Diastolic Dysfunction in Patients Undergoing Cardiac Surgery: The Role of Gender and Age-Gender Interaction Renata G. Ferreira, MD,*1 Alina Nicoara, MD,*1 Barbara G. Phillips-Bute, PhD,* Mani Daneshmand, MD,† Jochen D. Muehlschlegel, MD,‡ and Madhav Swaminathan, MD*† Objectives: To test the hypothesis that females presenting for coronary artery bypass graft (CABG) surgery are at a higher risk of left ventricular diastolic dysfunction (LVDD) and that age and gender interact to influence this risk. Design: Retrospective observational study. Setting: Tertiary university hospital. Participants: Eight hundred-ninety-five adult patients undergoing CABG surgery. Interventions: None. Measurements and Main Results: Baseline diastolic function was graded according to a predefined Doppler-based algorithm, which defined LVDD as a binary variable (grades 2 and 3 only) and as a continuous variable (E/e’ ratio). The authors found that women were more likely to present with LVDD in 2 multivariate regression models using both LVDD definitions (odds ratio ¼ 2.7; p o 0.0001 for logistic model, and parameter estimate (PE) ¼ 2.8; p o 0.0001 for the linear
model). In addition, there was a significant age and gender interaction on the risk of LVDD in the linear model (PE ¼ 0.08; p ¼ 0.01). A restricted cubic splines analysis revealed a progressively higher risk of LVDD (predicted E/e’ ratio) among older women. Conclusions: The authors confirmed that women undergoing CABG surgery are at higher risk of LVDD compared to men with a significant age-gender interaction suggesting a possible age-related differential effect on LVDD between the genders, a phenomenon previously demonstrated in preclinical studies. Therapies aimed at amelioration of diastolic dysfunction additionally should consider the higher risk in females, especially within the older subset of the patient population. & 2014 Elsevier Inc. All rights reserved.
L
complete intraoperative TEE data had been acquired in order to allow diastolic function grading. Patients who were determined by the clinical care team to be hemodynamically unstable, or had any arrhythmia were excluded. Patients dependent preoperatively on electrical pacing, inotropes, or intra-aortic balloon pump support also were excluded. Anesthesia and surgery were conducted based on an institutional protocol, which previously has been described.8 Data were extracted from the Duke Databank for Cardiovascular Diseases, a large, quality-assured database that prospectively gathers demographic, periprocedural, and followup data for all cardiovascular interventions performed at Duke University. Measurements pertaining to diastolic function were obtained at baseline after induction of general anesthesia. All measurements were made in real time and their accuracy verified offline by an independent, board-certified echocardiographer who was blinded to grading algorithm and outcome. Assessment of LV diastolic function was performed using a previously–described, simplified algorithm based on the peak early transmitral flow velocity (E) and early diastolic velocity of the lateral mitral annulus (e’). Diastolic function was defined in 4 grades: Normal (grade 0), impaired relaxation (grade 1), pseudonormal (grade 2), and restrictive (grade 3) as shown in Figure 1. Diastolic dysfunction was defined in 2 ways. First, as a binary variable (grades 2 or 3 only), and second, as a continuous variable using the E/e’ ratio as a measure of LV stiffness. While grades 0 and 1 were considered to reflect normal left atrial pressure, grades 2 (pseudonormal) and 3 (restrictive filling) reflected elevated left atrial
EFT VENTRICULAR DIASTOLIC DYSFUNCTION (LVDD) refers to an abnormality of diastolic distensibility, filling, or relaxation and is an important determinant of cardiovascular function. Indeed, LVDD is the predominant abnormality in patients with heart failure with preserved ejection fraction (HFpEF), which comprises approximately half of the patients with heart failure.1,2 In contrast to heart failure with reduced ejection fraction, the incidence of HFpEF has increased over the years and if these trends persist, HFpEF may become the most common type of heart failure in the near future.1,3 Studies indicate that women are at higher risk for developing HFpEF compared to men.4 Investigators have also shown that gender and age-gender interaction may play a role in the pathogenesis of LVDD.5,6 Possible mechanisms are changes in the hormonal milieu that accompany normal aging, which subsequently contribute to the development of hypertension and left ventricular hypertrophy and predispose women to a higher risk of LVDD.7 Left ventricular diastolic dysfunction also is common in patients with ischemic heart disease and has been shown to be associated with long-term adverse outcomes in patients undergoing coronary artery bypass grafting (CABG) surgery.8 Similar to their risk for developing HFpEF, women also have a higher risk for morbidity and mortality following CABG than their male counterparts.9 However, there are limited perioperative data on the association among gender, age, and LVDD in cardiac surgery patients. The authors, therefore, conducted a study to determine the association between gender and LVDD in the setting of CABG surgery. The authors tested the hypothesis that females undergoing CABG surgery are more likely than male patients to present with LVDD. The authors further hypothesized that age and gender interact to influence this risk. METHODS With approval from the authors’ institutional review board, detailed clinical and intraoperative transesophageal echocardiographic data were analyzed in a cohort of 897 adult patients undergoing elective, isolated CABG surgery with CPB at Duke University Medical Center in whom
KEY WORDS: coronary artery disease, diastole, surgery, women, aging, diastolic dysfunction
From the Departments of *Anesthesiology, and †Surgery, Duke University Medical Center, Durham, NC; and ‡Department of Anesthesiology, Perioperative, and Pain Medicine, Harvard Medical School, Boston, MA. 1 Both of the authors contributed to this work equally. Address reprint requests to Madhav Swaminathan, MD, Box 3094, DUMC Anesthesiology, 2301 Erwin Road, Durham, NC 27710. E-mail: [email protected] © 2014 Elsevier Inc. All rights reserved. 1053-0770/2601-0001$36.00/0 http://dx.doi.org/10.1053/j.jvca.2013.11.010
Journal of Cardiothoracic and Vascular Anesthesia, Vol ], No ] (Month), 2014: pp ]]]–]]]
1
2
DIASTOLIC DYSFUNCTION IN CARDIAC SURGERY PATIENTS
Fig 1. Algorithm used for grading diastolic dysfunction. Abbreviations: TDI, tissue Doppler imaging.
