Unique Features of Early Diastolic MitralAnnulus Velocity in Constrictive Pericarditis

Dae-Won Sohn, MD, Yong-Jin Kim, MD, Hyo-Soo Kim, MD, Ki-Bong Kim, MD,Young-Bae Park, MD, and Yun-Shik Choi, MD, Seoul, Korea

Background: In most primary myocardial diseases,early diastolic mitral annulus velocity (E’) decreaseswith disease progression. To our knowledge, con-strictive pericarditis (CP) is the only condition with-out this phenomenon.Objective: This study was performed to evaluate thediagnostic and pathophysiologic role of mitral an-nulus velocity in patients with CP.Methods: In all, 17 patients with CP (9 men; meanage 46.5 � 14.3 years), 8 patients with cardiactamponade (Tamp) (2 men; mean age 44.5 � 15.0years), and age- and sex-matched control subjectsfor CP and Tamp were recruited for the study. Earlymitral inflow velocity and E’ were obtained whilesimultaneously recording respiration. In 8 patientswith CP and in all patients with Tamp, these mea-surements were repeated after the relief of constric-tive physiology or after pericardiocentesis.Results: In patients with CP, E’ was significantlyhigher than it was for control subjects (12.9 � 3.0

222

cm/s vs 9.8 � 2.4 cm/s, P < .01). An E’ of 2 cm/shigher than the predicted normal E’ could differen-tiate patients with CP from control subjects with asensitivity of 76% and specificity of 82%. In 12 of 17patients (71%), inspiratory E’ was higher than expi-ratory E’–the opposite of mitral inflow variation. In8 patients, E’ decreased significantly after the reliefof constrictive physiology (13.8 � 2.5 cm/s vs 9.3 �3.1 cm/s, P < .05). In contrast, E’ in the Tamp groupwas significantly lower than in the control group(6.8 � 1.6 cm/s vs 10.2 � 2.5 cm/s, P < .01), did notshow significant respiratory variation, and in-creased significantly after pericardiocentesis (6.8 �1.6 cm/s vs 9.5 � 3.0 cm/s, P < .05).Conclusion: E’ is exaggerated in CP, which is helpfulfor diagnosis. The opposite phenomenon was notedin Tamp, a dissimilarity that might contribute todifferent hemodynamics. (J Am Soc Echocardiogr2004;17:222-6.)

Early diastolic mitral annulus velocity (E’) mea-sured by Doppler tissue imaging is a relativelyload-independent parameter that represents myocar-dial relaxation.1 Therefore, in most primary myocar-dial diseases, E’ decreases with disease progression.However, in constrictive pericarditis (CP), E’ is wellpreserved and this phenomenon is reported to beuseful in the differential diagnosis of CP and restric-tive cardiomyopathy.2 Also, in a recent report, Ha etal3 suggested that E’ is helpful for the diagnosis of CPwhen respiratory variation in mitral inflow velocityis absent. This study was performed to evaluate thediagnostic and pathophysiologic role of E’ in CP.

From the Clinical Research Institute and Division of Cardiology,Department of Internal Medicine, and Department of ThoracicSurgery (K-B.K.), Seoul National University College of Medicine.Supported in part by grant 2001-041-F00127 from the KoreaResearch Foundation.Reprint requests: Dae-Won Sohn, MD, Division of Cardiology,Department of Internal Medicine, Seoul National University Col-lege of Medicine, 28 Yongon-Dong, Chongno-Gu, Seoul 110-744, Korea (E-mail: dwsohn@snu.ac.kr).0894-7317/$30.00Copyright 2004 by the American Society of Echocardiography.doi:10.1067/j.echo.2003.11.015

METHODS

Study Patients

From November 2000 to September 2002, 17 patients withCP who were in sinus rhythm and admitted to our hospitalwere prospectively enrolled. The mean age of the patientswas 46.5 � 14.3 years and 9 were men. All patients showedsymptoms and signs of elevated central venous pressure andall except 1 showed thickened pericardium on computedtomography or magnetic resonance imaging. CP was diag-nosed on the basis of Doppler echocardiographic findings,4-6

thickened pericardium on computed tomography or mag-netic resonance imaging, and surgical findings in patientswho underwent pericardiectomy. Causes of the CP weretuberculosis in 7, idiopathic in 7, encasement as a result ofmalignant tumor in 2, and insufficient previous pericardiec-tomy in 1. In 8 patients with CP (7 patients who underwentpericardiectomy and 1 patient whose constrictive physiol-ogy resolved after antituberculosis therapy), Doppler echo-cardiographic examination was repeated after the constric-tive physiology had resolved.

