Left ventricular assist devices (LVAD) have emerged as the standard of care for treating patients with advanced heart failure refractory to maximal medical therapy. Numerous reports have demonstrated improved device durability, increased long-term survival, and decreased incidence of adverse events for both bridge to transplant (BTT) and destination therapy (DT) patients supported with continuous flow LVADs.1–5 However, postoperative stroke remains an important cause of morbidity and decreased quality of life.6–9 In this study, we reviewed our experience with patients on long-term HeartMate II (HM II) LVAD (Thoratec Corp., Pleasanton, CA) support to determine the prevalence and outcome of postoperative stroke at our institution and to identify predictors of its occurrence.
From March 2006 through November 2011, one hundred patients with chronic heart failure underwent implantation of a HM II LVAD with cumulative LVAD support time of 46,486 days. Patients were implanted as BTT in 65 cases and DT in 35 cases. Clinical records were retrospectively reviewed to identify those patients whose postimplant course was complicated by stroke, which was defined as a new global neurologic deficit ascertained in an examination by a neurologist lasting at least 24 hours and accompanied with head computed tomography scan findings of an ischemic or hemorrhagic area of affected brain.
Stroke patients were compared with nonstroke patients for numerous variables, including age, sex, race, etiology of heart failure, height, weight, albumin, diabetes, hypertension, chronic renal insufficiency, dialysis, creatinine, chronic obstructive pulmonary disease, mechanical ventilation, reoperation, history of stroke, hepatic function, inpatient versus outpatient, cardiopulmonary bypass time, whether cross-clamping with cardioplegic arrest was performed, and duration of support. The effect of stroke on survival was assessed. In addition, by performing Cox regression multivariate logistic analysis, independent predictors of postimplant stroke were identified. The procedures followed were in accordance with institutional guidelines, and this review was performed with institutional review board approval.
Postoperative Anticoagulation Protocol and Device Management
All patients were initially anticoagulated with aspirin 81 mg daily, as well as warfarin with a target international normalized ratio (INR) of 2.0–2.5 IU. Heart failure medications typically included a β-blocker, ace inhibitor, and diuretics, as well as sildenafil if there was significant residual pulmonary hypertension.
Patients underwent periodic echocardiograms to evaluate the degree of left ventricular (LV) decompression, aortic ejection, residual mitral regurgitation, position of the interventricular septum, and right ventricular size and function. Device speed was clinically adjusted to optimize flow, peripheral perfusion, organ function, and LV decompression.
Statistical analysis was conducted by a statistician from the Department of Biostatistics at Henry Ford Hospital. Data were represented as frequency distributions and percentages. Values of continuous variables were expressed as a mean ± standard deviation. Continuous variables were compared using independent sample t-tests. Categorical variables were compared using χ2 tests. Using univariate and multivariate Cox multivariate logistic regression analysis, stroke patients were compared with nonstroke patients with respect to clinical demographical variables to assess for possible independent predictors of stroke. For all analyses, a p value of <0.05 was considered statistically significant. Kaplan-Meier analysis was used to calculate survival along with a log-rank p value when comparing groups. Actuarial survival at 1, 3, and 5 years postimplant was calculated by constructing life tables. All data were analyzed using SPSS 11.5 (SPSS Inc., Chicago, IL).
There were 12 (12.0%) patients whose postimplant course was complicated by stroke. This included seven BTT patients and five DT patients, making the overall incidence of stroke 10.8% (7/65) for BTT patients and 14.3% (5/35) for DT patients. Etiology of stroke was embolic in four patients and hemorrhagic in eight patients. Median duration of support at the time of stroke was 340.5 days (range: 4–1,161 days); 281.0 days for embolic and 380.5 days for hemorrhagic strokes (p = 0.028). Table 1 characterizes each stroke with respect to embolic versus hemorrhagic, duration of support at the time of the stroke, anatomic cerebral location affected, and initial neurologic deficit.
Comparison of Patients With Stroke Versus Patients Without Stroke
A comparison of numerous demographical variables between patients with and without stroke is presented in Table 2. Patients with stroke had a significantly higher incidence of diabetes (66.7% vs. 40.9%; p = 0.024), history of previous stroke (16.7% vs. 4.5%; p = 0.046), and use of aortic cross-clamping with cardioplegic arrest during LVAD implantation (50.0% vs. 20.2%; p = 0.034) compared with patients without postoperative strokes.
