- Email: [email protected]
Prognostic Value of Right Ventricular Two-Dimensional Global Strain in Patients Referred for Cardiac Surgery
RIGHT AND LEFT VENTRICULAR FUNCTION
Prognostic Value of Right Ventricular Two-Dimensional Global Strain in Patients Referred for Cardiac Surgery Julien Ternacle, MD, Matthieu Berry, MD, Thomas Cognet, MD, Martin Kloeckner, MD, Thibaud Damy, MD, PhD, Jean-Luc Monin, MD, PhD, Jean-Paul Couetil, MD, PhD, Jean-Luc Dubois-Rande, MD, PhD, Pascal Gueret, MD, PhD, and Pascal Lim, MD, PhD, Creteil and Toulouse, France
Background: Right ventricular (RV) function is a strong predictor of patient outcome after cardiac surgery. Limited studies have compared the predictive value of RV global longitudinal strain (RV-GLS) with tricuspid annular plane systolic excursion (TAPSE) and RV fractional area change (RVFAC) in this setting. Methods: The study included 250 patients (66 6 13 years old, LVEF = 52% 6 12%) referred for cardiac surgery (EuroSCORE-II = 4.8% 6 8.0%). RV function before surgery was assessed by RV-GLS by using speckletracking analysis (3-segment from the RV free wall), RVFAC and TAPSE was compared with postoperative outcome defined by 1-month mortality. Results: Overall, 19 patients (7.6%) had RVFAC < 35%, 34 (13.6%) had TAPSE < 16 mm, and 99 (39.6%) had impaired RV-GLS > 21% (35% with normal RVFAC $ 35%). Postoperative death (n = 25) was higher in patients with abnormal RV-GLS > 21% (22% vs 3%; P < .0001), TAPSE < 16 mm (24% vs 8%; P = .007), and RVFAC < 35% (32% vs 9%; P = .001). Mortality was 3% in patients with preserved RV-GLS. In patients with preserved RVFAC $ 35% but abnormal RV-GLS, mortality was similar to that of those with RVFAC < 35% (20% vs 32%; P = .12). Among RV systolic indexes, only RV-GLS was associated with patient outcome by multivariate analysis adjusted to EuroSCORE-II and cardiopulmonary bypass duration. Conclusions: RV-GLS is a sensitive marker of RV dysfunction and correlates with postoperative mortality. (J Am Soc Echocardiogr 2013;26:721-6.) Keywords: Tricuspid annular plane systolic excursion, Right ventricular fractional area change, Speckletracking, Outcome, Cardiac surgery
Postoperative mortality risk is determinant in the decision to refer patients for cardiac surgery. Current guidelines recommend the use of surgery risk scores, such as the logistic EuroSCORE to predict mortality after cardiac surgery.1 These risk scores only include left ventricular systolic function in the risk assessment,2 whereas, in clinical practice, both right ventricular (RV) and left ventricular (LV) systolic function3 are considered in the decision to refer patients for cardiac surgery. The impact of RV function on postoperative outcome has been reported in previous studies, which mainly used RV fractional area change (RVFAC) to quantify RV function.4,5 RVFAC provides a global assessment of RV function but may be less sensitive than
longitudinal RV markers to detect early RV dysfunction. Recent studies demonstrated that longitudinal markers, i.e., tricuspid annular plane systolic excursion (TAPSE), peak systolic tricuspid annular velocity by tissue Doppler imaging (TDI)6 and, more recently, RV-global longitudinal strain (RV-GLS) by speckle-tracking may be superior to RVFAC in predicting outcome of a patient with heart failure.7-9 In the present study, we compared the accuracy of RV systolic function indexes for the assessment of patient outcome after cardiac surgery.
^ pitaux de Paris, Henri Mondor University From AP-HP for Assistance-Publique Ho Hospital, Cardiovascular Department and INSERM U955, Team #3, Creteil, France (J.T., M.K., T.D., J.-L.M., J.-P.C., J.-L.D.-R., P.G., P.L.); and CHU Rangueil, Cardiovascular Department, Toulouse, France (M.B., T.C.).
Study Population This retrospective study included 250 consecutive patients (mean [SD], 66 6 13 years) referred with a comprehensive echocardiography that included a modified 4C view for an RV function study. These patients were selected among the 467 patients referred for cardiac surgery from November 2010 to November 2011. Overall, 200 patients were previously included in a recent published study (n = 425) that aimed to demonstrate the impact of LV longitudinal global strain on the postoperative outcome. The protocol only included patients referred for mitral and/or aortic valve surgery and 721
Reprint requests: Pascal Lim, MD, PhD, Henri Mondor University Hospital, Department of Cardiovascular Medicine and INSERM U955, Team #3, 51 Av de Lattre de Tassigny, 94100 Creteil, France (E-mail: [email protected]). 0894-7317/$36.00 Copyright 2013 by the American Society of Echocardiography. http://dx.doi.org/10.1016/j.echo.2013.03.021
METHODS
722 Ternacle et al
Abbreviations
AUC = Area under curves CABG = Coronary artery bypass graft
CPB = Cardiopulmonary bypass
LV = Left ventricular LVEF = Left ventricular ejection fraction OR = Odds ratio RV-GLS = Right ventricular global longitudinal strain
Journal of the American Society of Echocardiography July 2013
those candidates for coronary artery bypass graft (CABG). Patients referred for cardiac tumor resection, isolated or associated severe tricuspid regurgitation (n = 18), tamponade, cardiac transplantation, or assistance for end-stage heart failure were excluded because the postoperative outcome of this population cannot be assimilated to the remaining cohort. All the patients provided written informed consent approved by our local ethics committee.
