Canadian Journal of Cardiology 31 (2015) 785e791

Systematic Review/Meta-analysis

Echocardiography vs Cardiac Magnetic Resonance Imaging for the Diagnosis of Left Ventricular Thrombus: A Systematic Review Idan Roifman, MD,a Kim A. Connelly, MBBS, PhD,a,b Graham A. Wright, PhD,a,c and Harindra C. Wijeysundera, MD, PhDa,d a

Schulich Heart Program, Sunnybrook Health Sciences Centre, University of Toronto, Toronto, Ontario, Canada b

Division of Cardiology, St Michael’s Hospital, University of Toronto, Toronto, Ontario, Canada c d

Department of Medical Biophysics, University of Toronto, Toronto, Ontario, Canada

Institute for Clinical Evaluative Sciences, University of Toronto, Toronto, Ontario, Canada

ABSTRACT

  RESUM E

Background: Left ventricular (LV) thrombi can occur in the setting of LV dysfunction especially in the acute postmyocardial infarction period. The ideal imaging strategy to detect LV thrombi is currently unknown. The objective of this study was to conduct a systematic review to compare the accuracy of transthoracic echocardiography (TTE) with that of cardiac magnetic resonance (CMR) imaging for the detection of LV thrombi. Methods: OvidMEDLINE, EMBASE, and Cochrane databases were searched for articles published between January 1, 1946 and July 31, 2013. After screening of all potentially relevant abstracts and articles, 7 studies were ultimately selected for this review. Results: Our results suggest that late gadolinium enhancement CMR imaging is the most accurate modality for the detection of LV thrombi (sensitivity 88%, specificity 99%), followed by cine-CMR imaging

Introduction : Les thrombus du ventriculaire gauche (VG) peuvent survenir dans le cadre de la dysfonction du VG, particulièrement au cours de la phase aiguë de l’infarctus postmyocardique. Nous ne gie ide ale d’imagerie pour connaissons pas actuellement de strate tecter les thrombus du VG. L’objectif de cette e tude e tait de mener de matique pour comparer la pre cision de une revue syste chocardiographie transthoracique (ETT) avec celle de l’imagerie l’e sonance magne tique (IRM cardiaque) dans la cardiaque par re tection des thrombus du VG. de thodes : Nous avons examine  les banques de donne es d’Ovid Me s MEDLINE, d’Embase et de Cochrane pour relever des articles publie pistage de entre le 1er janvier 1946 et le 31 juillet 2013. Après le de sume s et articles potentiellement pertinents, nous avons tous les re lectionne 7e tudes pour la pre sente revue. finalement se

Left ventricular (LV) thrombi can occur in the setting of LV dysfunction especially in the acute stage after myocardial infarction (MI).1,2 Such thrombi are clinically important because of their ability to embolize.3,4 In fact, patients who develop mural thrombi after MI have an overall poor prognosis including a 10% rate of systemic embolization.5 Early detection of LV thrombi is critical because it allows for early initiation of anticoagulation therapy to reduce the likelihood of embolization.6

Transthoracic echocardiography (TTE) has been used clinically in the detection of LV thrombi since the early 1980s.7-11 More recently, LV echocardiographic contrast agents have been used to increase accuracy of detection, especially in patients with difficult acoustic windows.12 TTE is currently considered the first-line diagnostic test for LV thrombus assessment.13 However, inherent limitations of TTE (eg, imaging patients with poor acoustic windows) might result in significant interobserver variability in clinical interpretation.14 Ultimately, this might result in reduced diagnostic accuracy in thrombus detection. Cardiac magnetic resonance (CMR) imaging can potentially provide a diagnostic advantage over echocardiography because of its ability to characterize the myocardium in addition to providing cine imaging (cine-CMR). Recent advances in sequence development, combined with the ability of paramagnetic contrast agents to enhance the ventricular blood pool, provide late gadolinium enhancement (LE)-CMR with a potential advantage in detecting LV thrombi.15-18

Received for publication December 17, 2014. Accepted January 21, 2015. Corresponding author: Dr Harindra C. Wijeysundera, Schulich Heart Centre, Sunnybrook Health Sciences Centre, 2075 Bayview Ave, Suite A202, Toronto, Ontario M4N 3M5, Canada. Tel.: þ1-416-480-4527; fax: þ1416-480-4657. E-mail: [email protected] See page 790 for disclosure information.

http://dx.doi.org/10.1016/j.cjca.2015.01.011 0828-282X/Ó 2015 Canadian Cardiovascular Society. Published by Elsevier Inc. All rights reserved.

