Refining the criteria for pulsed doppler diagnosis of mitral regurgitation by comparison with left ventricular angiography

Refiningthe Criteria for PulsedDopplerDiagnosis of Mitral Regurgitationby Comparison with left VentricularAngiography TZE-YU DANG, MD, JULIUS M. GARDIN, MD, SANDRA ALICE ALLFIE, and WALTER L. HENRY, MD

specificity was 92 % . Using the combined criteria of holosystolic increased spectral dispersion plus peak regurgitant blood flow velocity of >150 cm/s, sensitivity was 100 % and specificity was 97 % . Doppler studies in 20 clinically normal persons showed that the combined Doppler criteria were 100% specific. In conclusion, the presence of holosystolic increased spectral dispersion within the left atrium with a peak regurgitant flow vetocity > 150 cm/s appears to be as sensitive as and more specific than the presence of any systolic Doppler increased spectral dispersion in identifying patients with MR.

Doppler echocardiography is established as a useful method for detecting mitral regurgitation (MR). However, studies also show that Doppler echocardiography may be too sensitive for diagnosing pathologic MR. To determine whether the false-positive or clinically insignificant diagnosis of MR can be eliminated by defining more specific diagnostic criteria, pulsed Doppler findings were compared with left ventricular angiographic findings in 81 patients. Using the conventional Doppler criterion of any systolic increased spectral dispersion detected in the left atrium, sensitivity was 100% and specificity was 76 % . Using holosystolic increased spectral dispersion as the criterion, sensitivity was 100% and

T

(Am J Cardiol 1987;60:663-66

he usefulness of Doppler echocardiography for diagnosis of mitral regurgitation (MR) is well established.l-lz Comparisons of pulsed-wave Doppler to left ventricular angiography for detection of MR have showed an excellent correlation.11J2 In studies in which either M-mode or &ditiensionally directed pulsed Doppler sample volumes have been used, sensitivity and specificity range from 90 to 100%. Recent reports a&o suggest a high incidence of false-positive studies for MR in normal persons.13J4 Mitral regurgitation has been detected in as many as 10% of reportedly normal persons using continuouswave Doppler echocardiography. It has been suggested that Doppler echocardiography, especially the conFrom the Division of Cardiology, Department of Medicine, University of California, Irvine Medical Center, Orange, California. Manuscript received January 16, 1987; revised manuscript received and accepted May 4,1987. Address for reprints: Julius M. Gardin, MD, Division of Cardiology, University of California, University of California Irvine Medical Center, 101 City Drive South, Building 53, Route 81, Orange, California 92668.

CLARK,

tinuous-wave technique, may be too sensitive for diagnosing pathologic MR. Therefore, caution may be necessary in diagnosis of MR by the Doppler method. This study determines whether the false-positive or clinically insignificant diagnosis of MR can be eliminated by defining more specific diagnostic criteria.

Methods Patients: The study population consisted of 81 patients (49 men, 32 women, mean age 50 years) who underwent cardiac catheterization for a variety of clinical indications and had acceptable-quality contrast left ventriculograms and Doppler studies. Patients with atria1 fibrillation and prosthetic valves were excluded. Clinical diagnoses of the 81 study patients are summarized in Table I. Forty-nine patients had coronary artery disease and 8 were diagnosed as normal. Twenty other persons (16 men, 4 women, mean age 40 years) without clinical evidence of cardiovascular disease were studied by M-mode, &dimensional and pulsed Doppler echocardiography but not by cardiac catheterization. None of these subjects showed mitral valve prolapse by M-mode or a-dimensional echocardiography.15J6

663

664

DOPPLER DIAGNOSIS

TABLE I

OF MITRAL REGURGITATION

TABLE II

Patient Profile

Coronary artery disease Rheumatic mitral disease MS MS/AR MSITR MSIASIAR Cardiomyopathy Hypertrophic Idiopathic dilated Mitral valve prolapse Congenital heart disease Normal AR = aortic regurgitation; = tricuspid regurgitation.