RVFAC = Right ventricular fractional area change
Echocardiography Measurements S-lat-TDI = Systolic right All data were acquired before ventricular lateral wall velocity surgery by using a commercially tissue Doppler imaging available Vivid 7 system (GE Vingmed, Horton, Norway), and TAPSE = Tricuspid annular analysis was performed offline plane systolic excursion by using Echo-PAC software TDI = Tissue Doppler imaging (GE Vingmed, Horton, Norway). Comprehensive transthoracic echocardiography included LV apical views (4C, 2C, 3C) and a modified 4C view with a high frame rate (mean, 74 6 17) for RV strain quantification. All measurements were performed offline by an experienced investigator (J.T.), blinded to postoperative outcome. LV volumes and LV ejection fraction (LVEF) were computed by using Simpson biplane method from 2C and 4C apical views. Systolic pulmonary arterial pressure was computed by using a conventional Doppler method from the tricuspid regurgitation. Tricuspid annular plane systolic excursion (TAPSE) was measured from 4C apical view by using M-mode. RV fractional area change (RVFAC) was calculated from 4C apical view by using the conventional formula, i.e., ([end-diastolic area – end-systolic area]/end-diastolic area). According to recent guidelines, RV systolic dysfunction was defined by TAPSE < 16 mm, RVFAC < 35%.10 RV global longitudinal strain (RV-GLS) was computed from the 4C modified apical view focused on the RV free wall by using two-dimensional speckle-tracking software (automated function imaging, EchoPAC; GE Vingmed). Briefly, after manual initialization of the end-systolic endocardial border, the region of interest was automatically positioned to track frame by frame the RV free wall throughout the cardiac cycle. The endocardial contour and width were manually adjusted when necessary to provide optimal tracking. RV-GLS was calculated by averaging only strain curves from the RV free wall (3-segment model) to avoid interaction with LV function. According to a previous study,8 RVGLS > 21% was used to define RV dysfunction. The study was performed in a blinded fashion because RV strain and RVFAC were not included in the echocardiography report. End Point Outcome was evaluated by postoperative death (in-hospital death or 1-month postsurgical death after hospital discharge) and the need of prolonged inotropic support (>48 hours). The decision to initiate inotropic support after cardiac surgery in our institution is only based on the presence of hemodynamic instability (mean blood pressure
<65 mm Hg or systolic blood pressure < 90 mm Hg, oligoanuria). Outcome was obtained from medical records or direct patient interviews, or from the referring physician. Statistical Analysis Continuous variables with a normal distribution are expressed as mean (SD). Dichotomous data are expressed as percentages. To compare numerical data between two groups, paired and unpaired Student t tests were used, as appropriate. Nominal variables were compared by using either the c2 test or the Fisher test. Linear correlation was used to compare RV-GLS, RVFAC, and TAPSE. To identify independent factors associated with outcome, we included previously validated independent predictors (EuroSCORE-II and cardiopulmonary bypass [CPB] duration) and RV systolic indexes (TAPSE, RVFAC, and RV-GLS) in the multivariate stepwise analysis by using a logistic model. RV systolic indexes were included as continuous values in the stepwise analysis. Interaction test was used to compare the prognosis value of RV longitudinal strain in patients referred for CABG and valvular surgery. Reproducibility for TAPSE, RV strain, and RVFAC was performed by a second independent observer (T.C.) in 20 random patients. For intraobserver reproducibility, analysis was repeated with the 20 previous patients 1 month after the first measurement. Reproducibility was expressed by using the coefficient of variation. Two-tailed P values <.05 were considered statistically significant for all analyses, and, for adjustment comparison, P values were considered significant when <.1. RESULTS Of the 250 patients, 126 patients underwent CABG (83 without valvular surgery), 124 (49%) underwent a valvular surgery without CABG. The EuroSCORE- II averaged 4.8% 6 8.0% (range, 1.3%89%). After cardiac surgery, prolonged inotropic support was required in 82 patients (33%), and postoperative death was reported in 25 patients (10%). Most deaths were related to refractory heart failure or septicemia. Baseline characteristics are shown in Table 1. RV-GLS TAPSE, RVFAC, and RV-GLS averaged 21 6 5 mm (range, 5-33 mm), 51% 6 9% (range, 19%-72%) and 21% 6 7% (range, 46% to 3%), respectively. Of the 250 patients, only 19 had abnormal RVFAC < 35% and 34 had abnormal TAPSE, <16 mm. RVFAC and TAPSE correlated well (r = 0.66; P < .0001) and were concordant (kappa = 0.62): 90% of patients with RVFAC < 35% had abnormal TAPSE < 16 mm, and 93% of patients with normal RVFAC had normal TAPSE. In contrast, RV-GLS was less correlated with RVFAC (r = 0.50, P < .0001) (Figure 1A) and with TAPSE (r = 0.43, P < .0001) (Figure 1B). All patients with abnormal RVFAC < 35%, had abnormal RV-GLS (> 21%), but 34% of patients with preserved RVFAC > 35% (n = 80/231) or with TAPSE > 16 mm (n = 73/216) had abnormal RV-GLS. Patients with RVFAC < 35% had more severely impaired RV strain than those with normal RVFAC and abnormal RV strain ( 10% 6 4% vs 15% 6 3%; P < .0001) (mean [SD]). RV Function and Postoperative Hemodynamic Stability Patients who required prolonged inotropic support (n = 82 [33%]) had higher EuroSCORE-II and more prolonged CPB duration (Table 2). Prolonged inotropic support was more frequently required in patients with RVFAC<35% (Figure 2A), TAPSE < 16 mm
Ternacle et al 723
Journal of the American Society of Echocardiography Volume 26 Number 7
for those with preserved RVFAC but impaired RV-GLS (20%; P = .12) (Figure 3D).
Table 1 Baseline patient characteristics All patients (n = 250)
Mean (SD) age, y Women, no. (%) Smoker, no. (%) Diabetes, no. (%) Hypertension, no. (%) Permanent atrial fibrillation, no. (%) NYHA class III-IV, no. (%) Median (SD) creatinine, mmol/L Surgery Mean (SD) CPB duration, min Indications, no. (%) Aortic stenosis Aortic regurgitation Mitral regurgitation Isolated CABG Echocardiography Mean (SD) LVEF by Simpson biplane, % Mean (SD) RVFAC, % RVFAC < 35%, no. (%) Mean (SD) TAPSE, mm TAPSE < 16 mm, no. (%) Mean (SD) RV-GLS 3-segment model, % RV-GLS > 21%, no. (%) EuroSCORE-II LVEF, no. (%) <20% 21%-30% 31%-50% >50% Previous cardiac surgery, no. (%) Chronic lung disease, no. (%) Extracardiacarteriopathy, no. (%) Severe renal impairment, no. (%) Angina at rest, no. (%) Recent myocardial infarction, no. (%) Pulmonary hypertension > 55 mm Hg, no. (%) Critical preoperative state, no. (%) Emergency, no. (%) Active endocarditis, no. (%) Valve surgery, no. (%) Surgery on thoracic aorta, no. (%)
66 6 13 73 (29) 75 (30) 75 (30) 145 (58) 16 (6) 85 (34) 110 6 89 141 6 85 (range, 30-720) 90 (36) 15 (6) 55 (22) 83 (33) 51 6 12 51 6 9 19 (7.6) 21 6 5 34 (13.6) 21 6 7 99 (39.6)
6 (1) 8 (5) 72 (29) 164 (64) 28 (11) 18 (7) 53 (21) 68 (27) 6 (2.4) 43 (17) 19 (7.6) 14 (5.6) 35 (14) 13 (5.2) 160 (64) 25 (10)
NYHA, New York Heart Association.