pressure and were therefore categorized together as LVDD for the binary outcome. The E/e’ ratio is an established marker of abnormal LV filling pressures in patients with cardiovascular disease.10 Demographic and procedural variables were compared between the 2 gender groups using analysis of variance (ANOVA) and the Wilcoxon rank-sum tests for continuous variables and Chi square or Fischer’s exact test for categoric variables as appropriate. For the authors’ primary hypothesis, a multivariate logistic regression model was developed with LVDD as the dependent binary variable (grades 2 or 3 only) and female gender as the principal independent variable of interest. A second multivariate linear regression model was developed to test the independent association between female gender and E/e’. In both models, the authors adjusted for age, congestive heart failure, weight, diabetes, hypertension, LV ejection fraction, and cardiopulmonary bypass time. For the secondary hypothesis, the authors investigated the interaction between age and gender and their association with diastolic dysfunction by adding the ‘age-gender’ interaction term to both regression models. For further descriptive purposes, the authors employed a restricted cubic spline analysis that allowed them to explore the possibility of nonlinear relationships between age and the E/e’ ratio. The inclusion of the interaction term (age * gender) allows the slopes of the lines to vary between males and females. The spline analysis allows the smoothed fitted line to vary in slope at predetermined points and is constrained to be linear in the tails. The location of the splines, or knots (the points at which the slope of the line is free to vary) were chosen based on recommended quantiles described by Harrell.11 Predicted values were produced from the multivariate model, which was adjusted for the same covariables as in the primary analysis. The predictors of interest for the purpose of this analysis were age and gender and their interaction. A p value of less than 0.05 was considered statistically significant. All analyses were conducted using SAS statistical software (Version 9.1.3, SAS Inc, Cary, NC). RESULTS
Demographic and procedural variables for the cohort of 895 patients are described in Table 1. There were 158 patients (18%) with normal diastolic function while 309 (34%) had grade 1, 207 (30%) had grade 2, and 161 (18%) patients had grade 3 LVDD. Statistically significant differences between the two genders were noted in several variables as shown in table 1. A greater proportion of females were Caucasian (27 vs. 16%). Female patients were also older (66 vs. 63 years),
had higher rates of diabetes (44 vs. 32%), and chronic obstructive pulmonary disease (18 vs. 11%) compared to male patients. In addition, females had a lower average body weight (83 vs. 92 kg), and a lower average left ventricular ejection fraction (50 vs. 53%). Significant differences in diastolic function variables were also noted. Female patients had a higher average baseline E/e’ ratio (11.5 v 8.3) and a correspondingly higher percentage of cases with grades 2 or 3 LVDD (68 v 41%), indicating worse baseline diastolic function in this subset of patients. The authors’ primary hypothesis was confirmed in both multivariable models that defined LVDD as a binary variable and LV diastolic stiffness (E/e’) as a continuous variable. Female patients were more likely to present with LVDD defined as a binary variable (odds ratio ¼ 2.73; 95% confidence interval ¼ 1.9-3.8; p o 0.0001) (Table 2). Female gender remained significantly associated with E/e’ in the linear regression model (parameter estimate ¼ 2.78; p o 0.0001) (Table 3). Other significant co-variables were age, diabetes and low ejection fraction. For the secondary hypothesis, the addition of the age-gender interaction term to the logistic regression model showed marginal significance. (Linear regression model showed that there was a significant association between age-gender and E/e’ (parameter estimate ¼ 0.08; p ¼ 0.01). The results of the restricted cubic spline analysis of the interaction between age and gender and their association with diastolic dysfunction are shown in Figure 2. The specific ages that are knots (the points at which the lines are free to vary) in this analysis are 80, 72, 65, 57, and 47. The risk in diastolic dysfunction increases in slope between 60 and 70 for both males and females but is steeper in females and does not reach a plateau as it does in males.
Table 1. Demographic and Operative Characteristics
Variables
Age (years) Caucasian (%) Congestive heart failure (%) Hypertension, n (%) Diabetes mellitus (%) Weight (kg) Left ventricular ejection fraction (%) Preoperative creatinine (mg/dL) COPD (%) Previous MI (%) EuroSCORE Cardiopulmonary bypass time (min) Aortic cross-clamp time (min) Baseline E/e’ ratio LVDD (grade 2 or 3) (%)
Females (n ¼ 236)
Males (n ¼ 661)
p Value
66.6 (10.0) 26.7 18.5 77.6 44.0 83.4 (23.2) 49.9 (14.9)
63.3 (10.2) 15.8 17.6 74.8 32.2 92.1 (20.8) 53.3 (13.6)
o0.0001 0.0002 0.74 0.40 0.001 o0.0001 0.02
1.2 (1.1) 18.1 18.7 9.2 (3.0) 128 (61)
1.3 (1.2) 11.3 39.7 8.1 (3.3) 124 (42)
0.33 0.008 0.12 o0.0001 0.38
72 (25) 11.5 (5.8) 67.7
71 (23) 8.3 (4.1) 40.9
0.90 o0.0001 o0.0001
NOTE. Values are expressed as mean with standard deviation in parentheses unless specified otherwise. The p values are Wilcoxon rank sums or 2-Tailed Fisher’s exact test as appropriate. Abbreviations: COPD, chronic obstructive pulmonary disease; LVDD, left ventricular diastolic dysfunction; MI, myocardial infarction.
3
FERREIRA ET AL
Table 2. Multivariate Logistic Regression Model of Variables Independently Associated with Diastolic Dysfunction Wald 95% Variable
Parameter Estimate
Female gender Age CHF Weight COPD Diabetes Hypertension Ejection fraction CPB time
Standard Error
Odds Ratio
Confidence Limits
p Value
1.01
0.17
2.73
1.94-3.84
o0.0001
0.05 0.26 0.001 0.43 0.37 0.08 0.02
0.01 0.20 0.003 0.22 0.16 0.18 0.01
1.05 1.30 1.00 1.53 1.45 0.93 0.98
1.04-1.07 0.87-1.92 0.99-1.00 0.99-2.37 1.06-1.99 0.66-1.31 0.97-0.99
o0.0001 0.2 0.76 0.05 0.02 0.67 0.0039
0.003
0.002
1.00
1.00-1.01
0.05
NOTE. Diastolic dysfunction defined as a binary variable (only grades 2 or 3; see text for details). Abbreviations: CHF, congestive heart failure; CPB, cardiopulmonary bypass; COPD, chronic obstructive pulmonary disease.
