Ross Procedure in Children: 17‐Year Experience at a Single Institution

Background The Ross procedure in children carries substantial mortality and reoperation rate. Aortic root dilatation is of concern. To prevent dilatation of the neoaortic root, but permit normal growth, we began to apply an absorbable poly‐(p‐dioxanone)‐filaments (PDS) band at the sino‐tubular (ST)‐junction. Methods and Results All children (n=100) who underwent Ross procedure during 1995–2012 were studied. Mean age at operation was 8.6±6.1 years (median 8.3 years, range 3 days to 18 years); 19 patients were younger than 1 year of age. The root replacement (n=91, Ross‐Konno procedure in 29 patients), root inclusion (n=6), and subcoronary implantation (n=3) techniques were used. Operative mortality was 6% (6/100, 4 neonates, 2 infants). Age of <1‐year at time of operation was a risk factor for early death (P<0.001). Mean follow‐up time was 7.0±4.8 years (median 7.4 years, range 5 days to 16 years). Late mortality was 4.3% (4/94). Freedom from moderate or greater neoaortic valve insufficiency (AI) at 5 and 10 years was 89% and 83%, respectively. Freedom from neoaortic valve reoperation at 5 and 10 years was 96% and 86%, respectively. Aortic dilatation to Z‐score >4 was greatest at the ST‐junction (23%, 11/48) compared to the aortic annulus (17%, 11/66) and sinuses (14%, 7/50). Since 2001, a PDS band was placed around the ST‐junction in 19 patients. Survivors with the PDS band had less AI (0 versus 20%, P=0.043) compared to survivors (n=35) without the PDS at 4.1±3 years. Conclusions The Ross procedure in children can be performed with acceptable results. Children younger than 1 year of age have higher mortality, but not an increased autograft reoperation rate. Stabilization of the ST‐junction may reduce AI.

R eplacement of the aortic valve with a pulmonary autograft (Ross procedure) was introduced in the late 1960s, 1 and good results have been reported in both adults [2][3][4] and children. [5][6][7][8][9][10] However, younger age at the time of the Ross procedure appears to be a risk factor for mortality and reoperation. 11,12 Progressive neoaortic valve insufficiency (AI) and the need to replace the right ventricular outflow tract (RVOT) conduit in a growing child are key issues. 13 The main concern in children is the dilatation of the neoaortic root, which leads to progression of AI. 14,15 While stabilization of the aortic root can be achieved in most adult patients, 16,17 prevention of aortic root dilatation in a growing child can be problematic. To reduce aortic root dilatation and the rate of reoperations on the neoaortic root, we began to apply an absorbable poly-(p-dioxanone)-filaments (PDS) band to stabilize the sino-tubular (ST)-junction. Herein we describe our results with the Ross procedure and midterm outcomes of the ST-junction stabilization technique.

Patients and Methods Patients
This study was approved by the Human Research Ethics Committee at the Royal Children's Hospital (RCH). All patients who underwent a Ross procedure at the RCH between November 1995 and April 2012 were included in this retrospective study. Medical records were reviewed. Two researchers (S.P.T.T. and R.K.) reviewed all echocardiograms and recorded the diameters of the pulmonary and aortic annulus along with preoperative aortic insufficiency and stenosis. Follow-up echocardiogram reports and letters were obtained from the patients' respective cardiologists. At last follow-up, postoperative autograft measurements and Z-scores 18 were obtained at the annulus, ST-junction, and sinus of valsalva. Early death was defined as death occurring within 30 days of the Ross procedure having been performed or before discharge from the hospital.
Patient characteristics are outlined in Table 1. There were 63 male patients and the mean age at operation was 8.6AE6.1 years (range 3 days to 18 years). There were 6 neonates (≤30 days of age), 13 infants, and 81 children older than 1 year of age.

