ARTICLE OPEN ACCESS

Systematic review and meta-analysis of cardiacinvolvement in mitochondrial myopathyAsfia Quadir Carly Sabine Pontifex BSc Helen Lee Robertson MLIS Christopher Labos MD MSc and

Gerald Pfeffer MD PhD

Neurol Genet 20195e339 doi101212NXG0000000000000339

Correspondence

Dr Pfeffer

geraldpfefferucalgaryca

AbstractObjectiveOur goal was to perform a systematic review of the literature to demonstrate the prevalence ofcardiac abnormalities identified using cardiac investigations in patients with mitochondrialmyopathy (MM)

MethodsThis systematic review surveys the available evidence for cardiac investigations in MM froma total of 21 studies including 825 participants Data were stratified by genetic mutation andclinical syndrome

ResultsWe identified echocardiogram and ECG as the principal screening modalities that identifycardiac structural (29) and conduction abnormalities (39) in various MM syndromes ECGabnormalities were more prevalent in patients with m3243AgtG mutations than other genedefects and patients with mitochondrial encephalopathy lactic acidosis and stroke-like epi-sodes (MELAS) had a higher prevalence of ECG abnormalities than patients with other clinicalsyndromes Echocardiogram abnormalities were significantly more prevalent in patients withm3243AgtG or m8344AgtG mutations compared with other genetic mutations SimilarlyMELAS and MERRF had a higher prevalence compared with other syndromes We observeda descriptive finding of an increased prevalence of ECG abnormalities in pediatric patientscompared with adults

ConclusionsThis analysis supports the presence of a more severe cardiac phenotype in MELAS andmyoclonic epilepsy with ragged red fibres syndromes and with their commonly associatedgenetic mutations (m3243AgtG and m8344AgtG) This provides the first evidence basis onwhich to provide more intensive cardiac screening for patients with certain clinical syndromesand genetic mutations However the data are based on a small number of studies We rec-ommend further studies of natural history therapeutic response pediatric participants andcardiac MRI as areas for future investigation

From the Hotchkiss Brain Institute (AQ CSP GP) University of Calgary Health Sciences Library (HLR) University of Calgary Alberta Queen Elizabeth Health Complex (CL)Montreal Quebec and Department of Clinical Neurosciences (GP) Cumming School of Medicine University of Calgary Alberta Canada

Go to NeurologyorgNG for full disclosures Funding information is provided at the end of the article

The Article Processing Charge was funded by the authors

Study sponsorship Dr Pfeffer receives institutional support from the Department of Clinical Neurosciences the Cumming School of Medicine and Hotchkiss Brain Institute (Universityof Calgary)

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 40 (CC BY-NC-ND) which permits downloadingand sharing the work provided it is properly cited The work cannot be changed in any way or used commercially without permission from the journal

Copyright copy 2019 The Author(s) Published by Wolters Kluwer Health Inc on behalf of the American Academy of Neurology 1

Mitochondrial myopathies (MMs) are clinically heteroge-neous disorders resulting from defects in the respiratorychain preferentially affecting organs with high-energyrequirements Cardiac dysfunction is common as part of thephenotype of MM and is manifested by cardiomyopathycardiac conduction defects andor heart failure1 The iden-tification of cardiac dysfunction is particularly important be-cause of its impact on quality of life morbidity mortality andespecially because it is amenable to treatment2 Some mito-chondrial syndromes such as Kearns-Sayre syndrome (KSS)include cardiac dysfunction as a core feature2 However mostof the common mitochondrial syndromes may or may notinclude cardiac dysfunction which can occur at any point inthe disease course and as such screening investigations arerecommended for early detection At present the optimalinvestigations and time intervals for screening are not estab-lished and studies describing the natural history of cardiacdysfunction in MM have been limited3

This systematic review serves to summarize the available ev-idence regarding screening investigations to (1) provide dataindicating the diagnostic yield of different cardiac inves-tigations (2) compare the frequency of cardiac abnormalitiesin differing mitochondrial syndromes and genotypes (3)compare the frequency of diagnostic abnormalities betweenpediatric and adult patients and (4) identify limitations in theevidence and areas for future study The overall goal of thisstudy is to systematically demonstrate the prevalence of car-diac abnormalities across various MM syndromes

MethodsTwo investigators (AQ and GP) created a preliminary searchstrategy that was subsequently refined by a medical librarian(HLR) The search was conducted on June 4 2018 on thefollowing databases Epub Ahead of Print In-Process amp OtherNon-Indexed Citations Ovid MEDLINE Daily and OvidMEDLINE EMBASE Cochrane Central Register of Con-trolled Trials Scopus and Web of Science Combinations ofsubject headings keywords and synonyms used includedMMsmitochondrial encephalomyopathies Kearns-Sayre Syndromecardiomyopathy cardiac arrhythmia and tachycardia Fullsearch terms are available in supplemental data (linkslwwcomNXGA159) The authors (AQ and GP) independentlyscreened all titles and abstracts identified by the initial searchWe obtained full-text versions of studies identified as beingpotentially relevant which were then independently assessedOnly English-language publications were considered

We also performed a search through the bibliographies of allincluded articles to ensure that no other relevant articles weremissed although no additional publications for inclusion wereidentified by these means

