Transcatheter aortic valve implantation (TAVI) is invariably associated with the risk of clinically manifest transient or irreversible neurologic impairment. We sought to investigate the incidence and causes of clinically manifest stroke during TAVI. A total of 214 consecutive patients underwent TAVI with the Medtronic-CoreValve System from November 2005 to September 2011 at our institution. Stroke was defined according to the Valve Academic Research Consortium recommendations. Its cause was established by analyzing the point of onset of symptoms, correlating the symptoms with the computed tomography-detected defects in the brain, and analyzing the presence of potential coexisting causes of stroke, in addition to a multivariate analysis to determine the independent predictors. Stroke occurred in 19 patients (9%) and was major in 10 (5%), minor in 3 (1%), and transient (transient ischemic attack) in 6 (3%). The onset of symptoms was early (≤24 hours) in 8 patients (42%) and delayed (>24 hours) in 11 (58%). Brain computed tomography showed a cortical infarct in 8 patients (42%), a lacunar infarct in 5 (26%), hemorrhage in 1 (5%), and no abnormalities in 5 (26%). Independent determinants of stroke were new-onset atrial fibrillation after TAVI (odds ratio 4.4, 95% confidence interval 1.2 to 15.6), and baseline aortic regurgitation grade III or greater (odds ratio 3.2, 95% confidence interval 1.1 to 9.3). In conclusion, the incidence of stroke was 9%, of which >1/2 occurred >24 hours after the procedure. New-onset atrial fibrillation was associated with a 4.4-fold increased risk of stroke. In conclusion, these findings indicate that improvements in postoperative care after TAVI are equally, if not more, important for the reduction of periprocedural stroke than preventive measures during the procedure.
Transcatheter aortic valve implantation (TAVI) is increasingly used to treat patients with aortic stenosis who are considered too high a risk for surgical valve replacement (aortic valve replacement). Despite its clinical benefits, TAVI is invariably associated with the risk of clinically manifest transient or irreversible neurologic impairment. This can be explained by the various catheter and wire manipulations during TAVI that can result in a cerebral embolus but also by cerebral hypoperfusion due to episodes of hypotension during TAVI resulting from—for instance—rapid right ventricular pacing during aortic balloon valvuloplasty. Also gaseous and atherosclerotic microemboli can provoke ischemia and/or occlusion of deep penetrating arteries of the brain, as recently demonstrated. A neurologic deficit can also occur at some point after TAVI for reasons not directly related to the procedure itself, such as is seen in cardiac surgery. The understanding of the pathophysiology or cause of stroke during TAVI could help to determine which preventive strategies during and/or after TAVI will most effectively reduce the stroke rates. We, therefore, sought to elucidate the incidence and causes of stroke in a series of 214 consecutive patients by analyzing the time of symptom onset in relation to the procedure and by correlating the symptoms with the computed tomographic (CT)-detected defects in the brain, in addition to the assessment of independent predictors of stroke.
Methods
The study population consisted of all 214 patients (3 intraprocedural deaths excluded) who underwent transfemoral or transsubclavian TAVI with the Medtronic CoreValve System between November 2005 and September 2011 in the Erasmus Thoraxcenter (Rotterdam, The Netherlands). The patient selection criteria and the methods used for Doppler echocardiography have been previously described in detail. The treatment strategy (TAVI, aortic valve replacement, or medical therapy) was discussed at a joint cardiothoracic surgical and medical conference.
TAVI was performed with the patient under general anesthesia. The first 5 patients underwent TAVI with a 21F delivery catheter that was inserted into the common femoral (n = 4) or subclavian (n = 1) artery after a surgical cut down. All other patients underwent TAVI with an 18F compatible delivery catheter that was inserted into the common femoral artery using an ultrasound-guided Seldinger technique, except for 5 patients who underwent TAVI by way of the left subclavian artery (surgical exposure and closure). Extracorporal support (extracorporeal membrane oxygenation/TandemHeart, CardiacAssist, Pittsburgh, PA) was used in patients with impaired left ventricular function and a suspected increased risk of periprocedural hemodynamic instability. The subsequent phases of the transfemoral TAVI procedure have been described previously.
Patients who were not taking aspirin and/or clopidogrel received a dose of 80 and 600 mg, respectively, the day before TAVI. Patients who were receiving oral anticoagulant therapy were instructed to stop this treatment 3 days before the procedure. Anticoagulant therapy was replaced by enoxaprin until the day before TAVI in patients with a strict indication for anticoagulant therapy. At admission, a full blood examination was performed, including the prothombin time and international normalized ratio (INR).
