Residual root fate after aortic surgery in bicuspid aortic valve with right‐to‐left fusion: A comparative risk analysis

Abstract Background and Aim Although bicuspid aortic valve (BAV) anatomy might influence aortic aneurysm development, BAV‐related root involvement still lacks standardized surgical management. We aimed to evaluate late clinical outcomes and risk factors for root dilation after proximal aortic replacement in patients with BAV and right–left fusion (RL‐BAV). Methods Clinical and echocardiographic data of all patients with intraoperative RL‐BAV who underwent ascending aortic replacement with or without noncoronary sinus (NCS) replacement (Groups 1 and 2, respectively) between 1999 and 2017, were retrospectively revised. A multivariable analysis assessed hazard factors for root dilation during follow‐up (FU). Results Of 206 surgeries performed (M 81%; age: 57 ± 13 years, EuroSCORE II: 2.7 ± 1.9%), 79 (38%) required NCS replacement. One hundred fifty‐seven patients (76%) underwent aortic valve replacement (with aortic regurgitation predominating in Group 1, p = .04). The preoperative aortic root was larger in patients requiring NCS replacement (43.3 ± 5.1 vs. 39.2 ± 4.8 mm, p < .001). At a median FU time of 7 years (interquartile range: 4–10), no residual root dissections occurred, and only two patients (belonging to Group 2) required redo root surgery. Preoperative mild aortic regurgitation and aortic root diameter >35 mm at discharge were risk factors for root dilation >40 mm at FU (p = .02). Aortic root did not dilate over time, irrespective of NCS replacement (p = .06). Conclusions Aortic root in patients with RL‐BAV undergoing ascending aortic replacement (±NCS replacement) does not significantly dilate over time, even if patients with preoperative aortic regurgitation and postoperative root more than 35 mm might require more surveillance.


| INTRODUCTION
Bicuspid aortic valve (BAV) is the most common congenital heart defect, 1 it is associated with aortic valve disease and aneurysm formation, with an increased risk of developing acute aortic dissection. Aortic dilation involves either the arch and the proximal aorta (aortic root and/or ascending aorta) according to BAV anatomy. BAV with right-to-left fusion pattern (RL-BAV) is the most common anatomic pattern, 1 and it is frequently related to aortic root enlargement. Even if genetic and hemodynamic factors contribute to the heterogeneity of BAV aortopathy, there is no consensus about their degree of involvement, interaction, and implication for surgical management. Therefore, even if the most recent guidelines agree on thresholds for aortic repair 2 and many studies have recognized that aortic root does not dilate over time when not replaced, 3,4 it is still not well established which is the best surgical strategy and when a specific technique should be preferred above others in root surgery, especially regarding BAV anatomy. We aimed to evaluate late clinical outcomes and risk factors for root dilation after proximal aortic replacement in patients with RL-BAV.

| Patient selection
All subjects who underwent ascending aortic surgery between June 1999 and July 2017 at the Cardiac Surgery Department of the Padua University Hospital were identified through the operative reports database.
We selected all those cases with an intraoperative finding of RL-BAV (±aortic valve disease with surgical indication) and who underwent: ascending aortic replacement extended to the noncoronary sinus (NCS) (Group 1), or supracoronary ascending aortic replacement (Group 2). All operations were performed by means of full sternotomy. Technically, in Group 1 patients, NCS was replaced molding the dacron prosthesis used for ascending aortic replacement ( Figure 1A-C). NCS was replaced in patients with root diameters more than 45 mm and indications for aortic valve surgery, or in patients with root diameters less than 45 mm, associated aortic valve disease, and a thin and asymmetrically dilated root. Current guidelines for aortic surgery were used. 2 Patients who underwent a David (reimplantation or remodeling technique) or a Bentall operation were excluded. Emergent surgery (e.g., dissection), associated aortic arch surgery, active endocarditis, connective tissue disorders (e.g., Marfan syndrome, Loeys-Dietz syndrome, etc.), and redo operations were also exclusion criteria ( Figure 2).