Study selectionWe included all studies in which patients had a confirmed di-agnosis of MM using genetic testing andor muscle pathologyWe did not include patients with KSS because this syndromeincludes cardiac dysfunction as a core feature of its diagnosticcriteria Our objective was to identify all published studiesdocumenting cardiac complications of MM and using at leastone modality of cardiac investigation Studies that did not dis-close the investigative modality or presented focused or in-complete data were not included Cross-sectional studies cohortstudies or clinical trials were included in our review Casereports (ie small studies of 4 or fewer cases) were excluded

Data extractionData extraction was independently performed by 2 of theauthors (AQ and CSP) using a custom-designed data ex-traction form which was piloted on 5 randomly selected eli-gible studies before being finalized In cases of disagreement 3of the authors (AQ CSP and GP) discussed the dis-crepancies to reach a consensus

The information extracted included year of study duration offollow-up number of participants age the diagnostic testsused the number of subjects with abnormal results with ECGechocardiography (echo) Holter monitor cardiac MRI(CMRI) and nuclear medicine studies We also collected dataregarding treatment and clinical outcomes where availableWhen data were incomplete we attempted to contact authorsof the relevant articles to obtain more detailed data and re-ceived such data for 2 publications45

For purposes of our analysis data from child (aged lt18 years)and adult (aged ge18 years) patients were recorded and ana-lyzed separately Adult patients who had disease onset inchildhood were still included in the adult group Data werestratified using 2 separate approaches based on genetic sub-types of MM (m3243AgtG mutation m8344AgtG mutationother mitochondrial DNA (mtDNA) point mutations singlelarge-scale mtDNA deletions and nuclear gene mutations)and based on clinical syndromes (mitochondrial encepha-lopathy lactic acidosis and stroke-like episodes [MELAS]chronic progressive external ophthalmoplegia [CPEO] orother phenotypes of MM) Clinical syndromes were definedbased on how they had been identified within individual

GlossaryBBB = bundle branch block CI = confidence interval CMR = cardiac MRI CPEO = chronic progressive externalophthalmoplegia KSS = Kearns-Sayre syndrome LVH = left ventricular hypertrophy MELAS = mitochondrialencephalopathy lactic acidosis and stroke-like episodes MERRF = myoclonic epilepsy with ragged red fibres MM =mitochondrial myopathy mtDNA = mitochondrial DNA WPW = Wolff-Parkinson-White syndrome

2 Neurology Genetics | Volume 5 Number 4 | August 2019 NeurologyorgNG

articles In cases in which the specific syndrome was not de-fined but the genotype was available if sufficient in-dividualized data were presented we would infer the clinicalsyndrome for individual patients (eg a patient with them3243AgtG mutation with stroke-like episodes encepha-lopathy and lactic acidosis would be interpreted as a MELASsyndrome case) Patients with unspecified syndromes wereincluded in a category labeled ldquoother phenotypesrdquo

Based on this approach we allowed cases to be counted in boththe genetic group and the clinical syndrome group if in-dividualized data were reported for patients having both a de-fined clinical syndrome and genetic lesion The rationale for thisapproach was that there would be clinical value to providinga systematic analysis of data according to both genetic andclinical classifications However for overall prevalence estimateseach patient was counted only onceWe did not perform a risk ofbias assessment because the included studies were uncontrolledcross-sectional studies without an identified intervention

Statistical analysisThe proportion of patients with abnormal tests was summa-rized with a random effects meta-analysis6 with the estimateof heterogeneity being taken from the inverse-variance Toaccount for studies with no outcomes of interest ie zeroevents an arcsine transformation was applied7 Heterogeneitywas assessed using the I-squared statistic

Data availabilityAll data pertaining to this work (list of abstracts articlesreviewed data entry spreadsheet and statistical analysis) willbe made available upon request by any qualified investigator

ResultsWe identified 8601 articles in our literature search for title andabstract review for relevance After the primary screen weretained 171 articles for full-text review After assessing thesestudies for eligibility 21 studies were included in our meta-analysis The Preferred Reporting Items for Systematic ReviewsandMeta-Analyses flowchart is presented in figure 18 The basiccharacteristics of the included studies are presented in table 1

ECGIn our review 14 studies provided data on ECG findings inadult patients The overall prevalence of abnormalities was 039(95 confidence interval [CI] 028ndash050) (figure 2) The dataindicate a significantly higher prevalence of ECG abnormalitiesin patients with the m3243AgtG mutation (065 [CI033ndash092]) compared with the m8344AgtG mutation (026[CI 010ndash046]) single mtDNA deletions (CI 021[013ndash0305]) other mtDNA point mutations (021 [CI009ndash035]) and nuclear gene mutations (013 [CI002ndash029]) However the data for several of these groupsoriginate predominantly from a single study9 When analyzedby clinical syndrome patients with MELAS had significantlymore ECG abnormalities (088 [CI 071ndash099]) than patientswith CPEO (032 [CI 017ndash049]) or other phenotypes (017

[CI 000ndash049]) Given the heterogeneity in outcomereporting it was not possible to perform analysis of specificECG abnormalities with the exception of bundle branch block(BBB) which is presented in figure e-1 (linkslwwcomNXGA159) This abnormality was seldom reported in MM pop-ulations and the overall prevalence was zero in many studies