49

Detection

of Mitral Regurgitation Standard Doppler Criteria”

7

MR (+)

3 2 1 1

MR (3

MR (+) 5 Angio

1 4

MR C-1

6 6 a AS = aortic stenosis;

MS = mitral stenosis; TR

Doppler echocardiography: Each patient had a complete pulsed Doppler/Z-dimensional echocardiographic examination before or after cardiac catheterization. One of 2 commercially available ultrasound instruments (Biosound, Inc., or General Electric) with a spectrum analyzer-based Doppler velocimeter and a mechanicall (or phased-array) imaging transducer were used. Wall filters were set to reject signals with flow velocities of less than 10 f 5 cm/s. The time between!angiography and Doppler examination ranged from 0 to 21 days (mean 9). The left atrium was carefully mapped for increased spectral dispersion in both the apical 2- and 4-chamber views. Increased spectral dispersion of the flow signal was defined as a systolic left atria1 signal with a band widthgreater than 75 cm/s. The mapping technique consisted of placing the sample volume behind the mitral valve in the left atrium on the a-dimensional image and moving the sample volume in l-cm increments horizontally and vertically to search for the peak flow velocity. The duration and peak velocity of the regurgitant flow were measured using frame-by-frame analysis from the videotapes.

Refined Doppler Criteria1

MR (+)

MR(+)

MR(3

Angio MR (-)

*Doppler MR defined as the presence of any increased spectral dispersion in the left atrium during systole. TDoppler MR defined as the presence of holosystolic increased spectral dispersion plus a peak regurgitatant velocity >15Q cm/s. Angio = angiography; MR = mitral regurgitation; + = present; - = absent.

Left ventricular angiography: Left ventricular angiograms were recorded in the standard 30’ right anterior oblique view with iodinated contrast injections. Patients with inadequate studies because of premature beats or poor visualization of the left atrium were excluded. Mitral regurgitation was considered to be present when retrograde flow of the iodinated contrast material from the left ventricle to the left atrium was present on angiography. The angiographic severity of the regurgitation was not graded for purposes of this study. Data analysis: Both Doppler and angiographic studies were interpreted by 2 independent teams with 2 observers on each team. Doppler and angiographic results were compared to determine sensitivity and specificity of a positive Doppler study for predicting angiographic MR by the equations**: Sensitivity = (patients with both D and A for MR/all patients with angiographic MR) X 100%; specificity = (patients without D or A for MR)/all patients without angiographic MR] X 10070, where D = Doppler criteria and A = angiographic criteria.

Results

FIGURE 1. Doppler flow velocity recording in a patient with early systolic increased spectral dispersion in the left atrium. Arrows indicate the apparent peak (aliased) regurgitant velocity = 135 cm/s. Mitral regurgitation was not seen by left ventriculography.

Using the criterion that MR is present when any systolic increased spectral dispersion is detected in the left atrium, our results showed that pulsed Doppler correctly identified 10 of 10 patients with angiographitally documented MR and 54 patients without any evidence of regurgitation (Table II). No patients had positive findings for MR by angiography and negative findings by Doppler examination, i.e., Doppler was