786

Canadian Journal of Cardiology Volume 31 2015

(sensitivity 58%-79%, specificity 99%, accuracy 95%, positive predictive value 93%-95%, negative predictive value 95%-96%), contrast TTE (sensitivity 23%-61%, specificity 96%-99%, accuracy 92%, positive predictive value 93%, negative predictive value 91%), and, finally, noncontrast TTE (sensitivity 24%-33%, specificity 94%-95%, accuracy 82%, positive predictive value 57%, negative predictive value 85%). Accuracy of TTE might be improved if a clear clinical indication is provided and with routine use of LV opacifying contrast agents. Conclusions: Our findings indicate that late gadolinium enhancement CMR imaging is the most accurate sequence in the detection of LV thrombus, and should be favoured when there is a high index of suspicion. When CMR is contraindicated, unavailable, or impractical, our analysis argues for contrast-TTE in patients at high risk for developing LV thrombi.

sultats : Nos re sultats suggèrent que l’IRM cardiaque de Re  la rehaussement tardif après injection de gadolinium est la modalite cise pour la de tection de thrombus du VG (sensibilite  de 80 %, plus pre cificite  de 99 %), et qu’elle est suivie par la cine -IRM cardiaque spe  de 58 %-79 %, spe cificite  de 99 %, pre cision de (sensibilite dictive de 93 %-95 %, valeur pre dictive ne gative de 95 %, valeur pre  de 23 %-61 %, spe cificite  95 %-96 %), l’ETT de contraste (sensibilite cision de 92 %, valeur pre dictive positive de 93 %, de 96 %-99 %, pre dictive ne gative de 91 %) et finalement l’ETT sans contraste valeur pre  de 24 %-33 %, spe cificite  de 94 %-95 %, pre cision de 82 %, (sensibilite dictive positive de 57 %, valeur pre dictive ne gative de 85 %). valeur pre cision de l’ETT serait possible si une indication cliUne meilleure pre matique d’agents de nique claire est fournie et par l’utilisation syste contraste pour opacifier le VG. sultats indiquent que l’IRM cardiaque de Conclusions : Nos re quence la rehaussement tardif après injection de gadolinium est la se cise dans la de tection des thrombus du VG et qu’elle devrait plus pre gie e lorsque l’indice de suspicion est e leve . Lorsque l’IRM être privile e, non disponible ou irre alisable, notre cardiaque est contre-indique analyse plaide en faveur de l’ETT de contraste chez les patients s à un risque e leve  de de veloppement de thrombus du VG. expose

There is a paucity of data that compare these modalities in their ability to detect LV thrombi. Accordingly, the objective of our study was to conduct a systematic review of the published literature to determine which modality/sequence of TTE with no contrast, TTE with contrast, cine-CMR, and LE-CMR would have the highest diagnostic accuracy for the detection of LV thrombi.

search. Duplicate references were identified and removed using EndNote X5 Library (Thomson Reuters). Data to be extracted from the study were specified a priori. Extraction was achieved using piloted forms. The following data items were extracted: number of study participants, patient population, and type of study performed, reference standard used, and results reported. For types of diagnostic studies we categorized the following groups: (1) TTE without contrast; (2) TTE with contrast; (3) cine-CMR; and (4) LE-CMR. When provided, or when the data were available, sensitivity (Sn) and specificity (Sp) were reported/calculated, as was positive predictive value (PPV) and negative predictive value (NPV). Diagnostic accuracy, defined as the number of true positive plus true negative divided by the total number of observations, was reported if included in the original study. The quality of studies was assessed using criteria previously described for assessing diagnostic studies: (1) blinding; (2) consecutive recruitment of patients; and (3) using a single reference standard (rather than composite).20,21

