- Email: [email protected]
Preoperative Right Ventricular Dysfunction Indicates High Vasoactive Support Needed After Cardiac Surgery
ARTICLE IN PRESS Journal of Cardiothoracic and Vascular Anesthesia 000 (2018) 18
Contents lists available at ScienceDirect
Journal of Cardiothoracic and Vascular Anesthesia journal homepage: www.jcvaonline.com
Original Article
Preoperative Right Ventricular Dysfunction Indicates High Vasoactive Support Needed After Cardiac Surgery 1XD XPei-Chi Ting, D2X XMD, MSc*, D3X XVictor C-C. Wu, D4X XMDy, D5X XChia-Chih Liao, D6X XMD, PhD*, D7X XAn-Hsun Chou, D8X XMD, PhD*, D9X XFeng-Chun Tsai, D10X XMDz, D1X XPyng-Jing Lin, D12X XMDz, ,1 D13X XChun-Yu Chen, D14X XMD, PhD*, D15X XShao-Wei Chen, D16X XMDz *
Department of Anesthesiology, Chang Gung Memorial Hospital, Taoyuan, Taiwan y Department of Cardiology, Chang Gung Memorial Hospital, Taoyuan, Taiwan z Division of Thoracic and Cardiovascular Surgery, Chang Gung Memorial Hospital, Taoyuan, Taiwan
Objective: The aim of this study was to explore the relationship between preoperative right ventricular (RV) function and high vasoactive-inotropic score (VIS) after cardiac surgery. Design: Prospective observational study. Setting: A single medical center setting. Participants: One hundred three patients undergoing elective cardiac surgery. Interventions: None. Measurements and Main Results: Consecutive patients referred for cardiac surgery were enrolled prospectively. Comprehensive transesophageal echocardiography was performed before sternal incision. Specific RV indices, encompassing RV fractional area change, tricuspid annular plane systolic excursion, and RV global longitudinal strain (RVGLS), were measured offline. High VIS was defined as a maximum VIS of 20 in 24 hours postoperatively. Postoperative adverse events were recorded. One hundred three patients (mean age 61.2 § 11.0, 72 men) were included in this study, where 17 patients (16.5%) achieved high VIS with a mean maximum VIS of 39 in 24 hours postoperatively. Patients with high VIS encountered increased occurrence of extracorporeal membrane oxygenation placement, acute kidney injury, and mortality. Risk factors for high VIS included operation type, cardiopulmonary bypass duration, left atrium size, and pre-incisional RV indices. After adjustment for age, left ventricular ejection fraction, and the covariates, only RVGLS (odds ratio 1.19, p = 0.011) showed an independent association with high VIS. The optimal cutoff of RVGLS was ¡16.7% (sensitivity of 88.2%, specificity of 75.6%). Conclusion: Preoperative RV dysfunction is an independent risk factor for postoperative high VIS. Pre-incisional RVGLS is a reliable tool to predict high VIS after cardiac surgery. Patients with high VIS had increased adverse events postoperatively. Ó 2018 Elsevier Inc. All rights reserved. Key Words: cardiac surgery; right ventricular function; transesophageal echocardiography; vasoactive-inotropic score
STUDIES HAVE REPORTED the association between perioperative right ventricular dysfunction (RVD) and increased postoperative morbidity and mortality in patients 1 Address reprint requests to Shao-Wei Chen, MD, Division of Thoracic and Cardiovascular Surgery, Chang Gung Memorial Hospital, Linkou Medical Center, Taoyuan City, No. 5 Fuxing St, Guishan District, Taoyuan City 33305, Taiwan. E-mail address: [email protected] (S.-W. Chen).
https://doi.org/10.1053/j.jvca.2018.07.048 1053-0770/Ó 2018 Elsevier Inc. All rights reserved.
undergoing cardiac surgery.1-3 In particular, RVD consistently has been demonstrated to correlate with postoperative hemodynamic instability and extensive vasoactive and inotropic agent use. Maslow et al. reported that in patients undergoing coronary artery bypass grafting (CABG), prebypass RVD was associated with needing 2 or more inotropes in the postbypass period.1 Ternacle et al. showed that patients with preoperative RVD, diagnosed through transthoracic echocardiography (TTE) rather than intraoperative transesophageal echocardiography
ARTICLE IN PRESS 2
P.-C. Ting et al. / Journal of Cardiothoracic and Vascular Anesthesia 00 (2018) 18
(TEE), were at risk of prolonged inotropic support after CABG and valvular surgery.2 The vasoactive-inotropic score (VIS) was created as a weighted sum of various vasoactive medications typically used in contemporary clinical practice. It had been tested whether a mean or a maximum VIS at specific time frames (24 and 48 hours) after postoperative admission to the intensive care unit (ICU) had a strong association with morbidity and mortality in pediatric cohorts referred for cardiac surgery.4,5 After validating the optimal cutoff value in a multi-institutional data set, Gaies et al. found thatD17X X VIS of 20, defined as a maximum VIS in the first 24 hours of admission to the ICU, was associated strongly and significantly with a composite outcome (odds ratio 6.5, 95% confidence interval [CI] 2.9-14.6) of in-hospital death, need for mechanical life support, or renal replacement therapy.5 To contribute additional prognostic value to increased postoperative vasoactive agent requirements and quantify the clinical behavior of patients, the authors examined the relationship between intraoperative RV function and postoperative high VIS in adult cardiac surgery. Several echocardiographic methods have been proposed for evaluating RV function6; however, some methods have not been validated for use with TEE given the placement of the probe in the esophagus and the angle of acquisition in TEE.7 Studies have reported that 2-dimensional (2D) speckle-tracking echocardiography is a reliable method for RV assessment in perioperative settings.8,9 Therefore, in the present study, the authors hypothesize that intraoperative RV indices assessed through TEE are associated with postoperative high VIS in patients referred for elective cardiac surgery. Methods Study Population This prospective study was approved by the institutional review board of Chang Gung Memorial Hospital (IRB No. 101-1774B); prior informed written consent was obtained from all patients. The authors prospectively enrolled patients older than 20 years who underwent elective cardiac surgery between July 2012 and July 2013. Exclusion criteria included contraindications to TEE and a medical history of non-sinus rhythm. The authors also excluded patients who had received preoperative vasoactive agents or mechanical circulatory support. Comprehensive TEE and concurrent hemodynamic recording were performed in the operating room under general anesthesia before sternal incision. Echocardiographic Measurements All echocardiographic studies were conducted by an experienced echocardiographer who used a commercially available echocardiography system equipped with a 5.0-MHz transducer (Vivid 7; GE Healthcare, Milwaukee, WI). In the protocol,10 the authors acquired TEE data sets from 3 to 5 consecutive cardiac cycles, with temporary interruption of ventilator support at end-expiration. During acquisition, special care was
taken to ensure an adequate sector for the entire RV lateral wall and the interventricular septum throughout the cycle; moreover, a frame rate of >60 Hz was maintained to facilitate optimal tracking of the myocardium. The RV was visualized most efficiently from a slightly lower esophageal 4-chamber view with appropriate adjustment of the probe through retroflex and clockwise rotation. In addition, the authors changed the multiplane angle from 0˚ to 20˚ to obtain the widest RV chamber. The obtained images were transferred through a network for offline analysis by using dedicated software (EchoPAC PC version 110.0.0, GE Healthcare). The echocardiographic assessment of RV performance included RV fractional area change (RVFAC), tricuspid annular plane systolic excursion (TAPSE), and RV global longitudinal strain (RVGLS). From the modified RV-focused 4-chamber view, RVFAC was obtained by tracing the endocardial border of the RV and was calculated as (RV end-diastolic area ¡ RV end-systolic area)/RV end-diastolic area £ 100%. TAPSE was measured as the tricuspid annulus excursion in millimeters from end-diastole to end-systole by using a floating M-mode cursor (a technique of anatomical M-mode echocardiography) placed through the lateral tricuspid annulus and aligned with its motion.11 RVGLS was computed through 2D speckle-tracking analysis: a region of interest was generated by manually identifying the landmarks of the tricuspid annulus and the RV apex, encompassing 6 segments of the RV lateral wall and septum. The software detected the onset of QRS in the simultaneous electrocardiographic recording to define the zero-strain point and plotted both segmental and global strain curves. The global peak systolic strain was recorded (Fig 1).8 In addition, left ventricular ejection fraction (LVEF) was assessed through the biplane Simpson method. All measurements were performed offline and agreed upon by 2 investigators blinded to patient outcomes. Any disagreements were resolved by a third investigator. VasoactiveInotropic Scoring When the patient arrived in the ICU, medications were adjusted under the consensus of the ICU medical team. Inotropes (dobutamine, epinephrine, or milrinone) usually were used to maintain adequate cardiac output, and vasopressors (dopamine, norepinephrine, or vasopressin) were indicated for hypotension.12 The doses of the inotropic and vasoactive medications were recorded on an hourly basis for the first 24 hours in the ICU. As defined in the literature,4,5 VIS was calculated as (dopamine dose + dobutamine dose) (mg/kg/min) + 10 £ milrinone dose (mg/kg/min) + 100 £ (epinephrine dose + norepinephrine dose) (mg/kg/min) + 10000 £ vasopressin dose (unit/ kg/min). Patients with maximum VIS 20 within the first 24 hours in the ICU were recognized as having high VIS,5 and patients with maximum VIS <20 in the first 24 hours were defined as having low VIS. Patient Partition and Postoperative Adverse Events Patient demographics and comorbidities were recorded prospectively. Preoperative TTE was performed at admission for
ARTICLE IN PRESS P.-C. Ting et al. / Journal of Cardiothoracic and Vascular Anesthesia 00 (2018) 18
3
Fig 1. Speckle-tracking strain of RV. The range of interest was traced manually along the RV endocardial border. Numbers shown on the picture represent segmental and global peak systolic strain values. RVGLS is measured as ¡27.0%. RV, right ventricular; RVGLS, RV global longitudinal strain.
