Summary
Background
Patient selection for transcatheter aortic valve implantation (TAVI) remains a major concern. Indeed, despite promising results, it is still unclear which patients are most and least likely to benefit from this procedure.
Aims
To identify predictors of 6-month poor clinical outcomes after TAVI.
Methods
Patients who were discharged from our institution with a transcatheter-implanted aortic valve were followed prospectively. Our population was divided into two groups (‘good outcomes’ and ‘poor outcomes’) according to occurrence of primary endpoint (composite of all-cause mortality, all stroke, hospitalizations for valve-related symptoms or worsening heart failure from discharge to 6 months or 6-month New York Heart Association functional class III or IV). Patient characteristics were studied to find predictors of poor outcomes.
Results
We included 163 patients (mean age, 79.9 ± 8.8 years; 90 men [55%]; mean logistic EuroSCORE, 18.4 ± 11.4%). The primary endpoint occurred in 49 patients (mean age, 83 ± 5 years; 31 men [63%]). By multivariable analysis, atrial fibrillation (odds ratio [OR] 3.94), systolic pulmonary artery pressure ≥ 60 mmHg (OR 7.56) and right ventricular dysfunction (OR 3.55) were independent predictors of poor outcomes, whereas baseline aortic regurgitation ≥ 2/4 (OR 0.07) demonstrated a protective effect.
Conclusion
Atrial fibrillation, severe baseline pulmonary hypertension and right ventricular dysfunction (i.e. variables suggesting a more evolved aortic stenosis) were predictors of 6-month poor outcomes. Conversely, baseline aortic regurgitation ≥ 2/4 showed a protective effect, which needs to be confirmed in future studies. Our study highlights the need for a specific ‘TAVI risk score’, which could lead to better patient selection.
Résumé
Contexte
La sélection des patients pour l’implantation d’une valve aortique transcathéter (TAVI) demeure un challenge clinique. En effet, malgré des résultats prometteurs, il reste difficile de savoir quels patients sont les moins susceptibles de tirer bénéfice de cette procédure.
Objectif
Notre objectif était d’identifier des facteurs prédictifs d’un mauvais résultat 6 mois après TAVI.
Méthodes
Nous avons prospectivement suivi les patients sortis de l’hôpital avec une valve aortique implantée par voie transcathéter. Notre population a été divisée en 2 groupes, « bon résultat » et « mauvais résultat », en fonction de la survenue du critère primaire qui était un critère composite des décès toutes causes, des accidents vasculaires cérébraux, des hospitalisations pour insuffisance cardiaque ou symptômes en rapport avec la valve entre la sortie de l’hospitalisation et le suivi à 6 mois ou une classe fonctionnelle New York Heart Association III ou IV à 6 mois. Les caractéristiques des patients ont été étudiées afin de déterminer des facteurs prédictifs de mauvais résultat.
Résultats
Cent soixante-trois patients consécutifs (âge moyen : 79,9 ± 8,8 ans ; 90 hommes [55 %]) ont été inclus. L’EuroSCORE logistique moyen était de 18,4 ± 11,4 %. Quarante-neuf patients ont présenté le critère primaire. En analyse multivariée, la fibrillation atriale (OR 3,94), une pression artérielle pulmonaire systolique ≥ 60 mmHg (OR 7,56), une dysfonction ventriculaire droite (OR 3,55) étaient des facteurs prédictifs indépendants de mauvais résultat alors que l’insuffisance aortique préopératoire ≥ 2/4 (OR 0,07) présentait un effet protecteur.
Conclusion
La fibrillation atriale, une pression artérielle pulmonaire systolique ≥ 60 mmHg et une dysfonction ventriculaire droite, des variables évoquant un rétrécissement aortique plus évolué, étaient des facteurs prédictifs de mauvais résultat à 6 mois après TAVI. À l’inverse, une insuffisance aortique préopératoire ≥ 2/4 présentait un effet protecteur qui doit être confirmé dans des études futures. Notre étude souligne la nécessité de développer un score de risque spécifique du TAVI qui pourrait améliorer la sélection des patients.
