European Heart Journal (1989) 10, 334-340
V. DALL’AGLIO, G. D’ANGELO*, E. MORO, G. L. NICOLOSI, C. BURELLI, F. ZARDO, E. CERVESATO
AND D. ZANUTTINI
Servizio di Emodinamica and Divisione di Cardiologia, Ospedale di Pordenone and
*Divisione di Cardiologia, Ospedale di Eboli (SA), Italy
Colour Doppler flow mapping (CD) has proved to be ’a very sensitive and specific means of diagnosing valvular regurgitation and obtaining a rapid semiquantitative estimation of the severity of regurgitation itself. We tried to compare a semiquantitative evaluation of aortic and mitral regurgitation, without time-consuming calculations of regurgitant jet areas, with the conventional visual semiquantitative angiographic estimation. We have also evaluated in detail the interobserver variability of this type of semiquantitation.
Two independent observers (OB) have reviewed CD studies of a selected group of 47 consecutive patients who underwent both cineventriculography and aortography for aortic regurgitation (AR) and/or mitral regurgitation (MR), then graded as mild, moderate or severe. At CD, AR and MR were classified as present or absent and graded as mild, moderate or severe. The following interobserver percentage agreements were noted for AR presence or absence, AR grade, MR presence or absence, MR grade, respectively: 96%, 83%, 96%, 83%. Likewise, the respective echo-angio agreements were 90%, 58%, 80%, 70%. Agreement was significant (P< 0-001) in all cases.
Thus, good interobserver and echo-angio agreement was found in the CD assessment of AR and MR. However, under- or overestimation of CD vs. angio was noted in several cases (mostly by one grade). Underestimation of CD vs. angio was.27% for AR and. 18-5% for MR; overestimation of CD vs. angio was 15% for AR and 11-5% for MR.
CD has proved to be a useful technique not only for the qualitative but also for the semiquantitative evaluation of aortic and mitral regurgitation, as assessed in the same subjects, with good interobserver agreement. The disagreement between CD and angio has been generally by one grade of severity and is probably due mainly to the different type of information given by the two techniques.
Submitted for publication en 1.6 May 1988 and in revised form 7 October 1988.
Address for correspondence: Vittorio Dall’Aglio
Colour Doppler Flow mapping has proved to be a very sensitive and specific means of diagnosing valvular regurgitation[1-7]. Furthermore, colour Doppler allows immediate visualization of flow disturbances into the regurgitant chamber and rapid semiquantitative estimation of the severity of the regurgitation itself.
Several studies[2,6,8-11] have also demonstrated that colour Doppler can provide accurate semiquantitative assessment of aortic and mitral regurgitation by measuring the regurgitant jet area as a percentage of the regurgitant chamber area. For aortic regurgitation, the width of aortic regurgitant jet can be measured as a percentage of the left ventricular outflow diameter. However, despite these encouraging results, there remain important difficulties that limit our ability to quantify valvular regurgitation by colour Doppler[7,12-15,17].
We tried to compare a semiquantitative visual evaluation of aortic and mitral regurgitation, without time-consuming calculations of regurgitant jets areas, with the conventional visual semiquantitative angiographic estimation. We have also evaluated in detail the interl-observer variability of this type of semiquantitation in a selected group of consecutive patients in whom aortography and cineventriculography were performed.
Patients and methods
We analysed 52 consecutive patients who underwent cardiac catheterization for aortic and/or mitral valvular disease, and in whom the presence of aortic and/or mitral regurgitation was angiographically documented (41 A.R., 33 M.R.). In fact, the aim of this study was not the evaluation of sensitivity and specificity of colour Doppler examination, but the analysis of interobserver and echoangio variability in the semiquantitative assessment of aortic and mitral regurgitation in a selected group of patients with positive angiography for aortic and/or for mitral regurgitation.
Colour Doppler studies were made by using a commercially available system (Aloka SSD 860) and a 2-5 and/or 3-5 MHz transducer according to the acoustic access of the patient. For optimizing image quality, colour Doppler gain was first turned down completely and then increased very gradually until the static background noise just appeared[1,7,11]. In all patients, for each standard view (parasternal, apical and subcostal long and short axis) the transducer was angled and moved with small increments to obtain the maximal regurgitant signal in any given plane[1,7,11].
