Skip to main content
NCBI home page
Neurotherapeutics logoLink to Neurotherapeutics
. 2024 Jan 30;21(1):e00325. doi: 10.1016/j.neurot.2024.e00325

Biomarkers of mitochondrial disorders

Brian J Shayota 1
PMCID: PMC10903091  PMID: 38295557

Abstract

Mitochondrial diseases encompass a heterogeneous group of disorders with a wide range of clinical manifestations, most classically resulting in neurological, muscular, and metabolic abnormalities, but having the potential to affect any organ system. Over the years, substantial progress has been made in identifying and characterizing various biomarkers associated with mitochondrial diseases. This review summarizes the current knowledge of mitochondrial biomarkers based on a literature review and discusses the evidence behind their use in clinical practice. A total of 13 biomarkers were thoroughly reviewed including lactate, pyruvate, lactate:pyruvate ratio, creatine kinase, creatine, amino acid profiles, glutathione, malondialdehyde, GDF-15, FGF-21, gelsolin, neurofilament light-chain, and circulating cell-free mtDNA. Most biomarkers had mixed findings depending on the study, especially when considering their utility for specific mitochondrial diseases versus mitochondrial conditions in general. However, in large biomarker comparison studies, GDF-15 followed by FGF-21, seem to have the greatest value though they are still not perfect. As such, additional studies are needed, especially in light of newer biomarkers that have not yet been thoroughly investigated. Understanding the landscape of biomarkers in mitochondrial diseases is crucial for advancing early detection, improving patient management, and developing targeted therapies.

Keywords: Mitochondrial disease, Mitochondrial dysfunction, Biomarker

Introduction

Mitochondrial diseases encompass a heterogeneous group of disorders caused by dysfunction in the mitochondria. The term “mitochondrial disease” itself, is rather non-specific and can include numerous genetic and acquired etiologies. Dysfunction of these organelles can lead to a wide range of clinical manifestations, most classically resulting in neurological, muscular, and metabolic abnormalities, but having the potential to affect any organ system. The early and accurate diagnosis of mitochondrial diseases is challenging due to their variable onset and diverse clinical presentation.

Over the years, substantial progress has been made in identifying and characterizing various biomarkers associated with mitochondrial diseases. A biomarker, short for biological marker, is defined as a measurable and quantifiable substance that provides information about a particular biological process and can be found in various biological sources including but not limited to blood, urine, cerebral spinal fluid, and tissue samples. In some cases, biomarkers can also be measured by advanced imaging techniques like magnetic resonance spectroscopy. Biomarkers play a pivotal role aiding in the diagnosis and monitoring the metabolic status of a patient with a mitochondrial disease. The availability of advanced and comprehensive genetic testing has in some ways made it easier to diagnose mitochondrial disorders, but it also presents the frequent challenge of resolving variants of uncertain significance (VUS), meaning that in some cases, additional evidence with mitochondrial biomarkers can help support or rule-out a potential diagnosis. Furthermore, the consideration of heteroplasmy levels with different tissue segregation involving variants in mtDNA means next-generation sequencing may not be as powerful of a tool for mitochondrial DNA encoded conditions as it is for nuclear DNA encoded.

This review summarizes the current knowledge of mitochondrial biomarkers based on a literature review and discusses the evidence behind their use in clinical practice. This study will focus on the biomarkers found in commonly obtainable bodily fluids including blood/plasma, urine, and CSF. Understanding the landscape of biomarkers in mitochondrial diseases is crucial for advancing early detection, improving patient management, and developing targeted therapies.

Methods

This is a narrative review only, so no original data is being presented in this manuscript. Data for this review was collected based on pubmed searches for relevant articles, as of August 14th, 2023, using the search terms “mitochondrial disease” or “mitochondrial disorder” in conjunction with the terms “biomarker” or “diagnostic test.” Additional filters were set for human studies and English language. Prior review articles on this specific topic were also used to retrieve additional original studies that were not identified through the PubMed search.

Parameters for inclusion were (1) reports that include at least 3 human subjects, (2) confirmation of mitochondrial disease diagnosis, usually by molecular testing, and (3) inclusion of actual lab values, either in the main manuscript or in supplementary documents. In cases where lab values were reported for individual subjects in a cohort but not statistically summarized as a cohort, a simple statistical analysis was performed to determine the mean and standard deviation based on the provided data. In a few studies, the specific blood sample extracted and used for testing (i.e. plasma versus serum or venous versus arterial) was not made clear, so “blood” is used as a generalized term in this review to described any blood based testing. For each of the biomarkers described, a brief literature review was performed to describe the current knowledge of the underlying mechanisms in which the biomarker relates to mitochondrial function.

In total, 13 unique biochemical biomarker tests for mitochondrial disease that were studied more than once were identified in this review. Among them, 55 original manuscripts were found, many of which studied multiple biomarkers at once (Table 1). Significant variability was found in the reporting of clinical data and methods between the studies, particularly for some of the more recently described biomarkers.

Table 1.

Summary of biochemical biomarker studies for mitochondrial disorders in the medical literature.

