(Stroke. 2000;31:2628.)
© 2000 American Heart Association, Inc.
Original Contributions |
Predictors of Cerebrovascular Events and Death Among Patients With Valvular Heart Disease
A Population-Based Study
From the Division of Cerebrovascular Diseases (G.W.P., D.O.W.), Division of Cardiovascular Diseases and Internal Medicine (B.K.K.), and Department of Health Sciences Research (J.P.W., J.D.S., W.M.O., D.O.W.), Mayo Clinic and Mayo Foundation, Rochester, Minn.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Background and Purpose—There is little population-based information on cerebrovascular events and survival among valvular heart disease patients. We used the Kaplan-Meier product-limit method and the Cox proportional hazards model to determine rates and predictors of cerebrovascular events and death among valve disease patients.
Methods—This population-based historical cohort study in Olmsted County, Minnesota, reviewed residents with a first echocardiographic diagnosis of mitral stenosis (n=19), mitral regurgitation (n=528), aortic stenosis (n=140), and aortic regurgitation (n=106) between 1985 and 1992.
Results—During 2694 person-years of follow-up, 98 patients developed cerebrovascular events and 356 died. Compared with expected numbers, these observations are significantly elevated, with standardized morbidity ratio of 3.2 (95% CI, 2.6 to 3.8) and 2.5 (95% CI, 2.2 to 2.7), respectively. Independent predictors of cerebrovascular events were age, atrial fibrillation, and severe aortic stenosis. The risk ratio of severe aortic stenosis was 3.5 (95% CI, 1.4 to 8.6), with atrial fibrillation conferring greater risk at younger age. Predictors of death were age, sex, cerebrovascular events, ischemic heart disease, and congestive heart failure, the greatest risk being among those with both congestive heart failure and cerebrovascular events (risk ratio=8.8; 95% CI, 5.8 to 13.4). Valve disease type and severity were not independent determinants of death.
Conclusions—The risk of cerebrovascular events and death among patients with valve disease remains high. Age, atrial fibrillation, and severe aortic stenosis are independent predictors of cerebrovascular events, and age, sex, cerebrovascular events, congestive heart failure, and ischemic heart disease are independent predictors of death in these patients.
Key Words: atrial fibrillation • cerebral embolism and thrombosis • echocardiography • heart valve diseases • stroke
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Approximately 50 000 US residents with valvular heart disease develop a first stroke each year, and valve disease is an important risk factor for stroke recurrence.1 2 3 Despite the decline of rheumatic fever, valve disease will remain a frequent cause of stroke in the United States and especially abroad.4 5 6 Yet there is little modern information on the risk of cerebrovascular events among patients with hemodynamically significant mitral and aortic stenosis and regurgitation. Many previous studies appeared before advances in medical and surgical treatment and introduction of echocardiography as an accurate noninvasive means of diagnosing valve disease. Moreover, earlier longitudinal studies did not use multivariate analyses to identify independent determinants of morbidity and mortality and were derived largely from observations made on patients referred to tertiary care centers, an important source of bias.7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 We undertook a large population-based historical cohort study of all residents of Olmsted County, Minnesota, who had a first 2-dimensional color Doppler echocardiographic diagnosis of hemodynamically significant mitral or aortic stenosis or regurgitation made between 1985 and 1992 to estimate rates and model determinants of the development of cerebrovascular events and to quantify the influence of cerebrovascular events on survival among these patients.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This study used the population-based data resources of the Rochester Epidemiology Project medical records linkage system23 and the Mayo Echocardiography Laboratory database. Virtually all medical care in the community is supplied by the Mayo Clinic and its 2 hospitals or by Olmsted Medical Center and its hospital. All medical diagnoses made for a resident of Olmsted County are entered on a master sheet in the patient’s medical record, which is then entered into a central computer index. Funding of the Rochester Epidemiology Project >30 years ago permitted expanding this index to include other medical practices from surrounding communities, the University of Minnesota, and the Veterans Administration Hospital in Minneapolis, where Olmsted county residents may have received medical care. The medical record contains all inpatient and outpatient data, including information regarding emergency department visits, nursing home care, autopsies, and death certificates. All echocardiograms obtained at medical care facilities in Olmsted County, Minnesota, have been performed and interpreted by a single unified Mayo Clinic Echocardiography Laboratory since 1970. The echocardiography database includes detailed information on >300 000 echocardiographic examinations performed at the Mayo Clinic since 1975. From 1985 onward, virtually all patients studied at the Mayo Clinic had 2-dimensional echocardiograms, and >90% had Doppler and color-flow imaging. Approximately 1 in 50 residents of Olmsted County, Minnesota, underwent echocardiography each year for differential diagnostic assessment of cardiovascular-related conditions during the period of this study.24
Verification of residence was based on information obtained from city and county directories or earlier medical records. Patients were included only if their residence was established in Olmsted County, Minnesota, at least 1 year before their first echocardiographic diagnosis of valve disease. Those who apparently moved to the county to facilitate diagnosis or treatment were excluded. Population-based studies in Rochester and Olmsted County are approved by the Mayo and Olmsted Medical Center Institutional Review Boards.
The cohort in this study is defined as all residents of Olmsted County, Minnesota, who had a first-time 2-dimensional color Doppler echocardiographic diagnosis of moderate or severe mitral stenosis, mitral regurgitation, aortic stenosis, aortic regurgitation, or any combination thereof made from January 1, 1985, through December 31, 1992, who did not have a history of cerebrovascular events (stroke, transient ischemic attack, or amaurosis fugax) before or at the time of the echocardiogram. We did not include patients diagnosed with valve disease by 2-dimensional color Doppler echocardiography before 1985 to avoid bias that could accrue to differential selection of patients for referral for this type of echocardiogram during earlier years before the use of this technology was routine in our laboratory. Patients with a history of cerebrovascular events were excluded from the cohort because the aim of the study was to determine rates of occurrence of incident cerebrovascular events after diagnosis of valve disease and to assess the impact of the development of cerebrovascular events on survival among these patients. Such an approach also limits bias that could accrue to inclusion of patients with a history of cerebrovascular disease in the years before the period of the study who had characteristics that guaranteed survival. Patients with only trivial or mild mitral or aortic stenosis or regurgitation were excluded from this study.