Non-standardized off-axis intermediate views were also used as needed. Colour Doppler examination was technically adequate for detailed frame-by-frame analysis in 47 cases (90%) (28 men, 19 women). It was performed on the same day as angiography in 33 cases, within 24 h in 11 cases, within 4 days in 2 cases and within 7 days in 1 case.
Retrospective and blind evaluation of colour Doppler studies were made by two observers. All the echocardiographic sections were reviewed. Colour Doppler evaluation of aortic and mitral regurgitation was qualitative for the presence or absence of regurgitation. For the semiquantitative evaluation, three grades of severity were considered—mild, moderate and severe—according to the width and depth of regurgitant jets. Aortic or mitral regurgitations were evaluated as being mild if jet width was visually judged less than or equal to one-third of the regurgitant chamber width, moderate if jet width was visually judged more than one-third and less than half of the regurgitant chamber width, and severe if jet width was more than half of the regurgitant chamber width.
For this semiquantitative analysis, the maximal width of the jet was considered. The level at which this width was maximal could be variable, in individual cases, i.e. at the neck, the body or the tail of the jet. The maximal width of the jet was compared with the maximal width of the left ventricular out-flow tract for the aortic regurgitation, and with the maximal width of the left atrium for mitral regurgitation, even though this could occur in different views.
The colour Doppler evaluation of regurgitant jets was made not only in the large angle views, but also (and mainly) in the 28° angle format, in order to increase the frame rate. The relationship between the size of the jet and that of the receiving chamber was made by visual integration of all information deriving from the complete examination through different transducer positions, in order to ensure exploration of the entire regurgitant chamber.
The depth of the jet was only considered as a secondary criterion, mainly when the jet was received into a large chamber. In fact, severe regurgitation in this situation not only should have a large jet but also should extend deeper into the receiving chamber.
Blind evaluation of cineangiography was made by two observers; consensus had to be reached by the two angiographers in case of disagreement. By angiography, the presence or absence of aortic and mitral regurgitation was evaluated, and a semi-quantitative evaluation was made according to the modified Sellers criteria: grade 1 was considered mild, grade 2 moderate, grades 3 and 4 severe.
For statistical analysis. Spearman’s rank correlation method was used (Table 1).
For the evaluation of the presence or absence of aortic and mitral regurgitation, interobserver agreement was 96% for both aortic and mitral regurgitation, with disagreement only for mild regurgitation (Table 1). Agreement between colour Doppler and angiography was 90% for aortic and 80% for mitral regurgitation (Table 1). Disagreement was mostly by one grade; in a very small percentage of cases colour Doppler underestimated regurgitation by two grades (Table 2).
For the semiquantitative evaluation, interobserver agreement was 83% far both aortic and mitral regurgitaton (Table 1), with disagreement always by one grade for aortic regurgitation, and mostly by one grade for mitral regurgitation. Again, for the semiquantitative evaluation, agreement between colour Doppler and angiography was 58% for aortic and 70% for mitral regurgitation, with disagreement mostly by one grade (Tables 1 and 2). All instances of disagreement by two grades were due to underestimation of colour Doppler compared with angiography.
The percentage of under and over-estimation of colour Doppler vs. angiography are reported in Table 2. The underestimation was 27% for aortic and 18-5% for mitral regurgitation (mostly by one grade); the overestimation was 15% for aortic and 11-5% for mitral regurgitation (always by one grade).
Statistical analysis (Tables 1 and 2) always showed highly significant agreement in both interobserver and colour Doppler-angiographic comparisons. Fig 1-6 show some examples of our cases.
90% of mitral regurgitations that were evaluated as severe by angiography were also evaluated as severe by both colour Doppler observers. Only in one case (Figs 4 and 5) was there two-grade underestimation by colour Doppler (mild) vs. angiography (severe). In this case, at angiography, the contrast medium filled the left atrium slowly and unevenly (Fig.5).
65% of aortic regurgitations that were evaluated as severe by angiography were evaluated as severe by at least one of the colour Doppler observers (36% by both, 29% by one) with one grade of interobserver disagreement; 29% of all the angiographically severe aortic regurgitations were under-estimated by one grade by both colour Doppler observers. Only in one case of severe aortic regurgitation (6%) did one colour Doppler observer underestimate severity by two grades (Fig. 6).