cardiac surgery, and left atrium (LA) size measured in the parasternal long-axis view perpendicular to the axis of aortic root as an anteroposterior diameter was collected prospectively. Postoperative outcomes and complications, namely duration
of vasoactive agent use, time to first extubation, length of ICU and hospital stay, need for mechanical life support, occurrence of acute kidney injury (AKI), and mortality, also were recorded. Based on the Kidney Disease: Improving Global
ARTICLE IN PRESS 4
P.-C. Ting et al. / Journal of Cardiothoracic and Vascular Anesthesia 00 (2018) 18
Outcomes Clinical Practice Guideline,13 AKI was defined as a more than 1.5-fold increase in the serum creatinine compared with the baseline value, which was presumed to have occurred within 7 days of cardiac surgery. Statistical Analysis The parameter of primary interest in this study was RVGLS. According to a previous study,2 the RVGLS was ¡22% in patients with no inotropic support and ¡19% in patients with inotropic support for >48 hours. With an assumed standard deviation of 5%, alpha level of 5%, and power of 80%, a sample size of at least 90 patients was required. Normality of continuous data was tested using the Kolmogorov-Smirnov test. Continuous variables of the patients with low and high VISs were compared using the Student t test or Mann-Whitney U test, as appropriate. The proportions of categorical variables between the 2 study groups were compared using the Fisher exact test. To investigate the association of RV indices with the occurrence of high VIS, the authors developed 3 multivariable logistic regression models, into each of which an individual RV index was introduced and considered an independent variable. In each multivariable model, age and LVEF were considered covariates and were adjusted for other significant (p < 0.05) variables determined through prior univariate analyses. The performance of RV indices for diagnosing the occurrence of high VIS was evaluated through receiver operating characteristic curve analysis. Interobserver reproducibility for RVGLS was evaluated by an independent observer in 18 random patients. For intraobserver variability, the analysis was repeated 4 months after the first measurement in the same 18 patients. Reproducibility was expressed using the coefficient of variation and interclass correlation coefficients.
Postoperative Adverse Events in Patients With High VIS Compared with those with low VIS, patients with high VIS were subjected to the need for extracorporeal membrane oxygenation (ECMO) (p = 0.026) and AKI (p = 0.005) more frequently, and they encountered a significantly higher mortality rate (23.5% v 3.5%, p = 0.013). The detailed data are provided in Table 2. Association of RV Indices With High VIS After adjustment for age, LVEF, and the 3 characteristics (operation type, CPB duration, LA size), each 1% increase in pre-incisional RVGLS was associated significantly with 1.19fold odds of being high VIS (95% CI of odds ratio, 1.04-1.36, p = 0.011). By contrast, RVFAC and TAPSE showed no significant associations with high VIS (Table 3). As shown in Fig 2, the areas under the curves of these RV indices were greater than 0.7, indicating acceptable discrimination ability. Regarding RVGLS, the optimal cutoff value was ¡16.7%, which yielded a sensitivity of 88.2% and a specificity of 75.6% (positive likelihood ratio [LR+] 3.61). The optimal cutoff value of RVFAC was 30.8% (sensitivity 58.8%, specificity 76.7%), and that of TAPSE was 12.3 mm (sensitivity 88.2%, specificity 60.5%). Figure 3 shows that the addition of RVGLS markedly improved the predictive utility of the model that contained aforementioned covariates. The interobserver and intraobserver variabilities, expressed as coefficients of variation, for RVGLS were 7.1% and 6.0%, respectively. The interobserver and intraobserver interclass correlation coefficients for RVGLS were 0.93 (95% CI 0.82-0.98) and 0.95 (95% CI 0.87-0.98), respectively. Discussion
Results Patient Profile The study included 103 patients (mean age 61.2 § 11.0 years, 72 men), of whom 45 (43.7%) underwent CABG alone, 46 (44.7%) underwent isolated valvular surgery, and 12 (11.7%) underwent combined CABG and valvular surgery. Seventeen patients (16.5%) reached high VIS, with a mean maximum VIS of 39 (range 21 to 91) within 24 hours of postoperative admission to the ICU. Compared with patients with low VIS, those with high VIS had significantly larger LA size (p = 0.009) and poorer preincisional RV indices, namely RVFAC (p = 0.005), TAPSE (p = 0.001), and RVGLS (p < 0.001). Furthermore, patients with high VIS were more likely to have undergone isolated valvular surgery (p = 0.031) and to have experienced significantly longer cardiopulmonary bypass (CPB) duration (p < 0.001) than patients with low VIS. No differences were observed in age, sex, LVEF, medical history, and intraoperative hemodynamic measurements, as shown in Table 1.
In this prospective study, the authors used high VIS as a primary outcome in adult patients referred for elective cardiac surgery. Patients who sustained high VIS had impaired preincisional RVFAC, TAPSE, and RVGLS. After adjustments for covariates, RVGLS alone was associated with high VIS. In addition, high VIS was associated with postoperative adverse events such as AKI, the need for ECMO, and mortality. Several risk factors have been identified for predicting postoperative inotropic support.14-16 Advanced age and reduced preoperative LVEF were identified as significant predictors of postoperative inotropic support in patients undergoing CABG or cardiac valvular surgery.15,16 However, the risk factors recognized in this study were inconsistent with those in the aforementioned studies, possibly because of the differences in the primary outcome. Nevertheless, Imai et al. reported that patients who had undergone an extended CPB duration were more likely to require >12 hours of postoperative inotropic support irrespective of their preoperative LVEF,14 and Zaroff et al. demonstrated that not all patients with low preoperative LVEF required postoperative inotropic support.17 LA
ARTICLE IN PRESS P.-C. Ting et al. / Journal of Cardiothoracic and Vascular Anesthesia 00 (2018) 18
5
Table 1 Demographic Data and Clinical Characteristics of Patients With Low and High VISs All n = 103 Age (y) Sex, n (%) Female Male Body surface area (m2) Body mass index (kg/m2) ASA, n (%) III IV NYHA, n (%) I/ II III/ IV Comorbidity, n (%) Hypertension Diabetes mellitus Chronic kidney disease Chronic lung disease Cirrhosis Cerebral vascular attack Coronary artery disease Previous heart surgery Laboratory data Hemoglobin (g/dL) Creatinine (mg/dL) Operation type, n (%) CABG alone Valvular surgery alone Combined surgery Details of valvular surgery (n = 58) Aortic valve surgery (n = 22) Aortic stenosis Aortic regurgitation Mitral valve surgery (n = 40) Mitral stenosis, rheumatic Mitral regurgitation, degenerative Mitral regurgitation, ischemic Tricuspid valve surgery (n = 10) Preoperative echo Left atrium (mm) Preincisional echo LVEF (%) RVFAC (%) TAPSE (mm) RVGLS (%) Concurrent hemodynamics Cardiac index (L/min-m2) Mean arterial pressure (mmHg) MPAP (mmHg) PVRI (mmHg-min-m2/L) Bypass time (min)