DISCUSSION
The authors confirmed their primary hypothesis that women undergoing cardiac surgery are more likely to present with LVDD compared to men. They also found that older women are at a higher risk for LVDD. These findings add clinical significance to data from animal models that report a mechanistic link between female gender and impaired diastolic function. The authors’ study also suggests that aging may play a greater role in the risk of LVDD in women compared to men. Strategies for management of perioperative LVDD should consider the additional risk among elderly female patients. Left ventricular diastolic dysfunction is an ubiquitous finding in patients with heart failure, regardless of systolic function, and is a strong predictor of cardiovascular morbidity and mortality.12 In patients presenting for cardiovascular surgery, echocardiographically defined LVDD is a prognostic marker of adverse outcome.8,13,14 The association between female gender and cardiovascular risk has been extensively investigated, and studies suggest a complex relationship. A report from the Society of Thoracic Surgeons found that while Table 3. Multivariate Linear Regression Model of Variables Independently Associated with Left Ventricular Diastolic Stiffness Variable
Intercept Female gender Age CPB time CHF Weight COPD Diabetes Hypertension Ejection fraction
Parameter Estimate
Standard Error
t Value
p value
5.16 2.78 0.10 0.004 0.17 0.006 0.99 0.66 0.16 0.06
1.48 0.35 0.02 0.003 0.41 0.01 0.45 0.33 0.36 0.013
3.50 7.88 6.46 1.36 0.40 0.83 2.19 2.00 0.45 4.82
0.0005 o0.0001 o0.0001 0.17 0.69 0.41 0.0289 0.05 0.65 o0.0001
NOTE. Diastolic dysfunction defined as a continuous variable binary variable (E/e’ ratio; see text for details). Abbreviations: CHF, congestive heart failure; CPB, cardiopulmonary bypass; COPD, chronic obstructive pulmonary disease.
Fig 2. Cubic splines analysis demonstrating the age-gender interaction and risk of diastolic dysfunction. There is an increase in slope between the ages of 60 and 70 for both males and females, but it is steeper in females compared to males.
women have a higher mortality risk after CABG surgery, the difference in operative mortality rates between genders is absent in the high-risk sub-categories.15 Another cardiac surgery registry study reported disproportionately higher mortality rates among women compared to men undergoing CABG surgery. However, this risk was particularly significant among younger women, and the effect decreased with advancing age.16 An Australian cardiac surgery database study reported higher mortality rates among women undergoing CABG surgery that were likely attributable to worse risk profiles.17 The same group also reported no gender differences in outcomes following CABG with concurrent aortic valve surgery.18 Koch and colleagues found that the higher prevalence of comorbid conditions and not gender primarily influenced the increased risk of postoperative mortality among women undergoing CABG surgery.19 The higher perioperative risk among women may therefore be influenced more by gender differences in severity of comorbidities such as hypertension and heart failure. Heart failure, especially HFpEF, is more prevalent among older women.2,20 Population studies also suggest gender differences in incidence of arterial stiffness and hypertension.21,22 Animal models of cardiovascular disease support these observations and offer some insights into possible mechanisms. In ovariectomized rats, Groban and colleagues found that the altered hormonal environment was associated with an increase in the severity of diastolic stiffness.23 They also showed that the pathogenesis of diastolic dysfunction in female rats involved alterations in the myocardial interstitium that were strongly linked with metabolic changes resulting from estrogen depletion, such as shifts in the neuronal nitric oxide synthase pathway.24 They suggested that post menopausal women are at higher risk of diastolic dysfunction due to the metabolic consequences of estrogen loss on the myocardium. Taken
4
DIASTOLIC DYSFUNCTION IN CARDIAC SURGERY PATIENTS
together, these studies suggest that women are at higher risk of cardiovascular events compared to men due to structural differences in vascular and myocardial architecture. These differences appear to be associated, in part, with changes in the hormonal milieu that accompany natural ageing. The authors’ finding of higher incidence of diastolic dysfunction among women presenting for cardiac surgery provides clinical evidence to support mechanisms suggested by preclinical data. Although the authors found an interaction of age and female gender with diastolic stiffness only in the validation cohort, they feel that these findings may reflect the effect of post menopausal hormonal changes on the myocardium. Other factors such as diabetes and low ejection fraction that were significant independent variables in their regression analyses confirmed previous observations of their known association with LVDD.25,26 The authors’ study is strengthened by confirmation of their hypotheses in a large study population with well-defined echocardiographic and outcomes parameters. However, there are also limitations with respect to study design and data collection. The data was retrospectively collected, and the quality of data reflects the inherent limitations of this approach. The authors did not have hemodynamic validation of elevated filling pressures as this data was not collected. However, they are confident that the metrics used for defining diastolic dysfunction and LV stiffness previously have been validated and reflect filling pressures with reasonable accuracy in all their patients. Although the simplified algorithm they used for categorizing
LVDD was developed retrospectively and has not been prospectively validated, it has been endorsed by the American Society of Echocardiography in their recent guidelines for a comprehensive TEE examination.27 In the absence of any other algorithm validated against adverse outcomes in the perioperative CABG surgery population, the authors feel justified in using the simplified algorithm for their study. It also could be argued that hormone replacement therapy in postmenopausal patients could have altered the incidence of diastolic dysfunction and confounded their results. However, estrogen-based therapies have not consistently been shown to mitigate cardiovascular risk.28,29 Therefore, the authors feel that their findings would not be substantially different if they included data on preoperative hormone therapies in their patients. In summary, the authors found a significantly higher incidence of diastolic dysfunction among women presenting for cardiac surgery compared to men. They also noted that the gender difference in diastolic dysfunction was also influenced by age, a finding supported by pre-clinical disease models. Their findings imply that the higher incidence of LVDD among women should alert the perioperative team to consider this cohort of patients at higher risk for the adverse outcomes known to be associated with LVDD and tailor cardiovascular management and postoperative surveillance appropriately. Therapies aimed at amelioration of diastolic dysfunction should additionally consider the higher risk in females, especially within the older subset of the patient population.