(Figure 2B), and RV-GLS > 21% (Figure 2C). Inotropic support was similarly required in patients with preserved RVFAC $ 35% but impaired RV-GLS as in those with RVFAC < 35% (Figure 2D).
RV Systolic Function and Postoperative Mortality The patients who died (10%) had higher EuroSCORE-II and longer CPB duration (Table 2). A higher mortality was observed in patients with abnormal RVFAC < 35% (Figure 3A), TAPSE < 16 mm (Figure 3B), and RV-GLS > –21% (Figure 3C). The greatest mortality rate was observed for patients with RVFAC < 35% (32%) and
Independent Factors Associated with Postoperative Outcome Multivariate analysis (Table 3) demonstrated that RV strain, EuroSCORE-II, and CPB duration were associated with postoperative outcome. Results remained unchanged after excluding patients with permanent atrial fibrillation (n = 16). RV-GLS was similarly associated with postoperative death in valvular surgery patients (OR [odds ratio] 1.11 [95% CI, 1.02-1.20]) and in those referred for isolated CABG (OR 1.25 [95% CI, 1.03-1.51]). Patients with impaired RV-GLS had lower postoperative mortality when the CPB duration was <120 minutes (Figure 4). Reproducibility Intra- and interobserver variation coefficient was 9% and 11% for TAPSE, 9.7% and 12% for RVFAC, and 6% and 7% for RV-GLS, respectively (Figure 5). DISCUSSION This study confirms the importance of assessing RV function before cardiac surgery and underlines the need to consider both RVFAC and RV longitudinal strain. Abnormal RVFAC < 35% is associated to the greatest risk of postoperative mortality, probably because abnormal RVFAC reflects a severe and advanced RV dysfunction with both radial and longitudinal RV dysfunction. In patients with preserved RVFAC, RV speckle-tracking appears as a sensitive method to identify early RV dysfunction: the 34% of patients with normal RVFAC but abnormal RV longitudinal strain are at higher risk of postoperative mortality. Analysis of accumulating data suggests that RV function is a strong predictor of outcome in patients with heart failure.6,9,11 RV dysfunction plays a key role in renal and liver congestion and in the development of cardiogenic shock. In patients referred for cardiac surgery, small sample size studies demonstrated that abnormal RVFAC (cutoff between 32% and 35%) before cardiac surgery is a strong predictor of postoperative mortality.4,5 RVFAC requires an accurate delineation of RV borders but is a simple method to globally assess radial and longitudinal RV function. However, similar to RV ejection fraction, RVFAC is less sensitive to detect early impairment of RV function that seems to first affect the longitudinal component. RV longitudinal function usually is quantified from the RV free wall because this part of the RV contributes in 80% to the RV stroke volume. In addition, the inclusion of the septal wall in RV global strain calculation would overlap with LV longitudinal global strain, which also plays an important role on postoperative outcome. Indeed, in a larger population study (n = 425) without specific data for RV study, we showed that longitudinal strain is superior to LVEF for predicting postoperative outcome, especially in patients with preserved LVEF.3 Accuracy of longitudinal markers for assessing RV dysfunction in patients with heart failure has been reported by Damy et al.6 who showed that TAPSE (area under curves [AUC] = 0.80; P < .0001) and peak systolic RV lateral wall velocity by TDI (S-lat-TDI, AUC = 0.82; P < .0001) are more accurate than RVFAC (AUC = 0.76; P < .0001) to predict cardiac outcome.
724 Ternacle et al
Journal of the American Society of Echocardiography July 2013
Figure 1 Correlation between RV-GLS and RVFAC (A) and TAPSE (B). Table 2 Univariate predictors of death and the need of inotropic support
CPB duration, min* EuroSCORE-II, %* LVEF, %* <20% 21%-30% 31%-50% >50% RV-GLS, %* RV-GLS > 21%, % RVFAC, %* RVFAC < 35%, % TAPSE, mm* TAPSE < 16 mm, %
Death (n = 25)
Survival (n = 225)
P
220 6 139 13 6 14 47 6 14 1 3 8 13 15 6 8 84 46 6 12 24 18 6 6 31
132 6 71 466 52 6 12 5 5 63 151 22 6 7 34 52 6 9 6 21 6 5 12
<.001 <.001 .04 .05
<.001 <.001 .001 .001 .003 .007
Inotropic support >48 hours (n = 82)
No inotropic support (n = 168)
180 6 108 8.5 6 12 46.5 6 14 3 7 34 38 19 6 8 55 48 6 10 15 20 6 5 24
120 6 62 3.0 6 3.7 53.6 6 10 3 1 36 121 22 6 7 32 52 6 8 4 22 6 4 8
P
<.001 <.001 <.001 <.001
<.001 <.001 <.001 .005 .002 <.001
*Values are mean (SD).