Operative Techniques
All patients underwent surgery via sternotomy with standard cardiopulmonary bypass (CPB), hypothermia, and antegrade cardioplegia. Median weight at operation was 25.1 kg (range 12.7 to 49 kg) with a mean body surface area of 1.0AE0.54 m 2 (range 0.17 to 2.3 m 2 ). The median length of CPB was 215 minutes (IQR 178 to 253 minutes). Median aortic crossclamp time was 149 minutes (IQR 122 to 182 minutes). Circulatory arrest during operation was utilized for only 1 patient for 12 minutes. The median lowest temperature was 28°C (IQR 28 to 32°C). The root replacement (n=91, Ross-Konno procedure in 29 patients), root inclusion (n=6), and subcoronary implantation (n=3) techniques were used. The autograft was sutured to the left ventricular outflow tract using a running prolene suture. Right ventricular to pulmonary artery (RV-to-PA) continuity was established using a homograft (aortic or pulmonary) or a Contegra (Medtronic Inc.) conduit in all the patients. The median conduit size used was 20 mm (IQR 18 to 22 mm). Since 2001, 19 patients had PDS banding around the ST-junction with the anticipation of reducing progression of AI.

Statistical Analysis
All data were analyzed with Stata Statistical Software: Release 10 (StataCorp LP). Descriptive statistics for continuous variables were expressed as meanAEstandard deviation (range), whereas skewed continuous data are presented as medians (interquartile range). A Student's t test was used to compare continuous variables. Categorical variables were summarized as frequencies and percentages. A Fisher's exact test was used to compare categorical or binary variables.
Kaplan-Meier actuarial survival curves were used to analyze and plot time-related endpoints. Association of age at Ross procedure less than 1 year with early mortality and comparison of aortic insufficiency between PDS and control groups was assessed using a chi-squared test and confirmed using a Fisher's exact test. Cox proportional hazards regression analysis was used to identify risk factors for time-related outcomes. Variables tested in the univariate analysis included weight, age, syndrome, isolated aortic stenosis (unicuspid, bicuspid, tricuspid), isolated aortic insufficiency, combined aortic stenosis and insufficiency, endocarditis, rheumatic heart disease, root replacement technique, root inclusion technique, subcoronary technique, Ross-Konno procedure, PDS use, concomitant surgery, bypass time, and cross-clamp time. Of the endpoints examined, due to the number of outcomes, multivariable risk analysis was feasible only for any reoperation. For this endpoint, the number of risk factors included in the model was limited to the 3 most likely predictors from the univariate analysis (due to the limited number of reoperations in total).
At our institution, a PDS band was used in 19 patients. We compared them with a control group of 35 patients who underwent the Ross procedure without use of a PDS band (Table 4). These patients were then evaluated based on age, weight, year of surgery, and length of follow-up. There were no differences between operative mortality (P=0.31), late mortality (P=0. 19), and neoaortic reoperation (P=0.45). However, there was a reduction in moderate or greater AI (P=0.043) in the PDS group (Table 4). Kaplan-Meier comparison of AI in PDS versus non-PDS group is shown in Figure 6.

Follow-Up
Mean follow-up for the 90 survivors was 7.3AE4.8 years (1 month to 16 years). Concurrent follow-up (within the last 3 years) was obtained for 92.2% (83/90) of the patients, and the remainder were lost to follow-up. Of the survivors, 96.7% (87/90) of the patients were in New York Heart Association (NYHA) class 1 and 3.3% (3/90) were in class 2.