A total of 9 articles included ECG findings in pediatricpatients (figure 3) We cannot conclude that there is anysignificant difference between syndromes because of the smallnumber of reported participants The overall prevalence ofECG abnormalities was 055 [CI 038ndash071] We observeddescriptively that abnormalities in m3243AgtG patients (079[CI 048ndash100]) were higher than detected in other geneticor phenotypic categories In comparison to adults pediatricpatients with MM had a higher prevalence of ECG abnor-malities (055 [CI 038ndash071] compared with 039 [CI028ndash050]) as a descriptive finding which did not achievestatistical significance (p = 007)

EchocardiographyFourteen of the included studies provided data on echocardi-ography findings in adult patients The overall prevalence ofabnormalities was calculated as 029 (CI 017ndash042) (figure 4)The best-studied group was patients with the m3243AgtGmutation in which data from 246 participants were reportedHere it was expected that m3243AgtG andm8344AgtGwouldbe associated with a higher prevalence of abnormalities thanother genetic defects This was supported by the data whenm3243AgtG and m8344AgtG (041 (CI 021ndash062) and 044(CI 023ndash066)) were compared with single mtDNA deletions(005 [CI 001ndash012]) other mtDNA point mutations (008[CI 001ndash020]) and nuclear gene mutations (000 [CI000ndash007]) When analyzed by clinical syndrome MELAS(080 [CI 057ndash097]) had a higher prevalence compared withCPEO and other phenotypes It was possible to perform ananalysis for left ventricular hypertrophy (LVH) which hada prevalence of 053 in patients with MELAS (CI 025ndash080)and a prevalence of 018 (CI 010ndash027) in patients with MMoverall (figure e-2 linkslwwcomNXGA159)

Echocardiogram findings in child patients were reported ina total of 9 studies The overall prevalence of abnormalitieswas 035 (CI 023ndash047) (figure 5) There were no significantdifferences in prevalence between the syndromes or geno-types although the data consisted of a small number ofpatients and originated from a small number of studies Be-tween pediatric and adult patients there was no difference inthe overall prevalence of abnormalities

Cardiac MRICMRI was performed in adult patients in 5 studies and therewas substantial heterogeneity in the results Meta-analyzing all5 studies revealed an overall prevalence of abnormalities of035 (CI 009ndash066) but given the small number of partic-ipants and studies it is difficult to draw any further con-clusions (figure e-3 linkslwwcomNXGA159) Only one

NeurologyorgNG Neurology Genetics | Volume 5 Number 4 | August 2019 3

study reported CMRI results in children10 with data fromonly a single participant

HolterHolter monitoring was assessed in adults in only 4studies510ndash12 with very low diagnostic yield and with an overallprevalence estimate for abnormalities of 003 (CI 000ndash014)(figure e-4 linkslwwcomNXGA159) This suggests thatHolter monitor is a low-yield investigation for adult patientsOnly one study included Holter monitor data from children10

Common ECG and echo abnormalities were found in severalstudies of patients with various MM mutations and clinicalsyndromes but they were unsuitable for meta-analysis giventhe sparsity of the data They are presented additively in tablee-1 (linkslwwcomNXGA159) These data provide anoverall view of the diversity of findings in these common MMsyndromes Overall the most severe cardiac abnormalities

such as LVH and ventricular pre-excitation abnormalitieswere most prevalent in m3243AgtG or MELAS patients

DiscussionOverall patients with MM have a high prevalence of cardiacabnormalities although the data indicate that MELAS hasa higher prevalence of ECG abnormalities compared with otherclinical syndromes and the m3243AgtG mutation is associatedwith a higher prevalence of ECG abnormalities than other ge-netic defects Abnormalities detected by echo are more preva-lent in MELAS and myoclonic epilepsy with ragged red fibres(MERRF) compared with other syndromes and when analyzedby genetic category the m3243AgtG and m8344AgtG muta-tions were associated with a higher prevalence than othermutations Comparing children to adult patients we report thedescriptive finding of a higher prevalence of ECG abnormalitiesin children compared with adults with MM supporting the

Figure 1 PRISMA flowchart

PRISMA = Preferred Reporting Itemsfor Systematic Reviews and Meta-Analyses

4 Neurology Genetics | Volume 5 Number 4 | August 2019 NeurologyorgNG

concept that MM has a more severe cardiac phenotype inchildhood compared with adult patients However this differ-ence was not significant (p = 007) perhaps relating to the smallnumber of participants in pediatric studies

The majority of the available data are from ECG and echo andboth modalities demonstrated a high prevalence of detectedabnormalities We expected that a higher prevalence of abnor-malities would be detected with CMRI given that CMRI isrecognized as a structural imaging test with higher sensitivity andreproducibility1314 The available evidence shows similar di-agnostic yield of CMRI in comparison with echo however the

data are very heterogeneous and from a limited number ofpatients In studies including m3243AgtG andor MELASpatients 2 studies1516 reported no CMRI abnormalities and 1study17 reported CMRI abnormalities in nearly all patients Thisdiscrepancy can hopefully be addressed in future research Onestudy of CMRI in 14 participants showed no structural abnor-malities but an overall reduction of myocardial glucose uptakewhich may be a biomarker of interest for future study16 Onlya single study includedCMRI data from a pediatric participant10

Overall CMRI has thus far had very limited study in MM andfurther investigation will be required to understand the advan-tages and limitations of this modality in patients with MM

Table 1 Characteristics of included studies

Author year PMIDCohorttype

Cardiac modalitiesstudied

Total participantsincluded

Mean age ofparticipants (y)

