Ross Operation in Pediatric Population: Impact of the Surgical Timing and the Native Pulmonary Diameter on the Outcome

Aortic valve replacement early in life may be inevitable. Ross operation, until present day, remains the favorite surgical option in pediatrics with irreparable aortic valve disease. Nonetheless, the necessity for re-operation was always its principal limitation due to aortic valve failure or homograft degeneration. We present our 25 years of experience in the pediatric population. From August 1994 until June 2018, 157 children below 18 years underwent the Ross operation. This retrospective review aims at assessing the long-term outcomes, as well as the risk factors for re-operation after Ross procedure. Median age was 10.9 years, of which seven patients were infants, 79 children, and 71 adolescents. The median follow-up time was 14 years. Hospital mortality was 0.6%. Freedom from autograft re-operation for children was 96.7% and 94.1% at 10 and 20 years, respectively; whereas for adolescents, it was 92.6% and 74.9% at 10 and 20 years. For children, freedom from homograft re-operation was 92.5%, 83.5%, and 56.2% at 10, 15, and 20 years; while for adolescents, it was 96.8%, 91.8%, and 86.7% at 10, 15, and 20 years. Homograft size (p = 0.008) and childhood (p = 0.05) were risk factors for homograft re-operation. Pulmonary valve diameter > 24 mm (p = 0.044) and adolescence (p = 0.032) were risk factors for autograft re-operation. Our experience demonstrated excellent early and late survival. While children have preferential outcomes concerning autograft re-operation, those who received a smaller homograft had a higher right-sided re-intervention incidence than adolescents. Pulmonary diameter > 24 mm at surgery was an indicator of future autograft failure.


Introduction
The first successful Ross operation was carried out in 1967 by Ross [1] as an alternative to mechanical, biologic, and homograft valves for aortic valve disease. Patients undergoing Ross operation do not require anticoagulation, and in children, the neo-aortic autograft may follow the somatic growth [2,3]. Initially this surgical technique did not manifest great success, starting from the Nineties of the last century it reached the peak of its popularity, not only in the pediatric age group but also in the adult patients. The most evident criticism toward this procedure has been the fact of transforming a single-valve disease into a doublevalve disease; this has declined its application over the years [2][3][4][5]. Furthermore the procedure is a very complex one 1 3 which necessitates a long cross-clamp time, consequently augmenting cardiac ischemia chances. Another factor that contributed to this decline was the shortage of the pulmonary homografts, this resulted in a remarkable limitation of Ross operation to specific categories: children, women of childbearing age, young athletes, and those unsuitable to anticoagulation therapy [4,5].
The in-hospital outcome of the Ross procedure is generally characterized by low mortality and morbidity in both age groups, the adult and pediatric patients [6,7]. Progressive dilation of the neo-aortic root associated with valve regurgitation after Ross operation has been reported; as well as homograft calcification and stenosis, which are considerable drawbacks of this procedure, especially in pediatric patients [8][9][10][11][12].
Currently in the literature, a great attention is paid to the aortic annulus diameter at the time of the surgical operation [3][4][5][6][7]. This study is novel; there are no reported studies in the literature demonstrating any potential role of the native pulmonary autograft diameter and its impact on the outcome of Ross procedure.
The aims of this study can be summarized as follows: (i) to evaluate the long-term outcome of morbidity and mortality in our pediatric series following Ross operation and (ii) to investigate any variation in the long-term outcome among the three pediatric age groups (neonates, infants, and adolescents).

Patients and Methods
This study is a retrospective analysis of prospectively collected data from patients younger than 18 years who underwent Ross operation. The study was approved by the local Ethics Committee (19int/2020). Informed consent was obtained from all individual participants (or their legal guardians) included in the study.
Between August 1994 and June 2018, 157 pediatric patients who underwent Ross operation at our institution (IRCCS Policlinico San Donato) were included in this study. The age limit was settled at < 18 years at the operation time. Patients were considered as infants (< 12 months), children (1 to 12 years), and adolescents (13 to 17 years). Aortic regurgitation was graded as trivial, mild, moderate, or severe according to the American Heart Association guidelines [13], and homograft stenosis was graded as mild (peak gradient < 30 mmHg), moderate (30-60 mmHg), and severe (> 60 mmHg).
At the time of admission, transthoracic echocardiogram was used to measure the aortic and pulmonary diameters and then these data were confirmed by transesophageal echocardiogram in the operation theater. Furthermore the pulmonary diameter was verified with Hegar dilators after the autograft harvesting.
In November 2020, all patients or their relatives were contacted by telephone and invited to our institution for a follow-up control. Through these phone calls, deceased patients were identified.