Low VIS n = 86
High VIS n = 17
61.2 § 11.0
61.6 § 10.9
59.2 § 11.9
31 (30.1) 72 (69.9) 1.7 § 0.2 24.0 § 3.6
25 (29.1) 61 (70.9) 1.7 § 0.2 24.3 § 3.5
6 (35.3) 11 (64.7) 1.6 § 0.2 22.5 § 4.1
94 (91.3) 9 (8.7)
79 (91.9) 7 (8.1)
15 (88.2) 2 (11.8)
58 (56.3) 45 (43.7)
51 (59.3) 35 (40.7)
7 (41.2) 10 (58.8)
72 (69.9) 48 (46.6) 22 (21.4) 9 (8.7) 3 (2.9) 12 (11.7) 32 (31.1) 11 (10.7)
60 (69.8) 43 (50) 20 (23.3) 9 (10.5) 2 (2.3) 9 (10.5) 29 (33.7) 8 (9.3)
12 (70.6) 5 (29.4) 2 (11.8) 0 (0) 1 (5.9) 3 (17.6) 3 (17.6) 3 (17.6)
1.000 0.183 0.517 0.349 0.421 0.413 0.257 0.385
12.1 § 2.3 1.7 § 2.4
12.3 § 2.3 1.8 § 2.5
11.4 § 2.3 1.5 § 1.6
45 (43.7) 46 (44.7) 12 (11.7)
42 (48.8) 34 (39.5) 10 (11.6)
3 (17.6) 12 (70.6) 2 (11.8)
0.156 0.695 0.035 0.030 0.031 1.000
22 13 9 40 9 23 8 10
15 12 3 30 5 18 7 5
7 1 6 10 4 5 1 5
45.6 § 10.9
43.8 § 8.9
54.6 § 14.9
0.009
45.9 § 11.1 36.7 § 9.1 12.9 § 3.9 ¡18.8 § 6.1
46.0 § 11.3 37.8 § 8.7 13.5 § 3.8 ¡19.8 § 5.3
45.2 § 10.1 31.1 § 8.8 10.1 § 3.3 ¡13.3 § 6.9
0.796 0.005 0.001 <0.001
2.0 § 0.8 82.3 § 16.8 24.7 § 11.1 330.1 § 215.2 143.1 § 81.5
2.0 § 0.8 82.6 § 16.4 23.7 § 9.8 319.5 § 205.6 130.4 § 72.1
2.0 § 0.9 80.3 § 19.2 29.6 § 15.5 383.4 § 258.9 206.9 § 97.8
0.996 0.602 0.151 0.265 <0.001
p Value 0.429 0.579
0.396 0.063 0.641
0.190
NOTE. Continuous data are presented as mean § standard deviation. Abbreviations: ASA, American Society of Anesthesiologists physical status classification; CABG, coronary artery bypass graft; LVEF, left ventricular ejection fraction; MPAP, mean pulmonary arterial pressure; NYHA, New York Heart Association; PVRI, pulmonary vascular resistance index; RVFAC, right ventricular fractional area change; RVGLS, right ventricular global longitudinal strain; TAPSE, tricuspid annular plane systolic excursion.
enlargement has been reported to be an independent predictor for adverse cardiovascular outcomes18,19 and is a marker of both severity and chronicity of diastolic dysfunction, which may correlate with postoperative hemodynamic instability and heavy vasoactive agent use.20 However, the linear expression
of LA size did not show an association with high VIS in the multivariable analysis. In this study, RV function measured using RVGLS was an independent predictor of high VIS after adjustment for risk factors, a finding comparable with that of Ternacle et al.2 Furthermore, these results showed that
ARTICLE IN PRESS 6
P.-C. Ting et al. / Journal of Cardiothoracic and Vascular Anesthesia 00 (2018) 18
Table 2 Postoperative Adverse Events in Patients With Low and High VISs
Variable Vasoactive agent support (h) Intubation time (h) ECMO, n (%) Acute kidney injury, n (%) ICU stay (day) Hospital stay (d) Mortality, n (%)
All n = 103
Low VIS n = 86
High VIS n = 17
p Value
9 (0-57) 15 (5-23) 2 (1.9) 35 (34) 3 (2-5) 16 (10-25) 7 (6.8)
1 (0-40) 14 (5-22) 0 (0) 24 (27.9) 3 (2-5) 16 (10-24) 3 (3.5)
44 (27-97) 20 (12-87) 2 (11.8) 11 (64.7) 4 (3-8) 19 (11-30) 4 (23.5)
<0.001 0.018 0.026 0.005 0.065 0.321 0.013
NOTE. Continuous data are presented as median (25th to 75th percentiles). Abbreviations: ECMO, extracorporeal membrane oxygenation; ICU, intensive care unit.
considering RVGLS in addition to the previously identified risk factors (ie, age, LVEF, operation type, LA size, and CPB duration) in the prediction of postoperative high VIS significantly improved the predictive accuracy. Although many of the RV parameters were considered loaddependent, research of different loading conditions (normal or excessive) has revealed mixed results.21,22 Kjaergaard et al.21 reported that RV strain and TAPSE were unchanged after an increase in the preload and afterload, whereas da Costa Junior et al.,22 studying a group of patients with pulmonary hypertension, reported that all RV indices, including RVFAC, TAPSE, and RVGLS, decreased when RV was confronted by abnormally high pulmonary resistance. Under normal circumstances, RV is easily distensible; hence, the curvature of RV changes little because of the restraint of the pericardium, and the pressure does not rise significantly, rendering the RV indices relatively stable.23 In this study, on average, the concurrent cardiac index and pulmonary vascular resistance index was within the normal range and did not differ significantly between the low- and high-VIS patients; thus, the capacity of
RV indices to predict high VIS is independent of the loading conditions. The superiority of RVGLS over other RV indices in representing RV function can be explained by a global assessment of RV longitudinal performance, which constitutes the major component of RV systolic function,9,24 and by a characteristic that is not greatly affected by cardiac translation.25 By contrast, RVFAC is a simple method to delineate the RV borders that expresses both radial and longitudinal RV function; however, the longitudinal component is attenuated by the radial component. TAPSE focuses on the regional longitudinal movement of the RV lateral wall and is deemed problematic when measured using conventional M-mode through TEE because of its dependence on transducer orientation.7 The authors adopted a technique of anatomical M-mode in which a floating cursor is used to align freely the direction of the tricuspid motion during post-processing of the stored 2D digital images, thus eliminating the aforementioned TEE-related concern. Tousignant et al. reported that TAPSE measured using TEE through antomical M-mode (r = 0.48, p < 0.001) or the
Table 3 Associations Between Preincisional RV Indices and Risk of High VIS Characteristics Univariate logistic regression analysis Age, per 10 y Operation type CABG alone Valve alone Mixed Bypass time, per 10 min Left atrium, mm LVEF, per 5% RVFAC, % TAPSE, mm RVGLS, % Multivariable logistic models* RVFAC, % TAPSE, mm RVGLS, %
OR
95% CI
p Value
0.83
0.52 to 1.32
0.426 0.065
Ref. 4.94 2.80 1.13 1.08 0.97 0.91 0.74 1.22
1.29 to 18.94 0.41 to 19.05 1.05 to 1.23 1.03 to 1.14 0.76 to 1.23 0.85 to 0.98 0.61 to 0.90 1.09 to 1.35
0.020 0.293 0.001 0.002 0.794 0.008 0.002 <0.001
0.95 0.83 1.19
0.87 to 1.03 0.67 to 1.03 1.04 to 1.36
0.180 0.090 0.011
Abbreviations: CABG, coronary artery bypass graft; CI, confidence interval; LVEF, left ventricular ejection fraction; OR, odds ratio; RVFAC, right ventricular fractional area change; RVGLS, right ventricular global longitudinal strain; TAPSE, tricuspid annular plane systolic excursion. * Adjusted for age, operation type, bypass time, left atrium size, and left ventricular ejection fraction; each RV index was modeled separately.