Introduction
Aortic stenosis (AS) is the most common valvular disease, with an increasing incidence in the elderly population . Transcatheter aortic valve implantation (TAVI) was developed as an alternative to surgical aortic valve replacement (SAVR) in patients at prohibitive surgical risk. Several registries showed functional improvement in patients with severe symptomatic AS treated with TAVI. TAVI demonstrated a 2-year survival advantage over medical therapy in inoperable patients and non-inferiority against SAVR in high-risk patients ; it is now the standard of care for inoperable patients and a valid alternative to surgery for many high-risk patients .
Despite these promising results, a significant proportion of patients either die or have no functional benefits within the first months after TAVI . Numerous predictors of mortality have been identified, such as postprocedural aortic regurgitation (AR) , chronic obstructive pulmonary disease , chronic kidney disease, pulmonary hypertension and postprocedural complications .
Moreover, recently, postprocedural AR and severe mitral regurgitation (MR) were identified as independent predictors of poor treatment response . Nonetheless, data on predictors of functional outcomes after TAVI are scarce. Yet, given that this technique is generally intended for elderly patients, symptomatic improvement is as critical as the increase in life expectancy. A risk score to identify those patients who are least likely to benefit from TAVI should further improve the selection of TAVI candidates.
The goal of this prospective study was to identify predictors of 6-month poor outcomes after TAVI, defined as the clinical components of ‘clinical efficacy’, as outlined in the recommendations of the Valve Academic Research Consortium .
Methods
Patients
Patients with severe and symptomatic AS (effective orifice area [EOA] ≤ 1 cm 2 ) who underwent TAVI at our institution were prospectively enrolled. Exclusion criteria were death during the procedure or subsequent hospitalization, conversion to surgery or unsuccessful implantation (defined as impossibility to deliver and deploy a valve into the proper location for anatomical reasons). Before TAVI, these patients underwent an evaluation, which included a physical examination, blood tests, transthoracic and transoesophageal echocardiography (TTE and TEE, respectively) and computerized tomography. Indications, contraindications and anatomical requirements for TAVI have been described previously . SAVR risk of mortality was estimated using the logistic EuroSCORE and the Society of Thoracic Surgeons (STS) risk score . Finally, TAVI indication was retained by a multidisciplinary ‘Heart Team’ based on the evaluation cited above. Patients were followed on-site before discharge, 1 month after implantation and either on-site or by their cardiologist 6 months after TAVI. Follow-up information was also obtained by telephone contact with deceased patients’ physicians. Patients gave written informed consent before participation. The study was approved by the local ethics committee.
Endpoints
The primary endpoint was the clinical components of ‘clinical efficacy’ (i.e. a composite of all-cause mortality, all stroke [disabling and non-disabling], hospitalizations for valve-related symptoms or worsening heart failure from discharge to 6 months or a 6-month New York Heart Association [NYHA] class III or IV). Secondary endpoints were clinical efficacy, as defined in the recommendations of the Valve Academic Research Consortium (clinical components or valve-related dysfunction, i.e. mean aortic valve gradient ≥ 20 mmHg, EOA ≤ 0.9–1.1 cm 2 and/or Doppler velocity index < 0.35 m/s and/or moderate or severe prosthetic valve regurgitation), and 6-month all-cause mortality. The cohort was subsequently divided into two groups: ‘good outcomes’ and ‘poor outcomes’, according to the occurrence of the primary endpoint.
Atrial fibrillation (AF) was defined as any history of AF regardless of type of arrhythmia or presence of AF on at least one electrocardiogram during hospitalization for the preoperative assessment or the day before TAVI. Coronary artery disease was defined as presence of lesions with ≥ 50% diameter stenosis on pre-TAVI angiography and/or previous treatment with percutaneous coronary intervention or coronary artery bypass grafting. Complications were defined according to the recommendations of the Valve Academic Research Consortium .
Study devices and procedures
The two CE-approved prostheses and implantation techniques have been described previously . The procedure was performed in a catheterization laboratory in a sterile environment by at least two interventional cardiologists, a cardiac surgeon and an anaesthesiologist. The choice of whether to use local or general anaesthesia was left at the discretion of the anaesthesiologist in charge of the patient. The type of anaesthesia used was not recorded routinely in our database, but was known for 81% ( n = 132) of patients, 66% of whom ( n = 87) underwent local anaesthesia. TEE was used for transapical cases, to accurately define the apical surgical access site. Fluoroscopy was used for valve positioning in all cases, with the help of TEE guidance only in transapical cases.