To compare the difference between CP and cardiactamponade (Tamp), E’ was obtained before and after peri-cardiocentesis in 8 patients with Tamp. Mean age of thepatients was 44.5 � 15.0 years and 2 patients were men. Age-

Journal of the American Society of EchocardiographyVolume 17 Number 3 Sohn et al 223

and sex-matched control subjects for CP and Tamp wererecruited as control groups.

Echocardiography

Echocardiograms were obtained using a commerciallyavailable device (XP/10, Acuson, Mountain View, Calif)with a 2.5-MHz transducer. After the M-mode and 2-dimen-sional examination, a sample volume (2 mm in size) of thepulsed wave Doppler was placed between the tips of themitral leaflets on the apical 4-chamber view and in thehepatic vein on the subxiphoid view. Mitral inflow andhepatic vein flow were obtained while simultaneouslyrecording respiration using a nasal thermistor. For themitral inflow, early transmitral flow velocity (E) and latetransmitral flow velocity (A), the ratio of early to late peakvelocities (E/A), and the deceleration time of E wereobtained both during inspiration and expiration. Respira-tory variation of E was calculated from: (maximum E �minimum E)/minimum E � 100 (%) in 1 respiratory cycle.

Pulsed wave Doppler tissue imaging was performed byactivating the Doppler tissue imaging function in the samemachine. A sample volume (2-4 mm in size) was located atthe septal side of mitral annulus. Early diastolic mitralannulus velocity (E’) and late diastolic mitral annulusvelocity (A’), and the ratio of early to late peak velocities(E’/A’), were obtained while simultaneously recordingrespiration using a nasal thermistor. Similar to the mitralinflow, respiratory variation of E’ was calculated from:(maximum E’ � minimum E’)/minimum E’ � 100 (%) in 1respiratory cycle. Higher values between inspiratory E’and expiratory E’ were used for the comparison withcontrol subjects. Predicted normal E’ according to age wascalculated using the regression equation obtained fromthe control subjects included in our previous study1:predicted E’ � 16 � 0.14 � age (cm/s).

Doppler echocardiograms were recorded on a stripchart with a sweep speed of 100 mm/s and the meanvalues of 3 different cardiac cycles were obtained.

Statistics

Data are expressed as means � SD. The Mann-Whitneytest was used to compare echocardiographic continuousvariables between the patient and control groups. Wil-coxon signed rank test was used to compare echocardio-graphic parameters before and after pericardiectomy orpericardiocentesis, and to compare inspiratory and expi-ratory Doppler variables in each patient. All statisticalanalyses were performed with commercially availablesoftware (SPSS 10.0, SPSS Inc, Chicago, Ill). A P value ofless than .05 was considered statistically significant.

RESULTS

Mitral Inflow and Mitral Annulus Velocityin CP

Patients with CP showed a mean 35.4 � 18.0%(range 10-71) respiratory variation in E. In all pa-

tients, expiratory E was higher than inspiratory E.However, in 5 of 17 patients (29%), the respiratoryvariations were less than 25%. Patients with CP hadsignificantly higher E’ than the age- and sex-matchedcontrol subjects (12.9 � 3.0 cm/s vs 9.8 � 2.4 cm/s,P � .005). Moreover, an E’ of 2 cm/s higher thanpredicted differentiated patients with CP from con-trol subjects with a sensitivity of 76% and a specific-ity of 82%.

E’ showed respiratory variation. However, in themajority of patients (12 of 17; 71%), inspiratory E’was higher than expiratory E’–which was oppositeto that observed for E. Although statistically notsignificant, patients with this type of respiratory E’and E variation (“out of phase”) tended to havehigher respiratory E variation than those patientswho showed same respiratory E’ and E variations(“in phase”), (38.9 � 20.2% vs 25.8 � 4.9%, P � .21)(Figure 1).

In 8 patients in whom constrictive physiology wasrelieved, E’ decreased significantly after the relief ofthe constrictive physiology (13.8 � 2.5 cm/s vs 9.3� 3.1 cm/s, P � .005) (Figure 2).

Difference Between CP and Tamp

In contrast to CP, patients with Tamp had signifi-cantly lower E’ values than the age- and sex-matchedcontrol subjects (6.8 � 1.6 cm/s vs 10.2 � 2.5 cm/s,P � .01), and did not show significant respiratoryvariation. After pericardiocentesis, E’ increased sig-nificantly (6.8 � 1.6 cm/s vs 9.5 � 3.0 cm/s, P �.05) (Figure 3).