ARTICLE IN PRESS P.-C. Ting et al. / Journal of Cardiothoracic and Vascular Anesthesia 00 (2018) 18
Fig 2. Receiver operating characteristic curve analysis. Performance of preincisional echocardiographic parameters in discriminating the risk of high VIS. VIS, vasoactive-inotropic score.
speckle-tracking method (r = 0.44, p = 0.001) correlated modestly with stroke volume, which may be helpful in predicting a risk or an outcome.11 However, TAPSE measured by either
7
method was not sensitive enough to track output changes and thus was suggested as an incomplete assessment of RV function.7,11 In the univariate analysis, the authors did not apply specific cutoff values to designate RVD, because the values were derived from previous TTE studies6; nevertheless, continuous data of preincisional RVFAC, TAPSE, and RVGLS measured through TEE were decreased significantly in patients with high VIS. Moreover, a cutoff value was recommended to predict a specific outcome rather than an abnormal value that depended on the mean obtained from a group of healthy people.6 Thus, the receiver operating characteristic analysis in this study determined the optimal cutoff value of each RV index for high VIS. Although the areas under the curves of the RV indices did not differ significantly, the LR+ of RV indices helped discriminate their responsiveness. When RVGLS of ¡16.7% was used as the cutoff point, LR+ of 3.61 increased the prevalence of high VIS (16.5%) to the “post-test” prevalence of 41.5%, thus showing moderate responsiveness. When RVFAC of 30.8% and TAPSE of 12.8 mm were used as the cutoff values, the post-test prevalence of high VIS was 30% to 33%, showing less responsiveness than RVGLS. Based on previous studies,4,5 this study solicited high VIS as an effective primary outcome and found that patients with high VIS experienced more severe postoperative complications than did patients with low VIS. Notably, the finding that high VIS is associated with postoperative adverse events, including prolonged time to first extubation and increased occurrence of need for ECMO, AKI, and mortality, was consistent with the findings of Gaies et al.5 Although the length of ICU and hospital stay did not differ between study groups, the results easily may have been affected by domestic insurance regulations and other issues during the treatment course, as was the case in Gaies et al.4 Limitations
Fig 3. Incremental values of RV indices for predicting postoperative high VIS. The addition of RVGLS to age, left ventricular ejection fraction (LVEF), operation type (OP type), left atrium size (LA), and bypass time significantly improved the predictive value of the proposed model. RV, right ventricular; RVGLS, RV global longitudinal strain; VIS, vasoactive-inotropic score.
The study population was not homogenous and underwent a range of cardiac surgeries, including CABG, simple valvular surgery, and combined surgery. Considering the initial finding that patients referred for valvular surgery were subjected to postoperative high VIS, the authors made adjustments for the operation type in the regression models, and preincisional RVGLS remained an independent predictor of high VIS. For speckle-tracking measurement of RV, 2 methods currently are applied.6 RV free wall strain (RVFWS) averages longitudinal strains on the 3 segments of the lateral wall, and RVGLS encompasses the 6 segments of the whole RV, including the interventricular septum. The authors used RVGLS rather than RVFWS for RV measurement. Although RVFWS is considered relatively exclusive to the RV native function, 40% of the RV systolic function pertains to the septum and should not be ignored.24 A previous study demonstrated that impaired baseline RVGLS indicated poor outcomes in patients undergoing cardiac resynchronization therapy.26 This study showed that preincisional RVGLS is associated independently with postoperative high VIS and that RVGLS has an important
ARTICLE IN PRESS 8
P.-C. Ting et al. / Journal of Cardiothoracic and Vascular Anesthesia 00 (2018) 18
role in the evaluation of prognosis in patients referred for cardiac surgery. Because evidence offering guidance on the choice of vasoactive agents is limited, VIS could be affected by personal preference or misinterpretation of hemodynamic data. To eliminate personnel bias, postoperative care in the ICU is provided by a fixed medical team in the authors’ institute to ensure consensus on the choice of vasoactive and inotropic agents and to adjust the dosage in these patients. Despite this prudence, variation in practice was tolerated in a multicenter study of VIS,5 and high VIS retained its association with poor outcomes. Future studies to guide the most efficient strategy for addressing specific clinical situations should be conducted in collaboration with VIS.
8
9
10
11 12
13
Conclusion 14
This study concludes that preoperative depressed RV function is an independent risk factor for occurrence of postoperative high VIS. In the assessment of RV function, pre-incisional RVGLS measured using TEE is a reliable tool to predict postoperative high VIS in patients referred for elective cardiac surgery. In addition, high VIS acts as a practical outcome measurement, D18X Xpredicting early postoperative complications in the population of adult cardiac surgery. The authors thank Alfred Hsing-Fen Lin for his assistance in statistical analysis. References 1 Maslow AD, Regan MM, Panzica P, et al. 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. 2 Ternacle J, Berry M, Cognet T, et al. Prognostic value of right ventricular two-dimensional global strain in patients referred for cardiac surgery. J Am Soc Echocardiogr 2013;26:721–6. 3 Haddad F, Denault AY, Couture P, 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. 4 Gaies MG, Gurney JG, Yen AH, et al. Vasoactive-inotropic score as a predictor of morbidity and mortality in infants after cardiopulmonary bypass. Pediatr Crit Care Med 2010;11:234–8. 5 Gaies MG, Jeffries HE, Niebler RA, et al. Vasoactive-inotropic score is associated with outcome after infant cardiac surgery: An analysis from the Pediatric Cardiac Critical Care Consortium and Virtual PICU System Registries. Pediatr Crit Care Med 2014;15:529–37. 6 Lang RM, Badano LP, Mor-Avi V, et al. Recommendations for cardiac chamber quantification by echocardiography in adults: An update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr 2015;28:1–39;e14. 7 Markin NW, Chamsi-Pasha M, Luo J, et al. Transesophageal speckletracking echocardiography improves right ventricular systolic function
15
16
17
18
19
20
21
22
23 24
25
26
assessment in the perioperative setting. J Am Soc Echocardiogr 2017;30:180–8. Tousignant C, Desmet M, Bowry R, et al. Speckle tracking for the intraoperative assessment of right ventricular function: A feasibility study. J Cardiothorac Vasc Anesth 2010;24:275–9. Silverton N, Meineri M. Speckle tracking strain of the right ventricle: An emerging tool for intraoperative echocardiography. Anesth Analg 2017;125:1475–8. Ting PC, Chou AH, Wu VCC, et al. Relationship between right ventricular function and atrial fibrillation after cardiac surgery. J Cardiothorac Vasc Anesth 2017;31:1663–71. Tousignant C, Kim H, Papa F, et al. Evaluation of TAPSE as a measure of right ventricular output. Can J Anaesth 2012;59:376–83. Stephens RS, Whitman GJ. Postoperative critical care of the adult cardiac surgical patient. Part I: Routine postoperative care. Crit Care Med 2015;43:1477–97. Palevsky PM, Liu KD, Brophy PD, et al. KDOQI US commentary on the 2012 KDIGO clinical practice guideline for acute kidney injury. Am J Kidney Dis 2013;61:649–72. Imai H, Kurokawa S, Taneoka M, et al. Tissue Doppler imaging is useful for predicting the need for inotropic support after cardiac surgery. J Anesth 2011;25:805–11. Butterworth JF 4th, C Legault, Royster RL, et al. Factors that predict the use of positive inotropic drug support after cardiac valve surgery. Anesth Analg 1998;86:461–7. Royster RL, Butterworth JF 4th, Prough DS, et al. Preoperative and intraoperative predictors of inotropic support and long-term outcome in patients having coronary artery bypass grafting. Anesth Analg 1991;72:729–36. Zaroff J, Aronson S, Lee BK, et al. The relationship between immediate outcome after cardiac surgery, homogeneous cardioplegia delivery, and ejection fraction. Chest 1994;106:38–45. Kizer JR, Bella JN, Palmieri V, et al. Left atrial diameter as an independent predictor of first clinical cardiovascular events in middle-aged and elderly adults: The Strong Heart Study (SHS). Am Heart J 2006;151:412–8. Bouzas-Mosquera A, Broullon FJ, Alvarez-Garcia N, et al. Left atrial size and risk for all-cause mortality and ischemic stroke. CMAJ 2011;183: E657–64. Bernard F, Denault A, Babin D, et al. Diastolic dysfunction is predictive of difficult weaning from cardiopulmonary bypass. Anesth Analg 2001;92:291–8. Kjaergaard J, Snyder EM, Hassager C, et al. Impact of preload and afterload on global and regional right ventricular function and pressure: A quantitative echocardiography study. J Am Soc Echocardiogr 2006;19:515–21. da Costa Junior AA, Ota-Arakaki JS, Ramos RP, et al. Diagnostic and prognostic value of right ventricular strain in patients with pulmonary arterial hypertension and relatively preserved functional capacity studied with echocardiography and magnetic resonance. Int J Cardiovasc Imaging 2017;33:39–46. Greyson CR. Right heart failure in the intensive care unit. Curr Opin Crit Care 2012;18:424–31. Haddad F, Hunt SA, Rosenthal DN, et al. Right ventricular function in cardiovascular disease, Part I: Anatomy, physiology, aging, and functional assessment of the right ventricle. Circulation 2008;117:1436–48. Giusca S, Dambrauskaite V, Scheurwegs C, et al. Deformation imaging describes right ventricular function better than longitudinal displacement of the tricuspid ring. Heart 2010;96:281–8. Nagy VK, Szeplaki G, Apor A, et al. Role of right ventricular global longitudinal strain in predicting early and long-term mortality in cardiac resynchronization therapy patients. PLoS One 2015;10:e0143907.