Materials and Methods A systematic review was conducted based on guidelines from the Preferred Reporting Items for Systematic Reviews and Metaanalyses statement.19 OvidMEDLINE, EMBASE, and Cochrane databases were searched by a librarian at Sunnybrook Health Sciences Center, University of Toronto, for articles published between January 1, 1946 and July 31, 2013 using the following subject and text terms: left ventricle, thrombus, intracardiac, magnetic resonance imaging, computed tomography, and echocardiography (see the Supplemental Appendix S1 section of the Supplementary Material for an example of the full search string used in one of the databases). The initial search was limited to studies reporting on subjects  19 years of age and to those written in the English language. No other initial exclusion criteria were used. After the initial search, further screening and exclusion was performed according to standardized protocols.19 All abstracts from the selections derived using this search strategy were screened and studies that were clearly irrelevant were excluded at this point. Subsequently, potentially relevant fulltext articles were assessed for eligibility. Our search revealed several old studies that compared cardiac computed tomography (CCT) with echocardiography and/or magnetic resonance imaging. These studies were excluded because they used outdated computed tomography or echocardiographic and/or magnetic resonance imaging technology. For the sake of completeness, we included these studies in the supplemental material (see Supplemental Table S1). Bibliographies of all appropriate articles that were identified via this search were also scanned for other potential articles that might have been missed during our

Results Search results Figure 1 shows the results of our screening process. A total of 3315 citations were originally identified, of which only 7 studies were appropriate for inclusion into the systematic review. Reasons for exclusion included lack of assessment of diagnostic capabilities of TTE and CMR for detecting LV thrombi and lack of data provided for TTE and CMR. Study characteristics are summarized in Table 1. Study quality is summarized in Table 2. Only 2 studies met all 3 of the aforementioned quality criteria. Two other studies met 2 of 3 criteria. The remaining studies met only 1 criterion. Initial studies that compared echocardiography with CMR The first 2 studies that compared the accuracy of CMR and TTE for detecting LV thrombi were performed by

Roifman et al. TTE vs CMR for LV Thrombus Diagnosis

787

Figure 1. Flow diagram describing the literature review and subsequent assessment for eligibility of abstracts and articles in the systematic review. LV, left ventricular.

Barkhausen et al.22 and Mollet et al.23 These were published within a few months of each other in 2002. Barkhausen et al.22 compared the diagnostic accuracy achieved with different CMR techniques with that of TTE. They prospectively enrolled 24 patients with known or suspected intracardiac thrombi. Each patient had a TTE and CMR imaging examination. For LV thrombi, in LE-CMR imaging 9 patients were diagnosed as having thrombi. In cine-CMR sequences, 5 patients were diagnosed with thrombi. With TTE 6 patients were diagnosed with thrombi.22 Late enhancement images were acquired 20 minutes after administration of gadolinium. Inversion times were varied to null the normal myocardium and ranged from 180 to 300 msec. This was an exploratory study and the first that showed that CMR imaging might be more efficacious at detecting LV thrombi compared with TTE. However, there were significant limitations of this study. First, echocardiographic images were obtained without the use of LV opacifying contrast agents. Second, there was no gold standard that was used to validate the true presence of the thrombi. Mollet et al.23 prospectively enrolled 57 post-MI patients and those with ischemic cardiomyopathy and performed TTEs, cine-CMR, and LE-CMR imaging in these patients. Of 57 patients, with LE-CMR 12 ventricular thrombi were detected, with cine-CMR 6 thrombi were detected, and with TTE 5 thrombi were detected. Again, no gold standard for thrombus detection was used in this study. Further, TTE was not performed with the aid of contrast agents. Although these studies were important in generating the hypothesis that LE-CMR imaging might be superior to cineCMR imaging and noncontrast TTE in detecting LV thrombi, their validity was limited largely because of the lack of use of a gold standard. The first study to use a recognized gold standard was performed by Srichai et al., in 2006.24 They retrospectively assessed 361 patients who had surgical and/or pathological confirmation of the presence or absence of LV

