The Ross procedure, first described in 1967,1 involves replacement of the diseased aortic valve with a pulmonary autograft, excision of the coronary arteries from the native aortic root, and reimplantation into the neo-aortic root. A pulmonary or aortic homograft conduit is then placed between the right ventricle and pulmonary arteries. Many reports2–9 have documented the effective use of the Ross procedure for isolated aortic valvar disease, as well as for complex obstruction of the left ventricular outflow tract in neonates, infants, and children. Following the Ross procedure, patients do not require anti-coagulation, and the autograft has been shown to be durable, and to grow in proportion to somatic growth.10, 11 Due to these attributes the pulmonary autograft is an attractive alternative to mechanical, homograft and xenograft valves when treating aortic valvar disease in children and young adults. Despite the encouraging early results relative to other surgical options, however, specific long-term morbidities, including dysfunction of the neo-aortic valve, dilation of the aortic root, and periodic replacement of the homograft, have called into question whether the Ross procedure is the best choice for surgical palliation of significant aortic valvar disease.
Mechanical valves inserted in children frequently require replacement as somatic growth occurs, as well as lifelong anticoagulation. Although the haemodynamic result and long term durability may be acceptable after insertion of mechanical valvar prostheses, the haemodynamics may be better after the Ross procedure.12 Thromboembolism and haemorrhage in anticoagulated patients remain important complications, especially in children.13, 14 Additionally, placement of a mechanical valve in a female child or adolescent will make child bearing later in life problematic due to possible thromboembolism or haemorrhage, as well as the potential teratogenic effects of warfarin. Porcine bioprostheses, which do not require anticoagulation, deteriorate more rapidly in young patients and have limited durability.15, 16
Although both aortic valvoplasty and replacement of the aortic valve with a homograft generally result in acceptable post-operative haemodynamics, do not require anti-coagulation, and are associated with a low incidence of thromboembolic phenomenons, both have issues related to durability. Aortic homograft valves have been shown to have limited durability in children, and do not grow with the child.16, 17 Similar to the mechanical valve, the aortic homograft does not perform as well from a haemodynamic standpoint during exercise as the autograft.18 Techniques for aortic valvoplasty include extension of the leaflets using autologous pericardium, reconstruction of the zones of apposition between the leaflets, reduction of the size of the valvar orifice, reduction of the sinuses of Valsalva, remodelling of the sinutubular junction, thinning and excision of raphes in conjoint leaflets, and complete replacement of leaflets with autologous pericardium. Although the initial results of the newer techniques are encouraging, their durability is not yet clear, and likely depends on the specific abnormality of the valve, whether earlier surgical and catheterization based procedures have already been performed, and the strategy used to repair the valve.19, 20 Even if the durability of valvoplasty is ultimately deemed less than that of replacement with a pulmonary autograft, the potential advantage of attempting repair is that, if subsequent replacement is required, all the options, including the Ross procedure, remain available.
In the light of all these considerations, our review will consider the late mortality and specific morbidities after the Ross procedure, including the function of the neo-aortic valve, dilation of the aortic root, reintervention, issues of the coronary arteries, left ventricular size and function, exercise performance, and electrophysiologic abnormalities. We will conclude our discussion with a description of new modifications to the Ross procedure, and the state of other valves for use in future replacements.
Mortality
Similar to the techniques of valvoplasty and other strategies for replacement, early and late mortality after placement of the pulmonary autograft is rare, occurring in 2.5% of patients within 30 days of the Ross procedure, and between 1.7 and 2.5% late after placement of the pulmonary autograft.2–7, 21 Late mortality is often linked to progressive pulmonary hypertension secondary to left ventricular or mitral hypoplasia in patients with complex left ventricular disease. Given the low mortality after the Ross procedure, long-term outcomes have focused on specific morbidities related to the procedure.