Contents lists available at ScienceDirect
Journal of Cardiothoracic and Vascular Anesthesia journal homepage: www.jcvaonline.com
Original Article
Preoperative Right Ventricular Dysfunction Indicates High Vasoactive Support Needed After Cardiac Surgery 1XD XPei-Chi Ting, D2X XMD, MSc*, D3X XVictor C-C. Wu, D4X XMDy, D5X XChia-Chih Liao, D6X XMD, PhD*, D7X XAn-Hsun Chou, D8X XMD, PhD*, D9X XFeng-Chun Tsai, D10X XMDz, D1X XPyng-Jing Lin, D12X XMDz, ,1 D13X XChun-Yu Chen, D14X XMD, PhD*, D15X XShao-Wei Chen, D16X XMDz *
Department of Anesthesiology, Chang Gung Memorial Hospital, Taoyuan, Taiwan y Department of Cardiology, Chang Gung Memorial Hospital, Taoyuan, Taiwan z Division of Thoracic and Cardiovascular Surgery, Chang Gung Memorial Hospital, Taoyuan, Taiwan
Objective: The aim of this study was to explore the relationship between preoperative right ventricular (RV) function and high vasoactive-inotropic score (VIS) after cardiac surgery. Design: Prospective observational study. Setting: A single medical center setting. Participants: One hundred three patients undergoing elective cardiac surgery. Interventions: None. Measurements and Main Results: Consecutive patients referred for cardiac surgery were enrolled prospectively. Comprehensive transesophageal echocardiography was performed before sternal incision. Specific RV indices, encompassing RV fractional area change, tricuspid annular plane systolic excursion, and RV global longitudinal strain (RVGLS), were measured offline. High VIS was defined as a maximum VIS of 20 in 24 hours postoperatively. Postoperative adverse events were recorded. One hundred three patients (mean age 61.2 § 11.0, 72 men) were included in this study, where 17 patients (16.5%) achieved high VIS with a mean maximum VIS of 39 in 24 hours postoperatively. Patients with high VIS encountered increased occurrence of extracorporeal membrane oxygenation placement, acute kidney injury, and mortality. Risk factors for high VIS included operation type, cardiopulmonary bypass duration, left atrium size, and pre-incisional RV indices. After adjustment for age, left ventricular ejection fraction, and the covariates, only RVGLS (odds ratio 1.19, p = 0.011) showed an independent association with high VIS. The optimal cutoff of RVGLS was ¡16.7% (sensitivity of 88.2%, specificity of 75.6%). Conclusion: Preoperative RV dysfunction is an independent risk factor for postoperative high VIS. Pre-incisional RVGLS is a reliable tool to predict high VIS after cardiac surgery. Patients with high VIS had increased adverse events postoperatively. Ó 2018 Elsevier Inc. All rights reserved. Key Words: cardiac surgery; right ventricular function; transesophageal echocardiography; vasoactive-inotropic score
STUDIES HAVE REPORTED the association between perioperative right ventricular dysfunction (RVD) and increased postoperative morbidity and mortality in patients 1 Address reprint requests to Shao-Wei Chen, MD, Division of Thoracic and Cardiovascular Surgery, Chang Gung Memorial Hospital, Linkou Medical Center, Taoyuan City, No. 5 Fuxing St, Guishan District, Taoyuan City 33305, Taiwan. E-mail address: [email protected] (S.-W. Chen).
https://doi.org/10.1053/j.jvca.2018.07.048 1053-0770/Ó 2018 Elsevier Inc. All rights reserved.
undergoing cardiac surgery.1-3 In particular, RVD consistently has been demonstrated to correlate with postoperative hemodynamic instability and extensive vasoactive and inotropic agent use. Maslow et al. reported that in patients undergoing coronary artery bypass grafting (CABG), prebypass RVD was associated with needing 2 or more inotropes in the postbypass period.1 Ternacle et al. showed that patients with preoperative RVD, diagnosed through transthoracic echocardiography (TTE) rather than intraoperative transesophageal echocardiography
ARTICLE IN PRESS 2
P.-C. Ting et al. / Journal of Cardiothoracic and Vascular Anesthesia 00 (2018) 18
(TEE), were at risk of prolonged inotropic support after CABG and valvular surgery.2 The vasoactive-inotropic score (VIS) was created as a weighted sum of various vasoactive medications typically used in contemporary clinical practice. It had been tested whether a mean or a maximum VIS at specific time frames (24 and 48 hours) after postoperative admission to the intensive care unit (ICU) had a strong association with morbidity and mortality in pediatric cohorts referred for cardiac surgery.4,5 After validating the optimal cutoff value in a multi-institutional data set, Gaies et al. found thatD17X X VIS of 20, defined as a maximum VIS in the first 24 hours of admission to the ICU, was associated strongly and significantly with a composite outcome (odds ratio 6.5, 95% confidence interval [CI] 2.9-14.6) of in-hospital death, need for mechanical life support, or renal replacement therapy.5 To contribute additional prognostic value to increased postoperative vasoactive agent requirements and quantify the clinical behavior of patients, the authors examined the relationship between intraoperative RV function and postoperative high VIS in adult cardiac surgery. Several echocardiographic methods have been proposed for evaluating RV function6; however, some methods have not been validated for use with TEE given the placement of the probe in the esophagus and the angle of acquisition in TEE.7 Studies have reported that 2-dimensional (2D) speckle-tracking echocardiography is a reliable method for RV assessment in perioperative settings.8,9 Therefore, in the present study, the authors hypothesize that intraoperative RV indices assessed through TEE are associated with postoperative high VIS in patients referred for elective cardiac surgery. Methods Study Population This prospective study was approved by the institutional review board of Chang Gung Memorial Hospital (IRB No. 101-1774B); prior informed written consent was obtained from all patients. The authors prospectively enrolled patients older than 20 years who underwent elective cardiac surgery between July 2012 and July 2013. Exclusion criteria included contraindications to TEE and a medical history of non-sinus rhythm. The authors also excluded patients who had received preoperative vasoactive agents or mechanical circulatory support. Comprehensive TEE and concurrent hemodynamic recording were performed in the operating room under general anesthesia before sternal incision. Echocardiographic Measurements All echocardiographic studies were conducted by an experienced echocardiographer who used a commercially available echocardiography system equipped with a 5.0-MHz transducer (Vivid 7; GE Healthcare, Milwaukee, WI). In the protocol,10 the authors acquired TEE data sets from 3 to 5 consecutive cardiac cycles, with temporary interruption of ventilator support at end-expiration. During acquisition, special care was
taken to ensure an adequate sector for the entire RV lateral wall and the interventricular septum throughout the cycle; moreover, a frame rate of >60 Hz was maintained to facilitate optimal tracking of the myocardium. The RV was visualized most efficiently from a slightly lower esophageal 4-chamber view with appropriate adjustment of the probe through retroflex and clockwise rotation. In addition, the authors changed the multiplane angle from 0˚ to 20˚ to obtain the widest RV chamber. The obtained images were transferred through a network for offline analysis by using dedicated software (EchoPAC PC version 110.0.0, GE Healthcare). The echocardiographic assessment of RV performance included RV fractional area change (RVFAC), tricuspid annular plane systolic excursion (TAPSE), and RV global longitudinal strain (RVGLS). From the modified RV-focused 4-chamber view, RVFAC was obtained by tracing the endocardial border of the RV and was calculated as (RV end-diastolic area ¡ RV end-systolic area)/RV end-diastolic area £ 100%. TAPSE was measured as the tricuspid annulus excursion in millimeters from end-diastole to end-systole by using a floating M-mode cursor (a technique of anatomical M-mode echocardiography) placed through the lateral tricuspid annulus and aligned with its motion.11 RVGLS was computed through 2D speckle-tracking analysis: a region of interest was generated by manually identifying the landmarks of the tricuspid annulus and the RV apex, encompassing 6 segments of the RV lateral wall and septum. The software detected the onset of QRS in the simultaneous electrocardiographic recording to define the zero-strain point and plotted both segmental and global strain curves. The global peak systolic strain was recorded (Fig 1).8 In addition, left ventricular ejection fraction (LVEF) was assessed through the biplane Simpson method. All measurements were performed offline and agreed upon by 2 investigators blinded to patient outcomes. Any disagreements were resolved by a third investigator. VasoactiveInotropic Scoring When the patient arrived in the ICU, medications were adjusted under the consensus of the ICU medical team. Inotropes (dobutamine, epinephrine, or milrinone) usually were used to maintain adequate cardiac output, and vasopressors (dopamine, norepinephrine, or vasopressin) were indicated for hypotension.12 The doses of the inotropic and vasoactive medications were recorded on an hourly basis for the first 24 hours in the ICU. As defined in the literature,4,5 VIS was calculated as (dopamine dose + dobutamine dose) (mg/kg/min) + 10 £ milrinone dose (mg/kg/min) + 100 £ (epinephrine dose + norepinephrine dose) (mg/kg/min) + 10000 £ vasopressin dose (unit/ kg/min). Patients with maximum VIS 20 within the first 24 hours in the ICU were recognized as having high VIS,5 and patients with maximum VIS <20 in the first 24 hours were defined as having low VIS. Patient Partition and Postoperative Adverse Events Patient demographics and comorbidities were recorded prospectively. Preoperative TTE was performed at admission for
ARTICLE IN PRESS P.-C. Ting et al. / Journal of Cardiothoracic and Vascular Anesthesia 00 (2018) 18
3
Fig 1. Speckle-tracking strain of RV. The range of interest was traced manually along the RV endocardial border. Numbers shown on the picture represent segmental and global peak systolic strain values. RVGLS is measured as ¡27.0%. RV, right ventricular; RVGLS, RV global longitudinal strain.
