(Circulation. 1996;94:2472-2478.)
© 1996 American Heart Association, Inc.
Articles |
the Division of Cardiovascular Diseases and Internal Medicine (E.K., M.E.-S., A.J.T., J.B.S.), Section of Cardiovascular Surgery (C.J.M.), and Section of Biostatistics (K.R.B.), Mayo Clinic and Mayo Foundation, Rochester, Minn.
| Abstract |
|---|
|
|
|---|
Methods and Results Baseline characteristics and postoperative outcomes were compared between 51 women and 198 men undergoing surgery for isolated aortic regurgitation between 1980 and 1989. Compared with men, women had surgery rarely for severe left ventricular enlargement (systolic diameter
55 mm in 11% versus 27%, P=.031; diastolic diameter
80 mm in 0% versus 16%, P<.0001) and more often for class III to IV symptoms (59% versus 32%, P<.0001). Operative mortalities were similar in women and men (3.9% and 4.5%, respectively). Among operative survivors, 10-year survival was worse for women than for men (39±9% versus 72±4%, P=.0002) and, in contrast with men, was worse than expected for women (P<.0001). Independent predictors of late survival were different for men (age and ejection fraction) and women (age and concomitant coronary bypass grafting). By multivariate analysis, female sex was an independent predictor of worse late survival (adjusted relative risk, 1.80; 95% CI, 1.04 to 3.11).
Conclusions The generalization to women of the unadjusted left ventricular diameter surgical criteria established in men results in irrelevant criteria almost never reached in women, who often undergo surgery after developing severe symptoms. After surgery, women exhibit an excess late mortality, suggesting that surgical correction of aortic regurgitation should be considered at an earlier stage in women.
Key Words: aorta women prognosis regurgitation surgery
| Introduction |
|---|
|
|
|---|
Although very little information is available, recent data suggest that in aortic valve disease, sex may influence LV hypertrophy,14 15 response to AVR,3 operative mortality,16 and long-term survival.17 However, these results were not confirmed in multivariate analyses and were not focused on AR. Nevertheless, they raise the question about the generalizability to women of surgical criteria developed in men and based on LV size and function.7 8 9 10 12 13 Because of the small number of women with AR in previous series,2 3 4 5 6 7 8 9 10 12 13 this concern could not be addressed, and the outcome of AR surgery in women has not been analyzed.
Therefore, we examined a large cohort of women undergoing AVR for severe isolated AR and hypothesized (1) that women with AR constitute a unique patient population with different preoperative clinical characteristics and different postoperative outcomes compared with men and (2) that established indications for AVR based on LV size and function may not be valid in women.
| Methods |
|---|
|
|
|---|
The clinical, surgical, and echocardiographic variables that were compared between men and women and those that were used as potential predictors of outcome are presented in Table 1
. Echocardiography19 20 21 was performed 32±41 days before AVR and at least 6 months and closer to 1 year after surgery.5
|
Follow-up was complete in 98% of patients up to 1994 or death. Cardiac transplantation for terminal heart failure was performed in two male patients and was combined with mortality as an equivalent end point.
Statistical Methods
Group statistics were expressed as mean±SD. Group comparisons (women versus men) were carried out with a standard t test or
2 test. Outcome end points, evaluated separately in men and women, included overall survival, late survival, and postoperative LV ejection fraction. Operative mortality was compared between sexes by the
2 test. Analysis of overall and late survival used the Kaplan-Meier method (group comparison by the two-sample log-rank test) and was adjusted by the Cox proportional-hazards method. Because expected survival is markedly different in men and women, the survival curves were compared with the expected survival of age- and sex-matched populations, as defined by the US Census Bureau, by use of the one-sample log-rank test and were further analyzed by a generalization of a model described by Breslow et al.22 Multivariate analyses were performed with separate models for clinical, echocardiographic, and surgical variables independently in men and women. Significant variables from these analyses were then combined to determine the independent predictors of outcome for each sex. A common model grouping all significant predictors of outcome for either men or women was then performed, adding sex to analyze its independent predictive power for the end point. In addition, models including sex and interactions with the other predictive variables were fitted. The analyses were repeated with the substitution of LV diameters corrected for body surface area. For entry criteria in modeling for men and women, a P<.10 for men or <.15 for women was required. P<.10 was used as a cutoff in combined models. In the final analysis, P<.05 was considered statistically significant.
| Results |
|---|
|
|
|---|
|
Although the LV ejection fraction was similar between the two groups, women had smaller LV end-systolic and end-diastolic dimensions than men (P<.0001) (Table 2
). Only 11% of women had an LV end-systolic diameter
55 mm, compared with 27% of men (P=.031), and no woman had an end-diastolic diameter
80 mm, compared with 18% of men (P<.0001). However, these differences were not related to less severe regurgitation and were totally abolished when adjusted for body surface area.