Colour Doppler has recently been shown to be useful in the evaluation of intracardiac flows. Colour Doppler evaluation of regurgitant flows has demonstrated that regurgitant jets are very complex three-dimensional entities, which can show variable origin, size, outline, shape, direction and depth in different cross-sectional views, different cardiac cycles and also in different frames during the same cycle. For these reasons all standard and intermediate cross-sectional views must be used in order to obtain the maximum information about the jet. Several studies[2,6,8-11,17] have demonstrated that colour Doppler can provide accurate semi-quantitative estimation of the severity of aortic and mitral regurgitation by means of area and linear dimensions of jets. In our study, the depth of the jet was considered only as a secondary criterion, in accordance with previous reports[3,6,12] which have demonstrated that the depth of the jet is proportional not only to the area of the valvular incompetence but also to the magnitude of the imposed pressure gradient across the valve. In aortic regurgitation, for example, the depth of the regurgitant jet is largely dependent on the imposed systemic afterload; the width of the regurgitant jet, on the contrary, appears to be much more closely related to the severity of aortic regurgitation[3,6,12]. Grading of mitral regurgitation by measuring the maximum depth of the regurgitant jet into the left atrium showed no good correlation with angiographic grading. Nevertheless, in our experience, in cases of severe aortic and mitral regurgitations the depth of the jet also might be important, particularly when the receiving chamber is large. In fact,in this situation the jet should not only be large, but also extend deeper into the chamber.
Figure 1 Apical modified four-chamber view of a ease with severe aortic regurgitation. Diastolic stop frame. The regurgitant jet (arrow), with a typical ’mosaic pattern’, is very wide and deep. In this case there was complete echo-angio and interobserver agreement in the evaluation of the regurgitation as severe. LV = left ventricle; LA = left atrium.
Figure 2 Systolic stop frame, in a low parasternal long axis view, of a case of severe mitral regurgitation. The regurgitant jet (arrow), with a typical ’mosaic pattern’, is large and deep into the left atrium. In this case there was complete echo-angio and interobserver agreement in the evaluation of the regurgitation as severe. lv=left ventricle; la = left atrium.
Figure 3 Systolic stop frame of a subcostal four-chamber view is represented. A huge left atrium is evident. The regurgitant jet (arrow), with the typical ’mosaic pattern’, appears narrow compared with the size of the left atrium, with scattering of colour pixels deep into the atrium itself. The low velocity of the peripheral part of the jet into the huge left atrium and the wide discrepancy between jet width and dimensions of the extremely dilated left atrium can increase the probability of colour Doppler underestimation of the severity of regurgitation compared with angiography. In such a case the ratio between the regurgitantjet area and the left atrium could also be misleading. This mitral regurgitation was however evaluated as severe by both colour Doppler observers. This was in agreement with the angiographic evaluation, la=left atrium.
Figure 4 Systolic stop frame, in an apical four-chamber view, of a case of mitral regurgitation with two grades of disagreement between colour Doppler and angiographic evaluation. On colour Doppler, the regurgitantjet showed definable scattering of colour pixels into the left atrium, making it difficult to define the outline of regurgitantjet area. This mitral regurgitation was evaluated as mild by both colour Doppler observers and severe by angiographers. On angiography, the left atrium was slowly and unevenly filled by contrast medium (Fig. 5). This particular pattern of regurgitation could explain the two-grade underestimation by colour Doppler vs. angiography. lv = left ventricle; la = left atrium.
Figure 5 Four subsequent systolic frames, in a left anterior oblique projection, of the left cineventriculography in the same case as Fig. 4 (mitral regurgitation). On real time analysis, the left atrium (arrows) was slowly and unevenly filled by contrast medium. This mitral regurgitation was evaluated as severe by the angiographers.
Figure 6 Apical five-chamber view of a case with aortic regurgitation. Diastolic stop frame. The aortic regurgitantjet (arrow) appears very posterior and medial, running along the anterior mitral leaflet and merging early with the normal transmitral flow. In this case there was a two-grade disagreement between both colour Doppler observers (mild) and angiographers (severe). lv= left ventricle; la = left atrium.