cardiac surgery, and left atrium (LA) size measured in the parasternal long-axis view perpendicular to the axis of aortic root as an anteroposterior diameter was collected prospectively. Postoperative outcomes and complications, namely duration
of vasoactive agent use, time to first extubation, length of ICU and hospital stay, need for mechanical life support, occurrence of acute kidney injury (AKI), and mortality, also were recorded. Based on the Kidney Disease: Improving Global
ARTICLE IN PRESS 4
P.-C. Ting et al. / Journal of Cardiothoracic and Vascular Anesthesia 00 (2018) 18
Outcomes Clinical Practice Guideline,13 AKI was defined as a more than 1.5-fold increase in the serum creatinine compared with the baseline value, which was presumed to have occurred within 7 days of cardiac surgery. Statistical Analysis The parameter of primary interest in this study was RVGLS. According to a previous study,2 the RVGLS was ¡22% in patients with no inotropic support and ¡19% in patients with inotropic support for >48 hours. With an assumed standard deviation of 5%, alpha level of 5%, and power of 80%, a sample size of at least 90 patients was required. Normality of continuous data was tested using the Kolmogorov-Smirnov test. Continuous variables of the patients with low and high VISs were compared using the Student t test or Mann-Whitney U test, as appropriate. The proportions of categorical variables between the 2 study groups were compared using the Fisher exact test. To investigate the association of RV indices with the occurrence of high VIS, the authors developed 3 multivariable logistic regression models, into each of which an individual RV index was introduced and considered an independent variable. In each multivariable model, age and LVEF were considered covariates and were adjusted for other significant (p < 0.05) variables determined through prior univariate analyses. The performance of RV indices for diagnosing the occurrence of high VIS was evaluated through receiver operating characteristic curve analysis. Interobserver reproducibility for RVGLS was evaluated by an independent observer in 18 random patients. For intraobserver variability, the analysis was repeated 4 months after the first measurement in the same 18 patients. Reproducibility was expressed using the coefficient of variation and interclass correlation coefficients.
Postoperative Adverse Events in Patients With High VIS Compared with those with low VIS, patients with high VIS were subjected to the need for extracorporeal membrane oxygenation (ECMO) (p = 0.026) and AKI (p = 0.005) more frequently, and they encountered a significantly higher mortality rate (23.5% v 3.5%, p = 0.013). The detailed data are provided in Table 2. Association of RV Indices With High VIS After adjustment for age, LVEF, and the 3 characteristics (operation type, CPB duration, LA size), each 1% increase in pre-incisional RVGLS was associated significantly with 1.19fold odds of being high VIS (95% CI of odds ratio, 1.04-1.36, p = 0.011). By contrast, RVFAC and TAPSE showed no significant associations with high VIS (Table 3). As shown in Fig 2, the areas under the curves of these RV indices were greater than 0.7, indicating acceptable discrimination ability. Regarding RVGLS, the optimal cutoff value was ¡16.7%, which yielded a sensitivity of 88.2% and a specificity of 75.6% (positive likelihood ratio [LR+] 3.61). The optimal cutoff value of RVFAC was 30.8% (sensitivity 58.8%, specificity 76.7%), and that of TAPSE was 12.3 mm (sensitivity 88.2%, specificity 60.5%). Figure 3 shows that the addition of RVGLS markedly improved the predictive utility of the model that contained aforementioned covariates. The interobserver and intraobserver variabilities, expressed as coefficients of variation, for RVGLS were 7.1% and 6.0%, respectively. The interobserver and intraobserver interclass correlation coefficients for RVGLS were 0.93 (95% CI 0.82-0.98) and 0.95 (95% CI 0.87-0.98), respectively. Discussion
Results Patient Profile The study included 103 patients (mean age 61.2 § 11.0 years, 72 men), of whom 45 (43.7%) underwent CABG alone, 46 (44.7%) underwent isolated valvular surgery, and 12 (11.7%) underwent combined CABG and valvular surgery. Seventeen patients (16.5%) reached high VIS, with a mean maximum VIS of 39 (range 21 to 91) within 24 hours of postoperative admission to the ICU. Compared with patients with low VIS, those with high VIS had significantly larger LA size (p = 0.009) and poorer preincisional RV indices, namely RVFAC (p = 0.005), TAPSE (p = 0.001), and RVGLS (p < 0.001). Furthermore, patients with high VIS were more likely to have undergone isolated valvular surgery (p = 0.031) and to have experienced significantly longer cardiopulmonary bypass (CPB) duration (p < 0.001) than patients with low VIS. No differences were observed in age, sex, LVEF, medical history, and intraoperative hemodynamic measurements, as shown in Table 1.
In this prospective study, the authors used high VIS as a primary outcome in adult patients referred for elective cardiac surgery. Patients who sustained high VIS had impaired preincisional RVFAC, TAPSE, and RVGLS. After adjustments for covariates, RVGLS alone was associated with high VIS. In addition, high VIS was associated with postoperative adverse events such as AKI, the need for ECMO, and mortality. Several risk factors have been identified for predicting postoperative inotropic support.14-16 Advanced age and reduced preoperative LVEF were identified as significant predictors of postoperative inotropic support in patients undergoing CABG or cardiac valvular surgery.15,16 However, the risk factors recognized in this study were inconsistent with those in the aforementioned studies, possibly because of the differences in the primary outcome. Nevertheless, Imai et al. reported that patients who had undergone an extended CPB duration were more likely to require >12 hours of postoperative inotropic support irrespective of their preoperative LVEF,14 and Zaroff et al. demonstrated that not all patients with low preoperative LVEF required postoperative inotropic support.17 LA
ARTICLE IN PRESS P.-C. Ting et al. / Journal of Cardiothoracic and Vascular Anesthesia 00 (2018) 18
5
Table 1 Demographic Data and Clinical Characteristics of Patients With Low and High VISs All n = 103 Age (y) Sex, n (%) Female Male Body surface area (m2) Body mass index (kg/m2) ASA, n (%) III IV NYHA, n (%) I/ II III/ IV Comorbidity, n (%) Hypertension Diabetes mellitus Chronic kidney disease Chronic lung disease Cirrhosis Cerebral vascular attack Coronary artery disease Previous heart surgery Laboratory data Hemoglobin (g/dL) Creatinine (mg/dL) Operation type, n (%) CABG alone Valvular surgery alone Combined surgery Details of valvular surgery (n = 58) Aortic valve surgery (n = 22) Aortic stenosis Aortic regurgitation Mitral valve surgery (n = 40) Mitral stenosis, rheumatic Mitral regurgitation, degenerative Mitral regurgitation, ischemic Tricuspid valve surgery (n = 10) Preoperative echo Left atrium (mm) Preincisional echo LVEF (%) RVFAC (%) TAPSE (mm) RVGLS (%) Concurrent hemodynamics Cardiac index (L/min-m2) Mean arterial pressure (mmHg) MPAP (mmHg) PVRI (mmHg-min-m2/L) Bypass time (min)
