Follow-up of a simple method for aortic valve reconstruction with fixed pericardium in children

Abstract

In 2013, we published a simple method for aortic valve reconstruction in children using fixed pericardium in this journal. This was not expected to be a durable solution but a temporizing measure in the absence of other solutions, in order to buy time for growth that would allow subsequent prosthetic valve insertion. As such, this was implemented in 3 patients with excellent immediate results, as reported in our original publication. We are now writing to provide longer term information. In 2 cases, a newborn and a 12-year-old boy, the valve calcified heavily at 4 years’ follow-up. By then, the valvar annulus had grown sufficiently to accept a prosthetic valve in both cases. In these 2 cases, the material used for valve reconstruction was autologous pericardium fixed in 0.6% gluteraldehyde. In the third patient, a 3-year-old girl, the valve failed by dehiscence of the suture line 6 months following surgery. In this case, fixed bovine pericardium had been used, which is more rigid than autologous pericardium, thus possibly explaining the dehiscence. However, the pericardial patch was not calcified. Thus, the described method lived up to our expectation of buying time for growth for later insertion of a prosthetic valve in 2 of the 3 cases.

Aortic valve, Aortic valve reconstruction, Aortic valve repair, Aortic valvuloplasty

In 2013, we published a simple method for aortic valve reconstruction in children using fixed pericardium in this journal [1]. Essentially, the technique consists of first constructing a pericardial sleeve within the aortic root and then fashioning the commissures by tethering that sleeve to the aortic wall along 3 evenly spaced vertical lines. This was not expected to be a durable solution, given that fixed pericardium is prone to early calcification, especially in the young [2–4]. For this reason, it was presented as a technique that may be useful only when there are no other options, such as the patient being too small for a prosthetic valve and the Ross procedure being contraindicated for some reason. In such circumstances, the described method may rescue the patients and buy time for them to grow so that they may be offered a better solution later, i.e. insertion of a prosthetic valve. As such, this technique was used in 3 cases, as reported in our original publication: (i) a newborn boy with common arterial trunk (truncus arteriosus) and severe mixed truncal valve disease; (ii) a 12-year-old boy with severe mixed aortic valve disease and a small aortic valve annulus (14 mm) in whom a major coronary artery crosses the infundibulum, thus contraindicating the Ross procedure; and (iii) a 3-year-old girl previously operated on for correction of an aortoventricular tunnel, with severe mixed aortic valve disease and gross dilatation of the aortic root and ascending aorta that we repaired by plication in the same operation as that of the aortic valve reconstruction. In this case, the Ross procedure was an option but considered high risk in view of an unusual take-off of her right coronary artery—from the wall of the previously corrected aortoventricular tunnel, with an intramural course. Therefore, we decided to avoid the Ross procedure, even though this was not contraindicated, in favour of valvar reconstruction, which we considered safer. The immediate results were excellent in all the 3 cases, as reported in our original publication. We are now writing to provide longer term information.

In the first 2 cases, the reconstructed aortic valve functioned well in the first 3 years. During the fourth year, however, valvar function deteriorated rapidly. By 4.5 years, there was severe calcification and mean echocardiographic gradients of 60–70 mmHg. By that time, the valvar annulus had grown sufficiently to accept a prosthetic valve in both cases.

In the third patient, the 3-year-old girl, there was dehiscence of the suture line at the base of the pericardial sleeve (i.e. technical failure) at 6 months’ follow-up, such that valvar insufficiency was seen echocardiographically from then onwards. This increased in severity such that reintervention became necessary before the valvar annulus had grown as much as we would have liked, i.e. not sufficiently for the insertion of a prosthetic valve. Therefore, the Ross procedure was offered reluctantly for the want of other options, despite our efforts to avoid this in the previous operation in view of the above-mentioned coronary anomaly. This was carried out successfully, despite the coronary anomaly, 18 months after the original valvar reconstruction.

Interestingly, in the first 2 cases, valvar reconstruction had been done with autologous pericardium fixed in 0.6% gluteraldehyde. In the third case, there was no suitable autologous pericardium since the patient had undergone surgery before. Therefore, fixed bovine pericardium was used instead. This is noticeably more rigid than autologous pericardium, which may possibly explain the dehiscence. Theoretically, it is also possible that dehiscence was caused by poor suturing rather than by the material, but we consider this unlikely since failure due to poor suturing is usually immediate, whereas dehiscence in this case occurred 6 months later. However, the patch was not calcified, which is not surprising, given that technical failure did not allow this patch to remain in place long enough for calcification.

Thus, the described method lived up to our expectation of buying time for growth for later insertion of a prosthetic valve in 2 of the 3 cases. These findings also suggest that this method may buy no more than about 4 years for growth, although this is based on only 2 cases.

Conflict of interest: none declared.