Agency (MHRA). Bio- markers must demonstrate robust validity, including repro- ducibility, sensitivity and specificity and must be clearly defined within a specific “context of use” (e.g., diagnos- tic, monitoring, pharmacodynamic (response), predictive, etc.)[19]. This process typically requires evidence derived from multiple independent cohorts, standardized analytical methods, and harmonized protocols across centers. These challenges are amplified in rare diseases such as PMDs due to limited patient populations and heterogeneous disease progression [20]. Consequently, trial outcome measures have largely relied on established generic clinical or functional endpoints, rather than biomarkers that provide direct and short-term measures of target engagement or mitochondrial function. For example, studies on elamipretide in Barth Syndrome relied on echocardiography; idebenone in LHON stabilized and restored visual acuity; and deoxynucleoside supplemen- tation in TK2-deficiency improved respiratory function, the 6-min walk test and reduced GDF15 levels [7–9]. This review provides an overview of the most relevant current and emerging clinical biomarkers for PMDs, high- lighting their applicability and future utility for longitudinal disease monitoring. We summarize advances in laboratory biomarkers, neuroimaging modalities, and digital endpoints, and discuss possible future directions, such as molecular imaging and integrative multi-omics (Fig. 1, Table 1). As novel therapies advance toward clinical translation, the development and validation of sensitive, specific, and repro- ducible biomarkers will be critical to enable early diagnosis, optimize patient stratification, and reliably evaluate thera- peutic efficacy. Functional assessments: from motor tests to digital monitoring Fatigue is one of the most prevalent and disabling symptoms reported by individuals with PMDs, profoundly affecting daily functioning and quality of life and frequently co-occur- ring with muscle weakness, pain, and reduced mobility [21]. Consequently, the evaluation of fatigue and motor perfor- mance is central to both clinical monitoring and assessment of therapeutic efficacy. Traditional functional assessments, including manual muscle testing, functional motor scales, cardiopulmonary exercise test and timed performance measures, have been adapted from other neuromuscular diseases and validated in PMDs [22–26]. However, these tools are often limited by inter-rater variability, ceiling and floor effects, motivational dependence, and insufficient sensitivity to capture subclini- cal or slowly progressive changes, particularly in heteroge- neous and multisystem disorders [27]. Digital health technologies (DHTs) have emerged as a promising avenue to overcome these limitations. Wearable and non-wearable sensors enable objective and continuous quantification of motor and physiological function, either during structured in-clinic assessments or remotely during everyday activities [28]. Their successful implementation in other neuromuscular diseases, such as Duchenne muscular dystrophy, where digital gait biomarkers (e.g., Stride Veloc- ity 95th Centile) have been qualified by regulatory agencies, provides a compelling precedent for their potential applica- tion in PMDs [29]. Initial investigations in PMDs indicate that wearable- based assessments are both feasible and informative. One of the earliest cross-sectional studies employing actigra- phy demonstrated significantly reduced habitual physical activity in PMD patients, with daily step count inversely correlating with disease burden measured by the Newcas- tle Mitochondrial Disease Adult Scale (NMDAS) [30]. Complementary work using portable gait analysis sys- tems revealed discrete gait abnormalities in adults with m.3243A > G and m.8344A > G variants, even among mildly affected individuals, underscoring the sensitivity of accelerometry-based tools to detect subtle motor impair- ments [31]. In pediatric cohorts, accelerometry similarly distinguished affected children from age- and sex-matched controls across domains of physical activity [32]. More Journal of Neurology (2026) 273:263 Page 3 of 14  263 advanced gait quantification platforms have also shown excellent reproducibility. The GAITRite electronic walk- way has been validated in adults with the m.3243A > G variant, demonstrating high intra-class correlation coef- ficients across walking conditions and significant cor- relations between gait metrics and NMDAS scores [33]. Nonetheless, the responsiveness of wearable-derived out- comes remains an area of active investigation: in a longitu- dinal study assessing bezafibrate treatment, no significant change was observed in accelerometry-based activity lev- els, despite biochemical improvement, illustrating both the promise and challenges of incorporating digital endpoints into PMD trials [34]. Collectively, wearable and portable technologies can provide objective, reproducible, and clinically meaning- ful assessments of motor function and physical activity in PMDs. To advance their integration into clinical research, large-scale longitudinal studies, harmonized protocols, and disease-specific validation frameworks are needed. Future work should aim to develop multimodal digital biomarker profiles—combining movement data with cardiorespiratory, sleep, and fatigue-related metrics—to comprehensively cap- ture the complex and fluctuating functional impairments that characterize PMDs. Laboratory biomarkers: untargeted and targeted approaches Given the predominant bioenergetic impairment and fre- quent muscle involvement in PMDs, traditional CSF and plasma analytes, such as lactate, lactate/pyruvate ratio, amino acids, and creatine kinase (CK), have long served as adjunct biochemical indicators to support diagnosis, along- side muscle biopsy [35–37]. However, extensive clinical experience has demonstrated limited sensitivity and specific- ity for both diagnostic and longitudinal monitoring, as they may also reflect secondary mitochondrial dysfunction in a range of conditions including hypoxia, sepsis, seizures, or intense exercise [38, 39]. Therefore, their diagnostic utility has further declined with the widespread implementation of next-generation sequencing [3]. Omics-based approaches have gained considerable trac- tion over the past decade, deepening mechanistic under- standing of PMD pathophysiology and providing candidates for subsequent targeted evaluation [40]. Early transcrip- tomic and metabolomic work in mitochondrial disease models has shown profound oxidative stress remodeling converging on the induction of fibroblast growth factor 21 (FGF21), a metabolic cytokine secreted by stressed muscle Fig. 1   Non-invasive monitoring in primary mitochondrial disorders (PMDs). Biomarkers and tests for three major non-invasive moni- toring domains in PMDs include functional and performance motor tests, laboratory biomarkers, and imaging biomarkers. Emerging approaches include promising examples of biomarkers that have been tested in model systems or patients with other neurodegenera- tive or neuromuscular conditions with potential future applications for PMDs though further studies are still needed. DWI Diffusion- Weighted Imaging, FDG Fluorodeoxyglucose, FGF21 Fibroblast Growth Factor 21, GDF15 Growth Differentiation Factor 15, MRI Magnetic Resonance Imaging, MRS Magnetic Resonance Spectros- copy, OUES/BSA Oxygen Uptake Efficiency Slope/Body surface Area, PET Positron Emission Tomography, VO2 max Maximal Oxy- gen Consumption Journal of Neurology (2026) 273:263 263   Page 4 of 14 fibers [41]. FGF21 has since been validated as a biomarker for mitochondrial myopathies, especially those associated with impaired mitochondrial DNA maintenance or transla- tion [42–48]. Nonetheless, FGF21 can also be elevated in other conditions, including liver diseases and other meta- bolic neuromuscular disorders. Growth differentiation fac- tor 15 (GDF15), another stress-responsive cytokine, has subsequently emerged as a complementary marker, initially as differentially expressed gene in muscle from patients with TK2-deficiency [49] and later also associated with other PMDs [50, 51]. Although its diagnostic specificity is limited by elevations in cancer, inflammation, and other metabolic and neuromuscular conditions [52], GDF15 has demonstrated promise particularly in TK2 deficiency, where it correlates with disease severity and response to nucleoside supplementation [53, 54]. GDF15 detection in saliva might enable minimally invasive monitoring [55], and integrating measurements with plasma gelsolin was shown to improve diagnostic accuracy [56]. Proteomic profiles from a cohort of mitochondrial patient fibroblasts pointed out the expression of 5 pro- teins (GPX4, ORF4L1, MOXD1, MSRA, and TMED9), as putative novel biomarkers [57]. Hypocitrullinemia was found in MT-ATP6–related conditions, while tyrosine was shown to be elevated in deoxyguanosine kinase deficiency [58–60]. Metabolomic and multi-omics profiling of urine Table 1   Tentative classification of non-invasive biomarkers, according to both their clinical role (diagnostic, monitoring, pharmacodynamic) and their current level of validation “Exploratory” biomarkers refer to early-stage candidates supported primarily by preclinical studies or small clinical cohorts. “Partially vali- dated” biomarkers have shown reproducible