Figure 2 Prolonged inotropic support according to RV systolic function by RVFAC (A), TAPSE (B), and RV-GLS (C), and the combination of RVFAC and RV-GLS (D). In a recent study, the same group demonstrated that RV longitudinal strain quantified by speckle-tracking is superior to TAPSE and S-lat-TDI in characterizing RV dysfunction.8 The superiority
of RV-GLS over TAPSE and S-lat-TDI may be explained by a more global assessment of RV longitudinal function provided by the speckle-tracking approach. In addition, speckle-tracking
Ternacle et al 725
Journal of the American Society of Echocardiography Volume 26 Number 7
Table 3 Independent predictors of post-operative outcome Death (n = 25)
EuroSCORE. per % CPB duration > 120 min RV-GLS
Inotropic support >48 hours (n = 82)
OR (95% CI)
P
OR (95% CI)
P
1.06 (1.02-1.10) 3.53 (1.21-10.29) 1.13 (1.04-1.21)
.002 .02 .0006
1.11 (1.04-1.18) 2.82 (1.51-5.26) 1.08 (1.03-1.13)
.002 .001 .002
Figure 3 Postoperative mortality according to RV systolic function by RVFAC (A), TAPSE (B), and RV-GLS (C), and the combination of RVFAC and RV-GLS (D). is less angle dependent than TDI, and RV global strain is probably less influenced by radial motion than TAPSE, which correlated more closely with RVFAC. This may explain the greater sensitivity of RV-GLS in detecting abnormal longitudinal dysfunction, especially in the subgroup of patients with normal RVFAC. The superiority of RV strain over standard RV systolic indexes may be particularly interesting for detecting early RV dysfunction in patients commonly viewed as free of RV dysfunction, i.e., those without significant RV valvular disease or right coronary occlusion. Finally, one advantage of RV strain is its applicability both for valvular and coronary artery disease. Further studies should evaluate the clinical interest of assessing RV strain during and after cardiac surgery because this can be easily obtained from transesophageal echocardiography in the operating room.12 Identification of RV dysfunction before cardiac surgery should be carefully performed with an accurate method because of its strong impact on patient outcome. The presence of severe RV dysfunction may modify the medical strategy and the postoperative management. Alternative treatments, such as percutaneous procedures, that do not affect RV function may be considered in high-risk patients with severe RV dysfunction (RVFAC < 35%).13 Conversely, if cardiac surgery is maintained, then intensive care teams and surgeons should be aware of the risk of refractory postoperative shock, and cardiac assistance may be discussed and scheduled before cardiac surgery in case of an unfavorable outcome. In patients with subclinical RV dysfunction
Figure 4 Postoperative mortality (A) according to RV-GLS and CPB duration. P* for interaction term. with normal RVFAC and an abnormal strain value, pericardectomy, and CPB will aggravate the RV dysfunction and expose patients to a greater risk of hemodynamic instability and death.4,14 As suggested by our results, postoperative mortality may be improved by reducing CPB duration. Then, the presence of a subclinical RV dysfunction may lead to modifying the surgery strategy, and the risk-benefit balance should be individually discussed with the surgeon.
726 Ternacle et al
Journal of the American Society of Echocardiography July 2013
3.
4.
5.
6.
Figure 5 Intraobserver (A) and interobserver (B) Bland Altman analysis for RV strain. Limitations Patients with severe tricuspid regurgitation were excluded from the study to avoid bias related to the RV hyperkinesia and also because RV parameters optimal cutoff values probably differed between functional and organic tricuspid regurgitation. To assess longitudinal RV function, we only used RV strain and TAPSE and not TDI because of a poor Doppler signal quality in a large number of patients. Moreover, the cutoff of 21% may differ between vendors, and these results are valid only for the GE system. Lastly, specific criteria used by our physicians to discontinue inotropic support were well established but can be different in other centers.
7.
8.
9.
10.
CONCLUSIONS RV dysfunction strongly impacts on postoperative outcome in patients referred for cardiac surgery. Increased postoperative mortality is observed not only in patients with RVFAC < 35% but also in those with preserved RVFAC but impaired RV-GLS.
11.
REFERENCES
13.
1. Vahanian A, Baumgartner H, Bax J, Butchart E, Dion R, Filippatos G, et al. Guidelines on the management of valvular heart disease: the Task Force on the Management of Valvular Heart Disease of the European Society of Cardiology. Eur Heart J 2007;28:230-68. 2. Roques F, Nashef SA, Michel P, Gauducheau E, de Vincentiis C, Baudet E, et al. Risk factors and outcome in European cardiac surgery: analysis of the
12.
14.
EuroSCORE multinational database of 19030 patients. Eur J Cardiothorac Surg 1999;15:816-22. discussion 822-3. Ternacle J, Berry M, Alonso E, Kloeckner M, Couetil JP, Rande JL, et al. Incremental value of global longitudinal strain for predicting early outcome after cardiac surgery. Eur Heart J Cardiovasc Imaging 2013;14: 77-84. Maslow AD, Regan MM, Panzica P, Heindel S, Mashikian J, Comunale ME. Precardiopulmonary bypass right ventricular function is associated with poor outcome after coronary artery bypass grafting in patients with severe left ventricular systolic dysfunction. Anesth Analg 2002; 95:1507-18. Haddad F, Denault AY, Couture P, Cartier R, Pellerin M, Levesque S, et al. Right ventricular myocardial performance index predicts perioperative mortality or circulatory failure in high-risk valvular surgery. J Am Soc Echocardiogr 2007;20:1065-72. Damy T, Viallet C, Lairez O, Deswarte G, Paulino A, Maison P, et al. Comparison of four right ventricular systolic echocardiographic parameters to predict adverse outcomes in chronic heart failure. Eur J Heart Fail 2009; 11:818-24. Mor-Avi V, Lang RM, Badano LP, Belohlavek M, Cardim NM, Derumeaux G, et al. Current and evolving echocardiographic techniques for the quantitative evaluation of cardiac mechanics: ASE/EAE consensus statement on methodology and indications endorsed by the Japanese Society of Echocardiography. J Am Soc Echocardiogr 2011;24:277-313. Guendouz S, Rappeneau S, Nahum J, Dubois-Rande JL, Gueret P, Monin JL, et al. Prognostic significance and normal values of 2D strain to assess right ventricular systolic function in chronic heart failure. Circ J 2012;76:127-36. Donal E, Roulaud M, Raud-Raynier P, De Bisschop C, Leclercq C, Derumeaux G, et al. Echocardiographic right ventricular strain analysis in chronic heart failure. Eur J Echocardiogr 2007;8:449-56. Rudski LG, Lai WW, Afilalo J, Hua L, Handschumacher MD, Chandrasekaran K, et al. Guidelines for the echocardiographic assessment of the right heart in adults: a report from the American Society of Echocardiography endorsed by the European Association of Echocardiography, a registered branch of the European Society of Cardiology, and the Canadian Society of Echocardiography. J Am Soc Echocardiogr 2010;23: 685-713. quiz 786-8. Haddad F, Couture P, Tousignant C, Denault AY. The right ventricle in cardiac surgery, a perioperative perspective: II. Pathophysiology, clinical importance, and management. Anesth Analg 2009;108:422-33. Tousignant C, Desmet M, Bowry R, Harrington AM, Cruz JD, Mazer CD. Speckle tracking for the intraoperative assessment of right ventricular function: a feasibility study. J Cardiothorac Vasc Anesth 2010;24:275-9. Kempny A, Diller GP, Kaleschke G, Orwat S, Funke A, Schmidt R, et al. Impact of transcatheter aortic valve implantation or surgical aortic valve replacement on right ventricular function. Heart 2012;98: 1299-304. Salis S, Mazzanti VV, Merli G, Salvi L, Tedesco CC, Veglia F, et al. Cardiopulmonary bypass duration is an independent predictor of morbidity and mortality after cardiac surgery. J Cardiothorac Vasc Anesth 2008;22: 814-22.
