Author + information
- Received May 2, 2007
- Revision received November 20, 2007
- Accepted November 26, 2007
- Published online April 29, 2008.
- Julien Magne, MSc,
- Mario Sénéchal, MD, FRCPC,
- Patrick Mathieu, MD, FRCPC,
- Jean G. Dumesnil, MD, FRCPC, FACC,
- François Dagenais, MD, FRCPC and
- Philippe Pibarot, DVM, PhD, FACC, FAHA⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Philippe Pibarot, Québec Heart Institute, Department of Cardiology, 2725 Chemin Sainte-Foy, Québec City, Québec, Canada G1V 4G5.
Objectives The purpose of this study was to evaluate mitral valve hemodynamic performance and functional capacity in patients with ischemic mitral regurgitation (MR) who underwent restrictive mitral valve annuloplasty (MVA).
Background Restrictive MVA combined with coronary artery bypass graft is the conventional approach for the surgical management of patients with ischemic MR. We hypothesized that the restriction of the mitral annulus could cause an obstruction to antegrade mitral flow that may affect the patient's functional capacity.
Methods A dobutamine stress echocardiography (DSE) and a 6-min walk test (6MWT) were performed in 24 patients with ischemic MR 13 ± 3 months after restrictive MVA and coronary artery bypass graft and in 20 control patients with coronary artery disease matched for age, gender, and left ventricular ejection fraction.
Results None of the 24 MVA patients had significant MR after operation. Compared with control patients, MVA patients had significantly (p < 0.001) higher resting and stress peak gradients (rest: 13 ± 4 mm Hg vs. 4 ± 1 mm Hg; DSE: 19 ± 6 mm Hg vs. 6 ± 3 mm Hg) and systolic pulmonary arterial pressures (PAP) (rest: 42 ± 13 mm Hg vs. 27 ± 8 mm Hg; DSE: 58 ± 12 mm Hg vs. 38 ± 11 mm Hg) and lower (p = 0.01) 6MWT distance (358 ± 95 m vs. 433 ± 61 m). The resting peak mitral gradient correlated with systolic PAP (r = −0.67; p = 0.001) and 6MWT distance (r = −0.78; p < 0.0001) in the MVA group.
Conclusions The results suggest that performing a restrictive MVA in patients with ischemic MR may create a functional mitral stenosis. This hemodynamic sequel is associated with higher PAP and a worse functional capacity.
Ischemic mitral regurgitation (MR) is frequently observed in patients with ischemic left ventricular (LV) dysfunction and is associated with poor outcomes (1,2). Restrictive mitral valve annuloplasty (MVA) combined with coronary artery bypass graft (CABG) is the conventional approach for the surgical management of patients with ischemic MR (3). The size of the ring used for restrictive annuloplasty is generally selected by downsizing by 2 sizes the measured intertrigonal length. Despite this important restriction of the anteroposterior mitral annulus distance, MR usually persists or recurs in more than 20% of patients after MVA (4), and the presence of residual MR adversely affects the patient's post-operative outcome (5).
The restriction of the anteroposterior annulus distance by implanting a rigid or semiflexible ring may reduce the mitral valve effective orifice area (EOA) and the left atrial compliance, which may cause high transmitral pressure gradients and pulmonary arterial pressure (PAP). We therefore hypothesized that restrictive MVA in patients with ischemic MR could cause an obstruction to antegrade mitral flow that could, in turn, affect the patient's functional capacity. The objective of the present study was to evaluate the resting and exercise mitral valve hemodynamic performance and functional capacity in patients who underwent restrictive MVA and had no evidence of persistent or recurrent MR after operation.
From January 2004 to October 2006, 55 consecutive patients underwent restrictive MVA and CABG for ischemic MR in our institution (Fig. 1). The surgical protocol is described in our previous publication (5). From this cohort, we excluded patients with concomitant organic mitral valve lesions, >mild aortic regurgitation or stenosis, concomitant valve procedure, and >mild persistent or recurrent MR after MVA defined as a width of MR jet vena contracta >3 mm. The final study group was composed of 24 patients. These patients underwent a dobutamine stress echocardiography (DSE) followed by a 6-min walk test (6MWT) 13 ± 3 months after MVA. Nine of them also underwent an exercise stress echocardiography (ESE) that was performed within 30 days of the DSE. The beta-blocker medication was withdrawn for at least 24 h before DSE, 6MWT, and ESE. The study protocol was approved by the Institutional Review Board of Laval Hospital, and informed consent was obtained from all of the patients.
