DISCUSSION
The main finding of this single-centre, retrospective large-cohort study is that preoperative AF is independently predictive of long-term all-cause mortality at a median follow-up of 4 years and remained independently predictive after adjustment for other risk factors and in propensity score-matched analysis following first-time isolated surgical AVR.
Our study further lends support to previous studies investigating the relationship between preoperative AF and mortality after AVR. Ngaage and colleagues demonstrated an increased risk of major adverse cardiovascular events with preoperative AF in 381 patients undergoing surgical AVR, however preoperative AF was not predictive of all-cause long-term mortality after risk factors adjustment, which was probably due to the small sample size. Saxena and colleagues showed that preoperative AF significantly increases the risk of all-cause mortality by 36% after surgical AVR. However, intraoperative variables such as bypass time and aortic clamp time were not adjusted for in their analysis, despite the bypass time being significantly different between the study groups. These important variables have been shown to affect clinical outcomes following cardiac surgery. Of note, there was no significant difference between the bypass time and aortic clamp time between the two cohorts in our study (Table 1). Levy and colleagues found that preoperative AF was independently associated with more than 5-fold increase in long-term mortality (HR 5.5, 95% CI, 1.13-26.15; P=0.03), however, the study involved only 83 patients undergoing surgical AVR, all with poor left ventricular function. The remarkable hazard ratio expressed in Levy’s analysis likely reflects the well-documented fact that AF is poorly tolerated and portends a worse prognosis in patients with congestive heart failure. Transcatheter aortic valve implantation (TAVI) is currently the gold standard intervention for non-surgical candidates with aortic valve stenosis. A meta-analysis of observational studies found that preoperative AF significantly increases the risk of long-term all-cause and cardiovascular mortality after TAVI.
Wang and colleagues demonstrated that preoperative AF was a predictor of 30-day mortality after surgical AVR. In our study, patients who had preoperative AF were significantly older with more comorbidities such as hypertension, diabetes, peripheral vascular disease, previous myocardial infarction, renal disease and left ventricular dysfunction compared with patients who had preoperative SR. This is probably the reason for the prolonged in-hospital stay observed in the preoperative AF cohort. However, we observed no significant difference in the operative or 30-day mortality between the 2 study groups. One plausible explanation is that our study only included patients undergoing elective surgery whereas 62% of Wang et al.’s preoperative AF cohort had urgent or emergency surgery compared with 48% in the preoperative SR cohort (P=0.008). Urgent or emergency surgery was predictive of morbidity and mortality during follow-up.
Our results suggest that preoperative AF has no value in formalized surgical risk-stratifying tools, such as the EuroSCORE, as we did not find in-hospital mortality to be impacted by preoperative AF. There is scope, however, for preoperative AF to be incorporated into long-term surgical risk assessments, or at least to inform decision-making when counselling patients as to the risk of AVR in the outpatient setting. Of greater clinical interest is whether restoration of SR pre- or perioperatively reverses the risk attributable to preoperative AF. This is yet to be determined and represents a highly worthwhile area for further studies.