Previously, we reported the risk factors for dilation of the aorta over time after AVR for BAV which focused on the possible impact of valve fusion pattern [12]. In that study, the presence of aortic regurgitation and AAD > 40.0 mm at time of surgery were shown as significant predictors of dilation of the aorta after AVR, but valve fusion pattern did not. In this study, we investigated the actual long-term results, including the rate of reoperation related to the aorta.

The rate of later aorta operation after AVR for BAV remains unclear. Our major finding of the current study was that the cumulative rate of later aorta operation rate at 10 years was 16.3% (Fig. 1). This result does not indicate a need to perform concomitant ascending aorta replacement at the time of AVR for BAV if AAD is below 45 mm, as recommended in the current guidelines [2, 11]. Girdauskas and associates compared the risk of late aortic events after isolated AVR surgery for bicuspid versus tricuspid aortic valve stenosis with concomitant ascending aortic dilatation of 40 to 50 mm [13]. In 153 patients diagnosed with bicuspid aortic valve stenosis, 5 patients (3.3%) required later aorta operation and aortic dissection did not occur in the BAV group within follow-up period (mean 11.5 years) in their study, which was comparable to our results. However, in the present study, all the three patients (1.7%) who underwent reoperation during the follow-up period had preoperative AAD more than 44.0 mm, so patients with larger AAD at the time of AVR for BAV seems to be needed careful follow-up monitoring of aorta diameter.

Existing American College of Cardiology/American Heart Association and European Society of Cardiology guidelines do not specify a treatment strategy for the thoracic aorta in patients who have undergone AVR for BAV previously [2, 11]. AVR for BAV can eliminate hemodynamic factors but not genetic factors. Aortic problems may therefor occur sooner in patients who underwent AVR for BAV than in patients who underwent AVR of a tricuspid aortic valve. In fact, in the present study, one patient underwent emergency surgery for acute type A aortic dissection, who had had a CT follow-up as an outpatient every year. His most recent AAD before developing acute type A aortic dissection was 52.0 mm. The fact that the entry was located in the middle of the ascending aorta, not at the site where arterial line was inserted, also suggests that the aorta was fragile. In the ACC/AHA guidelines, surgery is indicated in patients with aneurysm of the aortic root or ascending aorta who have a maximum diameter of 55.0 mm or more (Class 1), and in patients with aneurysm of the aortic root or ascending aorta who have a maximum diameter of 50.0 mm or more when performed by experienced surgeons in a Multidisciplinary Aortic Team (Class 2a) [11]. In the ESC guidelines, surgery is indicated in patients with no elastopathy with ascending aortic aneurysm of 55.0 mm or more (Class 2a) and in patients with a bicuspid valve with risk factors with ascending aortic aneurysm of 50.0 mm or more (Class 2a) [2]. In a patient who has undergone AVR for BAV thoracic aorta surgery may be necessary before dilation to 55 mm, for example, at 50 mm.

In our institution, AAD in echocardiography results means the diameter of the ascending aorta in the long-axis view of the left ventricle. However, the maximum diameter of ascending aorta is often located in a more distal portion than that seen on the long-axis view. In our cohort, the mean preoperative AAD in echocardiography results was 32.6 ± 4.6 mm. On the other hand, the mean preoperative AAD in CT was 39.1 ± 3.9 mm. Therefore, it is problematic to measure the maximum diameter of ascending aorta with echocardiography alone. Considering the possible genetic factors, we may need to follow up with not only echocardiography but also CT every year in all patients who underwent AVR for BAV.

This study had several limitations. First, the sample size of the present study was small. There were only 3 patients who underwent aortic reoperation during the follow-up period, which may be associated with less statistical power. Second, all our cohorts included Japanese patients only, that may limit generalizability. Third, we could not know the cause of death in 12 patients. Of them, some patients may have died because of valve or aorta related event. Finally, only 29.8% in our cohort underwent CT as outpatients every year. Therefore, we could not completely follow aortic diameter after AVR for BAV.