We also recruited a control group of 20 patients with coronary artery disease who were matched with the MVA patients for age, gender, and LV ejection fraction. Patients with more than mild MR, mild aortic regurgitation, or aortic stenosis were excluded from the control group. These control patients underwent DSE and 6MWT using the same protocol as the MVA patients.
Doppler-echocardiography at rest
Two-dimensional and Doppler transthoracic echocardiography examination was first performed at rest as previously described (6).
LV Geometry and Function
The Doppler-echocardiographic measurements included the LV and left artial (LA) end-diastolic and -systolic diameters, the LV ejection fraction determined by the modified biplane Simpson method, and the LV stroke volume measured in the LV outflow tract by pulsed-wave Doppler.
Mitral Valve Hemodynamics
The mean transvalvular flow rate was calculated by dividing the left ventricular outflow tract (LVOT) stroke volume by the diastolic filling duration measured on the pulsed-wave Doppler signal of the mitral flow velocity. The peak and mean transmitral pressure gradients were calculated using the modified Bernoulli equation. Mitral valve EOA was determined by the continuity equation using the stroke volume measured in the LVOT divided by the integral of the mitral transvalvular velocity during diastole. The net atrioventricular compliance (Cn) was calculated from the valve EOA and mitral flow E-wave downslope as previously described (6,7).
Systolic PAP was calculated by adding the systolic right ventricular pressure derived from the tricuspid regurgitation to the estimated right atrial pressure (8).
The dobutamine infusion protocol was designed to obtain incremental increases in flow and a steady state at each level. It consisted of 5-min increments of 5 or 10 μg/kg/min up to a maximum dosage of 40 μg/kg/min. The predetermined end points for terminating DSE were as follows: 1) maximal heart rate reached (defined as the formula: 220 − age); 2) systolic blood pressure <80 mm Hg or >220 mm Hg or diastolic blood pressure >110 mm Hg; 3) ischemia detected by electrocardiogram (ECG) (>1 mm horizontal or downsloping ST depression 0.08 s after the J-point compared with the resting ECG); 4) ventricular arrhythmias (>6 premature ventricular contractions per minute or complex arrhythmia); 5) rapid new atrial arrhythmias; and 6) symptoms.
Left ventricular geometry and function, peak and mean transmitral pressure gradients, valve EOA, Cn, and systolic PAP were measured at rest and at each stage of the DSE protocol.
Nine of the MVA patients were submitted to a maximum ramp semisupine bicycle exercise with the ergometer table tilted to 20° (patient in a mild left lateral position). The initial workload was set at 0 W, followed by continuous increments of 10 to 25 W/min depending on the patient's physical condition. A 12-lead ECG was continuously recorded, and blood pressure was measured every 2 min. The patients were encouraged to exercise until exhaustion or the appearance of symptoms. The end points for terminating ESE were the same as those used for DSE. The Doppler-echocardiographic measurements were performed at rest in the semisupine position and at peak of exercise.
To evaluate the patient's functional capacity, a 6MWT was performed on the same day as the DSE protocol (9). The test was performed 2 times with a resting period of 15 minutes between the 2 tests, and the average of the distances achieved at these 2 consecutive tests was recorded.
Results are expressed as mean ± SD or percentages unless otherwise specified. Before analysis, normality distribution was tested using Kolmogorov-Smirnov test. In the MVA group, differences between pre- and post-operative data and between rest and ESE data were analyzed using paired t test or signed rank test when normality test failed. The resting and DSE data were analyzed using 2-way analysis of variance for repeated measures and a Tukey test to assess the effect of pharmacologic stress (DSE vs. rest) and group (MVA vs. control). When the normality test failed, the data were analyzed with the use of a Mann-Whitney rank sum test. Linear regression analyses were used to evaluate the relationship between echocardiographic parameters and 6MWT distance. Forward and backward multiple stepwise regression analyses were performed to identify independent predictors of shorter 6MWT distance. Age, gender, and variables with a p value of <0.05 on univariate analysis were entered in the multivariate model. Sensitivity, specificity, positive predictive value, and negative predictive value for the prediction of reduced 6MWT distance defined as a distance ≤ median value in the MVA group were determined for various cut-off values of the echocardiographic parameters using receiver-operating characteristic (ROC) curves.