Biomarker Manuscript Diagnosis P (n) C (n) Biologic Sample Patients (mean ​± ​SD/SEM) Controls (mean ​± ​SD) P-value AUC Sensitivity % (95 ​% CI) Specificity % (95 ​% CI) Threshold
Lactate Debray et al., 2007 [24] Mixed cohort 35 n.a. Blood 4.86 ​± ​3.11 ​mmol/l n.a. n.r.
Shaham et al., 2009 [32] Mixed cohort 16 25 Plasma +107 ​% vs controls n.r. <0.001
Heinicke et al., 2011 [18] Mitochondrial myopathy 5 4 Arterial blood Rest: 2.80 ​± ​1.38 ​mM Rest: 0.83 ​± ​0.25 ​mM n.s.
Peak exercise:
13.36 ​± ​2.26 ​mmol/l		 Peak exercise:
8.67 ​± ​0.47 ​mmol/l		 <0.05
5 ​min of recovery:
14.48 ​± ​1.87 ​mM		 5 ​min of recovery:
11.73 ​± ​1.42 ​mM		 n.s.
Kaufmann P et al., 2011 [5] MELAS (A3243G) 31 50 Venous 2.9 ​± ​2 ​mmol/l 1.9 ​± ​0.9 ​mmol/l n.r.
Magner et al., 2011 [11] Mixed cohort 24 16 Blood 3.87 ​± ​0.48 ​mmol/l 1.70 ​± ​0.13 ​mmol/l <0.001
CSF 4.43 ​± ​0.55 ​mmol/l 1.64 ​± ​0.07 ​mmol/l <0.001
Suomalainen et al., 2011 [25] Mixed cohort 35 n.a. Serum 2.37 ​± ​1.52 ​mmol/la Ref: <2.4 ​mmol/l n.r. 0.90 (0.84–0.96) 52.6 (39.0–66.0)
Combined adult & peds		 92.8 (83.9–97.6)
Combined adult & peds		 2.4 ​mmol/l
Yamada et al., 2012 [12] Mixed cohort 17 146 Blood 41.6 ​± ​25.4 ​mg/dl 14.6 ​± ​6.6 ​mg/dl <0.01 0.926 (0.852–1.000) 88.2 84.4 18.95 ​mg/dl
49 CSF 44.5 ​± ​19.4 ​mg/dl 12.8 ​± ​2.7 ​mg/dl <0.01 0.994 (0.981–1.008) 94.1 100 19.9 ​mg/dl
Minuz et al., 2012 [66] Mixed cohort 12 12 Plasma 2.35 ​± ​0.31 ​mmol/l 1.11 ​± ​0.15 ​mmol/l 0.01
Davis et al., 2013 [23] Mixed cohort 54 66 Serum 163 ​% vs controls n.r. <0.0001 0.76 (0.67–0.85) 15.09 (7.19–28.14) 98.46 (90.60–99.92) n.r.
Jeppesen et al., 2013 [19] Mitochondrial myopathy 10 10 Arterial blood Rest: 1.9 ​± ​0.3 ​mM Rest: 0.7 ​± ​0.1 ​mM <0.05
Exercise: 5.4 ​± ​1.2 ​mM Exercise: 0.7 ​± ​0.1 ​mM <0.05
Sofou et al. 2014 [15] Leigh syndrome 117 n.a Blood 4.62 ​± ​3.42 ​mmol/la n.a. n.r.
85 n.a. CSF 3.84 ​± ​2.41 ​mmol/la n.a. n.r.
Yatsuga et al., 2015 [22] Mixed cohort 48 n.a. Plasma 2.93 ​± ​1.58 ​mmol/l Ref: <1.9 ​mmol/l <0.001 0.882 (0.814–0.950 1.9 ​mmol/l
Zhang et al., 2015 [6] MELAS (A3243G) 40 n.a. Serum Range: 4.2–9.9 ​mmol/l Range: 2.3–10.8 ​mmol/l >0.05
Xia et al., 2016 [7] MELAS (A3243G) 100 n.a. Plasma Range: 1.4–19.0 ​mmol/l Ref: 0.7–2.0 ​mmol/l n.r.
Feldman et al., 2017 [67] Pediatric acute liver failure 8 n.a. Serum Median: 4.2 ​mmol/l Ref: <2.5 ​mmol/l n.s.
Lunsing et al. 2017 [14] Mixed cohort 17 44 Venous 3.7 ​± ​3.0 ​mmol/l 1.5 ​± ​0.9 ​mmol/l <0.001
12 25 CSF 2.3 ​± ​2.1 ​mmol/l 1.5 ​± ​0.3 ​mmol/l >0.05
Morovat et al., 2017 [68] Mixed cohort 57 11 Serum Median: 1.50 ​mmol/l (IQR ​± ​1.23) Median: 0.90 ​mmol/l (IQR ​± ​0.48) n.r.
Wang et al., 2017 [8] MELAS (A3243G) 34 34 Serum 3.0 ​± ​1.3 ​mmol/l 1.6 ​± ​0.6 ​mmol/l <0.001
MELAS (non-A3243G) 22 34 Serum 2.6 ​± ​0.7 ​mmol/l 1.6 ​± ​0.6 ​mmol/l <0.001
Yu et al., 2018 [16] mtDNA Leigh syndrome 13 n.a. Serum 2.58 ​± ​1.33 ​mmol/la Ref: <2.2 ​mmol/l n.r.
mtDNA Leigh syndrome 6 n.a. CSF 3.98 ​± ​1.13 ​mmol/la Ref: <2.2 ​mmol/l n.r.
Koga et al., 2019 [26] Mixed cohort 11 n.a. Blood 4.95 ​± ​1.1 ​mmol/l Ref: <1.2 ​mmol/l n.r.
Sofou et al., 2019 [13] Mixed cohort 46 22 Blood 1.5 ​mmol/l (0.7–2.4) 1.1 ​mmol/l (0.7–2.1) 0.21 0.6
Mixed cohort 46 22 CSF 2.8 ​mmol/l (2.1–3.7) 1.5 ​mmol/l (1.4–2.0) <0.001 0.88
Keshavan et al., 2020 [17] RRM2B associated MDDS 9 n.a. Serum 9.9 ​± ​11.3 ​mmol/la Ref: <2.2 ​mmol/l n.r.
RRM2B associated MDDS 5 n.a. CSF 3.8 ​± ​1.2 ​mmol/la Ref: <2.2 ​mmol/l n.r.
Huddar et al. 2021 [50] Mixed cohort 30 n.a. Serum 31.7 ​± ​16.11 ​mg/dl n.a. n.r.
Evangelisti et al., 2022 [9] MELAS (A3243G) 13 n.a. Plasma 22.0 ​± ​10.9 ​mg/dl n.a. n.r.
Guerrero-Molina et al., 2022 [10] MELAS 9 19 Plasma 3.37 ​± ​0.52 μmol/l 1.22 ​± ​0.10 μmol/l 0.003
MELAS 9 19 CSF 5.04 ​± ​0.46 μmol/l 1.68 ​± ​0.15 μmol/l <0.001
Wu et al., 2022 [69] Mixed cohort 112 n.a. Serum Median: 4.12 ​mmol/l
Range: 1.8–26 ​mmol/l		 Ref 1.4–1.9 ​mmol/l n.r.
Zhao et al., 2022 [20] Mitochondrial diabetes 10 n.a. Plasma Rest: Range 0.76–3.3 ​mmol/l Ref: 0.98–1.9 ​mmol/l n.r.
6 6 Plasma During exercise: 8.39 ​± ​2.75 ​mmol/l During exercise: 3.53 ​± ​1.47 ​mmol/l 0.003 0.94 (0.81–1) 83 100 5.5 ​mmol/l
6 6 Plasma Post exercise: 6.02 ​± ​2.65 ​mmol/l Post exercise: 2.17 ​± ​1.27 ​mmol/l 0.011 0.96 (0.85–1) 100 83 3.4 ​mmol/l
Pyruvate Debray et al., 2007 [24] Mixed cohort 35 n.a. Blood 0.149 ​± ​0.044 ​mmol/l n.a. n.r.
Suomalainen et al., 2011 [25] Mixed cohort 20 n.a. Serum 99.7 ​± ​60.7 μmol/la Ref: <70 μmol/l n.r. 0.80 (0.70–0.93) 75.0 (56.6–88.5)
Combined adult & peds		 87.2 (74.3–95.2)
Combined adult & peds		 70 μmol/l
Yamada et al., 2012 [12] Mixed cohort 17 146 Blood 2.5 ​± ​1.5 ​mg/dl 1.2 ​± ​0.3 ​mg/dl <0.01 0.907 (0.822–0.992) 88.2 81.2 1.45 ​mg/dl
Mixed cohort 17 49 CSF 2.3 ​± ​1.3 ​mg/dl 0.8 ​± ​0.2 ​mg/dl <0.01 0.983 (0.953–1.012) 88.2 100 1.45 ​mg/dl
Davis et al., 2013 [23] Mixed cohort 54 66 Serum 119 ​% vs controls n.r. 0.055 0.62 (0.52–0.72) 34.62 (22.33–49.16) 83.08 (71.31–90.85) n.r.
Yatsuga et al., 2015 [22] Mixed cohort 48 n.a. Serum 133.1 ​± ​55.1 μmol/l Ref: <106.8 μmol/l <0.001 0.706 (0.614–0.798) 106.8 μmol/l
Koga et al., 2019 [26] Mixed cohort 11 n.a. Blood 1.9 ​± ​0.5 ​mg/dl Ref 0.8–1.2 ​mg/dl
Lactate:Pyruvate Debray et al., 2007 [24] Mixed cohort 35 n.a. Blood 31.4 ​± ​13.0 n.a. n.r. 0.83 (0.70–0.80) 77.1 (59.8–89.5) 90.9 (58.7–98.5) >23.3
Suomalainen et al., 2011 [25] Mixed cohort 19 n.a. Serum 26.2 ​± ​18.7a Ref: <30 n.r. 0.90 (0.82–0.98) 31.3 (16.1–50.0)
Combined adult & peds		 100.0 (92.5–100.0)
Combined adult & peds		 >30
Yamada et al., 2012 [12] Mixed cohort 17 146 Blood 17.3 ​± ​7.8 11.4 ​± ​3.3 <0.01 0.768 (0.622–0.915) 82.4 68.7 12.4
49 CSF 20.7 ​± ​6.6 15.4 ​± ​2.8 <0.01 0.80 (0.638–0.962) 76.5 90.6 17.14
Davis et al., 2013 [23] Mixed cohort 54 66 Serum 135 ​% vs controls n.r. 0.0005 0.71 (0.60–0.81) 11.54 (4.78–24.13) 98.48 (90.73–99.92) n.r
Yatsuga et al., 2015 [22] Mixed cohort 48 n.a. Serum 22.0 ​± ​7.5 Ref: <30 <0.001 0.919 (0.865–0.973) 30
Feldman et al., 2017 [67] Pediatric acute liver failure 8 n.a. Serum Median: 20.6 Ref: <25 n.s.
Koga et al., 2019 [26] Mixed cohort 11 n.a. Blood 28.1 ​± ​8.6 Ref: <10 n.r.
Keshavan et al., 2020 [17] RRM2B associated MDDS 3 n.a. Serum 33.1 ​± ​19.3a n.a. n.r.
Creatine kinase Karppa et al., 2004 [70] MELAS 24 n.a. Serum 179.4 ​± ​119.3 U/La Ref: <150 (women),
<270 (men)		 n.r.
Suomalainen et al., 2011 [25] Mixed cohort 33 n.a. Serum 294.2 ​± ​467.9 U/La Variable ref range n.r. 0.63 (0.51–0.74) 34.1 (20.5–49.9) combined adult & peds 73.7 (62.3–83.1)
Combined adult & peds		 Variable ref range
Davis et al., 2013 [23] Mixed cohort 54 66 Serum 181 ​% vs controls n.r. 0.261 0.56 (0.44–0.68) 22.22 (12.48–35.95) 96.97 (88.52–99.47)
Yatsuga et al., 2015 [22] Mixed cohort 48 n.a. Serum 209.8 ​± ​309 U/L Variable ref range >0.01 0.609 (0.506–0.711) Variable ref range
Keshavan et al., 2020 [17] RRM2B associated MDDS 9 n.a. Serum 401.3 ​± ​308.5 U/La Variable ref range n.r.
Huddar et al. 2021 [50] Mixed cohort 28 n.a. Serum 291.14 ​± ​385.7 U/L n.a.
Creatine Shaham et al., 2009 [32] Mixed cohort 16 25 Plasma +233 ​% vs controls n.r.
14 4 Plasma +201 ​% vs controls <0.05
Pajares et al., 2013 [30] Mixed cohort 33 196 Plasma 146 ​± ​28 μmol/l 71 ​± ​2 μmol/l <0.0001 0.64 60 83 120 μmol/l
Maresca et al., 2020 [31] Mixed cohort 69 29 Serum 12.11 ​± ​10.83 μmol/l 6.92 ​± ​5.61 μmol/l <0.01 0.62 (0.51–0.73) 39 (28–50) 93 (84–102) 16 μmol/l
Amino acids Shaham et al., 2009 [32] Mixed cohort 16 25 Plasma Alanine: 146 ​% vs controls n.r. n.r.
Magner et al., 2011 [11] Mixed cohort 24 16 Blood Alanine: 558 ​± ​44 μmol/l n.r. n.r.
CSF Alanine: 51 ​± ​8 μmol/l n.r. n.r.
El-Hattab et al., 2012 [43] MELAS 10 10 Plasma Arginine: 57.4 ​± ​2.4 uM Arginine: 77.8 ​± ​4.4 uM <0.001
Citrulline: 23.2 ​± ​1.6 uM Citrulline: 28.1 ​± ​1.1 uM <0.05
Alanine: 534.0 ​± ​19.3 uM Alanine: 358.2 ​± ​25.0 uM <0.001
Valine: 228.1 ​± ​9.7 uM Valine: 181.8 ​± ​7.1 uM <0.005
Leucine: 143.4 ​± ​8.2 uM Leucine: 114.3 ​± ​8.3 uM <0.05
Isoleucine: 80.5 ​± ​6.7 uM Isoleucine: 62.9 ​± ​5.4 uM <0.05
ADMA: 0.53 ​± ​0.04 uM ADMA: 0.38 ​± ​0.02 uM <0.005
Salmi et al., 2012 [44] Mixed cohort 10 6 Plasma Total cysteine: 0.041 ​± ​0.025 μmol/g protein Total cysteine: 0.039 ​± ​0.022 μmol/g protein >0.05
Reduced cysteine: 12.7 μmol/l Reduced cysteine: 19.5 μmol/l 0.008
Reduced/Oxidized ratio: 0.063 Reduced/Oxidized ratio: 0.098 0.023
Clarke et al., 2013 [27] Mixed cohort 19 27 Plasma Alanine: 161.81 ​% median vs controls n.r. <0.01
Histidine: 17.82 ​% median vs controls n.r. <0.05
1-Methyl-histidine: 17.6 ​% median vs controls n.r. <0.01
Lysine: 40.13 ​% median vs controls n.r. <0.01
Asparagine: 19.67 ​% median vs controls n.r. <0.01
Crippa et al., 2015 [71] Pearson syndrome 4 n.a. Plasma Alanine: 730.8 ​± ​398.7 μmol/l Ref: 240–600 μmol/l n.r.
Arginine: 39.1 ​± ​29.1 μmol/l Ref: 40–160 μmol/l n.r.
El-Hattab et al., 2016 [42] MELAS 5 5 Plasma Arginine: 61 ​± ​5 μmol/l 151 ​± ​18 μmol/l <0.001
Citrulline: 22 ​± ​7 μmol/l 16 ​± ​2 μmol/l n.s.
ADMA: 0.66 ​± ​0.04 μmol/l ADMA: 0.81 ​± ​0.05 μmol/l <0.05
Koga et al., 2019 [26] Mixed cohort 11 n.a. Blood Alanine: 587.3 ​± ​66.6 μmol/l Ref: 270–420 μmol/l n.r.
Evangelisti et al., 2022 [9] MELAS (A3243G) 11 n.a. Plasma Alanine: 499.5 ​± ​114.3 ​mg/dl n.a. n.r.
Arginine: 63.5 ​± ​33 ​mg/dl n.a. n.r.
Guerrero-Molina et al., 2022 [10] MELAS 9 19 Plasma Aminobutirate: 13.67 ​± ​1.67 μmol/l Aminobutirate: 26.79 ​± ​2.90 μmol/l 0.003
Arginine: 142.56 ​± ​41.93 μmol/l Arginine: 59.26 ​± ​5.17 μmol/l 0.002
Asparagine: 46.11 ​± ​3.49 μmol/l Arginine: 59.26 ​± ​5.17 μmol/l 0.021
Cystine: 37.56 ​± ​2.59 μmol/l Cystine: 55.53 ​± ​3.10 μmol/l 0.002
Glutamate: 22.67 ​± ​3.37 μmol/l Glutamate: 45.05 ​± ​5.68 μmol/l 0.017