Agerange (y)

Sex(MF)

Akaike et al 199722 9068909 Adult ECG echo 5 480 20ndash64 23

Anan et al 199535 7850981 Childadult

ECG echo 15 410 12ndash54 96

Baik et al 201019 21189931 Child ECG 57 48 NA 2631

Baik et al 201220 23038991 Child ECG echo 27 50 NA 1116

Catteruicca et al201510

25559684 Childadult

ECG echo Holter cardiacMRI (CMRI)

15 460 8ndash71 510

Cordeiro et al200918

24209401 Child Echo 63 077 017ndash137 3033

Florian et al 201517 26001801 Adult ECG CMRI 64 467 35ndash65 2836

Galetta et al 201436 25139213 Adult ECG echo 20 553 NA 515

Hollingsworth et al201215

22513320 Adult ECG echo CMRI 10 425 30ndash55 55

Ikawa et al 200721 17280875 Childadult

ECG echo MIBI 5 398 16ndash51 23

Limongelli et al201011

20083621 Childadult

ECG echo Holter 30 392 16ndash62 1317

Lindroos et al201616

26112752 Adult CMRI 14 465 362ndash57 410

Nesbitt et al 201324 23355809 Adult Echo 129 NA 091ndash74 5079

Okajima et al199827

9875091 Childadult

ECG echo 11 165 6ndash23 83

Pfeffer and Mezei20123

22987704 Adult ECG echo 15 569 18ndash83 87

Ueno and Shiotani199923

10598894 Adult Echo 10 486 NA 64

Vydt et al 200712 17223431 Adult ECG echo Holter 12 350 18ndash57 57

Wahbi et al 20105 20177121 Childadult

ECG echo Holter 18 426 12ndash71 810

Wahbi et al 20159 26224072 Adult ECG echo 272 440 28ndash58 115157

Wortmann et al200728

17407476 Child ECG echo 5 56 0ndash14 32

Yilmaz et al 20124 22143423 Adult ECG CMRI 37 NA 45ndash61 1423

Abbreviation NA = not available

NeurologyorgNG Neurology Genetics | Volume 5 Number 4 | August 2019 5

Pediatric patients may characteristically be considered to havea more severe phenotype than adult patients with MM Basedon this it was expected that pediatric patients would havea higher prevalence of abnormalities which was suggested byour data comparing ECG abnormalities in children and adults

This suggests that pediatric patients with MM should be moreclosely monitored for cardiac conduction abnormalitiesHowever we cannot exclude the possibility that the prevalenceof abnormalities for pediatric MM could have been inflated bythe incidental inclusion of patients with KSS

Figure 2 Prevalence of ECG abnormalities with ES (95 CI) in adult participants

Note that patients are reported according to genetic defect (upper portion of the figure) and according to their clinical syndrome (lower portion of figure) andsome patients may be duplicated if both a genetic defect and clinical syndrome was provided This equally applies to figures 3ndash5 ES = effect size mtDNA =mitochondrial DNA

6 Neurology Genetics | Volume 5 Number 4 | August 2019 NeurologyorgNG

We observed that most studies of children included patientswith nonspecific MM syndromes18ndash20 Only 2 studies repor-ted m8344AgtG patients510 in which all patients had noabnormalities on ECG but 1 patient had abnormalitiesdetected by echo10 Further investigation into pediatric MMpatients with better described clinical syndromes and mo-lecular defects should be conducted to determine the severityof the cardiac phenotype in these patients In our analysisHolter monitoring was a very low-yield investigation in adultand child patients with MM and its use as a routine screeningprocedure should perhaps be reassessed

In our study 10 articles described patients with MM withdiabetes9111215ndash1721ndash24 Of the 122 patients with diabetes 47of them were diagnosed with MELAS Given that diabetes isa major risk factor for cardiac dysfunction it might beexpected that patients with MELAS have greater cardiac riskas a consequence This elevated risk was borne out by thedata and the interaction of diabetes on cardiac dysfunction inpatients with MM is worthy of future study

The goal of this study was to systematically review the prev-alence of cardiac abnormalities in MM syndromes it was nota goal of this work to characterize the natural history symp-toms sudden-death occurrence or therapies However these

might be areas for future study and in the course of thisreview some of the included articles contained data regardingsymptoms sudden cardiac death and natural history Wefound 8 articles reporting symptoms of cardiac dysfunction inpatients before diagnosis including chest pain41117 variousdegrees of dyspnea410111722 syncope11 palpitation1011 andexercise limitations11 Although 1 article found all patients tobe asymptomatic3 5 articles reported cardiac symptoms oc-curring before investigation for MM91117212526 One article9

reported 7 patients with a history of major adverse cardiacevents including heart failure and third-degree atrioventricularblock whereas another study25 reported a patient who hadpreviously underwent cardiac transplantation Twostudies2126 presented patients with cardiomyopathy pre-ceding their diagnosis of MM Of interest 1 study concludedthat 31 of their patients withMMwere previously diagnosedwith cardiac conditions particularly arrhythmia impaired leftventricular systolic function LVH coronary artery diseaseand left bundle branch block17 However it is unclear in 1study whether indicators of cardiac abnormalities were pres-ent before or after the diagnosis of MM22

An important limitation encountered in the literature is thatfew studies presented long-term follow-up data Only 6 of the21 included articles reported follow-up data regarding the