Overall Survival
Overall survival (operative and late deaths combined) at 5 and 10 years was 85±3% and 69±4%, respectively, in men and 72±6% and 38±8%, respectively, in women (P=.0008) (Fig 1
). In multivariate analysis, including age, ejection fraction, concomitant CABG, and presence of aortic aneurysm, female sex was a borderline predictor of worse outcome (P=.077), in addition to age (P=.0001). However, compared with age- and sex-matched reference populations, overall survival was significantly decreased in men (P=.002) and women (P<.0001) but was markedly decreased in women (representing, at 10 years, 52% of the expected survival, compared with 88% in men). This difference was confirmed in a multivariate proportional analysis adjusting for expected survival and including age, ejection fraction, concomitant CABG, and presence of aortic aneurysm in which female sex was associated with an excess mortality (adjusted RR [95% CI], 2.58 [1.51 to 4.44]).
|
To further analyze this excess mortality, (1) the outcome according to preoperative symptoms was examined because of the strong relationship between sex and symptoms (P<.0001). In the Cox proportional-hazards analysis combining men and women, NYHA class III or IV symptoms (P=.03) were independent predictors of worse overall survival, in addition to age (P=.0001). Women with class III or IV symptoms had a worse survival than those operated on with class I or II symptoms (at 8 years, 39±10% versus 75±10%, P=.005), suggesting that the large proportion of women operated on at a late stage with class III or IV symptoms was mainly responsible for the excess mortality observed in women. (2) The impact of sex on operative and late mortality was analyzed separately.
Operative Mortality
The operative mortality was slightly but not significantly lower in women (2 of 51, 3.9%) than in men (9 of 198, 4.5%, P=.70). Because of the small number of women affected by this end point, comparative analysis of the determinants of operative death was not performed.
Late Survival
At 5 and 10 years, late survival was 89±2% and 72±4%, respectively, in men and 75±6% and 39±9%, respectively, in women (P=.0002). The causes of late death were similar, but women tended to experience fatal rupture or dissection of the aorta more frequently than men (17% versus 2% of the causes of death; P=.04). Compared with expected, survival was worse in women (86% and 53% of expected survival at 5 and 10 years, respectively; P<.0001) but not different in men (99% and 92% of expected survival at 5 and 10 years, respectively; P=.074) (Figs 2
and 3).
This difference was confirmed in a multivariate analysis that adjusted for expected survival and included age, ejection fraction, concomitant CABG, and presence of aortic aneurysm, demonstrating an excess mortality in women (RR [95% CI], 3.84 [2.12 to 6.96], P=.0001).
|
|
Independent predictors of late survival were different in men (age, P=.0001; preoperative ejection fraction, P=.0087) and women (age, P=.015; CABG, P=.0080). These differences were confirmed by the presence of significant interactions between female sex and preoperative ejection fraction (P=.043) and CABG (P=.027) in the prediction of survival. In the standard multivariate analysis combining men and women, controlling for age, preoperative ejection fraction, and CABG, female sex remained a significant independent predictor of late mortality (P=.037; RR [95% CI], 1.80 [1.04 to 3.11]). LV dimensions corrected for body surface area were not predictive of late survival.
Excess mortality in women was confirmed in various subgroups examined. When stratified according to preoperative ejection fraction, men displayed significantly different late survivals (P=.0009), a phenomenon not observed in women. In fact, the late survival curves of men and women with preoperative ejection fraction <50% were similar, but women with a preoperative ejection fraction
50% displayed an excess late mortality compared with men (Fig 4
).
|
In patients with concomitant CABG compared with those without concomitant CABG, late survival tended to be worse in men (P=.052; RR [95% CI], 1.91 [0.98 to 3.69]), whereas in women, late survival was markedly worse (P=.0001; RR [95% CI], 5.54 [2.10 to 14.61]). However, late survival was worse in women whether CABG was performed (RR [95% CI], 7.45 [2.5 to 21.9]) or not performed (RR [95% CI], 2.1 [1.1 to 3.8]) (Fig 5
). When the small group (n=20) of women without CABG and without aortic aneurysm was analyzed, a trend for excess mortality was noted (RR, 1.64) compared with men but did not reach statistical significance because of the small size of the group.
|
Body surface area was smaller in women (Table 2
), but when forced in the multivariate model for late survival, it was not an independent predictor of outcome (P=.34) with or without sex in the model, whereas sex remained an independent predictor of survival (P=.037).