REFERENCES 1. Owan TE, Hodge DO, Herges RM, et al: Trends in prevalence and outcome of heart failure with preserved ejection fraction. N Engl J Med 355:251-259, 2006 2. Redfield MM, Jacobsen SJ, Burnett JC Jr, et al: Burden of systolic and diastolic ventricular dysfunction in the community: Appreciating the scope of the heart failure epidemic. JAMA 289:194-202, 2003 3. Paulus WJ, Tschope C, Sanderson JE, et al: How to diagnose diastolic heart failure: A consensus statement on the diagnosis of heart failure with normal left ventricular ejection fraction by the Heart Failure and Echocardiography Associations of the European Society of Cardiology. Eur Heart J 28:2539-2550, 2007 4. Lee DS, Gona P, Vasan RS, et al: Relation of disease pathogenesis and risk factors to heart failure with preserved or reduced ejection fraction: Insights from the Framingham Heart Study of the National Heart, Lung, and Blood Institute. Circulation 119:3070-3077, 2009 5. Wang H, Jessup JA, Lin MS, et al: Activation of GPR30 attenuates diastolic dysfunction and left ventricle remodelling in oophorectomized mRen2.Lewis rats. Cardiovasc Res 94:96-104, 2012 6. Jessup JA, Zhang L, Presley TD, et al: Tetrahydrobiopterin restores diastolic function and attenuates superoxide production in ovariectomized mRen2.Lewis rats. Endocrinology 152:2428-2436, 2011 7. Agabiti-Rosei E, Muiesan ML: Left ventricular hypertrophy and heart failure in women. J Hypertens Suppl 20:S34-S38, 2002 8. Swaminathan M, Nicoara A, Phillips-Bute BG, et al: Utility of a simple algorithm to grade diastolic dysfunction and predict outcome after coronary artery bypass graft surgery. Ann Thorac Surg 91: 1844-1850, 2011 9. Hannan EL, Racz MJ, Walford G, et al: Predictors of readmission for complications of coronary artery bypass graft surgery. JAMA 290: 773-780, 2003
10. Nagueh SF, Appleton CP, Gillebert TC, et al: Recommendations for the evaluation of left ventricular diastolic function by echocardiography. J Am Soc Echocardiogr 22:107-133, 2009 11. Harrell FE Jr, Lee KL, Pollock BG: Regression models in clinical studies: Determining relationships between predictors and response. J Natl Cancer Inst 80:1198-1202, 1988 12. Zile MR, Brutsaert DL: New concepts in diastolic dysfunction and diastolic heart failure: Part I: Diagnosis, prognosis, and measurements of diastolic function. Circulation 105:1387-1393, 2002 13. Matyal R, Hess PE, Subramaniam B, et al: Perioperative diastolic dysfunction during vascular surgery and its association with postoperative outcome. J Vasc Surg 50:70-76, 2009 14. Flu WJ, van Kuijk JP, Hoeks SE, et al: Prognostic implications of asymptomatic left ventricular dysfunction in patients undergoing vascular surgery. Anesthesiology 112:1316-1324, 2010 15. Edwards FH, Carey JS, Grover FL, et al: Impact of gender on coronary bypass operative mortality. Ann Thorac Surg 66: 125-131, 1998 16. Vaccarino V, Abramson JL, Veledar E, et al: Sex differences in hospital mortality after coronary artery bypass surgery: Evidence for a higher mortality in younger women. Circulation 105:1176-1181, 2002 17. Saxena A, Dinh D, Smith JA, et al: Sex differences in outcomes following isolated coronary artery bypass graft surgery in Australian patients: Analysis of the Australasian Society of Cardiac and Thoracic Surgeons cardiac surgery database. Eur J Cardiothorac Surg 41:755-762, 2012 18. Saxena A, Poh CL, Dinh DT, et al: Does patient gender affect outcomes after concomitant coronary artery bypass graft and aortic valve replacement? An Australian Society of Cardiac and Thoracic Surgeons Database study. Cardiology 119:116-123, 2011
FERREIRA ET AL
19. Koch CG, Weng YS, Zhou SX, et al: Prevalence of risk factors, and not gender per se, determines short- and long-term survival after coronary artery bypass surgery. J Cardiothorac Vasc Anesth 17:585-593, 2003 20. Masoudi FA, Havranek EP, Smith G, et al: Gender, age, and heart failure with preserved left ventricular systolic function. J Am Coll Cardiol 41:217-223, 2003 21. Russo C, Jin Z, Palmieri V, et al: Arterial stiffness and wave reflection: sex differences and relationship with left ventricular diastolic function. Hypertension 60:362-368, 2012 22. Deschepper CF, Llamas B: Hypertensive cardiac remodeling in males and females: From the bench to the bedside. Hypertension 49: 401-407, 2007 23. Groban L, Yamaleyeva LM, Westwood BM, et al: Progressive diastolic dysfunction in the female mRen(2). Lewis rat: Influence of salt and ovarian hormones. The journals of gerontology Series A, Biological sciences and medical sciences 63:3-11, 2008 24. Jessup JA, Zhang L, Chen AF, et al: Neuronal nitric oxide synthase inhibition improves diastolic function and reduces oxidative stress in ovariectomized mRen2.Lewis rats. Menopause 18:698-708, 2011
5
25. Poulsen MK, Henriksen JE, Dahl J, et al: Left ventricular diastolic function in type 2 diabetes mellitus: Prevalence and association with myocardial and vascular disease. Circ Cardiovasc Imaging 3: 24-31, 2010 26. Zile MR, Baicu CF, Gaasch WH: Diastolic heart failure— abnormalities in active relaxation and passive stiffness of the left ventricle. N Engl J Med 350:1953-1959, 2004 27. Hahn RT, Abraham T, Adams MS, et al: Guidelines for performing a comprehensive transesophageal echocardiographic examination: Recommendations from the American Society of Echocardiography and the Society of Cardiovascular Anesthesiologists. J Am Soc Echocardiogr 26:921-964, 2013 28. Rossouw JE, Anderson GL, Prentice RL, et al: Risks and benefits of estrogen plus progestin in healthy postmenopausal women: Principal results from the Women’s Health Initiative randomized controlled trial. JAMA 288:321-333, 2002 29. Wells G, Herrington DM: The Heart and Estrogen/Progestin Replacement Study: What have we learned, and what questions remain? Drugs Aging 15:419-422, 1999
model). In addition, there was a significant age and gender interaction on the risk of LVDD in the linear model (PE ¼ 0.08; p ¼ 0.01). A restricted cubic splines analysis revealed a progressively higher risk of LVDD (predicted E/e’ ratio) among older women. Conclusions: The authors confirmed that women undergoing CABG surgery are at higher risk of LVDD compared to men with a significant age-gender interaction suggesting a possible age-related differential effect on LVDD between the genders, a phenomenon previously demonstrated in preclinical studies. Therapies aimed at amelioration of diastolic dysfunction additionally should consider the higher risk in females, especially within the older subset of the patient population. & 2014 Elsevier Inc. All rights reserved.