thrombi. Of these, 106 also had preoperative transesophageal echocardiograms (TEEs), TTE, and CMR imaging. They found that the most accurate sequence was LE-CMR (Sn 88%, Sp 99%), followed by TEE (Sn 40%, Sp 96%), and finally TTE (Sn 23%, Sp 96%).24 LE imaging was performed with varying inversion times (240-280 msec) to null the normal myocardium. This study addressed 2 previously unresolved issues. First, it clearly showed superiority of LE-CMR to other techniques compared with a recognized gold standard with regard to Sn for the detection of LV thrombi. Second, LV opacification techniques with contrast agents were “occasionally” used in patients who were deemed to have technically difficult studies.24 However, the article did not mention how many patients received contrast agents nor did it describe the criteria for their administration. Routine use of contrast echocardiography vs CMR Weinsaft et al. performed CMR imaging and TTEs on 121 prospectively enrolled patients at high risk of developing thrombus because of previous MI or LV dysfunction.25 LECMR was used as the gold standard based largely on the study by Srichai et al.24 Late enhancement images were acquired 10 minutes after administration of gadolinium. Unlike previous studies, inversion time was increased from that needed to null the normal myocardium (200-350 msec) to 600 msec to selectively null the avascular thrombus. Further, 3 patients who had pathological evidence of thrombus all had thrombi detected using LE-CMR in the study population of Weinsaft et al.25 Contrast echocardiography was performed on all patients irrespective of noncontrast echocardiography findings. Patients with apical aneurysms on echocardiography showed a trend toward having more LV thrombi compared with those without an aneurysm (25% vs 10%; P ¼ 0.09). Cine-CMR had a Sn of 79%, Sp of 99%, accuracy of 95%, PPV of 95%, and NPV of 95%. Contrast TTE had a Sn of

788

Canadian Journal of Cardiology Volume 31 2015

Table 1. Summary of data abstracted for the articles included in the systematic review

Source

n

Mean age  SD, years

Barkhausen et al.22

24

57  14

Mollet et al.23

57

Srichai et al.24

106

59.7  12.5 (thrombus on LE-CMR); 62.2  9.9 (no thrombus on LE-CMR) 62  10

Weinsaft et al.25

121

61  13

Weinsaft et al.26

243

Joshi et al.27

Delewi et al.28

Study population (including specific inclusion and exclusion criteria if available) Patients referred to echocardiography laboratory with known or suspected myocardial thrombi; exclusion criteria: none specified All consecutive patients (over an 8month time frame) after MI or ischemic cardiomyopathy who underwent CMR for viability; exclusion criteria: none specified

Reference standard

Results

Patients who were clinically known or suspected to have thrombi

LE-CMR: 9 LV thrombi detected; Cine CMR: 5 LV thrombi detected; TTE (no contrast): 6 LV thrombi detected LE-CMR: detected 12 thrombi; Cine CMR: detected 6 thrombi TTE (no contrast): detected 5 thrombi

None declared

Consecutive patients (between November 1997 and December 2003) with surgical/pathological confirmation of the presence or absence of LV thrombi; exclusion criteria: none specified Patients required the following to be included in the study between August 2005 and November 2007: (1) either after MI and/or LV systolic dysfunction; (2) CMR and echocardiogram performed within 7 days of each other; exclusion criteria: none specified

Surgical, pathological

LE-CMR: Sn 88%, Sp 99%; TTE: Sn 23%, Sp 96%; TEE: Sn 40%, Sp 96%

LE-CMR

60  15

LV dysfunction registry patients (with inclusion criteria of LVEF < 50% for entry into the registry); exclusion criteria: none specified

LE-CMR

52

62  15

None

200

56  9

Patients with LV systolic dysfunction who underwent examination with CMR and echocardiogram within 30 days of each other for assessment of LVEF for consideration of ICD implantation between March 2007 and October 2010; exclusion criteria: none 200 patients after STEMI; inclusion criteria: successful primary PCI with stent insertion within 12 hours of symptom onset; exclusion criteria: hemodynamic instability, unsuccessful PCI, mild increase of cardiac enzyme levels (defined as an increase of creatine kinase < 10 times the upper limit of normal)

TTE (non contrast): Sn 33%, Sp 94%; accuracy 82%; PPV 57%; NPV 85%; TTE (contrast): Sn 61%, Sp 99%, accuracy 92%, PPV 93%, NPV 91%; Cine CMR: Sn 79%, Sp 99%, accuracy 95%, PPV 95%, NPV 95% TTE: Sn 33%, Sp 91%, accuracy 85%, PPV 29%, NPV 93%; Cine CMR: Sn 58%, Sp 99%, accuracy 95%, PPV 93%, NPV 96% LE-CMR: 6 thrombi detected; TTE (no contrast): 1 thrombus detected

LE-CMR

TTE (no contrast): Sn 24%, Sp 95%

CMR, cardiac magnetic resonance imaging; ICD, implantable cardioverter defibrillator; LE-CMR, late gadolinium enhancement cardiac magnetic resonance imaging; LV, left ventricular; LVEF, left ventricular ejection fraction; MI, myocardial infarction; NPV, negative predictive value; PPV, positive predictive value; PCI, percutaneous coronary intervention; Sn, sensitivity; Sp, specificity; STEMI, ST-elevation myocardial infarction; TEE, transesophageal echocardiography; TTE, transthoracic echocardiography.