September

15, 1987

100% sensitive for detection of angiographic MR. However, 17 patients had MR by this Doppler criterion but failed to demonstrate MR by angiography. Using angiography as the standard, these 17 patients with negative angiographic findings for MR constituted the false-positive Doppler group. Figures 1 and 2 are examples of Doppler recordings in this false-positive group. Of the 17 patients in this group, 11 showed early systolic Doppler increased spectral dispersion in the left atrium [Fig. 1). Only 1 of the 11 patients showed early systolic increased spectral dispersion with a peak regurgitant flow velocity greater than 150 cm/s. Four patients showed holosystolic increased spectral dispersion with a peak flow velocity of less than 150 cm/s (Fig. 21 and 2 patients with holosystolic increased spectral dispersion and a peak flow velocity of more than 150 cm/s. These observations suggest that the combined criteria of holosystolic increased spectral dispersion plus a peak regurgitant flow velocity of mmore than 150 cm/s is more specific than the conventional criterion, i.e., the mere presence of any increased spectral dispersion within the left atrium. To further examine the sensitivity of these Doppler criteria, 10 patients with angiographically documented MR were reexamined by Doppler echocardiography. All 10 showed holosystolic increased spectral dispersion with a peak flow velocity of more than 150 cm/s. An example of a Doppler recording from a patient with angiographically documented MR is shown in Figure 3. To examine the specificity of the criteria of holosystolic increased spectral dispersion and peak flow velocity of more than 150 cm/s for MR 20 normal persons were studied by pulsed Doppler echocardiography. Sixteen of the 20 normal persons showed no increased spectral dispersion in the left atrium. Three showed early systolic increased spectral dispersion and 1 sub-

FIGURE 2. Doppler flow velocity recording in a patient with previous myocardial infarction. Holosystolic increased spectral dispersion when a peak regurgitant flow velocity of less than 150 cm/s was recorded. Arrows indicate that apparent peak (aliased) regurgitant velocity = 125 cm/s. Mitral regurgitation was not seen by left ventriculography.

THE AMERICAN

JOURNAL

OF CARDIOLOGY

Volume 60

665

TABLE II1 Sensitivity and Specificity of Various Doppler Findings in the Diagnosis of Mitral Regurgitation

Any increased spectral dispersion Holosystoiic increased spectral dispersion kolosystolic increased spectral dispersion and PFV >I50 cm/s

Sensitivity

Specificity

100% 100%

76% 92%

100%

97%

PFV = peak flow velocity.

ject showed holosystolic increased spectral dispersion; all had peak flow velocity of less than 150 cm/s. None had holosystolic increased spectral dispersion with a peak flow velocity of more than 150 cm/s. In addition, none of these subjects demonstrated late systolic increased spectral dispersion. The sensitivity of all the following Doppler criteria for MR was 100%: presence of any increased spectral dispersion, holosystolic increased spectral dispersion or holosystolic increased spectral dispersion plus peak flow velocity more than 150 cm/s. However, the specificity for MR was significantly different using the different Doppler criteria: 36% for any increased spectral dispersion, 92% for holosystolic increased spectral dispersion and 97% for the combined criteria [Table III]. In 7 of the 17 patients in the Doppler false-positive group, increased spectral dispersion during systole could be mapped more than 2 cm behind the mitral valve plane into the left atrium. In 3 of these 7 patients, increased spectral dispersion could be mapped more than 3 cm (range 3 to 4.~4into the left atrium.

Discussion One method of assessing the severity of MR involves mapping the extension of the regurgitant signal behind the valve into the left atrium.ly5J1J2Jg How-

FIGURE 3. Pulsed Doppler flow velocity recording in the left atrium of a patient with holosystolic increased spectral dispersion (arrows) and a peak (aliased) regurgitant flow velocity of more than 150 cm/s. Mitral regurgitation was seen by left ventricutography.

