continuing medical education program in echocardiography

C© 2008, the AuthorsJournal compilation C© 2008, Wiley Periodicals, Inc.DOI: 10.1111/j.1540-8175.2008.00704.x

Color M-Mode Echocardiography Is MoreReproducible than Conventional M-ModeEchocardiography for Septal-to-Posterior WallMotion Delay Measurements

CME

Vincent L. Sorrell, M.D.,∗ William D. Ross,∗† Jeff Gregoire,† and Imran Ata, M.D.∗

Division of Cardiology, Department of Medicine, University of Arizona College of Medicine, SarverHeart Center, Tucson, Arizona

Objective: To compare the reproducibility of conventional gray-scale M-mode to color Doppler M-mode assessment of the septal-to-posterior wall motion delay (SPWMD) which is used as a markerof left ventricular mechanical dyssynchrony. Methods: Forty-two SPWMD datasets were interpretedby two experienced cardiologists and the inter- and intraobserver correlation coefficients were deter-mined. Subgroup analysis was performed on patients with severe septal wall motion abnormalities.Results: The correlation coefficients were significantly higher in color Doppler M-mode versus tradi-tional M-mode in all subgroups analyzed. Conclusions: Our findings indicate that the color Dopplertechnique is more reproducible than conventional gray-scale M-mode technique to assess SPWMD.These findings suggest that the color Doppler M-mode may be a superior marker of the SPWMD mea-surement and predictor of left ventricular (LV) remodeling after cardiac resynchronization therapy.(ECHOCARDIOGRAPHY, Volume 25, September 2008)

echocardiography, M-mode, color M-mode, reproducibility, cardiac resynchronization therapy

In patients with advanced nonischemic heartfailure (HF), cardiac resynchronization therapy(CRT) is thought to improve left ventricularejection fraction (LVEF) by reducing factors as-sociated with left ventricular delay, includingreverse remodeling, mitral regurgitation, andneurohormonal changes.1–5 Baseline septal-to-posterior wall motion delay (SPWMD), mea-sured with gray-scale M-mode echocardiogra-phy (MM), is a predictor of patients who arelikely to benefit from CRT.6,7 The conventionaldiagnostic method for determining the SPWMDhas been MM in which the delay between themotion of the septum and left posterior wall

Address for correspondence and reprint requests: VincentL. Sorrell, M.D., Professor of Clinical Medicine and Radi-ology, Allan C. Hudson and Helen Lovaas Chair of Car-diovascular Imaging, Department of Medicine/Cardiologyand the Sarver Heart Center, University of Arizona, 1501N. Campbell Ave., Tucson, AZ 85724. Fax: 520-626-4333;E-mail: [email protected]

†William D. Ross is in Medical Student Research Program.

is measured.8 A delay greater than 130 ms isconsidered abnormal and consistent with in-traventricular mechanical dyssynchrony. Otherinvestigators have suggested that MM is not ad-equate in predicting clinical or remodeling re-sponse to CRT and therefore should not be usedas a guide for CRT patient selection.9,10 Thesediffering conclusions may in part be due to thefact that MM has a high degree of inter- andintraobserver variability, especially in patientswith ischemic cardiomyopathies and associatedwall motion abnormalities.

Color Doppler M-mode echocardiography(CMM) is a relatively recent imaging tech-nique that has the ability to visualize subtlechanges in motion with high temporal and spa-tial resolution along the entire length of theLV throughout the cardiac cycle. This techniqueis displayed in a graphic parametric color for-mat. CMM has the potential to be the pre-ferred diagnostic tool for the assessment ofSPWMD to predict the response to the CRT in-tervention. This study compared the intra- and

Vol. 25, No. 8, 2008 ECHOCARDIOGRAPHY: A Jrnl. of CV Ultrasound & Allied Tech. 821

SORRELL, ET AL.

TABLE I

Patient Characteristics (n = 31)

Age (year) 75.5 ± 4LVEF (%) 25.5% ± 4Male, n (%) 23 (74)Ischemic etiology, n (%) 15 (48)QRS duration (ms ± s.d.) 169 ± 6Heart rate (bpm) 77 ± 3LV diameter (mm) 58 ± 12Prior cardiac surgery, n (%) 12 (39)Pulmonary hypertension, n (%) 8 (26)

interobserver variability of CMM to determineif this is a more reproducible method thanthe conventional gray-scale MM measurement.