However, there remain important difficulties that limit our ability to quantify valvular regurgitation by colour Doppler[7,12-15,17]. We tried to test the value of semiquantitative evaluation of aortic and mitral regurgitation in the same subject, compared with conventional angiographic evaluation. We have also tested the interobserver variability of colour Doppler interpretation of the severity of regurgitation. It was not the purpose of this study to compare separate colour flow studies made by different observers in the same patient; we do not know if this would have greatly increased variability. However we tried to obtain, in each patient, a very extensive examination from different approaches. This is a well-established procedure at our institution, where off-axis non-standard positions are also routinely searched in these situations.
Our results are consistent with good agreement, even if less than optimal, between colour Doppler and angiography for the semiquantitative evaluation of aortic and mitral regurgitation. Interobserver agreement and agreement between colour Doppler and angiography was very good in the qualitative detection of regurgitation. Interobserver agreement was also good in the, semiquantitative evaluation of regurgitation.
There are several possible causes of disagreement between colour Doppler and angiography in the evaluation of aortic and mitral regurgitation:
(1) A different type of information is given by the two methods. Colour Doppler is a tomographic colour imaging of flow velocities, while angiography gives the densitometric projection, on a single plane, of three-dimensional information coming from a dilution of the contrast medium into a cardiac chamber. In fact, colour Doppler displays flow velocity profiles, while angiography showsflow mass shifts. Nanna et al. have demonstrated that identical volumes of injection, through orifices of different areas, may generate different colour flow areas if driven by different pressure gradients (i.e. by different velocities).
(2) A huge left atrium and/or valvular-annular calcification are important factors in the assessment of mitral regurgitation. The evaluation of regurgitant jet area can be altered by low velocity of the peripheral part of the regurgitant jets with dispersion of colour pixels at the borders of the jets, deep inside the large atrial cavity. On the other hand, ultrasound energy can be absorbed by calcified structures.
(3) Large left atrial and left ventricular cavities decrease the possibility of adequate sampling of flows far from the transducer.
(4) Mitral prostheses can produce echo absorption, reverberations or phantom echoes. In these cases we have obtained useful information on the left atrium from the high parasternal long- and short-axis views, mostly in deep expiration.
(5) Colour Doppler may have the potential to identify even milder mitral regurgitant jets, compared with angiography, because the mitral valvular plane and the left atrium can be studied in more detail from multiple views using the former technique. In two of our cases, both colour Doppler observers identified the presence of a mild mitral regurgitation, while angiographers did not
(6) The coexistence of aortic regurgitation and mitral stenosis can cause disagreement between colour Doppler and angiography. We found 10 cases of coexistence of aortic regurgitation and mitral stenosis; in 40% of these there was agreement between colour Doppler and angiography; in 60% however, under- or overestimation by one or both colour Doppler observers was noted.
(7) The coexistence of aortic regurgitation and a mitral prosthesis can also cause disagreement. This can be due to the ’invasion’ of the left ventricular outflow tract by the prosthesis and its flow, when it is oriented towards the left ventricular outflow and the septum.
(8) The presence of aortic regurgitation in a narrow left ventricular outflow tract and/or coexistence of redundant anterior mitral leaflet can also cause disagreement. This can be due to the narrow left ventricular outflow tract which is used as a reference for judging the relative width of the jet.
(9) If the aortic regurgitant jet is posterior and medial, running along the anterior mitral leaflet, merging early with the normal transmitral diastolic flow, there is also a possibility of disagreement.
(10) Our choice of a simplified, semiquantitative evaluation of regurgitation in three grades of severity could also have increased the possibility of disagreement. On the other hand, a colour Doppler classification using four classes of severity can bedifficult to standardize and does not allow easy comparison with the four angiographic classes.
(11) A final possibility is that we did not measure the ratio between regurgitant jet area and regurgitant chamber area. This ratio evaluation, however, was not the aim of our study; we tried to find a simple, reproducible and not time-consuming semiquantitative evaluation of aortic and mitral regurgitation. Furthermore, the ratio itself could have been misleading in some of our cases, as shown in Fig. 3.
