Caution with explant of prestented percutaneous pulmonary valves with infective endocarditis

Ronald K. Woods, MD, PhD

Central Message.

Explanting the newer prestent percutaneous valves with infective endocarditis warrants heightened awareness and caution.

Many surgeons have had experience with explantation of Melody valves (Medtronic), with and without infective endocarditis, and with and without prestenting with a standard stent. We recently encountered a very unique situation—an infected SAPIEN (Edwards Lifesciences) pulmonary valve implanted within the Alterra (Edwards Lifesciences) prestent assembly. Our coauthor had also not seen this situation and was unaware of any reports of such a procedure. The Alterra device includes a polyethylene terephthalate skirt that does not cover the distal tines and is intended to be used with the SAPIEN valve with bovine pericardial leaflets—therefore, 2 layers of metal in all (Figure E1).¹ Moreover, the stent of the SAPIEN valve is designed to be stronger and not compressible compared with the typical stents we encounter as surgeons. The ends are flanged proximally and distally to allow anchoring. The intent and purpose are to provide a suitable size frame and mount in an outflow tract that would otherwise be too large to permit standard percutaneous deployment. In our specialty, the predominant focus is the post-transannular patch tetralogy of Fallot patient.²

Figure E1.

Picture of the Alterra Prestent and enclosed SAPIEN valve. The total height of the prestent is just less than 5 cm. Also note the size/gauge of the of the individual chromium–cobalt frame structures—thicker and less compressible than the typical strut material with which congenital heart surgeons are more familiar.

Case Detail

The patient was a 16-year-old young man with previous transannular patch repair of tetralogy of Fallot who met magnetic resonance imaging criteria for valve implantation and had the procedure performed by interventional cardiology with an excellent result—a 29-mm SAPIEN valve within the Alterra prestent, both positioned and seated well. He presented a few months later with staphylococcal bacteremia and obvious vegetations on the valve. Persistent fevers and laboratory evidence of mounting inflammation despite appropriate antibiotics were thought to indicate a need for surgical removal and replacement of the infected material. Sternal re-entry was uneventful; however, dense scar encased the pulmonary artery and left side of the aorta. A preoperative “bubble study” during transesophageal echocardiogram revealed no residual intracardiac shunt. Cardiopulmonary bypass was initiated with bicaval-aorto cannulation, a transright atrial right ventricular vent, and an active aortic root vent. The degree of inflammation and scarring on the left side of the aorta precluded separation of the aorta and pulmonary artery to allow placement of a crossclamp and cardioplegia administration. The right ventricle was opened just proximal to the palpable stent. Based on the computed tomography scan (Figure 1), there was concern that there may be a fractured stent strut near the right coronary artery, so consideration was given to potentially leaving some of the Altera stent in situ. However, when attempting to cut into the anterior double-stent structure, we found our usual instruments (heavy curved Mayo scissors, wire-cutting scissors) were ineffective. The only thing that worked was a sternal wire cutter. Even employing that device was very difficult and precluded a flush cut, consistently leaving tines protruding into the lumen. We then found a plane outside the skirted portion of the Alterra that permitted “safe” gentle dissection. Near the proximal right portion as we attempted to tease away the skirt, aortic blood entered the field. The skirt had covered the area where an Alterra frame joint (expanded by the SAPIEN valve stent) had compressed the aorta to such an extent as to cause remarkable thinning of the aortic wall, although no frank perforation was apparent before the dissection. The aortic bleeding was easily controlled with digital pressure, after which the cardiopulmonary bypass strategy was altered and cooling undertaken to permit deep hypothermic circulatory arrest and systemic hyperkalemia (as cardioplegia). After we achieved circulatory arrest, inspection through an anterior aortotomy revealed a 2-mm hole in the right coronary sinus of the aortic root anterior and proximal to the right–left commissural pillar, posterior to the right coronary ostium. The tissue appeared inflamed, consistent with extension of the regional endocarditis process. After gentle debridement, we placed a small bovine pericardial patch from within and then placed a much-larger bovine patch on the outside, carefully probing the coronary ostium to ensure patency. After deairing and aortotomy closure, flow was resumed and rewarming commenced. The extent of the inflammation mandated complete removal of the implanted system, which again was challenging distally because of the flared extension of the prestent system being imbedded in the wall of the pulmonary artery confluence. After complete removal of the involved native tissue and prosthetic material (Figure E2 is a picture of the explanted system), right ventricular outflow tract reconstruction was accomplished with a 29-mm pulmonary homograft. The patient recovered uneventfully and was discharged with ongoing antibiotic treatment. Assent and informed consent were obtained from the patient and parents specifically for including information about the case in this manuscript. Our report was reviewed by Children’s Wisconsin Research Compliance on February 1, 2023, and concluded it did not meet the requirement for institutional review board submission or oversight.

