Percutaneous Pulmonary Valve Implantation

Pulmonary Stenosis

Native PV or RVOT obstruction occurs in approximately 20% to 30% of patients with CHD, in isolation or in association with other congenital syndromes such as tetralogy of Fallot (TOF), Noonan syndrome, or congenital rubella. Up to 90% of isolated pulmonary stenosis (PS) is valvular, illustrated by systolic cusp doming caused by fusion of one or more of the valve commissures. Occasionally, this may be in the setting of a bicuspid, quadricuspid, or dysplastic valve (e.g., Noonan syndrome) where there is little valve fusion but associated valve thickening, a small annulus, and short main pulmonary artery (PA). In addition to the valvular lesion, obstruction can occur in the subvalvular region in the RVOT, such as in a double-chambered right ventricle resulting from muscle bundles, or a supravalvular region, as seen in Williams syndrome. Figure 8–1 illustrates the angiographic appearance of subvalvular, valvular, and supravalvular PS. TOF, the most common form of cyanotic heart disease, is represented by underdevelopment of the subpulmonary outflow caused by anterior deviation of the infundibular septum, resulting in the aorta overriding the ventricular septum and a malalignment ventricular septal defect (VSD).

Depending on the severity of the PS and adequacy of pulmonary blood flow, clinical symptoms may range from neonatal compromise with varying degrees of right ventricular (RV) hypoplasia requiring urgent neonatal intervention to the gradual development of PS symptoms as the obstruction worsens with time. Generally, PV gradients less than 20 mm Hg do not increase with the passage of time, whereas higher gradients may worsen. Increased RV systolic pressure leads to RV remodeling with either myocardial hyperplasia (when the exposure is during fetal or neonatal development) or hypertrophy outside the first months of life. This increase in RV mass decreases ventricular compliance and can progress to RV dilation and dysfunction in end stages. When PS manifests late, clinical manifestations include exertional intolerance, development of supraventricular or ventricular arrhythmias, endocarditis, cyanosis (in the presence of a concomitant interatrial shunt), and additional right-sided valvular disease (tricuspid or PR). Prognosis and symptoms are primarily determined by the severity of obstruction.

Severe valvular PS, when isolated, can be treated by either surgical valvotomy or as in the past several decades, by percutaneous balloon valvuloplasty, both with excellent short- and long-term results in children and adults. However, most patients with subvalvular, supravalvular, or multilevel obstruction require surgical correction. In children, surgical options for PVR have to be balanced with future growth in addition to the individual anatomy. Bioprosthetic valves, homograft, and synthetic conduits (both valved and nonvalved) have been employed with varying frequency as valve replacements. Complete surgical repair of TOF typically includes reduction or elimination of the valvular and subvalvular RVOT obstruction, often a transannular patch reconstruction of the RVOT (particularly in the symptomatic infant), and closure of the malalignment VSD. As such, TOF repair often leads to significant pulmonary regurgitation (PR).

Pulmonary Regurgitation

As noted, PR commonly occurs after surgical repair of TOF, especially with the use of a transannular patch. However, significant PR can also be found as an isolated condition, as a consequence of balloon or surgical pulmonary valvotomy, in the presence of bilateral PA or peripheral PA stenosis, or as a consequence of increased pulmonary vascular resistance. Over time, nearly all surgical implants (bioprosthetic valves and conduits) in the pulmonary position become dysfunctional and develop some combination of PR and PS requiring additional management.

For many years, the contribution of PR to patient morbidity and ventricular dysfunction were overlooked, partly because PR can be well tolerated for a long time. However, approximately 30% of patients with chronic severe PR develop symptoms by the fourth decade of life. In the repaired TOF population there is considerable variability in the anatomic dimension of the RVOT, pulmonary annulus, and branch PAs confounded by variations in TOF surgery that influence PR severity, the rate of RV dilation, and the development of symptoms. The regurgitant volume back across the PV is driven by the diastolic pressure differential between the PA and right ventricle. This pressure differential is in turn determined by: 1) the RV compliance related to the degree of RV hypertrophy, stiffness, and fibrosis; 2) PA afterload consisting of the pulmonary vascular resistance and the presence of any PA stenosis; and 3) intrathoracic and airway pressure changes, typically contributing in the acute rather than chronic setting.

Over time, significant PR results in RV dilation. As a consequence, pressure–volume hemodynamic studies have shown RV volumes to increase while left ventricular (LV) volumes decrease. In this setting, both ventricles have a diminished ability to augment ventricular output in response to inotropes. LV underfilling occurs as a consequence of RV dilation and secondary to septal bowing into the LV during diastole. The lack of mechanical–electrical synchrony between the ventricles also contributes to an overall reduction in biventricular function and efficiency. Further dilation of the right ventricle can lead to RV dysfunction with symptomatic heart failure and secondary complications such as cardiac cirrhosis. Secondary tricuspid regurgitation can develop from annular dilation, further exacerbating the volume load on the right heart. Valvular regurgitation is associated with inhomogeneous electrical activation and slowed conduction velocities that are proarrhythmic and promote ventricular and supraventricular arrhythmias.

The spectrum of clinical presentations secondary to chronic PR and RV dilation include dyspnea, fatigue, activity limiting palpitations, syncope, edema, diminished exercise capacity, and sudden cardiac death. Several observational studies have suggested an association between PR and long-term outcomes including mortality. In these studies, cardiac death was usually attributed to heart failure or sudden cardiac death.