Discussion
Although the Ross procedure was initially intended for the treatment of aortic valve disease in adults, 1 it is currently utilized in the pediatric population as well. Herein we reported a 17-year experience at a single institution using the Ross procedure in children.
An operative mortality rate of 6% was observed in the present study. This is consistent with other studies in children where operative mortality ranged from 0% to 16%. 5,[7][8][9]15,[19][20][21] Takkenberg et al 12    early mortality in children (4.2%) compared to adults (3.2%). 12 In our study, patients younger than 1 year were at a higher risk of early mortality (32%, 6/19). All of the early deaths occurred in the neonatal and infant age group. The patients who died presented in a critical state and necessitated urgent surgery. This is similar to the results reported by Alsoufi et al 11 (n=151, median age of 8.6 years at operation) who identified age younger than 1 year as a predictor of early mortality. In addition, Kadner et al 8 (n=52, median age of 5 years at operation) observed a high early mortality in neonates and infants who presented in poor preoperative condition. Overall, a 10-year survival rate of 91% in our study is comparable to the 90% reported by Horer et al 9 (n=152, median age of 10.1 years at operation). In our study, it appeared that younger and smaller children had worse outcomes, especially those who had preoperative endocarditis.
In this present study, the pulmonary autograft performed well with good intermediate-term durability. Autograft failure leading to AI was the cause of reoperation in all 9 patients. We report a 10-year rate of freedom from aortic valve or root reoperation of 86%, comparable to the 10-year rate reported by Alfousi et al 15  Increase in aortic root Z-scores with time, suggesting progressive autograft dilatation, is a key concern with the Ross procedure, and has been reported by several authors. 16,22,23 The challenge faced in the pediatric population is the potential for AI to develop as the neoaortic root increases in size becoming out of proportion to the somatic growth of the child or young adult. 24 Inclusion, sub-coronary, and reinforcement techniques to improve the durability of the autograft have been utilized in adult populations and mid-term results have been promising. 16,17,[25][26][27] However, stabilization of the neoaortic root that would, at the same time, permit somatic growth is challenging in children. In our study, 40% of patients did not have any AI, 53% had trivial to mild AI and 7% had moderate to severe AI.   The patients were not randomized prospectively to PDS band application, but the PDS band was applied by 2 surgeons as a personal preference. On analysis of our results, the technique did not significantly decrease the rate of autograft reoperation at mid-term follow-up. However, AI was reduced significantly in the PDS group. These results appear promising and further study of the PDS band technique is warranted in order to observe whether the reduction in AI translates into fewer autograft reoperations in the PDS group over time.
A weak point of the Ross procedure is the potential for an isolated aortic valve disease to be transformed into a double valve disease. The most common indication for reoperation in this cohort was for replacement of RV-to-PA conduit. As the patients grow, reoperation to replace the conduit can be expected. In our study, freedom from RV-to-PA conduit replacement was 76% at 10 years. This is comparable to the reported 10-year rate of freedom of conduit reoperations of other centers. 5,10,15 In the present study, risk factors on univariate analysis for any reoperation included younger age at surgery, lower operative weight, and smaller implanted conduit diameter. This is consistent with Clark et al 10 who reported that patient age and homograft size to influence the longevity of the pulmonary homograft. The findings of smaller homograft size and younger age at operation as risk factors for reoperation are apparent in other pediatric patients undergoing RV-to-PA conduit surgery. 28 At the RCH, the current policy is to delay the Ross procedure until the child is past infancy, if feasible. If the patient requires surgery during infancy, the preference is to perform an initial surgical aortic valve repair. 29 Due to excellent results with aortic valve repair, we tend to reserve Ross operation as the second line of management. The number of aortic valve repairs increased as compared to Ross operations during the recent years. 29,30 Only patients with small aortic annulus would have Ross procedure without contemplation for an aortic valve repair as initial operation. 29,30 Thus, postponing the Ross operation until later in childhood or, ideally, into adulthood may minimize both mortality and overall reoperation rates.

Limitations
Our study is retrospective in nature and is limited by its design. Over the study period, surgical and perioperative techniques varied between patients. Although the groups were retrospectively matched, the selection of patients who underwent PDS banding was not prospectively randomized. Dimensions of the neoaortic root were not available in all patients.

Conclusion
The Ross procedure in children can be performed with acceptable results. Children younger than 1 year of age have higher mortality, but not an increased autograft reoperation rate. Stabilization of the ST-junction may reduce AI. The majority of patients are asymptomatic at last follow-up. A smaller homograft inserted at the Ross procedure is a significant predictor for reoperation.

Sources of Funding
Dr d'Udekem is a Career Development Fellow of The National Heart Foundation of Australia (CR 10M 5339). This research project was supported by the Victorian Government's Operational Infrastructure Support Program.