Low VIS n = 86
High VIS n = 17
61.2 § 11.0
61.6 § 10.9
59.2 § 11.9
31 (30.1) 72 (69.9) 1.7 § 0.2 24.0 § 3.6
25 (29.1) 61 (70.9) 1.7 § 0.2 24.3 § 3.5
6 (35.3) 11 (64.7) 1.6 § 0.2 22.5 § 4.1
94 (91.3) 9 (8.7)
79 (91.9) 7 (8.1)
15 (88.2) 2 (11.8)
58 (56.3) 45 (43.7)
51 (59.3) 35 (40.7)
7 (41.2) 10 (58.8)
72 (69.9) 48 (46.6) 22 (21.4) 9 (8.7) 3 (2.9) 12 (11.7) 32 (31.1) 11 (10.7)
60 (69.8) 43 (50) 20 (23.3) 9 (10.5) 2 (2.3) 9 (10.5) 29 (33.7) 8 (9.3)
12 (70.6) 5 (29.4) 2 (11.8) 0 (0) 1 (5.9) 3 (17.6) 3 (17.6) 3 (17.6)
1.000 0.183 0.517 0.349 0.421 0.413 0.257 0.385
12.1 § 2.3 1.7 § 2.4
12.3 § 2.3 1.8 § 2.5
11.4 § 2.3 1.5 § 1.6
45 (43.7) 46 (44.7) 12 (11.7)
42 (48.8) 34 (39.5) 10 (11.6)
3 (17.6) 12 (70.6) 2 (11.8)
0.156 0.695 0.035 0.030 0.031 1.000
22 13 9 40 9 23 8 10
15 12 3 30 5 18 7 5
7 1 6 10 4 5 1 5
45.6 § 10.9
43.8 § 8.9
54.6 § 14.9
0.009
45.9 § 11.1 36.7 § 9.1 12.9 § 3.9 ¡18.8 § 6.1
46.0 § 11.3 37.8 § 8.7 13.5 § 3.8 ¡19.8 § 5.3
45.2 § 10.1 31.1 § 8.8 10.1 § 3.3 ¡13.3 § 6.9
0.796 0.005 0.001 <0.001
2.0 § 0.8 82.3 § 16.8 24.7 § 11.1 330.1 § 215.2 143.1 § 81.5
2.0 § 0.8 82.6 § 16.4 23.7 § 9.8 319.5 § 205.6 130.4 § 72.1
2.0 § 0.9 80.3 § 19.2 29.6 § 15.5 383.4 § 258.9 206.9 § 97.8
0.996 0.602 0.151 0.265 <0.001
p Value 0.429 0.579
0.396 0.063 0.641
0.190
NOTE. Continuous data are presented as mean § standard deviation. Abbreviations: ASA, American Society of Anesthesiologists physical status classification; CABG, coronary artery bypass graft; LVEF, left ventricular ejection fraction; MPAP, mean pulmonary arterial pressure; NYHA, New York Heart Association; PVRI, pulmonary vascular resistance index; RVFAC, right ventricular fractional area change; RVGLS, right ventricular global longitudinal strain; TAPSE, tricuspid annular plane systolic excursion.
enlargement has been reported to be an independent predictor for adverse cardiovascular outcomes18,19 and is a marker of both severity and chronicity of diastolic dysfunction, which may correlate with postoperative hemodynamic instability and heavy vasoactive agent use.20 However, the linear expression
of LA size did not show an association with high VIS in the multivariable analysis. In this study, RV function measured using RVGLS was an independent predictor of high VIS after adjustment for risk factors, a finding comparable with that of Ternacle et al.2 Furthermore, these results showed that
ARTICLE IN PRESS 6
P.-C. Ting et al. / Journal of Cardiothoracic and Vascular Anesthesia 00 (2018) 18
Table 2 Postoperative Adverse Events in Patients With Low and High VISs
Variable Vasoactive agent support (h) Intubation time (h) ECMO, n (%) Acute kidney injury, n (%) ICU stay (day) Hospital stay (d) Mortality, n (%)
All n = 103
Low VIS n = 86
High VIS n = 17
p Value
9 (0-57) 15 (5-23) 2 (1.9) 35 (34) 3 (2-5) 16 (10-25) 7 (6.8)
1 (0-40) 14 (5-22) 0 (0) 24 (27.9) 3 (2-5) 16 (10-24) 3 (3.5)
44 (27-97) 20 (12-87) 2 (11.8) 11 (64.7) 4 (3-8) 19 (11-30) 4 (23.5)
<0.001 0.018 0.026 0.005 0.065 0.321 0.013
NOTE. Continuous data are presented as median (25th to 75th percentiles). Abbreviations: ECMO, extracorporeal membrane oxygenation; ICU, intensive care unit.
considering RVGLS in addition to the previously identified risk factors (ie, age, LVEF, operation type, LA size, and CPB duration) in the prediction of postoperative high VIS significantly improved the predictive accuracy. Although many of the RV parameters were considered loaddependent, research of different loading conditions (normal or excessive) has revealed mixed results.21,22 Kjaergaard et al.21 reported that RV strain and TAPSE were unchanged after an increase in the preload and afterload, whereas da Costa Junior et al.,22 studying a group of patients with pulmonary hypertension, reported that all RV indices, including RVFAC, TAPSE, and RVGLS, decreased when RV was confronted by abnormally high pulmonary resistance. Under normal circumstances, RV is easily distensible; hence, the curvature of RV changes little because of the restraint of the pericardium, and the pressure does not rise significantly, rendering the RV indices relatively stable.23 In this study, on average, the concurrent cardiac index and pulmonary vascular resistance index was within the normal range and did not differ significantly between the low- and high-VIS patients; thus, the capacity of
RV indices to predict high VIS is independent of the loading conditions. The superiority of RVGLS over other RV indices in representing RV function can be explained by a global assessment of RV longitudinal performance, which constitutes the major component of RV systolic function,9,24 and by a characteristic that is not greatly affected by cardiac translation.25 By contrast, RVFAC is a simple method to delineate the RV borders that expresses both radial and longitudinal RV function; however, the longitudinal component is attenuated by the radial component. TAPSE focuses on the regional longitudinal movement of the RV lateral wall and is deemed problematic when measured using conventional M-mode through TEE because of its dependence on transducer orientation.7 The authors adopted a technique of anatomical M-mode in which a floating cursor is used to align freely the direction of the tricuspid motion during post-processing of the stored 2D digital images, thus eliminating the aforementioned TEE-related concern. Tousignant et al. reported that TAPSE measured using TEE through antomical M-mode (r = 0.48, p < 0.001) or the
Table 3 Associations Between Preincisional RV Indices and Risk of High VIS Characteristics Univariate logistic regression analysis Age, per 10 y Operation type CABG alone Valve alone Mixed Bypass time, per 10 min Left atrium, mm LVEF, per 5% RVFAC, % TAPSE, mm RVGLS, % Multivariable logistic models* RVFAC, % TAPSE, mm RVGLS, %
OR
95% CI
p Value
0.83
0.52 to 1.32
0.426 0.065
Ref. 4.94 2.80 1.13 1.08 0.97 0.91 0.74 1.22
1.29 to 18.94 0.41 to 19.05 1.05 to 1.23 1.03 to 1.14 0.76 to 1.23 0.85 to 0.98 0.61 to 0.90 1.09 to 1.35
0.020 0.293 0.001 0.002 0.794 0.008 0.002 <0.001
0.95 0.83 1.19
0.87 to 1.03 0.67 to 1.03 1.04 to 1.36
0.180 0.090 0.011
Abbreviations: CABG, coronary artery bypass graft; CI, confidence interval; LVEF, left ventricular ejection fraction; OR, odds ratio; RVFAC, right ventricular fractional area change; RVGLS, right ventricular global longitudinal strain; TAPSE, tricuspid annular plane systolic excursion. * Adjusted for age, operation type, bypass time, left atrium size, and left ventricular ejection fraction; each RV index was modeled separately.