associations with disease biology or severity in independent cohorts but lack widespread clinical implementation or regulatory qualification. “Clinically actionable” biomarkers are currently used in clinical diagnostic workflows or patient management ccf-mtDNA Circulating Cell-Free mtDNA; CK Creatine kinase; FGF21 Fibroblast Growth Factor 21; GDF15 Growth Differentiation Factor 15; DWI Diffusion-Weighted Imaging; FDG Fluorodeoxyglucose; FLAIR Fluid-Attenuated Inversion Recovery; MRI Magnetic Resonance Imag- ing; MRS Magnetic Resonance Spectroscopy; MSOT Multi-Spectral Optoacoustic Tomography; NfL Neurofilament Light Chains; OEF Oxygen Extraction Fraction; PET Positron Emission Tomography; SV95C Stride Velocity 95th Centile; TSPO Translocator protein; VO2 max Maximal Oxygen Consumption *In specific diseases (Thymidine Kinase 2 deficiency and Mitochondrial Neurogastrointestinal Encephalopathy) Clinical application in PMDs Exploratory Partially validated Clinically actionable Diagnostic Cardiopulmonary exercise testing with Oxygen Uptake Efficiency Slope/Body Surface Area (OUES/BSA) miRNA; ‘omics signatures Gelsolin; NAD+; CHIT1 Lactate; Pyruvate; Lactate/Pyruvate ratio; CK; amino acids; citrulline; Thd; dUrd; GDF15; FGF21 Raman spectroscopy; Near Infrared Spec- troscopy (NIRS) with vascular occlusion test 31P-MRS; PET-FDG CT and MRI standard sequences (T2/ FLAIR, DWI); 1H-MRS (with lactate peak) Monitoring Smartphone and/or wearable-based moni- toring tools; SV95C; Cardiopulmonary exercise testing with Oxygen Uptake Efficiency Slope/Body Surface Area (OUES/BSA) Gait analysis metrics (e.g., GAITrite) Traditional functional/performance tests (6- and 12-min walk test, 30-s Sit- to-Stand test, Timed Up and Go test, Cardiopulmonary exercise testing with VO2 max) miRNA; cytokines; ‘omics signatures FGF21; NfL; pNfH; ccf-mtDNA Nucleosides (dThd, dUrd); GDF15 MSOT; Near Infrared Spectroscopy (NIRS); Chemical Exchange Saturation Transfer (CEST); Raman spectroscopy TSPO PET; 31P-MRS; 1H-MRS (with lactate peak) Pharmaco- dynamic (response biomarker) SV95C; digital health-based composite endpoints Nucleosides (dThd, dUrd); GDF15* ‘omics signatures PET with new mitochondrial tracers (18F-BCPP-EF, TPSO); Oxygen Extrac- tion Fraction (OEF); Near Infrared Spec- troscopy (NIRS) with vascular occlusion test; MSOT; Raman spectroscopy; MRI with Deuterium Metabolic Imaging 31P-MRS Journal of Neurology (2026) 273:263 Page 5 of 14  263 samples in carriers of the common MELAS m.3243A > G variant [61] or in patients with progressive external oph- thalmoplegia [62], have demonstrated systemic molecular alterations, supporting the sensitivity of urinary profiling in PMDs. Metabolomics have also been applied in plasma from Maternally Inherited Diabetes and Deafness (MIDD) patients [63], and more recently, in Mitochondrial Neu- rogastrointestinal Encephalomyopathy (MNGIE) [64], showing disease-specific patterns of organ involvement and supporting development of integrated multi-omics signatures rather than reliance on single biomarkers for future clinical trials. Circulating signatures of NADH-driven reductive stress have been associated with disease burden in m.3243A > G, including well-established (lactate, alanine, GDF15, α-hydroxybutyrate) as well as new groups of analytes such N-lactoyl-amino acids, β-hydroxy acylcarnitines, and β-hydroxy fatty acids [65]. Reduced NAD⁺ levels have been proposed as a biomarker with direct relationship to the underlying biochemical OXPHOS defect in PMDs, and are being explored as a therapeutic target, for example in trials of niacin supplementation [42, 66]. Among the other metabolites with direct relevance for the underlying genetic or biochemical defect, thymidine/deoxyuridine levels are well-established as diagnostic biomarkers in MNGIE due to thymidine phosphorylase deficiency [67], where they also act as response biomarkers to measure therapeutic efficiency of liver or bone marrow transplantation [68, 69], and novel advanced therapies [70–72]. Whereas the aforementioned biomarkers broadly reflect primary mitochondrial dysfunction and may reflect tissue- specific disease phenotypes, several other biomarkers are more broadly associated with organ damage and acute dis- ease activity in other conditions. Several of these have also been explored in PMDs. Neurofilament light chain (NfL), widely used in neurodegenerative disorders, has shown promise as a marker of neuroaxonal injury in PMDs in serum and in the cerebrospinal fluid, where it correlates with clinical severity in MELAS [73–77]. Elevated cir- culating cell-free mitochondrial DNA (ccf-mtDNA) may act as an indicator of mitochondrial stress during acute MELAS episodes [78, 79]. Chitotriosidase (CHIT1), his- torically used in lysosomal storage diseases, was found to be elevated in PMDs versus other neuromuscular diseases (NMDs) and healthy controls [80]. Finally, recent studies have identified inflammatory activation and upregulation of interferon-stimulated genes in pediatric PMD cohorts [81, 82], and have explored changes in cytokines [83] or circulat- ing microRNAs as possible stable, disease-related indicators in MNGIE [84], MELAS [85], or patients with PMDs and sensorineural hearing loss [86]. At present, most of these remain adjunctive tools for diagnosis, stratification, and trial readiness, pending longitudinal validation. A summary of the main circulating biomarkers that have been tested in patients with mitochondrial disease, is provided in Table 2. Neuroimaging for diagnosis and monitoring Imaging biomarkers provide non-invasive, spatial read- outs of tissue bioenergetics of the brain, heart and skeletal muscle, and potentially detect metabolic derangement and structural injury [87–89]. In PMDs, imaging can support diagnosis, map affected organs and pathways, and monitor progression or treatment response as outcome measures in trials. Classic MRI remains key to diagnose Leigh Syn- drome with symmetric T2/FLAIR hyper-intensities in basal ganglia, brainstem, and thalamus [90]. Other characteristic MRI patterns affecting deep gray matter nuclei, thalami, brainstem, cerebellum and cerebral white matter frequently guide clinicians toward a PMD and their differential diagno- sis [91–94], while CT may sometimes reveal bilateral basal ganglia calcification. Apart from these classical PMDs, it is important to always consider treatable differential diagnoses, that may present with MRI features similar to for example Leigh Syndrome. MRI is also used in the heart to visualize cardiomyopathies and fibrosis [95], which has been reviewed recently elsewhere [96]. Magnetic resonance spectroscopy (MRS) provides non- invasive metabolic information about the biochemical com- position of tissues. Proton MRS (1H-MRS) has long been used to measure metabolic state, with elevated cerebral lac- tate representing one of the first biomarkers of mitochondrial dysfunction, for example in MELAS [97] or Kearns–Sayre syndrome/Pearson syndrome, where lactate is increased in damaged white matter regions; and Leigh syndrome, where lactate peaks are seen in gray and white matter alongside high choline levels [98, 99]. The detection of lactate in the brain is more consistent than in serum, correlates with dis- ease activity in several cohorts and may provide a CNS- based indicator of OXPHOS failure [100–102], but cannot necessarily be distinguished from cellular hypoxia seen in other conditions such as seizures or hypoxic-ischemic inju- ries [103]. Phosphorus MRS (31P-MRS) enables non-invasive assessment of cardiac and skeletal muscle bioenergetics. By quantifying high-energy phosphate metabolites, such as phosphocreatine (PCr), inorganic phosphate (Pi), and ATP, 31P-MRS allows real-time evaluation of oxidative phospho- rylation capacity [104, 105]. Impaired oxidative metabolism typically manifests as prolonged post-exercise PCr recov- ery [106]. Abnormalities, such as altered phosphate ratios and delayed recovery kinetics, have been demonstrated in patients with mitochondrial myopathies due to mtDNA variants although diagnostic sensitivity is limited [107]. Exercise-based 31P-MRS provides a quantitative in vivo Journal of Neurology (2026) 273:263 263   Page 6 of 14 Ta bl e  2   L ist o f t he m ai n ci rc ul at in g an al yt es /b io m ar ke rs te ste d in P rim ar y M ito ch on dr ia l D is ea se s ( PM D s) O rg an A na ly te /B io m ar ke r M ito ch on dr ia l c on di tio n( s) Fi nd in gs a nd a pp lic at io ns /li m its Re fe re nc es B lo od La ct at e* Py ru va te *; L ac ta te /p yr uv at e ra tio Te ste d in d iff er en t P M D s El ev at ed in d iff er en t P M D s b ut o ve ra ll lim ite d di ag no sti c se ns iti vi ty /s pe ci fic ity . P yr uv at e an d la ct at e/ py ru va te c ou ld b e us ef ul a s a c om pl e- m en ta ry b io m ar ke r i n Py ru va te D eh yd ro ge na se an d Py ru va te C ar bo xy la se d efi ci en ci es Pa re de s- Fu en te s e t a l. [3 9] C re at in e ki na se (C K ) Te ste d in d iff er en t P M D s El ev at ed in m yo pa th ic fo rm s, bu t l im ite d di ag - no sti c se ns iti vi ty /s pe ci fic ity , a pa rt fro m sp ec ifi c co nt ex ts (T K 2 de fic ie nc y) D eb ra y et  a l. [3 6] A m in o ac id s ( al an in e* , a sp ar ta te ) Te ste d in d iff er en t P M D s O cc as io na lly e le va te d, e sp ec ia lly in e ar ly o ns et fo rm s. O ve ra ll lo w se ns iti vi ty a nd sp ec ifi ci ty Pa re de s- Fu en te s e t a l. [3 9] C itr ul lin e A TP 6- re la te d di so rd er s Re du ce d in A TP 6- re la te d di so rd er s Li e t a l., [5 8] ; C ar li et  a l. [5 9] Th ym id in e (T hd ), de ox yu rid in e (d U rd ) M N G IE Th d an d dU rd le ve ls a re u se fu l i n M N G IE pa tie nt s H ira no e t a l. [6 7] FG F2 1 Te ste d in d iff er en t P M D s, m os tly m ito ch on dr ia l m yo pa th ie s El ev at ed in m tD N A m ai nt en an ce /tr an sl at io n de fe ct s. Va lid at ed a s b io m ar ke r o f m ito ch on - dr ia l m yo pa th y bu t l ow d ia gn os tic sp ec ifi ci ty Fo rs str om e t a l. [4 2] G D F1 5 Te ste d in d iff er en t P M D s, pa rti cu la rly T K 2 de fic ie nc y El ev at ed in P M D s b ut lo w sp ec ifi ci ty ; c or re - la te s w ith d is ea se se ve rit y an d de cr ea se s w ith th er ap y in T K 2 de fic ie nc y Ts yg nk ov a et  a l. [5 2] , B er - m ej o- G ue rr er o et  a l. [5 3] G el so lin Te ste d in a c oh or t o f d iff er en t P M D s, in cl ud in g C PE O , M EL A S Re du ce d. E va lu at ed in a ss oc ia tio n w ith G D F1 5, ca n im pr ov e di ag no sti c ac cu ra cy Pe na s e t a l.  [5 6] N A D H -r el at ed re du ct iv e str es s m ar ke rs a nd N A D ⁺ Te ste d in a c oh or t o f d iff er en t P M D s, pa rti cu la rly w ith m ito ch on dr ia l m yo pa th y C irc ul at in g si gn at ur es c or re la te w ith d is ea se bu rd en ; R ed uc ed N A D ⁺, ex pl or ed a s b io m ar ke r an d ta rg et in n ia ci n tri al Pi rin en e t a l., [6 6] N eu ro fil am en t l ig ht c ha in (N fL )* Te ste d in d iff er en t d is or de rs , p ar tic ul ar ly in M EL A S Se ru m N fL e le va te d in n eu ro ax on al in ju ry ; bi om ar ke r p ot en tia l. In C SF c or re la te d w ith di se as e se ve rit y Va rh au g et  a l. [7 4] ; So fo u et  a l. [7 3] Ph os ph or yl at ed n eu ro fil am en t h ea vy c ha in (p N fH ) LH O N Su gg es te d as m ar ke r o f a xo na l d eg en er at io n af te r vi su al lo ss fu nc tio n G uy e t a l. [7 5] In te rfe ro n- sti m ul at ed g en es Te ste d in a c oh or t o f p ed ia tri c PM D s U pr eg ul at ed ; m ay re fle ct im m un e dy sr eg ul at io n K es ha va n et  a l. [8 1] C yt ok in es PM D s M ay re fle ct in fla m m at or y ac tiv at io n, c yt ok in e al te ra tio ns . N ee d fu rth er v al id at io n Pr im ia no e t a l. [8 3] C hi to tri os id as e (C H IT 1) PM D s El ev at ed ; c an di da te b io m ar ke r f or P M D s Fo er ste r e t a l. [8 0] A ra ch id on ic a ci d m et ab ol is m , b ile a ci d bi os yn - th es is M N G IE Si gn ifi ca nt u pr eg ul at io n; d is ea se -s pe ci fic m et a- bo lic p at te rn , p ot en tia lly sp ec ifi c fo r M N G IE B ax e t a l. [6 4] m iR N A M N G IE m iR -1 92 -5 p, m iR -1 93 a- 5p , m iR -1 94 -5 p, m iR - 21 5- 5p a nd m iR -3 4a -5 p in p la sm a; m iR -1 92 -5 p, m iR -1 94 -5 p, m iR -3 4a -5 p: p ot en tia lly sp ec ifi c fo r M N G IE M en ci as e t a l. [8 4] ; PM D p at ie nt s w ith se ns or in eu ra l h ea rin g lo ss m iR -3 4a , m iR -2 9b M ar oz zo e t a l. [8 6] Journal of Neurology (2026) 273:263 Page 7 of 14  263 measure of skeletal muscle mitochondrial oxidative capac- ity and has been explored as pharmacodynamic biomarker in preliminary interventional studies [108]. Despite this poten- tial, broader clinical implementation remains limited by the need for specialized hardware, standardized protocols, and expertise in spectral analysis [109, 110] and it is not widely used in longitudinal studies. Diffusion MRI and arterial spin labeling (ASL) con- tribute complementary information about microstructure, cytotoxic stress and perfusion, particularly relevant for stroke-like lesions in MELAS, but insufficiently specific for PMDs [111]. Diffusion-weighted MRI (DWI) reveals characteristics abnormalities in MELAS, reflecting meta- bolic edema rather than vascular ischemia [112]. In Leigh syndrome and Kearns–Sayre syndrome, DWI hyperintensi- ties with abnormal ADC values have also been noted in the brainstem [113]. Molecular PET imaging approaches enhance biomarker sensitivity and specificity, with a potential for directly measuring disease-relevant aspects of mitochondrial biol- ogy: brain translocator protein (TSPO) PET scans are typi- cally used to measure neuro-inflammation and microglial activation in a wide range of neurologic conditions [114, 115]. Interestingly, TSPO is an outer mitochondrial mem- brane protein, and could therefore be affected by impaired mitochondrial function or mass. TSPO PET with the 11C- PK11195 ligand appears to show signal changes that cor- relate with disease severity, and may detect abnormality in absence of MRI changes [116]. Multiple case reports on patients with MELAS indicate that FDG-PET reveals regional glucose hypo- or hyper-metabolism [117, 118]. Challenges persist, in particular for PET, with substantial costs associated with tracer production and imaging. Genetic variants can influence tracer binding [119], there are partial volume effects [120], and mitochondrial dysfunction can vary across tissue types, which together complicate signal interpretation and routine clinical application. Neverthe- less, as highlighted in recent reviews [87, 103, 121], many of these imaging approaches may ultimately support ear- lier diagnosis, stratify disease subtypes, or track treatment response. A summary of the main imaging techniques used in patients with PMDs is presented in Table 3. Emerging non‑invasive biomarkers and imaging approaches with promise in PMDs Beyond traditional serum and neuroimaging markers, several innovative methodologies are emerging with the potential to provide sensitive, real-time readouts of mitochondrial dys- function, metabolic capacity, and tissue-level bioenergetics. These approaches may offer a viable path toward objective Ta bl e  2   (c on tin ue d) O rg an A na ly te /B io m ar ke r M ito ch on dr ia l c on di tio n( s) Fi nd in gs a nd a pp lic at io ns /li m its Re fe re nc es C irc ul at in g ce ll- fr ee m ito ch on dr ia l D N A (c cf - m tD N A )* M EL A S, M ER R F El ev at ed d ur in g ac ut e ep is od es ; i nd ic at es m ito - ch on dr ia l s tre ss M ar es ca e t a l. [7 0] Sa liv a G D F1 5 Te ste d in d iff er en t P M D s D et ec ta bl e an d dy na m ic ; c or re la te s w ith st re ss , ut ili ty m ig ht st an d in it s m in im al ly in va si ve ne ss H ua ng e t a l. [5 5] O rg an ic a ci ds a nd a cy lc ar ni tin es (w hi ch c ou ld b e us ef ul in th e di ffe re nt ia l d ia gn os is w ith o th er m et ab ol ic c on di tio ns ) a re n ot m en tio ne d in th e ta bl e CK C re at in e ki na se , C PE O C hr on ic P ro gr es si ve E xt er na l O ph th al m op le gi a, F G F2 1 Fi br ob la st G ro w th F ac to r 2 1, G D F1 5 G ro w th D iff er en tia tio n Fa ct or 1 5, L H O N L eb er H er ed ita ry O pt ic N eu - ro pa th y, M EL AS M ito ch on dr ia l E nc ep ha lo m yo pa th y, L ac tic A ci do si s, an d St ro ke -li ke E pi so de s, M ER RF M yo cl on ic E pi le ps y w ith R ag ge d- Re d Fi be rs , M N G IE M ito ch on dr ia l N eu ro ga str oi nt es - tin al E nc ep ha lo pa th y, T K 2 Th ym id in e K in as e 2 *T es te d al so in C er eb ro sp in al fl ui d (C SF ) Journal of Neurology (2026) 273:263 263   Page 8 of 14 stratification and functional endpoints for upcoming clini- cal trials, although many are still constrained to preclinical models or other diseases and have not been tested in patients with PMDs. Among laboratory biomarkers, circulating axonal cytoskeletal proteins such as peripherin have emerged as promising biomarkers of peripheral nervous damage [122, 123], paving the way for PMDs presenting with peripheral neuropathy. Ongoing studies, for example in asymptomatic mtDNA variant carriers for LHON, aim to identify signa- tures predictive of phenotypic conversion [124]. A wide range of imaging techniques is also emerging, with promise as novel biomarkers in PMDs. Near-Infrared Spectroscopy (NIRS) with vascular occlusion testing pro- vides a non-invasive method to assess muscle oxygenation, hemodynamics, and oxidative metabolism. In pediatric Table 3   List of the main central nervous system imaging techniques used in patients with Primary Mitochondrial Diseases (PMDs) CPEO Chronic Progressive External Ophthalmoplegia, DWI Diffusion-Weighted Imaging, FDG Fluorodeoxyglucose, FLAIR Fluid-Attenuated Inversion Recovery, MELAS Mitochondrial Encephalomyopathy, Lactic Acidosis, and Stroke-like Episodes, MRI Magnetic Resonance Imaging, MRS Magnetic Resonance Spectroscopy, PET Positron Emission Tomography, POLG DNA Polymerase, Gamma, TSPO Translocator protein Organ Technique Sequence Mitochondrial condition Observation References Brain MRI T2/FLAIR Wide range of PMDs Symmetric hyper-intensities in white and deep gray matter; Cerebel- lar atrophy, for example in POLG mutations; Prominent leukoenceph- alopathy, for example in MNGIE; Symmetric hyper-intensity of basal ganglia, for example in Leigh Syn- drome; Focal gray matter lesions, for example in stroke-like episodes caused by MELAS, typically in temporal, parietal, and occipital lobes; Developmental abnormali- ties, for example in Pyruvate Dehy- drogenase deficiency, etc Extensively reviewed in Saneto et al. [88] DWI MELAS (stroke-like episodes) Stroke-like lesions typically demon- strate normal or increased apparent diffusion coefficient (ADC) values, indicative of vasogenic edema, rather than