61%, Sp of 99%, accuracy of 92%, PPV of 93%, and NPV of 91%. Noncontrast TTE had a Sn of only 33%, Sp of 94%, accuracy of 82%, PPV of 57% and NPV of 85%. TTE with contrast nearly doubled the Sn (61% vs 33%; P < 0.05) and yielded improved accuracy vs noncontrast TTE (92% vs 82%; P < 0.001). Further, there was a close level of agreement between cine-CMR imaging and contrast TTE (k statistic ¼ 0.79; P < 0.001). Finally, LE-CMR was found more likely to yield incremental diagnostic value over noncontrast TTE and cine-CMR imaging in patients with lower ejection fraction values.25 This was the first study to compare diagnostic accuracies of cine-CMR imaging, contrast TTE, and noncontrast TTE. The findings suggest that routine

administration of contrast agents greatly enhances the detection rate of LV thrombi. In fact, results of this study suggest that the overall accuracy of contrast TTE and cine-CMR imaging are similar. However, even with administration of echocardiographic contrast agents, LE-CMR imaging was still a considerably more accurate modality in terms of thrombus detection. ‘Real-life’ registry data Weinsaft et al. used patients from a CMR registry consisting of 243 consecutive patients with LV systolic dysfunction (defined as left ventricular ejection fraction < 50%).26

Roifman et al. TTE vs CMR for LV Thrombus Diagnosis

789

Table 2. Assessment of study quality

Blinding

Consecutive enrollment of patients

Mollet et al.23 Barkhausen et al.22

Yes Yes

Yes Yes

Srichai et al.24 Weinsaft et al.25 Weinsaft et al.26 Joshi et al.27 Delewi et al.28

Yes Yes Yes Yes Yes

No Yes Yes No Unknown

Single reference standard No No imaging or pathological reference standard Yes Yes Yes No No

CMR imaging consisted of cine-CMR and LE-CMR imaging for tailored assessment of LV thrombus. Late enhancement imaging was performed 10 minutes after administration of gadolinium. Similar to the previous study by the same group, inversion time was increased from that needed to null the normal myocardium (200-350 msec) to 600 msec to selectively null the avascular thrombus. Clinical records of registry patients were then assessed to identify patients who had TTE performed within 1 week of CMR imaging. TTE had an accuracy of 85%, Sn of 33%, Sp of 91%, PPV of 29%, and NPV of 93% compared with LE-CMR. LV opacifying echocardiographic contrast was administered in only 4% of patients who underwent TTE. In these cases, TTE had an accuracy of 80% compared with LE-CMR. Further, among patients with possible LV thrombus as an indication for the TTE, Sn and PPV were dramatically improved (60% and 75%, respectively). TTEs with lower image quality ratings were more likely to be discordant with LE-CMR images (P < 0.02). Cine-CMR imaging had a Sn of 58%, Sp of 99%, accuracy of 95%, PPV of 93%, and NPV of 96% compared with LE-CMR. Of the 243 patients included in this study, 219 underwent a 6-month follow-up for end points supporting imaging diagnosis of an LV thrombus. These consisted of the development of stroke, transient ischemic attack, or pathology verification. Patients who had thrombus detected using LE-CMR imaging had a 5-fold greater rate of these end points compared with those who did not have a thrombus detected using LE-CMR (16.7% vs 3.1%; P ¼ 0.02). In contrast, patients who were stratified according to presence or absence of thrombus using TTE did not show a significant difference in the development of clinical end points (7.7% vs 4.2%; P ¼ 0.34). This study was an important addition to the literature, because it highlighted that in ‘real-life’ practice at a large tertiary care centre, TTE can lead to inaccurate results with respect to LV thrombus detection. Further, in contrast to LE-CMR, thrombi detected using TTE were not associated with significantly more clinical end points.