Figure 3 Prevalence of ECG abnormalities with ES (95 CI) in pediatric participants

ES = effect size

NeurologyorgNG Neurology Genetics | Volume 5 Number 4 | August 2019 7

Figure 4 Prevalence of echocardiography abnormalities with ES (95 CI) in adult participants

ES = effect size mtDNA = mitochondrial DNA

8 Neurology Genetics | Volume 5 Number 4 | August 2019 NeurologyorgNG

evolution of cardiac dysfunction over time359112728 Five ofthese reported detailed follow-up data39112728 are as follows

One study examined the occurrence of major cardiac events ina population of 260 patients retrospectively followed fora median of 7 years9 Twenty-seven of the patients died ofcardiac events during the follow-up period and the likelihoodof death was associated with the presence of abnormalities oncardiac screening tests (42 in patients with 2 or more ab-normalities on ECGecho compared with 10 in the totalstudy population) Patients with large-scale single mtDNAdeletions and the m3243AgtG mutation had the most severecardiac phenotype (although this former category may haveincidentally included patients with KSS)

Another publication presented follow-up data on 15 patientsdiagnosed with adult-onset CPEO for a mean of 65 years withECG and echo3 A total of 5 patients developed new abnor-malities but in 4 of these cases the abnormalities were un-related to MM only one patient developed a mildcardiomyopathy attributed to MM

A previous investigation followed 32 patients for an average of41 years undergoing serial echocardiograms and clinical as-sessment11 Two patients with initially normal ECG de-veloped hypertrophic cardiomyopathy during follow-upSerial echocardiography in a patient with CPEO and a patient

with MM showed LVH development whereas 1 patient withMELAS developed systolic impairment

A study reported 6 pediatric m3243AgtG patients of which 2died at an early stage of disease28 One patient had died be-cause of multiorgan failure whereas the other had died be-cause of cardiorespiratory failure One patient had developedmild LVH and tricuspid regurgitation without pulmonaryhypertension in the first year of follow-up One patient haddeveloped stroke-like episodes in adulthood The remainingpatients did not develop clinical symptoms of cardiac dys-function over 3 years of follow-up

Last one study examined the development of cardiac dys-function in adult and pediatric patients withMELAS (10 of 11harbored the m3243AgtG mutation) during subsequent fol-low-up27 They observed 11 patients 4 adults and 7 childrenover an average time span of 69 years whereas 6 patients werefollowed up for more than 5 years Three patients showeda worsening ejection fraction whereas 2 patients died a childand an adult one from causes unrelated to MM and the otherdue to congestive heart failure respectively

There are other works that did not meet our criteria for in-clusion in this systematic analysis that have also providedimportant contributions to our understanding of the naturalhistory of cardiac dysfunction in MM A previous study of 228

Figure 5 Prevalence of echocardiography abnormalities with ES (95 CI) in pediatric participants

ES = effect size

NeurologyorgNG Neurology Genetics | Volume 5 Number 4 | August 2019 9

patients with single mtDNA deletion syndromes presenteddata from initial presentation and after an average of 187years of follow-up29 Over this time interval the prevalence ofcardiac conduction abnormalities increased from 13 to53 and the prevalence of cardiomyopathy increased from0 to 26 This demonstrates a lower prevalence of cardiacdysfunction than seen in our systematic analysis which isa surprising finding given the inclusion of numerous KSSpatients (which prevented it from being included in our sys-tematic analysis) In contrast another study of patients withMELAS and the m3243AgtG mutation had a mean follow-upof 38 years and identified death due to cardiac causes in 4 of55 patients which appears to be a more severe phenotypethan expected for this short duration of follow-up30 This highlevel of variability between publications emphasizes the needfor further study of natural history after extended follow-up

Based on the above data the following observations arepossible (1) cardiac abnormalities are very common acrossvarious MM syndromes (2) analysis of clinical syndromesindicates that MELAS and MERRF appear to have the mostsevere cardiac phenotype in adult patients and (3) analysis ofgenetic subtypes reveals a more severe cardiac phenotype forthe m3243AgtG and m8344AgtG mutations compared withother MM genetic defects Cardiac screening investigationsshould be made on a case-by-case basis but the followingsuggestions based on the evidence may be helpful and aresummarized in table 2 ECG and echo are the best-establishedscreening tests for MM the role for CMRI requires furtherinvestigation Generally patients with MELAS and MERRFare considered to be at a higher risk of cardiac complicationsand annual screening has been recommended31 A commonand well-described cardiac complication is Wolff-Parkinson-White syndrome (WPW) which is overrepresented in

MELAS32 In contrast adult-onset CPEO is generally con-sidered to have a milder phenotype and a previous study hassuggested a screening interval as long as 3ndash5 years3 Theresults of our analysis appear to be consistent with theserecommendations but it should be emphasized that the dataare based on a small number of studies and the prevalence ofabnormalities is high for most syndromes

Some cardiac findings that have been associated with MMwere rarely or not identified in this systematic analysis Leftventricular noncompaction has been associated with severalmtDNA point mutations3133 and we identified a single casewith a mtDNA point mutation at position m1123211 Re-strictive or histiocytoid cardiomyopathies have also been as-sociated with MM34 but these were also not identified as partof this analysis This was likely due to the inclusion criteria forthis review which excluded individual case reports and smallcase series