restricted diffusion in ischemic stroke Oppenheim et al. [112] Arterial Spin Labeling (ASL) MELAS (stroke-like episodes) Peripheral hyper-perfusion observed in stroke-like lesions assists in differentiating from acute ischemic stroke Li et al. [111] MRS 1H-MRS MELAS An elevated lactate peak and reduced N-acetyl-aspartate (NAA) constitute a metabolic profile indicative of mitochondrial dysfunction Abe et al. [100] Kearns–Sayre syndrome/Pearson syndrome An increase in lactate within the affected white matter Kapeller et al. [99] Leigh syndrome Elevated lactate is detected in both affected gray and white matter, often accompanied by increased choline levels Sijens et al. [98] 31P-MRS Mitochondrial myopathies Prolonged post-exercise Phosphocre- atine (PCr) recovery Jeppesen et al. [107] PET TSPO PET Tested in a cohort of different PMDs Regional alterations in TSPO ([11C] PK11195) binding may correspond to phenotypes or clinical severity van den Ameele et al. [116] FDG-PET MELAS (stroke-like episodes) Regional hypo-metabolism or hyper- metabolism, dependent on disease stage or presence of seizures Liu et al. [117] Eye MRI CPEO Marked atrophy of the extraocular muscles Yu-Wai-Man et al. [89] Journal of Neurology (2026) 273:263 Page 9 of 14  263 populations with PMDs and neuro-genetic disorders, NIRS- derived parameters have shown sensitivity to impaired OXPHOS in small clinical cohorts of patients with mito- chondrial or other neuromuscular disorders, correlating with clinical severity and functional limitations [125]. The portability and tolerability of NIRS could make it attractive for longitudinal clinical monitoring, including in pediatric populations [126]. Other imaging techniques seek to enhance spatial speci- ficity and provide mechanistic insights into metabolic altera- tions. Chemical Exchange Saturation Transfer (CEST), has demonstrated potential in an Ndufs4 knockout mouse model for mapping intracellular lactate with higher resolution than MRS [127], but has not yet been systematically evaluated in patients. Deuterium Metabolic Imaging (DMI) is an MRI- based technology that provides a dynamic map of glucose utilization and downstream metabolites such as glutamate/ glutamine or lactate by tracing deuterated glucose, offer- ing insight into real-time tissue bioenergetics. Though only recently applied in neurological disorders, such as Alz- heimer’s disease and glioblastoma [128, 129], DMI holds promise for PMDs by enabling direct assessment of gly- colytic flux and OXPHOS that could serve as quantitative biomarkers in trials targeting cerebral energy metabolism [129, 130]. However, studies in PMDs are currently lacking. Another emerging MRI-based technique is Muscle Oxygen Extraction Fraction (OEF), which measures the efficiency of oxygen utilization in muscle and might detect changes in early disease stages [132, 133] although these tracers are currently being evaluated primarily in research settings and have not yet been systematically applied in PMD cohorts. Mitochondrial-targeted tracers for PET imaging, such as 18F-BCPP-EF, are in development to enable in vivo quanti- fication of complex I activity [131]. Multi-Spectral Optoacoustic Tomography (MSOT) enables high-resolution real-time visualization of tissue structure and function based on photoacoustic contrast. MSOT allows non-invasive interrogation of physiological parameters linked to mitochondrial function, such as tissue oxygen extraction and microvascular dynamics. Although experience in PMDs remains limited, studies in related neuromuscular diseases as Duchenne Muscular Dystrophy [115], Spinal Muscular Atrophy [134] or Pompe Disease [135], support its potential as a future PMD monitoring or response biomarker. Preclinical studies have demonstrated its feasibility in mouse models for quantification of optoa- coustic signatures from metabolically active organs with excellent temporal and spatial resolution [136, 137], ulti- mately appearing a potential bedside-compatible functional imaging biomarker for tissue-level bioenergetics [138], as demonstrated in a recent exploratory study in patients with m.3243 A>G pathogenic variants [REF]. Finally, Raman spectroscopy of tissue samples or in vivo with fiber-optics may capture protein conformation, meta- bolic signatures, and structural alterations that precede or accompany muscle pathology. Fiber-optic Raman spec- troscopy was shown to differentiate between healthy and diseased muscle in mouse, offering a minimally invasive alternative to standard histopathology [139], and might even be combined with muscle electrophysiology [140]. Overall, while these emerging imaging approaches pro- vide exciting opportunities to interrogate mitochondrial biol- ogy in vivo, most remain at an early stage of development. Systematic validation in well-characterized PMD cohorts, standardization of acquisition protocols, and demonstration of reproducibility across centers will be essential before these techniques can be considered robust biomarkers for clinical settings and trials. Conclusion In clinical practice, the management of PMDs often remains challenging, in particular due to marked phenotypic and genetic heterogeneity and fluctuating disease trajectories. To date, no single biomarker has proven sufficiently sensi- tive, specific, and broadly applicable to meet all the diverse diagnostic and longitudinal needs of these heterogeneous disorders. Traditional imaging, biochemical and func- tional measures remain essential in routine clinical care, yet their limitations highlight the need for complementary approaches. Recent advances in non-invasive laboratory bio- markers, digital and wearable technologies, and metabolic and molecular imaging, although still requiring validation in many cases, provide clinicians with an expanding toolbox capable of capturing complementary dimensions of mito- chondrial dysfunction, tissue injury, and functional impair- ment. At this stage, rather than replacing traditional meas- ures, these biomarkers should be viewed as adjunct options that can refine diagnostic confidence and support clinical monitoring. They are likely to play an increasingly important role in patient stratification and in providing objective read- outs of disease activity and treatment response, particularly in clinical trials. Looking ahead, validation of partially established bio- markers remains a priority, such as NfL in serum, brain, or muscle MRS imaging, or PET tracers. These should be validated in larger cohorts, together with harmonization of image acquisition protocols, and improved reproduc- ibility of digital health technology endpoints. In parallel, exploratory technologies, including Raman spectroscopy, MSOT, DMI, and novel PET tracers should be progres- sively translated from preclinical studies into clinical eval- uation. Over the medium term, prospective longitudinal Journal of Neurology (2026) 273:263 263   Page 10 of 14 studies will be essential to determine biomarker respon- siveness, predictive value, and utility as pharmacodynamic endpoints in clinical trials. This should occur alongside efforts to develop disorder- and treatment-specific bio- marker profiles, leveraging deep phenotyping and machine learning to inform patient stratification and therapeutic decision-making. Finally, long-term progress will likely rely on integration of both generic and disorder-specific biomarkers into com- pound measures, for which large-scale cohorts, deep pheno- typing and machine learning approaches may prove useful. Given the distinct pathophysiology and therapeutic strate- gies of individual syndromic groups, the pursuit of universal biomarkers across all PMDs is likely to remain challenging. Biomarker development should therefore align closely with disease pathogenesis and the key metabolic and biochemical pathways amenable to therapy development, or aim to meas- ure clinical phenotypes with direct relevance for patients and their quality of life [15]. As targeted treatments progress toward clinical application, establishing and harmonizing such multidimensional, disorder- and treatment-specific bio- marker profiles will be essential to enable rigorous clinical trials and to advance clinical patient care. Acknowledgements  We thank all MitoCamb team members for helpful discussions and insights. Funding  This work was supported by a Wellcome Clinical Research Career Development Fellowship (219615/Z/19/Z), a Medical Research Council (MRC) award (MC_PC_21046) to establish a National Mouse Genetics Network Cluster in Mitochondrial Diseases (MitoCluster) and core funding from the UKRI MRC to the MRC Mitochondrial Biology Unit (MC_UU_00028/8) to JvdA. RH is supported by the Medical Research Council (UK) (MR/V009346/1), the Hereditary Neuropathy Foundation, AFM-Telethon, Ataxia UK, Action for AT, Muscular Dys- trophy UK, and the UKRI/Horizon Europe Guarantee MSCA Doctoral Network Programme (Project 101120256: MMM). She is also sup- ported by an MRC strategic award to establish an International Centre for Genomic Medicine in Neuromuscular Diseases (ICGNMD) MR/ S005021/1. RH and JvdA are further supported by a Wellcome Discov- ery Award (226653/Z/22/Z), the Rosetrees Trust (PGL23/100048), and the LifeArc Centre to Treat Mitochondrial Diseases (LAC-TreatMito) under grant no. 10748. LifeArc is a charity registered in England and Wales under no. 1015243 and in Scotland under no. SC037861. This research was supported by the NIHR Cambridge Biomedical Research Centre (NIHR203312). The views expressed are those of the authors and not necessarily those of the NIHR or the Department of Health and Social Care. Data availability  Not applicable. Declarations  Conflicts of interest  The authors declare no competing interests. Open Access  This article is licensed under a Creative Commons Attri- bution 4.0 International License, which permits use, sharing, adapta- tion, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. References 1. Gorman GS, Schaefer AM, Ng Y et al (2015) Prevalence of nuclear and mitochondrial DNA mutations related to adult mito- chondrial disease. Ann Neurol 77:753–759. https://​doi.