Morbidity
Function of the neo-aortic valve and dilation of the aortic root
The morphologic development and fibroelastic structure of the pulmonary and aortic valves are different, with the leaflets of the aortic valve being supported more extensively by fibrous structures.22–24 After placement of the pulmonary valve in the aortic position there is some evidence that there is down regulation of the smooth muscle cells in the pulmonary autograft, along with fracture of the elastin fibres and progressive structural disorganization.24, 25
Several studies have evaluated the function of the neo-aortic valve after the Ross procedure. Freedom from severe neo-aortic regurgitation at 10 years ranges from 75 to 90%.9, 26–28 This is in contrast to children undergoing replacement of the aortic valve with a mechanical prosthesis, where failure is usually due to thrombosis or ingrowth of pannus,16, 29, 30 or “outgrowth” of the valve resulting in obstruction of the left ventricular outflow tract.16, 29, 30 In patients who have undergone the Ross procedure, failure due to significant neo-aortic regurgitation may occur early in the post-operative period, or develop over time. Early failure is rare, and has been associated with subtle abnormalities of the native pulmonary valve, such as mild thickening of the leaflets, a quadrifoliate arrangement, or a discrepancy in size of the leaflets, that may be congenital or acquired after repair of a ventricular septal defect, even in the absence of significant pre-operative insufficiency of the native pulmonary valve.7 The development of neo-aortic regurgitation over time may be related to dilation of the newly constructed aortic root.9, 31, 32
The pulmonary valve also functions as the neo-aortic valve after the arterial switch procedure for repair of transposition, and staged reconstruction for hypoplastic left heart syndrome. Studies in both these groups of patient have shown progressive dilation of the new aortic root over time.33–35 Unlike patients who have undergone the arterial switch operation and staged reconstruction for hypoplastic left heart syndrome, however, there are two important differences for those undergoing the Ross procedure. Prior to the Ross procedure, the pulmonary valve has been subjected to the low-pressure pulmonary circulation, sometimes for decades, and needs to adapt acutely to an increased afterload. In addition, patients who have undergone the Ross procedure have suture lines above and below the neo-aortic valve. It is not surprising, therefore, that the pulmonary autograft, when placed under systemic pressure and resistance, dilates out of proportion to somatic growth.21, 36–38 Dilation has been found at both the basal level of the neo-aortic valve and the valvar sinuses.36, 37 Tantengco et al.36 showed a significant increase in the z-score for the diameter of the valvar orifice from 1.4 at the time of hospital discharge to 2.6 after follow-up of 6 months, and an increase for the diameter of the sinuses from 2.0 to 3.3 over the same period of time. Follow-up data on dilation of the root beyond 6 months is conflicting, with some studies demonstrating stabilization of the dimensions, and others suggesting continued dilation over the next 1 to 2 years.11, 36–39 Serial echocardiographic measurements in children are lacking beyond the early post-operative period. Data in adults suggests that dilation progresses over time.40–42 Simon-Kupilik et al.42 found that those showing freedom from dilation of the newly constructed aortic root decreased from 80% after follow up of 1 year, to 45% after 7 years. Predictors of such dilation include younger age at surgery, pre-operative dilation of the ascending aorta in presence of a bifoliate aortic valve, and pre-operative dilation associated with aortic insufficiency.28, 41
Reintervention
Although the pulmonary autograft has the potential for growth after the Ross procedure, the homograft placed in the pulmonary position does not. Hence, reintervention must be expected when the procedure is performed in young patients. Right ventricular reintervention is nearly always due to stenosis of the pulmonary homograft, with or without pulmonary insufficiency, and may require replacement or augmentation of the conduit, balloon dilation, and/or stenting. Right-sided reintervention is necessary in up to one-fifth of patients at midterm follow-up.7, 9, 26, 27, 43 Predictors of reintervention in the right ventricular outflow tract of children include utilization of an aortic homograft, which tends to calcify more quickly than pulmonary homograft, small size of the initially inserted homograft, and younger age at placement, longer storage time of the homograft, and immune mediated reactions.43, 44
Reintervention in the left ventricular outflow tract is as frequent as reintervention on the right side, occurring in approximately one-tenth of patients after follow-up of 10 years, and in almost two-fifths after follow-up of 25 years.9, 21, 26, 27 Rates of reintervention for other types of valvar replacements and aortic valvoplasty are higher. Mechanical valves have a reintervention rate of up to one-fifth at follow-up of 10 years. For those with mechanical valves, the reintervention rate is dependent on the age of the patients in the cohort. The younger the patients, the higher the rate of reintervention.16, 29, 30, 45 Of patients undergoing aortic valvoplasty one-fifth have required reintervention at 3 years,20 while half and two-thirds, respectively, of those having replacements with aortic homografts and xenografts have needed reintervention at 10 years.16
The majority of reinterventions on the left side for patients after the Ross procedure are performed due to significant neo-aortic insufficiency, as discussed above, and involve replacement of the valve, usually with a mechanical prosthesis.9, 26, 27 Predictors of left sided reintervention include an abnormal native pulmonary valve, and pre-operative aortic insufficiency with dilation of the native aortic root.7, 46 Valve sparing replacements have also been reported in patients who develop severe dilation without significant neo-aortic insufficiency.26 It is not known whether, in some patients, dilation and insufficiency of the newly constructed aortic root will continue to progress over time, and whether this will lead to higher rates of reintervention in the second to third decades of follow-up. It is also unclear if the same factors that put the patient at risk for dilation or insufficiency of the neo-aortic root also apply to the root subsequent to reintervention, or whether other risk factors will be identified.
Issues relating to the coronary arteries
There are few studies assessing flow in the myocardium and coronary arterial flow reserve after reimplantation of the coronary arteries in the Ross operation. Hauser et al.,47 in a small series, showed that coronary arterial flow reserve was significantly reduced relative to normal controls in all patients undergoing the arterial switch, while it was normal in those having the Ross procedure, with no exercise induced defects of perfusion. It is not clear, however, whether the differences related to the extent of follow-up between the two groups, the age at surgery, or inherent abnormalities of the coronary arteries in those undergoing the arterial switch that predisposed them to abnormalities of flow.