Despite these limitations, colour Doppler has proven to be a useful technique for the qualitative and semiquantitative evaluation of aortic and mitral regurgitation in the same subjects, with good interobserver agreement. Disagreement between colour Doppler and angiography has been mostly by one grade of severity and it is probably due mainly to the different type of information given by the two techniques.
Several problems still remain to be defined, for example, interinstrument, interpatient and interexamination variability. We feel that since regurgitant jets are very complex three-dimensional entities very extensive examination, including all standard and intermediate cross-sectional views, should be routinely performed.
Further carefully designed studies are certainly needed to define better the variability of methods for semiquantitative evaluation of regurgitant jets.
 S-tewart WJ. Levine RA, Main J, King M. Initial experience with color-coded Doppler flow mapping. Echocardiography 1985;2: 511.
 Omoto R, Yokote Y, Takomoto S et al. The development of real-time two-dimensional Doppler echocardiography and its clinical significance in acquired valvular diseases. Jpn Heart J 1984; 325: 25-3.
 Switzer DF, Nanda NC. Color Doppler evaluation of valvular regurgitation. Echocardiography 1985; 2: 553.
 Miyatake K, Okamoto M, Kinoshita N et al. Clinical application of a new type of real-time two-dimensional Doppler flow imaging system. Am J Cardiol 1984; 54: 857.
 Perry GJ, Helmcke F, Nanda NC et al. Evaluation of aortic insufficiency by Doppler color flow mapping. J Am Coll Cardiol 1987; 9: 952-9.
 Helmcke F, Nanda NC, Hsiung MC et al. Color Doppler assessment of mitral regurgitation with orthogonal planes. Circulation 1987; 75: 175.
 Perry GJ, Nanda NC. Recent advances in color Doppler evaluation of valvular regurgitation. Echocardiography 1987;4:503.
 Byard CE, Perry GJ, Roitman DI, Nanda NC. Quantitative assessment of aortic regurgitation by color Doppler. Circulation 1985; 72: 111^6 (Abstr).
 Saenz CB, Deumite J, Roitman DI, Moos S, Nanda NC, Soto B. Limitations of color Doppler in quantitative assessment of mitral regurgitation. Circulation, 1985; 72: III-99.
 Perry GJ, Helmcke F, Nanda NC. Color Doppler assessment of aortic insufficiency in two orthogonal planes. J Am Coll Cardiol 1986; 7: 101A (Abstr).
 Miyatake K, Izumi S, Okamoto M et al. Semiquantitative grading of severity of mitral regurgitation by realtime two-dimensional Doppler flow imaging technique. J Am Coll Cardiol 1986; 7: 82.
 Switzer DF, Yoganathan AP, Nanda NG. Woo YR, Ridgway AJ. Calibration of color Doppler flow mapping during extreme hemodynamic conditions in vitro: a foundation for a reliable quantitative grading system for aortic incompetence. Circulation 1987; 75: 837.
 Otsuji Y, Tei C, Kisanuki A, Natsugoe K, Kawazoe Y. Color Doppler echocardiographic assessment of the change in the mitral regurgitant volume. Am Heart J 1987; 114:349.
 Goldman ME. Real-time two-dimensional Doppler flow imaging: a word of caution (editorial comment). J Am Coll Cardiol 1986; 7: 89.
 Nanna M, Feng X, Holen J, Gramiak R, Meltzer RS. Analysis of color Doppler: an in vitro study of the effects of jet velocity on Doppler color flow mapping. In: Proceedings of International Workshop on Patient management decisions in valvular heart disease: the role of echo-Doppler versus cardiac catheterization. Bari, Italy, May 18/20, 1987, 112. Rizzon P., Roelandt J., Iliceto S., Chiddo A. Ed. ’La Nuova Tecnografica’, Bari, Italy, 1987, May 18/20, p. 112
 Sellers RD, Levy MJ, Amplatz K, Lillehei CW. Left retrograde cardioangiography in acquired cardiac disease. Am J Cardiol 1964; 14: 437.
 Pearlman AS, Otto MC. Quantification of valvular regurgitation. Echocardiography 1987; Vol 4, n. 4: 271.
KEY WORDS: Colour Doppler, aortic and mitral regurgitation.