Pre-operative and operative clinical data
Patients of the MVA group were operated for moderate to severe ischemic MR (vena contracta width 5.2 ± 0.8 mm, range 4.0 to 7.0 mm), and restrictive MVA was performed with the use of a Carpentier-Edwards Physio ring 24 mm in 8 patients (33%), 26 mm in 12 patients (50%), 28 mm in 3 patients (13%), and 30 mm in 1 patient (4%). All of these patients underwent CABG, with an average of 3 ± 1.5 vessels grafted.
As expected, there was no significant difference between the MVA and control groups regarding gender distribution (16 [66%] male vs. 12 [60%] male; p = NS), mean age (65 ± 8 [range 50 to 79] years vs. 66 ± 11 [range 50 to 88] years; p = NS), body surface area (1.8 ± 0.2 m2 vs. 1.7 ± 0.2 m2; p = NS), and LV ejection fraction (43 ± 11% [range 20% to 64%] vs. 45 ± 10% [range 25% to 64%]; p = NS). The prevalence of comorbidities was also similar in both groups.
Comparison of post-operative versus pre-operative data in the MVA group
Mitral regurgitation severity and LV geometry and function improved significantly after MVA and CABG (Table 1). On the other hand, resting mitral peak and mean transmitral gradients and systolic PAP increased significantly after surgery. There was also no significant improvement in the distribution of New York Heart Association functional class.
Resting and stress echocardiographic data
The maximum DSE dose (40 μg/kg/min) was reached in 4 (17%) of the 24 MVA patients, whereas DSE infusion was stopped at a lower dosage owing to chest discomfort in 4 (17%), new rapid atrial arrhythmias in 3 (13%), and ventricular arrhythmias in 13 (54%). In the control group, the maximum dose was reached in 12 (60%) of the 20 patients, and it was stopped at a lower dosage owing to the occurrence of symptoms or arrhythmias in 8 patients.
LV Geometry and Function
Compared with the patients of the control group, those of the MVA group had significantly larger LV end-diastolic (p < 0.001) and end-systolic (p < 0.001) diameters at rest and during DSE (Table 2). However, these differences were no longer significant when the diameters were indexed for body surface area. The differences in LV geometry between the 2 groups (Table 2) are likely related to the fact that, in contrast to control patients, MVA patients had a history of moderate/severe ischemic MR and therefore of marked LV dilation that was only partially reversed by surgery (Table 1).
There was a significant decrease in LV end-diastolic diameter and a marked increase in heart rate, LV ejection fraction, LVOT stroke volume, cardiac output, and mean mitral flow rate during DSE in both control and MVA patients (Table 2). Patients of the MVA group had similar LV ejection fraction but higher stroke volume at peak DSE compared with the patients of the control group. However, they had lower peak heart rate and thus longer diastolic filling time on DSE. Consequently, the maximum cardiac output and mean mitral flow rate reached during DSE were similar in both groups.
In the subset of 9 patients of the MVA group who also underwent ESE, there was no significant change in LV dimensions, whereas heart rate, LV ejection fraction, stroke volume, cardiac output, and mean mitral flow rate increased markedly (Table 3).
Mitral Valve Hemodynamics
Compared with the patients of the control group, those of the MVA group had significantly lower mitral valve EOA and indexed EOA and higher mitral gradients at rest and during DSE (Table 2). Among the 24 MVA patients, 13 (54%) had a resting valve EOA ≤1.5 cm2, 15 (62.5%) had an indexed EOA ≤0.9 cm2/m2, 10 (42%) had peak gradient ≥15 mm Hg, and 13 (54%) had a mean gradient ≥5 mm Hg at rest. Therefore, a large proportion of the MVA patients had at least moderate mitral stenosis.
The vast majority of patients of the MVA group had a marked increase in mitral mean and peak gradients during DSE and ESE (Fig. 2). This higher increase in mitral gradients observed during ESE is likely related to the higher increase in heart rate and mean mitral flow rate achieved during ESE compared to DSE.
Systolic PAP at rest was 42 ± 13 mm Hg and increased up to 58 ± 12 mm Hg during DSE (p < 0.001) (Table 2) and up to 69 ± 14 mm Hg during ESE (p = 0.008) (Table 3) in the MVA group. Moreover, the resting and DSE systolic PAPs of the MVA patients were significantly higher than those of the control patients (Table 2).
In the MVA group, the prevalence of pulmonary arterial hypertension defined as a resting systolic PAP ≥40 mm Hg was 22% before MVA and 45% at the time of this post-operative study. In contrast, only 15% and 10% of the control patients had pulmonary hypertension at rest and during DSE, respectively.