Accurate identification of LV thrombi is important because it often directs subsequent anticoagulation therapy to prevent embolic events. Although it is true that current guidelines consider it “reasonable” to initiate anticoagulation in patients highly suspected to have thrombus because of apical akinesis or dyskinesis in the absence of an identifiable thrombus (class 2b recommendation), a 2a recommendation is afforded if a thrombus is visualized.6 These recommendations are also supported by the results of our systematic review. The only study to explicitly assess apical aneurysms found that there was a nonsignificant increase in the prevalence of thrombi when an aneurysm was present on echocardiography.25 Although our results indicate a superiority of LE-CMR imaging in the detection of LV thrombi, it is conceivable that some of the studies that were assessed underestimated its diagnostic accuracy. Most of the studies in this systematic review assessed LE-CMR imaging using ‘standard’ techniques by adjusting the inversion time to null the normal myocardium. Two studies that were reviewed also used the “long inversion time” approach to selectively null the thrombus.25,26 This approach has been described as being potentially beneficial in the detection of thrombi because it was designed to render the thrombus black and surrounding myocardium white.29 Thus, it is conceivable that the diagnostic accuracy of LE-CMR imaging might have been further enhanced if this ‘long inversion time’ approach would have been used in the other studies. The results of our systematic review are significant because they strongly argue for a more prominent role for CMR imaging in the clinical assessment of LV thrombi. This recommendation needs to be weighed against known drawbacks of LE-CMR imaging, such as contraindications (eg, patients with renal dysfunction with indexed estimated glomerular filtration rate < 30 mL/min/1.73 m2), increased acquisition cost, availability, and the requirement for patient breathholding (which can dramatically affect image quality). Many of these limitations do not apply to TTE. When CMR is contraindicated, unavailable, or impractical, these results argue for routine administration of LV opacifying echocardiographic contrast agents in patients at high risk for developing LV thrombi. Ideally, a meta-analysis would be performed to quantify the difference in diagnostic accuracies among the different modalities. However, the number of studies included in the systematic review was small and used multiple gold standards (or none at all). Thus, we believe that any attempts to combine the individual study results in meta-analysis to provide a single summary estimate might have led to inaccurate or misleading results. Limitations

Discussion Results of our review suggest that LE-CMR is the most accurate modality for the detection of LV thrombi, followed by cine-CMR imaging, contrast TTE, and, finally, noncontrast TTE. The evidence also suggests that under ‘real-life’ conditions, use of TTE can allow a large number of LV thrombi to be undetected. Accuracy of TTE might be improved if a clear clinical indication is provided for the TTE and with routine use of LV-opacifying contrast agents.

Results of this study need to be interpreted in the context of its limitations. First, the quality of the studies included in the review varied, with only a few of them meeting all 3 criteria for quality assessment of diagnostic tests described herein (Table 2). However, it is important to note that none of the studies showed that an alternate modality or sequence was superior to LE-CMR imaging in the detection of thrombi. Second, there was only 1 study that exclusively used surgical or pathological evidence of LV thrombus as the gold

790

standard.24 In this study the authors concluded that LE-CMR imaging was the most accurate sequence compared with this standard. Based largely on this study, subsequent studies considered LE-CMR as the gold standard. Despite this, 2 other studies used pathological verification in a subset of patients.25,26 In these small subsets, LE-CMR imaging remained the most accurate sequence for thrombus detection. Third, in only 1 study the routine use of contrast TTE vs LE-CMR imaging was assessed. It is known that routine use of LVopacifying contrast agents increases the accuracy of thrombus detection. Thus, in the studies that did not assess patients who were routinely administered contrast, the accuracy that they attributed to TTE might have been underestimated. Future directions To address some of the limitations, more studies are needed that directly compare LE-CMR, cine-CMR imaging, noncontrast TTE, and contrast TTE with a pathological or surgical gold standard for the detection of LV thrombi. This work might help solidify the status of LE-CMR imaging as the nonpathological gold standard for the detection of LV thrombi. More studies are also needed that specifically compare routine use of contrast TTE with LE-CMR imaging to confirm the results of the study from Weinsaft et al.25 Moreover, newer novel CMR sequences (such as T2* to identify ferrous products of hemoglobin breakdown) should be examined to determine if they provide further diagnostic utility above and beyond LE-CMR imaging. In addition, further study is required to determine if use of long inversion time LE-CMR imaging to selectively null the thrombus results in enhanced diagnostic yield vs more traditional T1 times set to null the normal myocardium. Finally, CCT has become an increasingly used and important cardiac imaging modality in North America and is capable of detecting LV thrombi in a time-efficient manner.30 Although our search identified a small number of studies that compared CCT with echocardiography and CMR, they were all published between 1983 and 1990 (we included a summary of these studies in the Supplemental Table S1). As such, they used imaging technology that is not currently used clinically and we believe that it would not be appropriate to compare their results with the newer studies (published from 2002 to 2012) and sequences that were summarized in this review. Future head to head studies that compare contemporary CCT, CMR imaging, and echocardiography are required to assess the relative diagnostic accuracy of CCT for LV thrombus detection. Conclusions The current literature suggests that LE-CMR is the most accurate modality for detecting LV thrombi. Further, routine use of LV-opacifying echocardiographic contrast agents increases accuracy to levels similar to that of cine-CMR imaging. Routine noncontrast echocardiograms are often performed in patients after MI and provide invaluable clinical information. We do not recommend a change in this practice. In balancing the available evidence with practical clinical realities, we recommend the following: in patients in whom there is a clinical suspicion for LV thrombus and a noncontrast echocardiography is negative, contrast should be administered. If a