666

DOPPLER DIAGNOSIS

OF MITRAL

REGURGITATION

ever, despite the high sensitivity of Doppler echocardiography in detecting MR, false-positive findings are not infrequent .13J4JoIt has been suggested that the Doppler technique is so sensitive that clinically insignificant regurgitation can be detected.13J4t20Whether this systolic flow pattern represents true or artifactual regurgitation is still uncertain.16,20 The current study is an attempt to refine the Doppler criteria for diagnosing MR so that the false-positive diagnoses [or the diagnosis of clinically insignificant regurgitation] can be eliminated. The combined criteria of left atria1 holosystolic increased spectral dispersion plus peak regurgitant flow velocity of more than 150 cm/s are more specific (97%) than other Doppler criteria for diagnosing MR The peak flow velocity of 150 cm/s was chosen because the true peak velocity could not be determined with the standard pulsed Doppler, either because of aliasing [exceeding the Nyquist limit) or high-velocity signal dropout.13 Use of the continuous-wave Doppler technique to measure true peak velocity in the regurgitant jet should pinpoint more precisely the velocity cutoff point. Nevertheless, peak regurgitant flow velocity must be more than 150 cm/s before the Doppler diagnosis of MR can be made reliably. In fact, Hatle showd that peak velocity of the regurgitant jet in MR is frequently as high as 400 to 600 cm/s. One may expect such high velocities because the peak systolic pressure gradient between the left ventricle and left atrium is often in 64 to 144 mm Hg. Our findings are consistent, in part, with the recent report of Kostucki et all3 that in 10 of 25 normal suban early systolic to midsystolic regurgitant jects (4O%), flow pattern was recorded at or within 1 cm of the mitral valve with a duration up to 60% (mean 15%) of systole. However, in 3 of our patients with increased systolic spectral dispersion in the left atrium by Doppler but no evidence of MR by angiography, the increased spectral dispersion was recorded more than 3 cm (range 3 to 4.5 cm] behind the mitral valve plane. We do not know the etiology of pulsed Doppler recordings of nonholosystolic increased spectral dispersion with low apparent peak regurgitant flow velocities. We cannot exclude the possibility that these Doppler recordings indicate true minimal MR, even in In normal persons, diastolic our normal subjects. 13,14~20 flow moving retrograde from the coaptation point of the pulmonic valve into the right ventricular outflow tract has been reported by Doppler color flow mapping.21 Angiography, which is clearly less than 100% sensitive for detection of MR, may simply not be sensitive enough for detecting MR.22 The limitations of left ventricular angiography in detection of MR include inability to visualize small amounts of contrast medium in the left atrium, especially in a large left atrium, and the possibility of masking regurgitation by the surrounding tissues.23In contrast to pulsed Doppler echocardiography, angiography is a dilutional method in which the concentration of regurgitant contrast material may build up over several cardiac cycles, especially in trivial regurgitation. Doppler echocardiography permits beat-to-beat detection of variation in regurgitant flow patterns.

Our Doppler criteria for diagnosing MR are based on use of pulsed Doppler echocardiography. The more sensitive continuous-wave Doppler technique may reveal velocities above 150 cm/s, and even a holosystolic signal [albeit 1 of generally low intensity] in patients who do not exhibit these characteristics by pulsed Doppler echocardiography. Acknowledgment: We thank Cynthia McDaniel for her expert preparation of the manuscript.