Figure 1 Patients showing in phase respiratory variation inearly mitral inflow velocity and early diastolic mitral annu-lus velocity (expiratory velocities higher than inspiratoryvelocities) tended to show small respiratory variation thanthose in whom these variations were out of phase.

Journal of the American Society of Echocardiography224 Sohn et al March 2004

DISCUSSION

Previous studies suggested that E’ in CP was notblunted.2,3 However, it has not been clearly demon-strated whether E’ in CP is just preserved or exag-gerated. In our study, E’ in CP was found to besignificantly higher than in age- and sex-matchedcontrol subjects. In addition, E’ in these patientsdecreased after pericardiectomy. Therefore, E’ in CPis exaggerated rather than just preserved. In thisstudy, when E’ was more than 2 cm/s higher than itspredicted value, it could differentiate patients withCP from the age- and sex-matched control subjectswith a sensitivity of 76% and a specificity of 82%.Because E’ is reduced in most primary myocardialdiseases, its diagnostic power would be higher in thedifferentiation of CP from other disease conditions.

As might be expected, exaggerated E’ resulted inan exaggerated mitral annulus excursion in CP,which decreased after pericardiectomy. In contrast,patients with Tamp showed reduced E’ and annulusexcursion, which increased after pericardiocentesis(Figure 4). In CP, as Ha et al7 suggested, lateralexpansion is limited and there is a compensatoryincrease in longitudinal expansion. However, forpatients with Tamp, the heart is pushed in the basal

Figure 2 Mitral annulus velocity (E’) for patients withconstrictive pericarditis before (A) and after (B) pericardi-ectomy. Before pericardiectomy, inspiratory E’ was higherthan expiratory E’–which was opposite to that of respira-tory mitral inflow variation. Significant decrease was ob-served in E’ after pericardiectomy.

direction by pericardial effusion and there is nocompensatory increase in longitudinal expansion.Invasive hemodynamics were not obtained duringour study, but a previous study7 showed an inverserelationship between the E/E’ ratio and pulmonarycapillary wedge pressure, which suggests that theremight be an accentuated compensatory mechanismon disease progression.

In the majority of patients with CP (13 of 17;76%), E’ varied in the opposite manner to mitralinflow with respect to respiratory variation. In ouropinion, this out of phase respiratory variation has 2clinical implications. First, although paradoxicalpulse was originally described in CP by Kussmaul,8

paradoxical pulse is observed less frequently in CPthan Tamp.9,10 Various plausible theories have beensuggested but the issue remains unresolved. In 1theory, an inspiratory decrease in intrathoracic pres-sure is transmitted to the interior of the heart inTamp, and leads to an inspiratory increase in volumeof the right ventricle. This increase in the volume ofthe right ventricle decreases the volume of the leftventricle, which leads to reduced stroke volumeduring inspiration. However, in CP, an inspiratorydecrease in intrathoracic pressure is not transmittedto the interior of the heart. Therefore, increasedsystemic venous return during inspiration reflects as

Figure 3 Mitral annulus velocity (E’) for patients withcardiac tamponade before (A) and after (B) pericardiocen-tesis. Respiratory variation of E’ was not apparent beforepericardiocentesis. There was increase in E’ after pericar-diocentesis.

Journal of the American Society of EchocardiographyVolume 17 Number 3 Sohn et al 225

an increase in the systemic venous pressure, theso-called Kussmaul’s sign, and does not lead to anincrease in the volume of the right ventricle. Wehypothesize that the out of phase respiratory varia-tions between E’ and E, which were observed in themajority of patients with CP, contribute to theabsence of a paradoxical pulse in CP. An inspiratorydecrease in short-axis dimension is well recognized.Therefore, an opposite change in the long-axis di-mension could lessen the respiratory variation ofstroke volume by reducing the respiratory variationof diastolic volume. Second, because the Doppler-measured mitral flow velocity is the vectorial sum ofthe cardiac base velocity directed to the atrial sidewith the mitral inflow velocity directed to thecardiac apex11 (Figure 5), the inspiratory increase inE’ will lower the inspiratory mitral E, thus, accentu-ating respiratory mitral inflow variation. In thisstudy, patients with a respiratory E’ variation out ofphase with mitral E variation tended to show higherrespiratory variation than patients whose variationswere in phase (Figure 6). Although statistically sig-nificance was not reached, probably because of thesmall number of patients in whom respiratory E’ andE variations were in phase and also because otherfactors are involved in the mechanism of respiratoryvariation, this phenomenon could contribute to theabsence of respiratory variation in a small number ofpatients6 despite applying the preload reductionmaneuver.12