After insertion of the arterial sheath, a bolus of 70 U/kg IU unfractionated heparin was administered, followed by additional doses to maintain the activated clotting time at 250 to 350 seconds. The activated clotting time was checked every 30 minutes. The activated partial thromboplastin time was checked within 6 hours after the procedure.
After completion of the procedure (percutaneous or surgical closure of the access site), sedation was stopped, followed by extubation. All patients were transferred to the intensive care unit/cardiac care unit for 12 to 24 hours, or longer if clinically indicated. They were then transferred to the medium care unit until hospital discharge. According to the TAVI protocol, rhythm monitoring by telemetry was performed during the hospital stay. All patients received aspirin 80 mg and clopidogrel 75 mg for 6 months. Patients with an indication for oral anticoagulant therapy only received clopidogrel. In these patients, unfractionated heparin was continued after TAVI until adequate INR levels were obtained by acenocoumarol. In-hospital anticoagulant treatment was guided by the prothombin time, INR and activated partial thromboplastin time.
Stroke was defined according to the Valve Academic Research Consortium end point definitions. This implies the following: (1) exclusion of metabolic or toxic encephalopathy or pharmacologic influences explaining the symptoms, in addition to a solely nonfocal neurologic syndrome, (2) execution of a CT study to confirm the clinical diagnosis, (3) the distinction between stroke and transient ischemic attack, and (4) classification of stroke as major or minor according to the degree of disability (modified Rankin score after the procedure and at 30 and 90 days). For patients in whom a modified Rankin score was not documented during the 3 intervals, a detailed chart review was performed to estimate this and accurately classify strokes as major or minor events.
The brain CT scan findings were analyzed using a standard protocol. The cause of stroke was established by (1) analyzing the time of symptom onset, (2) correlating the symptoms with CT-detected defects in the brain, and (3) analyzing the presence of potential coexisting causes of stroke, in addition to multivariate analysis to determine the independent predictors of stroke. Infarcts were categorized as old or new, with the latter further defined as cortical (territorial), cortical watershed, or lacunar infarct.
With respect to the timing of stroke, a distinction was made between stroke that occurred during versus after TAVI. The first was considered directly related to the procedure itself (e.g., due to catheter manipulations or hemodynamic changes) and the second was considered indirectly related to the procedure but not to the procedure itself. Stroke during TAVI was defined if the first symptoms and/or signs were detected ≤24 hours after termination of TAVI. Stroke after TAVI was defined when the first symptoms and/or signs were detected >24 hours after termination of TAVI. The termination of TAVI was defined by the time of vascular closure and hemostasis by either a percutaneous closure device or surgically.
All pre-, intra-, and postprocedural and follow-up data were prospectively collected and entered in a dedicated database as previously described. Porcelain aorta was defined as an extensive circumferential calcification of the thoracic aorta, as assessed by computed tomography and/or fluoroscopy. The blood coagulant status was assessed by collecting the prothombin time, INR, and thrombocyte levels before the procedure. The maximum and minimum activated clotting time levels were documented during the procedure, and the activated partial thromboplastin time was checked within 6 hours after the procedure. Data on red blood cell transfusions were recorded by the institution’s blood bank laboratory and used to determine the corrected hemoglobin decrease within 24 hours after TAVI according to the modified Landefeld equation. In this equation, 1 U of packed red blood cells is considered to represent 1 g/dl of hemoglobin; therefore, the net hemoglobin decrease corresponds to the addition of the number of packed red blood cells to the baseline minus the measured nadir hemoglobin level.
The occurrence and timing of new atrial fibrillation (AF) after TAVI—defined as any episode of AF lasting >30 seconds in patients with no history of chronic/paroxysmal AF—was determined by collecting the baseline and all postoperative 12-lead ECGs and 24-hour telemetry rhythm strips. Follow-up information was prospectively collected during the structured outpatient clinic visits after hospital discharge. In addition, the survival and cause of death was obtained every 6 months by contacting the Dutch Civil Register.
The categorical variables are presented as frequencies and percentages and were compared using the chi-square test or Fisher’s exact test. The normality of distributions was assessed with the Shapiro-Wilk test. Normal and skewed continuous variables are presented as the mean ± SD and median (interquartile range [IQR]), respectively. A comparison of continuous variables was done using Student t tests or Wilcoxon’s rank sum test, when appropriate. Univariate analysis was performed to characterize the patients with and without stroke. Multivariate logistic regression analysis was performed to determine the predictive factors for stroke or transient ischemic attack, taking into account the restricted number of events. Preprocedural AF rhythm was included in the model. A 2-sided p value <0.05 was considered to indicate significance, and all statistical analyses were performed with SPSS software, version 17 (SPSS, Chicago, IL).