Echocardiographic diagnosis of valve disease was specifically chosen as the criterion for diagnosis and inclusion in the cohort because of the superiority of this technology over the clinical examination in accurately identifying and quantifying the type and severity of valvular disease.25 26 27 28 29 However, patients with a nonechocardiographic diagnosis of valve disease before the date of the echocardiographic diagnosis were included, in which case the date of first diagnosis was noted. Individual definitions of valve disease are as follows.
Aortic regurgitation was defined as the presence of diastolic flow through the aortic valve detected on color flow imaging or Doppler echocardiography. Aortic regurgitation was graded semiquantitatively by looking at the width of the jet in the left ventricular outflow tract as well as the jet area in the short-axis view of the aortic valve.30
Moderate aortic stenosis was defined as a mean pressure gradient of 20 to 30 mm Hg. Severe aortic stenosis was defined as a mean pressure gradient of >30 mm Hg. Mean gradient was used to judge the severity of aortic stenosis since aortic valve area calculations were not available in all patients.
Mitral regurgitation was defined as the presence of systolic flow through the mitral valve detected on color flow imaging and Doppler echocardiography. Semiquantitative assessment of severity of mitral regurgitation was made by visual estimate of the regurgitant jet size as a percentage of the left atrial size.31
Moderate mitral stenosis was defined as mitral valve area
of 1.1 to 1.4 cm. Severe mitral stenosis was defined as mitral
valve area 
1.0 cm2.
Definitions of cerebrovascular events, as follows, are identical to those used in previous Rochester Epidemiology Project population-based studies of stroke incidence and recurrence.1 2 32
Ischemic stroke was defined as the acute onset, over minutes to hours, of a focal neurological deficit persisting for >24 hours, with or without CT or MRI documentation, and due to altered circulation to a limited region of the cerebral hemispheres, brain stem, or cerebellum. Persons with only persistent sensory symptoms and minimal sensory signs or mild impairment of dexterity with preservation of normal muscle strength were included if the patient was aware of such symptoms being present for >24 hours. Patients with only deep tendon reflex changes or other minor signs without any functional impairment or awareness of the deficit were excluded. CT, MRI, or autopsy did not show evidence of intracerebral hemorrhage. Hemorrhagic infarction found on radiological imaging was classified as infarction. Without clinical evidence of stroke, a case of ischemic stroke detected at autopsy was excluded unless noted pathologically as a recent infarct, in which case the date of onset was estimated. Nonhemorrhagic infarctions from hematologic cause, vasculitis, or hemostatic factors were included. Persons with an area of probable infarction on CT without any associated clinical symptoms were not included.
Intracerebral hemorrhage was defined as the acute or progressive onset of a focal neurological deficit possibly associated with headache, vomiting, altered level of consciousness, signs of meningeal irritation, or blood-stained cerebrospinal fluid. If performed, CT, MRI, or autopsy demonstrated a parenchymal hemorrhage. Rupture of a lesion resulting in parenchymal hemorrhage that was not associated with hemorrhage into the subarachnoid space was classified as an intracerebral hemorrhage. A case of intracerebral hemorrhage detected at autopsy was included if described as a recent intracerebral hemorrhage, in which case a date of onset could be estimated. Traumatic and neonatal intracerebral hemorrhages were excluded.
Transient ischemic attack was defined as an episode of focal neurological symptoms with abrupt onset and rapid resolution, lasting <24 hours, and due to altered circulation to a limited region of the brain. Transient visual disturbances associated with monocular retinal ischemia were defined as amaurosis fugax (see below). Transient symptoms such as syncope, unexplained loss of consciousness, and dizziness or wooziness were excluded unless associated with other symptoms of brain stem ischemia. Symptoms such as vertigo, dysarthria, or diplopia, which occur in isolation without other symptoms of brain stem ischemia, were excluded. Focal symptoms associated with migraine were excluded.
Amaurosis fugax (transient monocular blindness) was defined as an episode of transient monocular visual disturbance, with abrupt onset and rapid resolution, lasting <24 hours, and due to altered circulation to the retina. The patient may have had total or partial loss of visual acuity affecting all or part of the visual field in that eye. Visual symptoms associated with migraine were excluded.
Syncope alone was not considered a cerebrovascular end point.
The medical records of all residents of Olmsted County, Minnesota, who had a 2-dimensional color Doppler echocardiogram and a diagnosis of valve disease during the period of the study were screened by a nurse abstractor under the supervision of a cardiologist and neurologist to identify individuals meeting criteria for inclusion in the cohort as well as to verify exclusion criteria. The entire medical record of those meeting inclusion criteria was then abstracted to record the presence or absence of hypertension, congestive heart failure, ischemic heart disease (myocardial infarction, angina pectoris), atrial fibrillation at the time of the echocardiogram, the date of subsequent development of these risk factors if not present at the time of the echocardiogram, and the date of subsequent development of transient ischemic attack, amaurosis fugax, ischemic stroke, intracerebral hemorrhage, subarachnoid hemorrhage, death, or date of last follow-up or date of last contact before migration from Olmsted County. Treatment with oral anticoagulant agents after echocardiographic diagnosis of valvular heart disease was also recorded, as were dates of aortic or mitral valve surgery, replacement, or valvuloplasty.