Methods

Conventional echocardiograms were per-formed using commercially available echocar-diographic systems (Philips Sonos 7500 or IE33,Philips Medical Systems, Andover, MA, USA;or GE Vivid 7 Dimension 2006, GE Healthcare,Milwaukee, WI, USA). Retrospective CMM ul-trasound recordings in the parasternal short-axis view with a sweep speed of 100 mm/secwere selected in 31 patients diagnosed with acardiomyopathy (Table I; patient characteris-tics). All patients had both CMM as well as MMimages obtained and 11 patients, had two com-parative datasets available for each (n = 42).SPWMD measurements were performed by twoexperienced cardiologists (readers 1 and 2; inFigs. 3, 4, 5, and 6). CMM images were inter-preted without knowledge of the MM measure-ments and vice versa. Eight randomly selectedimage datasets were reanalyzed by both read-ers blinded to the earlier results for analysisof intraobserver reproducibility. This study wasapproved by the University IRB committee forhuman research.

Conventional M-Mode Measurementsof SPWMD

As previously described, parasternal short-axis images were utilized to determine theSPWMD.5 This delay was calculated as theshortest interval between the initial, maximalposterior deflection, after the QRS complex, ofthe anterior-septal wall segment and the maxi-mal inward deflection of the posterior wall seg-ment (Fig. 1). When more than one SPWMDmeasurement (one for each QRS complex) was

Figure 1. MM SPWMD technique. Parasternal short-axistwo-dimensional guided conventional gray-scale M-modeimage. The SPWMD is shown. There is severe hypokine-sia/akinesia of the septum resulting in an imprecise pointof measure.

available, the results were averaged and usedfor subsequent analysis.

Color M-Mode Measurements of SPWMD

Ten initial CMM image datasets of patientsdiagnosed with cardiomyopathy and meetingthe study population criteria (Table I), were in-terpreted jointly with both readers to create astandardized protocol to minimize potential dis-crepancies in this novel technique. These im-ages were not included for analysis in this in-vestigation. CMM images were then analyzedas follows: (1) SPWMD was measured as thetime difference between the maximum poste-rior deflection of the septum as indicated bythe change in color from blue to red (blue/redinterface), and maximum anterior deflection ofthe posterior wall as indicated by the changein color from red to blue (red/blue interface);(2) when the images were inconsistent, andcardiac translation from a respiratory artifactcould be confirmed (shifting of the entire heartwithin the imaging plane), the cardiac cyclewas excluded (n = 7); (3) when there weretwo or more blue/red septal interfaces, the ini-tial interface after the end of the QRS com-plex was utilized; (4) if the interface was notcompletely linear within the myocardium, butinstead was tangential, the measurement wastaken from the endocardial interface and notthe mid-myocardial or epicardial origin; and (5)if the interface was not transmural, but had re-gions of interrupted colors, cursor position wasdetermined by the color transition at a depthof 1 cm from the wall with interrupted color(Fig. 2).

822 ECHOCARDIOGRAPHY: A Jrnl. of CV Ultrasound & Allied Tech. Vol. 25, No. 8, 2008

COLOR M-MODE REPRODUCIBILITY

Figure 2. CMM SPWMD technique. Parasternal short-axistwo-dimensional guided color Doppler M-mode paramet-ric display in the same patient as Figure 1. The SPWMDis shown. Despite the wall motion abnormality of the sep-tum, the point of measurement is readily identified by theblue/red interface.

As a SPWMD >130 ms is consideredthe threshold for determining intramyocar-dial dyssynchrony, analysis of the interob-server agreement of this specific value was alsoassessed.

Calculations and Statistics

Inter- and intraobserver reproducibility wasevaluated using linear regression. Correlationsbetween variables were assessed using Pear-son’s linear correlation. All P-values were two-sided and a P-value <0.01 was considered sta-tistically significant. All statistical tests wereperformed using the SPSS 15.0.1 statisticalpackage (SPSS, Inc., Chicago, IL, USA).

Results

Reproducibility of LV dyssynchrony assess-ment: MM versus CMM, a total of 31 patientswere evaluated and this provided 42 SPWMDmeasurements. Eight randomly selected im-ages were also remeasured by each of the ex-pert cardiologists. In total, therefore, 100 imagedatasets were utilized for this investigation.