​org/​10.​ 1002/​ana.​24362 2. Ng YS, Gorman GS, Mcfarland R et al (2021) Mitochondrial disease in adults: recent advances and future promise. Lancet Neurol 20(7):573–84 3. Schon KR, Ratnaike T, van den Ameele J et al (2020) Mito- chondrial diseases: a diagnostic revolution. Trends Genet 36:702–717 4. Suomalainen A, Battersby BJ (2018) Mitochondrial diseases: the contribution of organelle stress responses to pathology. Nat Rev Mol Cell Biol 19:77–92 5. Stewart JB, Chinnery PF (2021) Extreme heterogeneity of human mitochondrial DNA from organelles to populations. Nat Rev Genet 22:106–118 6. Ball M, van Bergen NJ, Compton AG et al (2025) Therapies for mitochondrial disease: past, present, and future. J Inherit Metab Dis 48:e70065 7. Yu-Wai-Man P, Carelli V, Newman NJ et al (2024) Therapeutic benefit of idebenone in patients with Leber hereditary optic neu- ropathy: the LEROS nonrandomized controlled trial. Cell Rep Med. https://​doi.​org/​10.​1016/j.​xcrm.​2024.​101437 8. Thompson WR, Manuel R, Abbruscato A et al (2024) Long-term efficacy and safety of elamipretide in patients with Barth syn- drome: 168-week open-label extension results of TAZPOWER. Genet Med. https://​doi.​org/​10.​1016/j.​gim.​2024.​101138 9. Domínguez-González C, Chiang C, Colson AO et al (2025) Pyrimidine nucleos(t)ide therapy in patients with thymidine kinase 2 deficiency a multicenter retrospective chart review study. Neurology. https://​doi.​org/​10.​1212/​WNL.​00000​00000​ 213908 10. U.S. Food and Drug Administration. FDA approves 1st drug for thymidine kinase 2 deficiency, a very raremitochondrial disease [Internet]. Silver Spring (MD): FDA; 2025 Nov 3 [cited 2026 Apr 7]. Available from: https://​www.​fda.​gov/​drugs/​news-​events-​ human-​drugs/​fda-​appro​ves-​1st-​drug-​thymi​dine-​kinase-​2-​defic​ iency-​very-​rare-​mitoc​hondr​ial-​disea​se 11. Karaa A, Klopstock T (2023) Clinical trials in mitochondrial dis- eases. Handbook of clinical neurology. Elsevier B.V, Amsterdam, pp 229–250 12. Newman NJ, Biousse V, Yu-Wai-Man P et  al (2025) Meta- analysis of treatment outcomes for patients with m.11778G>A MT-ND4 Leber hereditary optic neuropathy. Surv Ophthalmol 70:283–295 13. Leslie M (2025) First approved drug for mitochondrial disease could pave way for more treatments. Science (1979) 390:114–115 14. Mancuso M, Lopriore P, Semmler L, Kornblum C (2025) 280th ENMC International Workshop: The ERN EURO-NMD mitochondrial diseases working group; diagnostic criteria and outcome measures in primary mitochondrial myopathies. http://creativecommons.org/licenses/by/4.0/ https://doi.org/10.1002/ana.24362 https://doi.org/10.1002/ana.24362 https://doi.org/10.1016/j.xcrm.2024.101437 https://doi.org/10.1016/j.gim.2024.101138 https://doi.org/10.1212/WNL.0000000000213908 https://doi.org/10.1212/WNL.0000000000213908 https://www.fda.gov/drugs/news-events-human-drugs/fda-approves-1st-drug-thymidine-kinase-2-deficiency-very-rare-mitochondrial-disease https://www.fda.gov/drugs/news-events-human-drugs/fda-approves-1st-drug-thymidine-kinase-2-deficiency-very-rare-mitochondrial-disease https://www.fda.gov/drugs/news-events-human-drugs/fda-approves-1st-drug-thymidine-kinase-2-deficiency-very-rare-mitochondrial-disease Journal of Neurology (2026) 273:263 Page 11 of 14  263 Hoofddorp, the Netherlands, 22–24 November 2024. Neuromus- cul Disord 50:105340 15. Thomas RH, Hunter A, Butterworth L et al (2022) Research pri- orities for mitochondrial disorders: current landscape and patient and professional views. J Inherit Metab Dis 45:796–803. https://​ doi.​org/​10.​1002/​jimd.​12521 16. Atkinson AJ, Colburn WA, DeGruttola VG et al (2001) Biomark- ers and surrogate endpoints: preferred definitions and conceptual framework. Clin Pharmacol Ther 69:89–95 17. Steele HE, Horvath R, Lyon JJ, Chinnery PF (2017) Monitor- ing clinical progression with mitochondrial disease biomarkers. Brain 140:2530–2540. https://​doi.​org/​10.​1093/​brain/​awx168 18. Hubens WHG, Vallbona-Garcia A, de Coo IFM et al (2022) Blood biomarkers for assessment of mitochondrial dysfunction: an expert review. Mitochondrion 62:187–204 19. FDA-NIH Biomarker Working Group. BEST (Biomarkers, EndpointS, and other Tools) Resource [Internet]. SilverSpring (MD): Food and Drug Administration (US); 2016. Co-published by National Institutes of Health (US), Bethesda(MD). Available from: https://​www.​ncbi.​nlm.​nih.​gov/​books/​NBK32​6791/​toc/ 20. Pizzamiglio C, Vernon HJ, Hanna MG, Pitceathly RDS (2022) Designing clinical trials for rare diseases: unique challenges and opportunities. Nat Rev Methods Primers 2:13 21. Karaa A, Johnson N, Clarkson I et al (2023) Characterization of fatigue in primary mitochondrial myopathies: findings from a qualitative interview study. Neurol Clin Pract. https://​doi.​org/​ 10.​1212/​CPJ.​00000​00000​200229 22. Flickinger J, Fan J, Wellik A et al (2021) Development of a mito- chondrial myopathy-composite assessment tool. JCSM Clin Rep 6(4):109–27 23. Jeppesen TD, Madsen KL, Poulsen NS et al (2021) Exercise test- ing, physical training and fatigue in patients with mitochondrial myopathy related to mtDNA mutations. J Clin Med. https://​doi.​ org/​10.​3390/​jcm10​081796 24. Barroso de Queiroz Davoli G, Bartels B, Mattiello-Sverzut AC, Takken T (2021) Cardiopulmonary exercise testing in neuromus- cular disease: a systematic review. Expert Rev Cardiovasc Ther 19:975–991 25. Montano V, Lopriore P, Gruosso F et al (2022) Primary mito- chondrial myopathy: 12-month follow-up results of an Ital- ian cohort. J Neurol 269:6555–6565. https://​doi.​org/​10.​1007/​ s00415-​022-​11324-3 26. Løkken N, Khawajazada T, Storgaard JH et al (2021) No effect of resveratrol in patients with mitochondrial myopathy: a cross-over randomized controlled trial. J Inherit Metab Dis 44:1186–1198. https://​doi.​org/​10.​1002/​jimd.​12393 27. Bergs PMJ, Maas DM, Janssen MCH, Groothuis JT (2022) Feasible and clinical relevant outcome measures for adults with mitochondrial disease. Mol Genet Metab 135:102–108. https://​ doi.​org/​10.​1016/j.​ymgme.​2021.​12.​010 28. Bortolani S, Brusa C, Rolle E et al (2022) Technology outcome measures in neuromuscular disorders: a systematic review. Eur J Neurol 29:1266–1278 29. Servais L, Yen K, Guridi M et al (2022) Stride velocity 95th centile: insights into gaining regulatory qualification of the first wearable-derived digital endpoint for use in Duchenne Muscular Dystrophy trials. J Neuromuscul Dis 9:335–346 30. Apabhai S, Gorman GS, Sutton L et al (2011) Habitual physical activity in mitochondrial disease. PLoS ONE. https://​doi.​org/​10.​ 1371/​journ​al.​pone.​00222​94 31. Galna B, Newman J, Jakovljevic DG et  al (2014) Discrete gait characteristics are associated with m.3243A>G and m.8344A>G variants of mitochondrial disease and its pathologi- cal consequences. J Neurol 261:73–82. https://​doi.​org/​10.​1007/​ s00415-​013-​7129-2 32. Koene S, Stolwijk NM, Ramakers R et al (2018) Quantification of gait in children with mitochondrial disease. J Inherit Metab Dis 41:731–740. https://​doi.​org/​10.​1007/​s10545-​018-​0148-5 33. Ramakers R, Koene S, Groothuis JT et al (2017) Quantification of gait in mitochondrial m.3243A > G patients: a validation study. Orphanet J Rare Dis. https://​doi.​org/​10.​1186/​s13023-​017-​0644-y 34. Steele H, Gomez‐Duran A, Pyle A et al (2020) Metabolic effects of bezafibrate in mitochondrial disease. EMBO Mol Med. https://​ doi.​org/​10.​15252/​emmm.​20191​1589 35. Haas RH, Parikh S, Falk MJ et al (2008) The in-depth evaluation of suspected mitochondrial disease. Mol Genet Metab 94:16–37. https://​doi.​org/​10.​1016/j.​ymgme.​2007.​11.​018 36. Debray FG, Mitchell GA, Allard P et al (2007) Diagnostic accu- racy of blood lactate-to-pyruvate molar ratio in the differential diagnosis of congenital lactic acidosis. Clin Chem 53:916–921. https://​doi.​org/​10.​1373/​clinc​hem.​2006.​081166 37. Parikh S, Goldstein A, Koenig MK et  al (2015) Diagnosis and management of mitochondrial disease: a consensus state- ment from the Mitochondrial Medicine Society. Genet Med 17:689–701 38. Shayota BJ (2024) Biomarkers of mitochondrial disorders. Neu- rotherapeutics 21:e00325 39. Paredes-Fuentes AJ, Oliva C, Urreizti R et al (2023) Laboratory testing for mitochondrial diseases: biomarkers for diagnosis and follow-up. Crit Rev Clin Lab Sci 60:270–289 40. Gervasoni J, Primiano A, Cicchinelli M et al (2024) Mitochon- drial biomarkers in the omics era: a clinical-pathophysiological perspective. Int J Mol Sci. https://​doi.​org/​10.​3390/​ijms2​50948​55 41. Tyynismaa H, Carroll CJ, Raimundo N et al (2010) Mitochon- drial myopathy induces a starvation-like response. Hum Mol Genet 19:3948–3958. https://​doi.​org/​10.