L
complete intraoperative TEE data had been acquired in order to allow diastolic function grading. Patients who were determined by the clinical care team to be hemodynamically unstable, or had any arrhythmia were excluded. Patients dependent preoperatively on electrical pacing, inotropes, or intra-aortic balloon pump support also were excluded. Anesthesia and surgery were conducted based on an institutional protocol, which previously has been described.8 Data were extracted from the Duke Databank for Cardiovascular Diseases, a large, quality-assured database that prospectively gathers demographic, periprocedural, and followup data for all cardiovascular interventions performed at Duke University. Measurements pertaining to diastolic function were obtained at baseline after induction of general anesthesia. All measurements were made in real time and their accuracy verified offline by an independent, board-certified echocardiographer who was blinded to grading algorithm and outcome. Assessment of LV diastolic function was performed using a previously–described, simplified algorithm based on the peak early transmitral flow velocity (E) and early diastolic velocity of the lateral mitral annulus (e’). Diastolic function was defined in 4 grades: Normal (grade 0), impaired relaxation (grade 1), pseudonormal (grade 2), and restrictive (grade 3) as shown in Figure 1. Diastolic dysfunction was defined in 2 ways. First, as a binary variable (grades 2 or 3 only), and second, as a continuous variable using the E/e’ ratio as a measure of LV stiffness. While grades 0 and 1 were considered to reflect normal left atrial pressure, grades 2 (pseudonormal) and 3 (restrictive filling) reflected elevated left atrial
EFT VENTRICULAR DIASTOLIC DYSFUNCTION (LVDD) refers to an abnormality of diastolic distensibility, filling, or relaxation and is an important determinant of cardiovascular function. Indeed, LVDD is the predominant abnormality in patients with heart failure with preserved ejection fraction (HFpEF), which comprises approximately half of the patients with heart failure.1,2 In contrast to heart failure with reduced ejection fraction, the incidence of HFpEF has increased over the years and if these trends persist, HFpEF may become the most common type of heart failure in the near future.1,3 Studies indicate that women are at higher risk for developing HFpEF compared to men.4 Investigators have also shown that gender and age-gender interaction may play a role in the pathogenesis of LVDD.5,6 Possible mechanisms are changes in the hormonal milieu that accompany normal aging, which subsequently contribute to the development of hypertension and left ventricular hypertrophy and predispose women to a higher risk of LVDD.7 Left ventricular diastolic dysfunction also is common in patients with ischemic heart disease and has been shown to be associated with long-term adverse outcomes in patients undergoing coronary artery bypass grafting (CABG) surgery.8 Similar to their risk for developing HFpEF, women also have a higher risk for morbidity and mortality following CABG than their male counterparts.9 However, there are limited perioperative data on the association among gender, age, and LVDD in cardiac surgery patients. The authors, therefore, conducted a study to determine the association between gender and LVDD in the setting of CABG surgery. The authors tested the hypothesis that females undergoing CABG surgery are more likely than male patients to present with LVDD. The authors further hypothesized that age and gender interact to influence this risk. METHODS With approval from the authors’ institutional review board, detailed clinical and intraoperative transesophageal echocardiographic data were analyzed in a cohort of 897 adult patients undergoing elective, isolated CABG surgery with CPB at Duke University Medical Center in whom
KEY WORDS: coronary artery disease, diastole, surgery, women, aging, diastolic dysfunction
From the Departments of *Anesthesiology, and †Surgery, Duke University Medical Center, Durham, NC; and ‡Department of Anesthesiology, Perioperative, and Pain Medicine, Harvard Medical School, Boston, MA. 1 Both of the authors contributed to this work equally. Address reprint requests to Madhav Swaminathan, MD, Box 3094, DUMC Anesthesiology, 2301 Erwin Road, Durham, NC 27710. E-mail: [email protected] © 2014 Elsevier Inc. All rights reserved. 1053-0770/2601-0001$36.00/0 http://dx.doi.org/10.1053/j.jvca.2013.11.010
Journal of Cardiothoracic and Vascular Anesthesia, Vol ], No ] (Month), 2014: pp ]]]–]]]
1
2
DIASTOLIC DYSFUNCTION IN CARDIAC SURGERY PATIENTS
Fig 1. Algorithm used for grading diastolic dysfunction. Abbreviations: TDI, tissue Doppler imaging.
