Colombian Experience with Transcatheter Aortic Valve Implantation of Medtronic CoreValve

Abstract

At our institutions, increasing numbers of aortic stenosis patients were not candidates for surgical aortic valve replacement. Accordingly, we initiated the Cali Colombian Transcatheter Aortic Valve Implantation (TAVI) program. From March 2008 through January 2011, 53 consecutive patients (mean age, 79 ± 6 yr; men, 58%) underwent TAVI with the Medtronic CoreValve System, and data were prospectively collected. Our study's endpoints conformed with Valve Academic Research Consortium recommendations. We report our clinical results.

Predicted mortality rates were 25% (interquartile range, 17%–34%) according to logistic EuroSCORE and 6% (interquartile range, 3%–8%) according to the Society of Thoracic Surgeons score. The 30-day mortality rate was 9% (3 intraprocedural deaths, 5 total). The combined 30-day safety endpoint was 30% (major vascular sequelae, 23%; life-threatening bleeding, 12%; myocardial infarction, 4%; major stroke, 4%; and acute kidney injury [stage 3], 2%). Eight patients (15%) required post-implantation balloon dilation and 2 (4%) required valve-in-valve implantation, for a technical device success rate of 77%. Mean peak transvalvular gradient decreased from 74 ± 29 to 17 ± 8 mmHg and mean transvalvular gradient from 40 ± 17 to 8 ± 4 mmHg (both P=0.001). Moderate or severe aortic regurgitation decreased from 32% to 18% (P=0.12) and mitral regurgitation from 32% to 13% (P=0.002). The 1-year survival rate was 81%.

We found that TAVI with the CoreValve prosthesis was safe and feasible, with sustained long-term results, for treating aortic stenosis in patients at excessive surgical risk; nonetheless, serious adverse events occurred in 30% of the patients.

Transcatheter aortic valve implantation (TAVI) is a relatively new catheter-based, minimally invasive procedure performed on a beating heart to treat patients with aortic stenosis (AS) who are considered to be at too high a risk for surgical aortic valve replacement. The procedure has proved to be superior to medical therapy (including aortic balloon valvuloplasty) for such patients.¹ Since the first use of TAVI in 2002,² the number of patients thus treated has increased exponentially: an estimated 30,000 procedures had been performed by 2011, mostly in Europe and Canada.

Transcatheter aortic valve implantation is still evolving from an initially experimental therapy into an established treatment for high-risk patients with AS, and very little experience with TAVI has been reported in Latin America. At our institutions, increasing numbers of AS patients were at high surgical risk because of age or comorbidities. Therefore, we initiated the Cali Colombian TAVI program in 2007. After a nearly 3-year experience in performing TAVI with use of the Medtronic CoreValve® System (Medtronic CV Luxembourg S.a.r.l.; Luxembourg), we report our clinical results in 53 patients.

Patients and Methods

When our TAVI program was formed in November 2007, our multidisciplinary team consisted of 2 interventional cardiologists, a cardiothoracic surgeon, a vascular surgeon, an anesthetist, an imaging specialist, and 2 research nurses. The team took theoretical courses, paid multiple visits to different experienced centers in Europe, and performed 21 TAVI procedures in the presence of a proctor.

Inclusion and Exclusion Criteria

Eligibility criteria for treatment with the CoreValve have been published.^3,4 Our study's inclusion criteria were as follows: aortic valve area, ≤1 cm² or 0.6 cm²/m²; native aortic valve annular sizes, ≤20 mm or ≥27 mm; femoral artery diameter, ≥6 mm; severe symptoms (New York Heart Association [NYHA] functional class III or IV); and age ≥75 years plus a logistic EuroSCORE ≥20% or age ≥65 years with one of the following major complicating factors: liver cirrhosis, severe pulmonary disease (forced expiratory volume in 1 s, <1 L), severe pulmonary hypertension (>60 mmHg), previous cardiac surgery, porcelain aorta, recurrent pulmonary emboli, right ventricular dysfunction, contraindication to open-heart surgery (previous chest radiation), or cachexia (body mass index, <18 kg/m²).