Anticoagulation at the Time of Stroke
At the time of stroke, 11 of 12 patients were on warfarin. Mean INR at the time of stroke was subtherapeutic in all four patients with embolic strokes (mean: 1.5 ± 0.1 IU; range 1.3–1.6 IU) and supratherapeutic in four of eight patients with hemorrhagic strokes (mean: 3.2 ± 2.2 IU, range: 1.4–7.0 IU; p = 0.024). This included INRs of 1.3, 1.5, 1.5, and 1.6 IU in patients with embolic strokes, as well as INRs of 1.3, 1.4, 1.4, 1.8, 3.0, 3.3, 6.0, and 7.0 IU in patients with hemorrhagic strokes. Regarding antiplatelet therapy, 11 of 12 patients were on 81 mg of daily aspirin at the time of stroke.
Effect of Stroke on Survival
Mortality within 30 days of stroke was 25.0%, with three patients expiring at 8, 14, and 18 days after stroke. The cause of death was refractory vasodilatory shock in one patient and multisystem organ failure in two patients. Among the nine surviving patients, two were transplanted at 5 and 20 months after stroke, six are on ongoing mechanical support, and one died 17 months after stroke from unrelated causes.
Length of Intensive Care Unit and Overall Hospital Stay After Stroke
Stroke patients were in the hospital for 13.3 ± 9.0 days; median of 13 days. Among the nine patients who survived the first 30 days after stroke, five were discharged home and four were discharged to subacute rehabilitation facilities.
Predictors of Stroke
Using Cox multivariate logistic regression analysis, diabetes (OR 6.36; p = 0.029), aortic cross-clamping with cardioplegic arrest (OR 4.75; p = 0.025), duration of support (OR 1.00; p = 0.008), and INR (OR 4.42; p = 0.020) were independent predictors of stroke. There was a trend toward significance for history of stroke (OR 6.25; p = 0.075) (Tables 3 and 4). Age, sex, race, preoperative hepatic function, postoperative bleeding, driveline infection, and right ventricular failure were not significant predictors of postoperative stroke (p = NS).
Freedom from Stroke
There was a significantly greater freedom from stroke in patients who did not undergo aortic cross-clamping with cardioplegic arrest compared with those patients who underwent aortic cross-clamping with cardioplegic arrest as a part of their LVAD implant (p = 0.018; Figure 1). In addition, there was a trend toward significance for greater freedom from stroke in people without diabetes versus with diabetes (p = 0.055; Figure 2).
Relationship of Complete Aortic Cross-Clamping with Cardioplegic Arrest and Postoperative Stroke
Complete aortic cross-clamping with cardioplegic arrest was performed in 23 of 100 (23.0%) patients as a part of their initial LVAD implant. Strokes occurred in six of 23 (26.1%) of these patients compared with six of 77 (7.8%) of patients who did not undergo complete aortic cross-clamping with cardioplegic arrest (p < 0.001). Overall, 50.0% (6/12) of all strokes occurred in patients who underwent complete aortic cross-clamping with cardioplegic arrest. Only one of these six patients underwent an aortic valve procedure and was cross-clamped for this indication. The other five patients/aortas were cross-clamped because of surgeon preference of having a nonbeating heart while performing the LVAD implant. Strokes occurred at a median duration of support of 131.0 days for clamped patients versus 474.5 days for unclamped patients (p = 0.016).
Management of Anticoagulation After Stroke
For the four patients with embolic strokes, the dose of aspirin was increased from 81 to 325 mg daily. In addition, the target INR range was increased from 1.5–2.0 IU to 2.0–2.5 IU. For the eight patients with hemorrhagic strokes, one patient was restarted on 81 mg aspirin only, two patients on 81 mg aspirin and warfarin with a target INR range of 1.5–2.0 IU, and five patients were taken off anticoagulation completely.
Recurrence of Stroke or Other Thromboembolic Events After Modification of Anticoagulation Protocol
There were no recurrences of stroke. In addition, there were no pump thromboses or other thromboembolic events after anticoagulation was modified in any of the 12 stroke patients.