Glutamine: 393.67 ​± ​25.19 μmol/l Glutamine: 517.37 ​± ​23.25 μmol/l 0.005
Glycine: 132.33 ​± ​9.86 μmol/l Glycine: 230.95 ​± ​12.55 μmol/l <0.001
Lysine: 148.33 ​± ​12.59 μmol/l Lysine: 220.95 ​± ​11.96 μmol/l 0.002
Methionine: 17.67 ​± ​2.07 μmol/l Methionine: 26.21 ​± ​2.40 μmol/l 0.013
Phosphoserine: 4.00 ​± ​0.41 μmol/l Phosphoserine: 6.47 ​± ​0.50 μmol/l 0.004
Serine: 83.67 ​± ​6.37 μmol/l Serine: 132.63 ​± ​6.27 μmol/l <0.001
Threonine: 85.33 ​± ​1.74 μmol/l Threonine: 168.63 ​± ​12.19 μmol/l <0.001
CSF Alanine: 62.30 ​± ​6.83 μmol/l Alanine: 26.58 ​± ​1.38 μmol/l <0.001
Arginine: 37.83 ​± ​3.07 μmol/l Arginine: 18.03 ​± ​1.12 μmol/l <0.001
Citrulline: 2.81 ​± ​0.58 μmol/l Citrulline: 0.66 ​± ​0.05 μmol/l 0.001
Glutamate: 18.48 ​± ​1.34 μmol/l Glutamate: 5.31 ​± ​1.09 μmol/l <0.001
Glutamine: 336.31 ​± ​12.92 μmol/l Glutamine: 407.06 ​± ​15.74 μmol/l 0.017
Histidine: 12.66 ​± ​0.82 μmol/l Histidine: 9.85 ​± ​0.46 μmol/l 0.011
Isoleucine: 14.23 ​± ​1.42 μmol/l Isoleucine: 8.83 ​± ​0.33 μmol/l <0.001
Leucine: 24.44 ​± ​3.33 μmol/l Leucine: 15.52 ​± ​1.14 μmol/l 0.002
Taurine: 8.43 ​± ​0.56 μmol/l Taurine: 5.78 ​± ​0.41 μmol/l 0.004
Valine: 26.61 ​± ​5.88 μmol/l Valine: 15.46 ​± ​1.47 μmol/l 0.032
Glutathione Piccolo G et al., 1991 [46] CPEO 11 25 Plasma GSH: 8.64 ​± ​1.82 uM Plasma GSH: 11 ​± ​3 uM <0.01
Erythrocytes GSH: 12.41 ​± ​2.37 ​nmol/mg protein GSH: 18 ​± ​2.2 ​nmol/mg protein <0.01
Salmi et al., 2012 [44] Mixed cohort 10 6 Plasma Total GSH: 6.4 ​± ​1.1 μmol/g protein Total GSH: 6.7 ​± ​0.56 μmol/g protein >0.05
Pastore et al., 2013 [47] Leigh syndrome 10 20 Lymphocytes Tot GSH: Median 18.6 ​nmol/mg proteins, CI 12.55–37.89 Tot GSH: Median 44.02 ​nmol/mg proteins, CI 21.96–60.42 <0.05
GSH: Median 5.28 ​nmol/mg proteins, CI 1.90–13.40 GSH: Median 39.10 ​nmol/mg proteins, CI 19.31–54.55 <0.001
GSSG: Median 7.25 ​nmol/mg proteins, CI 2.76–21.25 GSSG: Median 2.94 ​nmol/mg proteins, CI 1.33–3.82 <0.001
GS-Pro: Median 4.93 ​nmol/mg proteins, CI 2.11–12.80 GS-Pro: Median 1.64 ​nmol/mg proteins, CI 1.08–3.41 <0.001
Plasma Total thiols: 0.15 ​± ​0.05 ​mM 0.22 ​± ​0.06 ​mM <0.05
Enns et al., 2014 [48] Mixed cohort 58 59 Blood GSH: 808 ​± ​149 uM GSH: 900 ​± ​141 uM 0.0008
GSSG: 2.23 ​± ​1.84 uM GSSG: 1.17 ​± ​0.43 uM <0.0001
GSH/GSSG: 596 ​± ​424 GSH/GSSG: 881 ​± ​374 0.0002
Malondialdehyde Piccolo G et al., 1991 [46] CPEO 11 25 Plasma 0.38 ​± ​0.1 ​nmol/ml 0.3 ​± ​0.1 ​nmol/ml >0.05
Salmi et al., 2012 [44] Mixed cohort 10 4 Erythrocytes 1.8 ​± ​1.4 μmol/l 1.52 ​± ​1.52 μmol/l >0.05
GDF-15 Kalko et al., 2014 [72] TK2 mito myopathy 13 37 Serum 3562 ​± ​3973 ​pg/ml 380.5 ​pg/ml <0.0001
Koene et al., 2015 [73] MELAS (A3243G) 97 30 Serum 1525 ​pg/ml (CI 411–5691) 490 ​pg/ml (CI 272–1616) <0.001
Yatsuga et al., 2015 [22] Mixed cohort 48 146 Serum 2760 ​± ​2461 ​pg/ml 462.5 ​± ​141.0 ​pg/ml <0.001 0.997 (0.993–1.000) 97.90 95.20 710.0 ​pg/ml
Davis et al., 2016 [74] Mixed cohort 54 66 Serum 3956.0 ​pg/ml (IQR 2516.5–8676.8) 1183.5 ​pg/mL (IQR 1037.5–1475.3) <0.0001 0.941 (0.893–0.989) 77.8 (64.1–87.5) 95.5 (86.4–98.8) 2330 ​pg/ml
Lehtonen et al., 2016 [75] Mt translation defect 32 49 Serum Median: 3092 ​pg/ml (IQR 1844–4868) 328 ​pg/ml (IQR 235–474) <0.0001 Combined: 0.89 (0.84–0.94) 76.0 (64.5–85.4) 86.4 (77.4–92.8) 1014 ​pg/ml
mtDNA deletions 31 49 Serum Median: 1520 ​pg/ml (IQR 852–3403) 328 ​pg/ml (IQR 235–474) <0.0001
RC structure and assembly defect 8 49 Serum Median: 512 ​pg/ml (IQR 348–1178) 328 ​pg/ml (IQR 235–474) >0.05
Montero et al., 2016 [76] Mixed cohort 16 33 Serum 7593 ​± ​3870 pg/ul 350.3 ​± ​20.69 pg/ul <0.001 0.844 72.7 (49.8–89.2) 92.3 (81.5–97.9) 550 pg/ul
Ji et al., 2017 [77] Mixed cohort 42 50 Serum Median: 2521 ​pg/ml Median: 312 ​pg/ml <0.001 0.999 (0.963–1) Unclear Unclear 508 ​pg/ml
Koga et al., 2019 [26] Mixed cohort 11 n.a. Blood 3508.1 ​± ​1721.5 ​pg/ml Ref: <707.4 ​pg/ml n.r.
Tsygankova et al., 2019 [78] Mixed cohort 80 127 Plasma Pediatric: 4640 ​pg/ml (IQR 1896–14064) Pediatric: Median 907 ​pg/ml (IQR 575–1345) <0.00025 0.88 66 64 2400 ​pg/ml
Mixed cohort 42 127 Plasma Adult: 2553 ​pg/ml (IQR 1556–7615) Adult: Median 1985 ​pg/ml (IQR 627.5–2886) <0.0125 0.69 5000 ​pg/ml
Formichi et al., 2020 [79] Mixed cohort 70 55 Serum Median: 1456.5 ​pg/ml (IQR 818.3–2532.0) Median: 449.0 ​pg/ml (IQR 379.5–604.0) <0.0001 0.90 (0.84–0.95) 77 96 790.8 ​pg/ml
Lehtonen et al., 2021 [52] Mixed cohort 42 n.a. Serum 1585.4 ​± ​1526.7 ​pg/mla n.a. n.r. 1014 ​pg/ml
Maresca et al., 2020 [31] Mixed cohort 70 24 Serum 3712.74 ​± ​4004.98 ​pg/ml 459.39 ​± ​164.13 ​pg/ml <0.0001 0.96 (0.93–0.99) 89 (81–96) 92 (81–103) 711 ​pg/ml
Huddar et al. 2021 [50] Mixed cohort 30 36 Serum 4085.15 ​± ​2672.94 ​pg/ml 942.79 ​± ​478.8 ​pg/ml <0.0001 0.895 73.33 97.22 1883 ​pg/ml
Peñas et al., 2021 [58] Mixed cohort 60 56 Plasma 1757 ​pg/ml (range 329–5047) 588 ​pg/ml (range 153–1563) <0.001 0.87 (0.82–0.94) 71.93 (60.26–83.59) 92.86 (86.11–99.60) 975 ​pg/ml
Trifunov et al. 2021 [65] Mixed cohort 16 15 CSF 1882.5 ​± ​2637.3 ​pg/mla 158.5 ​± ​109.9 ​pg/mla n.r.
Varhaug et al., 2021 [80] Mixed cohort 26 n.a. Serum 2802.43 ​± ​2628.04 ​pg/ml Ref: <2330 ​pg/ml n.r.
Evangelisti et al., 2022 [9] MELAS (A3243G) 13 n.a. Serum 5223 ​± ​6587 ​pg/ml n.a. n.r.
FGF-21 Suomalainen et al., 2011 [25] Pediatric mixed cohort 40 25 Serum 1983 ​± ​1550 ​pg/ml 114 ​pg/ml (range 42–244) <0.0001 0.97 (0.94–0.99) 82.8 (71.3–91.1) combined adult & peds 91.7 (84.8–96.1) combined adult & peds 200 ​pg/ml
Adult mixed cohort 41 49 Serum 820 ​± ​1151 ​pg/ml 70 ​pg/ml (range 15–309) <0.0001 Unclear
Su et al., 2012 [81] Mito ataxia 23 53 Serum Median: 294.56 ​pg/ml
Range: 28–2600 ​pg/ml		 Median: 74.25 ​pg/ml
Range: 0–386 ​pg/ml		 <0.05
Davis et al., 2013 [23] Mixed cohort 54 66 Serum 617.4 ​pg/ml (IQR 281.7–1098.9) 82.5 ​pg/ml (IQR 51.3–133.8) <0.001 0.91 (0.86–0.97) 68.52 (54.31–80.09) 95.45 (86.44–98.82) 350 ​pg/ml
Salehi et al., 2013 [82] Mixed cohort 32 30 Serum 346.2 ​pg/ml (range 17.5–1707) 87.1 ​pg/ml (range 32–269.1) <0.05
Koene et al., 2014 [83] A3243G carriers 99 n.a. Serum Median: 263 ​pg/ml (IQR 140–523) n.a. n.r.
Koene et al., 2015 [73] MELAS (A3243G) 93 25 Serum Median: 263 ​pg/ml (IQR 142–534) 13 ​pg/ml (CI 0–444) <0.001
Yatsuga et al., 2015 [22] Mixed cohort 48 146 Serum 1244 ​± ​1502 ​pg/ml 156.0 ​± ​203.5 ​pg/ml <0.001 0.889 (0.837–0.962) 77.1 87.7 350.0 ​pg/ml
Davis et al., 2016 [74] Mixed cohort 54 66 Serum Unclear Unclear <0.0001 0.911 (0.855–0.968) 68.5 (54.3–80) 95.5 (86.4–98.8) 350 ​pg/ml
Lehtonen et al., 2016 [75] Mt translation defect 51 87 Serum Median: 675 ​pg/ml (IQR 437–1504) 66 ​pg/ml (IQR 48–104) <0.0001 Combined: 0.88 (0.84–0.93) 67.3 (57.8–75.8) 89.3 (83.2–93.8) 331 ​pg/ml
mtDNA deletions 51 87 Serum Median: 347 ​pg/ml (IQR 206–1062) 66 ​pg/ml (IQR 48–104) <0.0001
RC structure and assembly defect 11 87 Serum Median: 335 ​pg/ml (IQR 54–604) 66 ​pg/ml (IQR 48–104) <0.05
Montero et al., 2016 [76] Mixed cohort 16 33 Serum 966.5 ​± ​231 pg/ul 77.59 ​± ​10.3 pg/ul <0.001 0.792 59.1 (36.3–79.3) 96.2 (87–99.5) 300 pg/ul
Alban et al., 2017 [84] RC defect 14 n.a. Plasma Range: 331–10,700 ​pg/ml Ref: <250 ​pg/ml n.r.
Ji et al., 2017 [77] Mixed cohort 42 50 Serum Median: 4.5× higher than controls Unclear <0.001 0.935 (0.869–0.974)) 84.5 57.5 Unclear
Morovat et al., 2017 [68] Mixed cohort 104 27 Serum Median Z-score: 1.72 (CI 1.46–1.99) Median Z-score: 0.86 (CI -0.39–1.80) 0.0037 0.68 65 (55–74) 70 (50–86) Z-score: 1.35
20 (13–29) 93 (76–99) Z-score: 2.82
Koga et al., 2019 [26] Mixed cohort 11 n.a. Blood 1183.5 ​± ​507.1 ​pg/ml <281.0 ​pg/ml n.r.
Tsygankova et al., 2019 [78] Mixed cohort 122 127 Plasma Median: 253 ​pg/ml (IQR 69.5–889.5) Median: 43 ​pg/ml (IQR 30–102) <0.00025 0.78 51 76 400 ​pg/ml
Formichi et al., 2020 [79] Mixed cohort 70 55 Serum Median: 204 ​pg/ml (IQR 105.3–375.5) Median: 47.0 ​pg/ml (IQR 31.0–101.5)
Lehtonen et al., 2021 [52] Mixed cohort 42 n.a. Serum 482.7 ​± ​564.2 ​pg/mla n.a. n.r. 331 ​pg/ml
Maresca et al., 2020 [31] Mixed cohort 70 27 Serum 1765.76 ​± ​1374.26 ​pg/ml 990.61 ​± ​909.29 ​pg/ml <0.01 0.75 (0.65–0.85) 41 (30–53) 92 (81–103) 1947 ​pg/ml
Huddar et al., 2021 [50] Mixed cohort 30 36 Serum 2501.78 ​± ​2032.5 ​pg/ml 469.11 ​± ​453 ​pg/ml <0.0001 0.849 63.33 97.22 1475 ​pg/ml
Peñas et al., 2021 [58] Mixed cohort 60 56 Plasma 440 ​pg/ml (range 55–2908) 187 ​pg/ml (range 24–727) <0.001 0.77 (0.68–0.86) 61.67 (49.36–73.97) 83.93 (74.31–93.55) 300 ​pg/ml
Varhaug et al., 2021 [80] Mixed cohort 26 n.a. Serum 679.77 ​± ​630.07 ​pg/ml Ref: <350.0 ​pg/ml n.r.
Evangelisti et al., 2022 [9] MELAS (A3243G) 13 n.a. Serum 2376 ​± ​1810 ​pg/ml n.r. n.r.
Gelsolin Garcia-Bartolome et al., 2020 [60] Mixed cohort 9 9 Plasma mGSN relative to the mean control value:
2.33 ​± ​1.75a 		mGSN relative to the mean control value:
0.87 ​± ​0.69a
pGSN relative to the mean control value:
0.94 ​± ​0.21a 		pGSN relative to the mean control value:
1.02 ​± ​0.17a
mGSN/pGSN ratio:
2.36 ​± ​1.52		 mGSN/pGSN ratio:
0.81 ​± ​0.58
Peñas et al., 2021 [58] Mixed cohort 60 56 Plasma pGSN: 233 μg/ml (range 34–960) 432 μg/ml (range 213–830) <0.001 0.83 (0.75–0.92) 66.10 (54.02–78.18) 98.21 (94.75–100) 252 μg/ml
Neurofilament light-chain Sofou et al., 2019 [13] Mixed cohort 46 22 CSF 2011 (746–7194) 210 (127–379) <0.001 0.9
Varhaug et al., 2021 [80] Mixed cohort 26 n.a. Serum 25.70 ​± ​23.4 ​pg/ml n.a.
Circulating cell free mtDNA (ccf-mtDNA) Maresca et al., 2020 [31] Mixed cohort 123 35 Serum 123.00 ​± ​118.93 copies
MT-ND2/ul		 72.94 ​± ​47.74 copies
MT-ND2/ul		 n.r. 0.61 (0.50–0.72) 25 (17–34) 94 (86–102) 147 copies MT-ND2/ul
Trifunov et al. 2021 [65] Mixed cohort 25 18 CSF 61.1 ​± ​72.2 copies/uL 11.7 ​± ​9.5 copies/uL <0.0001
Evangelisti et al., 2022 [9] MELAS (A3243G) 14 n.a. Plasma 141 ​± ​136 cp/ul n.a. n.r.
Open in a new tab