In addition to ischemic cerebrovascular events (ischemic stroke, transient ischemic attack, amaurosis fugax), we included intracerebral hemorrhage as a cerebrovascular end point because case-control studies have demonstrated that treatment with anticoagulant agents (which were frequently administered because of the diagnosis of valvular disease in our cohort) is an independent risk factor for intracerebral hemorrhage in our community (D.O. Wiebers, MD, et al, unpublished data, 2000). Subarachnoid hemorrhage was not treated as an end point because it has no etiologic relationship to valvular heart disease or its treatment.
All valve disease diagnoses were validated by the study echocardiographer, and all cerebrovascular events were validated by the study neurologists using the definitions outlined above. Because of the retrospective nature of the study, the study cardiologist and neurologist were not masked to the type of valve disease or other cardiovascular event or to the subsequent development of cerebrovascular events or death. Information regarding possible systemic (non–central nervous system) embolization was not collected because of the unreliability of clinical diagnosis of these events and infrequent validation of clinical diagnoses using radiographic or surgical techniques in our community.20
Statistical Analysis
The Kaplan-Meier product-limit method33 was
used to estimate rates of cerebrovascular events and death after the
date of first echocardiographic diagnosis of valve
disease. Patients were censored at the date of last information in the
medical record as an Olmsted County resident in the survival
analysis. Patients were censored at the time of death in the
analysis of rates of development of cerebrovascular events. We
calculated standardized morbidity ratios (SMRs) for cerebrovascular
events and survival by dividing the number of observed events in the
cohort by the expected number using age- and sex-appropriate rates from
the population. CIs on the SMRs were obtained by assuming that the
number of observed events has a Poisson distribution. Rates of observed
events were compared with reported rates by a 1-sample log-rank
test.
The Cox proportional hazards model34 was used to assess the impact of several potential risk factors on the occurrence of cerebrovascular events and death. Both baseline (at diagnosis of valvular heart disease) and time-dependent (possibly developing during follow-up) risk factors were considered. The baseline risk factors were age, sex, time from clinical to echo diagnosis of valvular heart disease, and treatment of valvular heart disease. The time-dependent variables included angina, myocardial infarction, hypertension, congestive heart failure, atrial fibrillation, and, in the survival model only, cerebrovascular events.
Univariate analyses of each of these variables were followed by multivariate analyses to identify risk factors making independent contributions to the development of cerebrovascular events or survival. This was done with the use of stepwise and backward selection procedures that involved not only the basic variables but also their interactions.
To compare death and cerebrovascular event rates among the different valve disease types (aortic stenosis, aortic regurgitation, mitral stenosis, mitral regurgitation), 4 variables indicating the presence or absence of each of the 4 types of valve diagnoses and 8 variables to differentiate moderate and severe hemodynamic designations among the 4 valve disease types were added to the multivariate proportional hazards models. The results of the variable selection process for the final models incorporating valve disease type and severity were validated by resampling (bootstrapping) methods,35 in which 500 multivariate proportional hazards regression analyses were performed on 500 data sets generated by random resampling from the original observations. Variables were validated for inclusion in the final model if they were included in >70% of these analyses at the 0.05 level.
The influence of anticoagulants on the development of cerebrovascular events was assessed with the use of a nested case-control design in which patients in the primary cohort who developed events (subsequent transient ischemic attack, amaurosis fugax, ischemic stroke, or intracerebral hemorrhage) served as cases. Each case was matched to a member of the cohort of the same age and sex who survived event free for the same length of time as the case had been event free. The odds ratio for treatment with warfarin at any time before the event and at the time of the event was derived by conditional logistic regression analyses.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
During the period of the study, 740 residents had a first 2-dimensional color Doppler echocardiographic diagnosis of moderate or severe valve disease. Eleven patients declined to give consent to have medical records reviewed for this study, resulting in a cohort of 729 patients. Nineteen patients had mitral stenosis, 528 had mitral regurgitation, 140 had aortic stenosis, and 106 had aortic regurgitation, with 63 having >1 type of valve disease. Valve disease was classified as hemodynamically severe in 2 patients (10.5%) with mitral stenosis, 46 (8.7%) with mitral regurgitation, 55 (39.3%) with aortic stenosis, and 7 (6.6%) with aortic regurgitation. Combinations of mixed valve disease were mitral regurgitation and aortic stenosis in 29 patients; mitral regurgitation and aortic regurgitation, 22; aortic stenosis and aortic regurgitation, 8; mitral stenosis and mitral regurgitation, 1; mitral stenosis and aortic stenosis, 1; mitral stenosis and aortic regurgitation, 1; and mitral regurgitation, aortic stenosis, and aortic regurgitation, 1.
The mean (±SD) age of patients in the entire cohort was 72.3±15.8
years, and the age range was 13 to 107 years. Four hundred twenty-four
of these patients (58.2%) had a nonechocardiographic
diagnosis of valve disease before the date of the first
echocardiographic diagnosis.