Statistical Analysis
Data collected are reported in categorical variables as numbers and percentages, while the continuous variables are reported as median and Interquartile Range (IQR). The timerelated freedom event was assessed non-parametrically using the Kaplan-Meier method; while, differences were evaluated with a log-rank test for equality. Univariate risk factors analysis was concluded using cox regression. For survival analysis, patients were censored at death; otherwise at the last-known date to be alive for freedom from re-operation or re-intervention. The first event that occurred after Ross completion was taken into consideration. Statistical differences were considered significant when the p-value < 0.05. Data analysis was performed with Stata Statistical Software (Release 12; StataCorp 2011 College Station. TX: StataCorp LP), SAS software, version 9.4 (SAS Institute, Inc., Cary, NC).

Patient Characteristics
At the time of surgery the median age was 10.9 years (IQR 8-14 years). Of all the 157 patients, 51 patients (32.5%) had undergone at least one previous surgery; altogether 66 procedures before the Ross operation. Eleven of these patients underwent two previous operations, while other four patients underwent three previous operations. A total of 68 patients (43.3%) underwent 75 procedures on the aortic valve before the Ross operation: in 41 patients (26.1%), aortic balloon valvuloplasty was achieved, while in other 27 (17.2%) surgery was the favored option (19 commissurotomies and eight repairs). Sixteen patients (10.2%) underwent balloon aortic valvuloplasty and aortic valve repair. Indications for Ross operation were as follows: aortic insufficiency (38.9%), aortic stenosis (26.8%), and mixed stenosis and insufficiency (34.4%). In the case of aortic insufficiency and those with the mixed pathology, congenital insufficiency was encountered in 74 patients, whereas other 36 patients developed regurgitation after valvuloplasty or valvulotomy. The median aortic annulus diameter and pulmonary annulus diameter were 20 mm (IQR 17-24) and 22 mm (IQR 18-24.5). All preoperative details are reported in Tables 1 and 2.

Surgical Procedure
All patients underwent surgery via median sternotomy with standard cardiopulmonary bypass (CPB) under moderate hypothermia (32 °C) and cold intermittent antegrade blood cardioplegia with topical cardiac cooling.
Pulmonary autograft was harvested taking into consideration the length and the dimension of the autograft, therefore utilizing the minimum indispensable length of the autograft; the pulmonary autograft was excised from the right ventricular outflow tract with a 2-to 3-mm muscle rim with a beating heart. However, we have modified our strategy over the last five years, harvesting the autograft only after cardioplegia injection and complete cardiac activity cessation. The right and left coronary arteries were mobilized with sufficient buttons and then the aortic root was removed. The adopted surgical techniques were as follows: root replacement 122 (77.7%), Ross-Konno 17 (10.8%), root inclusion 6 (3.8%), and subcoronary implantation 12 (7.7%). Right ventricle-pulmonary artery continuity was established using a pulmonary homograft in 153 patients; Contegra conduit (Medtronic, Minneapolis, MN) was used in the remaining four cases. The median conduit diameter size was 23 mm (IQR 21-24 mm). The median CPB time was 177 min (IQR 161-193 min), and the median aortic cross-clamp time was 124 min (IQR 110-143 min). Additional cardiac pathologies were addressed at the same time. In the root replacement group (a total of 122 patients), in 106 cases, the autograft was sutured to the annulus with 4-0 polyester interrupted sutures, of which in 64 patients, the autograft was knotted over a thin strip of autologous pericardium. In the remaining 16 cases, the anastomosis was done using three running sutures with a 5-0 or 6-0 polypropylene. As reinforcement against future aortic root dilatation and progressive aortic insufficiency, 15 adolescent patients had a 0.1-mm GORE® PRECLUDE® Pericardial Membrane sutured around the aortic root with a running suture before coronary reimplantation. Our current surgical tips for autograft implantation are reported in Table 3, and our decision algorithm is summarized in Fig. 1. The coronary artery buttons were implanted into their respective sinuses using a running suture of 5-0 or 6-0 polypropylene. After completing the left-sided part of the operation, a cryopreserved pulmonary homograft was implanted on the right ventricular outflow tract using two running sutures of 5-0 polypropylene for the distal anastomosis and 4-0 polypropylene for the proximal anastomosis. Operative data are summarized in Table 4.

Hospital Outcomes
There was one in-hospital death (0.6%). The patient was a neonate with aortic valve endocarditis, required extracorporeal membrane oxygenation (ECMO) support and died after 4 days due to multi-organ failure (MOF). Surgical morbidity was around 5.1% (8/156 patients). In three cases (1.9%) following a rise in the troponin levels and electrocardiographic changes after surgery, we performed coronary angiography, which was negative in all three cases. Two patients required surgical revision for deep sternal infection (1.3%). Bleeding requiring re-exploration was encountered in two patients (1.3%). One patient presented specific neurological deficits after the surgery. The median hospital stay was nine days (IQR 7-12 days). At discharge, the echocardiographic examination of two patients revealed mild to moderate autograft insufficiency.