pressure and were therefore categorized together as LVDD for the binary outcome. The E/e’ ratio is an established marker of abnormal LV filling pressures in patients with cardiovascular disease.10 Demographic and procedural variables were compared between the 2 gender groups using analysis of variance (ANOVA) and the Wilcoxon rank-sum tests for continuous variables and Chi square or Fischer’s exact test for categoric variables as appropriate. For the authors’ primary hypothesis, a multivariate logistic regression model was developed with LVDD as the dependent binary variable (grades 2 or 3 only) and female gender as the principal independent variable of interest. A second multivariate linear regression model was developed to test the independent association between female gender and E/e’. In both models, the authors adjusted for age, congestive heart failure, weight, diabetes, hypertension, LV ejection fraction, and cardiopulmonary bypass time. For the secondary hypothesis, the authors investigated the interaction between age and gender and their association with diastolic dysfunction by adding the ‘age-gender’ interaction term to both regression models. For further descriptive purposes, the authors employed a restricted cubic spline analysis that allowed them to explore the possibility of nonlinear relationships between age and the E/e’ ratio. The inclusion of the interaction term (age * gender) allows the slopes of the lines to vary between males and females. The spline analysis allows the smoothed fitted line to vary in slope at predetermined points and is constrained to be linear in the tails. The location of the splines, or knots (the points at which the slope of the line is free to vary) were chosen based on recommended quantiles described by Harrell.11 Predicted values were produced from the multivariate model, which was adjusted for the same covariables as in the primary analysis. The predictors of interest for the purpose of this analysis were age and gender and their interaction. A p value of less than 0.05 was considered statistically significant. All analyses were conducted using SAS statistical software (Version 9.1.3, SAS Inc, Cary, NC). RESULTS
Demographic and procedural variables for the cohort of 895 patients are described in Table 1. There were 158 patients (18%) with normal diastolic function while 309 (34%) had grade 1, 207 (30%) had grade 2, and 161 (18%) patients had grade 3 LVDD. Statistically significant differences between the two genders were noted in several variables as shown in table 1. A greater proportion of females were Caucasian (27 vs. 16%). Female patients were also older (66 vs. 63 years),
had higher rates of diabetes (44 vs. 32%), and chronic obstructive pulmonary disease (18 vs. 11%) compared to male patients. In addition, females had a lower average body weight (83 vs. 92 kg), and a lower average left ventricular ejection fraction (50 vs. 53%). Significant differences in diastolic function variables were also noted. Female patients had a higher average baseline E/e’ ratio (11.5 v 8.3) and a correspondingly higher percentage of cases with grades 2 or 3 LVDD (68 v 41%), indicating worse baseline diastolic function in this subset of patients. The authors’ primary hypothesis was confirmed in both multivariable models that defined LVDD as a binary variable and LV diastolic stiffness (E/e’) as a continuous variable. Female patients were more likely to present with LVDD defined as a binary variable (odds ratio ¼ 2.73; 95% confidence interval ¼ 1.9-3.8; p o 0.0001) (Table 2). Female gender remained significantly associated with E/e’ in the linear regression model (parameter estimate ¼ 2.78; p o 0.0001) (Table 3). Other significant co-variables were age, diabetes and low ejection fraction. For the secondary hypothesis, the addition of the age-gender interaction term to the logistic regression model showed marginal significance. (Linear regression model showed that there was a significant association between age-gender and E/e’ (parameter estimate ¼ 0.08; p ¼ 0.01). The results of the restricted cubic spline analysis of the interaction between age and gender and their association with diastolic dysfunction are shown in Figure 2. The specific ages that are knots (the points at which the lines are free to vary) in this analysis are 80, 72, 65, 57, and 47. The risk in diastolic dysfunction increases in slope between 60 and 70 for both males and females but is steeper in females and does not reach a plateau as it does in males.
Table 1. Demographic and Operative Characteristics
Variables
Age (years) Caucasian (%) Congestive heart failure (%) Hypertension, n (%) Diabetes mellitus (%) Weight (kg) Left ventricular ejection fraction (%) Preoperative creatinine (mg/dL) COPD (%) Previous MI (%) EuroSCORE Cardiopulmonary bypass time (min) Aortic cross-clamp time (min) Baseline E/e’ ratio LVDD (grade 2 or 3) (%)
Females (n ¼ 236)
Males (n ¼ 661)
p Value
66.6 (10.0) 26.7 18.5 77.6 44.0 83.4 (23.2) 49.9 (14.9)
63.3 (10.2) 15.8 17.6 74.8 32.2 92.1 (20.8) 53.3 (13.6)
o0.0001 0.0002 0.74 0.40 0.001 o0.0001 0.02
1.2 (1.1) 18.1 18.7 9.2 (3.0) 128 (61)
1.3 (1.2) 11.3 39.7 8.1 (3.3) 124 (42)
0.33 0.008 0.12 o0.0001 0.38
72 (25) 11.5 (5.8) 67.7
71 (23) 8.3 (4.1) 40.9
0.90 o0.0001 o0.0001
NOTE. Values are expressed as mean with standard deviation in parentheses unless specified otherwise. The p values are Wilcoxon rank sums or 2-Tailed Fisher’s exact test as appropriate. Abbreviations: COPD, chronic obstructive pulmonary disease; LVDD, left ventricular diastolic dysfunction; MI, myocardial infarction.
3
FERREIRA ET AL
Table 2. Multivariate Logistic Regression Model of Variables Independently Associated with Diastolic Dysfunction Wald 95% Variable
Parameter Estimate
Female gender Age CHF Weight COPD Diabetes Hypertension Ejection fraction CPB time
Standard Error
Odds Ratio
Confidence Limits
p Value
1.01
0.17
2.73
1.94-3.84
o0.0001
0.05 0.26 0.001 0.43 0.37 0.08 0.02
0.01 0.20 0.003 0.22 0.16 0.18 0.01
1.05 1.30 1.00 1.53 1.45 0.93 0.98
1.04-1.07 0.87-1.92 0.99-1.00 0.99-2.37 1.06-1.99 0.66-1.31 0.97-0.99
o0.0001 0.2 0.76 0.05 0.02 0.67 0.0039
0.003
0.002
1.00
1.00-1.01
0.05
NOTE. Diastolic dysfunction defined as a binary variable (only grades 2 or 3; see text for details). Abbreviations: CHF, congestive heart failure; CPB, cardiopulmonary bypass; COPD, chronic obstructive pulmonary disease.
