This was a single-center retrospective study of consecutive patients who underwent isolated SAVR for severe aortic stenosis (functional area <1.0 cm2 or mean gradient >40 mmHg) in 2005 and 2009 (respectively two years before and two years after implementation of the TAVI program).

Patients’ demographic, clinical and echocardiographic data were collected from medical and hospital records. Surgical risk was estimated for each patient based on the logistic European System for Cardiac Operative Risk Evaluation (EuroSCORE), using the online calculator. Surgical procedures were classified as elective, urgent or emergency. Elective patients were scheduled for surgery and admitted accordingly, while non-elective patients were hospitalized as a result of clinical instability and/or life-threatening situations (such as congestive heart failure [HF], low cardiac output and cardiogenic shock) and were operated either immediately (emergency surgery) or during the same hospitalization (urgent surgery).

Operative mortality was defined as all-cause death within 30 days of surgery (including during hospital stay). Operative morbidity was defined as the occurrence during the post-operative period of any of the following: death, prolonged (>24 hours) invasive mechanical ventilation, severe arrhythmias (ventricular fibrillation or recurrent ventricular tachycardia, new-onset atrial flutter or fibrillation, or bradyarrhythmia requiring pacemaker), myocardial infarction (elevation of serum troponin I to levels five times higher than pre-surgery values), severe HF (NYHA class III-IV), moderate to large pericardial or pleural effusion, pneumonia, pyelonephritis, acute renal failure (serum creatinine >2.0 mg/dl or twice pre-surgery values or need for de novo dialysis), stroke, major bleeding (resulting in hemodynamic compromise or need for transfusion of more than 2–3 units of red cells or urgent surgical intervention), cardiac reoperation, mediastinitis or surgical wound infection.

One-year follow-up, achieved in 94% of patients, was based on data from medical consultations, urgent hospitalizations and telephone interviews. The following events were recorded: all-cause and cardiac death, myocardial infarction, stroke, rehospitalization for cardiac cause, cardiac re-operation, and thromboembolic and bleeding complications.

During the two years analyzed in the study, 214 patients underwent isolated SAVR; of these 103 were operated two years before implementation of the TAVI program and 111 two years after, an increase of 8% in the volume of surgical procedures.

The mean age of the population was 70±9.9 years (range 35–86), and most patients were women (56.1%). The most prevalent cardiovascular risk factors were hypertension (72%), dyslipidemia (63%), obesity (39%) and diabetes (27%). Almost half (47%) had at least one comorbidity, the most common being chronic obstructive pulmonary disease (COPD) (18%), significant coronary artery disease (15%), anemia (15%) and atrial fibrillation or flutter (12%). Of the 32 patients with significant coronary artery disease, eight had previously undergone percutaneous (n=7) or surgical (n=1) revascularization; one underwent single-vessel angioplasty in the post-operative period. The remainder had not been revascularized for the following reasons: absence of ischemia and/or documented myocardial viability, distal lesions, lesion in a minor branch, or a vascular bed distal to a minor lesion. Half the patients had significant symptoms at the time of surgery (41% with NYHA class III-IV HF, and 9% with CCS class III-IV angina). In hemodynamic terms, mean aortic functional area was 0.6±0.16 cm2 (range 0.2–1.2 cm2) and mean valve gradient was 60±16.9 mmHg (range 20–130 mmHg). Most patients had preserved ejection fraction (mean 55±8.9%, range 20–79%). Around 13% of patients (n=27) had associated mitral valve disease, including four with grade II/IV mitral stenosis, 22 with grade II/IV mitral regurgitation and one with grade III/IV, predominantly functional. Echocardiographic signs suggestive of pulmonary hypertension (pulmonary artery systolic pressure >40 mmHg higher than central venous pressure) were observed in 16% of patients. The mean logistic EuroSCORE was 7.5±6.4% (range 1.5–44.6), reflecting a population with high surgical risk (EuroSCORE >6).

Patients operated in the post-TAVI period were older, with a higher proportion above the 75th percentile for age (18.4% of patients operated pre-TAVI were aged ≥77 years vs. 37.8% post-TAVI, p=0.006). They also tended to have more comorbidities (40.8% pre-TAVI vs. 52.3% post-TAVI, p=0.061), particularly COPD (14.9% vs. 19.8%, p=NS), acute coronary syndrome (2.0% vs. 5.4%, p=NS) and previous coronary angioplasty (0% vs. 6.3%, p=NS). A lower proportion of patients operated in the post-TAVI period had echocardiographic signs of pulmonary hypertension (21.4% vs. 11.5%, p=0.043). The two groups are compared in Table 1.

