Assessment of left ventricular non-compaction in adults ... · PDF filephased array cardiac...

Click here to load reader

  • date post

  • Category


  • view

  • download


Embed Size (px)

Transcript of Assessment of left ventricular non-compaction in adults ... · PDF filephased array cardiac...

  • Archives of Cardiovascular Disease (2010) 103, 150159


    Assessment of left ventricular non-compaction inadults: Side-by-side comparison of cardiac magneticresonance imaging with echocardiography

    valuation de la non-compaction isole du ventricule gauche chez ladulte :comparaison de limagerie par rsonance magntique nuclaire aveclchographie cardiaque

    Franck Thunya,, Alexis Jacquierb, Bertrand Jopa,Roch Giorgi c, Jean-Yves Gaubertb, Jean-MichelBartoli b, Guy Moulinb, Gilbert Habiba

    a Department of Cardiology, CHU la Timone, University of Marseille Mditerrane ,264, rue Saint-Pierre, 13385 Marseille cedex 05, Franceb Department of Radiology, CHU la Timone, University of Marseille Mditerrane ,264, rue Saint-Pierre, 13385 Marseille cedex 05, Francec LERTIM, EA 3283, service de sant publique et dinformation mdicale, hpital de laTimone, Assistance publiqueHpitaux de Marseille, facult de mdecine, universitdAix-Marseille, Marseille, France

    Received 19 November 2009; received in revised form 10 January 2010; accepted 15 January2010Available online 10 March 2010


    SummaryBackground. Two-dimensional echocardiography images obtained at end-diastole and end-systole and cardiac magnetic resonance (CMR) images obtained at end-diastole represent the

    Echocardiography;Cardiac magneticresonance

    three imaging methodologies validated for diagnosis of left ventricular non-compaction (LVNC).No study has compared these methodologies in assessing the magnitude of non-compaction.Aims. To compare two-dimensional echocardiography with CMR in the evaluation of patientswith suspected LVNC.

    Abbreviations: C, compacted epicardial layer thickness; CMR, cardiac magnetic resonance; LV, left ventricular; LVNC, left ventricularnon-compaction; NC, non-compacted endocardial layer thickness; Echo-2D, two-dimensional echocardiography.

    Corresponding author. Fax: +33 491 384 764.E-mail address: [email protected] (F. Thuny).

    1875-2136/$ see front matter 2010 Elsevier Masson SAS. All rights reserved.doi:10.1016/j.acvd.2010.01.002

    mailto:[email protected]/10.1016/j.acvd.2010.01.002

  • Mutimodality imaging of non-com

    MOTS CLSNon-compaction ;Cardiomyopathie ;chocardiographie ;IRM cardiaque

    interobservateur du rapport NC/C maximal tait similaire pour les trois modalits.Conclusion. LIRMc apparat suprieure lEcho-2D standard dans lvaluation de lextension


    ddbtebiThus, there is no clear consensus for LVNC diagnosis. More-

    de la non-compaction en fo 2010 Elsevier Masson SAS.


    Left ventricular non-compaction (LVNC) is a recently rec-ognized cardiomyopathy characterized by a distinctive(spongy) morphological appearance of the left ventricu-lar (LV) myocardium [1]. Prominent LV trabeculae and deepintertrabecular recesses are present and the myocardial wallis often thickened with a thin compacted epicardial layer

    and a thickened non-compacted endocardial layer. In somepatients, LVNC is associated with LV dilatation and systolicdysfunction, leading to heart failure, ventricular arrhythmiaand thrombo-embolic complications [28]. This myocardial


    sant une information morphologique supplmentaire.s droits rservs.

    isorder has been described by two-dimensional echocar-iography (Echo-2D) and although many echo criteria haveeen proposed [4,8,9], a poor correlation exists betweenhem [10]. These discrepancies may be explained by differ-nces not only in the definition of abnormal trabeculation,ut also in the echo planes and phase of the cardiac cyclen which they are applied (end-diastole or end-systole) [10].

    paction 151

    Methods. Sixteen patients (48 17 years) with LVNC underwent echocardiography and CMRwithin the same week. Echocardiography images obtained at end-diastole and end-systolewere compared in a blinded fashion with those obtained by CMR at end-diastole to assessnon-compaction in 17 anatomical segments.Results. All segments could be analysed by CMR, whereas only 238 (87.5%) and 237 (87.1%)could be analysed by echocardiography at end-diastole and end-systole, respectively (p = 0.002).Among the analysable segments, a two-layered structure was observed in 54.0% by CMR, 42.9%by echocardiography at end-diastole and 41.4% by echocardiography at end-systole (p = 0.006).Similar distribution patterns were observed with the two echocardiographic methodologies.However, compared with echocardiography, CMR identified a higher rate of two-layered struc-tures in the anterior, anterolateral, inferolateral and inferior segments. Echocardiographyat end-systole underestimated the NC/C maximum ratio compared with CMR (p = 0.04) andechocardiography at end-diastole (p = 0.003). No significant difference was observed betweenCMR and echocardiography at end-diastole (p = 0.83). Interobserver reproducibility of the NC/Cmaximum ratio was similar for the three methodologies.Conclusion. CMR appears superior to standard echocardiography in assessing the extent ofnon-compaction and provides supplemental morphological information beyond that obtainedwith conventional echocardiography. 2010 Elsevier Masson SAS. All rights reserved.