Canadian Journal of Cardiology Volume 31 2015

thrombus is still not seen on contrast echocardiography and clinical suspicion is high, LE-CMR should be performed as the definitive test. Funding Sources Dr Harindra C. Wijeysundera is supported by a Heart and Stroke Foundation of Canada Distinguished Clinician Scientist Award. Dr Kim A. Connelly is supported by a Canadian Institutes of Health Research New Investigator Award. Dr Graham A. Wright is supported by a Canada Research Chair in Imaging for Cardiovascular Therapeutics. Disclosures The authors have no conflicts of interest to disclose. References 1. Nihoyannopoulos P, Smith GC, Maseri A, Foale RA. The natural history of left ventricular thrombus in myocardial infarction: a rationale in support of masterly inactivity. J Am Coll Cardiol 1989;14:903-11. 2. Asinger RW, Mikell FL, Elsperger J, Hodges M. Incidence of leftventricular thrombosis after acute transmural myocardial infarction. Serial evaluation by two-dimensional echocardiography. N Engl J Med 1981;305:297-302. 3. Stratton JR, Resnick AD. Increased embolic risk in patients with left ventricular thrombi. Circulation 1987;75:1004-11. 4. Vaitkus PT, Barnathan ES. Embolic potential, prevention and management of mural thrombus complicating anterior myocardial infarction: a meta-analysis. J Am Coll Cardiol 1993;22:1004-9. 5. Barbera S, Hillis LD. Echocardiographic recognition of left ventricular mural thrombus. Echocardiography 1999;16:289-95. 6. O’Gara PT, Kushner FG, Ascheim DD, et al. 2013 ACCF/AHA guideline for the management of ST-elevation myocardial infarction: executive summary: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation 2013;127:529-55. 7. Stratton JR, Lighty GW Jr, Pearlman AS, Ritchie JL. Detection of left ventricular thrombus by two-dimensional echocardiography: sensitivity, specificity, and causes of uncertainty. Circulation 1982;66:156-66. 8. Greenland P, Knopman DS, Mikell FL, et al. Echocardiography in diagnostic assessment of stroke. Ann Intern Med 1981;95:51-3. 9. Chiarella F, Santoro E, Domenicucci S, Maggioni A, Vecchio C. Predischarge two-dimensional echocardiographic evaluation of left ventricular thrombosis after acute myocardial infarction in the GISSI-3 study. Am J Cardiol 1998;81:822-7. 10. Asinger RW, Mikell FL, Sharma B, Hodges M. Observations on detecting left ventricular thrombus with two dimensional echocardiography: emphasis on avoidance of false positive diagnoses. Am J Cardiol 1981;47:145-56. 11. Visser CA, Kan G, David GK, Lie KI, Durrer D. Two dimensional echocardiography in the diagnosis of left ventricular thrombus. A prospective study of 67 patients with anatomic validation. Chest 1983;83: 228-32. 12. Mansencal N, Nasr IA, Pilliere R, et al. Usefulness of contrast echocardiography for assessment of left ventricular thrombus after acute myocardial infarction. Am J Cardiol 2007;99:1667-70.