Predicted mortality based on the logistic EuroSCORE was similar for both groups (7.4% pre-TAVI vs. 7.6% post-TAVI, p=NS), but far higher than observed operative mortality (3.9% vs. 1.8%) (Figure 1).

Figure 1.

Comparison of operative mortality predicted by the EuroSCORE and observed mortality for the two patient groups.

(0.06MB).

Operative and one-year mortality rates in the total population were 2.8% (n=6) and 7% (n=14), respectively. Causes of death in the post-operative period were septic shock (n=3), refractory HF (n=2) and unknown cause (n=1). Only two of the 14 deaths during one-year follow-up were attributed to cardiac cause.

Observed operative and one-year mortality were lower after implementation of TAVI, but not significantly (3.9% vs. 1.8% and 10% vs. 4.5%, respectively).

Kaplan-Meier survival curves are shown in Figure 2. At one-year follow-up, survival among patients operated in the pre-TAVI period was 89.9%, with a higher survival rate among those operated in the post-TAVI period (95.5%), but without statistical significance (log-rank test p=0.126).

Figure 2.

Kaplan-Meier one-year survival curves.

(0.06MB).

Significant differences were found in morbidity between the two groups (Table 2), with more complications in patients operated in the pre-TAVI era, both in the post-operative period (23.3% pre-TAVI vs. 13.5% post-TAVI, p=0.047) and at one year (20.4% vs. 9.9%, p=0.032).

Patients who underwent SAVR in 2005 had more postoperative complications, particularly need for prolonged invasive mechanical ventilation (24.5% pre-TAVI vs. 6.3% post-TAVI, p<0.001), severe bradyarrhythmias (17.5% vs. 10%; p=NS) and major bleeding (10.1% vs. 4.5%, p=NS). This trend continued during one-year follow-up, with a greater number of rehospitalizations for cardiac cause (19.8% vs. 9.0%, p=0.027).

On univariate regression analysis, the following were predictors of operative mortality: history of dialysis (odds ratio [OR] 51.5, 95% confidence interval [CI] 2.7–977.2, p=0.009); significant mitral regurgitation (OR 8.4, 95% CI 1.1–63.0, p=0.038); urgent cardiac surgery (OR 16.8, 95% CI 2.6–107.3, p=0.003); prolonged invasive mechanical ventilation (OR 19.0, 95% CI 1.9–188.8, p=0.012); post-operative ventricular fibrillation or recurrent ventricular tachycardia (OR 67.7, 95% CI 3.4–1355.1, p=0.006), severe HF (OR=28.1, 95% CI 3.4–22.8, p=0.002), pneumonia (OR 33, 95% CI 3.9–275.4, p=0.001), surgical wound infection or mediastinitis (OR 13.7, 95% CI 1.2–155.4, p=0.035), major bleeding (OR 9.8, 95% CI 1.5–64.2, p=0.017), cardiac re-intervention (OR 7.9, 95% CI 1.2–50.9, p=0.029); and prolonged hospital stay (≥14 days) (OR 18, 95% CI 2.8–115.7, p=0.002) (Table 3).

In multivariate analysis, urgent cardiac surgery (86-fold greater risk of death, 95% CI 2.6–2800, p=0.012) and need for prolonged invasive mechanical ventilation (33-fold greater risk of death, 95% CI 1.3–847, p=0.036) were independent predictors of operative mortality (Table 3).

There are various algorithms for preoperative risk stratification of cardiac surgery, one of the most commonly used being the logistic EuroSCORE. We analyzed the discriminatory power of this algorithm for operative and one-year mortality, as presented in Table 4.

The logistic EuroSCORE did not show discriminatory power for operative mortality (AUC 0.51). It showed satisfactory predictive value for one-year mortality (AUC 0.79, 95% CI 0.64–0.93, p=0.004), particularly in male patients (AUC 0.94, 95% CI 0.8–1.0, p=0.003) and in those aged under 75 (AUC 0.86, 95% CI 0.7–1.0, p=0.001) (Table 5).

The authors declare that no experiments were performed on humans or animals for this study.

The authors have obtained the written informed consent of the patients or subjects mentioned in the article. The corresponding author is in possession of this document.