    RsumContexte. Les images dchocardiographie bidimensionnelle (Echo-2D) obtenues la fois entldiastole et en tlsystole, ainsi que celles fournies par lIRM cardiaque (IRMc) en tldias-tole, sont les trois modalits valides pour le diagnostic de non-compaction isole du ventriculegauche (NCVG). Aucune tude na compar ces trois modalits entre elles dans lvaluation dela distribution et de limportance de la non-compaction.Objectifs. Comparer lEcho-2D standard avec lIRMc dans lvaluation des patients atteintsde NCVG.Mthodes. Seize patients (48 17 ans) avec une NCVG ont eu une Echo-2D et une IRMc durantla mme semaine. Les images chocardiographiques, obtenues en tldiastole et tlsystole,ont t compares, en insu, avec celles obtenues par IRMc en tldiastole, pour valuer lanon-compaction dans 17 segments anatomiques (soit n = 272 segments).Rsultats. Tous les segments ont pu tre analyss par lIRMc, alors que seuls 238 (87,5 %)et 237 (87,1 %) ont pu ltre par lEcho-2D, respectivement, en tldiastole et en tlsystole(p = 0,002). Parmi les segments analysables, une structure en double couche a t observe chez54 % des patients en IRMc, 42,9 % par Echo-2D en tldiastole et 41,4 % par Echo-2D en tlsys-tole (p = 0,006). Une distribution similaire de la non-compaction a t observe par les deuxmodalits chocardiographiques. Cependant, en comparaison avec lEcho-2D, lIRMc identifiaitun taux de structure en double couche plus important dans les segments antrieur, antrolatralet infrieur. LEcho-2D en tlsystole sous-estime le rapport NC/C maximal en comparaisonavec lIRMc (p = 0,04) et lEcho-2D en tldiastole (p = 0,003). En revanche, aucune diffrencesignificative na t observe entre lIRMc et lEcho-2D en tldiastole. La reproductibilit

    ver, Echo-2D evaluation can be limited by poor acousticindows. Recently, cardiac magnetic resonance (CMR) imag-

    ng has been proposed for the diagnosis of LVNC [1113] andew criteria have been proposed [14]. This technique has the

  • 1





















    apability of acquiring tomographic images of the LV cham-ers, with border definition that is often superior to thatchievable by echocardiography. However, while Echo-2Driteria have been validated either at end-diastole [8,9,15]r end-systole [4], CMR criteria have only been validatedt end-diastole [14]. Thus, no previous study has comparedhe two imaging techniques at different times of the cardiacycle in the assessment of patients with suspected LVNC. Theim of the present study was to compare standard Echo-2Dith CMR in the evaluation of patients with isolated LVNC.

    aterials and methods

    atient selection

    etween January 2003 and January 2008, 16 consecutiveatients for whom diagnosis of LVNC was suspected onchocardiographic criteria and without CMR contraindica-ions were enrolled. Each patient was assessed clinicallyy echocardiography first and CMR studies were performeduring the same week.

    Among the 16 patients included, 12 fulfilled Jennisriteria [4,5]. These specific echocardiographic diagnosticriteria are (in the absence of coexisting cardiac anomalies):

    a typical two-layered myocardial structure with a thincompacted outer (epicardial) band and a much thicker,non-compacted inner (endocardial) layer consisting oftrabecular meshing with deep endocardial spaces;a maximum end-systolic non-compacted epicardial layerthickness/compacted epicardial layer thickness (NC/C)ratio >2;colour Doppler evidence of deeply perfused intertrabec-ular recesses.

    Four other patients were also included, despite having aaximum NC/C ratio 2. In these patients, LVNC was highly

    uspected due to the presence of a marked two-layeredtructure associated with more than three prominent tra-eculations [8] and a family history of LVNC.

    This study complies with the declaration of Helsinki andas approved by our institutional review board. Written

    nformed consent was obtained from all patients.

    wo-dimensional echocardiographic imaging

    cho-2D were obtained using commercially available instru-ents (GE Medical Systems, Milwaukee, Wis) with 3.5-MHz

    ransducers equipped with harmonic imaging. Images werecquired in the long-axis parasternal view, the three short-xis views (basal, mid, apical), and the 2-, 3- and 4-chamberpical views. These views were used to visualize all 17egments according to American Heart Association recom-endations [16]. Offline analysis was then performed onigital