Major exclusion criteria were life expectancy of less than 12 months because of comorbid conditions; an existing bioprosthesis; myocardial infarction within the preceding 14 days; unprotected left main coronary artery stenosis >70%; hemodynamic instability or cardiogenic shock; history of, or active, endocarditis; active peptic ulcer or upper gastrointestinal bleeding within the prior 6 months; active infections requiring current antibiotic therapy, or clinical suspicion of active infection; contraindication to antiplatelet or anticoagulative therapy or contrast media; and hypersensitivity to nitinol.

All patients underwent a structured evaluation consisting of a formal interview, physical examination, laboratory tests, 12-lead electrocardiography, transesophageal echocardiography, and angiographic evaluation of the coronary and peripheral arteries. Thirty-three of the patients also underwent this arterial study by means of multislice computed tomography.

From March 2008 through January 2011, we screened 91 consecutive patients with suspected severe AS for potential TAVI with the CoreValve. Three were referred for surgical aortic valve replacement. Fourteen continued a medical regimen without cardiac intervention: 4 by choice, 4 whose annular dimensions did not match the CoreValve criteria, 3 with nonsevere AS, 1 with unprotected left main coronary artery stenosis >70%, and 2 for unknown reasons. Of 74 patients in whom TAVI was indicated, 14 were hemodynamically unstable and needed aortic balloon valvuloplasty before TAVI. Of these 14 patients, 7 died after valvuloplasty and before TAVI could be performed, and 7 were bridged to TAVI at a median of 38 days after valvuloplasty (interquartile range [IQR], 2–57 d). In total, 21 of the 74 patients died while on the waiting list, so our study population comprised 53 patients.

Interventional Procedure

Details of the device and the procedure have been published.^4,5 In brief, the CoreValve is a trileaflet porcine pericardial tissue valve mounted on a self-expanding nitinol frame and currently available in sizes with 26- and 29-mm inflow diameters. All 53 of our patients were treated with the 3rd-generation CoreValve, which we implanted with use of an 18F disposable delivery catheter inserted into the common femoral artery or subclavian artery via a pigtail approach. We performed closure through a preclose technique with use of a Prostar® XL 10F Percutaneous Vascular Surgical System⁶ (Abbott Vascular, part of Abbott Laboratories; Abbott Park, Ill), or we used surgical closure in the presence of severe circumferential femoral vascular calcification. Valve implantation was performed under fluoroscopic and angiographic guidance. We placed most patients under local anesthesia with sedation but without mechanical ventilation. Patients with a suspected increased risk of developing complications during or immediately after the procedure were placed under general anesthesia with mechanical ventilation.

Data Collection

All endpoints were selected and defined according to the Valve Academic Research Consortium (VARC) recommendations, available in 2010 and published in 2011.⁷ For patients treated before 2011, we applied the VARC definitions, as other investigators have done. The VARC—consisting of representatives of academic research organizations in Europe and the United States and representatives of the European and American societies of Cardiology and Cardiothoracic Surgery—was formed in 2009. The VARC established standardized endpoint definitions for TAVI clinical trials, in order to increase comparability between studies.

Transesophageal echocardiography, including continuous pulsed-wave Doppler study, was performed at a median of 43 days before TAVI (IQR, 32–91 d) and within 7 days after TAVI, to calculate aortic valve area and mean transvalvular gradient in conformity with recommendations from the American Society of Echocardiography.⁸ Paraprosthetic aortic regurgitation was determined in accordance with the following VARC-proposed criteria: jet width in central jets (percentage of left ventricular diameter), jet density, jet deceleration rate (pressure half time, ms), diastolic flow reversal in the descending aorta, and the circumferential extent of paraprosthetic aortic regurgitation.

Endpoint data collected during or immediately after the procedure included death; myocardial infarction; cerebrovascular, vascular, and bleeding sequelae; and acute kidney injury. All cerebrovascular sequelae were diagnosed by a neurologist, with evaluation of such patients daily and then at least once during a later outpatient clinical visit. Full hematologic and chemistry blood samples were collected daily from all patients during the first 3 days, to determine the severity of vascular, bleeding, and renal sequelae. The use of red blood cell transfusions was documented by our institution's blood-bank laboratory. The occurrence of new-onset 3rd-degree atrioventricular (AV) block and the timing of permanent pacemaker implantation were recorded during the patients' hospital stays.