Cardiovascular risk factors among these patients are
presented in Table 1
. The entire
cohort was followed for a total of 2694 person-years. Median follow-up
was 3.5 years.
|
Cerebrovascular Events
Cerebrovascular events developed in 98 patients (13.4%) during
follow-up, including 4 with amaurosis fugax, 20 with transient
ischemic attack, 68 with ischemic stroke, and 6 with
intracerebral hemorrhage. Table 2 presents the Kaplan-Meier estimates
of rates of development of cerebrovascular events for the entire
cohort. These rates were significantly greater than the corresponding
age- and sex-adjusted rates for the community (SMR=3.2; 95% CI, 2.6 to
3.8; log-rank P<0.001; Figure 1). Rate estimates for the development of
cerebrovascular events for each valve disease subtype are
presented in Table 3. Differences
in rates among the valve groups were not significant when adjusted for
age and sex (P=0.6).
|
|
|
Time-dependent proportional hazards regression analysis
identified age and an interaction between age and intermittent or
persistent atrial fibrillation as independent determinants of
cerebrovascular events in this cohort, with atrial fibrillation
imparting a greater risk of cerebrovascular events at younger ages
(Table 4). Adjusted for age and atrial
fibrillation, severe aortic stenosis (compared with moderate
aortic stenosis) was also an independent determinant of the
development of cerebrovascular events (risk ratio [RR]=3.5; 95% CI,
1.4 to 8.6). The time-dependent proportional hazards regression
analysis was repeated with only ischemic
cerebrovascular events used as end points (excluding the 6
intracerebral hemorrhages), and the model was
unchanged.
|
Clinical diagnosis of valve disease before echocardiographic diagnosis was not a determinant of the subsequent development of cerebrovascular events, either in age-adjusted univariate analysis (RR=0.95; 95% CI, 0.63 to 1.43; P=0.8) or in the final multivariate model. One hundred twenty-two members of the cohort (16.7%) had aortic or mitral valve repair or replacement during follow-up. Valve repair or replacement was not a determinant of survival free of cerebrovascular events either in the univariate analysis (RR=1.64; 95% CI, 0.95 to 2.84) or in the final model.
Three hundred forty-five members of the cohort (47.3%) received anticoagulation after the diagnosis of valve disease, including 12 of the patients (63.2%) with mitral stenosis, 236 (44.7%) with mitral regurgitation, 67 (47.9%) with aortic stenosis, and 41 (38.7%) with aortic regurgitation. Nested case-control analysis, adjusted for age and atrial fibrillation, demonstrated no effect on the occurrence of cerebrovascular events for either current (odds ratio=1.21; 95% CI, 0.57 to 2.58) or prior (odds ratio=1.24; 95% CI, 0.64 to 2.39) treatment with warfarin. Three of the 6 intracerebral hemorrhages (50.0%) and 14 of the 68 ischemic strokes (20.6%) occurred among patients treated with warfarin.
Survival
During follow-up, 356 patients died. The cause of death was
congestive heart failure in 90, myocardial infarction in 47, stroke in
22, and sudden unexplained death in 3. The other deaths were due to
other noncardiovascular causes (193 patients) or
unknown cause (1 patient). The Kaplan-Meier estimates of rates of death
after first echocardiographic diagnosis of valve
disease are presented in Table 2
. These rates were
significantly worse than the age- and sex-adjusted rates for the
population (SMR=2.5; 95% CI, 2.2 to 2.7; P<0.001, Figure 2). Estimates of rates of death for the
individual types of valve disease, adjusted for age and sex, were not
significantly different (P=0.1) (Table 5).
|
|
Time-dependent Cox proportional hazards regression analysis
univariately identified ischemic heart disease
(RR=1.71; 95% CI, 1.38 to 2.12), atrial fibrillation (intermittent
RR=1.46; 95% CI, 1.13 to 1.90; persistent RR=1.35; 95% CI, 1.05 to
1.74), congestive heart failure (RR=4.19; 95% CI, 3.18 to 5.52),
cerebrovascular events (RR=2.44; 95% CI, 1.82 to 3.27), and clinical
diagnosis of valve disease before echocardiogram (RR=0.77; 95% CI,
0.62 to 0.95) as possible determinants of survival after adjustment for
age and sex. Multivariate time-dependent proportional
hazards regression analysis resulted in a final model in which
independent predictors of death were age, sex, cerebrovascular events,
ischemic heart disease, and congestive heart failure, which
interacted with each other as described in Table 6. After adjustment for these
variables, the type and severity of valve disease were not
important determinants of survival. Valve repair or replacement was not
a determinant of survival either in the univariate
analysis (RR=1.01; 95% CI, 0.72 to 1.41) or in the final
model.
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The significance of our study is that it reports rates, identifies determinants of cerebrovascular events, and quantifies the impact of cerebrovascular events on survival among all residents with newly echocardiographically diagnosed valve disease in a defined population. The population-based nature of our study limits bias that could have affected measurements of cerebrovascular events and survival in previous natural history studies involving patients referred to tertiary centers from outside a community. In contrast to many earlier studies addressing the risk of stroke among patients with valve disease, our study included only patients with newly echocardiographically diagnosed valve disease and without a history of cerebrovascular events, thus eliminating possible bias from inclusion of patients referred for echocardiographic evaluation of single or multiple cerebrovascular events or from inclusion of patients with a history of stroke in earlier years who had guaranteed survival to the time of echocardiography. In addition to providing new insights into the relationship between valve disease and cerebrovascular events, the uniform echocardiographic diagnostic criteria and large number of patients in our study allowed us to develop and validate multivariate models, which more precisely quantify the relative risks of cerebrovascular events and death attributable to various cardiovascular comorbidities among patients with valvular heart disease.
Studies of the relationship between valve disease and stroke derived
from patients followed at referral centers during earlier decades
focused on the importance of mitral stenosis as a source of
thromboembolism, particularly when accompanied by atrial
fibrillation.18 19 20 22 36 37 Aortic stenosis is
presumed to be uncommonly associated with stroke.11 36 37 38
In our study of patients with valvular heart disease newly
diagnosed by echocardiography, aortic
stenosis was, in fact, associated with rates of development of
cerebrovascular events similar to those for mitral stenosis
(Table 3) and was an independent determinant of the development
of cerebrovascular events, even after adjustment for age and atrial
fibrillation (RR=3.5). Our study suggests that
hemodynamically significant aortic valve disease may be
an underrecognized source of embolism among patients with
valvular heart disease. Alternatively,
hemodynamically severe aortic valve disease may be a
marker for other conditions that increase stroke risk in valve disease
patients, such as atherosclerosis39 40 41 42 43 44 or
prothrombotic tendencies.45 46
Another important finding of our study that differs from some earlier
studies of stroke risk among patients with valvular heart
disease is the interaction between age and atrial fibrillation. Prior
reports indicated either no difference in stroke risk among young and
old patients with valve disease and atrial fibrillation22
or an increased risk with age.20 38 Our
multivariate model demonstrates that atrial
fibrillation has a far stronger impact on the risk of cerebrovascular
events among younger patients with valve disease (Table 4).