Results
The baseline characteristics and procedural details are listed in Tables 1 and 2 . The incidence of stroke was 9% (19 patients) and—in accordance with the Valve Academic Research Consortium criteria—consisted of major stroke in 10 patients (5%), minor in 3 (1%), and transient ischemic attack in 6 (3%).
| Variable | Entire Cohort (n = 214) | No Stroke (n = 195) | Stroke (n = 19) | p Value |
|---|---|---|---|---|
| Age (years) | 80 ± 8 | 80 ± 8 | 82 ± 6 | 0.48 |
| Men | 107 (50%) | 101 (52%) | 6 (32%) | 0.093 |
| Height (cm) | 167 ± 11 | 167 ± 12 | 166 ± 8 | 0.75 |
| Weight (kg) | 74 ± 13 | 74 ± 13 | 76 ± 12 | 0.50 |
| Body mass index (kg/m 2 ) | 26.2 ± 4.1 | 26.1 ± 4.1 | 27.3 ± 4.3 | 0.25 |
| Body surface area (m 2 ) | 1.85 ± 0.19 | 1.85 ± 0.19 | 1.86 ± 0.17 | 0.78 |
| New York Heart Association class III or greater | 175 (82%) | 158 (81%) | 17 (90%) | 0.54 |
| Previous cerebrovascular event | 49 (23%) | 47 (24%) | 2 (11%) | 0.26 |
| Previous myocardial infarction | 51 (24%) | 49 (25%) | 2 (11%) | 0.26 |
| Previous coronary artery bypass graft surgery | 58 (27%) | 55 (28%) | 3 (16%) | 0.25 |
| Previous percutaneous coronary intervention | 56 (26%) | 51 (26%) | 5 (26%) | 1.0 |
| Diabetes mellitus | 50 (24%) | 46 (24%) | 4 (21%) | 1.0 |
| Hypertension | 126 (59%) | 116 (60%) | 10 (53%) | 0.56 |
| Peripheral vascular disease | 26 (21%) | 24 (12%) | 2 (11%) | 1.0 |
| Chronic obstructive pulmonary disease | 60 (28%) | 53 (27%) | 7 (37%) | 0.37 |
| Creatinine | 95 (76–123) | 96 (77–123) | 84 (66–112) | 0.27 |
| Glomerular filtration rate | 57 ± 20 | 57 ± 20 | 58 ± 18 | 0.86 |
| Hemoglobin (g/dl) | 12.3 ± 1.7 | 12.3 ± 1.7 | 12.1 ± 1.2 | 0.62 |
| Thrombocyte count | 225 ± 67 | 223 ± 67 | 244 ± 63 | 0.21 |
| Prothrombin time (s) | 14 ± 8 | 14 ± 8 | 14 ± 3 | 0.89 |
| International normalized ratio | 1.21 ± 1.01 | 1.22 ± 1.05 | 1.16 ± 0.28 | 0.83 |
| Atrial fibrillation | ||||
| All | 64 (30%) | 57 (29%) | 7 (37%) | 0.49 |
| Chronic | 44 (21%) | 38 (20%) | 6 (32%) | 0.24 |
| Paroxysmal | 20 (9%) | 19 (10%) | 1 (5%) | 1.0 |
| Preprocedural rhythm | ||||
| Atrial fibrillation | 48 (23%) | 44 (23%) | 4 (21%) | 1.0 |
| Paced | 10 (5%) | 8 (4%) | 2 (11%) | 0.22 |
| Porcelain aorta | 45 (21%) | 38 (20%) | 7 (37%) | 0.084 |
| Aortic valve area (cm 2 ) | 0.66 ± 0.21 | 0.66 ± 0.22 | 0.61 ± 0.16 | 0.33 |
| Peak velocity | 4.3 ± 0.8 | 4.3 ± 0.8 | 4.2 ± 0.8 | 0.76 |
| Mean aortic gradient | 45 ± 17 | 46 ± 17 | 44 ± 18 | 0.68 |
| Left ventricular ejection fraction ≤35% | 24 (14%) | 23 (14%) | 1 (6%) | 0.70 |
| Aortic regurgitation grade III or greater | 42 (20%) | 34 (17%) | 8 (42%) | 0.016 |
| Mitral regurgitation grade III or greater | 26 (12%) | 23 (12%) | 3 (16%) | 0.71 |