Left ventricular size and function
At the time of the Ross procedure, myocardial ischaemia secondary to cardioplegic arrest may exacerbate pre-existing left ventricular dysfunction. There is often a mismatch between ventricular mass and volume, with acute reduction of end-diastolic volume in patients with severe aortic insufficiency and/or mitral regurgitation in the presence of a hypertrophied myocardium. As long as significant residual neo-aortic regurgitation is not present, the left ventricle remodels over time, with a reduction in left ventricular end diastolic dimension, left ventricular end systolic dimension, and left ventricular mass.12, 36, 48, 49 The majority of patients have a normal or increased shortening fraction or ejection fraction before, and have normal function after the Ross procedure.12, 49 Gauthier et al.,48 for example, showed that two-thirds of patients with severely depressed left ventricular function before the Ross procedure had an increase in ejection fraction of greater than one-fifth after the procedure.
Exercise performance
There is limited data on exercise performance after the Ross procedure is carried out in chidlren. In a study performed by ourselves,50 we found no significant change in maximal uptake of oxygen indexed to ideal body weight before and after the Ross procedure. In four-fifths of the patients, aerobic capacity improved or was stable after the Ross procedure, and there was no post-operative chronotropic impairment. The patients, however, had significantly increased adiposity after the Ross procedure.
Electrophysiologic abnormalities
The early electrophysiologic abnormalities after the Ross procedure are well documented, and include complete heart block in up to one-twentieth, and ventricular tachycardia in one-quarter, both arrhythmias being seen in the first and second days after surgery.4, 7, 51, 52 After the combined Ross and Konno procedure, however, complete heart block has been reported in up to one-fifth of patients.5, 7, 51 Systematic long-term data on the electrophysiologic outcome, however, is sparse. Reports of persistent ventricular tachycardia requiring medication or automatic internal cardiac defibrillator are rare, and the incidence of placement of pacemakers after the early post-operative period is very low.51
Modifications of the procedure, and other potential valves
So as to minimize dilation of the newly constructed aortic root, some surgeons have described “wrapping” the autograft with pericardium or the ascending aorta, or with Dacron or Teflon felt at the level of insertion of the pulmonary autograft, albeit with mixed results.9, 53, 54 These modifications may have had limited success because dilation of the root is not limited to just the base of the autograft, but instead involves its full length. In an attempt to support fully the entire autograft, Slater et al.55 suggested that, for adults, the autograft could be encased in a Dacron tube, since subsequent growth is not necessary. They performed their modification in a small cohort of patients, and as yet follow-up data is not available.
Tissue engineered valves, with autologous cell seeding on a polymer leaflet scaffold, may prove to be ideal for replacement of the aortic valve. Efforts thus far to perfect the development of durable valves have been difficult. Loss of cells, and deficient polymers resulting in destruction of the scaffold, have limited the ability for researchers to put these valves into the systemic circulation.56, 57 Instead, some companies and researchers have focused on improving the performance of bioprosthetic valves with technology to reduce calcification. Kanter et al.,58 for example, inserted the Medtronic Freestyle Porcine Aortic Root in 56 children, with no deaths, mild or no pulmonary insufficiency in nine-tenths of the patients, with a mean peak systolic gradient across the right ventricular outflow tract of 19.7 plus or minus 15.4 millimetres of mercury after two to three years. Gleason et al.59 achieved comparable results using the porcine St. Jude Toronto Bioprosthesis.
Conclusion
Insertion of the pulmonary autograft provides superior haemodynamics, does not require anti-coagulation, grows with the patient, and has the best long-term durability of the present options available for replacement of the aortic valve using a tissue prosthesis. The Ross procedure, therefore, is the procedure of choice when it is necessary to replace the aortic valve in the neonate, infant, and young child. The procedure is contra-indicated, however, when the native pulmonary valve is structurally abnormal. Whether or not the newer techniques of aortic valvoplasty will have equal or better durability than the pulmonary autograft is not known. Durability of the various techniques for aortic valvoplasty may increase as there is improvement in matching appropriate candidates with specific surgical techniques. To date, tissue engineered valves have not realized the potential that was initially theorized, and it is not clear whether modification of the technique used for implantation of the pulmonary autograft will change the incidence over time of dilation and insufficiency of the neo-aortic root, and left sided reintervention. Given the issues of dilation and insufficiency, and the potential increasing risk of left sided reintervention over time, we suggest that the Ross procedure should be performed only if the patient is not suitable for aortic valvoplasty. Given that all options for valvar replacement are still available after aortic valvoplasty, in appropriate patients an attempt at valvoplasty should be considered prior to undertaking the Ross procedure. Longer term issues of life long neo-pulmonary regurgitation, obesity, exercise performance, coronary arterial perfusion and flow reserve, and risks of atherosclerosis remain to be defined.