ADMA: asymmetric dimethylarginine; AUC: area under the curve; C: controls; CI: confidence interval; CPEO: Chronic progressive external ophthalmoplegia; CSF: cerebral spinal fluid; GSH: Glutathione; GSN: gelsolin; GSSG: glutathione disulfide; IQR: interquartile range; MDDS: Mitochondrial DNA depletion syndrome; MELAS: Mitochondrial encephalomyopathy, lactic acidosis and stroke-like episodes; mt: mitochondria; n.a.: not applicable; n.r.: not reported; n.s.: not statistically significant; P: patient; peds: pediatric; RC: respiratory chain; ref: reference value; SD: standard deviation; SEM: standard error of the mean.

a

Statistical calculation performed by using raw data from the manuscript or supplemental information from the study made available online.

Lactate

Lactate plays a vital role in mitochondrial metabolism, particularly in the context of energy production and the maintenance of metabolic homeostasis. Lactate is formed by the reduction of pyruvate catalyzed by lactate dehydrogenase, classically during anaerobic respiration, and now appreciated to occur during aerobic respiration as well [1]. In the context of mitochondrial dysfunction, impaired passage through the electron transport chain (ETC) and subsequent decreased adenosine triphosphate (ATP) production leads to activation of glycolysis and therefore increased production of pyruvate. Altered NADH/NAD ​+ ​ratios can also inhibit the conversion of pyruvate to acetyl-CoA, leading to further buildup of pyruvate and therefore lactate as well. Although previously thought to simply be a waste product, there is growing evidence to suggest the role of lactate as a major energy source, gluconeogenic precursor, and as an important signaling molecule [2]. The role of hyperlactatemia and whether it is in some cases a beneficial compensatory mechanism in mitochondrial disease, for example shifting the oxyhemoglobin dissociation curve rightward to drive diffusion of O2 into tissues, remains a topic of debate [3].

Clinically, elevated lactate levels are known to have a variety of potential causes not limited to mitochondrial dysfunction. For example, spurious elevations may be caused by improper sample collection methods such as prolonged tourniquet application or difficult to obtain venous samples from a non-cooperative patient. Lactic acidosis is also a well-known sequela of hypoperfusion, commonly a result of cardiogenic, hemorrhagic, or septic shock [4]. Therefore, the sensitivity and specificity are supportive, but not ideal for a disease biomarker.

In this review, lactate as a sole biomarker was evaluated in 27 different studies, including 935 blood samples and 204 CSF samples. Among them, 13 reported statistical significance in at least one cohort, 5 reported non-significance in at least one of cohort, and 13 did not include measures of statistical significance in at least one cohort. Measures of sensitivity and specificity for lactate as a biomarker ranged from 15.1 to 100 ​% and 83.0–100 ​% respectively. Notably, while many studies included a mixed cohort of various mitochondrial diseases, 6 of the cohorts only included patients with MELAS [[5], [6], [7], [8], [9], [10]], in which lactic acidosis is a well-known feature of the disease, thereby suggesting some bias if trying to generalize for all mitochondrial diseases. In the studies that compared blood and CSF lactate levels, 2 reported similar p-values [11,12], 2 found CSF levels to have greater statistical significance [10,13], 1 reported blood levels to have greater statistical significance [14], and 3 did not report p-values for either method [[15], [16], [17]]. Interestingly, 3 studies were interventional experiments in which lactate levels were measured before and during exercise, and in 2 of the 3 studies, lactate levels compared to controls were not statistically different when measured at rest but became statistically significant when measured during peak exercise [[18], [19], [20]]. This suggests that the circumstances under which a lactate sample is drawn and whether pre- and post-exercise sample may improve the diagnostic utility remains a possibility that requires further investigation.

Pyruvate

As described above, pyruvate also plays an important role in mitochondrial energy metabolism, serving as a link between the glycolytic pathway and the tricarboxylic acid (TCA) cycle. Under normal circumstances pyruvate dehydrogenase converts pyruvate into acetyl-CoA. However, due to similar pathophysiological mechanisms as seen with elevated lactate levels in mitochondrial disease, pyruvate may accumulate as well. Therefore, pyruvate has also long been used as a biomarker of mitochondrial disease, although it can be elevated in other conditions as well, such as pyruvate dehydrogenase complex deficiency, pyruvate carboxylase deficiency, and biotinidase deficiency [21].

Pyruvate as a sole biomarker was evaluated in 6 different studies, including 185 blood samples and 17 CSF samples. Two of these studies reported statistical significance [12,22], 1 did not find statistical significance [23], and the remaining 3 did not report statistical significance [[24], [25], [26]]. Sensitivity and specificity of pyruvate as a biomarker ranged from 34.6 to 88.2 ​% and 81.2–87.2 ​% respectively, in the 3 studies that included them. The single study of CSF pyruvate levels did find statistical significance and high sensitivity and specificity 88.2 ​% and 100 ​% respectively [12].

Lactate:Pyruvate ratio

Given the close association between the two biomarkers, both lactate and pyruvate are frequently drawn together and a lactate to pyruvate (L:P) ratio generated as a more specific biomarker for mitochondrial disease. The L:P ratio is a well-established indicator of the cytoplasmic NADH/NAD ​+ ​redox ratio [21] and is clinically advantageous in that it can differentiate disorders of pyruvate metabolism like pyruvate dehydrogenase complex deficiency, in which patients would classically have proportional elevations of both biomarkers and therefore a normal ratio, as opposed to mitochondrial conditions in which lactate is comparatively higher than pyruvate [27].