DISCUSSION
The authors confirmed their primary hypothesis that women undergoing cardiac surgery are more likely to present with LVDD compared to men. They also found that older women are at a higher risk for LVDD. These findings add clinical significance to data from animal models that report a mechanistic link between female gender and impaired diastolic function. The authors’ study also suggests that aging may play a greater role in the risk of LVDD in women compared to men. Strategies for management of perioperative LVDD should consider the additional risk among elderly female patients. Left ventricular diastolic dysfunction is an ubiquitous finding in patients with heart failure, regardless of systolic function, and is a strong predictor of cardiovascular morbidity and mortality.12 In patients presenting for cardiovascular surgery, echocardiographically defined LVDD is a prognostic marker of adverse outcome.8,13,14 The association between female gender and cardiovascular risk has been extensively investigated, and studies suggest a complex relationship. A report from the Society of Thoracic Surgeons found that while Table 3. Multivariate Linear Regression Model of Variables Independently Associated with Left Ventricular Diastolic Stiffness Variable
Intercept Female gender Age CPB time CHF Weight COPD Diabetes Hypertension Ejection fraction
Parameter Estimate
Standard Error
t Value
p value
5.16 2.78 0.10 0.004 0.17 0.006 0.99 0.66 0.16 0.06
1.48 0.35 0.02 0.003 0.41 0.01 0.45 0.33 0.36 0.013
3.50 7.88 6.46 1.36 0.40 0.83 2.19 2.00 0.45 4.82
0.0005 o0.0001 o0.0001 0.17 0.69 0.41 0.0289 0.05 0.65 o0.0001
NOTE. Diastolic dysfunction defined as a continuous variable binary variable (E/e’ ratio; see text for details). Abbreviations: CHF, congestive heart failure; CPB, cardiopulmonary bypass; COPD, chronic obstructive pulmonary disease.
Fig 2. Cubic splines analysis demonstrating the age-gender interaction and risk of diastolic dysfunction. There is an increase in slope between the ages of 60 and 70 for both males and females, but it is steeper in females compared to males.
women have a higher mortality risk after CABG surgery, the difference in operative mortality rates between genders is absent in the high-risk sub-categories.15 Another cardiac surgery registry study reported disproportionately higher mortality rates among women compared to men undergoing CABG surgery. However, this risk was particularly significant among younger women, and the effect decreased with advancing age.16 An Australian cardiac surgery database study reported higher mortality rates among women undergoing CABG surgery that were likely attributable to worse risk profiles.17 The same group also reported no gender differences in outcomes following CABG with concurrent aortic valve surgery.18 Koch and colleagues found that the higher prevalence of comorbid conditions and not gender primarily influenced the increased risk of postoperative mortality among women undergoing CABG surgery.19 The higher perioperative risk among women may therefore be influenced more by gender differences in severity of comorbidities such as hypertension and heart failure. Heart failure, especially HFpEF, is more prevalent among older women.2,20 Population studies also suggest gender differences in incidence of arterial stiffness and hypertension.21,22 Animal models of cardiovascular disease support these observations and offer some insights into possible mechanisms. In ovariectomized rats, Groban and colleagues found that the altered hormonal environment was associated with an increase in the severity of diastolic stiffness.23 They also showed that the pathogenesis of diastolic dysfunction in female rats involved alterations in the myocardial interstitium that were strongly linked with metabolic changes resulting from estrogen depletion, such as shifts in the neuronal nitric oxide synthase pathway.24 They suggested that post menopausal women are at higher risk of diastolic dysfunction due to the metabolic consequences of estrogen loss on the myocardium. Taken
4
DIASTOLIC DYSFUNCTION IN CARDIAC SURGERY PATIENTS
together, these studies suggest that women are at higher risk of cardiovascular events compared to men due to structural differences in vascular and myocardial architecture. These differences appear to be associated, in part, with changes in the hormonal milieu that accompany natural ageing. The authors’ finding of higher incidence of diastolic dysfunction among women presenting for cardiac surgery provides clinical evidence to support mechanisms suggested by preclinical data. Although the authors found an interaction of age and female gender with diastolic stiffness only in the validation cohort, they feel that these findings may reflect the effect of post menopausal hormonal changes on the myocardium. Other factors such as diabetes and low ejection fraction that were significant independent variables in their regression analyses confirmed previous observations of their known association with LVDD.25,26 The authors’ study is strengthened by confirmation of their hypotheses in a large study population with well-defined echocardiographic and outcomes parameters. However, there are also limitations with respect to study design and data collection. The data was retrospectively collected, and the quality of data reflects the inherent limitations of this approach. The authors did not have hemodynamic validation of elevated filling pressures as this data was not collected. However, they are confident that the metrics used for defining diastolic dysfunction and LV stiffness previously have been validated and reflect filling pressures with reasonable accuracy in all their patients. Although the simplified algorithm they used for categorizing
LVDD was developed retrospectively and has not been prospectively validated, it has been endorsed by the American Society of Echocardiography in their recent guidelines for a comprehensive TEE examination.27 In the absence of any other algorithm validated against adverse outcomes in the perioperative CABG surgery population, the authors feel justified in using the simplified algorithm for their study. It also could be argued that hormone replacement therapy in postmenopausal patients could have altered the incidence of diastolic dysfunction and confounded their results. However, estrogen-based therapies have not consistently been shown to mitigate cardiovascular risk.28,29 Therefore, the authors feel that their findings would not be substantially different if they included data on preoperative hormone therapies in their patients. In summary, the authors found a significantly higher incidence of diastolic dysfunction among women presenting for cardiac surgery compared to men. They also noted that the gender difference in diastolic dysfunction was also influenced by age, a finding supported by pre-clinical disease models. Their findings imply that the higher incidence of LVDD among women should alert the perioperative team to consider this cohort of patients at higher risk for the adverse outcomes known to be associated with LVDD and tailor cardiovascular management and postoperative surveillance appropriately. Therapies aimed at amelioration of diastolic dysfunction should additionally consider the higher risk in females, especially within the older subset of the patient population.