When pooling rest, DSE, and ESE data, the systolic PAP of the MVA patients correlated well with mitral peak gradient (r = 0.70; p < 0.001) (Fig. 3) and mean gradient (r = 0.62; p < 0.001). Systolic PAP also correlated with Cn (r = 0.44; p = 0.0003). In contrast, there was no significant correlation between systolic PAP and these hemodynamic parameters in the control group.
The 6MWT was completed by 22 patients of the MVA group and 17 patients of the control group. The average 6MWT distance was 358 ± 95 m in the MVA group, which was markedly lower (Fig. 4) than the age and gender-predicted distance calculated using the equation of Gibbons et al. (10) (492 ± 59 m; p < 0.001). Moreover, MVA patients had a significantly lower average 6MWT distance compared with that of the control patients (433 ± 61 m; p = 0.01) despite similar age and gender-predicted distance in both groups (Fig. 4). Accordingly, the percentage of predicted 6MWT distance was also significantly lower in the MVA group than in the control group (74 ± 21% vs. 87 ± 13%; p = 0.036).
In the MVA group, the 6MWT distance correlated with LV end-systolic diameter (r = −0.51; p = 0.015), Cn (r = 0.54, p = 0.01), mitral peak gradient (rest: r = −0.78; p < 0.0001; DSE: r = −0.58; p = 0.004) (Fig. 5A), mean gradient (rest: r = −0.65; p = 0.001; DSE: r = −0.26; p = NS), and systolic PAP (rest: r = −0.68; p = 0.001; DSE: r = −0.60; p = 0.006). Similar results were obtained with the percentage of age- and gender-predicted 6MWT distance. In the control group, there was a trend for correlation between 6MWT distance and resting LV ejection fraction (r = 0.49; p = 0.09) and LA diameter (r = −0.54; p = 0.06), whereas the percentage of age and gender-predicted 6MWT distance correlated with LV end-systolic diameter (rest: r = −0.73; p = 0.005).
In forward and backward multiple stepwise regression analyses, the resting mitral peak gradient (Δr2 = 0.59; p = 0.0003) and systolic PAP (Δr2 = 0.16; p = 0.004) were independently associated with reduced 6MWT distance (model r2 = 0.75) in the MVA group.
In ROC curve analysis (Table 4), the best predictors of reduced 6MWT distance (median value ≤350 m) in the MVA patients were mitral peak and mean gradient (area under the ROC curve = 0.87 for both parameters). A mitral peak gradient ≥13 mm Hg had the best performance for the prediction of reduced 6MWT distance (sensitivity 92%, specificity 82%, positive predictive value 83%, negative predictive value 90%).
The most important finding of this study is that a large proportion (>50%) of the patients who underwent successful restrictive MVA in terms of correction of MR nonetheless had a significant obstruction to antegrade mitral flow. As a matter of fact, MVA patients had markedly higher transmitral gradients and PAP at rest and during DSE compared with control patients matched for age, gender, and LV ejection fraction. Moreover, the presence of a high post-operative transmitral pressure gradient was associated with reduced 6MWT distance in the MVA patients. These findings suggest that restrictive MVA may cause a functional mitral stenosis and that this hemodynamic abnormality may negatively impact the patient's functional capacity.
Hemodynamic outcome of restrictive MVA
Restrictive MVA aims at improving leaflet coaptation by correcting posterior annular dilatation. Bolling et al. (11) were the first to propose the use of undersized annuloplasty, which consists in implanting a prosthetic ring downsized by 1 or 2 sizes (i.e., 2 or 4 mm) relative to the intertrigonal distance. However, mild to moderate MR may persist or recur in more than 20% of patients after restrictive MVA (4). Furthermore, patients having a residual MR with a vena contracta width >3 mm have worse clinical outcomes (5).
In the present study, we focused on the subset of patients who had no persistence or recurrence of clinically significant MR 1 year after operation (Fig. 1). Moreover, none of these patients had a worsening of MR severity during DSE or ESE, thus excluding the presence of dynamic changes in MR with exercise in this series. Therefore, although all patients included in this study had a successful and persistent correction of MR, a large proportion nonetheless exhibited significant pulmonary hypertension and reduced functional capacity. Importantly, these hemodynamic and functional abnormalities can in large part be explained by the presence of increased post-operative transmitral pressure gradients.