The L:P ratio as a biomarker was evaluated in 8 different studies, including 195 blood samples and 17 CSF samples. Among these, all 3 studies that reported statistical significance for the L:P ratio [12,22,23]. Sensitivity and specificity of pyruvate as a biomarker ranged from 11.5 to 82.4 ​% and 68.7–100 ​% respectively. The L:R ratio from CSF samples was evaluated in one study, meeting statistical significance as a biomarker, and having a sensitivity of 76.5 ​% and specificity of 90.6 ​% [12]. Several studies in this review compared both lactate and pyruvate as individual markers as well as the L:R ratio, with the latter generally proving to be the most useful biomarker based on AUC, sensitivity, and specificity.

Creatine kinase

Creatine kinase (CK), also known as creatine phosphokinase (CPK), is an enzyme found in most tissues, but is particularly concentrated in skeletal and cardiac muscle and the brain. The primary role of CK is to replenish ATP by catalyzing the transfer of a phosphate group from phosphocreatine to adenosine diphosphate, thereby creating ATP. This system serves as an energy buffer during times of quick and intense bursts of energy consumption [28]. When damage occurs in any of these tissues, there is a quick release of CK into the bloodstream that is measurable. Although myopathy and therefore hyperCKemia is a common feature of some mitochondrial diseases, it is important to note that there are numerous other causes of hyperCKemia including several other genetic conditions like muscular dystrophies as well as non-genetic conditions like dehydration.

CK as a biomarker was evaluated in 6 different studies, including 196 blood samples. Among these, statistical significance was found in 1 study [22], not met in another [23], and the remaining not reporting any p-values. Sensitivity and specificity were relatively poor compared to the other biomarkers, with ranges of 22.2–34.1 ​% and 73.7–97.0 ​% respectively. It is perhaps unsurprising that the utility of CK as a mitochondrial disease biomarker is limited, likely owing to the fact that elevations in CK can be seen for so many other medical reasons and that many mitochondrial diseases do not result in increased CK levels.

Creatine

Creatine is closely related to CK, where creatine is stored in the form of phosphocreatine, serving as a short-term reservoir of phosphate groups to donate to adenosine diphosphate, catalyzed by the enzyme CK. Clinically, creatine has been proposed as a biomarker for mitochondrial disease due to its inverse relation with intracellular phosphocreatine:creatine ratio, thereby suggesting that elevated creatine levels may represent a low energetic state in tissues [29].

Creatine as a biomarker was evaluated in 3 different studies with 4 different cohorts, including 132 blood samples. Statistical significance was found in 3 of the 4 cohorts [30,31], with the 4th not reporting [32]. Despite the strong association, the sensitivity was relatively low at 39–60 ​% and specificity 83–93 ​% among these studies. So, while statistical significance was noted in these studies, the sensitivity and specificity were relatively weak compared to other biomarkers. Similar to CK as a biomarker, this may be related to the fact that creatine can also be released in any condition causing a myopathy.

Amino acids

Plasma amino acids are not a specific biomarker on their own, as this biochemical test typically includes approximately 40 different individual amino acids depending on the lab performing the test. Therefore, amino acid testing can be used to analyze several different metabolic processes at once. Mitochondria have an important role in the synthesis and catabolism of several amino acids, so under circumstances of mitochondrial dysfunction there may be certain markers/patterns of abnormal amino acid levels. Alanine, for example, is a reversible product of pyruvate transamination catalyzed by the enzyme alanine aminotransferase and is therefore often correlated with both lactate and pyruvate levels. Alanine can also help discriminate primary mitochondrial disorders from pyruvate dehydrogenase deficiency, in which there is greater elevations of pyruvate and alanine, as well as branched chain amino acids [27]. Notably, elevated alanine is the only amino acid biomarker used for the Nijmegen mitochondrial disease criteria scoring system with an alanine level greater than 450 μM being highly specific for mitochondrial dysfunction [33,34]. However, hyperalaninemia occurs in several other non-mitochondrial conditions including lactic acidosis in general, which has many causes as described above, as well as hyperinsulinism, chronic thiamine deficiency, and use of certain medication like valproic acid [35,36].

Other amino acids abnormalities may be more specific to certain kinds of mitochondrial diseases. For example, low arginine and citrulline levels have been reported in MELAS [37,38]. These amino acids often work together with citrulline being converted to argininosuccinate and then arginine by the enzymes argininosuccinate synthetase and arginiosuccinate lyase as part of the urea cycle. Although the specific mechanism by which this deficiency occurs in MELAS is not well understood at this time, it is recognized as being part of the pathophysiological mechanism by which the stroke-like episodes occur and explains why improved outcomes and even prevention of episodes can occur with supplementation [[37], [38], [39]]. MT-ATP6 mitochondrial disease can also present with low citrulline levels and on occasion may be found incidentally when trying to identify proximal urea cycle defects on newborn screening [40]. Similarly, the mechanism by which the hypocitrullinemia occurs is currently unknown. Other reports of amino acid abnormalities are highly variable with the occasional report suggesting an association of amino acids like proline, glycine, sarcosine, tyrosine, and the branched chain amino acids (leucine, valine, and isoleucine), but none a consistent finding among most studies [21,41]. Another drawback for plasma amino acid analysis is how easily environmental factors like a patient's nutritional status can affect the values, making them more difficult to interpret, especially when limited clinical information is available.

Fewer studies have evaluated the utility of urine amino acid levels, which clinically may be helpful to identify aminoaciduria as a sign of renal disease in patients with a mitochondrial disorder, but not helpful as a diagnostic biomarker. Other urine studies such as urine organic acids are also of limited utility as a diagnostic biomarker despite the inclusion of tricarboxylic acids, methylmalonic acid, and ethylmalonic acid in prior diagnostic criteria [33,41].

The use of plasma amino acids as a biomarker was evaluated in 10 different studies, including 119 blood samples and 33 CSF samples. Among them, at least 5 studies that reported p-values found at least 1 ​amino acid to have statistical significance compared to controls. There was some variability in the exact amino acids that were measured and reported in these studies and 4 of the studies were specifically done on patients with MELAS, which as described above has a known amino acid abnormality that may not necessarily be the case with other mitochondrial conditions. Among the MELAS cohorts, reported amino acid abnormalities included low citrulline, aminobutyrate, asparagine, cystine, glutamate, glutamine, glycine, lysine, methionine, phosphoserine, serine, and threonine, high alanine, valine, leucine, and isoleucine, and conflicting results for arginine and asymmetric dimethylarginine (ADMA) being high in one study and low in another [9,10,42,43]. The single study of CSF amino acids in MELAS patients also showed multiple abnormalities with statistical significance for elevated alanine, arginine, citrulline, glutamate, histidine, isoleucine, leucine, taurine, and valine, as well as low glutamine [10].

Among the studies that included a more diverse cohort of mitochondrial disease patients, elevated alanine in blood and CSF samples was a common finding [11,26,27,32], while less consistent abnormalities included reduced cystine, histidine, 1-methyl-histidine, lysine, and asparagine [27,44]. Nonetheless, with the exception of alanine and citrulline in the setting of specific mitochondrial diagnoses, most of the patterns observed seem to have limited diagnostic utility.

Glutathione

Glutathione is an intracellular thiol tripeptide molecule composed of the amino acids cysteine, glutamate, and glycine and serves multiple functions relating to mitochondrial health, most notably being a key defense against reactive oxygen and nitrogen species (RONS) created by the ETC. It has been shown that glutathione depletion occurs in ETC-deficient tissues [45], which may be due to over consumption as a result of increased RONS production, deficient glutathione production owing to its synthesis being an ATP-dependent process, or a combination of the two. As such, glutathione is sometimes supplemented in clinical practice, the efficacy of which is yet to be determined.

Glutathione as a biomarker was evaluated in 4 different studies, including 79 serum/plasma samples, 11 erythrocyte samples, and 10 lymphocyte samples. Among them, statistical significance for decreased glutathione was found in all cohorts including all sample types [[46], [47], [48]], except one [44]. Although most labs measure free glutathione levels, two studies evaluated all forms of glutathione with particular statistical significance noted for glutathione disulfide (GSSG) and protein-bound glutathione (GS-Pro) [47,48]. None of the studies performed statistical analysis with AUC, sensitivity, or specificity. Therefore, it is difficult to assess the true diagnostic utility of glutathione testing. There may still be clinical utility in testing glutathione levels though, as a determining factor for supplementation if deficient.

Malondialdehyde

Malondialdehyde (MDA) is a byproduct of lipid peroxidation that occurs when RONS react with polyunsaturated fatty acids in the cell membrane. Elevated MDA levels has been associated with aging and arterial stiffness [49], but the association with mitochondrial disease has only been weakly suggested. MDA as a biomarker was only evaluated in 2 different studies, including 11 blood samples and 10 erythrocyte samples and neither found statistical significance compared to controls [44,46]. It would therefore suggest that MDA is a poor diagnostic biomarker for mitochondrial disease.

GDF-15

The cytokine stress markers of mitochondrial dysfunction, growth differentiation factor 15 (GDF-15) and fibroblast growth factor 21 (FGF-21), have gained attention in recent years as biomarkers of mitochondrial disease. GDF-15 is a member of the transforming growth factor β superfamily that is produced as a response to cellular stress. It had previously been associated with numerous other conditions like cardiovascular disease, sepsis, cancer, and diabetes [50]. GDF-15 is known to have immunoregulatory and anti-inflammatory functions, but its exact role in mitochondrial diseases is not well understood [51].

GDF-15 as a biomarker was evaluated in 17 studies with 20 different cohorts, including 785 blood samples and 16 CSF samples. Among them, 14 out of 15 studies that reported p-values found statistical significance. The only exception was a cohort of 8 patients specifically with respiratory chain structure and assembly defects with a p-value of >0.05 compared to 49 controls, whereas the other cohorts in the same study with mitochondrial translation defects and mtDNA deletions each found strong associations with p-values <0.0001 [52]. It is unclear why a respiratory chain defect mitochondrial disease may not cause as significant of a rise in GDF-15 compared to other mitochondrial conditions in this one study. None of the other studies found in this review subcategorized subjects in this way to support this finding. The AUC for GDF-15 ranged from 0.69 to 0.99 and sensitivity and specificity among the 9 cohorts that included them had ranges of 66–97.9 ​% and 64–97 ​% respectively, making it one of the stronger diagnostic biomarkers.

FGF-21

FGF-21 is another hormone that has been associated with mitochondrial stress in general. It is produced in the liver and expressed in adipose tissue, skeletal muscle, and the pancreas, with pleiotropic actions to help regulate both glucose and lipid metabolism [53]. Interestingly, unlike the other fibroblast growth factors, FGF-21 does not promote cell proliferation or tumorigenesis [54]. Rather, in the fasting state, it stimulates gluconeogenesis, fatty acid oxidation, and ketogenesis, and in the fed state has an autocrine function of regulating the activity of Peroxisome Proliferator-Activated Receptor γ (PPAR-γ) [55]. It has been shown that the administration of FGF-21 in mice and monkeys has beneficial metabolic effects on obesity and diabetes, with increased fat utilization, energy expenditure, insulin sensitivity, and HDL-C levels, as well as decreased weight, blood glucose, triglycerides, insulin, and LDL-C levels [56,57]. Elevated FGF-21 levels have also been associated with several medical conditions in humans including obesity, metabolic syndrome, type 2 diabetes, coronary heart disease, non-alcoholic fatty liver, and chronic kidney disease [55]. The association with mitochondrial disease had not been made until 2011, followed by several more studies supporting its use as a mitochondrial disease biomarker.