In the subset of patients who proscribe progressive and/or debilitating symptoms secondary to significant PR and RV dilation, the choice to offer a PVR is an easy one. The clinical decision making is less straightforward in those patients who deny symptoms or a change in functional status. Objective evidence of severe RV dilation or RV dysfunction attributable to PR is an indication for PVR in asymptomatic patients, though there remains diagnostic uncertainty as to the acceptable degree of RV dilation before proceeding with PVR. Small nonrandomized studies have looked at imaged-based parameters and suggested several cut-off points (i.e., an RV–end diastolic volume index [RVEDVI] of greater than 150 to 170 mL/m ) to guide decision making for surgical intervention, in addition to objective findings on formal exercise testing and Holter monitoring. Regardless of the exact timing for intervention, a majority of patients with severe PR require a PVR at some point.

Types of Prosthesis

The choice of prosthesis choice must take into account age, potential for growth, and the need for future reoperations or interventions. Mechanical valves are not implanted in the pulmonary position secondary to a high thrombosis risk observed during initial clinical experience. although this has been contested in more recent retrospective series. Homografts are commonly used in the primary correction of children with complex RVOT obstruction (e.g., TOF with pulmonary atresia or truncus arteriosus) because of the availability of small diameters. Retrospective surgical series of such implants suggest 75% to 85% 10-year freedom from reoperation, although significant implant dysfunction may be present in up to 50% at 10 years.

Bioprosthetic implants are xenograft-based valves, constructed from either animal valves (porcine, bovine, or equine) or pericardial tissue. In the basic design, these valves are either “stented” when constructed within a sewing ring (usually three metal struts) or “stentless.” There is no evidence that either bioprosthetic design is advantageous, although one retrospective series suggested that stentless valves pose an independent risk factor for redo PVR at 10 years. In this study, other risk factors for bioprosthetic failure included younger age of valve implantation and a diagnosis of pulmonary atresia with VSD. Surgical series for bioprosthetic PVR show rates of 10-year valve dysfunction to be as high as 80%, with 10-year freedom from redo-PVR ranging from 52 to 86%.

Monocusp valves are constructed individually by the surgeon to fit the patient’s anatomy using pericardial tissue or polytetrafluoroethylene (PTFE). Although they are described as having good short- to medium-term functional characteristics, their longevity has been questioned, due to variable surgical technique. As they must be modeled in the operating room, monocusp valve implantation may increase bypass and operative times. One published surgical series reported a 52% incidence of significant PR, although freedom from reoperation at 10 years was 82%.

Prosthesis Durability

Whereas some study cohorts have suggested improved longevity for homografts compared with bioprosthetic valves, there is little prospective evidence. Choice of valve is influenced by local availability and operator preference. The largest published pediatric surgical PVR series originates from Toronto and includes 945 valves, including homografts, bioprosthetic valves (porcine and pericardial), and bioprosthetic valved conduits. Overall, there was an 81% freedom from reoperation at 5 years, 58% at 10 years, and 41% at 15 years. The risk factor associated with valve deterioration in the entire study group was a younger age at valve implantation (median age for surgery was 6.2 years). Risk factors in the 260-patient subgroup aged 13 to 65 years included smaller normalized valve prosthesis size for patient size, placement of endovascular stents, and increased number of previous valve placements. There was also a suggestion that homografts, conduits, and pericardial based bioprosthetic valves may be preferable to porcine bioprosthetic valves for long-term durability.

Valve degradation or homograft conduit dysfunction over time is theorized to be secondary to gradual calcification and fibrosis of the chemically or cryopreserved valve leaflets, in addition to an autoimmune response from the host. Alternate means of valve preservation, such as fresh decellularized pulmonary homografts may eventually result in increased freedom from explanation. Although anti-calcification treatments and newer valve designs focus on improved valve durability, surgical valve replacements have limited durability, with the majority of patients requiring reoperation between 5 and 15 years.

Rationale for Percutaneous Pulmonary Valve Implantation

With low morbidity and mortality rates, surgical PVR represents one of the medical success stories of cardiothoracic surgery. However, there are many instances when operative risks are high or surgery is prohibitive. Patients requiring RVOT reconstruction require multiple sternotomies and cardiac surgical procedures throughout their lives. A second, third, or even fourth reoperation is associated with increased surgical risk and perioperative mortality from adhesions, scar tissue, and anatomic disruption, and each subsequent ventriculotomy increases the risk of developing scar-related ventricular arrhythmias. Redo sternotomies can be especially challenging when the RV-to-PA conduit or valve lies in an anterior position immediately behind the sternum. Other patients are simply not surgical candidates because of comorbidities, and those with concomitant physical or mental challenges can make recovery after traditional surgery a daunting task.

The inspiration leading to the development of surgical alternatives arises from an innate human desire to minimize risk. Intuitively patients are looking for resolution and treatment of their medical problems but with minimal discomfort and recovery. There remain constraints to cardiac surgery, including incisional pain, wound healing, cosmetic appearance, length of rehabilitation, and the complications related to cardiopulmonary bypass. These concerns are compounded by the reality that many patients requiring RVOT reconstruction are young and will need multiple reoperations in their lifetimes as they outgrow their implant or for prosthesis failure. A balance between the deleterious effects of RVOT dysfunction and the need to minimize the total number of lifetime surgeries for a given patient is a clinically challenging task prompting the need to develop minimally invasive valve therapies.