Roifman et al. TTE vs CMR for LV Thrombus Diagnosis 13. Douglas PS, Garcia MJ, Haines DE, et al. ACCF/ASE/AHA/ASNC/ HFSA/HRS/SCAI/SCCM/SCCT/SCMR 2011 appropriate use criteria for echocardiography. A report of the American College of Cardiology Foundation Appropriate Use Criteria Task Force, American Society of Echocardiography, American Heart Association, American Society of Nuclear Cardiology, Heart Failure Society of America, Heart Rhythm Society, Society for Cardiovascular Angiography and Interventions, Society of Critical Care Medicine, Society of Cardiovascular Computed Tomography, Society for Cardiovascular Magnetic Resonance American College of Chest Physicians. J Am Soc Echocardiogr 2011;24:229-67. 14. Berger AK, Gottdiener JS, Yohe MA, Guerrero JL. Epidemiological approach to quality assessment in echocardiographic diagnosis. J Am Coll Cardiol 1999;34:1831-6. 15. Simonetti OP, Kim RJ, Fieno DS, et al. An improved MR imaging technique for the visualization of myocardial infarction. Radiology 2001;218:215-23. 16. Kim RJ, Fieno DS, Parrish TB, et al. Relationship of MRI delayed contrast enhancement to irreversible injury, infarct age, and contractile function. Circulation 1999;100:1992-2002.

791

23. Mollet NR, Dymarkowski S, Volders W, et al. Visualization of ventricular thrombi with contrast-enhanced magnetic resonance imaging in patients with ischemic heart disease. Circulation 2002;106:2873-6. 24. Srichai MB, Junor C, Rodriguez LL, et al. Clinical, imaging, and pathological characteristics of left ventricular thrombus: a comparison of contrast-enhanced magnetic resonance imaging, transthoracic echocardiography, and transesophageal echocardiography with surgical or pathological validation. Am Heart J 2006;152:75-84. 25. Weinsaft JW, Kim RJ, Ross M, et al. Contrast-enhanced anatomic imaging as compared to contrast-enhanced tissue characterization for detection of left ventricular thrombus. JACC Cardiovasc Imaging 2009;2:969-79. 26. Weinsaft JW, Kim HW, Crowley AL, et al. LV thrombus detection by routine echocardiography: insights into performance characteristics using delayed enhancement CMR. JACC Cardiovasc Imaging 2011;4:702-12.

17. Kim RJ, Hillenbrand HB, Judd RM. Evaluation of myocardial viability by MRI. Herz 2000;25:417-30.

27. Joshi SB, Connelly KA, Jimenez-Juan L, et al. Potential clinical impact of cardiovascular magnetic resonance assessment of ejection fraction on eligibility for cardioverter defibrillator implantation. J Cardiovasc Magn Reson 2012;14:69.

18. Kim RJ, Wu E, Rafael A, et al. The use of contrast-enhanced magnetic resonance imaging to identify reversible myocardial dysfunction. N Engl J Med 2000;343:1445-53.

28. Delewi R, Nijveldt R, Hirsch A, et al. Left ventricular thrombus formation after acute myocardial infarction as assessed by cardiovascular magnetic resonance imaging. Eur J Radiol 2012;81:3900-4.

19. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med 2009;151:264-9. W64.

29. Goyal P, Weinsaft JW. Cardiovascular magnetic resonance imaging for assessment of cardiac thrombus. Methodist Debakey Cardiovasc J 2013;9:132-6.

20. Romagnuolo J, Bardou M, Rahme E, et al. Magnetic resonance cholangiopancreatography: a meta-analysis of test performance in suspected biliary disease. Ann Intern Med 2003;139:547-57.

30. Kassop D, Donovan MS, Cheezum MK, et al. Cardiac masses on cardiac CT: a review. Curr Cardiovasc Imaging Rep 2014;7:9281.

21. Shiga T, Wajima Z, Apfel CC, Inoue T, Ohe Y. Diagnostic accuracy of transesophageal echocardiography, helical computed tomography, and magnetic resonance imaging for suspected thoracic aortic dissection: systematic review and meta-analysis. Arch Intern Med 2006;166:1350-6. 22. Barkhausen J, Hunold P, Eggebrecht H, et al. Detection and characterization of intracardiac thrombi on MR imaging. AJR Am J Roentgenol 2002;179:1539-44.

Supplementary Material To access the supplementary material accompanying this article, visit the online version of the Canadian Journal of Cardiology at www.onlinecjc.ca and at http://dx.doi.org/10. 1016/j.cjca.2015.01.011.