| Data collection
Perioperative, intraoperative, and postoperative clinical data were retrieved from our institutional database. Patients were F I G U R E 1 Surgical image of the noncoronary sinus replacement technique: first, the bicuspid valve is exposed and its anatomy is defined (LC, left-coronary cusp; NC, noncoronary cusp; RC, right coronary cusp) (A); the Dacron prosthesis is prepared and sutured starting from the noncoronary sinus (B and C)   Table 1. Among this cohort, 79 patients underwent ascending aortic replacement extended to the NCS (Group 1), and 127 patients underwent supracoronary ascending aortic surgery (Group 2). Patients were predominantly male (>75%), and did not significantly differ in terms of age and EuroSCORE II risk among the two groups (p = .52 and .78, respectively). Group 1 and Group 2 required associated aortic valve replacement in more than three-fourth of patients (75% and 77%, respectively); predominant indications were aortic regurgitation in Group 1, and aortic stenosis in Group 2 (p < .01). Group 1 showed a low rate of postoperative reoperations for bleeding (n = 2, 3%) with no significant difference compared to Group 2 (p = .57).
There were no differences between the two groups in terms of hospital stay and in-hospital mortality (p = .44 and .43, respectively).
T A B L E 2 Intraoperative and early postoperative outcomes  Table 4. Only two patients belonging to Group 2 required a root replacement due to aortic root enlargement (none in Group 1).
The cumulative incidence rate for cardiac reoperations did not find any significant difference between the two groups (p = .42) (Figure 3).

| Aortic root echocardiographic analysis
One hundred forty-three (74%) FU echocardiograms were collected within the study period (with a median length of FU echo studies = 7 years, IQR: 4-9). When analyzing aortic root diameters, we found  The differential progression rate between the two groups was not significant (p = .06) (Figure 4). Multivariable analysis of risk factors for root dilation at FU is summarized in Table 5 and shows that preoperative aortic regurgitation and dilated root at discharge are the only factors related to increased risk of dilation at FU (p = .02).

| DISCUSSION
BAV, the most common congenital heart defect in adults, 5,6 is usually complicated by the development of aortic stenosis or regurgitation.
However, aortic dilation from the aortic root to the aortic arch (bicuspid aortopathy) is also present in approximately 50% of affected people. 1 Hemodynamic and genetic factors seem to play a combined role in the development of BAV aortopathy. 7,8 Even if current guidelines clearly establish limit thresholds for aortic replacement in BAV patients, 2 aortic root management still remains unresolved.
Many studies have advocated supracoronary ascending aortic replacement as a treatment (preserving the intact moderately dilated sinus segment and coronary ostia), 9,10 whereas others suggest removal of the sinus segment by means of a valve-sparing procedure or a Bentall operation. 11 However, there is limited comparative analysis regarding the stability of the residual root after aortic surgery (either with or without aortic valve replacement) according to BAV morphology. We selectively analyzed RL-BAV patients because the most common fusion pattern clinically encountered involves the right and left cusps. 1 Even if RL-BAV pattern is most frequently related to type 1 BAV aortopathy (which preferentially includes patients >50 years of age, with associated aortic stenosis and different degrees of aortic root dilation), 7,12 BAV regurgitation is typically associated with root phenotype. 13,14 Our results confirm this evidence: in fact, patients with a more dilated aortic root and who underwent an associated partial root procedure (Group 1) suffered more frequently of moderate-severe aortic regurgitation rather than stenosis. The more "aggressive" approach of replacing NCS arises from the  23,24 ). Similarly to these previous studies, 3,4 we also showed a reduction of root diameter in patients undergoing supracoronary ascending aorta replacement: this might be explained by the fact that when the vascular prosthesis is anastomosed to the sinotubular junction (STJ), the mismatch between STJ and the prosthesis might indirectly reduce the root diameter.
Differently from previous references, our multivariable analysis found that larger diameters at discharge and preoperative aortic regurgitation were significant risk factors of aortic root dilation at FU. As aortic regurgitation is strongly correlated to root phenotype, we hypothesize that there are some unknown genetic features of the aortic root when associated with regurgitation which make it more prone to enlargement. 25 Patients who did not undergo AVR were also at increased risk of dilation, despite a not significant p value. We think that the prosthetic stent might stabilize the annulus and the root; therefore, BAV, even if continent and not stenotic, produces abnormal flows and shear stress into the sinusal portion, and this might explain a higher predisposition to dilation. 8,26,27 This study has the following limitations. First, it is a retrospective study, and surgeries were performed by several surgeons working at our center. Aside from current guidelines, the decision to replace NCS or not was mostly based on surgeon preference. Clinical and echocardiographic FU was not 100% complete, as well as median clinical FU is short. The three groups show a disproportion in the total amount of patients included, and larger patients groups would be mandatory. Aortic measurements were performed by different cardiologists in different centers with different machines; a systematic evaluation of all the available images by a single examiner (blinded to patients group) would have reduced the error margin.
In conclusion, our study shows that aortic root in patients with RL-BAV who undergo proximal aortic replacement does not significantly dilate over time (irrespective of SNC replacement). Patients with preoperative aortic regurgitation and postoperative residual aortic root dilation (>35 mm) seem more prone to dilation at FU and might require surveillance. However, surgical reoperations for aortic root dissection or dilation are extremely rare at FU, so that full replacement of the aortic root when only mildly dilated still does not appear justified. NCS replacement in the asymmetrically dilated root might be easier and less risky ( Figure 4).