Ann Thorac Surg 2009;88:672-674. doi:10.1016/j.athoracsur.2009.01.023
© 2009 The Society of Thoracic Surgeons
Case Reports
Severe, Transient Aortic Insufficiency After Bovine Pericardial Aortic Valve Replacement: A Cautionary Note
Thomas J. Vander Salm, MDa,*,
Ann J. Toran, MDa,
Wilfred Lewis, MDb,
Johanna O'Connorb
a Department of Cardiac Surgery, North Shore Medical Center, Salem, Massachusetts
b Department of Anesthesia, North Shore Medical Center, Salem, Massachusetts
Accepted for publication January 12, 2009.
* Address correspondence to Dr Vander Salm, North Shore Medical Center, 81 Highland Ave, Salem, MA 01970 (Email: tvandersalm{at}partners.org).
 |
Abstract
|
|---|
Three patients demonstrated severe, central aortic insufficiency noted immediately upon removal of the aortic cross clamp after aortic valve replacement with a bovine pericardial Edwards Magna valve. After maintaining left ventricular decompression, the aortic insufficiency resolved in less than 1 hour. Knowledge of this phenomenon will prevent unnecessary prosthetic valve explantation and re-replacement.
|
Introduction
|
|---|
Prosthetic valves are expected to be intrinsically competent, and to demonstrate that competency immediately upon implantation. We had 3 patients who had four bovine pericardial aortic valves (Magna, Carpentier-Edwards Perimount, model 3000; Edwards Lifesciences, Irvine, CA) implanted that were severely incompetent immediately after aortic valve replacement (AVR), of whom the first patient had two valves implanted. Within 1 hour, the aortic insufficiency (AI) disappeared spontaneously. This phenomenon does not appear to be widely recognized.
|
Case Reports
|
|---|
Patient 1
Patient 1, a 59-year-old woman who had severe aortic stenosis of a bicuspid valve, was unwilling to commit to permanent anticoagulation. At the completion of a routine AVR implantation using a 23-mm Magna pericardial valve (Carpentier-Edwards Perimount), it appeared that one of the struts might interfere with the left main coronary ostium, so the valve was explanted, a new but identical valve was reimplanted with a different orientation, and the aorta was closed. Upon release of the aortic cross clamp, severe central (transvalvular) AI was present, as ascertained by transesophageal echocardiography and by the amount of blood returning from the left ventricular (LV) drain.
Disconnecting the LV drain demonstrated a strong jet of return during the initial asystole, and as a normal rhythm returned, during both systole and diastole. With a left ventricular (LV) drain on suction, distention did not occur, and the patient had normal LV function and ejection, albeit while still on cardiopulmonary bypass. With no improvement in 40 minutes, the aorta was cross clamped, and reopened. This second prosthesis was explanted and replaced with a 21-mm Magna valve. The downsizing was necessary because we had no further inventory of the 23-mm valves; the previous valves were implanted with ease and were not oversized. Examination of the explanted valve demonstrated no obvious defect, although at the top of one of the commissures, one of two adjacent leaflets was approximately 0.05 mm lower than the other. This did not seem to be causal to the AI. After aortic closure and release of the aortic cross clamp, severe AI was demonstrated again, and it caused LV distention until the LV drain suction was increased to greater than 600 mL/min. The LV function seemed to be normal, and after waiting approximately 20 minutes, the AI lessened and then disappeared. Her convalescence was difficult, but it was eventually complete.
Patient 2
Patient 2, an 86-year-old man with a recent myocardial infarction and aortic stenosis (valve area 0.4 cm2) had a double coronary artery bypass, and an AVR using a 23-mm Magna bovine pericardial valve. The implanted valve appeared normal in configuration and in its position in the aortic root. Upon removal of the aortic cross clamp, severe, central AI was present as ascertained by the transesophageal echocardiography and the magnitude of return from the LV drain. After approximately 20 minutes, the AI lessened and then disappeared.
Patient 3
Patient 3, a 79-year-old man with progressive dyspnea had three-vessel coronary artery disease, aortic stenosis with a gradient of 60 mm Hg, and an area of 0.7 cm2. His operation consisted of a triple coronary artery bypass and an AVR using a 23-mm Magna bovine pericardial valve. With release of the aortic cross clamp, ventricular fibrillation occurred, and LV distention occurred despite LV drain suction of approximately 600 mL/min. Suction was increased to lessen the LV distention. The transesophageal echocardiography confirmed central AI. After 30 minutes, AI lessened and then disappeared.