Echocardiography
TTE was performed according to the American Society of Echocardiography guidelines by an experienced echocardiographer using a digital ultrasound scanner (Vivid7, GE Healthcare, Little Chalfont, UK; or iE33, Philips Healthcare, Andover, MA, USA).
In apical five-chamber view, peak and mean pressure gradients across the aortic valve were calculated using the Bernoulli equation. EOA was calculated using the continuity equation.
A multiparametric approach with both semiquantitative and quantitative variables was used to grade valvular regurgitation on a scale from 0 to 4, with higher grades indicating greater severity (0, no; 1, mild; 2, moderate; 3/4, severe). Baseline and postprocedural AR were graded in accordance with the European Society of Cardiology guidelines for native valves . However, given the frequent eccentric and irregular jet of postprocedural AR, we also gave a heavy weighting to the circumferential extent of prosthetic AR in parasternal short-axis view, to provide an integrated assessment of postprocedural AR . Thresholds were as follows: none, no regurgitant colour flow; mild extent, < 10%; moderate extent, 10–29%; severe extent, ≥ 30%. Before TAVI, we used TEE to measure the annulus diameter accurately and sometimes to grade AR or MR when TTE was not conclusive.
Pulmonary hypertension was defined as systolic pulmonary artery pressure (sPAP) ≥ 40 mmHg at rest, estimated using tricuspid regurgitation (TR) velocity . Right atrial pressure was assessed using inferior vena cava diameter (in its long axis) and inspiratory collapse in the subcostal view : a diameter ≤ 21 mm and a collapse > 50% with a sniff were used as cut-offs for normal right atrial pressure (i.e. 3 mmHg, range 0–5 mmHg), whereas a diameter > 21 mm and a collapse < 50% defined high right atrial pressure (15 mmHg, range 10–20 mmHg). In indeterminate cases, in which the inferior vena cava diameter and collapse did not fit these definitions, an intermediate value of 8 mmHg (range 5–10 mmHg) was used. Right ventricular (RV) function was assessed in apical four-chamber view using tricuspid annular plane systolic excursion measured by M-mode, with a reference value for impaired RV systolic function of < 16 mm and an RV peak systolic velocity of the tricuspid annulus measured by tissue Doppler, with a value of ≥ 10 cm/s defining normal RV function . Left ventricular ejection fraction (LVEF) was measured by Simpson’s method from the four- and two-chamber views . Left atrial end-systolic area was measured from the four-chamber apical view. LV end-diastolic and end-systolic diameters and end-diastolic septal thickness were measured by M-mode from parasternal views.
TTE was performed the day before TAVI, before discharge and 1 month and 6 months after TAVI.
Blood tests
Venous blood samples were obtained on the day before TAVI to determine concentrations of N-terminal pro B-type natriuretic peptide and serum creatinine. The estimated glomerular filtration rate was calculated using the abbreviated Modification of Diet in Renal Disease Study equation. Kidney disease was defined as moderate when the glomerular filtration rate was between 30 and 59 mL/min/1.73 m 2 and as severe when < 30 mL/min/1.73 m 2 .
Statistical analysis
Numeric values are expressed as mean ± standard deviation. Normality was tested using the Kolmogorov-Smirnov test. Continuous variables were compared using the unpaired t -test or the Mann-Whitney U test, as appropriate. Chi-square analysis or Fisher’s exact test was used to compare categorical variables. Patient characteristics were evaluated for poor outcomes. All baseline variables with a P value ≤ 0.2 in univariate analysis were entered into an ascending stepwise multivariable logistic regression analysis to identify independent predictors of poor outcomes and into an ascending stepwise Cox multivariable analysis to identify predictors of all-cause mortality. The likelihood ratio statistic was used at each step to define which variable should be included in or excluded from the model. Variables with a P value < 0.05 were added to or remained in the model, whereas variables with a P value ≥ 0.1 were removed. Results are presented as odds ratios (ORs) and hazard ratios (HRs). A P value ≤ 0.05 was considered significant. All probability values reported are two-sided. Statistical analysis was performed with the use of SPSS 21.0 software (SPSS, Inc., Chicago, IL, USA).