The average SPWMD measured 0.145 ± 0.06second for reader 1 and 0.138 ± 0.06 sec-ond for reader 2. In MM displays, the SP-WMD was 0.143 ± 0.06 second and 0.137± 0.05 second and for CMM displays it was0.145 ± 0.07 second and 0.140 ± 0.06 sec-ond, for readers 1 and 2 (in Figs. 3, 4,5, and 6), respectively. For all patients, theinter- and intraobserver correlation coefficient

Figure 3. A, B. MM variability: correlation between SP-WMD measurements made by different observers: (a) inter-observer and (b) intraobserver. C, D. CMM variability: cor-relation between SPWMD measurements made by differentobservers: (c) interobserver and (d) intraobserver.


SORRELL, ET AL.

Figure 4. A, B. Interobserver variabilityBland–Altman display of the interobservervariability for MM (A) and CMM (B) mea-surements.

was 0.575 (P < 0.01) and 0.714 (P < 0.01) forMM and 0.929 (P < 0.01) and 0.967 (P < 0.01)for CMM (Figs. 3 and 4). In patients with severehypokinesia or akinesia of the septal myocar-dial segment (n = 20), the interobserver corre-lation coefficient was 0.510 (P < 0.01) for MMdisplays and 0.914 (P < 0.01) for CMM displays(Fig. 5).

Disagreement on the candidacy for CRT (SP-WMD >130ms) was found in 26.19% of MM as-sessments, but only 2.38% of CMM assessments(Fig. 6).

Discussion

CRT, using biventricular pacer implanta-tion, has become a reproducible method to im-prove the LV function and symptoms of se-vere HF in patients with a markedly reducedLVEF and a left bundle branch block pat-tern electrocardiogram. With current eligibil-ity criteria using a wide QRS complex to es-tablish ventricular dyssynchrony, 30–40% ofpatients do not respond to CRT. Given thishigh rate of CRT failure, a new and more


COLOR M-MODE REPRODUCIBILITY

Figure 5. A, B. Akinetic subgroup: correlation between SP-WMD measurements made by two different observers for pa-tients with severe hypokinesia or akinesia of the interventric-ular septal wall: (A) MM and (B) CMM.

predictable method to quantify mechanicaldyssynchrony is required.14,15 A number ofechocardiographic markers, both simple andcomplex, have been recommended to improveupon the QRS determination of LV mechani-cal dyssynchrony. Until outcome data assess-ing each of these parameters in the same pa-tient population is published, the most accurateparameter in predicting LV remodeling and re-covery of function or significant symptomaticimprovement remains controversial.11 Prelim-inary reports from the PROSPECT trial revealthat the echocardiographic measurements ofdyssynchrony were unable to accurately pre-dict response to CRT.12,13 A critical finding inthis large, multicenter study was the high in-tercore lab variability which ranged from 6.5%to 72.1%. Given the expertise of these core labs,simplified methods with higher reproducibility

Figure 6. Agreement: instances of agreement between read-ers 1 and 2 for detecting a prolonged SPWMD >130 ms.

seem indicated and studies into technique vari-ability seem warranted. One of the earliest pub-lished and relatively simple echo markers ofsuccessful CRT is the M-mode technique of mea-suring the SPWMD, commonly referred to asthe Pitzalis method. Recent studies, however,have suggested that MM-derived SPWMD val-ues are less effective predictors of clinical re-sponse to CRT than earlier reports.9,10 This maybe attributed to the different clinical popula-tions studied, slight differences in technique,differences in image quality, or the reportedpoor reproducibility of this measurement. Thepresent study suggests that, in general, the re-producibility of the SPWMD value can be im-proved with CMM compared to MM.

A significant limitation to accurate measure-ment of SPWMD in MM has been the presenceof regional wall motion abnormalities, namelyseptal wall severe hypokinesia or akinesia. Al-though this may adversely affect MM SPWMDevaluation, the ability of CMM to color code thedirection of even subtle myocardial contractionsallows for visualization of septal wall inward de-flection even in instances of severe hypokinesiaor akinesia. Whether this will contribute to aprediction of successful CRT remains unknown.

This investigation evaluated the CMM tech-nique that is increasingly being recommendedby experts in this field.16 The specifics of thisCMM methodology have not been publishedand a review of the medical literature failed toidentify the reproducibility of this color-coded,Doppler technique. Although it remains to beproven if this technique is reproducible at otherlaboratories, our technique described above ap-pears valid.


SORRELL, ET AL.

Limitations

This investigation should be considered pre-liminary and hypothesis-generating given thesmall number of patients and the retrospectivedesign. The patient population was not selectedand included patients with both ischemic andnonischemic cardiomyopathies. The use of colorDoppler MM displays is very sensitive to trans-lation artifacts and datasets should be obtainedduring breath hold acquisition. It is possiblethat translation artifacts were not excluded andwere inadvertently included in this analysis.The CMM protocol used for the measurementof the SPWMD was determined without valida-tion or clinical outcome data.