In all cases, valve suture techniques were identical. The valves were implanted in a supra-annular position using Teflon felt pledget (Ethicon Inc, Somerville, NJ) buttressed horizontal mattress sutures, which were passed from the ventricular surface to the aortic surface of the annulus. Maximum spacing within each mattress suture was 5 mm. In each case, the valve seated comfortably onto the tissue annulus, and there appeared to be no distortion of the leaflets or of the prosthetic annular stent.
|
Comment
|
|---|
Prior to using the Magna bovine pericardial valves, we had not encountered the problem of severe but spontaneously resolving AI immediately after AVR. The AI encountered in the first patient caused us to re-replace the valve. With the third valve in the initial patient, and the valves in the second and third patients, waiting for the AI to disappear spontaneously allowed us to avoid valve re-replacement. There seems to be little information on this occurrence. In a 1984 case report, a porcine mitral valve replacement was performed without incident [1]. Two months later, severe mitral regurgitation required valve re-replacement where the explanted valve leaflets all seemed to resist maintaining a closed position, despite otherwise normal appearance of the leaflets. A recent publication by Buklas [2] described a patient only slightly more similar to ours. Five days after a routine AVR using a Carpentier-Edwards pericardial valve, severe AI was noted. Echocardiography showed one of the three valve leaflets fixed in an open position. Upon valve explantation, no abnormalities were found. Fleisher [3], in 2004, reported two cases of AVR with a Carpentier-Edwards Perimount (model 2700) aortic valve in which AI was noticed at 1 and 3 days postoperatively, respectively for each case [3]. But the AI progressed and both valves were explanted and re-replaced. In both removed valves, one elongated, prolapsing cusp was found. In our cases, a different valve (model 3000 Magna) was used, and the AI was present and severe immediately upon release of the aortic cross clamp, but completely resolved in less than an hour with no subsequent valve malfunction.
The package insert accompanying the Magna valve states, "Due to the relative flexibility of the frame, care must be exercised to prevent folding or deformation of the stent that may lead to regurgitation, altered hemodynamics, and/or leaflet disruption rendering the valve incompetent. In this regard oversizing must be avoided. The spacing of the sutures in the remnant of the valvular orifice and the prosthesis suture ring must be carefully matched to avoid folding of the leaflets or distortion of the orifice" (Carpentier-Edwards Perimount Magna, Edwards Lifesciences, Irvine, CA). These instructions also recommend against sutures spanning a distance of more than 10 to 15 mm. In none of our patients was the valve oversized; in fact, in each case, a one size larger valve could have been inserted. As the package insert notes, the stent is flexible and can be deformed. Also, the stent is not monoplanar, and dips down to better fit the native annular contour. We hypothesize that valve sutures lacking congruent spacing between the tissue and prosthetic annuli might exert torque or stress in a direction tangential to the prosthetic annulus and thereby slightly distort the leaflets, leading to abnormal leaflet coaptation and AI. With time, the tissue might conform better to the prosthesis, thus returning competence to the valve. This hypothesis could not be tested in these patients, but it is consistent with the time course of resolution of the AI.
The importance of this phenomenon is in its recognition, thus enabling surgeons to avoid the need to re-replace such valves as we believed we were forced to do in the first such patient we encountered.
|
References
|
|---|
- Tolis GA, Michalis A, Pouliou A, Ivros C, Bolos C, Sfyras N. Unusual complication of a Carpentier-Edwards porcine valve Tex Heart Inst J 1986;13:337-339.[Medline]
- Buklas D, Massetti M, Neri E, Chocron S. Built-in defect of a biological pericardial aortic prosthesis? Interact Cardiovasc Thorac Surg 2008;7:1110-1111.[Abstract/Free Full Text]
- Fleisher AG, Lafaro RJ, Moggio RA. Immediate structural valve deterioration of 27-mm Carpentier-Edwards aortic pericardial bioprosthesis Ann Thorac Surg 2004;77:1443-1445.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
K. Orihashi, T. Kurosaki, and T. Sueda
Everted leaflet of a bovine pericardial aortic valve
Interact CardioVasc Thorac Surg,
June 1, 2010;
10(6):
1059 - 1060.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