Our results thus suggest that the reduction of the mitral annulus area induced by restrictive MVA may often cause a functional mitral stenosis. Several mechanisms may explain this adverse hemodynamic consequence. First, the narrowing of the mitral annulus area combined with the presence of redundant mitral valve tissue within this narrowed orifice could contribute to reduce the mitral valve EOA available for antegrade mitral blood flow. Second, this procedure may reduce the mobility of the posterior mitral valve leaflet. Indeed, previous studies have shown that restrictive MVA markedly reduces the mobility of the posterior leaflet and alters the valve closure in a single leaflet process. Third, the implantation of a rigid or semiflexible prosthetic ring may considerably limit the expansion of the mitral annulus during diastole. In this regard, previous studies have demonstrated that, as opposed to the aortic valve, the diastolic expansion of the mitral annulus is one of the key mechanisms involved in the opening of the mitral valve orifice during diastole (12).
Moreover, the constriction of the mitral annulus and the limitation of its expansion during diastole may also reduce the LA compliance. The relation between LA compliance and LA pressure is exponential and, in the presence of moderate-severe valvular obstruction, the loss of the buffering function caused by a reduction in LA compliance may result in an augmentation of the LA and PAPs. In the present study, atrioventricular compliance was lower in the MVA group than in the control group and it correlated with both PAP and 6MWT distance in the MVA group. Net atrioventricular compliance is, however, not only dependent on LA compliance but also on LV compliance and could therefore also be affected by a decrease in LV compliance. In contrast to what has been reported in previous series with mitral stenosis or mitral prosthesis-patient mismatch with normal LV function (6,7,13), impaired LV compliance may also have contributed to the increased PAP and reduced 6MWT distance in the present series. Nonetheless, the fact that the systolic PAP correlated much better with peak mitral gradient than with Cn suggests that the impact of the MVA-induced valve obstruction on the pulmonary hemodynamics was relatively more important than that of LV compliance.
Clinical impact of functional mitral stenosis caused by restrictive MVA
The mitral valve hemodynamic abnormalities observed in the present series of patients after restrictive MVA are similar to those observed in patients with mitral stenosis or mitral prosthesis-patient mismatch (Table 5) (6,7,13,14). Interestingly, a large proportion (>50%) of the patients included in the present series had stress mitral gradients and PAP similar to those observed in patients with moderate to severe mitral stenosis under DSE or ESE (Table 5) (15,16).
The stress hemodynamics of these patients also tended to be worse compared with those reported in patients with mitral prosthetic valves (17,18). The level of systolic PAP measured at DSE or ESE in the present series was also markedly higher than that observed in normal subjects under maximum exercise stress test (peak exercise systolic PAP <30 to 35 mm Hg) (19). Moreover, the stress hemodynamics of these patients also tended to be worse compared with those reported in patients with mitral prosthetic valves (17,18).
It remains to be determined to what extent these hemodynamic abnormalities have a negative impact on clinical outcomes. By inference, it should, however, be noted that the high post-operative transmitral gradients correlated with a reduced 6MWT distance and that several studies have shown that a reduced 6MWT distance is a powerful predictor of morbidity and mortality in patients with heart failure (9). Moreover, in a recent study of patients with mitral prosthesis-patient mismatch and similar hemodynamic abnormalities (Table 5), severe mismatch was associated with a 3.2-fold (95% confidence interval 1.5 to 6.8; p = 0.003) increase in the risk of mortality after mitral valve replacement after adjustment for age, gender, and other relevant risk factors (14).
The results of the present study therefore suggest that, when ischemic MR is treated by restrictive MVA, MR is often traded for mitral stenosis, which may obviate or outweigh the benefits of this procedure on the hemodynamic, functional, and clinical outcomes.
In light of the relatively high postoperative rates of persistent or recurrent MR recently reported in the literature (4,5), one could be tempted to use an even more restrictive annuloplasty with downsizing of more than 2 sizes. A recent study, however, suggested that a more restrictive annuloplasty would likely further worsen posterior leaflet angulation and therefore may not be able to restore adequate leaflet coaptation and MR correction (5). Moreover, the results of the present study suggest such a strategy would further increase the risk and severity of post-operative mitral stenosis. It remains to be determined if the newer prosthetic rings specifically designed to treat ischemic MR can improve the success rate for the correction of MR without inducing mitral stenosis. Moreover, there are at present many research efforts aimed at finding procedures that can improve LV geometry thus allowing the treatment of ischemic MR on the basis of its causal mechanism (20). Further studies are needed to determine if such approaches will succeed in both correcting MR and avoiding mitral stenosis. Meanwhile, a potential alternative, especially in the patients who are identified as being at high risk for persistent or recurrent MR if treated by restrictive MVA (5), would probably be mitral valve replacement with chordal sparing. If such a procedure is envisioned, it is, however, important to attempt to avoid prosthesis-patient mismatch (14).