Because of the discrepancy between the number of patients with advanced heart failure and the limited availability of donor organs, the number of LVADs being implanted has steadily increased. In addition, there has been rapid growth in the DT patient population. With advances in technology, improved device design, patient selection, and postoperative management, outcomes for patients on long-term LVAD support have significantly improved.1–5 However, postimplant stroke remains an important complication, with significant associated morbidity and mortality.
In our single-center study of 100 HM II LVAD patients, 65 implanted as BTT and 35 as DT, and the overall incidence of stroke was 12.0%. This included a 10.8% incidence of stroke for BTT patients and a 14.3% incidence for DT patients. The 10.8% incidence of stroke in our BTT patients is similar to data published by other BTT series including Starling et al.1 with an 11% (8% embolic, 2% hemorrhagic, and 1% unknown) incidence of stroke in 169 HM II BTT patients, Pagani et al.2 with an 8% (5% embolic and 3% hemorrhagic) incidence of stroke in 281 HM II BTT patients, and Miller et al.3 with an 8% (6% embolic and 2% hemorrhagic) incidence of stroke in 133 HM II BTT patients. The 14.3% incidence of stroke in our DT patients is also similar and possibly an improvement to the incidence of stroke published by other DT series,4 with an 18% incidence of stroke (8% embolic and 11% hemorrhagic) in 134 patients who underwent HM II implantation as DT.
Stroke adversely affected survival with a 25% 30 day mortality after its occurrence. In addition to its adverse effect on mortality, given the median length of stay of 13 days for the hospital readmission, there was significant additional resource utilization for stroke patients. Furthermore, four of the nine surviving patients were discharged to rehabilitation facilities where they received additional therapy. Follow-up studies are underway to calculate the precise cost of the additional admission along with subacute rehabilitation fees.
Diabetes, aortic clamping, duration of support, and INR values were independent predictors of stroke. More specifically, people with diabetes were 6.36 times more likely to have a stroke than those without diabetes, patients who were cross-clamped were 4.75 times more likely to have a stroke than those who were not clamped, and the odds of a stroke increased by 0.2% with every day of LVAD support endured. Our finding of a higher incidence of stroke among patients with diabetes is consistent with reports from the non-LVAD literature that have also identified people with diabetes as higher risk for stroke than those without diabetes.10–12 The etiology may be related to increased cerebrovascular atherosclerotic burden present in people with diabetes.
It is unclear why complete aortic clamping with cardioplegic arrest compared with partial aortic clamping with a side-biting clamp was associated with a significantly higher incidence of stroke and was an independent predictor of stroke especially because the strokes in the clamped subgroup occurred at a median duration of 131.0 days. It is possible that aortic cross-clamping plays a larger role in early, perioperative strokes, rather than later ones. However, our ability to analyze this point in a statistically valid manner was limited because of a small number of patients requiring subgroup analysis (early vs. late strokes and clamped vs. unclamped). Possible explanations for the increased stroke rate with aortic cross-clamping and cardioplegic arrest include endothelial injury to the aorta created by the cross-clamp with subsequent thrombin deposition, platelet aggregation, and delayed embolization. It is possible that there is greater endothelial injury to the aorta with complete cross-clamping compared with partial clamping because a partial clamp does not directly compress the posterior aspect of the aorta, thus minimizing the area of injury and potential source of embolization. As this series is the first one to evaluate the effect of aortic cross-clamping on postoperative stroke, it is possible that additional clarification of the mechanism of stroke from aortic cross-clamping may arise from a larger sample size of patients who underwent cross-clamping during their LVAD implant. Collaborations with other institutions along with additional data analysis are underway.