FGF-21 as a biomarker was evaluated in 22 studies with 25 different cohorts, including 1217 blood samples. Among them, all 18 cohorts that reported p-values found statistical significance when compared to controls. Additionally, AUC analysis of 12 reporting cohorts had a range of 0.75–0.97 and the reported sensitivity and specificities were 20–82.8 ​% and 57.5–97.2 ​% respectively. The sensitivity and specificity for FGF-21 was perhaps lower than one might expect, which may be related to the variability of threshold levels used in these studies. While most used a number between 300 and 350 ​pg/ml as a cutoff, others were as low as 200 ​pg/ml or as high as 1947 ​pg/ml. This may however be due to different testing techniques, suggesting a greater need for standardization of FGF-21 lab testing methods. Similar to GDF-15, the single study that subcategorized mitochondrial diseases by the type of defect, found that respiratory chain defective disorders had the weakest, although still statistically significant in this case, association with FGF-21 levels [52].

Gelsolin

Gelsolin (GSN) is a protein encoded by the GSN gene located on chromosome 9, with two primary isoforms depending on alternative transcription initiation and mRNA processing, creating cytoplasmic GSN (cGSN) and plasma GSN (pGSN) [58]. The primary known function of GSN is to regulate actin filaments involved in the cellular cytoskeleton, but more recently has been found to also be a marker of oxidative phosphorylation (OXPHOS) dysfunction. Cell models of respiratory chain dysfunction had been found to downregulate pGSN and upregulate cGSN into the mitochondrial outer membrane (mGSN) in order to bind to voltage-dependent anion channels and prevent apoptotic cell death [59,60]. As such, a mGSN:pGSN protein ratio had been proposed as a hallmark of OXPHOS dysfunction [60]. In clinical use, one study had found that decreased levels of pGSN alone can be an effective biomarker of mitochondrial disease with statistical significance found in 1 study with 60 patients, but its greatest benefit was seen when used alongside GDF-15 levels with a reported sensitivity of 89.29 ​% and specificity of 92.86 ​% [58]. Nonetheless, additional studies are needed to validate the utility of GSN as a mitochondrial biomarker.

Neurofilament light-chain

Neurofilament light-chain (NF-L) is one of three subunits that form neurofilaments, which are an important structural component of neurons. CSF NF-L levels have already been associated with other neurological conditions including Alzheimer's disease and traumatic brain injury [61,62]. Given the neurodegenerative course of some mitochondrial conditions, it has also been proposed as a potential biomarker of mitochondrial disease. Only 2 studies have examined NF-L levels in patients with mitochondrial disease, including one with CSF and the other with plasma levels tested. Statistical significance was found with increased CSF NF-L levels, but the authors do note that it was found to be of greater value among those who have multisystem disease presentation, versus those that primarily have a myopathy presentation [13]. Therefore, NF-L would perhaps not be the ideal biomarker except in rare cases with a primary neurological presentation.

Circulating cell-free mtDNA

Circulating cell-free mtDNA (ccfmtDNA) refers to fragments of mitochondrial DNA that can be found in the bloodstream or other bodily fluids that are not associated with intact mitochondria or cells. The simplistic explanation for the presence of ccfmtDNA is the result of any trauma or stress to the cell that would induce cellular apoptosis/death, so it would seem sensible that abnormal levels would be seen in mitochondrial disease. Elevated ccfmtDNA levels in plasma have thus far been associated with medical conditions such as necrosis, acute respiratory distress syndrome, tumors, and inflammation [63,64].

ccfmtDNA as a biomarker was evaluated in only 3 studies, including 137 blood samples and 25 CSF samples. Among them, a p-value was only reported in the single study that examined CSF levels of ccfmtDNA, in which strong statistical significance was found [65]. The utility of ccfmtDNA was less convincing in plasma though with one of the studies reporting a sensitivity of 25 ​% and specificity of 94 ​%, suggesting that most patients with mitochondrial disease would not be found to have abnormal ccfmtDNA levels in plasma [31]. Nonetheless, additional studies would be necessary to more fully assess the utility of such testing. If truly only remarkable levels can be seen on CSF studies, then measuring ccfmtDNA would be less useful than other biomarkers given the invasiveness of obtaining a CSF sample.

In this review, information was gathered from decades of research to present the current spectrum of mitochondrial disease biomarkers that have been studied to-date. Presented here are 13 different biochemical tests for mitochondrial disease, obtainable through either blood or CSF samples, some of which have been studied rather extensively and others that are fairly new with limited data. While some appear to have greater utility with higher sensitivity and specificity when used in large comparison studies, none of them is a perfect biomarker for mitochondrial disease, which is expected given the complexity and variability in which mitochondrial diseases can manifest. As such, some biomarkers may be of greater utility in certain mitochondrial conditions, like lactate in MELAS. The evidence suggest that the cellular stress biomarkers GDF-15 and FGF-21 might have the greatest overall utility for mitochondrial diseases in general, with GDF-15 surpassing FGF-21 slightly in most comparison studies.

Novel approaches to identifying additional biomarkers of mitochondrial disease have been suggested, including metabolomic and lipidomic testing, with insufficient data at this time. The challenge of such testing is the extensive amount of data that is gathered, so more experience interpreting such results, potentially with the use of artificial intelligence, may help solve this issue and identify more complex patterns involving multiple metabolites at once. Nonetheless, the diagnosis of mitochondrial disease based on biochemical testing remains a challenge, especially in light of other conditions that may cause secondary mitochondrial dysfunction. As such, molecular testing will remain the standard for diagnosing mitochondrial conditions, though biochemical testing can still be useful when presented with unclear sequencing results or in the clinical management and monitoring of those with an already diagnosed primary mitochondrial disease.

Author contributions

As the sole author, Brian J. Shayota was responsible for the entirety of this review article, including study design, material preparation, data collection, analysis, and writing of the manuscript.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