| Logistic European system for cardiac operative risk evaluation | 13.8 (10.0–22.0) | 13.8 (10.0–22.8) | 12.0 (8.4–16.5) | 0.14 |
| Society of Thoracic Surgeon score | 5.0 (3.4–7.5) | 5.0 (3.4–7.3) | 4.3 (3.5–7.5) | 0.96 |
| Antiplatelets | 105 (49%) | 96 (50%) | 9 (47%) | 1.0 |
| Anticoagulants | 67 (32%) | 63 (33%) | 4 (21%) | 0.30 |
| Variable | No Stroke (n = 195) | Stroke (n = 19) | p Value |
|---|---|---|---|
| Procedural results | |||
| Vascular access | |||
| Surgical—femoral artery | 4 (2%) | 0 | 1.0 |
| Surgical—subclavian artery | 6 (3%) | 0 | 1.0 |
| Percutaneous—femoral artery | 185 (95%) | 19 (100%) | 0.25 |
| Circulatory support | 15 (8%) | 3 (16%) | 0.21 |
| Additional interventions during TAVI | |||
| Percutaneous transluminal angioplasty iliac artery | 6 (3%) | 0 | 1.0 |
| Percutaneous coronary intervention | 15 (8%) | 2 (11%) | 0.65 |
| Prosthesis size ⁎ (mm) | |||
| 26 | 59 (30%) | 9 (47%) | 0.13 |
| 29 or 31 | 135 (69%) | 10 (53%) | 0.14 |
| Valve/annulus ratio | 1.15 ± 0.08 | 1.16 ± 0.08 | 0.73 |
| Life-threatening arrhythmia | 9 (5%) | 0 | 1.0 |
| Any complication leading to severe hypotension | 4 (2%) | 0 | 1.0 |
| Highest activated clotting time (s) | 284 ± 87 | 283 ± 64 | 0.97 |
| Lowest activated clotting time (s) | 221 ± 72 | 231 ± 73 | 0.65 |
| Red blood cell transfusions | 1.2 ± 2.2 | 1.3 ± 1.3 | 0.86 |
| Hemoglobin decrease—uncorrected for red blood cell transfusion (g/dl) | 2.0 ± 1.3 | 2.1 ± 1.0 | 0.82 |
| Hemoglobin decrease—corrected for red blood cell transfusion (g/dl) | 3.2 ± 2.6 | 3.4 ± 1.7 | 0.77 |
| Thrombocyte decrease | 60 ± 45 | 60 ± 36 | 0.98 |
| Therapy-specific results | |||
| Postimplantion balloon dilation | 34 (17%) | 1 (5%) | 0.33 |
| Valve dislodgement † | 19 (10%) | 1 (5%) | 1.0 |
| Valve-in-valve implantation | 10 (5%) | 1 (5%) | 1.0 |
| Duration of procedure (min) | 215 ± 75 | 197 ± 83 | 0.32 |
| Postprocedural results | |||
| Activated partial thromboplastin time (s) ‡ | 128 ± 92 | 133 ± 92 | 0.85 |
| Prosthetic-valve associated results | |||
| Permanent atrial fibrillation § | 40 (21%) | 6 (33%) | 0.23 |
| New atrial fibrillation ∥ | 17 (9%) | 5 (26%) | 0.032 |
| New left bundle branch block | 85 (46%) | 6 (32%) | 0.31 |
| New permanent pacemaker | 41 (21%) | 2 (11%) | 0.38 |
| Echocardiography | |||
| Peak velocity | 2.0 ± 0.5 | 2.1 ± 0.7 | 0.59 |
| Mean aortic gradient | 9 ± 4 | 11 ± 6 | 0.19 |
| Aortic regurgitation grade III or greater | 24 (12%) | 3 (16%) | 0.72 |
| Mitral regurgitation grade III or greater | 20 (10%) | 3 (16%) | 0.44 |
⁎ One patient did not receive a valve because of aborted TAVI after failed introduction of 18F sheath.
† In all cases, the valve was recaptured and successfully implanted in a second attempt.
‡ Checked within 6 hours after the procedure.
§ Atrial fibrillation before, during, and after TAVI.