Limitations of this review relate to the heterogeneity in studydesign ascertainment methods cardiac investigations andoutcome reporting The majority of included publicationswere case series focused on cardiologic issues which may havebiased the reported cases toward a higher prevalence of car-diac involvement The majority of the data pertain to patientswith mtDNA-related diseases and patients with nuclear mi-tochondrial disorders were not well represented Details re-garding the genetic investigations which were performed(specific assays their sensitivity and other negative geneticinvestigations) were not generally reported Details regardingmuscle pathology interpretation were not generally reportedbut our assumption was that a statement regarding diagnosticmuscle pathology would be reliable in the reported litera-ture In 2 studies417 data from patients with CPEO were

Table 2 Summary of findings from systematic review of the evidence

Recommendation

Principal screening tests ECGEchocardiography

Other screening tests requiring further study Cardiac MRI (CMRI)

Recommended screening Available evidence did not allow for a specific recommendation although we provide 2 generalrecommendationsIn adults patients withMELAS andMERRF should be screenedmore frequently than patients withother syndromesIn patients with mitochondrial myopathy (MM) ECG screening should be more frequent forchildren than adults

Common abnormalities in CPEO Bundle branch block (BBB)Left ventricular dysfunction

Common abnormalities in m3243AgtGMELAS andm8344AgtGMERRF

Left ventricular hypertrophy (LVH)Wolff-Parkinson-White syndrome (WPW)

Areas for future study Natural history studiesStudies of treatment outcomesInvestigations of new modalities (CMRI and nuclear medicine) in comparison with establishedmodalities (ECGecho)Investigation into diabetes as a risk factor for cardiac dysfunction associated with MELAS andpossibly other mitochondrial diseasesAdditional studies in pediatric patients

10 Neurology Genetics | Volume 5 Number 4 | August 2019 NeurologyorgNG

aggregated with data from a few patients with Kearns-SayreSyndrome (KSS) so it is possible that the prevalence ofcardiac abnormalities was inflated by the incidental inclusionof some patients with KSS to the CPEO group (given thatpatients with KSS necessarily have cardiac conduction defectsas part of their syndrome) In some of the studies includingpatients with the m3243AgtG mutation data from mutationcarriers who were asymptomatic or paucisymptomatic wereincluded which could also have affected the results9121524

This illustrates another limitation of our study that patientswith the m3243AgtG mutation included aggregate data frompatients with MELAS syndrome or other phenotypes whichdoes not provide an accurate picture of the cardiac findings inphenotypic subgroups with the m3243AgtG mutation Asimilar limitation is present for patients with the m8344AgtGmutation in which only a subset would have had MERRFsyndrome There may also have been some overlap in thegroups (for example CPEO could overlap with single large-scale mtDNA deletions mtDNA point mutations and nucleargene mutation groups) based on the available informationthe groups were assigned based on how they had been des-ignated in their respective publications Based on this classi-fication it is also possible that differences in the definition ofeach syndrome existed between the included publicationswhich is also a limitation of this study In 1 article it wasnecessary to assume which abnormalities were from ECG(BBB andWPW) compared with Holter monitor (ventriculardysrhythmias)10 In 1 article studying CMRI only aggregatedata were presented but as this was in the normal range weassumed a normal result for all patients15 Finally our analysiscould not assess the severity of individual reported results soeven if the prevalence of results was similar between syn-dromes we cannot rule out the possibility that some of theabnormalities may have been more severe in some syndromesthan others

This systematic analysis provides evidence to support a moreaggressive cardiac screening for subsets of patients with MMbased on a more severe cardiac phenotype for MELAS andMERRF among clinical syndromes and a more severe phe-notype for m3243AgtG and m8344AgtG mutations amonggenetic defects We also report a descriptive finding of moreprevalent ECG abnormalities in pediatric MM compared withadult MM The prevalence of abnormalities with Holtermonitoring is low and suggests that it may not be necessaryfor routine screening More research is needed to determinehow CMRI compares with echo as a modality for structuralcardiac imaging There are overall limited data on naturalhistory and treatment response and further longitudinal studyof these conditions will be of benefit to the field Further studyincluding pediatric participants would also be of benefit giventhe small numbers of studies and participants identified in thisreview

AcknowledgmentThe authors specially thank University of Calgary libraries andTheresa Connolly for their assistance with this review They

also thank Dr Ali Yilmaz and Dr Karim Wahbi who providedadditional information regarding their published data for thisreview

Study fundingAsfia Quadir is the recipient of OrsquoBrien Centre SummerStudent Awards in 2017 and 2018

DisclosureDisclosures available NeurologyorgNG

Publication historyReceived by Neurology Genetics November 2 2018 Accepted in finalform April 30 2019

References1 Meyers DE Basha HI Koenig MKMitochondrial cardiomyopathy pathophysiology

diagnosis and management Texas Heart Inst J 201340385ndash3942 Pfeffer G Chinnery PF Diagnosis and treatment of mitochondrial myopathies Ann

Med 2013454ndash163 Pfeffer G Mezei MM Cardiac screening investigations in adult-onset progressive

external ophthalmoplegia patients Muscle Nerve 201246593ndash596

Appendix Authors

Name Location Role Contribution

Asfia Quadir University ofCalgaryCalgaryAlbertaCanada

First author Data collectiondatainterpretation anddrafting andediting of themanuscript forintellectualcontent