Prognostic Value of Right Ventricular Two-Dimensional Global Strain in Patients Referred for Cardiac Surgery Julien Ternacle, MD, Matthieu Berry, MD, Thomas Cognet, MD, Martin Kloeckner, MD, Thibaud Damy, MD, PhD, Jean-Luc Monin, MD, PhD, Jean-Paul Couetil, MD, PhD, Jean-Luc Dubois-Rande, MD, PhD, Pascal Gueret, MD, PhD, and Pascal Lim, MD, PhD, Creteil and Toulouse, France
Background: Right ventricular (RV) function is a strong predictor of patient outcome after cardiac surgery. Limited studies have compared the predictive value of RV global longitudinal strain (RV-GLS) with tricuspid annular plane systolic excursion (TAPSE) and RV fractional area change (RVFAC) in this setting. Methods: The study included 250 patients (66 6 13 years old, LVEF = 52% 6 12%) referred for cardiac surgery (EuroSCORE-II = 4.8% 6 8.0%). RV function before surgery was assessed by RV-GLS by using speckletracking analysis (3-segment from the RV free wall), RVFAC and TAPSE was compared with postoperative outcome defined by 1-month mortality. Results: Overall, 19 patients (7.6%) had RVFAC < 35%, 34 (13.6%) had TAPSE < 16 mm, and 99 (39.6%) had impaired RV-GLS > 21% (35% with normal RVFAC $ 35%). Postoperative death (n = 25) was higher in patients with abnormal RV-GLS > 21% (22% vs 3%; P < .0001), TAPSE < 16 mm (24% vs 8%; P = .007), and RVFAC < 35% (32% vs 9%; P = .001). Mortality was 3% in patients with preserved RV-GLS. In patients with preserved RVFAC $ 35% but abnormal RV-GLS, mortality was similar to that of those with RVFAC < 35% (20% vs 32%; P = .12). Among RV systolic indexes, only RV-GLS was associated with patient outcome by multivariate analysis adjusted to EuroSCORE-II and cardiopulmonary bypass duration. Conclusions: RV-GLS is a sensitive marker of RV dysfunction and correlates with postoperative mortality. (J Am Soc Echocardiogr 2013;26:721-6.) Keywords: Tricuspid annular plane systolic excursion, Right ventricular fractional area change, Speckletracking, Outcome, Cardiac surgery
Postoperative mortality risk is determinant in the decision to refer patients for cardiac surgery. Current guidelines recommend the use of surgery risk scores, such as the logistic EuroSCORE to predict mortality after cardiac surgery.1 These risk scores only include left ventricular systolic function in the risk assessment,2 whereas, in clinical practice, both right ventricular (RV) and left ventricular (LV) systolic function3 are considered in the decision to refer patients for cardiac surgery. The impact of RV function on postoperative outcome has been reported in previous studies, which mainly used RV fractional area change (RVFAC) to quantify RV function.4,5 RVFAC provides a global assessment of RV function but may be less sensitive than
longitudinal RV markers to detect early RV dysfunction. Recent studies demonstrated that longitudinal markers, i.e., tricuspid annular plane systolic excursion (TAPSE), peak systolic tricuspid annular velocity by tissue Doppler imaging (TDI)6 and, more recently, RV-global longitudinal strain (RV-GLS) by speckle-tracking may be superior to RVFAC in predicting outcome of a patient with heart failure.7-9 In the present study, we compared the accuracy of RV systolic function indexes for the assessment of patient outcome after cardiac surgery.
^ pitaux de Paris, Henri Mondor University From AP-HP for Assistance-Publique Ho Hospital, Cardiovascular Department and INSERM U955, Team #3, Creteil, France (J.T., M.K., T.D., J.-L.M., J.-P.C., J.-L.D.-R., P.G., P.L.); and CHU Rangueil, Cardiovascular Department, Toulouse, France (M.B., T.C.).
Study Population This retrospective study included 250 consecutive patients (mean [SD], 66 6 13 years) referred with a comprehensive echocardiography that included a modified 4C view for an RV function study. These patients were selected among the 467 patients referred for cardiac surgery from November 2010 to November 2011. Overall, 200 patients were previously included in a recent published study (n = 425) that aimed to demonstrate the impact of LV longitudinal global strain on the postoperative outcome. The protocol only included patients referred for mitral and/or aortic valve surgery and 721
Reprint requests: Pascal Lim, MD, PhD, Henri Mondor University Hospital, Department of Cardiovascular Medicine and INSERM U955, Team #3, 51 Av de Lattre de Tassigny, 94100 Creteil, France (E-mail: [email protected]). 0894-7317/$36.00 Copyright 2013 by the American Society of Echocardiography. http://dx.doi.org/10.1016/j.echo.2013.03.021
METHODS
722 Ternacle et al
Abbreviations
AUC = Area under curves CABG = Coronary artery bypass graft
CPB = Cardiopulmonary bypass
LV = Left ventricular LVEF = Left ventricular ejection fraction OR = Odds ratio RV-GLS = Right ventricular global longitudinal strain
Journal of the American Society of Echocardiography July 2013
those candidates for coronary artery bypass graft (CABG). Patients referred for cardiac tumor resection, isolated or associated severe tricuspid regurgitation (n = 18), tamponade, cardiac transplantation, or assistance for end-stage heart failure were excluded because the postoperative outcome of this population cannot be assimilated to the remaining cohort. All the patients provided written informed consent approved by our local ethics committee.