In all patients, except 1, who experienced a subdural hemorrhage (event number 18), the stroke was ischemic ( Table 3 ) and occurred early (≤24 hours after TAVI) in 8 patients (42%) and was delayed (>24 hours, mean 3.5 days after TAVI) in 11 (58%). CT scan analysis of the brain revealed that stroke consisted of a cortical infarct in 8 patients (of which 4 were territorial and 2 were watershed) and a lacunar infarct in 5. In descending order of odds, new AF (odds ratio 4.4, 95% confidence interval 1.2 to 15.6) and baseline aortic regurgitation grade III or greater (odds ratio 3.2, 95% confidence interval 1.1 to 9.3) were identified as independent predictors. New AF occurred in 22 (14%) at a median of 2 days (IQR 1 to 4.5) after TAVI and resolved spontaneously within 12 hours in 8 patients (36%). Seven patients (32%) received pharmacologic treatment (n = 6) or electric (n = 1) conversion. Antithrombotic therapy (aspirin and clopidogrel) without anticoagulant therapy was maintained in 7 patients (36%) in whom the risk of bleeding was considered greater than the risk of thromboembolism. None of the 5 patients with new AF who experienced a stroke had received anticoagulant therapy.
| Event No. | Pt. No. | Clinical Symptoms | CT Analysis | Classification | AF | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Timing(d) | Symptoms | Duration >24 h | Rankin Score 3 Intervals ⁎ | Timing (d) | Infarct Type † | Localization (Hemisphere) | Stroke Type | Ischemic Subtype § | New Onset or Permanent ∥ | ||
| 1 | 7 | 0 | Left-sided hemiparesis, dysarthria-clumsy hand syndrome | No | 2; 0; 0 | 1 | Old lacunar | Left | TIA | Lacunar | No |
| 2 | 16 | 3 | Left-sided hemiparesis, left-sided neglect | Yes | 4; 2; 2 | 3 | New cortical (territorial) | Right | Major stroke | Cortical territorial | Permanent |
| 3 | 17 | 1 ¶ | Right-sided hemianopia, minimal motor aphasia | Yes | 4; 3; 3 | 2 | New cortical (watershed) | Left | Major stroke | Cortical watershed | No |
| 4 | 18 | 4 | Right leg paresis | No | 4; 0; 0 | 7 | Old lacunar | Right + stem | TIA | Uncertain | No |
| 5 | 19 | 1 ¶ | Pure motor right hemiparesis | No | 4; 0; 0 | 1 | Old lacunar | Left | TIA | Lacunar | Permanent |
| 6 | 22 | 5 | Left-sided hemiparesis | No | 4; 0; 0 | 6 | Negative imaging | Na | TIA | Uncertain | New-onset Day 4 |
| 7 | 26 | 1 | Right-sided hemiparesis | Yes | 4; 3; 2 | 2 | Old lacunar (multiple) | Bilateral | Major stroke | Lacunar | No |
| 8 | 44 | 0 | Buccofacial apraxia | Yes | 3; 3; 2 | 0 | Old lacunar | Left | Major stroke | Cortical | Permanent |
| 9 | 53 | 0 | Right-sided hemiparesis, aphasia | Yes | 6; 6; 6 | 0 | Old lacunar; new cortical (watershed) | Left; left | Major stroke | Cortical watershed | Permanent |
| 10 | 54 | 2 | Right-sided hemiparesis | Yes | 2; 0; 0 | 2 | Negative imaging | Na | Minor stroke | Uncertain | Permanent |
| 11 | 64 | 6 | Right-sided hemiparesis, aphasia | Yes | 6; 6; 6 | 6 | Old lacunar; new cortical | Left; left | Major stroke | Cortical territorial | New-onset Day 4 |
| 12 | 86 | 3 | Dysarthria | No | 1; 0; 0 | 3 | Old lacunar | Right | TIA | Lacunar | No |
| 13 | 105 | 6 | Left arm paresis | Yes | 3; 2; 2 | 7 | Old lacunar; old cortical (watershed) | Left; right | Major stroke ‡ | Uncertain | Permanent |
| 14 | 120 | 2 | Left-sided hemiparesis, dysarthria | Yes | 3; 2; 2 | 2 | Old lacunar (multiple) | Left | Major stroke | Lacunar | No |
| 15 | 137 | 5 | Blurry vision, loss of balance | Yes | 4; 4; 3 | 6 | New cortical | Right | Major stroke | Cortical | No |
| 16 | 146 | 0 | Right-sided hemiparesis | Yes | 6; 6; 6 | 1 | New cortical (territorial) | Left | Major stroke | Cortical territorial | New-onset Day 0 |
| 17 | 184 | 1 ¶ | Right arm paresis | No | 2; 0; 0 | 1 | Old cortical | Left | TIA | Uncertain | New-onset Day 0 |
| 18 | 194 | 0 | Right arm paresis, aphasia | Yes | 2; 0; 0 | 6 | Subdural hemorrhage | Left | Minor stroke | Na | New-onset Day 12 |
| 19 | 205 | 2 | Hemianopia | Yes | 2; 2; 1 | 2 | New cortical (territorial) | Right | Minor stroke | Cortical territorial | No |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