Survival
The median follow-up time was 14 years (IQR 6-17 years) and it was completed in 95% of the cases. At follow-up, the median age of the patients was 24.6 years (IQR 17.8-30.6). There were eight late deaths. The overall 20-year survival rate was 95%. Ross-related deaths were five, of which two occurred early. The first patient died after two years, due to myocardial infarction as a consequence of acute closure of the left anterior descending coronary artery; while the second patient, who was a known case of Williams syndrome, died 5 months after the operation due to an acute episode of ischemic myocardial infarction, in which the autopsy revealed severe hyperplasia of the aortic intima that occluded the coronary ostia. The other three deaths were late: the first as a result of accidental homograft laceration during a re-intervention for right ventricle outflow tract (RVOT) stenosis, in another institution. The second case died 16 years following the Ross operation due to heart failure after other two re-interventions in the left and right side; the last patient was a neonate with autograft failure who was re-operated on for aortic homograft implantation and then after 6 years for a new left-side re-operation, consequently developed severe biventricular dysfunction then died in the intensive care for end-stage heart failure.  (Fig. 2a). The univariable analysis identified the following two risk factors for mortality: infants (HR 0.84, 95%CI 0.7-0.9, p = 0.015) and Ross-Konno procedure (HR 5.86, 95%CI 1.4-24.7, p = 0.016). Particularly if we consider infants as a reference age group, then children (HR 0.13, p = 0.021) and adolescents (HR 0.07, p = 0.008) (Fig. 2b). This indicates the promising results of Ross operation when it is performed after the first year of life. Considering the low number of outcomes at follow-up, the multivariate analysis was not significant for the above-mentioned risk factors.

Re-interventions
During the study period, 28 patients (19.1%) required surgical or percutaneous procedures for either autograft or homograft failure. Among them, 17 patients (11.6%) needed a pulmonary autograft re-operation, while 15 patients (10.3%) required treatment for RVOT failure with surgical homograft replacement or percutaneous pulmonary valve implantation (PPVI). Among the 28 patients, in one case a concurrent surgical operation was performed on the left and the right side. Whereas in three other patients, right and left surgical interventions were performed in two separate occasions. Two patients required two re-interventions after Ross operation, Fig. 1 Decision algorithm while another one necessitated three re-operations. One patient underwent heart transplantation five years after the Ross procedure due to myocarditis. The different surgical procedures after Ross operation are summarized in Table 5.
The risk factors that were identified on the univariable analysis for autograft re-operation were as follows: adolescence (HR 3.9, 95%CI 1.12-13.8, p = 0.032) and pulmonary autograft diameter size in mm (HR 1.1, 95% CI 1.0034-1.2876, p = 0.044). The main risk factors are summarized in Table 6.
During the follow-up, three children and 13 adolescents were re-operated for autograft failure, showing a significant difference in freedom from re-operation between the two age groups (p = 0.021). Freedom from autograft re-operation in the children was 96.7% (95%CI 86.  (Fig. 4). Regarding the native pulmonary autograft, a diameter of 24 mm results to be the cut-off value for the risk of subsequent left-side re-operation (p = 0.015), (Fig. 5).