Carly SabinePontifexBSc

University ofCalgaryCalgaryAlbertaCanada

Author Data collectiondatainterpretation andediting of themanuscript forintellectualcontent

Helen LeeRobertsonMLIS

University ofCalgaryCalgaryAlbertaCanada

Author Development ofsearch strategyand editing of themanuscript forintellectualcontent

ChristopherLabos MDMSc

QueenElizabethHealthComplexMontrealQuebecCanada

Author Statistical analysisdatainterpretation andediting of themanuscript forintellectualcontent

GeraldPfeffer MDPhD

University ofCalgaryCalgaryAlbertaCanada

Correspondingauthor

Design andconceptualizationof the study datacollection datainterpretation anddrafting andediting of themanuscript forintellectualcontent

NeurologyorgNG Neurology Genetics | Volume 5 Number 4 | August 2019 11

4 Yilmaz A Gdynia H-J Ponfick M et al Cardiovascular magnetic resonance imaging(CMR) reveals characteristic pattern of myocardial damage in patients with mito-chondrial myopathy Clin Res Cardiol 2012101255ndash261

5 Wahbi K Larue S Jardel C et al Cardiac involvement is frequent in patients with them8344AgtG mutation of mitochondrial DNA Neurology 201074674ndash677

6 DerSimonian R Laird N Meta-analysis in clinical trials Control Clin Trials 19867177ndash188

7 Rucker G Schwarzer G Carpenter J Olkin I Why add anything to nothing Thearcsine difference as a measure of treatment effect in meta-analysis with zero cells StatMed 200928721ndash738

8 Moher D Liberati A Tetzlaff J Altman DG PRISMA Group Preferred reportingitems for systematic reviews and meta-analyses the PRISMA statement Ann InternMed 2009151264ndash269

9 Wahbi K Bougouin W Behin A et al Long-term cardiac prognosis and risk strati-fication in 260 adults presenting with mitochondrial diseases Eur Heart J 2015362886ndash2893

10 Catteruccia M Sauchelli D DellaMarca G et al ldquoMyo-cardiomyopathyrdquo is commonlyassociated with the A8344G ldquoMERRFrdquo mutation J Neurol 2015262701ndash710

11 Limongelli G Tome-Esteban M Dejthevaporn C Rahman S Hanna MG Elliott PMPrevalence and natural history of heart disease in adults with primary mitochondrialrespiratory chain disease Eur J Heart Fail 201012114ndash121

12 Vydt TCG de Coo RFM Soliman OII et al Cardiac involvement in adults withm3243AgtG MELAS gene mutation Am J Cardiol 200799264ndash269

13 Grothues F Smith GC Moon JC et al Comparison of interstudy reproducibility ofcardiovascular magnetic resonance with two-dimensional echocardiography in normalsubjects and in patients with heart failure or left ventricular hypertrophy Am J Cardiol20029029ndash34

14 Gardner BI Bingham SE Allen MR Blatter DD Anderson JL Cardiac magneticresonance versus transthoracic echocardiography for the assessment of cardiac vol-umes and regional function after myocardial infarction an intrasubject comparisonusing simultaneous intrasubject recordings Cardiovasc Ultrasound 2009738

15 Hollingsworth KG Gorman GS Trenell MI et al Cardiomyopathy is common inpatients with the mitochondrial DNA m3243AgtG mutation and correlates withmutation load Neuromuscul Disord 201222592ndash596

16 LindroosMM Parkka JP TaittonenMT et al Myocardial glucose uptake in patients withthe m3243AgtG mutation in mitochondrial DNA J Inherit Metab Dis 20163967ndash74

17 Florian A Ludwig A Stubbe-Drager B et al Characteristic cardiac phenotypes aredetected by cardiovascular magnetic resonance in patients with different clinicalphenotypes and genotypes of mitochondrial myopathy J Cardiovasc Magn Reson20151740

18 Cordeiro M Scaglia F Lopes Da Silva S et al The brain-heart connection in mito-chondrial respiratory chain diseases Neuroradiol J 200922558ndash563

19 Baik R Chae JH Lee YM Kang HC Lee JS KimHD Electrocardiography as an earlycardiac screening test in children with mitochondrial disease Korean J Pediatr 201053644ndash647

20 Baik R Yu R Lee YM Kang HC Lee JS KimHD Early cardiac evaluation in childrenwith non-specific mitochondrial disease with isolated mitochondrial respiratory chaincomplex I defect J Paediatr Child Health 2012481016ndash1020

21 Ikawa M Kawai Y Arakawa K et al Evaluation of respiratory chain failure in mito-chondrial cardiomyopathy by assessments of 99mTc-MIBI washout and 123I-BMIPP99mTc-MIBI mismatch Mitochondrion 20077164ndash170

22 Akaike M Kawai H Yokoi K et al Cardiac dysfunction in patients with chronicprogressive external ophthalmoplegia Clin Cardiol 199720239ndash243

23 Ueno H Shiotani H Cardiac abnormalities in diabetic patients with mutation in themitochondrial tRNA(Leu(UUR)) gene Jpn Circ J 199963877ndash880

24 Nesbitt V Pitceathly RDS Turnbull DM et al The UKMRCMitochondrial DiseasePatient Cohort Study clinical phenotypes associated with the m3243AgtGmutationmdashimplications for diagnosis and management J Neurol Neurosurg Psy-chiatry 201384936ndash938