In conformity with VARC recommendations, technical (device success) and hierarchical composite endpoints (combined safety endpoint) were collected. Device success was achieved through 1) successful vascular access, successful delivery and deployment of the device, and successful retrieval of the delivery system; 2) correct positioning of the device in the proper anatomic location with the prosthetic heart valve performing as intended (aortic valve area >1.2 cm² and mean aortic valve gradient <20 mmHg or peak velocity <3 m/s, without moderate or severe paraprosthetic aortic regurgitation); and 3) the need to implant only one valve in the proper anatomic location. The combined safety endpoint was defined as a composite of all-cause death, major stroke, major vascular sequelae, life-threatening or disabling bleeding, acute kidney injury (stage 3), periprocedural myocardial infarction, and repeat surgical or interventional procedures for valve-related dysfunction. Structured follow-up involved the confirmation of vital status by contacting referring hospitals or each patient's family.

Statistical Analysis

Statistical analysis was performed with use of SPSS software version 17.0 (IBM Corporation; Somers, NY). Categorical variables were compared by means of the χ² or Fisher exact test and are presented as numbers and percentages. Normality of distribution for continuous variables, determined by means of the Shapiro-Wilk test, is presented as mean ±SD or as median and IQR. Continuous variables were compared by using the Student t or Wilcoxon rank sum test. The Wilcoxon signed rank test (for continuous variables) and the McNemar test conducted by exact methods (for binomial variables) were used to perform paired comparisons between pre-treatment and post-treatment echocardiographic results. A Kaplan-Meier survival curve was constructed. Two-sided P values <0.05 indicated statistical significance.

Results

Table I shows the baseline characteristics of the 53 patients who underwent TAVI. Their mean age was 79 ±6 years, 40 were in NYHA class III or IV (75%), and 31 were men (58%). The predicted surgical risk was 25% according to logistic EuroSCORE and 6% according to the Society of Thoracic Surgeons score. Left ventricular ejection fraction was <0.35 in 9 of the patients (17%), and mean aortic valve area was 0.69 ±0.19 cm² with a mean transvalvular gradient of 40 ±17 mmHg. Aortic regurgitation and mitral regurgitation were each moderate in 16 patients (30%) and severe in 1 patient (2%).

TABLE I. Baseline Clinical and Echocardiographic Characteristics of the 53 Patients Who Underwent TAVI

Thirty-Day Clinical Outcomes

Transcatheter aortic valve implantation was performed through the femoral artery in 50 patients, and through the left subclavian artery in 3 who had severe peripheral vascular disease. Table II summarizes the clinical outcomes. Five patients died. The 30-day mortality rate was 9% for all-cause death and 6% for cardiovascular death (3 patients). Of note, 4 of the 5 deaths occurred either during the original procedure (n=3) or during cardiac re-intervention (n=1). The intraprocedural deaths were due to coronary obstruction, retroperitoneal hemorrhage, and electromechanical complications, respectively. All deaths occurred in the earlier half of the cohort (March 2008 through November 2009). The technical composite endpoint (device success rate) was 77%. This is explained by paraprosthetic aortic regurgitation in 8 patients, valve-in-valve implantation in 2 patients, and failed vascular closure associated with fatal bleeding in 2 patients. The device success rate in the earlier half of the cohort was 74%, compared with 81% in the later half.

TABLE II. Thirty-Day Cardiovascular and Noncardiovascular Sequelae, Prosthetic Valve-Associated Endpoints, and Therapy-Specific Endpoints in the 53 Patients Who Underwent TAVI*

Cerebrovascular sequelae were diagnosed in 3 patients: major stroke in 2 (on day 1 and day 2, respectively) and a transient ischemic attack in one (on day 30). Vascular sequelae in 13 patients were associated with problematic functioning of the Prostar device in 10 patients, retroperitoneal hemorrhage in 2, and an access-site hematoma leading to a significant drop in hemoglobin in one. All these sequelae were associated with overt bleeding events. In total, there were 4 life-threatening or disabling, 7 major, and 4 minor overt bleeding events within the first 24 hours. Another 2 severe bleeding events occurred more than 24 hours after TAVI: one caused by cardiac tamponade, and one by a fatal retroperitoneal hemorrhage during re-intervention that involved balloon dilation of an underexpanded CoreValve prosthesis.