​1093/​hmg/​ddq310 42. Forsström S, Jackson CB, Carroll CJ et al (2019) Fibroblast Growth Factor 21 drives dynamics of local and systemic stress responses in mitochondrial myopathy with mtDNA deletions. Cell Metab 30:1040-1054.e7. https://​doi.​org/​10.​1016/j.​cmet.​ 2019.​08.​019 43. Lehtonen JM, Forsström S, Bottani E et al (2016) FGF21 is a bio- marker for mitochondrial translation and mtDNA maintenance disorders. Neurology 87:2290–2299. https://​doi.​org/​10.​1212/​ WNL.​00000​00000​003374 44. Buzkova J, Nikkanen J, Ahola S et al (2018) Metabolomes of mitochondrial diseases and inclusion body myositis patients: treatment targets and biomarkers. EMBO Mol Med. https://​doi.​ org/​10.​15252/​emmm.​20180​9091 45. Sevastianova K, Yki-Järvinen H, Smeitink J, et al (2011) FGF-21 as a biomarker for muscle-manifesting mitochondrial respiratory chain defi ciencies: a diagnostic study. www.​thela​ncet.​com/​neuro​ logy https://​doi.​org/​10.​1016/​S1474 46. Khan NA, Nikkanen J, Yatsuga S et al (2017) mTORC1 regulates mitochondrial integrated stress response and mitochondrial myo- pathy progression. Cell Metab 26:419-428.e5. https://​doi.​org/​10.​ 1016/j.​cmet.​2017.​07.​007 47. Riley LG, Nafisinia M, Menezes MJ et al (2022) FGF21 out- performs GDF15 as a diagnostic biomarker of mitochondrial disease in children. Mol Genet Metab 135:63–71. https://​doi.​ org/​10.​1016/j.​ymgme.​2021.​12.​001 48. Tekin Neijmann S, Gunes D, Karaca M et al (2025) Evaluation of serum FGF21 levels in patients with mitochondrial aminoacyl- tRNA synthetase deficiency. Int J Mol Sci. https://​doi.​org/​10.​ 3390/​ijms2​61995​25 49. Kalko SG, Paco S, Jou C et al (2014) Transcriptomic profiling of TK2 deficient human skeletal muscle suggests a role for the p53 signalling pathway and identifies growth and differentiation factor-15 as a potential novel biomarker for mitochondrial myo- pathies. BMC Genom. https://​doi.​org/​10.​1186/​1471-​2164-​15-​91 https://doi.org/10.1002/jimd.12521 https://doi.org/10.1002/jimd.12521 https://doi.org/10.1093/brain/awx168 https://www.ncbi.nlm.nih.gov/books/NBK326791/toc/ https://doi.org/10.1212/CPJ.0000000000200229 https://doi.org/10.1212/CPJ.0000000000200229 https://doi.org/10.3390/jcm10081796 https://doi.org/10.3390/jcm10081796 https://doi.org/10.1007/s00415-022-11324-3 https://doi.org/10.1007/s00415-022-11324-3 https://doi.org/10.1002/jimd.12393 https://doi.org/10.1016/j.ymgme.2021.12.010 https://doi.org/10.1016/j.ymgme.2021.12.010 https://doi.org/10.1371/journal.pone.0022294 https://doi.org/10.1371/journal.pone.0022294 https://doi.org/10.1007/s00415-013-7129-2 https://doi.org/10.1007/s00415-013-7129-2 https://doi.org/10.1007/s10545-018-0148-5 https://doi.org/10.1186/s13023-017-0644-y https://doi.org/10.15252/emmm.201911589 https://doi.org/10.15252/emmm.201911589 https://doi.org/10.1016/j.ymgme.2007.11.018 https://doi.org/10.1373/clinchem.2006.081166 https://doi.org/10.3390/ijms25094855 https://doi.org/10.1093/hmg/ddq310 https://doi.org/10.1016/j.cmet.2019.08.019 https://doi.org/10.1016/j.cmet.2019.08.019 https://doi.org/10.1212/WNL.0000000000003374 https://doi.org/10.1212/WNL.0000000000003374 https://doi.org/10.15252/emmm.201809091 https://doi.org/10.15252/emmm.201809091 http://www.thelancet.com/neurology http://www.thelancet.com/neurology https://doi.org/10.1016/S1474 https://doi.org/10.1016/j.cmet.2017.07.007 https://doi.org/10.1016/j.cmet.2017.07.007 https://doi.org/10.1016/j.ymgme.2021.12.001 https://doi.org/10.1016/j.ymgme.2021.12.001 https://doi.org/10.3390/ijms26199525 https://doi.org/10.3390/ijms26199525 https://doi.org/10.1186/1471-2164-15-91 Journal of Neurology (2026) 273:263 263   Page 12 of 14 50. Steele HE, Horvath R, Taylor RW (2016) The swinging pen- dulum of biomarkers in mitochondrial disease. Neurology 87:2286–2287 51. Yatsuga S, Fujita Y, Ishii A et al (2015) Growth differentiation factor 15 as a useful biomarker for mitochondrial disorders. Ann Neurol 78:814–823. https://​doi.​org/​10.​1002/​ana.​24506 52. Tsygankova PG, Itkis YS, Krylova TD et al (2019) Plasma FGF- 21 and GDF-15 are elevated in different inherited metabolic dis- eases and are not diagnostic for mitochondrial disorders. J Inherit Metab Dis 42:918–933. https://​doi.​org/​10.​1002/​jimd.​12142 53. Bermejo-Guerrero L, Hernández-Voth A, Serrano-Lorenzo P et al (2024) Remarkable clinical improvement with oral nucleo- side treatment in a patient with adult-onset TK2 deficiency: a case report. Mitochondrion. https://​doi.​org/​10.​1016/j.​mito.​2024.​ 101879 54. Martín-Jimenez P, Bermejo-Guerrero L, Navarro-Riquelme M et al (2025) Comprehensive analysis of GDF15 as a biomarker in primary mitochondrial myopathies. Mol Genet Metab. https://​ doi.​org/​10.​1016/j.​ymgme.​2025.​109023 55. Huang Q, Trumpff C, Monzel AS et al (2025) The mitochondrial disease biomarker GDF15 is dynamic, quantifiable in saliva, and correlates with disease severity. Mol Genet Metab. https://​doi.​ org/​10.​1016/j.​ymgme.​2025.​109179 56. Peñas A, Fernández-De la Torre M, Laine-Menéndez S et al (2021) Plasma gelsolin reinforces the diagnostic value of FGF-21 and GDF-15 for mitochondrial disorders. Int J Mol Sci. https://​ doi.​org/​10.​3390/​ijms2​21263​96 57. Correia SP, Moedas MF, Taylor LS et al (2024) Quantitative pro- teomics of patient fibroblasts reveal biomarkers and diagnostic signatures of mitochondrial disease. JCI Insight. https://​doi.​org/​ 10.​1172/​jci.​insig​ht.​178645 58. Li Y, Wang D, Zhou M et al (2025) Hypocitrullinemia as an early diagnostic biomarker for MT-ATP6 mitochondrial diseases. J Mol Neurosci. https://​doi.​org/​10.​1007/​s12031-​025-​02440-6 59. Carli S, Levarlet A, Diodato D et al (2025) Natural history of patients with mitochondrial ATPase deficiency due to pathogenic variants of MT-ATP6 and MT-ATP8. Neurology. https://​doi.​org/​ 10.​1212/​WNL.​00000​00000​213462 60. Manzoni E, Carli S, Gaignard P et al (2024) Deoxyguanosine kinase deficiency: natural history and liver transplant outcome. Brain Commun. https://​doi.​org/​10.​1093/​brain​comms/​fcae1​60 61. Hall AM, Vilasi A, Garcia-Perez I et al (2015) The urinary proteome and metabonome differ from normal in adults with mitochondrial disease. Kidney Int 87:610–622. https://​doi.​org/​ 10.​1038/​ki.​2014.​297 62. Cicchinelli M, Primiano G, Canu F et al (2025) Urinary multi- omics profiling reveals systemic molecular alterations in Progres- sive External Ophthalmoplegia. Int J Mol Sci 26:11257. https://​ doi.​org/​10.​3390/​ijms2​62311​257 63. Cao B, Lu H, Liu P et al (2025) Serum metabolomics signature of maternally inherited diabetes and deafness by gas chroma- tography–time of flight mass spectrometry. J Diabetes Investig 16:146–153. https://​doi.​org/​10.​1111/​jdi.​14334 64. Bax BE, Uçar SK (2025) Untargeted plasma metabolomics extends the biomarker profile of mitochondrial neurogastroin- testinal encephalomyopathy. Int J Mol Sci. https://​doi.​org/​10.​ 3390/​ijms2​61891​07 65. Sharma R, Reinstadler B, Engelstad K et al (2021) Circulating markers of NADH-reductive stress correlate with mitochondrial disease severity. J Clin Invest. https://​doi.​org/​10.​1172/​JCI13​6055 66. Pirinen E, Auranen M, Khan NA et al (2020) Niacin cures sys- temic NAD+ deficiency and improves muscle performance in Adult-Onset Mitochondrial Myopathy. Cell Metab 31:1078- 1090.e5. https://​doi.​org/​10.​1016/j.​cmet.​2020.​04.​008 67. Hirano M, Carelli V, De Giorgio R et al (2021) Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE): position paper on diagnosis, prognosis, and treatment by the MNGIE International Network. J Inherit Metab Dis 44:376–387 68. De Giorgio R, Pironi L, Rinaldi R et al (2016) Liver transplanta- tion for mitochondrial neurogastrointestinal encephalomyopathy. Ann Neurol 80:448–455. https://​doi.​org/​10.​1002/​ana.​24724 69. Halter JP, Michael W, Schüpbach M et al (2015) Allogeneic hae- matopoietic stem cell transplantation for mitochondrial neurogas- trointestinal encephalomyopathy. Brain 138:2847–2858. https://​ doi.​org/​10.​1093/​brain/​awv226 70. Bax BE, Bain MD, Scarpelli M et al (2013) Clinical and bio- chemical improvements in a patient with MNGIE following enzyme replacement. Neurology 81:1269–1271. https://​doi.​org/​ 10.​1212/​WNL.​0b013​e3182​a6cb4b 71. Torres-Torronteras J, Viscomi C, Cabrera-Pérez R et al (2014) Gene therapy using a liver-targeted AAV vector restores nucleo- side and nucleotide homeostasis in a murine model of MNGIE. Mol Ther 22:901–907. https://​doi.​org/​10.​1038/​mt.​2014.6 72. Brevini T, Swift L, Reynolds H et al (2025) Successful AAV8 gene therapy on hepatic ex situ machine perfusion for mito- chondrial neurogastrointestinal encephalomyopathy. J Hepatol 83:1218–1225. https://​doi.​org/​10.​1016/j.​jhep.​2025.​07.​022 73. Sofou K, Shahim P, Tulinius M et al (2019) Cerebrospinal fluid neurofilament light is associated with survival in mitochondrial disease patients. Mitochondrion 46:228–235. https://​doi.​org/​10.​ 1016/j.​mito.​2018.​07.​002 74. Varhaug KN, Hikmat O, Nakkestad HL et al (2021) Serum biomarkers in primary mitochondrial disorders. Brain Com- mun. https://​doi.​org/​10.​1093/​brain​comms/​fcaa2​22 75. Guy J, Shaw G, Ross-Cisneros FN et al (2008) Phosphorylated neurofilament heavy chain is a marker of neurodegeneration in Leber hereditary optic neuropathy (LHON). Mol Vis 14:2443 76. Zheng YS, Sun C, Wang R et al (2021) Neurofilament light is a novel biomarker for mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes. Sci Rep. https://​doi.