Postoperative Status of the Left Ventricle
Twenty-four women and 125 men underwent echocardiography at least 6 months after AVR (Table 3
). The difference in postoperative ejection fraction between sexes was not statistically significant.
|
By multivariate analysis, only preoperative ejection fraction was a significant independent predictor of postoperative LV function in both men (P=.0001) and women (P=.0009); sex was not a significant predictor, and no interaction was noted.
Significant regression was noted in LV systolic and diastolic dimensions after surgery in both men and women, and when corrected for body surface area, no significant differences were noted in the final LV dimensions between the two groups (Table 3
). Only 1 woman (5.6%) displayed an end-diastolic dimension
60 mm after surgery, compared with 34 men (27.5%) (P=.045).
| Discussion |
|---|
|
|
|---|
Preoperative Characteristics and Indication for Surgery
The timing of AVR for AR had not been evaluated specifically in women but traditionally relied on symptoms23 24 25 or indexes of LV size or function reportedly associated with poor outcome (end-systolic dimension
55 mm, LV end-diastolic dimension
80 mm, diminished ejection fraction7 8 9 10 12 13 ). In the present series, a striking difference between men and women was noted in the condition leading to the surgical indication: (1) Nearly two thirds of women experienced class III or IV dyspnea, whereas only one third of men experienced such symptoms. (2) LV dilatation reaching the above-mentioned criteria was uncommon in women: only 1 in 10 women attained a preoperative end-systolic dimension
55 mm, compared with nearly 1 in 3 men, and no woman achieved an end-diastolic dimension
80 mm. This undoubtedly was related to the fact that women have smaller body sizes. Nevertheless, it means that generalizing to women the LV diameter criteria for surgery established in men results in irrelevant criteria almost never observed in women. (3) In terms of preoperative LV function, no difference was noted between men and women in the present series.
Postoperative Outcome
Few reports have examined the influence of the sex of the patient on late outcome after surgical correction of AR. Although early series suggested that male sex was significantly associated with late mortality,1 that association was not reproduced in other studies.2 26 More recently, analysis of a large patient cohort undergoing AVR at our institution suggested that postoperative survival may be worse in women, although the finding was not confirmed by multivariate analysis.17 However, the population examined was pooled, with aortic stenosis representing the overwhelming majority of patients. Sex-related differences have been shown recently to be less prominent in aortic stenosis27 and, therefore, may have blunted the statistical result, mandating a specific analysis of women with AR.
Operative mortality rates in men and women were similar in the present study and representative of those reported in previous series.1 4 26 28 29 30 31 32 33 34 However, in analysis of long-term survival, women exhibited an excess age-adjusted mortality, even after stratification for significant predictors of outcome, in comparison with men. This excess mortality is sex-specific and not related to body surface area. This observation of an excess mortality in women late after aortic valve surgery is an essential observation, whatever its explanation may be, and should lead to a reappraisal of the surgical criteria used in women with AR.
Indeed, the explanation for this excess mortality is complex, in part because women represent a minority of patients with AR, which limits the extent of statistical analysis. Some factors do not appear to play a role. The late incidence of heart failure and the determinants of postoperative LV function were similar in men and women, and survival rates in patients with preoperative ejection fractions <50% were poor irrespective of sex. Thus, poor preoperative LV function was not a contributing factor to excess late mortality in women.
The excess mortality in women is in part due to factors unrelated to the standards used for timing valve replacement in women. First, women tended to experience a higher rate of fatal disruptions of the aorta, paralleling the higher preoperative prevalence of aortic root disease. Second, in the present series, concomitant CABG was a major determinant of survival35 in women and was associated with a marked excess risk of late death compared with men. Although conflicting data have been reported about the effect of sex on the outcome of coronary interventions,36 37 38 several studies have emphasized the worse prognosis of women undergoing CABG.39 40 In the present series, the RR of long-term death of women compared with men when CABG was associated with valve replacement was considerable, at 7.45. However, women had an increased mortality in the absence of CABG (RR, 2.1) compared with men, which suggests that other factors may be involved.