Limitations

Our hypothesis that there is less respiratory changein stroke volume in CP than in Tamp cannot be

Figure 4 Mitral annulus motion for patients wit(bottom). As expected from mitral annulus velocittomy for patients with constrictive pericarditis bpericardiocentesis.

verified by measuring stroke volume because mea-suring respiratory variation in time velocity integralof left ventricular outflow tract velocity was notfeasible, as the sample volume location cannot bestabilized during respiration. In CP, we only in-

trictive pericarditis (top) and cardiac tamponadeern, annulus excursion reduced after pericardiec-eased for patients with cardiac tamponade after

Figure 5 Doppler-measured mitral inflow velocity (V2) isvectorial sum of mitral inflow net velocity (V1) and mitralannulus velocity (V3): V2 � V1 � V3. Therefore, opposingrespiratory variations of mitral annulus and mitral inflowvelocities increase respiratory variation of mitral inflowmeasured by Doppler. LA, Left atrium; LV, left ventricle.

h consy pattut incr

Journal of the American Society of Echocardiography226 Sohn et al March 2004

cluded patients with in sinus rhythm, therefore, thediagnostic role of E’ cannot be extended to patientswith atrial fibrillation.

Conclusions

This study shows that exaggerated E’ is helpful inthe diagnosis of CP. In addition, respiratory variationof E’ in CP could possibly explain the absence ofrespiratory mitral inflow variation in CP and the lessfrequent observation of paradoxical pulse in CP thanin Tamp.

REFERENCES

1. Sohn DW, Chai IH, Lee DJ, Kim HC, Kim HS, Oh BH, et al.Assessment of mitral annulus velocity by Doppler tissue imag-ing in the evaluation of left ventricular diastolic function. J AmColl Cardiol 1997;30:474-80.

2. Garcia MJ, Rodriquez L, Ares M, Griffin BP, Thomas JD,Klein AL. Differentiation of constrictive pericarditis from re-strictive cardiomyopathy: assessment of left ventricular dia-stolic velocities in longitudinal axis by Doppler tissue imaging.J Am Coll Cardiol 1996;27:108-14.

3. Ha JW, Oh JK, Omnen SR, Ling LH, Tajik AJ. Diagnosticvalue of mitral annular velocity for constrictive pericarditis inthe absence of respiratory variation in mitral inflow velocity.

Figure 6 Mitral inflow and mitral annulus velorespiratory mitral annulus velocity variation is ininflow variation. Respiratory variation of mitral infof phase.

J Am Soc Echocardiogr 2002;15:1468-71.

4. Hatle LK, Appleton CP, Popp RL. Differentiation of constric-tive pericarditis and restrictive cardiomyopathy by Dopplerechocardiography. Circulation 1989;79:357-70.

5. Appleton CP, Hatle LK, Popp RL. Demonstration of restric-tive ventricular physiology by Doppler echocardiography.J Am Coll Cardiol 1988;11:757-68.

6. Oh JK, Hatle LK, Seward JB, Danielson GK, Schaff HV, ReederGS, et al. Diagnostic role of Doppler echocardiography in con-strictive pericarditis. J Am Coll Cardiol 1994;23:154-62.

7. Ha JW, Oh JK, Ling LH, Nishimura RA, Seward JB, Tajik AJ.Annulus paradoxus: transmitral flow velocity to mitral annulusvelocity ratio is inversely proportional to pulmonary capillarywedge pressure in patients with constrictive pericarditis. Cir-culation 2001;104:976-8.

8. Kussmaul A. Uber schwielige mediastinopericarditis und denparadoxen pulse. Klin Wochenschr 1873;10:433-5.

9. Lange RL, Botticelli JT, Tsagaris TJ, Walker JA, Gani M,Bustamante RA. Diagnostic signs in compressive cardiac dis-orders: constrictive pericarditis, pericardial effusion and tam-ponade. Circulation 1966;33:763-77.

10. Hancock EW. On the elastic and rigid form of constrictivepericarditis. Am Heart J 1980;100:917-23.

11. Isaaz K, del Romeral LM, Lee E, Schiller NB. Quantitation ofthe motion of the base in normal subjects by Doppler echo-cardiography. J Am Soc Echocardiogr 1993;6:166-76.

12. Oh JK, Tajik AJ, Appleton CP, Hatle LK, Nishimura RA,Seward JB. Preload reduction to unmask the characteristicDoppler features of constrictive pericarditis: a new observa-

or patients with constrictive pericarditis whose(A) and out of phase (B) with respiratory mitrallocity is much lower for patients in phase than out

city fphaselow ve

tion. Circulation 1997;95:796-9.