ARTICLE IN PRESS P.-C. Ting et al. / Journal of Cardiothoracic and Vascular Anesthesia 00 (2018) 18
Fig 2. Receiver operating characteristic curve analysis. Performance of preincisional echocardiographic parameters in discriminating the risk of high VIS. VIS, vasoactive-inotropic score.
speckle-tracking method (r = 0.44, p = 0.001) correlated modestly with stroke volume, which may be helpful in predicting a risk or an outcome.11 However, TAPSE measured by either
7
method was not sensitive enough to track output changes and thus was suggested as an incomplete assessment of RV function.7,11 In the univariate analysis, the authors did not apply specific cutoff values to designate RVD, because the values were derived from previous TTE studies6; nevertheless, continuous data of preincisional RVFAC, TAPSE, and RVGLS measured through TEE were decreased significantly in patients with high VIS. Moreover, a cutoff value was recommended to predict a specific outcome rather than an abnormal value that depended on the mean obtained from a group of healthy people.6 Thus, the receiver operating characteristic analysis in this study determined the optimal cutoff value of each RV index for high VIS. Although the areas under the curves of the RV indices did not differ significantly, the LR+ of RV indices helped discriminate their responsiveness. When RVGLS of ¡16.7% was used as the cutoff point, LR+ of 3.61 increased the prevalence of high VIS (16.5%) to the “post-test” prevalence of 41.5%, thus showing moderate responsiveness. When RVFAC of 30.8% and TAPSE of 12.8 mm were used as the cutoff values, the post-test prevalence of high VIS was 30% to 33%, showing less responsiveness than RVGLS. Based on previous studies,4,5 this study solicited high VIS as an effective primary outcome and found that patients with high VIS experienced more severe postoperative complications than did patients with low VIS. Notably, the finding that high VIS is associated with postoperative adverse events, including prolonged time to first extubation and increased occurrence of need for ECMO, AKI, and mortality, was consistent with the findings of Gaies et al.5 Although the length of ICU and hospital stay did not differ between study groups, the results easily may have been affected by domestic insurance regulations and other issues during the treatment course, as was the case in Gaies et al.4 Limitations
Fig 3. Incremental values of RV indices for predicting postoperative high VIS. The addition of RVGLS to age, left ventricular ejection fraction (LVEF), operation type (OP type), left atrium size (LA), and bypass time significantly improved the predictive value of the proposed model. RV, right ventricular; RVGLS, RV global longitudinal strain; VIS, vasoactive-inotropic score.
The study population was not homogenous and underwent a range of cardiac surgeries, including CABG, simple valvular surgery, and combined surgery. Considering the initial finding that patients referred for valvular surgery were subjected to postoperative high VIS, the authors made adjustments for the operation type in the regression models, and preincisional RVGLS remained an independent predictor of high VIS. For speckle-tracking measurement of RV, 2 methods currently are applied.6 RV free wall strain (RVFWS) averages longitudinal strains on the 3 segments of the lateral wall, and RVGLS encompasses the 6 segments of the whole RV, including the interventricular septum. The authors used RVGLS rather than RVFWS for RV measurement. Although RVFWS is considered relatively exclusive to the RV native function, 40% of the RV systolic function pertains to the septum and should not be ignored.24 A previous study demonstrated that impaired baseline RVGLS indicated poor outcomes in patients undergoing cardiac resynchronization therapy.26 This study showed that preincisional RVGLS is associated independently with postoperative high VIS and that RVGLS has an important
ARTICLE IN PRESS 8
P.-C. Ting et al. / Journal of Cardiothoracic and Vascular Anesthesia 00 (2018) 18
role in the evaluation of prognosis in patients referred for cardiac surgery. Because evidence offering guidance on the choice of vasoactive agents is limited, VIS could be affected by personal preference or misinterpretation of hemodynamic data. To eliminate personnel bias, postoperative care in the ICU is provided by a fixed medical team in the authors’ institute to ensure consensus on the choice of vasoactive and inotropic agents and to adjust the dosage in these patients. Despite this prudence, variation in practice was tolerated in a multicenter study of VIS,5 and high VIS retained its association with poor outcomes. Future studies to guide the most efficient strategy for addressing specific clinical situations should be conducted in collaboration with VIS.
8
9
10
11 12
13
Conclusion 14
This study concludes that preoperative depressed RV function is an independent risk factor for occurrence of postoperative high VIS. In the assessment of RV function, pre-incisional RVGLS measured using TEE is a reliable tool to predict postoperative high VIS in patients referred for elective cardiac surgery. In addition, high VIS acts as a practical outcome measurement, D18X Xpredicting early postoperative complications in the population of adult cardiac surgery. The authors thank Alfred Hsing-Fen Lin for his assistance in statistical analysis. References 1 Maslow AD, Regan MM, Panzica P, et al. 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. 2 Ternacle J, Berry M, Cognet T, et al. Prognostic value of right ventricular two-dimensional global strain in patients referred for cardiac surgery. J Am Soc Echocardiogr 2013;26:721–6. 3 Haddad F, Denault AY, Couture P, 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. 4 Gaies MG, Gurney JG, Yen AH, et al. Vasoactive-inotropic score as a predictor of morbidity and mortality in infants after cardiopulmonary bypass. Pediatr Crit Care Med 2010;11:234–8. 5 Gaies MG, Jeffries HE, Niebler RA, et al. Vasoactive-inotropic score is associated with outcome after infant cardiac surgery: An analysis from the Pediatric Cardiac Critical Care Consortium and Virtual PICU System Registries. Pediatr Crit Care Med 2014;15:529–37. 6 Lang RM, Badano LP, Mor-Avi V, et al. Recommendations for cardiac chamber quantification by echocardiography in adults: An update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr 2015;28:1–39;e14. 7 Markin NW, Chamsi-Pasha M, Luo J, et al. Transesophageal speckletracking echocardiography improves right ventricular systolic function
15
16
17
18
19
20
21
22
23 24
25
26
assessment in the perioperative setting. J Am Soc Echocardiogr 2017;30:180–8. Tousignant C, Desmet M, Bowry R, et al. Speckle tracking for the intraoperative assessment of right ventricular function: A feasibility study. J Cardiothorac Vasc Anesth 2010;24:275–9. Silverton N, Meineri M. Speckle tracking strain of the right ventricle: An emerging tool for intraoperative echocardiography. Anesth Analg 2017;125:1475–8. Ting PC, Chou AH, Wu VCC, et al. Relationship between right ventricular function and atrial fibrillation after cardiac surgery. J Cardiothorac Vasc Anesth 2017;31:1663–71. Tousignant C, Kim H, Papa F, et al. Evaluation of TAPSE as a measure of right ventricular output. Can J Anaesth 2012;59:376–83. Stephens RS, Whitman GJ. Postoperative critical care of the adult cardiac surgical patient. Part I: Routine postoperative care. Crit Care Med 2015;43:1477–97. Palevsky PM, Liu KD, Brophy PD, et al. KDOQI US commentary on the 2012 KDIGO clinical practice guideline for acute kidney injury. Am J Kidney Dis 2013;61:649–72. Imai H, Kurokawa S, Taneoka M, et al. Tissue Doppler imaging is useful for predicting the need for inotropic support after cardiac surgery. J Anesth 2011;25:805–11. Butterworth JF 4th, C Legault, Royster RL, et al. Factors that predict the use of positive inotropic drug support after cardiac valve surgery. Anesth Analg 1998;86:461–7. Royster RL, Butterworth JF 4th, Prough DS, et al. Preoperative and intraoperative predictors of inotropic support and long-term outcome in patients having coronary artery bypass grafting. Anesth Analg 1991;72:729–36. Zaroff J, Aronson S, Lee BK, et al. The relationship between immediate outcome after cardiac surgery, homogeneous cardioplegia delivery, and ejection fraction. Chest 1994;106:38–45. Kizer JR, Bella JN, Palmieri V, et al. Left atrial diameter as an independent predictor of first clinical cardiovascular events in middle-aged and elderly adults: The Strong Heart Study (SHS). Am Heart J 2006;151:412–8. Bouzas-Mosquera A, Broullon FJ, Alvarez-Garcia N, et al. Left atrial size and risk for all-cause mortality and ischemic stroke. CMAJ 2011;183: E657–64. Bernard F, Denault A, Babin D, et al. Diastolic dysfunction is predictive of difficult weaning from cardiopulmonary bypass. Anesth Analg 2001;92:291–8. Kjaergaard J, Snyder EM, Hassager C, et al. Impact of preload and afterload on global and regional right ventricular function and pressure: A quantitative echocardiography study. J Am Soc Echocardiogr 2006;19:515–21. da Costa Junior AA, Ota-Arakaki JS, Ramos RP, et al. Diagnostic and prognostic value of right ventricular strain in patients with pulmonary arterial hypertension and relatively preserved functional capacity studied with echocardiography and magnetic resonance. Int J Cardiovasc Imaging 2017;33:39–46. Greyson CR. Right heart failure in the intensive care unit. Curr Opin Crit Care 2012;18:424–31. Haddad F, Hunt SA, Rosenthal DN, et al. Right ventricular function in cardiovascular disease, Part I: Anatomy, physiology, aging, and functional assessment of the right ventricle. Circulation 2008;117:1436–48. Giusca S, Dambrauskaite V, Scheurwegs C, et al. Deformation imaging describes right ventricular function better than longitudinal displacement of the tricuspid ring. Heart 2010;96:281–8. Nagy VK, Szeplaki G, Apor A, et al. Role of right ventricular global longitudinal strain in predicting early and long-term mortality in cardiac resynchronization therapy patients. PLoS One 2015;10:e0143907.