In 16 patients, a new permanent pacemaker was implanted because of 3rd-degree AV block (15 patients) and 2nd-degree AV block (1 patient). The indication for pacemaker implantation occurred during the procedure in 6 patients, after the procedure in 8, and after discharge from the hospital in 2 (day 10 and day 24).

The combined 30-day safety endpoint of 30% did not change between the 2 halves of the cohort.

Figure 1 shows the periprocedural echocardiographic results. CoreValve implantation resulted in a significant increase in aortic valve area from 0.7 ±0.2 to 2.5 ±0.6 cm² (P <0.001). This in turn reduced the peak transvalvular velocity from 4.2 ±0.8 to 2 ±0.5 m/s and the peak gradient from 74 ±29 to 17 ±8 mmHg (both P <0.001). After TAVI, fewer patients had moderate or severe mitral regurgitation (13% vs 32% at baseline, P=0.002).

Fig. 1 Comparisons before and after transcatheter aortic valve implantation in A) aortic valve area, B) peak transvalvular velocity, C) peak transvalvular gradient, D) mean transvalvular gradient, and E) aortic and mitral regurgitation.

Clinical Follow-Up

Clinical follow-up was available for all patients and ranged from 0 to 16 months (median, 9 mo). During follow-up, 10 patients died at a median of 87 days after TAVI (IQR, 1–358 d). Of these, 5 died after discharge from the hospital at a median of 324 days after TAVI (IQR, 166–414 d). Three in-hospital deaths (60%) and 1 death after hospital discharge (20%) were cardiac-related. At 1 year, the survival rate was 81% (Fig. 2).

Fig. 2 Kaplan-Meier survival curve after transcatheter aortic valve implantation.

Discussion

In our early experience with CoreValve TAVI in a Latin American population, the short-term mortality rate was 9%—comparable to that in previous studies.1,4,9–11 The fact that most of our patients' deaths occurred during the procedure itself is most likely associated with a learning curve; our device success rates improved over time in parallel with lower mortality rates in the second half of the cohort.

Upon enrollment, our high-risk study population faced the poor prognosis associated with standard medical therapy alone. The patients presented with a median EuroSCORE of 25% and poor left ventricular function in 17% of cases. The 9% short-term mortality rate after TAVI (with a decrease to 4% in the second half of the cohort) seems to confirm the safety of this treatment in very high-risk patients. Nonetheless, 30% of the patients experienced a severe adverse event. The frequency of adverse events (the combined 30-day safety endpoint) did not decrease over time and was most often driven by problems at the access site during vascular closure. The importance of a learning curve has been described in previous studies, in which procedural success rates improved as a function of time and were subsequently associated with improved early survival rates.^12,13

Vascular sequelae reportedly occur in 4% to 32% of patients and are associated with a 2- or 3-fold higher mortality rate.^14-16 In our study, all vascular sequelae were associated with a bleeding component (fatal in 2 patients) and were most often due to problematic functioning of the Prostar device. For these reasons, we currently prefer a more controlled and limited surgical cutdown to the femoral artery, with exposure of its surrounding tissues when peripheral calcification is present at the site of access. Also, the Prostar device was originally designed for puncture holes up to 10F in size.¹⁷ Therefore, it seems reasonable to perform a limited surgical cutdown, because “off-label” application of the Prostar for the CoreValve requires the insertion of a large 18F sheath in the femoral artery.¹⁸ If we are at all uncertain about hemostasis, we will routinely perform follow-up angiography to pinpoint contrast-medium leakage that might not be clinically visible at the end of the procedure. Given the importance of adequately preventing and managing vascular and bleeding sequelae during TAVI, it is prudent to visit institutions that have expertise in percutaneous femoral closure techniques before embarking upon a percutaneous transfemoral TAVI program.

In our study, a mean of 1.6 ±2.3 units of red blood cells per patient were transfused, and 28 of the patients required 1 or more units while hospitalized; this is in accordance with previous reports.^19,20 Red blood cell transfusion is a predictor of acute kidney injury after TAVI, along with baseline renal dysfunction and contrast-medium administration during the procedure.^19-23 Stage 1, 2, or 3 acute kidney injury occurred in 17% of our patients, similar to findings of 12% to 28% in previous studies.^19-23 Measures to reduce the risk of acute kidney injury include strict application of a pre-hydration protocol preprocedurally when contrast-enhanced multislice computed tomography and left-sided heart catheterization are performed, and maintaining sufficient time between these imaging procedures and TAVI itself. Renal damage can be avoided by reducing contrast-medium administration and red blood cell transfusions, and by controlling hemodynamic status during and after the procedure. Finally, patients with acute kidney injury require close periprocedural monitoring of renal function and vital signs (heart rate, blood pressure, and urinary production), given their 2- to 3-fold higher risk of short- and long-term death.^19-21