Because we did not collect information on duration of atrial
fibrillation before the echocardiographic diagnosis of
valve disease, we cannot determine whether the greater impact on the
risk of stroke attributable to valvular atrial fibrillation at
younger ages is due to an increased risk primarily among patients with
new-onset atrial fibrillation, as suggested by Szekely,19
or whether atrial fibrillation simply becomes a relatively less
important mechanism of stroke in older patients with valve disease as
other stroke risk factors become more prevalent, despite the increase
in atrial fibrillation prevalence with age in the general
population.47
Our study also provides new quantitative evidence on the relative
importance of cardiovascular comorbidities that
influence survival in patients with valve
disease7 8 11 12 13 48 : age, ischemic heart disease,
and congestive heart failure. Our multivariate survival
model (Table 6) demonstrates that congestive heart failure is a
far more important determinant of death among patients with newly
diagnosed valvular heart disease than ischemic heart
disease. Congestive heart failure alone confers a >4-fold risk of
death among patients with valve disease, regardless of sex, whereas
ischemic heart disease confers a <2-fold increased risk of
death among men with valve disease and no significant increase in risk
of death among women with valve disease.
A unique and striking aspect of our study is the documentation of a nearly 9-fold increased risk of death among valvular heart disease patients with congestive heart failure who develop cerebrovascular events. This finding is an illustration of the power of multivariate modeling to identify biological interactions that strongly influence disease outcomes and suggests that prospective randomized studies to identify interventions to prevent cerebrovascular events and death among valve disease patients with congestive heart failure may be especially indicated, regardless of the presence or absence of atrial fibrillation.
In contrast to the findings in the cerebrovascular event analysis, we found that among the various types and severities of aortic and mitral disease, none were independent predictors of death after adjustment for age, sex, ischemic heart disease, congestive heart failure, and cerebrovascular events. These findings suggest that clinical comorbidities of ischemic heart disease, congestive heart failure, and cerebrovascular events are more important determinants of survival among patients with valvular heart disease than the particular type and severity of valve disease.
Despite the decline in rheumatic valve disease and recent advances in
medical and surgical therapy for cardiac disease, our study
demonstrates that the relative risks of cerebrovascular events and
death in valve disease patients remain disturbingly high, at 3.2 and
2.5 times, respectively, the age- and sex-adjusted rates for the
general population. In fact, death rates in our community for
individual types of valve disease (Table 5) are similar to those
reported from a referral center by Rapaport8 25 years ago.
Although earlier non-population-based studies may have overestimated
death rates by inclusion of patients selected for referral because of
comorbidity profiles that negatively influence survival, our findings
suggest that randomized treatment trials for patients with valve
disease may be indicated to determine interventions that will improve
outcomes in these patients.
Because we included amaurosis fugax and transient ischemic attack as cerebrovascular end points, the question could arise as to whether patients with known valvular heart disease might be more vigilant regarding medical symptoms and thus more likely to notice and seek medical attention for transient symptoms than individuals in the general population, thus resulting in an overestimation of the standardized morbidity ratio for cerebrovascular events. We believe this is not the case for 2 reasons. First, we have shown previously that our methods of ascertainment identify virtually all instances of transient ischemic attack and amaurosis fugax that occur in our community.49 Second, the proportion (24 of 92, 26.1%) of amaurosis fugax and transient ischemic attack among all incident ischemic events in our cohort is actually lower than the proportion in the community (34.4%).50 51 This observation is also consistent with the findings of previous case-control studies in our community, which suggest that valvular heart disease is less likely to be associated with cerebral ischemic episodes that are brief enough to be called transient ischemic attack than other mechanisms of cerebral ischemia.52
Our study does have some limitations. The community is largely white, and the prevalence of cardiovascular risk factors may not be the same as in more ethnically diverse communities. Although demographic factors may influence overall assessment of determinants of cerebrovascular events and survival for the population in general, we believe they are probably less important in assessing the determinants of these end points within a defined cohort of individuals with valve disease. Although our study is population based, most patients were evaluated by physicians in a tertiary referral center. Replication of our study in a setting remote from a tertiary referral center could give different results. Inclusion of autopsy-ascertained cerebrovascular events could have biased our findings toward a higher RR for cerebrovascular disease as a determinant of death. On the other hand, exclusion of autopsy-ascertained events would not have been consistent with the population-based design of our study and could have resulted in underestimation of cerebrovascular event rates.