Results
Patients
From January 2009 to June 2012, 514 consecutive patients with severe and symptomatic AS were referred to our institution for pre-TAVI evaluation. After meetings of the Heart Team, TAVI indication was confirmed in 180 patients who underwent the procedure from February 2009 to July 2012. A total of 17 patients died during the procedure ( n = 3) or the initial hospitalization ( n = 6), were converted to surgery ( n = 4: two annulus ruptures; two embolizations of the prosthesis in the left ventricle) or had unsuccessful implantation ( n = 6: five non-fatal vascular access complications, one insufficient distance between valvular plane and a circumflex artery with an anomalous origin from the right sinus of Valsalva) and thus were excluded. Our study cohort included 163 surviving patients ( Fig. 1 ). No patient was lost to follow-up.
The mean age of the study patients was 79.9 ± 8.8 years, 90 patients (55.2%) were men, the mean logistic EuroSCORE was 18.4 ± 11.4%, the mean STS risk score was 5.8 ± 3.1%, 118 patients (72.4%) were NYHA functional class III or IV, 86 patients (52.8%) had a history of acute heart failure and 71 patients (44%) had AF. The baseline characteristics of the study population are summarized in Table 1 .
| Characteristics | All patients | Good outcomes | Poor outcomes | P |
|---|---|---|---|---|
| ( n = 163) | ( n = 114) | ( n = 49) | ||
| Age (years) | 79.9 ± 8.8 | 78.6 ± 9.7 | 82.8 ± 5.0 | 0.01 |
| Men | 90 (55.2) | 59 (51.8) | 31 (63.3) | 0.24 |
| Body surface area (m 2 ) | 1.78 ± 0.2 | 1.77 ± 0.2 | 1.81 ± 0.3 | 0.63 |
| Logistic EuroSCORE (%) | 18.4 ± 11.4 | 17.4 ± 10.6 | 20.7 ± 12.7 | 0.12 |
| STS risk score (%) | 5.8 ± 3.1 | 5.4 ± 3.0 | 6.0 ± 3.0 | 0.002 |
| NYHA class III or IV | 118 (72.4) | 76 (66.7) | 42 (85.7) | 0.02 |
| Angina pectoris | 32 (19.6) | 25 (21.9) | 7 (14.3) | 0.36 |
| Syncope | 21 (12.9) | 19 (16.7) | 2 (4.1) | 0.04 |
| Previous acute heart failure | 86 (52.8) | 57 (50.0) | 29 (59.2) | 0.36 |
| Clinical history | ||||
| CAD | 85 (52.1) | 64 (56.1) | 21 (42.9) | 0.17 |
| Previous PCI | 24 (14.7) | 18 (15.8) | 6 (12.2) | 0.73 |
| Previous BAV | 29 (17.8) | 16 (14.0) | 13 (26.5) | 0.09 |
| Previous CABG | 28 (17.2) | 21 (18.4) | 7 (14.3) | 0.68 |
| Previous SAVR | 3 (1.8) | 2 (1.8) | 1 (2.0) | 1.0 |
| Cerebrovascular disease | 24 (14.7) | 17 (14.9) | 7 (14.3) | 1.0 |
| PVD | 32 (19.6) | 26 (22.8) | 6 (12.4) | 0.18 |
| Porcelain aorta | 11 (6.7) | 8 (7.0) | 3 (6.1) | 1.0 |
| AF | 71 (43.5) | 37 (32.5) | 34 (69.4) | < 0.0001 |
| Chest wall irradiation | 21 (12.9) | 19 (16.7) | 2 (4.1) | 0.04 |
| Hypertension | 108 (66.3) | 69 (60.5) | 39 (79.6) | 0.03 |
| Diabetes mellitus | 27 (16.6) | 20 (17.5) | 7 (14.3) | 0.78 |
| COPD | 63 (38.9) | 41 (36.3) | 22 (44.9) | 0.39 |
| Chronic kidney disease | 0.47 | |||
| Moderate | 66 (40.5) | 43 (37.7) | 23 (46.9) | |
| Severe | 6 (3.7) | 5 (4.4) | 1 (2.0) | |
| NT-proBNP (pg/mL) | 4281.3 ± 4378.4 | 4329.6 ± 4724.1 | 4172.1 ± 3518.2 | 0.36 |
Procedural outcomes
The aortic valve prosthesis was inserted using the retrograde femoral artery approach ( n = 132), the subclavian artery approach ( n = 10), the transapical approach ( n = 12) or the transaortic approach ( n = 9). The implanted prosthesis was an Edwards SAPIEN ( n = 8), an Edwards SAPIEN XT ( n = 91) or a Medtronic CoreValve ( n = 64). Valve size was 23 mm ( n = 32), 26 mm ( n = 63) or 29 mm ( n = 4) for the Edwards devices and 26 mm ( n = 20), 29 mm ( n = 36) or 31 mm ( n = 8) for the Medtronic CoreValve.