​org/​10.​ 1038/​s41598-​021-​81721-7 77. Bermejo-Guerrero L, Restrepo-Vera JL, Martin-Jimenez P et al (2026) Clinical heterogeneity and candidate biomarkers in POLG-related mitochondrial disease. Neurol Genet. https://​ doi.​org/​10.​1212/​NXG.​00000​00000​200365 78. Maresca A, Del Dotto V, Romagnoli M et al (2020) Expand- ing and validating the biomarkers for mitochondrial dis- eases. J Mol Med 98:1467–1478. https://​doi.​org/​10.​1007/​ s00109-​020-​01967-y 79. Trifunov S, Paredes-Fuentes AJ, Badosa C et al (2021) Circu- lating cell-free mitochondrial DNA in cerebrospinal fluid as a biomarker for mitochondrial diseases. Clin Chem 67:1113– 1121. https://​doi.​org/​10.​1093/​clinc​hem/​hvab0​91 80. Foerster L, Scholle L, Mayer T et al (2025) Serum chitotriosi- dase-1 (CHIT1) as candidate biomarker for mitochondriopa- thies. J Neurol. https://​doi.​org/​10.​1007/​s00415-​025-​12916-5 81. Keshavan N, Mhaldien L, Gilmour K, Rahman S (2024) Inter- feron stimulated gene expression is a biomarker for primary mitochondrial disease. Ann Neurol 96:1185–1200. https://​doi.​ org/​10.​1002/​ana.​27081 82. Warren EB, Gordon-Lipkin EM, Cheung F et al (2023) Inflam- matory and interferon gene expression signatures in patients with mitochondrial disease. J Transl Med. https://​doi.​org/​10.​ 1186/​s12967-​023-​04180-w 83. Primiano G, Plantone D, Piro G et al (2023) Inflammatory profile in mitochondrial diseases: A cohort study. Eur J Neurol 30:3409–3410 84. Mencias M, Levene M, Blighe K, Bax BE (2021) Circulating mirnas as biomarkers for mitochondrial neuro-gastrointestinal encephalomyopathy. Int J Mol Sci. https://​doi.​org/​10.​3390/​ ijms2​20736​81 https://doi.org/10.1002/ana.24506 https://doi.org/10.1002/jimd.12142 https://doi.org/10.1016/j.mito.2024.101879 https://doi.org/10.1016/j.mito.2024.101879 https://doi.org/10.1016/j.ymgme.2025.109023 https://doi.org/10.1016/j.ymgme.2025.109023 https://doi.org/10.1016/j.ymgme.2025.109179 https://doi.org/10.1016/j.ymgme.2025.109179 https://doi.org/10.3390/ijms22126396 https://doi.org/10.3390/ijms22126396 https://doi.org/10.1172/jci.insight.178645 https://doi.org/10.1172/jci.insight.178645 https://doi.org/10.1007/s12031-025-02440-6 https://doi.org/10.1212/WNL.0000000000213462 https://doi.org/10.1212/WNL.0000000000213462 https://doi.org/10.1093/braincomms/fcae160 https://doi.org/10.1038/ki.2014.297 https://doi.org/10.1038/ki.2014.297 https://doi.org/10.3390/ijms262311257 https://doi.org/10.3390/ijms262311257 https://doi.org/10.1111/jdi.14334 https://doi.org/10.3390/ijms26189107 https://doi.org/10.3390/ijms26189107 https://doi.org/10.1172/JCI136055 https://doi.org/10.1016/j.cmet.2020.04.008 https://doi.org/10.1002/ana.24724 https://doi.org/10.1093/brain/awv226 https://doi.org/10.1093/brain/awv226 https://doi.org/10.1212/WNL.0b013e3182a6cb4b https://doi.org/10.1212/WNL.0b013e3182a6cb4b https://doi.org/10.1038/mt.2014.6 https://doi.org/10.1016/j.jhep.2025.07.022 https://doi.org/10.1016/j.mito.2018.07.002 https://doi.org/10.1016/j.mito.2018.07.002 https://doi.org/10.1093/braincomms/fcaa222 https://doi.org/10.1038/s41598-021-81721-7 https://doi.org/10.1038/s41598-021-81721-7 https://doi.org/10.1212/NXG.0000000000200365 https://doi.org/10.1212/NXG.0000000000200365 https://doi.org/10.1007/s00109-020-01967-y https://doi.org/10.1007/s00109-020-01967-y https://doi.org/10.1093/clinchem/hvab091 https://doi.org/10.1007/s00415-025-12916-5 https://doi.org/10.1002/ana.27081 https://doi.org/10.1002/ana.27081 https://doi.org/10.1186/s12967-023-04180-w https://doi.org/10.1186/s12967-023-04180-w https://doi.org/10.3390/ijms22073681 https://doi.org/10.3390/ijms22073681 Journal of Neurology (2026) 273:263 Page 13 of 14  263 85. Meseguer S, Martínez-Zamora A, García-Arumí E, Andreu AL, Armengod ME (2015) The ROS-sensitive microRNA- 9/9*controls the expression of mitochondrial tRNA-modifying enzymes and is involved in the molecular mechanism ofME- LAS syndrome. Hum Mol Genet 24(1):167-184. https://​doi.​ org/​10.​1093/​hmg/​ddu427 86. Marozzo R, Pegoraro V, Dipietro L et al (2020) Can miR- 34a be suitable for monitoring sensorineural hearing loss in patients with mitochondrial disease? A case series. Int J Neu- rosci 130:1272–1277. https://​doi.​org/​10.​1080/​00207​454.​2020.​ 17315​05 87. Distelmaier F, Klopstock T (2023) Neuroimaging in mitochon- drial disease. Handbook of clinical neurology. Elsevier B.V, Amsterdam, pp 173–185 88. Saneto RP, Friedman SD, Shaw DWW (2008) Neuroimaging of mitochondrial disease. Mitochondrion 8:396–413 89. Yu-Wai-Man C, Smith FE, Firbank MJ et al (2013) Extraocu- lar muscle atrophy and central nervous system involvement in chronic progressive external ophthalmoplegia. PLoS One. https://​doi.​org/​10.​1371/​journ​al.​pone.​00750​48 90. Ardissone A, Bruno C, Diodato D et  al (2021) Clinical, imaging, biochemical and molecular features in Leigh syn- drome: a study from the Italian network of mitochondrial diseases. Orphanet J Rare Dis. https://​doi.​org/​10.​1186/​ s13023-​021-​02029-3 91. Manuel Barragán-Campos H, Vallée J-N, Lô D et al (2005) Brain magnetic resonance imaging findings in patients with mitochon- drial cytopathies. Arch Neurol 62(5):737–42 92. Lerman-Sagie T, Leshinsky-Silver E, Watemberg N et  al (2005) White matter involvement in mitochondrial diseases. Mol Genet Metab 84:127–136 93. Tzoulis C, Engelsen BA, Telstad W (2006) The spectrum of clinical disease caused by the A467T and W748S POLG muta- tions: a study of 26 cases. Brain 129:1685–1692. https://​doi.​ org/​10.​1093/​brain/​awl097 94. Farina L, Chiapparini L, Uziel G et al (2002) MR Findings in Leigh Syndrome with COX Deficiency and SURF-1 Mutations. Am J Neuroradiol 23(7):1095–100 95. Florian A, Ludwig A, Stubbe-Dräger B et al (2015) Character- istic cardiac phenotypes are detected by cardiovascular mag- netic resonance in patients with different clinical phenotypes and genotypes of mitochondrial myopathy. J Cardiovasc Magn Reson. https://​doi.​org/​10.​1186/​s12968-​015-​0145-x 96. Gray R, Kantartzı V, Lopes LR et al (2025) Cardiac magnetic resonance findings in mitochondrial disease: a guide for clini- cians. Eur Heart J Imaging Methods Pract. https://​doi.​org/​10.​ 1093/​ehjimp/​qyaf1​34 97. Gramegna LL, Evangelisti S, Di Vito L et al (2021) Brain MRS correlates with mitochondrial dysfunction biomarkers in MELAS-associated mtDNA mutations. Ann Clin Transl Neurol 8:1200–1211. https://​doi.​org/​10.​1002/​acn3.​51329 98. Sijens PE, Smit GPA, Rödiger LA et al (2008) MR spectros- copy of the brain in Leigh syndrome. Brain Dev 30:579–583. https://​doi.​org/​10.​1016/j.​brain​dev.​2008.​01.​011 99. Kapeller P, Fazekas F, Offenbacher H et al (1996) Magnetic resonance imaging and spectroscopy of progressive cer- ebral involvement in Keams Sayre Syndrome. J Neurol Sci 135(2):126–30 100. Abe K, Yoshimura H, Tanaka H et al (2004) Comparison of con- ventional and diffusion-weighted MRI and proton MR spectros- copy in patients with mitochondrial encephalomyopathy, lactic acidosis, and stroke-like events. Neuroradiology 46:113–117. https://​doi.​org/​10.​1007/​s00234-​003-​1138-2 101. Lunsing RJ, Strating K, de Koning TJ, Sijens PE (2017) Diag- nostic value of MRS-quantified brain tissue lactate level in identifying children with mitochondrial disorders. Eur Radiol 27:976–984. https://​doi.​org/​10.​1007/​s00330-​016-​4454-8 102. Chi CS, Lee HF, Tsai CR et al (2011) Lactate peak on brain MRS in children with syndromic mitochondrial diseases. J Chin Med Assoc 74:305–309. https://​doi.​org/​10.​1016/j.​jcma.​2011.​05.​006 103. Kielty K, Collyer J, Ganesh KD et al (2025) Spectrum of clini- cal neuroimaging in mitochondrial disorders: a neuroanatomical approach. Pediatr Radiol 55:1350–1365 104. Lin A, Andronesi O, Bogner W et al (2021) Minimum reporting standards for in vivo magnetic resonance spectroscopy (MRS- inMRS): experts’ consensus recommendations. NMR Biomed. https://​doi.​org/​10.​1002/​nbm.​4484 105. Lodi R, Rajagopalan B, Blamire AM et al (2004) Abnormal cardiac energetics in patients carrying the A3243G mtDNA mutation measured in vivo using phosphorus MR spectroscopy. Biochim Biophys Acta Bioenerg 1657:146–150. https://​doi.​org/​ 10.​1016/j.​bbabio.​2004.​05.​003 106. Kemp GJ, Meyerspeer M, Moser E (2007) Absolute quantifi- cation of phosphorus metabolite concentrations in human mus- cle in vivo by 31P MRS: a quantitative review. NMR Biomed 20:555–565 107. Jeppesen TD, Quistorff B, Wibrand F, Vissing J (2007) 31P-MRS of skeletal muscle is not a sensitive diagnostic test for mitochon- drial myopathy. J Neurol 254:29–37. https://​doi.​org/​10.​1007/​ s00415-​006-​0229-5 108. Roshanravan B, Liu SZ, Ali AS et al (2021) In vivo mitochon- drial ATP production is improved in older adult skeletal muscle after a single dose of elamipretide in a randomized trial. PLoS One. https://​doi.​org/​10.​1371/​journ​al.​pone.​02538​49 109. Kumar V, Chang H, Reiter DA et al (2017) Phosphorus-31 mag- netic resonance spectroscopy: A tool for measuring in vivo mito- chondrial oxidative phosphorylation capacity in human skeletal muscle. J Vis Exp. https://​doi.​org/​10.​3791/​54977 110. Meyerspeer M, Boesch C, Cameron D et al (2021) 31P magnetic resonance spectroscopy in skeletal muscle: experts’ consensus recommendations. NMR Biomed 34:e4246 111. Li R, Xiao HF, Lyu JH et al (2017) Differential diagnosis of mito- chondrial encephalopathy with lactic acidosis and stroke-like epi- sodes (MELAS) and ischemic stroke using 3D pseudocontinu- ous arterial spin labeling. J Magn Reson Imaging 45:199–206. https://​doi.