RVFAC = Right ventricular fractional area change
Echocardiography Measurements S-lat-TDI = Systolic right All data were acquired before ventricular lateral wall velocity surgery by using a commercially tissue Doppler imaging available Vivid 7 system (GE Vingmed, Horton, Norway), and TAPSE = Tricuspid annular analysis was performed offline plane systolic excursion by using Echo-PAC software TDI = Tissue Doppler imaging (GE Vingmed, Horton, Norway). Comprehensive transthoracic echocardiography included LV apical views (4C, 2C, 3C) and a modified 4C view with a high frame rate (mean, 74 6 17) for RV strain quantification. All measurements were performed offline by an experienced investigator (J.T.), blinded to postoperative outcome. LV volumes and LV ejection fraction (LVEF) were computed by using Simpson biplane method from 2C and 4C apical views. Systolic pulmonary arterial pressure was computed by using a conventional Doppler method from the tricuspid regurgitation. Tricuspid annular plane systolic excursion (TAPSE) was measured from 4C apical view by using M-mode. RV fractional area change (RVFAC) was calculated from 4C apical view by using the conventional formula, i.e., ([end-diastolic area – end-systolic area]/end-diastolic area). According to recent guidelines, RV systolic dysfunction was defined by TAPSE < 16 mm, RVFAC < 35%.10 RV global longitudinal strain (RV-GLS) was computed from the 4C modified apical view focused on the RV free wall by using two-dimensional speckle-tracking software (automated function imaging, EchoPAC; GE Vingmed). Briefly, after manual initialization of the end-systolic endocardial border, the region of interest was automatically positioned to track frame by frame the RV free wall throughout the cardiac cycle. The endocardial contour and width were manually adjusted when necessary to provide optimal tracking. RV-GLS was calculated by averaging only strain curves from the RV free wall (3-segment model) to avoid interaction with LV function. According to a previous study,8 RVGLS > 21% was used to define RV dysfunction. The study was performed in a blinded fashion because RV strain and RVFAC were not included in the echocardiography report. End Point Outcome was evaluated by postoperative death (in-hospital death or 1-month postsurgical death after hospital discharge) and the need of prolonged inotropic support (>48 hours). The decision to initiate inotropic support after cardiac surgery in our institution is only based on the presence of hemodynamic instability (mean blood pressure
<65 mm Hg or systolic blood pressure < 90 mm Hg, oligoanuria). Outcome was obtained from medical records or direct patient interviews, or from the referring physician. Statistical Analysis Continuous variables with a normal distribution are expressed as mean (SD). Dichotomous data are expressed as percentages. To compare numerical data between two groups, paired and unpaired Student t tests were used, as appropriate. Nominal variables were compared by using either the c2 test or the Fisher test. Linear correlation was used to compare RV-GLS, RVFAC, and TAPSE. To identify independent factors associated with outcome, we included previously validated independent predictors (EuroSCORE-II and cardiopulmonary bypass [CPB] duration) and RV systolic indexes (TAPSE, RVFAC, and RV-GLS) in the multivariate stepwise analysis by using a logistic model. RV systolic indexes were included as continuous values in the stepwise analysis. Interaction test was used to compare the prognosis value of RV longitudinal strain in patients referred for CABG and valvular surgery. Reproducibility for TAPSE, RV strain, and RVFAC was performed by a second independent observer (T.C.) in 20 random patients. For intraobserver reproducibility, analysis was repeated with the 20 previous patients 1 month after the first measurement. Reproducibility was expressed by using the coefficient of variation. Two-tailed P values <.05 were considered statistically significant for all analyses, and, for adjustment comparison, P values were considered significant when <.1. RESULTS Of the 250 patients, 126 patients underwent CABG (83 without valvular surgery), 124 (49%) underwent a valvular surgery without CABG. The EuroSCORE- II averaged 4.8% 6 8.0% (range, 1.3%89%). After cardiac surgery, prolonged inotropic support was required in 82 patients (33%), and postoperative death was reported in 25 patients (10%). Most deaths were related to refractory heart failure or septicemia. Baseline characteristics are shown in Table 1. RV-GLS TAPSE, RVFAC, and RV-GLS averaged 21 6 5 mm (range, 5-33 mm), 51% 6 9% (range, 19%-72%) and 21% 6 7% (range, 46% to 3%), respectively. Of the 250 patients, only 19 had abnormal RVFAC < 35% and 34 had abnormal TAPSE, <16 mm. RVFAC and TAPSE correlated well (r = 0.66; P < .0001) and were concordant (kappa = 0.62): 90% of patients with RVFAC < 35% had abnormal TAPSE < 16 mm, and 93% of patients with normal RVFAC had normal TAPSE. In contrast, RV-GLS was less correlated with RVFAC (r = 0.50, P < .0001) (Figure 1A) and with TAPSE (r = 0.43, P < .0001) (Figure 1B). All patients with abnormal RVFAC < 35%, had abnormal RV-GLS (> 21%), but 34% of patients with preserved RVFAC > 35% (n = 80/231) or with TAPSE > 16 mm (n = 73/216) had abnormal RV-GLS. Patients with RVFAC < 35% had more severely impaired RV strain than those with normal RVFAC and abnormal RV strain ( 10% 6 4% vs 15% 6 3%; P < .0001) (mean [SD]). RV Function and Postoperative Hemodynamic Stability Patients who required prolonged inotropic support (n = 82 [33%]) had higher EuroSCORE-II and more prolonged CPB duration (Table 2). Prolonged inotropic support was more frequently required in patients with RVFAC<35% (Figure 2A), TAPSE < 16 mm
Ternacle et al 723
Journal of the American Society of Echocardiography Volume 26 Number 7
for those with preserved RVFAC but impaired RV-GLS (20%; P = .12) (Figure 3D).
Table 1 Baseline patient characteristics All patients (n = 250)
Mean (SD) age, y Women, no. (%) Smoker, no. (%) Diabetes, no. (%) Hypertension, no. (%) Permanent atrial fibrillation, no. (%) NYHA class III-IV, no. (%) Median (SD) creatinine, mmol/L Surgery Mean (SD) CPB duration, min Indications, no. (%) Aortic stenosis Aortic regurgitation Mitral regurgitation Isolated CABG Echocardiography Mean (SD) LVEF by Simpson biplane, % Mean (SD) RVFAC, % RVFAC < 35%, no. (%) Mean (SD) TAPSE, mm TAPSE < 16 mm, no. (%) Mean (SD) RV-GLS 3-segment model, % RV-GLS > 21%, no. (%) EuroSCORE-II LVEF, no. (%) <20% 21%-30% 31%-50% >50% Previous cardiac surgery, no. (%) Chronic lung disease, no. (%) Extracardiacarteriopathy, no. (%) Severe renal impairment, no. (%) Angina at rest, no. (%) Recent myocardial infarction, no. (%) Pulmonary hypertension > 55 mm Hg, no. (%) Critical preoperative state, no. (%) Emergency, no. (%) Active endocarditis, no. (%) Valve surgery, no. (%) Surgery on thoracic aorta, no. (%)
66 6 13 73 (29) 75 (30) 75 (30) 145 (58) 16 (6) 85 (34) 110 6 89 141 6 85 (range, 30-720) 90 (36) 15 (6) 55 (22) 83 (33) 51 6 12 51 6 9 19 (7.6) 21 6 5 34 (13.6) 21 6 7 99 (39.6)
6 (1) 8 (5) 72 (29) 164 (64) 28 (11) 18 (7) 53 (21) 68 (27) 6 (2.4) 43 (17) 19 (7.6) 14 (5.6) 35 (14) 13 (5.2) 160 (64) 25 (10)
NYHA, New York Heart Association.