The mean total procedural time was 96 ± 31 minutes and the mean contrast agent volume was 238.8 ± 92.7 mL. Valve embolization in the aorta was observed in two cases and was managed with implantation of a second prosthesis. Acute kidney injury stage 2 or 3 arose in 10 patients (6.1%), including one who required temporary dialysis. Sixteen patients (9.8%, 15 CoreValve) received a new permanent pacemaker. Procedural outcomes are summarized in Table 2 .
| Variables | All patients | Good outcomes | Poor outcomes | P |
|---|---|---|---|---|
| ( n = 163) | ( n = 114) | ( n = 49) | ||
| Valve type | 0.03 | |||
| Edwards (SAPIEN and SAPIEN XT) | 99 (60.7) | 76 (66.7) | 23 (46.9) | |
| Medtronic CoreValve | 64 (39.2) | 38 (33.3) | 26 (53.1) | |
| Valve diameter | 0.14 | |||
| 23 mm | 32 (19.6) | 23 (20.2) | 9 (18.4) | |
| 26 mm | 83 (50.9) | 62 (54.4) | 21 (42.9) | |
| 29 mm | 40 (24.5) | 26 (22.8) | 14 (28.6) | |
| 31 mm | 8 (4.9) | 3 (2.6) | 5 (10.2) | |
| Vascular access | 0.88 | |||
| Transfemoral | 132 (80.9) | 92 (80.7) | 40 (81.6) | |
| Subclavian | 10 (6.1) | 6 (5.3) | 4 (8.2) | |
| Transapical | 12 (7.4) | 9 (7.9) | 3 (6.1) | |
| Transaortic | 9 (5.5) | 7 (6.1) | 2 (4.1) | |
| Procedural time (minutes) | 96 ± 31 | 97 ± 34 | 95 ± 24 | 0.90 |
| Total amount of contrast agent (mL) | 238.8 ± 92.7 | 233.2 ± 85.1 | 252.8 ± 109.0 | 0.25 |
| Need for second valve | 2 (1.2) | 1 (0.9) | 1 (2.0) | 0.51 |
| Hospital stay (days) | 9.2 ± 5.5 | 8.4 ± 4.6 | 11.0 ± 6.8 | 0.001 |
| ICU stay (days) | 3.5 ± 2.1 | 3.2 ± 1.9 | 4.2 ± 2.6 | 0.03 |
| Bleeding | ||||
| Life-threatening or disabling | 4 (2.5) | 2 (1.8) | 2 (4.1) | 0.58 |
| Major | 34 (20.9) | 24 (21.1) | 10 (20.4) | 1.0 |
| Myocardial infarction | 3 (1.8) | 3 (2.6) | 0 (0.0) | 0.55 |
| Stroke | 4 (2.5) | 3 (2.6) | 1 (2.0) | 1.0 |
| Major vascular complication | 16 (9.8) | 10 (8.8) | 6 (12.2) | 0.57 |
| Acute kidney injury, RIFLE stage 2 or 3 | 10 (6.1) | 3 (2.6) | 7 (14.3) | 0.009 |
| Need for permanent pacemaker implantation | 16 (9.8) | 6 (5.3) | 10 (20.4) | 0.007 |
| Postoperative treatment | ||||
| Aspirin | 143 (87.7) | 102 (90.3) | 41 (83.7) | 0.35 |
| Clopidogrel | 93 (57.1) | 74 (65.5) | 19 (38.8) | 0.003 |
| Vitamin K antagonists | 56 (34.4) | 29 (25.7) | 27 (55.1) | 0.0006 |
| Diuretics | 94 (57.7) | 61 (53.9) | 33 (67.3) | 0.16 |
| Beta-blockers | 87 (53.4) | 60 (53.6) | 27 (55.1) | 0.99 |
| ACE inhibitors/ARBs | 80 (49.1) | 62 (54.9) | 18 (36.7) | 0.05 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