This excess mortality in women may also be related to the standards used in determining the timing of valve replacement. Note, women and men undergoing surgery at the same stage (ejection fraction <50%) had similar survival. But because most women failed to reach the accepted criteria of LV dilatation, they were more often referred to surgery after class III or IV symptoms developed. In the present study, such symptoms were associated with a higher risk for overall postoperative mortality and, therefore, are risk factors for excess mortality in women. Performance of valve replacement at a severely symptomatic stagebecause of the failure of women to reach the objective male-based unduly generalized criteria of LV size for AVRin our opinion represents a bias in the management of women with severe AR. This issue has not been addressed because in previous series, even with large numbers of patients, a small number of women,* seldom more than 30,47 were included, thus limiting sound conclusions about sex-related differences.
Clinical Implications for Women With Severe Chronic AR
The generalization to women of the unadjusted LV diameter surgical criteria established in men results in irrelevant criteria almost never reached by women.
Alternatively, LV diameters normalized to body surface area have been used but do not provide additional prognostic information.48 49 50 Currently, no specific recommendation for surgery can be based on those variables. The assessment of the severity of regurgitation based on LV dimensions is hazardous and, preferentially, should be based on quantification of regurgitation.
An ejection fraction <50% is associated with a similarly poor prognosis in men and women and, despite the controversy about the prognostic usefulness of LV variables,42 43 should, in our opinion, remain an indication for surgical correction of AR.5 8 10
Severe symptoms (class III or IV) should continue to be an indication for surgery in men and women.8 However, the facts that in women the indication for surgery is based mostly on symptoms and that an excess long-term mortality is noted suggest that even mild symptoms (class II) should lead to consideration of surgical correction of AR. The issue of performing surgery in patients with minimal symptoms and no LV dysfunction, whether female or male, deserves attention in future studies.
Study Limitations
The use of echocardiography to determine ventricular dimensions and function may be disputed. However, the methods used were stable, and measurements of LV dimensions were guided by two-dimensional echocardiography. Furthermore, similar echocardiographic data have been used in previous studies and, as such, have served as the basis for recommending the timing of AVR.7 8
Data from a referral center may introduce a referral bias. Because of the limited number of women with AR, a population-based study is not possible. In addition, not all patients were followed at our institution; therefore, the present series represents the routine cardiological practice.
The number of women in the present series was small and did not allow for analysis of specific predictors of end points such as operative mortality. Nevertheless, despite its size, this series represents the largest population of women with chronic severe AR reported and allowed the demonstration of significant sex-specific differences in patient characteristics and outcomes.
Conclusions
In women with AR, the generalization of the unadjusted LV diameter surgical criteria established in men results in irrelevant criteria almost never reached by women. Women are more likely to undergo surgery after developing severe symptoms and, after surgery, to show an independent excess long-term mortality. These results suggest that in women with AR, surgical correction should be considered at an earlier stage before severe symptoms develop.
| Selected Abbreviations and Acronyms |
|---|
|
| Acknowledgments |
|---|
| Footnotes |
|---|
*References 4, 7, 10, 12, 13, 16, 32, 33, 41-46.
Received March 4, 1996; revision received May 29, 1996; accepted June 16, 1996.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
M. Enriquez-Sarano, V. T. Nkomo, and H. Michelena Principles and Practice of Echocardiography in Cardiac Surgery Card. Surg. Adult, January 1, 2008; 3(2008): 315 - 348. [Full Text] |
||||
![]() |
