Association between Heart Transplantation and Subsequent Risk of Stroke among Patients with Heart Failure

Abstract

Background and Purpose:

It is uncertain whether heart transplantation decreases the risk of stroke. The objective of our study was to determine whether heart transplantation is associated with a decreased risk of subsequent stroke among patients with heart failure awaiting transplantation.

Methods:

We performed a retrospective cohort study using administrative data from New York, California, and Florida between 2005-2015. Individuals with heart failure awaiting heart transplantation were identified using previously validated ICD-9-CM diagnosis codes for heart failure in combination with code V49.83 for “awaiting organ transplant” status. Individuals with prior stroke were excluded. Our primary exposure variable was heart transplantation, modeled as a time-varying covariate and defined by procedure code 37.51. The primary outcome was stroke, defined as the composite of ischemic and hemorrhagic stroke. Survival statistics were used to calculate stroke incidence, and Cox proportional hazards analysis was used to determine the association between heart transplantation and stroke while adjusting for demographics, stroke risk factors, Elixhauser comorbidities, and implantation of a left ventricular assist device.

Results:

We identified 7,848 patients with heart failure awaiting heart transplantation, of whom 1,068 (13.6%) underwent heart transplantation. Over a mean follow-up of 2.7 years, we identified 428 strokes. The annual incidence of stroke was 0.7% (95% CI, 0.5-1.0%) after heart transplantation versus 2.4% (95% CI, 2.2-2.6%) among those awaiting heart transplantation. After adjustment for potential confounders, heart transplantation was associated with a lower risk of stroke (HR 0.4; 95% CI, 0.2-0.6).

Conclusions:

Heart transplantation is associated with a decreased risk of stroke among patients with heart failure awaiting transplantation.

Heart transplantation remains the optimal long-term therapy for patients with severe heart failure refractory to medical therapy.¹ Approximately 25,000 patients have undergone heart transplantation in the United States over the last 10 years.¹ Without heart transplantation, patients with severe heart failure face a high risk of stroke.^{2, 3} Furthermore, although left ventricular assist devices (LVAD) are increasingly being used as a bridge therapy to heart transplantation, these devices are associated with an approximately 9% annual incidence of stroke.⁴ Moreover, data from prior studies and the United Network of Organ Sharing indicate that cerebrovascular disease remains a leading cause of death following transplantation.^{1, 5–7} Hence, we evaluated whether heart transplantation is associated with a decreased risk of subsequent stroke in patients with severe heart failure awaiting transplantation.

Methods

Design

We performed a retrospective cohort study using data on all emergency department (ED) visits and hospitalizations in New York, California, and Florida between 2005-2015. These states comprise approximately 25% of the total US population and were chosen because they are the three largest states with publicly available deidentified data that allow individual patients to be followed anonymously through subsequent ED and hospital visits.⁸ Data were available for 2006–2014 in New York, 2005–2011 in California, and 2005–2015 in Florida. These data are collected from the New York Statewide Planning and Research Cooperative System, the California Office of Statewide Health Planning and Development, and the Florida Agency for Health Care Administration by the Agency for Healthcare Research and Quality for its Healthcare Cost and Utilization Project (HCUP).⁹ The data used in this analysis includes restricted claims data and therefore cannot be shared directly with other investigators because of the terms of the data use agreement. However, investigators can obtain access to these data by application to the HCUP. We adhered to guidelines for research using administrative claims data.¹⁰ The Weill Cornell Medicine institutional review board approved this study and waived the requirement for informed consent.

Patient Population

We identified all adult patients with heart failure awaiting heart transplantation using previously validated International Classification of Diseases, 9^th Revision, Clinical Modification (ICD-9-CM) diagnosis codes for heart failure¹¹ in combination with code V49.83 for “awaiting organ transplant” status. We excluded patients who were documented as having a prior stroke as early as their first recorded hospitalization or ED visit in 2005 in California and Florida and in 2006 in New York. We also excluded patients who had a stroke during the index hospitalization for heart transplantation. In addition, we excluded non-residents of New York, California, and Florida in order to maximize the completeness of follow-up.