REFERENCES 1. Owan TE, Hodge DO, Herges RM, et al: Trends in prevalence and outcome of heart failure with preserved ejection fraction. N Engl J Med 355:251-259, 2006 2. Redfield MM, Jacobsen SJ, Burnett JC Jr, et al: Burden of systolic and diastolic ventricular dysfunction in the community: Appreciating the scope of the heart failure epidemic. JAMA 289:194-202, 2003 3. Paulus WJ, Tschope C, Sanderson JE, et al: How to diagnose diastolic heart failure: A consensus statement on the diagnosis of heart failure with normal left ventricular ejection fraction by the Heart Failure and Echocardiography Associations of the European Society of Cardiology. Eur Heart J 28:2539-2550, 2007 4. Lee DS, Gona P, Vasan RS, et al: Relation of disease pathogenesis and risk factors to heart failure with preserved or reduced ejection fraction: Insights from the Framingham Heart Study of the National Heart, Lung, and Blood Institute. Circulation 119:3070-3077, 2009 5. Wang H, Jessup JA, Lin MS, et al: Activation of GPR30 attenuates diastolic dysfunction and left ventricle remodelling in oophorectomized mRen2.Lewis rats. Cardiovasc Res 94:96-104, 2012 6. Jessup JA, Zhang L, Presley TD, et al: Tetrahydrobiopterin restores diastolic function and attenuates superoxide production in ovariectomized mRen2.Lewis rats. Endocrinology 152:2428-2436, 2011 7. Agabiti-Rosei E, Muiesan ML: Left ventricular hypertrophy and heart failure in women. J Hypertens Suppl 20:S34-S38, 2002 8. Swaminathan M, Nicoara A, Phillips-Bute BG, et al: Utility of a simple algorithm to grade diastolic dysfunction and predict outcome after coronary artery bypass graft surgery. Ann Thorac Surg 91: 1844-1850, 2011 9. Hannan EL, Racz MJ, Walford G, et al: Predictors of readmission for complications of coronary artery bypass graft surgery. JAMA 290: 773-780, 2003
10. Nagueh SF, Appleton CP, Gillebert TC, et al: Recommendations for the evaluation of left ventricular diastolic function by echocardiography. J Am Soc Echocardiogr 22:107-133, 2009 11. Harrell FE Jr, Lee KL, Pollock BG: Regression models in clinical studies: Determining relationships between predictors and response. J Natl Cancer Inst 80:1198-1202, 1988 12. Zile MR, Brutsaert DL: New concepts in diastolic dysfunction and diastolic heart failure: Part I: Diagnosis, prognosis, and measurements of diastolic function. Circulation 105:1387-1393, 2002 13. Matyal R, Hess PE, Subramaniam B, et al: Perioperative diastolic dysfunction during vascular surgery and its association with postoperative outcome. J Vasc Surg 50:70-76, 2009 14. Flu WJ, van Kuijk JP, Hoeks SE, et al: Prognostic implications of asymptomatic left ventricular dysfunction in patients undergoing vascular surgery. Anesthesiology 112:1316-1324, 2010 15. Edwards FH, Carey JS, Grover FL, et al: Impact of gender on coronary bypass operative mortality. Ann Thorac Surg 66: 125-131, 1998 16. Vaccarino V, Abramson JL, Veledar E, et al: Sex differences in hospital mortality after coronary artery bypass surgery: Evidence for a higher mortality in younger women. Circulation 105:1176-1181, 2002 17. Saxena A, Dinh D, Smith JA, et al: Sex differences in outcomes following isolated coronary artery bypass graft surgery in Australian patients: Analysis of the Australasian Society of Cardiac and Thoracic Surgeons cardiac surgery database. Eur J Cardiothorac Surg 41:755-762, 2012 18. Saxena A, Poh CL, Dinh DT, et al: Does patient gender affect outcomes after concomitant coronary artery bypass graft and aortic valve replacement? An Australian Society of Cardiac and Thoracic Surgeons Database study. Cardiology 119:116-123, 2011
FERREIRA ET AL
19. Koch CG, Weng YS, Zhou SX, et al: Prevalence of risk factors, and not gender per se, determines short- and long-term survival after coronary artery bypass surgery. J Cardiothorac Vasc Anesth 17:585-593, 2003 20. Masoudi FA, Havranek EP, Smith G, et al: Gender, age, and heart failure with preserved left ventricular systolic function. J Am Coll Cardiol 41:217-223, 2003 21. Russo C, Jin Z, Palmieri V, et al: Arterial stiffness and wave reflection: sex differences and relationship with left ventricular diastolic function. Hypertension 60:362-368, 2012 22. Deschepper CF, Llamas B: Hypertensive cardiac remodeling in males and females: From the bench to the bedside. Hypertension 49: 401-407, 2007 23. Groban L, Yamaleyeva LM, Westwood BM, et al: Progressive diastolic dysfunction in the female mRen(2). Lewis rat: Influence of salt and ovarian hormones. The journals of gerontology Series A, Biological sciences and medical sciences 63:3-11, 2008 24. Jessup JA, Zhang L, Chen AF, et al: Neuronal nitric oxide synthase inhibition improves diastolic function and reduces oxidative stress in ovariectomized mRen2.Lewis rats. Menopause 18:698-708, 2011
5
25. Poulsen MK, Henriksen JE, Dahl J, et al: Left ventricular diastolic function in type 2 diabetes mellitus: Prevalence and association with myocardial and vascular disease. Circ Cardiovasc Imaging 3: 24-31, 2010 26. Zile MR, Baicu CF, Gaasch WH: Diastolic heart failure— abnormalities in active relaxation and passive stiffness of the left ventricle. N Engl J Med 350:1953-1959, 2004 27. Hahn RT, Abraham T, Adams MS, et al: Guidelines for performing a comprehensive transesophageal echocardiographic examination: Recommendations from the American Society of Echocardiography and the Society of Cardiovascular Anesthesiologists. J Am Soc Echocardiogr 26:921-964, 2013 28. Rossouw JE, Anderson GL, Prentice RL, et al: Risks and benefits of estrogen plus progestin in healthy postmenopausal women: Principal results from the Women’s Health Initiative randomized controlled trial. JAMA 288:321-333, 2002 29. Wells G, Herrington DM: The Heart and Estrogen/Progestin Replacement Study: What have we learned, and what questions remain? Drugs Aging 15:419-422, 1999