25 Malfatti E Laforet P Jardel C et al High risk of severe cardiac adverse events inpatients with mitochondrial m3243AgtG mutation Neurology 201380100ndash105

26 Holmgren D Wahlander H Eriksson BO Oldfors A Holme E Tulinius M Car-diomyopathy in children with mitochondrial disease clinical course and cardiologicalfindings Eur Heart J 200324280ndash288

27 Okajima Y Tanabe Y Takayanagi M Aotsuka H A follow up study of myocardialinvolvement in patients with mitochondrial encephalomyopathy lactic acidosis andstroke-like episodes (MELAS) Heart 199880292ndash295

28 Wortmann SB Rodenburg RJ Backx AP Schmitt E Smeitink JA Morava E Earlycardiac involvement in children carrying the A3243GmtDNAmutation Acta Paediatr200796450ndash451

29 Mancuso M Orsucci D Angelini C et al Redefining phenotypes associated withmitochondrial DNA single deletion J Neurol 20152621301ndash1309

30 Kaufmann P Engelstad K Wei Y et al Natural history of MELAS associated withmitochondrial DNA m3243AgtG genotype Neurology 2011771965ndash1971

31 Bates MG Bourke JP Giordano C drsquoAmati G Turnbull DM Taylor RW Cardiacinvolvement in mitochondrial DNA disease clinical spectrum diagnosis and man-agement Eur Heart J 2012333023ndash3033

32 Sproule DM Kaufmann P Engelstad K Starc TJ Hordof AJ De Vivo DC Wolff-Parkinson-White syndrome in patients with MELAS Arch Neurol 2007641625ndash1627

33 Arbustini E Favalli V Narula N Serio A Grasso M Left ventricular noncompactiona distinct genetic cardiomyopathy J Am Coll Cardiol 201668949ndash966

34 El-Hattab AW Scaglia F Mitochondrial cardiomyopathies Front Cardiovasc Med2016325

35 Anan R NakagawaMMiyataM et al Cardiac involvement in mitochondrial diseasesA study on 17 patients with documented mitochondrial DNA defects Circulation199591955ndash961

36 Galetta F Franzoni F Mancuso M et al Cardiac involvement in chronic progressiveexternal ophthalmoplegia J Neurol Sci 2014345189ndash192

12 Neurology Genetics | Volume 5 Number 4 | August 2019 NeurologyorgNG

DOI 101212NXG000000000000033920195 Neurol Genet

Asfia Quadir Carly Sabine Pontifex Helen Lee Robertson et al myopathy

Systematic review and meta-analysis of cardiac involvement in mitochondrial

This information is current as of June 12 2019

reserved Online ISSN 2376-7839Published by Wolters Kluwer Health Inc on behalf of the American Academy of Neurology All rightsan open-access online-only continuous publication journal Copyright Copyright copy 2019 The Author(s)

is an official journal of the American Academy of Neurology Published since April 2015 it isNeurol Genet

ServicesUpdated Information amp

httpngneurologyorgcontent54e339fullhtmlincluding high resolution figures can be found at

References httpngneurologyorgcontent54e339fullhtmlref-list-1

This article cites 36 articles 4 of which you can access for free at

Subspecialty Collections

httpngneurologyorgcgicollectionmuscle_diseaseMuscle disease

netics-mitochondrial_disordershttpngneurologyorgcgicollectionmitochondrial_disorders_see_geMitochondrial disorders see GeneticsMitochondrial disorders

httpngneurologyorgcgicollectionmitochondrial_disordersMitochondrial disorders

httpngneurologyorgcgicollectioncardiacCardiacfollowing collection(s) This article along with others on similar topics appears in the

Permissions amp Licensing

httpngneurologyorgmiscaboutxhtmlpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in

Reprints

httpngneurologyorgmiscaddirxhtmlreprintsusInformation about ordering reprints can be found online

reserved Online ISSN 2376-7839Published by Wolters Kluwer Health Inc on behalf of the American Academy of Neurology All rightsan open-access online-only continuous publication journal Copyright Copyright copy 2019 The Author(s)

is an official journal of the American Academy of Neurology Published since April 2015 it isNeurol Genet

ServicesUpdated Information amp

httpngneurologyorgcontent54e339fullhtmlincluding high resolution figures can be found at

References httpngneurologyorgcontent54e339fullhtmlref-list-1

This article cites 36 articles 4 of which you can access for free at

Subspecialty Collections

httpngneurologyorgcgicollectionmuscle_diseaseMuscle disease

netics-mitochondrial_disordershttpngneurologyorgcgicollectionmitochondrial_disorders_see_geMitochondrial disorders see GeneticsMitochondrial disorders

httpngneurologyorgcgicollectionmitochondrial_disordersMitochondrial disorders

httpngneurologyorgcgicollectioncardiacCardiacfollowing collection(s) This article along with others on similar topics appears in the

Permissions amp Licensing

httpngneurologyorgmiscaboutxhtmlpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in

Reprints

httpngneurologyorgmiscaddirxhtmlreprintsusInformation about ordering reprints can be found online

reserved Online ISSN 2376-7839Published by Wolters Kluwer Health Inc on behalf of the American Academy of Neurology All rightsan open-access online-only continuous publication journal Copyright Copyright copy 2019 The Author(s)

is an official journal of the American Academy of Neurology Published since April 2015 it isNeurol Genet