A. I. Duncan, J. Lin, C. G. Koch, A. M. Gillinov, M. Xu, and N. J. Starr The Impact of Gender on In-Hospital Mortality and Morbidity After Isolated Aortic Valve Replacement Anesth. Analg., October 1, 2006; 103(4): 800 - 808. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. O. Bonow, B. A. Carabello, K. Chatterjee, A. C. de Leon Jr, D. P. Faxon, M. D. Freed, W. H. Gaasch, B. W. Lytle, R. A. Nishimura, P. T. O'Gara, et al. ACC/AHA 2006 Guidelines for the Management of Patients With Valvular Heart Disease: A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 1998 Guidelines for the Management of Patients With Valvular Heart Disease) Developed in Collaboration With the Society of Cardiovascular Anesthesiologists Endorsed by the Society for Cardiovascular Angiography and Interventions and the Society of Thoracic Surgeons J. Am. Coll. Cardiol., August 1, 2006; 48(3): e1 - e148. [Full Text] [PDF] |
||||
![]() |
R. O. Bonow, B. A. Carabello, K. Chatterjee, A. C. de Leon Jr, D. P. Faxon, M. D. Freed, W. H. Gaasch, B. W. Lytle, R. A. Nishimura, P. T. O'Gara, et al. ACC/AHA 2006 Practice Guidelines for the Management of Patients With Valvular Heart Disease: Executive Summary: A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 1998 Guidelines for the Management of Patients With Valvular Heart Disease) Developed in Collaboration With the Society of Cardiovascular Anesthesiologists Endorsed by the Society for Cardiovascular Angiography and Interventions and the Society of Thoracic Surgeons J. Am. Coll. Cardiol., August 1, 2006; 48(3): 598 - 675. [Full Text] [PDF] |
||||
![]() |
A. Evangelista, P. Tornos, A. Sambola, G. Permanyer-Miralda, and J. Soler-Soler Long-term vasodilator therapy in patients with severe aortic regurgitation. N. Engl. J. Med., September 29, 2005; 353(13): 1342 - 1349. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Enriquez-Sarano and A. J. Tajik Aortic Regurgitation N. Engl. J. Med., October 7, 2004; 351(15): 1539 - 1546. [Full Text] [PDF] |
||||
![]() |
B. A. Carabello Is it ever too late to operate on the patient with valvular heart disease? J. Am. Coll. Cardiol., July 21, 2004; 44(2): 376 - 383. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. J. Carpenter and M. Camacho Valvular heart disease in women: The surgical perspective J. Thorac. Cardiovasc. Surg., January 1, 2004; 127(1): 4 - 6. [Full Text] [PDF] |
||||
![]() |
Y. Misawa, K. Fuse, H. P. Chaliki, D. Mohty, J.-F. Avierinos, A. J. Tajik, M. Enriquez-Sarano, C. G. Scott, and H. V. Schaff Left Ventricular Remodeling After Valve Replacement in Patients With Isolated Aortic Regurgitation * Response Circulation, June 10, 2003; 107 (22): e208 - e209. [Full Text] [PDF] |
||||
![]() |
W. H. Gaasch and E. C. Schick Symptoms and left ventricular size and function in patients with chronic aortic regurgitation J. Am. Coll. Cardiol., April 16, 2003; 41(8): 1325 - 1328. [Full Text] [PDF] |
||||
![]() |
B. Iung, C. Gohlke-Barwolf, P. Tornos, C. Tribouilloy, R. Hall, E. Butchart, and A. Vahanian Recommendations on the management of the asymptomatic patient with valvular heart disease Eur. Heart J., August 2, 2002; 23(16): 1253 - 1266. [PDF] |
||||
![]() |
D Vinereanu, A A Ionescu, and A G Fraser Assessment of left ventricular long axis contraction can detect early myocardial dysfunction in asymptomatic patients with severe aortic regurgitation Heart, January 1, 2001; 85(1): 30 - 36. [Abstract] [Full Text] |
||||
![]() |
M. L. McDonald, N. G. Smedira, E. H. Blackstone, R. A. Grimm, B. W. Lytle, and D. M. Cosgrove REDUCED SURVIVAL IN WOMEN AFTER VALVE SURGERY FOR AORTIC REGURGITATION: EFFECT OF AORTIC ENLARGEMENT AND LATE AORTIC RUPTURE J. Thorac. Cardiovasc. Surg., June 1, 2000; 119(6): 1205 - 1215. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. S. Dujardin, M. Enriquez-Sarano, H. V. Schaff, K. R. Bailey, J. B. Seward, and A. J. Tajik Mortality and Morbidity of Aortic Regurgitation in Clinical Practice : A Long-Term Follow-Up Study Circulation, April 13, 1999; 99(14): 1851 - 1857. [Abstract] [Full Text] [PDF] |
||||
![]() |
Aortic Regurgitation: Women May Need Different Criteria for Surgery Journal Watch Women's Health, December 1, 1996; 1996(1201): 5 - 5. [Full Text] |
||||
![]() |
DO WOMEN WITH AORTIC REGURGITATION NEED DIFFERENT CRITERIA FOR SURGERY? Journal Watch (General), November 22, 1996; 1996(1122): 2 - 2. [Full Text] |
||||
![]() |
B. A. Carabello Aortic Regurgitation in Women: Does the Measuring Stick Need a Change? Circulation, November 15, 1996; 94(10): 2355 - 2357. [Full Text] |
||||
| |||||||||||||||||||||||||