Conclusions

In a random clinical sample of 100 datasetsblindly reviewed by two independent observers,the SPWMD interval can be more reproduciblymeasured using color Doppler guided M-modethan conventional gray scale. This study sug-gests that the parametric display of color M-mode is superior to traditional gray-scale M-mode in precisely measuring the SPWMD.These findings warrant further investigation ofthe use of color Doppler M-mode as a potentialsuperior marker of the SPWMD and predictor ofLV remodeling after CRT. It remains unknownif this will be a more accurate predictor of LVremodeling after CRT.

Acknowledgments: The authors greatly appreciate Dr.Frank Marcus’ critical review or our manuscript. Dr. Sor-rell’s work is in part supported from the Allan C. Hudsonand Helen Lovaas Endowment.

References

1. Cazeau S, Leclercq C, Lavergne T, et al: Effects ofmulti-site biventricular pacing in patients with heartfailure and intraventricular conduction delay. N EnglJ Med 2001;344:873–880.

2. Saxon LA, De Marco T, Schafer J, et al: Effects of long-term biventricular stimulation for resynchronization

on echocardiographic measures of remodeling. Circu-lation 2002;105:1304–1310.

3. Auricchio A, Stellbrink C, Sack S, et al: Long-termclinical effect of hemodynamically optimized cardiacresynchronization therapy in patients with heart fail-ure and ventricular conduction delay. J Am Coll Car-diol 2002;39:2026–2033.

4. Abraham WT, Fisher WG, Smith AL, et al: Cardiacresynchronization in chronic heart failure. N Engl JMed 2002;346:1845–1853.

5. Young JB, Abraham WT, Smith AL, et al: Combinedcardiac resynchronization and implantable cardiover-sion defibrillation in advanced chronic heart failure:The MIRACLE ICD trial. JAMA 2003;289:2685–2694.

6. Pitzalis MV, Iacoviello M, Romito R, et al: Car-diac resynchronization therapy tailored by echocar-diographic evaluation of ventricular asynchrony. J AmColl Cardiol 2002;40:1615–1622.

7. Pitzalis MV, Iacoviello M, Romito R, et al: Ventricularasynchrony predicts a better outcome in patients withchronic heart failure receiving cardiac resynchroniza-tion therapy. J Am Coll Cardiol 2005;45:65–69.

8. Pitzalis MV, Iacoviello M, Romito R, et al: Echocardio-graphic left ventricular asynchrony predicts long-termbenefit of resynchronization therapy (abstr). Circula-tion 2003;108:IV347.

9. Marcus GM, Rose E, Viloria EM, et al: Septal to poste-rior wall motion delay fails to predict reverse remod-eling or clinical improvement in patients undergoingcardiac resynchronization therapy. J Am Coll Cardiol2005;46:2209–2214.

10. Gabriel RS, Bakshi TK, Scott AG, et al: Reliability ofechocardiographic indices of dyssynchrony. Echocar-diography 2007;24:40–46.

11. Dı́az-Infante E, Sitgres M, Vidal B, et al: Usefulnessof ventricular dyssynchrony measured using M-modeechocardiography to predict response to resynchro-nization therapy. Am J Cardiol 2007;100:84–89.

12. Yu C, Abraham WT, Bax J, et al: Predictors of responseto cardiac resynchronization therapy (PROSPECT)–study design. Am Heart J 2005;149:600–605.

13. Ghio S: Results of the predictors of response to CRT(PROSPECT) trial. European Society of CardiologyCongress, September 4, 2007; Vienna. Hotline 3.

14. Kass DA: An epidemic of dyssynchrony: But what doesit mean? J Am Coll Cardiol 2008;51:12–17.

15. Nagueh SF: Mechanical dyssynchrony in congestiveheart failure: Diagnostic and therapeutic Implica-tions. J Am Coll Cardiol 2008;51:18–22.

16. Gorcsan J, Abraham T, Agler D, et al: Echocardiog-raphy for cardiac resynchronization therapy: Recom-mendations for performance and reporting–a reportfrom the ASE dyssynchronny writing group. ASE Ex-pert Consensus Statement. J Am Soc Echocardiogra-phy 2008;21:191–213.


continuing medical education program in echocardiography

Documents