Measurements

Our exposure variable was heart transplantation, defined as per previous studies by ICD-9-CM procedure code 37.51.¹² Heart transplantation was modeled as a time-varying covariate. The primary outcome was a hospital admission for any stroke, defined as the composite of ischemic or hemorrhagic stroke. Secondary outcomes were hospital admission for ischemic or hemorrhagic stroke (intracerebral hemorrhage [ICH] or subarachnoid hemorrhage [SAH]). Ischemic stroke was defined by ICD-9-CM codes 433.x1, 434.x1, or 436 in any discharge diagnosis position in the absence of any concomitant code for trauma or intracranial hemorrhage or a primary code for rehabilitation. This algorithm has been found to be 86% sensitive and 95% specific for ischemic stroke.¹³ Hemorrhagic stroke was defined as any diagnosis code for ICH (ICD-9-CM code 431) or SAH (430), in the absence of any concomitant code for trauma or a primary code for rehabilitation. This has been shown to be 82% sensitive and 93% specific for ICH, and 98% sensitive and 92% specific for SAH.¹³

We identified the following demographic characteristics and stroke risk factors: age, sex, race, hypertension, diabetes, coronary heart disease, peripheral vascular disease, chronic obstructive pulmonary disease, chronic kidney disease, atrial fibrillation, tobacco use, and alcohol abuse. In addition, we measured the overall burden of medical illness using the Elixhauser comorbidity index, which is often used as a surrogate for overall medical illness used in studies using large administrative datasets.¹⁴ Finally, because LVAD implantation is frequently used as a bridge to heart transplantation and is itself associated with a significant stroke risk,¹⁵ we used ICD-9-CM procedure code 37.66 to ascertain LVAD implantation status and modeled it as a time-varying covariate.^{4, 16}

Statistical Analysis

Baseline characteristics were reported with standard descriptive statistics and compared using the chi-squared test or t-test, as appropriate. Kaplan Meier survival statistics were used to calculate the cumulative rate of stroke. Patients were censored at the time of death or at the last available follow-up date. After visually inspecting the log-log plots to confirm that the proportional hazards assumption was met, Cox regression models were used to calculate cause-specific hazard ratios (HR) for the association between heart transplantation and stroke while adjusting for demographics, stroke risk factors, Elixhauser comorbidities, and implantation of a LVAD. These models were also used to identify clinical variables associated with stroke among patients awaiting heart transplantation. All analyses were performed using Stata/MP version 15 (StataCorp, TX).

Results

We identified 7,848 patients with heart failure awaiting heart transplantation, of whom 1068 (13.6%) eventually underwent transplantation. Patients who received a heart transplant were slightly younger than those who did not undergo transplantation (mean age, 52.4 versus 54.2 years), were more often male, and were more likely to have received a LVAD (Table 1). Over a mean follow-up of 2.7 (±2.5) years, we identified 428 (5.4%) patients with a stroke. Of these, 304 (71.0%) were ischemic strokes and 124 (29.0%) were hemorrhagic strokes. Patients with stroke were of similar age (mean age, 54.8 versus 53.9 years) and had a higher burden of stroke risk factors (Table 2).

Table 1.

Characteristics of Heart Failure Patients, Stratified by Heart Transplantation

Characteristic*	Heart Transplantation (N = 1068)	No Heart Transplantation (N = 6,780)	P-Value
Age, mean (SD), y	52.4 (12.8)	54.2 (12.0)	<0.001
Female	242 (22.7)	2,459 (36.3)	<0.001
Race†			<0.001
White	643 (60.2)	3,064 (45.2)
Black	176 (16.5)	1,401 (20.7)
Other	239 (22.4)	2,202 (32.5)
Hypertension	739 (69.2)	5,478 (80.8)	<0.001
Diabetes	454 (42.5)	3,638 (53.7)	<0.001
Coronary heart disease	734 (68.7)	3,745 (55.2)	<0.001
Peripheral vascular disease	119 (11.1)	1,010 (14.9)	0.001
Chronic obstructive pulmonary disease	252 (23.6)	1,667 (24.6)	0.49
Chronic kidney disease	434 (40.6)	4,175 (61.6)	<0.001
Atrial fibrillation	557 (52.2)	1,833 (27.0)	<0.001
Tobacco use	418 (39.1)	2,173 (32.1)	<0.001
Alcohol use	219 (20.5)	1,673 (24.7)	0.004
Left Ventricular Assist Device	188 (17.6)	348 (5.1)	<0.001
Elixhauser comorbidities, mean (SD)	2.8 (2.0)	3.1 (1.9)	<0.001

Abbreviations: SD, standard deviation

^*Data are presented as number (%) unless otherwise specified.