(Figure 2B), and RV-GLS > 21% (Figure 2C). Inotropic support was similarly required in patients with preserved RVFAC $ 35% but impaired RV-GLS as in those with RVFAC < 35% (Figure 2D).
RV Systolic Function and Postoperative Mortality The patients who died (10%) had higher EuroSCORE-II and longer CPB duration (Table 2). A higher mortality was observed in patients with abnormal RVFAC < 35% (Figure 3A), TAPSE < 16 mm (Figure 3B), and RV-GLS > –21% (Figure 3C). The greatest mortality rate was observed for patients with RVFAC < 35% (32%) and
Independent Factors Associated with Postoperative Outcome Multivariate analysis (Table 3) demonstrated that RV strain, EuroSCORE-II, and CPB duration were associated with postoperative outcome. Results remained unchanged after excluding patients with permanent atrial fibrillation (n = 16). RV-GLS was similarly associated with postoperative death in valvular surgery patients (OR [odds ratio] 1.11 [95% CI, 1.02-1.20]) and in those referred for isolated CABG (OR 1.25 [95% CI, 1.03-1.51]). Patients with impaired RV-GLS had lower postoperative mortality when the CPB duration was <120 minutes (Figure 4). Reproducibility Intra- and interobserver variation coefficient was 9% and 11% for TAPSE, 9.7% and 12% for RVFAC, and 6% and 7% for RV-GLS, respectively (Figure 5). DISCUSSION This study confirms the importance of assessing RV function before cardiac surgery and underlines the need to consider both RVFAC and RV longitudinal strain. Abnormal RVFAC < 35% is associated to the greatest risk of postoperative mortality, probably because abnormal RVFAC reflects a severe and advanced RV dysfunction with both radial and longitudinal RV dysfunction. In patients with preserved RVFAC, RV speckle-tracking appears as a sensitive method to identify early RV dysfunction: the 34% of patients with normal RVFAC but abnormal RV longitudinal strain are at higher risk of postoperative mortality. Analysis of accumulating data suggests that RV function is a strong predictor of outcome in patients with heart failure.6,9,11 RV dysfunction plays a key role in renal and liver congestion and in the development of cardiogenic shock. In patients referred for cardiac surgery, small sample size studies demonstrated that abnormal RVFAC (cutoff between 32% and 35%) before cardiac surgery is a strong predictor of postoperative mortality.4,5 RVFAC requires an accurate delineation of RV borders but is a simple method to globally assess radial and longitudinal RV function. However, similar to RV ejection fraction, RVFAC is less sensitive to detect early impairment of RV function that seems to first affect the longitudinal component. RV longitudinal function usually is quantified from the RV free wall because this part of the RV contributes in 80% to the RV stroke volume. In addition, the inclusion of the septal wall in RV global strain calculation would overlap with LV longitudinal global strain, which also plays an important role on postoperative outcome. Indeed, in a larger population study (n = 425) without specific data for RV study, we showed that longitudinal strain is superior to LVEF for predicting postoperative outcome, especially in patients with preserved LVEF.3 Accuracy of longitudinal markers for assessing RV dysfunction in patients with heart failure has been reported by Damy et al.6 who showed that TAPSE (area under curves [AUC] = 0.80; P < .0001) and peak systolic RV lateral wall velocity by TDI (S-lat-TDI, AUC = 0.82; P < .0001) are more accurate than RVFAC (AUC = 0.76; P < .0001) to predict cardiac outcome.
724 Ternacle et al
Journal of the American Society of Echocardiography July 2013
Figure 1 Correlation between RV-GLS and RVFAC (A) and TAPSE (B). Table 2 Univariate predictors of death and the need of inotropic support
CPB duration, min* EuroSCORE-II, %* LVEF, %* <20% 21%-30% 31%-50% >50% RV-GLS, %* RV-GLS > 21%, % RVFAC, %* RVFAC < 35%, % TAPSE, mm* TAPSE < 16 mm, %
Death (n = 25)
Survival (n = 225)
P
220 6 139 13 6 14 47 6 14 1 3 8 13 15 6 8 84 46 6 12 24 18 6 6 31
132 6 71 466 52 6 12 5 5 63 151 22 6 7 34 52 6 9 6 21 6 5 12
<.001 <.001 .04 .05
<.001 <.001 .001 .001 .003 .007
Inotropic support >48 hours (n = 82)
No inotropic support (n = 168)
180 6 108 8.5 6 12 46.5 6 14 3 7 34 38 19 6 8 55 48 6 10 15 20 6 5 24
120 6 62 3.0 6 3.7 53.6 6 10 3 1 36 121 22 6 7 32 52 6 8 4 22 6 4 8
P
<.001 <.001 <.001 <.001
<.001 <.001 <.001 .005 .002 <.001
*Values are mean (SD).