Figure 1.

Computed tomography axial frame depicting concerning appearance of the stent near the right coronary artery.

Figure E2.

Picture of the explanted prestent and SAPIEN valve system.

Discussion

We completely understand the motivation underlying this newer technology based on the hope that it will provide decades, if not a lifetime, of surgery-free valve-in-valve approaches. Although this may be a step in that direction, it remains a hope and far from proven. As this is newer technology, extensive long-term data are lacking. Also, we could not find any published reports describing explants of the Alterra/SAPIEN system with or without infective endocarditis.

We believe that this case is an example of how the ostensibly reasonable desire to avoid surgery may have unanticipated and unfortunate consequences. In particular, the dictum primum non nocere may be violated if the powerful radial forces delivered by a percutaneous stent against the wall of the delicate aortic root can cause potentially lethal injury if complicated by a destructive infectious process. To be clear, it is not the radial force of the Alterra prestent, per se, but rather the radial force of the SAPIEN valve compressing the Alterra stent against the aortic wall. For this particular patient, abundant data support both the early effectiveness, and, more importantly, the long-term safety of a second-time sternotomy for a patient with low surgical risk and minimal medical comorbidity undergoing placement of a large pulmonary homograft or stented bioprosthetic valve. Such a clinical scenario comes at a cost of a postsurgical patient with both inpatient and outpatient convalescence but has an excellent prognosis for endocarditis-free durability. For such a patient, we believe it is an as-yet-unjustified assertion that equipoise exists for the alternative of employing the strategy of prestenting with the flared prosthesis employed in this patient. The experience with these prestent mechanisms and newer-generation pulmonary valves is simply too early to characterize the longer-term risk of infective endocarditis, particularly if there is likely to be involvement of the aorta with proximal flare or the pulmonary confluence (and also potentially the aorta) with the distal flare. For teenager or a young adult with no major surgical contraindications, it may be reasonable to avoid an initial flared prestent strategy, at least outside the bounds of a clinical trial.

As has been the experience with aortic valve technology, which similarly uses a flared prosthesis, surgical explant has proven to be challenging and associated with nontrivial mortality.³ We recognize ours is a single case and may simply be a one-off. However, with extensive experience explanting other pulmonary valves and stents, we believe that explanting this prestent valve combination is entirely different and far more challenging because of the proximal and distal flared stent extension combined with the noncompressible stented frame of the valve itself. Although proceduralists are very careful to avoid coronary compression, there is less concern about aortic wall compression in the absence of new aortic insufficiency—the computed tomography of our patient (Figure 2) clearly shows some effacement of the aortic wall indicating nontrivial radial expansive force of the SAPIEN valve. Attempting to dissect this space would be imprudent, given the length of the Alterra stent and the radial force. Furthermore, even the slightest fracture in the stent with minimal protrusion can easily penetrate the aortic wall. Since these tend to be large valves and amenable to subsequent valve-in-valve approaches, for the most part surgical reintervention will either be for infective endocarditis or complications of a subsequent valve-in-valve approach, the forethought of both of which we shudder to contemplate. At a very minimum, these patients should be discussed in a multidisciplinary format and not simply be cardiology-to-cardiology referrals, a process that lacks gatekeeper function. Perhaps another 10 years of experience will prove us wrong, but we believe these devices should be reserved for patients with surgical risk factors other than a single previous sternotomy.

Figure 2.

Computed tomography axial frame depicting mild effacement of the left side of the ascending aorta indicating a nontrivial degree of radial force from the prestent and valve system. Although fracture of the SAPIEN stent is highly improbable, fracture of the less-rigid Alterra prestent could lead to aortic penetration. If not acute, and if fracture occurs in skirt-covered region, the skirt would likely prevent significant shunt and would only be detectable after removal of the system.