​org/​10.​1002/​jmri.​25354 112. Oppenheim C, Galanaud D, Samson Y et al (2000) Can diVusion weighted magnetic resonance imaging help diVerentiate stroke from stroke-like events in MELAS? J Neurol Neurosurg Psychia- try 69(2):248–50 113. Sakai Y, Kira R, Torisu H, Ihara K, Yoshiura T, Hara T (2006) Persistent diffusion abnormalities in the brain stem of three- children with mitochondrial diseases. AJNR Am J Neuroradiol 27(9):1924-1926 114. De Picker LJ, Morrens M, Branchi I et al (2023) TSPO PET brain inflammation imaging: a transdiagnostic systematic review and meta-analysis of 156 case-control studies. Brain Behav Immun 113:415–431. https://​doi.​org/​10.​1016/j.​bbi.​2023.​07.​023 115. Wijesinghe SS, Rowe JB, Mason HD et al (2025) Post-mortem validation of in vivo TSPO PET as a microglial biomarker. Brain 148:1904–1910. https://​doi.​org/​10.​1093/​brain/​awaf0​78 116. Van Den Ameele J, Hong YT, Manavaki R et al (2021) PK11195- PET brain imaging of the mitochondrial translocator protein in mitochondrial disease. Neurology 96:E2761–E2773. https://​doi.​ org/​10.​1212/​WNL.​00000​00000​012033 117. Liu F, Ruan W, Wang Y, Lan X (2019) Simultaneous 18 F-FDG PET/MRI assists diagnosis of a rare disease, MELAS. Clin Nucl Med 44:81–82. https://​doi.​org/​10.​1097/​RLU.​00000​00000​002344 118. Ikawa M, Okazawa H, Arakawa K et al (2009) PET imaging of redox and energy states in stroke-like episodes of MELAS. https://doi.org/10.1093/hmg/ddu427 https://doi.org/10.1093/hmg/ddu427 https://doi.org/10.1080/00207454.2020.1731505 https://doi.org/10.1080/00207454.2020.1731505 https://doi.org/10.1371/journal.pone.0075048 https://doi.org/10.1186/s13023-021-02029-3 https://doi.org/10.1186/s13023-021-02029-3 https://doi.org/10.1093/brain/awl097 https://doi.org/10.1093/brain/awl097 https://doi.org/10.1186/s12968-015-0145-x https://doi.org/10.1093/ehjimp/qyaf134 https://doi.org/10.1093/ehjimp/qyaf134 https://doi.org/10.1002/acn3.51329 https://doi.org/10.1016/j.braindev.2008.01.011 https://doi.org/10.1007/s00234-003-1138-2 https://doi.org/10.1007/s00330-016-4454-8 https://doi.org/10.1016/j.jcma.2011.05.006 https://doi.org/10.1002/nbm.4484 https://doi.org/10.1016/j.bbabio.2004.05.003 https://doi.org/10.1016/j.bbabio.2004.05.003 https://doi.org/10.1007/s00415-006-0229-5 https://doi.org/10.1007/s00415-006-0229-5 https://doi.org/10.1371/journal.pone.0253849 https://doi.org/10.3791/54977 https://doi.org/10.1002/jmri.25354 https://doi.org/10.1016/j.bbi.2023.07.023 https://doi.org/10.1093/brain/awaf078 https://doi.org/10.1212/WNL.0000000000012033 https://doi.org/10.1212/WNL.0000000000012033 https://doi.org/10.1097/RLU.0000000000002344 Journal of Neurology (2026) 273:263 263   Page 14 of 14 Mitochondrion 9:144–148. https://​doi.​org/​10.​1016/j.​mito.​2009.​ 01.​011 119. Yoder KK, Nho K, Risacher SL et al (2013) Influence of TSPO genotype on 11C-PBR28 standardized uptake values. J Nucl Med 54:1320–1322. https://​doi.​org/​10.​2967/​jnumed.​112.​118885 120. Azimi MS, Rahmim A, Arabi H et al (2025) Critical review of partial volume correction methods in PET and SPECT imaging: benefits, pitfalls, challenges, and future outlook. Eur J Nucl Med Mol Imaging. https://​doi.​org/​10.​1007/​s00259-​025-​07612-5 121. Kandemirli SG, Al-Dasuqi K, Aslan B et al (2025) Overview of neuroimaging in primary mitochondrial disorders. Pediatr Radiol 55:765–791 122. Bellanti R, Keh RYS, Keddie S et al (2025) Plasma periaxin is a biomarker of peripheral nerve demyelination. Brain. https://​doi.​ org/​10.​1093/​brain/​awaf2​34 123. Keddie S, Smyth D, Keh RYS et al (2023) Peripherin is a bio- marker of axonal damage in peripheral nervous system disease. Brain 146:4562–4573. https://​doi.​org/​10.​1093/​brain/​awad2​34 124. Orssaud C, Reynier P (2025) MetabOCT: a clinical trial look- ing for a metabolomic signature predicting the onset of Leber’s hereditary optic neuropathy in healthy MtDNA mutations carri- ers. Metabolomics. https://​doi.​org/​10.​1007/​s11306-​025-​02328-x 125. Raman S, Chentouf L, DeVile C et al (2018) Near infrared spec- troscopy with a vascular occlusion test as a biomarker in chil- dren with mitochondrial and other neuro-genetic disorders. PLoS ONE. https://​doi.​org/​10.​1371/​journ​al.​pone.​01997​56 126. Khaksari K, Chen WL, Chanvanichtrakool M et al (2024) Appli- cations of near-infrared spectroscopy in epilepsy, with a focus on mitochondrial disorders. Neurotherapeutics 21:e00323 127. Saito S, Takahashi Y, Ohki A et al (2019) Early detection of elevated lactate levels in a mitochondrial disease model using chemical exchange saturation transfer (CEST) and magnetic reso- nance spectroscopy (MRS) with 7T MR imaging. Radiol Phys Technol 12:232–233 128. Khan AS, Peterson KA, Vittay OI et  al (2024) Deuterium metabolic imaging of Alzheimer Disease at 3-T magnetic field strength: a pilot case-control study. Radiology. https://​doi.​org/​ 10.​1148/​radiol.​232407 129. Ahmadian N, Gosselink M, Otto S et  al (2025) Dynamic deuterium metabolic imaging in glioblastoma at 7T. Magn Reson Mater Phys Biol Med. https://​doi.​org/​10.​1007/​ s10334-​025-​01299-3 130. De Feyter HM, de Graaf RA (2021) Deuterium metabolic imag- ing – back to the future. J Magn Reson. https://​doi.​org/​10.​1016/j.​ jmr.​2021.​106932 131. Tsukada H, Kanazawa M, Ohba H et al (2016) PET imaging of mitochondrial complex i with 18F-BCPP-EF in the brains of MPTP-treated monkeys. J Nucl Med 57:950–953. https://​doi.​org/​ 10.​2967/​jnumed.​115.​169615 132. Engle J, Saberi P, Bain P et al (2024) Oxygen extraction fraction (OEF) values and applications in neurological diseases. Neurol Sci 45:3007–3020 133. Engle J, Saberi P, Bain P, Ikram A, Selim M, Soman S (2024) Oxygen extraction fraction (OEF) values and applicationsin neu- rological diseases. Neurol Sci 45(7):3007-3020. https://​doi.​org/​ 10.​1007/​s10072-​024-​07362-6 134. Wylezinska-Arridge M, Pizzamiglio C, Küppers F, Shah NJ, Mancini L, Thomas D, Pitceathly RDS, Thornton JS (2025) Calfmuscle oxygen extraction fraction measurement using the multi contrast EPI with keyhole sequence – feasibilitystudy. In: Proceedings of the international society for magnetic resonance in medicine (ISMRM). Honolulu, HI, USA. Abstract 2476. Available from: https://​archi​ve.​ismrm.​org/​2025/​2476.​html 135. Tan L, Zschüntzsch J, Meyer S et al (2024) Non-invasive optoa- coustic imaging of glycogen-storage and muscle degeneration in late-onset Pompe disease. Nat Commun. https://​doi.​org/​10.​1038/​ s41467-​024-​52143-6 136. Xia W, Need E, Schiavone C et al (2025) Image-guided targeting of mitochondrial metabolism sensitizes pediatric malignant rhab- doid tumors to low-dose radiotherapy. Sci Adv 11(21):eadv930 137. Zhang C, Qiu Z, Zhang L et al (2022) Mitochondria-targeted fluorescence/photoacoustic dual-modality imaging probe tailored for visual precise diagnosis of drug-induced liver injury. Anal Chem 94:6251–6260. https://​doi.​org/​10.​1021/​acs.​analc​hem.​ 2c000​41 138. Else TR, Wright L, Schon K, Tiet MY, Seikus C, Ashby E, et al. Photoacoustic imaging in mitochondrial disease. medRxiv.202 6:2026.03.10.26347962.https://​www.​medrx​iv.​org/​conte​nt/​10.​ 64898/​2026.​03.​10.​26347​962v1 139. Alix JJP, Plesia M, Stockholm D et al (2024) In vivo Raman spectroscopy of muscle is highly sensitive for detection of healthy muscle and highly specific for detection of disease. Anal Chem 96:15991–15997. https://​doi.​org/​10.​1021/​acs.​analc​hem.​ 4c034​30 140. Alix JJP, Plesia M, Shaw PJ et al (2023) Combining electromyo- graphy and Raman spectroscopy: optical EMG. Muscle Nerve 68:464–470. https://​doi.​org/​10.​1002/​mus.​27937 https://doi.org/10.1016/j.mito.2009.01.011 https://doi.org/10.1016/j.mito.2009.01.011 https://doi.org/10.2967/jnumed.112.118885 https://doi.org/10.1007/s00259-025-07612-5 https://doi.org/10.1093/brain/awaf234 https://doi.org/10.1093/brain/awaf234 https://doi.org/10.1093/brain/awad234 https://doi.org/10.1007/s11306-025-02328-x https://doi.org/10.1371/journal.pone.0199756 https://doi.org/10.1148/radiol.232407 https://doi.org/10.1148/radiol.232407 https://doi.org/10.1007/s10334-025-01299-3 https://doi.org/10.1007/s10334-025-01299-3 https://doi.org/10.1016/j.jmr.2021.106932 https://doi.org/10.1016/j.jmr.2021.106932 https://doi.org/10.2967/jnumed.115.169615 https://doi.org/10.2967/jnumed.115.169615 https://doi.org/10.1007/s10072-024-07362-6 https://doi.org/10.1007/s10072-024-07362-6 https://archive.ismrm.org/2025/2476.html https://doi.org/10.1038/s41467-024-52143-6 https://doi.org/10.1038/s41467-024-52143-6 https://doi.org/10.1021/acs.analchem.2c00041 https://doi.org/10.1021/acs.analchem.2c00041 https://www.medrxiv.org/content/10.64898/2026.03.10.26347962v1 https://www.medrxiv.org/content/10.64898/2026.03.10.26347962v1 https://doi.org/10.1021/acs.analchem.4c03430 https://doi.org/10.1021/acs.analchem.4c03430 https://doi.org/10.1002/mus.27937 Non-invasive biomarkers for diagnosis and monitoring of primary mitochondrial diseases Abstract Introduction Functional assessments: from motor tests to digital monitoring Laboratory biomarkers: untargeted and targeted approaches Neuroimaging for diagnosis and monitoring Emerging non-invasive biomarkers and imaging approaches with promise in PMDs Conclusion Acknowledgements References