References 1. Miyatake K, Kinoshita N, Nagata S, Beppu S, Park YD, Sakakibara H, Nimura Y. Intracardiac flow pattern in mitral regurgitation studied with combined use of the ultrasonic pulsed Doppler technique and cross-sectional echocardiography. Am J Cardiol 1980;45:155-162. 2. Diebold B, Theroux P, Bourassa MG, Thuillez C, Peronneau P, Guermonprez JL, Xhaard M, Waters DD. Non-invasive Doppler study of mitral stenosis and mitral regurgitation: preliminary study. Br Heart J 1979;42:168-175. 3. Richards KL, Cannon SR, Crawford MH, Sorensen SG. Non-invasive diagnosis of aortic and mitral valve disease with pulsed-Doppler spectral analysis. Am J Cardiol 1983;51:1122-1127. 4. Quinones MA, Young JB, Waggoner Ad, Ostojic MC, Ribeiro LGT, Miller RR. Assessment of pulsed Doppler echocardiography in detection and quantification of aortic and mitral regurgitation. Br Heart J 1980;44:612-620, 5. Abbasi AS, Allen MW, DeCristofaro D, Ungar I. Detection and estimation of the degree of mitral regurgitation by range gated pulsed Doppler echocardiography. Circulation 1980;61:143-147. 6. Blanchard D, Diebold B, Peronneau P, Fault JM, Nee M, Guermonprez J, Maurice P. Non-invasive diagnosis of mitral regurgitation by Doppler echocardiography. Br Heart J 1981;45:589-593. 7. Areias JC, Goldberg SJ, de Villeneuve VH. Use and limitations of time interval histogram output from echo Doppler to detect mitral regurgitation. Am Heart J 1981;101:805-809. 6. Knutsen KM, Bae EA, Sivertssen E, Hoiseth A. Detection of mitral regurgitation by Doppler ultrasound: a comparison with left ventricular angiography. Acta Med Stand 1981;210:349-351. 9. Mivatake K. Nimura Y. Sakakibara H. Kinoshita N. Okamota M. Naeata S. Kawa”zoe K, Fujita T. Loc&ation and direction of mitral regurgitant 80, ii mitral orifice studied with combined use of ultrasonic pulsed Doppler technique and 2-dimensional echocardiography. Br Heart J 1982;48:449-458. 10. Kalmanson D, Veyrat C, Bouchareine F, Degroote A. Non-invasive recording of mitral valve flow velocity pattern using pulsed Doppler echocardiography. Br Heart J 1977;39:517-528. 11. Kalmanson D, Veyrat C, Abitbol G, Farjon M. Doppler echocardiography and valvular reguraitation with soecial emphasis on mitral insufficiencv. Advantages of 2-&iensionaI echocardiograpiy with real-time spediral analysis. In: Rijsterborghy H, ed. Echocardiography. The Hague: Martinus Nijhoff, 1981:279-29012. Miyatake K, Izumi S, Okamoto M, Kinoshita N, Asonuma H, Nakagawa H, Yamamoto K, Takamiya M, Sakakibara H, Nimura Y. Semiquantitative grading of severity of mitral regurgitation by real-time 2.dimensional Doppler flow imaging technique. JACC 1986;7:82-88. 13. Kostucki W. Vandenbossche JL, Friart A, Englert M. Pulsed Doppler regurgitant flow patterns of normal valves. Am J Cardiol 1986;58:309-313. 14. Meltzer RS, Abovich L, Finkelstein M. Regurgitation of all four cardiac valves detected by Doppler echocardiography. Am J CardioI1986;58:169-171. 15. Morganroth J, Mardelli TJ, Naito M, Chen CC. Apical cross sectional echocardiography. Standard for the diagnosis of idiopathic mitral valve prolapse syndrome. Chest 1981;79:23-28. 16. Markiwicz W, Stoner J, London E, Hunt SA, Popp RL. Mitral valve prolapse in one hundred presumably healthy young females. Circulation 1976;53:464-473, 17. Griffith JM, Henry WL. An ultrasound system for cardiac imaging and Doppler blood flow measurement in man. Circulation 1978;57:925-929. 18. Ellestad M. Stress Testing. Principles and Practice. 2nd ed. Philadelphia: F.A. Davis, 1980:304-306. 19. Hatle L, Annelsen B. Doppler Ultrasound in Cardiology. Physical Principles and CIinicd Applications. 2nd Ed. Philadelphia: Lea 6 Febiger, 1986:153; 176-188. 20. Feigenbaum H. Echocardiography. 4th Ed. Philadelphia: Lea B Febiger, 19863262-266. 21. Takao S, Miyatake K, Izumi S, Kinoshita N, Sakakibara H, Nimura Y. Physiological pulmonary regurgitation detected by the Doppler technique and its differential diagnosis (abstr). JACC 1985;5:499. 22. Baron M. Angiographic evaluation of valvular insufficiency. Circulation 1971;43:599-605. 23. Croft CH, Lipscomb K, Mathis K, Firth BG, Nicod P, Tilton G, Hillis LD. Limitations of quantitative angiographic grading in aortic or mitral regurgitation. Am J Cardiol 1984;53:1593-1598.