Discussion
Ross procedure is an excellent choice for treating the pediatric population with aortic valve disease. This article presents a long-term follow-up of patients operated on with Ross procedure. Our data reflect how time passes and things change and how the Ross operation was reconsidered regularly over the recent decades. Going back to our previous manuscript about the pediatric Ross procedure [14], we have noticed considerable alterations in terms of risk factors for autograft and homograft failure, with the emergence of new risk factors and the regression of others. Our data confirm an excellent long-term survival. The overall in-hospital mortality of patients was (0.6%), which was a neonate who had devastating endocarditis in the postoperative period with MOF. The incidence of hospital mortality was very low, but it is correlated to the small number of neonatal Ross operations performed at our institution. Our   Fig. 2 a Long-term survival after Ross procedure. b Distribution of long-term survival by age class first-line choice for aortic stenosis in neonates is the percutaneous approach, which reduces the gradient noticeably. However, in some cases, the percutaneous approach results in a degree of insufficiency, which is anyhow well tolerated in the neonates. This approach allows us to evade a surgical intervention in the neonatal age, unless the percutaneous procedure is ineffective or there is an associated subaortic stenosis component, where the surgical intervention is inevitable in such cases.
Our results in long-term survival are quite satisfying. Compared to other studies [11,[15][16][17][18], overall long-term mortality in our study is the same. Patients below one year of age are at a higher risk of late mortality, mainly if associated with left ventricular dysfunction and previous interventions, as demonstrated in other studies [11,[16][17][18].
Pulmonary autograft is a durable substitute. In our study, the long-term freedom from autograft re-operation corresponded to other studies and was even superior to data reported by other authors [11,[16][17][18][19]. It is widely proven that the incidence of re-operation due to autograft failure is unlikely in the first few years following the Ross procedure. This incidence becomes significant starting from the fifth year after the operation [20][21][22][23][24][25][26].
Regarding autograft failure, it emerged from our analysis that the autograft has a better longevity if the surgery was performed during childhood rather than in adolescence.
While in children the freedom from re-operation was 94.1% (95%CI 82.2-98.1) at 15 years follow-up and in adolescents, it was 77.8% (95%CI 63.3-87.1) with a p-value of 0.021. This is an exciting finding, since in one of our previous reports, ten years ago [14], the age class was not identified as a risk factor for the left-side re-operation. Our results are probably attributed to a better adaptation and performance of the autograft in young children than adolescents. Other reports have confirmed varying age-related outcomes after Ross operation in the pediatric population [11,22,27]. Similarly, in another series evaluating various age groups, it demonstrated a notable discrepancy in freedom from left ventricle outflow tract (LVOT) re-operation at tenyear follow-up between the 1-to 12-year age group and the 13-to 17-year age group (73.3% and 46.1%, respectively); but this was just a trend since the difference in freedom from re-operation did not reach any statistical significance [28]. Moreover, in our series, the second important risk factor that emerged for LVOT re-operation was the pulmonary diameter before the Ross procedure. A pulmonary diameter > 24 mm is a probable risk factor for late autograft failure. Until now, in the literature the focus was on the aortic/ pulmonary size, with a mismatch that should not exceed 2 mm [29]; in addition, the focus has always been on the aortic annulus diameter, on which the autograft should be implanted. David et al. described the importance of the aortic annulus diameter as a risk factor for re-operation, whenever the latter is over 27 mm or indexed > 16 mm/m2 [22]. Horer et al. evidenced that the patient's age at the time of Ross operation substantially impacts the outcome, more significantly than the initial aortic annulus size [30]. Indeed, while a dilated annulus in children tends to normalize over time, it tends to grow further in adolescents, resulting in autograft failure [5].
The autograft dimension before the Ross procedure is probably also a considerable risk factor for dilatation and failure. When its size is above a specific value, it has a greater tendency to dilate, regardless of the aortic annulus diameter in which it is implanted.
This datum of our study is a tendency, since it disappears in the multivariate. Still, we consider it an exciting starting point to reflect on and possibly deepen further, given that other authors in the previous papers did not highlight this original finding.
Thus, we believe that when the diameter of the pulmonary autograft is > 24 mm, it should be taken into account the possibility of reinforcing the autograft at the moment of Ross operation, independently of the patient's age.
Our experience in the autograft reinforcement applied to 15 adolescent patients demonstrated optimum distant results [31]. Still, in the current series, it did not result to be a protective factor against future re-intervention on the autograft.
Regarding the evolution of RVOT, in our cohort study, 18 patients required a subsequent surgical procedure and in 10 cases, we treated homograft failure with PPVI. In followup, freedom from RVOT re-operation was 94% at ten years, 88% at 15 years, and 73% at 20 years. In our institution, currently, the first choice is always percutaneous treatment for homograft failure and only when it is unavoidable, we proceed with surgery. The surgical treatment is preferred in the presence of a contemporary left-side surgical indication or in the presence of endocarditis. Failure of the pulmonary homograft is a significant drawback in young children, but it is less remarkable in young adults and in older patients [32]. Our data confirm this concept. The analysis of risk factors has shown that young children are more vulnerable to homograft failure than adolescents. However, the difference becomes more evident after the 15th year of follow-up. This outcome correlates to the size of the implanted homograft; actually, the homograft size was also a risk factor. There is considerable heterogeneity concerning RVOT re-operation in other series, with freedom from re-operation ranging between 71 and 53% at 15-year follow-up [11,[15][16][17]33]. We have always used cryopreserved pulmonary homograft from a single donor bank and Contegra Medtronic conduit in small patients in our praxis. This study's freedom from reoperation is higher than other pediatric Ross reports, since we have a minor percentage of infants included in the study. Still, the freedom from re-operation is much the same as other studies that analyzed the procedure's outcome in adult patients [7].

Study Limitations
Few neonates were included in the study population and thus the indicated mortality rate could be little reliable. Nevertheless, we inserted them in the study to involve in the analysis the entire population operated at our institution. Absence of the risk factors on the multivariate, owing to the fact that we had few adverse events.

Conclusion
Our experience demonstrated excellent early and late survival rates following Ross operation. While children have preferential outcomes concerning autograft re-operation, those who received a smaller homograft had a higher incidence of right-sided re-intervention than the adolescents; this risk became statistically significant after ten years of follow-up. Pulmonary annulus diameter > 24 mm at the time of the surgical operation was an indicator of subsequent autograft failure.