Deviation from our institutional anticoagulation protocol also influenced the development of stroke. Our INR goal is generally 2.0–2.5 IU. The INR at the time of stroke was subtherapeutic in each of the four patients with embolic strokes. In addition, among the eight patients with hemorrhagic strokes, four had supratherapeutic INR values at the time of stroke. These data strongly support regulating patient’s anticoagulation more strictly. A subtherapeutic INR is a significant risk factor for an embolic stroke, whereas a supratherapeutic INR is a significant risk factor for a hemorrhagic stroke. It should be noted, however, that four of the patients with hemorrhagic strokes had subtherapeutic INRs at the time of stroke, indicating that there are certainly other risk factors for hemorrhagic stroke other than a supratherapeutic INR. Improvements in anticoagulation regulation may be possible using a multidisciplinary approach that includes more frequent blood draws, as well as additional patient education focused on the impact of dietary intake on the potency of warfarin.
As LVADs become a more widely used and accepted treatment option for patients with refractory, end-stage heart failure, it will be imperative to further improve outcomes and reduce the incidence of postoperative stroke. This complication is a devastating one as it significantly decreases survival and adds significant morbidity. Constructing updated statistical models based on a larger number of patients to more accurately predict the development of stroke may affect both preoperative patient selection and intraoperative technique. In addition, stricter control of postoperative anticoagulation to avoid INR values <1.7 IU or >3.0 IU may significantly impact the incidence of this devastating postoperative complication.
This is a retrospective review, and limitations include potential inaccuracy of data retrieved from medical records. In addition, the number of patients in this study was relatively small, thus limiting the statistical power of the analysis and conclusions. Further studies with additional patients and longer follow-up will be useful and are currently underway.
1. Starling RC, Naka Y, Boyle AJ, et al. Results of the post-U.S. Food and Drug Administration-approval study with a continuous flow left ventricular assist device
as a bridge to heart transplantation: A prospective study using the INTERMACS (Interagency Registry for Mechanically Assisted Circulatory Support). J Am Coll Cardiol. 2011;57:1890–1898
2. Pagani FD, Miller LW, Russell SD, et al.HeartMate II Investigators. Extended mechanical circulatory support
with a continuous-flow rotary left ventricular assist device
. J Am Coll Cardiol. 2009;54:312–321
3. Miller LW, Pagani FD, Russell SD, et al.HeartMate II Clinical Investigators. Use of a continuous-flow device in patients awaiting heart transplantation. N Engl J Med. 2007;357:885–896
4. Slaughter MS, Rogers JG, Milano CA, et al.HeartMate II Investigators. Advanced heart failure treated with continuous-flow left ventricular assist device
. N Engl J Med. 2009;361:2241– 2251
5. Slaughter MS, Pagani FD, Rogers JG, et al.HeartMate II Clinical Investigators. Clinical management of continuous-flow left ventricular assist devices in advanced heart failure. J Heart Lung Transplant. 2010;29(4 suppl):S1–S39
6. John R, Kamdar F, Liao K, Colvin-Adams M, Boyle A, Joyce L. Improved survival and decreasing incidence of adverse events with the HeartMate II left ventricular assist device
as bridge-to-transplant therapy. Ann Thorac Surg. 2008;86:1227–1234; discussion 1234
7. John R, Kamdar F, Liao K, et al. Low thromboembolic risk for patients with the Heartmate II left ventricular assist device
. J Thorac Cardiovasc Surg. 2008;136:1318–1323
8. Slaughter MS, Naka Y, John R, et al. Post-operative heparin may not be required for transitioning patients with a HeartMate II left ventricular assist system to long-term warfarin therapy. J Heart Lung Transplant. 2010;29:616–624
9. Lahpor J, Khaghani A, Hetzer R, et al. European results with a continuous-flow ventricular assist device for advanced heart-failure patients. Eur J Cardiothorac Surg. 2010;37:357–361
10. Jørgensen H, Nakayama H, Raaschou HO, Olsen TS. Stroke
in patients with diabetes. The Copenhagen Stroke
11. Karapanayiotides T, Piechowski-Jozwiak B, van Melle G, Bogousslavsky J, Devuyst G. Stroke
patterns, etiology, and prognosis in patients with diabetes mellitus. Neurology. 2004;62:1558–1562
12. Stöllberger C, Exner I, Finsterer J, Slany J, Steger C. Stroke
in diabetic and non-diabetic patients: Course and prognostic value of admission serum glucose. Ann Med. 2005;37:357–364
Keywords:Copyright © 2014 by the American Society for Artificial Internal Organs
left ventricular assist device; mechanical circulatory support; stroke; complications