  • 1.Bendahan D., Chatel X.B., Jue T. Comparative NMR and NIRS analysis of oxygen-dependent metabolism in exercising finger flexor muscles. Am J Physiol Regul Integr Comp Physiol. 2017;313:740–753. doi: 10.1152/ajpregu.00203.2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Brooks G.A. The science and translation of lactate shuttle theory. Cell Metab. 2018:757–785. doi: 10.1016/j.cmet.2018.03.008. [DOI] [PubMed] [Google Scholar]
  • 3.Glancy B., Kane D.A., Kavazis A.N., Goodwin M.L., Willis W.T., Gladden L.B. Mitochondrial lactate metabolism: history and implications for exercise and disease. J Physiol. 2021;599:863–888. doi: 10.1113/JP278930. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Andersen L.W., Mackenhauer J., Roberts J.C., Berg K.M., Cocchi M.N., Donnino M.W. Etiology and therapeutic approach to elevated lactate levels. Mayo Clin Proc. 2013:1127–1140. doi: 10.1016/j.mayocp.2013.06.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Kaufmann P., Engelstad K., Wei B.Y., Kulikova R., Oskoui M., Sproule D.M., et al. Natural history of MELAS associated with mitochondrialDNA m.3243A>G genotype. Neurology. 2011;77(22):1965–1971. doi: 10.1212/WNL.0b013e31823a0c7f. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Zhang J., Guo J., Fang W., Jun Q., Shi K. Clinical features of MELAS and its relation with A3243G gene point mutation [Internet] Int J Clin Exp Pathol. 2015;8(10):13411–13415. [PMC free article] [PubMed] [Google Scholar]
  • 7.Xia C.Y., Liu Y., Liu H., Zhang Y.C., Ma Y.N., Qi Y. Clinical and molecular characteristics in 100 Chinese pediatric patients with m.3243a>G mutation in mitochondrial DNA. Chin Med J. 2016;129:1945–1949. doi: 10.4103/0366-6999.187845. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Wang W., Zhuang Q., Ji K., Wen B., Lin P., Zhao Y., et al. Identification of miRNA, lncRNA and mRNA-associated ceRNA networks and potential biomarker for MELAS with mitochondrial DNA A3243G mutation. Sci Rep. 2017;7:41639. doi: 10.1038/srep41639. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Evangelisti S., Gramegna L.L., La Morgia C., Di Vito L., Maresca A., Talozzi L., et al. Molecular biomarkers correlate with brain grey and white matter changes in patients with mitochondrial m.3243A > G mutation. Mol Genet Metabol. 2022;135:72–81. doi: 10.1016/j.ymgme.2021.11.012. [DOI] [PubMed] [Google Scholar]
  • 10.Guerrero-Molina M.P., Morales-Conejo M., Delmiro A., Morán M., Domínguez-González C., Arranz-Canales E., et al. Elevated glutamate and decreased glutamine levels in the cerebrospinal fluid of patients with MELAS syndrome. J Neurol. 2022;269:3238–3248. doi: 10.1007/s00415-021-10942-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Magner M., Szentiványi K., Švandová I., Ješina P., Tesařová M., Honzík T., et al. Elevated CSF-lactate is a reliable marker of mitochondrial disorders in children even after brief seizures. Eur J Paediatr Neurol. 2011;15:101–108. doi: 10.1016/j.ejpn.2010.10.001. [DOI] [PubMed] [Google Scholar]
  • 12.Yamada K., Toribe Y., Yanagihara K., Mano T., Akagi M., Suzuki Y. Diagnostic accuracy of blood and CSF lactate in identifying children with mitochondrial diseases affecting the central nervous system. Brain Dev. 2012;34:92–97. doi: 10.1016/j.braindev.2011.08.004. [DOI] [PubMed] [Google Scholar]
  • 13.Sofou K., Shahim P., Tulinius M., Blennow K., Zetterberg H., Mattsson N., et al. Cerebrospinal fluid neurofilament light is associated with survival in mitochondrial disease patients. Mitochondrion. 2019;46:228–235. doi: 10.1016/j.mito.2018.07.002. [DOI] [PubMed] [Google Scholar]
  • 14.Lunsing R.J., Strating K., de Koning T.J., Sijens P.E. Diagnostic value of MRS-quantified brain tissue lactate level in identifying children with mitochondrial disorders. Eur Radiol. 2017;27:976–984. doi: 10.1007/s00330-016-4454-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Sofou K., De Coo I.F.M., Isohanni P., Ostergaard E., Naess K., De Meirleir L. A multicenter study on Leigh syndrome: disease course and predictors of survival. Orphanet J Rare Dis. 2014;9:52. doi: 10.1186/1750-1172-9-52. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Yu X.L., Yan C.Z., Ji K.Q., Lin P.F., Xu X.B., Dai T.J., et al. Clinical, neuroimaging, and pathological analyses of 13 Chinese leigh syndrome patients with mitochondrial DNA mutations. Chin Med J (Engl) 2018:2705–2712. doi: 10.4103/0366-6999.245265. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Keshavan N., Abdenur J., Anderson G., Assouline Z., Barcia G., Bouhikbar L., et al. The natural history of infantile mitochondrial DNA depletion syndrome due to RRM2B deficiency. Genet Med. 2020;22(1):199–209. doi: 10.1038/s41436-019-0613-z. [DOI] [PubMed] [Google Scholar]
  • 18.Heinicke K., Taivassalo T., Wyrick P., Wood H., Babb T.G., Haller R.G. Exertional dyspnea in mitochondrial myopathy: clinical features and physiological mechanisms. Am J Physiol Regul Integr Comp Physiol. 2011;301:873–884. doi: 10.1152/ajpregu.00001.2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Jeppesen T.D., Orngreen M.C., Van Hall G., Vissing J. Lactate metabolism during exercise in patients with mitochondrial myopathy. Neuromuscul Disord. 2013;23:629–636. doi: 10.1016/j.nmd.2013.05.007. [DOI] [PubMed] [Google Scholar]
  • 20.Zhao Y., Zhang Y., Qi M., Ping F., Zhang H., Xu L., et al. The role of lactate exercise test and fasting plasma C-peptide levels in the diagnosis of mitochondrial diabetes: analysis of clinical characteristics of 12 patients with mitochondrial diabetes in a single center with long-term follow-up. Front Endocrinol. 2022;13:835570. doi: 10.3389/fendo.2022.835570. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Haas R.H., Parikh S., Falk M.J., Saneto R.P., Wolf N.I., Darin N., et al. The in-depth evaluation of suspected mitochondrial disease. Mol Genet Metabol. 2008;94:16–37. doi: 10.1016/j.ymgme.2007.11.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Yatsuga S., Fujita Y., Ishii A., Fukumoto Y., Arahata H., Kakuma T., et al. Growth differentiation factor 15 as a useful biomarker for mitochondrial disorders. Ann Neurol. 2015;78:814–823. doi: 10.1002/ana.24506. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Davis R.L., Liang C., Edema-Hildebrand F., Riley C., Needham M., Sue C.M. Fibroblast growth factor 21 is a sensitive biomarker of mitochondrial disease. Neurology. 2013;81:1819–1826. doi: 10.1212/01.wnl.0000436068.43384.ef. [DOI] [PubMed] [Google Scholar]
  • 24.Debray F.G., Mitchell G.A., Allard P., Robinson B.H., Hanley J.A., Lambert M. Diagnostic accuracy of blood lactate-to-pyruvate molar ratio in the differential diagnosis of congenital lactic acidosis. Clin Chem. 2007;53:916–921. doi: 10.1373/clinchem.2006.081166. [DOI] [PubMed] [Google Scholar]
  • 25.Suomalainen A., Elo J.M., Pietiläinen K.H., Hakonen A.H., Sevastianova K., Korpela M., et al. FGF-21 as a biomarker for muscle-manifesting mitochondrial respiratory chain deficiencies: a diagnostic study. Lancet Neurol. 2011;10:806–818. doi: 10.1016/S1474-4422(11)70155-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Koga Y., Povalko N., Inoue E., Nashiki K., Tanaka M. Biomarkers and clinical rating scales for sodium pyruvate therapy in patients with mitochondrial disease. Mitochondrion. 2019;48:11–15. doi: 10.1016/j.mito.2019.02.001. [DOI] [PubMed] [Google Scholar]
  • 27.Clarke C., Xiao R., Place E., Zhang Z., Sondheimer N., Bennett M., et al. Mitochondrial respiratory chain disease discrimination by retrospective cohort analysis of blood metabolites. Mol Genet Metabol. 2013;110:145–152. doi: 10.1016/j.ymgme.2013.07.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Schlattner U., Tokarska-Schlattner M., Wallimann T. Mitochondrial creatine kinase in human health and disease. Biochim Biophys Acta. 2006:164–180. doi: 10.1016/j.bbadis.2005.09.004. [DOI] [PubMed] [Google Scholar]
  • 29.Ostojic S.M. Plasma creatine as a marker of mitochondrial dysfunction. Med Hypotheses. 2018:52–53. doi: 10.1016/j.mehy.2018.02.022. [DOI] [PubMed] [Google Scholar]
  • 30.Pajares S., Arias A., García-Villoria J., Briones P., Ribes A. Role of creatine as biomarker of mitochondrial diseases. Mol Genet Metabol. 2013;108:119–124. doi: 10.1016/j.ymgme.2012.11.283. [DOI] [PubMed] [Google Scholar]
  • 31.Maresca A., Del Dotto V., Romagnoli M., La Morgia C., Di Vito L., Capristo M., et al. Expanding and validating the biomarkers for mitochondrial diseases. J Mol Med. 2020;98:1467–1478. doi: 10.1007/s00109-020-01967-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Shaham O., Slate N.G., Goldberger O., Xu Q., Ramanathan A., Souza A.L., et al. A plasma signature of human mitochondrial disease revealed through metabolic profiling of spent media from cultured muscle cells. Proc Natl Acad Sci U S A. 2010;107:1571–1575. doi: 10.1073/pnas.0906039107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Morava E., Van Den Heuvel L., Hol F., de Vries M.C., Hogeveen M., Rodenburg R.J., et al. Mitochondrial disease criteria Diagnostic applications in children. Neurology. 2006;67(10):1823–1826. doi: 10.1212/01.wnl.0000244435.27645.54. [DOI] [PubMed] [Google Scholar]
  • 34.Wolf N.I., Smeitink J.A.M. Mitochondrial disorders A proposal for consensus diagnostic criteria in infants and children. Neurology. 2002;59(9):1402–1405. doi: 10.1212/01.wnl.0000031795.91814.d8. [DOI] [PubMed] [Google Scholar]
  • 35.Thauvin-Robinet C., Faivre L., Barbier M.L., Chevret L., Bourgeois J., Netter J.C., et al. Severe lactic acidosis and acute thiamin defciency: a report of 11 neonates with unsupplemented total parenteral nutrition. J Inherit Metab Dis. 2004;27(5):700–704. doi: 10.1023/b:boli.0000043017.90837.93. [DOI] [PubMed] [Google Scholar]
  • 36.Thabet H., Brahmi N., Amamou M., Ben Salah N., Hédhili A. Hyperlactatemia and hyperammonemia as secondary effects of valproic acid poisoning. Am J Emerg Med. 2000;18(4):508. doi: 10.1053/ajem.2000.3981. [DOI] [PubMed] [Google Scholar]
  • 37.Koga Y., Akita Y., Junko N., Yatsuga S., Povalko N., Fukiyama R., et al. Endothelial dysfunction in MELAS improved by l-arginine supplementation. Neurology. 2006;66(11):1766–1769. doi: 10.1212/01.wnl.0000220197.36849.1e. [DOI] [PubMed] [Google Scholar]
  • 38.Naini A., Kaufmann P., Shanske S., Engelstad K., De Vivo D.C., Schon E.A. Hypocitrullinemia in patients with MELAS: an insight into the “MELAS paradox”. J Neurol Sci. 2005;229-230:187–193. doi: 10.1016/j.jns.2004.11.026. [DOI] [PubMed] [Google Scholar]
  • 39.Argudo J.M., Astudillo Moncayo O.M., Insuasti W., Garofalo G., Aguirre A.S., Encalada S., et al. Arginine for the treatment of mitochondrial encephalopathy, lactic acidosis, and stroke-like episodes: a systematic review. Cureus. 2022;14(12) doi: 10.7759/cureus.32709. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Tise C.G., Verscaj C.P., Mendelsohn B.A., Woods J., Lee C.U., Enns G.M., et al. MT-ATP6 mitochondrial disease identified by newborn screening reveals a distinct biochemical phenotype. Am J Med Genet A. 2023;191(6):1492–1501. doi: 10.1002/ajmg.a.63159. [DOI] [PubMed] [Google Scholar]
  • 41.Shatla H.M., Tomoum H.Y., Elsayed S.M., Elagouza I.A., Shatla R.H., Mohsen M.M., et al. Role of plasma amino acids and urinary organic acids in diagnosis of mitochondrial diseases in children. Pediatr Neurol. 2014;51:820–825. doi: 10.1016/j.pediatrneurol.2014.08.009. [DOI] [PubMed] [Google Scholar]