Because of the retrospective and observational nature of the study, there was no standard echocardiographic follow-up, and we were therefore unable to assess the impact of hemodynamic progression of valvular disease or severity of congestive heart failure on outcomes. The smaller number of patients and end points within each subgroup of valve disease, especially mitral stenosis, did not permit modeling of interactions between different types and severities of valve disease53 or modeling of determinants of cerebrovascular events and survival within individual valve disease subgroups. We did not examine the effect of the specific etiology of valve disease on survival and cerebrovascular events in the cohort, although we did control for the presence or absence of other cardiovascular diseases that might play a role in the genesis of valve disease in some patients. The relative infrequency with which new diagnoses of mitral stenosis were made in our community during the period of our study is probably attributable in part to the decline in the incidence of rheumatic fever and rheumatic mitral valve disease. We do not have information on ischemic stroke subtypes and therefore cannot determine whether the increased cerebrovascular risk attributable to severe aortic stenosis is due to embolism or due to associated atherosclerotic arterial occlusive disease or atherosclerotic aortic debris. The observational nature of our study somewhat limits our ability to speculate on why, despite treatment advances, survival in these patients remains so poor. For example, we did not find any influence on outcomes for valve replacement or repair in time-dependent multivariate analysis, nor could we detect a protective effect of treatment with warfarin with respect to the occurrence of cerebrovascular events (adjusting for age and atrial fibrillation) in the nested case-control analysis. However, we do not believe that the results of this historical cohort study can be interpreted as indications of lack of efficacy or safety of these treatments, because patients were selected and treated in a nonrandomized, nonstandardized fashion, using criteria for which our models cannot completely control. For example, patients who were not selected for anticoagulation may have had risk factor profiles that guaranteed better outcomes than those who were selected for treatment. Therefore, inferences regarding treatment recommendations should not be made on the basis of these observations. Finally, because ours is a historical cohort study of patients with hemodynamically significant valvular heart disease diagnosed echocardiographically, the qualitative and quantitative information that our study provides about risk factors and their interactions is applicable to patients with valvular heart disease, not to patients in the general population without valvular heart disease.
In summary, despite advances in medical and surgical therapy for valve disease and its complications over the last several decades, the risk of cerebrovascular events and death among these patients remains high. Age, atrial fibrillation, and the presence of severe aortic stenosis are the most important determinants of cerebrovascular events in patients with valvular heart disease. Age, sex, ischemic heart disease, congestive heart failure, and cerebrovascular events are the most important determinants of death, with the poorest survival being among those valvular heart disease patients with congestive heart failure who develop cerebrovascular events. These findings predict a continuing significant burden of morbidity and mortality attributable to valve disease, especially in developing countries where the prevalence of valve disease is highest.
|
Acknowledgments |
|---|
This study was supported by the National Institute of Neurological Disorders and Stroke (grant NS06663).
|
Footnotes |
|---|
Reprint requests to George W. Petty, MD, Mayo Clinic, 200 First St SW, Rochester, MN 55905.
Received February 22, 2000; revision received July 11, 2000; accepted August 2, 2000.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
1. Brown RD, Whisnant JP, Sicks JD, O’Fallon WM, Wiebers DO. Stroke incidence, prevalence, and survival: secular trends in Rochester, Minnesota, through 1989. Stroke. 1996;27:373–380.
2.
Broderick JP, Phillips SJ, O’Fallon WM, Frye RL,
Whisnant JP. Relationship of cardiac disease to stroke occurrence,
recurrence, and mortality. Stroke. 1992;23:1250–1256.
3.
Broderick J, Brott T, Kothari R, Miller R, Khoury J,
Pancioli A, Gebel J, Mills D, Minneci L, Shukla R. The Greater
Cincinnati/Northern Kentucky Stroke Study: preliminary first-ever and
total incidence rates of stroke among blacks. Stroke. 1998;29:415–421.
4. Stollerman GH. Rheumatic fever. Lancet. 1997;349:935–942.[Medline] [Order article via Infotrieve]
5.
Marcus RH, Sareli P, Pocock WA, Barlow JB. The
spectrum of severe rheumatic mitral valve disease in a developing
country: correlations among clinical presentation, surgical
pathologic findings, and hemodynamic sequelae.
Ann Intern Med. 1994;120:177–183.
6.
Bisno AL, Shulman ST, Dajani AS. The rise and fall
(and rise?) of rheumatic fever. JAMA. 1988;259:728–729.
7. Goldschlager N, Pfeifer J, Cohn K, Popper R, Selzer A. The natural history of aortic regurgitation: a clinical and hemodynamic study. Am J Med. 1973;54:577–588.[Medline] [Order article via Infotrieve]
8. Rapaport E. Natural history of aortic and mitral valve disease. Am J Cardiol. 1975;35:221–227.[Medline] [Order article via Infotrieve]
9.
Adams GF, Merrett JD, Hutchinson WM, Pollock AM.
Cerebral embolism and mitral stenosis: survival with and
without anticoagulants. J Neurol Neurosurg Psychiatry. 1974;37:378–383.
10. Nakajima K, Naito M, Nakanishi N, Shimomura K. Studies on non-operative cases of acquired valvular heart disease. Jpn Circ J. 1984;48:1118–1122.[Medline] [Order article via Infotrieve]
11. Shimada S. A 13-year follow-up study of rheumatic valvular diseases. Jpn Circ J. 1986;50:1304–1308.[Medline] [Order article via Infotrieve]
12. Kelly TA, Rothbart RM, Cooper CM, Kaiser DL, Smucker ML, Gibson RS. Comparison of outcome of asymptomatic to symptomatic patients older than 20 years of age with valvular aortic stenosis. Am J Cardiol. 1988;61:123–130.[Medline] [Order article via Infotrieve]
13.
Turina J, Hess O, Sepulcri F, Krayenbuehl HP.
Spontaneous course of aortic valve disease. Eur Heart
J. 1987;8:471–483.
14. Chizner MA, Pearle DL, deLeon AC Jr. The natural history of aortic stenosis in adults. Am Heart J. 1980;99:419–424.[Medline] [Order article via Infotrieve]
15. Rowe JC, Bland EF, Sprague HB, White PD. The course of mitral stenosis without surgery: ten- and twenty-year perspectives. Ann Intern Med. 1960;52:741–749.