^†Self-reported by patients or their surrogates. Numbers do not sum to group totals because of missing race/ethnicity data in a small fraction of patients.

Table 2.

Characteristics of Patients, Stratified by Stroke

Characteristic*	Stroke (N = 428)	No Stroke (N = 7,420)	P-Value
Age, mean (SD), y	54.8 (11.4)	53.9 (12.2)	0.11
Female	145 (33.9)	2,556 (34.5)	0.81
Race†			<0.001
White	167 (39.2)	3,540 (48.5)
Black	124 (29.1)	1,453 (19.9)
Other	135 (31.7)	2,305 (31.6)
Hypertension	364 (85.1)	5,853 (78.9)	0.002
Diabetes	258 (60.3)	3,834 (51.7)	0.001
Coronary heart disease	297 (69.4)	4,182 (56.4)	<0.001
Peripheral vascular disease	72 (16.8)	1,057 (14.3)	0.14
Chronic obstructive pulmonary disease	111 (25.9)	1,808 (24.4)	0.46
Chronic kidney disease	276 (64.5)	4,333 (58.4)	0.01
Atrial fibrillation	159 (37.2)	2,231 (30.1)	0.002
Tobacco use	132 (30.8)	2,459 (33.1)	0.38
Alcohol use	94 (22.0)	1,798 (24.2)	0.29
Left Ventricular Assist Device	81 (18.9)	455 (6.1)	<0.001
Elixhauser comorbidities, mean (SD)	3.0 (2.0)	3.0 (1.9)	0.98

Abbreviations: SD, standard deviation

^*Data are presented as number (%) unless otherwise specified.

^†Self-reported by patients or their surrogates. Numbers do not sum to group totals because of missing race/ethnicity data in a small fraction of patients.

The annual incidence of stroke was 0.7% (95% Confidence Interval [CI], 0.5-1.0%) after heart transplantation versus 2.4% (95% CI, 2.2-2.6%) among those awaiting heart transplantation. The annual incidence of ischemic stroke was 0.4% (95% CI, 0.2-0.6%) after heart transplantation versus 1.7% (95% CI, 1.5-1.9%) among those awaiting heart transplantation. The annual incidence of hemorrhagic stroke was 0.3% (95% CI, 0.2-0.5%) after heart transplantation versus 0.7% (95% CI, 0.6-0.8%) among those awaiting heart transplantation (Table 3).

Table 3.

Rate of Stroke among Patients Awaiting Heart Transplantation, Stratified by Heart Transplantation Status

Transplant Status*	Annual Incidence	5-Year Cumulative Rate
Awaiting heart transplant
All stroke	2.4% (2.2-2.6%)	9.2% (8.2-10.2%)
Ischemic stroke	1.7% (1.5-1.9%)	6.8% (5.9-7.7%)
Hemorrhagic stroke	0.7% (0.6-0.8%)	2.6% (2.1-3.1%)
Heart transplant
All stroke	0.7% (0.5-1.0%)	2.9% (1.8-4.4%)
Ischemic stroke	0.4% (0.2-0.6%)	1.5% (0.8-2.7%)
Hemorrhagic stroke	0.3% (0.2-0.5%)	1.4% (0.7-2.6%)

^*Data are presented as percent or ratios with the 95% confidence interval in parentheses.

By 5 years, the cumulative rate of stroke was 2.9% (95% CI, 1.8%-4.4%) among patients who underwent heart transplantation, as compared to 9.2% (95% CI, 8.2%-10.2%) among those awaiting heart transplantation. The cumulative rates of stroke were lower for both ischemic and hemorrhage stroke subtypes (Figure, Table 3).

Figure 1:

Cumulative Incidence of Stroke Stratified by Presence or Absence of Heart Transplantation.

Kaplan-Meier curve showing cumulative rates of (A) all stroke, (B) ischemic stroke, and (C) hemorrhagic stroke stratified by presence or absence of heart transplantation.

In the primary analysis after adjustment for demographics, stroke risk factors, Elixhauser comorbidities, and use of a LVAD, heart transplantation was associated with a lower risk of stroke (HR, 0.4; 95% CI, 0.2-0.6). The association appeared stronger for ischemic stroke (HR, 0.3; 95% CI, 0.2-0.5) than for hemorrhagic stroke (HR, 0.6; 95% CI, 0.3-1.1).