  • 42.El-Hattab A.W., Emrick L.T., Hsu J.W., Chanprasert S., Almannai M., Craigen W.J., et al. Impaired nitric oxide production in children with MELAS syndrome and the effect of arginine and citrulline supplementation. Mol Genet Metabol. 2016;117:407–412. doi: 10.1016/j.ymgme.2016.01.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.El-Hattab A.W., Hsu J.W., Emrick L.T., Wong L.J., Craigen W.J., Jahoor F., et al. Restoration of impaired nitric oxide production in MELAS syndrome with citrulline and arginine supplementation. Mol Genet Metabol. 2012;105:607–614. doi: 10.1016/j.ymgme.2012.01.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Salmi H., Leonard J.V., Rahman S., Lapatto R. Plasma thiol status is altered in children with mitochondrial diseases. Scand J Clin Lab Invest. 2012;72:152–157. doi: 10.3109/00365513.2011.646299. [DOI] [PubMed] [Google Scholar]
  • 45.Hargreaves I.P., Sheena Y., Land J.M., Heales S.J. Glutathione deficiency in patients with mitochondrial disease: implications for pathogenesis and treatment. J Inherit Metab Dis. 2005;28:81–88. doi: 10.1007/s10545-005-4160-1. [DOI] [PubMed] [Google Scholar]
  • 46.Piccolo G., Banfi P., Azan G., Rizzuto R., Bisson R., Sandoná D., et al. Biological markers of oxidative stress in mitochondrial myopathies with progressive external ophthalmoplegia. J Neurol Sci. 1991;105(1):57–60. doi: 10.1016/0022-510x(91)90118-q. [DOI] [PubMed] [Google Scholar]
  • 47.Pastore A., Petrillo S., Tozzi G., Carrozzo R., Martinelli D., Dionisi-Vici C., et al. Glutathione: a redox signature in monitoring EPI-743 therapy in children with mitochondrial encephalomyopathies. Mol Genet Metabol. 2013;109:208–214. doi: 10.1016/j.ymgme.2013.03.011. [DOI] [PubMed] [Google Scholar]
  • 48.Enns G.M., Moore T., Le A., Atkuri K., Shah M.K., Cusmano-Ozog K., et al. Degree of glutathione deficiency and redox imbalance depend on subtype of mitochondrial disease and clinical status. PLoS One. 2014;9(6) doi: 10.1371/journal.pone.0100001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Ito F., Sono Y., Ito T. Measurement and clinical significance of lipid peroxidation as a biomarker of oxidative stress: oxidative stress in diabetes, atherosclerosis, and chronic inflammation. Antioxidants (Basel) 2019;8(3):72 doi: 10.3390/antiox8030072. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Huddar A., Govindaraj P., Chiplunkar S., Deepha S., Jessiena Ponmalar J.N., Philip M., et al. Serum fibroblast growth factor 21 and growth differentiation factor 15: two sensitive biomarkers in the diagnosis of mitochondrial disorders. Mitochondrion. 2021;60:170–177. doi: 10.1016/j.mito.2021.08.011. [DOI] [PubMed] [Google Scholar]
  • 51.Pence B.D. Growth differentiation factor-15 in immunity and aging. Front Aging. 2022;3:837575. doi: 10.3389/fragi.2022.837575. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Lehtonen J.M., Auranen M., Darin N., Sofou K., Bindoff L., Hikmat O., et al. Diagnostic value of serum biomarkers FGF21 and GDF15 compared to muscle sample in mitochondrial disease. J Inherit Metab Dis. 2021;44:469–480. doi: 10.1002/jimd.12307. [DOI] [PubMed] [Google Scholar]
  • 53.Tacer K.F., Bookout A.L., Ding X., Kurosu H., John G.B., Wang L., et al. Research resource: comprehensive expression atlas of the fibroblast growth factor system in adult mouse. Mol Endocrinol. 2010;24:2050–2064. doi: 10.1210/me.2010-0142. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Kharitonenkov A., Shiyanova T.L., Koester A., Ford A.M., Micanovic R., Galbreath E.J., et al. FGF-21 as a novel metabolic regulator. J Clin Invest. 2005;115:1627–1635. doi: 10.1172/JCI23606. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Woo Y.C., Xu A., Wang Y., Lam K.S. Fibroblast Growth Factor 21 as an emerging metabolic regulator: clinical perspectives. Clin Endocrinol. 2013:489–496. doi: 10.1111/cen.12095. [DOI] [PubMed] [Google Scholar]
  • 56.Coskun T., Bina H.A., Schneider M.A., Dunbar J.D., Hu C.C., Chen Y., et al. Fibroblast growth factor 21 corrects obesity in mice. Endocrinology. 2008;149:6018–6027. doi: 10.1210/en.2008-0816. [DOI] [PubMed] [Google Scholar]
  • 57.Kharitonenkov A., Wroblewski V.J., Koester A., Chen Y.F., Clutinger C.K., Tigno X.T., et al. The metabolic state of diabetic monkeys is regulated by fibroblast growth factor-21. Endocrinology. 2007;148:774–781. doi: 10.1210/en.2006-1168. [DOI] [PubMed] [Google Scholar]
  • 58.Peñas A., Fernández-De la Torre M., Laine-Menéndez S., Lora D., Illescas M., García-Bartolomé A., et al. Plasma gelsolin reinforces the diagnostic value of FGF-21 and GDF-15 for mitochondrial disorders. Int J Mol Sci. 2021;22(12):6396. doi: 10.3390/ijms22126396. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Garćıa-Bartolomé A., Peñas A., Maŕın-Buera L., Lobo-Jarne T., Pérez-Pérez R., Morán M., et al. Respiratory chain enzyme deficiency induces mitochondrial location of actin-binding gelsolin to modulate the oligomerization of VDAC complexes and cell survival. Hum Mol Genet. 2017;26:2493–2506. doi: 10.1093/hmg/ddx144. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.García-Bartolomé A., Peñas A., Illescas M., Bermejo V., López-Calcerrada S., Pérez-Pérez R., et al. Altered expression ratio of actin-binding gelsolin isoforms is a novel hallmark of mitochondrial OXPHOS dysfunction. Cells. 2020;9(9):1922. doi: 10.3390/cells9091922. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Zetterberg H., Skillbäck T., Mattsson N., Trojanowski J.Q., Portelius E., Shaw L.M., et al. Association of cerebrospinal fluid neurofilament light concentration with Alzheimer disease progression. JAMA Neurol. 2016;73:60–67. doi: 10.1001/jamaneurol.2015.3037. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62.Shahim P., Gren M., Liman V., Andreasson U., Norgren N., Tegner Y., et al. Serum neurofilament light protein predicts clinical outcome in traumatic brain injury. Sci Rep. 2016;6:36791. doi: 10.1038/srep36791. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 63.McClintock C.R., Mulholland N., Krasnodembskaya A.D. Biomarkers of mitochondrial dysfunction in acute respiratory distress syndrome: a systematic review and meta-analysis. Front Med (Lausanne) 2022;9:1011819. doi: 10.3389/fmed.2022.1011819. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 64.Rahat B., Ali T., Sapehia D., Mahajan A., Kaur J. Circulating cell-free nucleic acids as epigenetic biomarkers in precision medicine. Front Genet. 2020;11:844. doi: 10.3389/fgene.2020.00844. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65.Trifunov S., Paredes-Fuentes A.J., Badosa C., Codina A., Montoya J., Ruiz-Pesini E., et al. Circulating cell-free mitochondrial DNA in cerebrospinal fluid as a biomarker for mitochondrial diseases. Clin Chem. 2021;67:1113–1121. doi: 10.1093/clinchem/hvab091. [DOI] [PubMed] [Google Scholar]
  • 66.Minuz P., Fava C., Vattemi G., Arcaro G., Riccadonna M., Tonin P., et al. Endothelial dysfunction and increased oxidative stress in mitochondrial diseases. Clin Sci. 2012;122:289–297. doi: 10.1042/CS20110199. [DOI] [PubMed] [Google Scholar]
  • 67.Feldman A.G., Sokol R.J., Hardison R.M., Alonso E.M., Squires R.H., Narkewicz M.R. Lactate and lactate: pyruvate ratio in the diagnosis and outcomes of pediatric acute liver failure. J Pediatr. 2017;182:217–222.e3. doi: 10.1016/j.jpeds.2016.12.031. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68.Morovat A., Weerasinghe G., Nesbitt V., Hofer M., Agnew T., Quaghebeur G., et al. Use of FGF-21 as a biomarker of mitochondrial disease in clinical practice. J Clin Med. 2017;6(8):80. doi: 10.3390/jcm6080080. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 69.Wu T., He F., Xiao N., Han Y., Yang L., Peng J. Phenotype-genotype analysis based on molecular classification in 135 children with mitochondrial disease. Pediatr Neurol. 2022;132:11–18. doi: 10.1016/j.pediatrneurol.2022.04.013. [DOI] [PubMed] [Google Scholar]
  • 70.Kärppä M., Mahjneh I., Karttunen A., Tolonen U., Majamaa K. Muscle computed tomography patterns in patients with the mitochondrial DNA mutation 3243A>G. J Neurol. 2004;251:556–563. doi: 10.1007/s00415-004-0363-x. [DOI] [PubMed] [Google Scholar]
  • 71.Crippa B.L., Leon E., Calhoun A., Lowichik A., Pasquali M., Longo N. Biochemical abnormalities in Pearson syndrome. Am J Med Genet A. 2015;167:621–628. doi: 10.1002/ajmg.a.36939. [DOI] [PubMed] [Google Scholar]
  • 72.Kalko S.G., Paco S., Jou C., Rodríguez M.A., Meznaric M., Rogac M., et al. 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 myopathies. BMC Genom. 2014;15:91. doi: 10.1186/1471-2164-15-91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 73.Koene S., de Laat P., van Tienoven D.H., Weijers G., Vriens D., Sweep F.C., et al. Serum GDF15 levels correlate to mitochondrial disease severity and myocardial strain, but not to disease progression in adult m.3243A>G carriers. JIMD Rep. 2015;24:69–81. doi: 10.1007/8904_2015_436. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 74.Davis R.L., Liang C., Sue C.M. A comparison of current serum biomarkers as diagnostic indicators of mitochondrial diseases. Neurology. 2016;86:2010–2015. doi: 10.1212/WNL.0000000000002705. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 75.Lehtonen J.M., Forsstrom S., Bottani E., Viscomi C., Baris O.R., Isoniemi H., et al. FGF21 is a biomarker for mitochondrial translation and mtDNA maintenance disorders. Neurology. 2016;87(22):2290–2299. doi: 10.1212/WNL.0000000000003374. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 76.Montero R., Yubero D., Villarroya J., Henares D., Jou C., Rodríguez M.A., et al. GDF-15 is elevated in children with mitochondrial diseases and is induced by mitochondrial dysfunction. PLoS One. 2016;11(2) doi: 10.1371/journal.pone.0148709. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 77.Ji X., Zhao L., Ji K., Zhao Y., Li W., Zhang R., et al. Growth differentiation factor 15 is a novel diagnostic biomarker of mitochondrial diseases. Mol Neurobiol. 2017;54:8110–8116. doi: 10.1007/s12035-016-0283-7. [DOI] [PubMed] [Google Scholar]
  • 78.Tsygankova P.G., Itkis Y.S., Krylova T.D., Kurkina M.V., Bychkov I.O., Ilyushkina A.A., et al. Plasma FGF-21 and GDF-15 are elevated in different inherited metabolic diseases and are not diagnostic for mitochondrial disorders. J Inherit Metab Dis. 2019;42:918–933. doi: 10.1002/jimd.12142. [DOI] [PubMed] [Google Scholar]
  • 79.Formichi P., Cardone N., Taglia I., Cardaioli E., Salvatore S., Gerfo A.L., et al. Fibroblast growth factor 21 and grow differentiation factor 15 are sensitive biomarkers of mitochondrial diseases due to mitochondrial transfer-RNA mutations and mitochondrial DNA deletions. Neurol Sci. 2020;41:3653–3662. doi: 10.1007/s10072-020-04422-5. [DOI] [PubMed] [Google Scholar]
  • 80.Varhaug K.N., Hikmat O., Nakkestad H.L., Vedeler C.A., Bindoff L.A. Serum biomarkers in primary mitochondrial disorders. Brain Commun. 2021;3(1):fcaa222. doi: 10.1093/braincomms/fcaa222. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 81.Su S.L., Wang W.F., Wu S.L., Wu H.M., Chang J.C., Huang C.S., et al. FGF21 in ataxia patients with spinocerebellar atrophy and mitochondrial disease. Clin Chim Acta. 2012;414:225–227. doi: 10.1016/j.cca.2012.09.019. [DOI] [PubMed] [Google Scholar]
  • 82.Salehi M.H., Kamalidehghan B., Houshmand M., Aryani O., Sadeghizadeh M., Mossalaeie M.M. Association of fibroblast growth factor (FGF-21) as a biomarker with primary mitochondrial disorders, but not with secondary mitochondrial disorders (Friedreich Ataxia) Mol Biol Rep. 2013;40:6495–6499. doi: 10.1007/s11033-013-2767-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 83.Koene S., De Laat P., Van Tienoven D.H., Vriens D., Brandt A.M., Sweep F.C., et al. Serum FGF21 levels in adult m.3243A.G carriers: clinical implications. Neurology. 2014;83(2):125–133. doi: 10.1212/WNL.0000000000000578. [DOI] [PubMed] [Google Scholar]
  • 84.Alban C., Fatale E., Joulani A., Ilin P., Saada A. The relationship between mitochondrial respiratory chain activities in muscle and metabolites in plasma and urine: a retrospective study. J Clin Med. 2017;6(3):31. doi: 10.3390/jcm6030031. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Neurotherapeutics are provided here courtesy of Elsevier

RESOURCES