Figure 2 Prolonged inotropic support according to RV systolic function by RVFAC (A), TAPSE (B), and RV-GLS (C), and the combination of RVFAC and RV-GLS (D). In a recent study, the same group demonstrated that RV longitudinal strain quantified by speckle-tracking is superior to TAPSE and S-lat-TDI in characterizing RV dysfunction.8 The superiority
of RV-GLS over TAPSE and S-lat-TDI may be explained by a more global assessment of RV longitudinal function provided by the speckle-tracking approach. In addition, speckle-tracking
Ternacle et al 725
Journal of the American Society of Echocardiography Volume 26 Number 7
Table 3 Independent predictors of post-operative outcome Death (n = 25)
EuroSCORE. per % CPB duration > 120 min RV-GLS
Inotropic support >48 hours (n = 82)
OR (95% CI)
P
OR (95% CI)
P
1.06 (1.02-1.10) 3.53 (1.21-10.29) 1.13 (1.04-1.21)
.002 .02 .0006
1.11 (1.04-1.18) 2.82 (1.51-5.26) 1.08 (1.03-1.13)
.002 .001 .002
Figure 3 Postoperative mortality according to RV systolic function by RVFAC (A), TAPSE (B), and RV-GLS (C), and the combination of RVFAC and RV-GLS (D). is less angle dependent than TDI, and RV global strain is probably less influenced by radial motion than TAPSE, which correlated more closely with RVFAC. This may explain the greater sensitivity of RV-GLS in detecting abnormal longitudinal dysfunction, especially in the subgroup of patients with normal RVFAC. The superiority of RV strain over standard RV systolic indexes may be particularly interesting for detecting early RV dysfunction in patients commonly viewed as free of RV dysfunction, i.e., those without significant RV valvular disease or right coronary occlusion. Finally, one advantage of RV strain is its applicability both for valvular and coronary artery disease. Further studies should evaluate the clinical interest of assessing RV strain during and after cardiac surgery because this can be easily obtained from transesophageal echocardiography in the operating room.12 Identification of RV dysfunction before cardiac surgery should be carefully performed with an accurate method because of its strong impact on patient outcome. The presence of severe RV dysfunction may modify the medical strategy and the postoperative management. Alternative treatments, such as percutaneous procedures, that do not affect RV function may be considered in high-risk patients with severe RV dysfunction (RVFAC < 35%).13 Conversely, if cardiac surgery is maintained, then intensive care teams and surgeons should be aware of the risk of refractory postoperative shock, and cardiac assistance may be discussed and scheduled before cardiac surgery in case of an unfavorable outcome. In patients with subclinical RV dysfunction
Figure 4 Postoperative mortality (A) according to RV-GLS and CPB duration. P* for interaction term. with normal RVFAC and an abnormal strain value, pericardectomy, and CPB will aggravate the RV dysfunction and expose patients to a greater risk of hemodynamic instability and death.4,14 As suggested by our results, postoperative mortality may be improved by reducing CPB duration. Then, the presence of a subclinical RV dysfunction may lead to modifying the surgery strategy, and the risk-benefit balance should be individually discussed with the surgeon.
726 Ternacle et al
Journal of the American Society of Echocardiography July 2013
3.
4.
5.
6.
Figure 5 Intraobserver (A) and interobserver (B) Bland Altman analysis for RV strain. Limitations Patients with severe tricuspid regurgitation were excluded from the study to avoid bias related to the RV hyperkinesia and also because RV parameters optimal cutoff values probably differed between functional and organic tricuspid regurgitation. To assess longitudinal RV function, we only used RV strain and TAPSE and not TDI because of a poor Doppler signal quality in a large number of patients. Moreover, the cutoff of 21% may differ between vendors, and these results are valid only for the GE system. Lastly, specific criteria used by our physicians to discontinue inotropic support were well established but can be different in other centers.
7.
8.
9.
10.
CONCLUSIONS RV dysfunction strongly impacts on postoperative outcome in patients referred for cardiac surgery. Increased postoperative mortality is observed not only in patients with RVFAC < 35% but also in those with preserved RVFAC but impaired RV-GLS.
11.
REFERENCES
13.
1. Vahanian A, Baumgartner H, Bax J, Butchart E, Dion R, Filippatos G, et al. Guidelines on the management of valvular heart disease: the Task Force on the Management of Valvular Heart Disease of the European Society of Cardiology. Eur Heart J 2007;28:230-68. 2. Roques F, Nashef SA, Michel P, Gauducheau E, de Vincentiis C, Baudet E, et al. Risk factors and outcome in European cardiac surgery: analysis of the
12.
14.
EuroSCORE multinational database of 19030 patients. Eur J Cardiothorac Surg 1999;15:816-22. discussion 822-3. Ternacle J, Berry M, Alonso E, Kloeckner M, Couetil JP, Rande JL, et al. Incremental value of global longitudinal strain for predicting early outcome after cardiac surgery. Eur Heart J Cardiovasc Imaging 2013;14: 77-84. Maslow AD, Regan MM, Panzica P, Heindel S, Mashikian J, Comunale ME. Precardiopulmonary bypass right ventricular function is associated with poor outcome after coronary artery bypass grafting in patients with severe left ventricular systolic dysfunction. Anesth Analg 2002; 95:1507-18. Haddad F, Denault AY, Couture P, Cartier R, Pellerin M, Levesque S, et al. Right ventricular myocardial performance index predicts perioperative mortality or circulatory failure in high-risk valvular surgery. J Am Soc Echocardiogr 2007;20:1065-72. Damy T, Viallet C, Lairez O, Deswarte G, Paulino A, Maison P, et al. Comparison of four right ventricular systolic echocardiographic parameters to predict adverse outcomes in chronic heart failure. Eur J Heart Fail 2009; 11:818-24. Mor-Avi V, Lang RM, Badano LP, Belohlavek M, Cardim NM, Derumeaux G, et al. Current and evolving echocardiographic techniques for the quantitative evaluation of cardiac mechanics: ASE/EAE consensus statement on methodology and indications endorsed by the Japanese Society of Echocardiography. J Am Soc Echocardiogr 2011;24:277-313. Guendouz S, Rappeneau S, Nahum J, Dubois-Rande JL, Gueret P, Monin JL, et al. Prognostic significance and normal values of 2D strain to assess right ventricular systolic function in chronic heart failure. Circ J 2012;76:127-36. Donal E, Roulaud M, Raud-Raynier P, De Bisschop C, Leclercq C, Derumeaux G, et al. Echocardiographic right ventricular strain analysis in chronic heart failure. Eur J Echocardiogr 2007;8:449-56. Rudski LG, Lai WW, Afilalo J, Hua L, Handschumacher MD, Chandrasekaran K, et al. Guidelines for the echocardiographic assessment of the right heart in adults: a report from the American Society of Echocardiography endorsed by the European Association of Echocardiography, a registered branch of the European Society of Cardiology, and the Canadian Society of Echocardiography. J Am Soc Echocardiogr 2010;23: 685-713. quiz 786-8. Haddad F, Couture P, Tousignant C, Denault AY. The right ventricle in cardiac surgery, a perioperative perspective: II. Pathophysiology, clinical importance, and management. Anesth Analg 2009;108:422-33. Tousignant C, Desmet M, Bowry R, Harrington AM, Cruz JD, Mazer CD. Speckle tracking for the intraoperative assessment of right ventricular function: a feasibility study. J Cardiothorac Vasc Anesth 2010;24:275-9. Kempny A, Diller GP, Kaleschke G, Orwat S, Funke A, Schmidt R, et al. Impact of transcatheter aortic valve implantation or surgical aortic valve replacement on right ventricular function. Heart 2012;98: 1299-304. Salis S, Mazzanti VV, Merli G, Salvi L, Tedesco CC, Veglia F, et al. Cardiopulmonary bypass duration is an independent predictor of morbidity and mortality after cardiac surgery. J Cardiothorac Vasc Anesth 2008;22: 814-22.