16. Olesen KH. The natural history of 271 patients with mitral stenosis under medical treatment. Br Heart J. 1962;24:349–357.
17. Munoz S, Gallardo J, Diaz-Gorrin JR, Medina O. Influence of surgery on the natural history of rheumatic mitral and aortic valve disease. Am J Cardiol. 1975;35:234–242.[Medline] [Order article via Infotrieve]
18. Abernathy WS, Willis PW III. Thromboembolic complications of rheumatic heart disease. Cardiovasc Clin. 1973;5:131–175.[Medline] [Order article via Infotrieve]
19. Szekely P. Systemic embolism and anticoagulant prophylaxis in rheumatic heart disease. BMJ. 1964;1:1209–1212.
20.
Fleming HA, Bailey SM. Mitral valve disease, systemic
embolism and anticoagulants. Postgrad Med J. 1971;47:599–604.
21. Wilson JK, Greenwood WF. The natural history of mitral stenosis. Can Med Assoc J. 1954;71:323–331.
22.
Coulshed N, Epstein EJ, McKendrick CS, Galloway RW,
Walker E. Systemic embolism in mitral valve disease. Br
Heart J. 1970;32:26–34.
23. Melton LJ III. History of the Rochester Epidemiology Project. Mayo Clin Proc. 1996;71:266–274.[Abstract]
24. Ballard DJ, Khandheria BK, Tajik AJ, Seward JB, Weber VP, Melton LJ III. Population-based study of echocardiography: time trends in utilization and diagnostic profile of an evolving technology, 1975–1987. Int J Technol Assess Health Care. 1989;5:249–261.[Medline] [Order article via Infotrieve]
25. Aronow WS, Kronzon I. Correlation of prevalence and severity of valvular aortic stenosis determined by continuous-wave Doppler echocardiography with physical signs of aortic stenosis in patients aged 62 to 100 years with aortic systolic ejection murmurs. Am J Cardiol. 1987;60:399–401.[Medline] [Order article via Infotrieve]
26. Grayburn PA, Smith MD, Handshoe R, Friedman BJ, Demaria AN. Detection of aortic insufficiency by standard echocardiography, pulsed Doppler echocardiography, and auscultation: a comparison of accuracies. Ann Intern Med. 1986;104:599–605.
27. Meyers DG, McCall D, Sears TD, Olson TS, Felix GL. Duplex pulsed Doppler echocardiography in mitral regurgitation. CU J Clin Ultrasound. 1986;14:117–121.
28. Vargas-Barron J, Bialostozky D, Attie F, Pop G, Keirns C, Gil-Moreno M, Esquivel-Avila J. Differential diagnosis of various causes of systolic-diastolic murmurs using pulsed Doppler echocardiography. Am Heart J. 1984;108:1507–1513.[Medline] [Order article via Infotrieve]
29. Saal AK, Gross BW, Franklin DW, Pearlman AS. Noninvasive detection of aortic insufficiency in patients with mitral stenosis by pulsed Doppler echocardiography. J Am Coll Cardiol. 1985;5:176–181.[Abstract]
30. Perry GJ, Helmcke F, Nanda NC, Byard C, Soto B. Evaluation of aortic insufficiency by Doppler color flow mapping. J Am Coll Cardiol. 1987;9:952–959.[Abstract]
31.
Helmcke F, Nanda NC, Hsiung MC, Soto B, Adey CK, Goyal
RG, Gatewood RP Jr. Color Doppler assessment of mitral
regurgitation with orthogonal planes.
Circulation. 1987;75:175–183.
32.
Petty GW, Brown RD Jr, Whisnant JP, Sicks JD, O’Fallon
WM, Wiebers DO. Survival and recurrence after first cerebral
infarction: a population-based study in Rochester, Minnesota, 1975
through 1989. Neurology. 1998;50:208–216.
33. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53:457–481.
34. Cox DR. Regression models and life-tables. J R Stat Soc [B]. 1972;34:187–202.
35. Hjorth JSU. Computer Intensive Statistical Methods: Validation Model Selection and Bootstrap. London, UK: Chapman & Hall; 1994.
36.
Bonow RO, Carabello B, de Leon AC Jr, Edmunds LH Jr,
Fedderly BJ, Freed MD, Gaasch WH, McKay CR, Nishimura RA, O’Gara PT,
O’Rourke RA, Rahimtoola SH, Ritchie JL, Cheitlin MD, Eagle KA, Gardner
TJ, Garson A Jr, Gibbons RJ, Russell RO, Ryan TJ, Smith SC Jr. ACC/AHA
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 (Committee on Management of Patients With
Valvular Heart Disease). Circulation. 1998;98:1949–1984.
37. Salem DN, Levine HJ, Pauker SG, Eckman MH, Daudelin DH. Antithrombotic therapy in valvular heart disease. Chest. 1998;114(suppl):590S–601S.
38.
Cerebral Embolism Task Force. Cardiogenic brain
embolism. Arch Neurol. 1986;43:71–84.
39.
Boon A, Lodder J, Cheriex E, Kessels F. Risk of stroke
in a cohort of 815 patients with calcification of the aortic valve with
or without stenosis. Stroke. 1996;27:847–851.
40.
Boon A, Cheriex E, Lodder J, Kessels F. Cardiac valve
calcification: characteristics of patients with calcification of the
mitral annulus or aortic valve. Heart. 1997;78:472–474.
41. Stewart BF, Siscovick D, Lind BK, Gardin JM, Gottdiener JS, Smith VE, Kitzman DW, Otto CM, for the Cardiovascular Health Study. Clinical factors associated with calcific aortic valve disease. J Am Coll Cardiol. 1997;29:630–634.[Abstract]
42.