The main limitation of this study is the relatively small number of patients. The apparent lack of significant association between PAP or 6MWT distance and some clinical or echocardiographic variables may be related to a type II error due to the small sample size. Nonetheless, this limitation does not affect the validity of the main results of this study, which is the demonstration that there is a high prevalence of mitral stenosis after restrictive MVA and that there is a strong association between the transmitral pressure gradient and 6MWT distance.
One could argue that the regression of pulmonary arterial hypertension and the improvement of symptomatic status were incomplete at 1 year after MVA and that a longer-term follow-up would have shown better hemodynamic and functional results. However, previous studies have demonstrated that the post-operative improvement in pulmonary arterial hemodynamics and functional capacity essentially occur during the first year after mitral valve surgery (21).
In our experience, DSE is superior to ESE with respect to the feasibility and reproducibility of the Doppler-echocardiographic measurements of mitral valve hemodynamics and PAP. On the other hand, ESE is a more physiological means to assess the behavior of LV function and mitral valve hemodynamics under stress conditions. In this regard, one could argue that although the patients selected in this study had no significant residual MR at rest, a dynamic appearance or worsening of MR during physical exercise could have contributed to the impairment of functional capacity. To address this issue, we randomly selected a subset of patients for ESE, and none of these patients had a dynamic occurrence or worsening of MR under exercise. In fact, the data obtained in this subgroup further confirmed those obtained by DSE.
The present results show that restrictive MVA in patients with ischemic MR may create some degree of functional mitral stenosis. This hemodynamic sequela is associated with higher PAP and worse functional capacity. Further studies are needed, however, to determine to what extent the mitral stenosis caused by restrictive MVA negatively impacts on other clinical outcomes. The present findings also provide an impetus toward the development of new surgical procedures that would allow for efficient and long-term correction of ischemic MR without inducing mitral stenosis.
The authors thank Jocelyn Beauchemin, Brigitte Dionne, Stéphanie Dionne, Michelle Dubois, Dominique Labrèche, and Isabelle Laforest for the data collection and their technical assistance.
Supported by a grant from the Canadian Institutes of Health Research (MOP 67123), Ottawa, Ontario, Canada. Dr. Pibarot holds the Canada Research Chair in Valvular Heart Diseases, Canadian Institutes of Health Research, Ottawa, Ontario, Canada.
- Abbreviations and Acronyms
- 6-min walk test
- coronary artery bypass graft
- net atrioventricular compliance
- dobutamine stress echocardiography
- effective orifice area
- exercise stress echocardiography
- left ventricular
- left ventricular outflow tract
- mitral valve annuloplasty
- pulmonary arterial pressure
- receiver-operating characteristic
- Received May 2, 2007.
- Revision received November 20, 2007.
- Accepted November 26, 2007.
- American College of Cardiology Foundation
- Grigioni F.,
- Enriquez-Sarano M.,
- Zehr K.J.,
- Bailey K.R.,
- Tajik A.J.
- Lamas G.A.,
- Mitchell G.F.,
- Flaker G.C.,
- et al.,
- Survival and Ventricular Enlargement Investigators
- Magne J.,
- Pibarot P.,
- Dagenais F.,
- Hachicha Z.,
- Dumesnil J.G.,
- Sénéchal M.
- Li M.,
- Dumesnil J.G.,
- Mathieu P.,
- Pibarot P.
- Schwammenthal E.,
- Vered Z.,
- Agranat O.,
- Kaplinsky E.,
- Rabinowitz B.,
- Feinberg M.S.
- Magne J.,
- Mathieu P.,
- Dumesnil J.G.,
- et al.
- Reis G.,
- Motta M.S.,
- Barbosa M.M.,
- Estives W.A.,
- Souza S.F.,
- Bocchi E.A.
- Leavitt J.I.,
- Coats M.H.,
- Falk R.H.
- Bossone E.,
- Rubenfire M.,
- Bach D.S.,
- Ricciardi M.,
- Armstrong W.F.
- Levine R.A.,
- Schwammenthal E.