Finally, among patients awaiting heart transplantation, clinical variables that were associated with a higher risk of stroke included black race (HR, 1.8; 95% CI, 1.4-2.3), coronary heart disease (HR, 1.5; 95% CI, 1.2-1.9), atrial fibrillation (HR, 1.5; 95% CI, 1.2-1.9), and diabetes (HR, 1.2; 95% CI, 1.01-1.5).

Discussion

In a large, heterogeneous population of patients with heart failure awaiting transplantation, receipt of a heart transplant was associated with a significantly lower subsequent risk of stroke. Heart transplantation appeared to be more strongly associated with a decreased risk of ischemic stroke as compared to hemorrhagic stroke.

Previous studies have reported rates of stroke following heart transplantation that range from 4% to 43%.^5–7 Most recently, van de Beek et al reported a 6% rate of stroke following heart transplantation in 313 patients at Mayo Clinic in Rochester between 1998-2006.⁵ Our findings in this broadly representative cohort build upon prior case series and registries by comparing stroke risk after transplantation to stroke risk among patients awaiting transplantation.

The mechanism of decreased stroke risk after heart transplantation is uncertain. Patients with severe heart failure face an increased risk of stroke,^{2, 3} and transplantation may mitigate this risk by reversal of systolic dysfunction which may prevent cerebrovascular disease by decreasing the propensity for left ventricular clot formation and subsequent embolism. Improving efforts to reverse systolic dysfunction may therefore be a viable strategy of reducing stroke risk independent of anticoagulation therapy.^17–21 Additionally, prior data suggest that heart transplantation may decrease the risk of subsequent atrial fibrillation compared to other major cardiac surgeries due to denervation of the autonomic nervous system.^{22, 23} Finally, heart transplantation leads to a rapid normalization of intracardiac pressures and cardiac output which may also minimize the risk of clot formation and subsequent stroke.²⁴

Although the number of patients awaiting heart transplantation has steadily increased since 2007, the rate of heart transplantation per waitlist years has gradually declined.¹ Moreover, until 2015, the number of donor hearts used for heart transplant had not increased in the past 20 years; the United Network of Organ Sharing reported that 2,177 heart transplantations occurred in 1994 and 2,166 heart transplantations occurred in 2014.¹ Meanwhile, approximately 2.5 million individuals in the US have heart failure with reduced ejection fraction, the most common diagnosis for which heart transplantation is considered.¹⁸ In addition, between 100,000 to 200,000 of these patients are sick enough to require heart transplant and healthy enough to otherwise benefit from it.^{18, 25} Furthermore, long-term prognosis and survival following heart transplantation continues to improve. The 5-year survival between 2008-2010 was 77.0%.¹ Due to the dearth of available donor hearts, LVADs are increasingly being used to support cardiac function; however, patients receiving LVADs have only a 50% chance of receiving a heart transplant within 1 year of device implantation and face an almost 9% incident risk of stroke per year.^{4, 15} We found that heart transplantation was associated with a significantly decreased risk of stroke in patients with heart failure awaiting heart transplantation even after adjustment for demographics, stroke risk factors and LVAD use in our multivariable analysis. Our results argue that further emphasis on strategies to reduce systolic dysfunction may in turn reduce the burden of cerebrovascular disease. Although efforts to increase the number of heart transplantations are ongoing,^{17, 18} other strategies to reduce systolic dysfunction such as better uptake of proven heart failure therapies or development of novel medications, improved mechanical support, or xenotransplantation may prove worthwhile.^19–21

There are several limitations to our study. First, our findings may be partly attributable to residual confounding since patients who ultimately received a transplant may have been less severely ill in ways that were not documented in these claims data. However, given the strength of the inverse association between transplantation and ischemic stroke, such an unmeasured confounder would need to be very strongly associated with a decreased ischemic stroke risk (HR ~0.2) to fully explain away our findings.²⁶ Second, we lacked information regarding heart failure severity or etiology. Similarly, we lacked information regarding the severity, distribution, or neuroimaging characteristics of stroke. In addition, we lacked information regarding the type of LVAD used – continuous versus pulsatile flow. Third, we lacked data regarding medications and medication adherence, particularly antithrombotic agents. Fourth, our use of administrative data may have led to misclassification of stroke events. However, we used diagnosis code algorithms with high sensitivity and specificity.¹³

In summary, we found that heart transplantation was associated with a decreased risk of stroke among patients with heart failure. These data highlight one important potential benefit of broadening access to heart transplantation.