Wilmshurst PT, Stevenson RN, Griffiths H, Lord JR. A
case-control investigation of the relation between hyperlipidaemia and
calcific aortic valve stenosis. Heart. 1997;78:475–479.
43.
Juvonen J, Juvonen T, Laurila A, Kuusisto J, Alarakkola
E, Sarkioja T, Bodian CA, Kairaluoma MI, Saikku P. Can degenerative
aortic valve stenosis be related to persistent Chlamydia
pneumoniae infection? Ann Intern Med. 1998;128:741–744.
44.
Otto CM, Lind BK, Kitzman DW, Gersh BJ, Siscovick DS.
Association of aortic-valve sclerosis with
cardiovascular mortality and morbidity in the elderly.
N Engl J Med. 1999;341:142–147.
45. Barbut D, Borer JS, Gharavi A, Wallerson D, Devereux RB, Supino P, Suite ND. Prevalence of anticardiolipin antibody in isolated mitral or aortic regurgitation, or both, and possible relation to cerebral ischemic events. Am J Cardiol. 1992;70:901–905.[Medline] [Order article via Infotrieve]
46. Lip GY, Rumley A, Dunn FG, Lowe GD. Thrombogenesis in mitral regurgitation and aortic stenosis. Angiology. 1996;47:1117–1125.
47. Phillips SJ, Whisnant JP, O’Fallon WM, Frye RL. Prevalence of cardiovascular disease and diabetes mellitus in residents of Rochester, Minnesota. Mayo Clin Proc. 1990;65:344–359.[Medline] [Order article via Infotrieve]
48. Pellikka PA, Nishimura RA, Bailey KR, Tajik AJ. The natural history of adults with asymptomatic, hemodynamically significant aortic stenosis. J Am Coll Cardiol. 1990;15:1012–1017.[Abstract]
49. Whisnant JP, Melton LJ III, Davis PH, O’Fallon WM, Nishimaru K, Schoenberg BS. Comparison of case ascertainment by medical record linkage and cohort follow-up to determine incidence rates for transient ischemic attacks and stroke. J Clin Epidemiol. 1990;43:791–797.[Medline] [Order article via Infotrieve]
50.
Petty GW, Brown RD Jr, Whisnant JP, Sicks JD, O’Fallon
WM, Wiebers DO. Frequency of major complications of aspirin, warfarin,
and intravenous heparin for secondary stroke prevention: a
population-based study. Ann Intern Med. 1999;130:14–22.
51.
Brown RD Jr, Petty GW, O’Fallon WM, Wiebers DO,
Whisnant JP. Incidence of transient ischemic attack in
Rochester, Minnesota, 1985–1989. Stroke. 1998;29:2109–2113.
52.
Whisnant JP, Brown RD, Petty GW, O’Fallon WM, Sicks
JD, Wiebers DO. Comparison of population-based models of risk factors
for TIA and ischemic stroke. Neurology. 1999;53:532–536.
53.
Chiang CW, Lo SK, Ko YS, Cheng NJ, Lin PJ, Chang CH.
Predictors of systemic embolism in patients with mitral
stenosis: a prospective study. Ann Intern Med. 1998;128:885–889.
This article has been cited by other articles:
 |
|
 |
|
  E. Doufekias, A. Z. Segal, and J. R. Kizer Cardiogenic and aortogenic brain embolism. J. Am. Coll. Cardiol., March 18, 2008; 51(11): 1049 - 1059. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. W. Petty, B. K. Khandheria, I. Meissner, J. P. Whisnant, W. A. Rocca, T. J. H. Christianson, J. D. Sicks, W. M. O'Fallon, R. L. McClelland, and D. O. Wiebers Population-Based Study of the Relationship Between Patent Foramen Ovale and Cerebrovascular Ischemic Events Mayo Clin. Proc., May 1, 2006; 81(5): 602 - 608. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. W. Petty, B. K. Khandheria, I. Meissner, J. P. Whisnant, W. A. Rocca, J. D. Sicks, T. J. H. Christianson, W. M. O'Fallon, R. L. McClelland, and D. O. Wiebers Population-Based Study of the Relationship Between Atherosclerotic Aortic Debris and Cerebrovascular Ischemic Events Mayo Clin. Proc., May 1, 2006; 81(5): 609 - 614. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. R. Kizer, D. O. Wiebers, J. P. Whisnant, J. M. Galloway, T. K. Welty, E. T. Lee, L. G. Best, H. E. Resnick, M. J. Roman, and R. B. Devereux Mitral Annular Calcification, Aortic Valve Sclerosis, and Incident Stroke in Adults Free of Clinical Cardiovascular Disease: The Strong Heart Study Stroke, December 1, 2005; 36(12): 2533 - 2537. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. W. Petty, B. K. Khandheria, J. P. Whisnant, J. D. Sicks, W. M. O'Fallon, and D. O. Wiebers Outcomes Among Valvular Heart Disease Patients Experiencing Ischemic Stroke or Transient Ischemic Attack in Olmsted County, Minnesota Mayo Clin. Proc., August 1, 2005; 80(8): 1001 - 1008. [Abstract] [PDF]
|
|
|
|
 |
|
 C. S. Fox, R. S. Vasan, H. Parise, D. Levy, C. J. O'Donnell, R. B. D'Agostino, and E. J. Benjamin Mitral Annular Calcification Predicts Cardiovascular Morbidity and Mortality: The Framingham Heart Study Circulation, March 25, 2003; 107(11): 1492 - 1496. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. L. Tirschwell, W.T. Longstreth Jr, K. J. Becker, R. E. Gammans Sr, L. A. Sabounjian, S. Hamilton, and L. B. Morgenstern Shortening the NIH Stroke Scale for Use in the Prehospital Setting Stroke, December 1, 2002; 33(12): 2801 - 2806. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||