Although TAVI can be safe in very high-risk patients with AS, there may be an increased risk of perioperative stroke and encephalopathy during the immediate postprocedural period. The reported prevalence of stroke after TAVI ranges between 1% and 10%; however, the frequency of encephalopathy has rarely been documented.^10,24 Two of our patients had major ischemic strokes within 2 days postprocedurally, immediately after recovery from general anesthesia. Another patient had a transient ischemic attack on day 30, after hospital discharge. Encephalopathy, defined as delirium, coma, or seizures at any time during the postprocedural period, occurred in 4 patients (8%). Stroke and encephalopathy reportedly share similar pathophysiologic mechanisms, including microembolic formation or hypoperfusion induced by the procedure itself or occurring secondary to persistent atrial fibrillation. Well-established predictive factors after open-heart surgery include advanced age, hypertension, diabetes mellitus, and peripheral vascular disease²⁵—all common comorbid conditions in our cohort. Existing patient-related predictive factors (such as septal wall thickness, noncoronary cusp thickness, and existing right bundle branch block) can be distinguished from procedure-related factors (including depth of valve implantation within the left ventricular outflow tract, prosthesis expansion after implantation, and type of prosthesis).

Conduction abnormalities frequently occur during TAVI, and with use of the CoreValve in particular. After CoreValve implantation, the prevalence of new left bundle branch block has ranged from 29% to 65%, 3rd-degree AV block from 15% to 44%, and permanent pacemaker implantation from 9% to 49%.^4,26–29 In comparison, after implantation of the SAPIEN valve (Edwards Lifesciences LLC; Irvine, Calif), the corresponding ranges in prevalence have been 6% to 18%, 0 to 27%, and 0 to 27%.^1,30–32 A new permanent pacemaker was implanted in 16 of our patients (30%) during hospitalization or within 30 days after TAVI. The new conduction abnormality for which a new permanent pacemaker was implanted occurred during TAVI in 6 of these patients and after TAVI in the others. In view of the large number of patients who developed 3rd-degree AV block after the procedure, we have substantially lowered our threshold for implanting a pacemaker.

From a technical standpoint, TAVI significantly increased aortic valve area and reduced the transvalvular gradient. The subsequent reduction in afterload may explain our observed reduction in severe mitral regurgitation, from 32% at baseline to 13% after TAVI. These findings suggest that coexisting severe mitral regurgitation—currently a contraindication for CoreValve implantation, according to the manufacturer's guidelines—might not always preclude TAVI in high-risk AS patients. Dedicated prospective echocardiographic studies are warranted to determine which patients with combined AS and severe mitral regurgitation might benefit from TAVI.

In our study, actuarial survival at 1 year was 81% (10 deaths), similar to that in the most recent registries.^1,10,33 Five of the deaths can be attributed to the procedure itself, whereas the other 5 patients died approximately 1 year after TAVI and predominantly of noncardiovascular causes. This suggests that short-term survivors of TAVI have a more favorable prognosis than do patients remaining on medical therapy, with its annual mortality rates of 25%. We are disappointed that many of our patients could not be offered TAVI, chiefly because of limited patient accessibility to the treatment. In addition, aortic balloon valvuloplasty does not seem to meet the clinical needs of high-risk AS patients. In our patients, the mortality rate after balloon valvuloplasty was 50%.

Limitations

The data used for this study were prospectively collected, but the analyses were performed on a relatively small sample of 53 patients. Therefore, this study does not permit firm conclusions, despite the fact that the direction of outcomes is in accordance with those of previous reports.

Conclusion

Transcatheter aortic valve implantation with the Medtronic CoreValve System in this Latin American population was associated with promising results, as evidenced by a short-term mortality rate of 9% and sustained 1-year results. Nevertheless, 30% of the patients experienced a severe adverse event, which is most likely explained by the nature of TAVI, the baseline risk of the patients under treatment, and the learning curve of the medical personnel.