| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | SEARCH |
|
MMCTS
(June 26, 2009). doi:10.1510/mmcts.2009.003905 Copyright © 2009 European Association for Cardio-thoracic Surgery
Procedure Aortic valve and/or aortic root replacement using an aortic homograftClinic for Cardiovascular Surgery, University Hospital Bern, 3010 Bern, Switzerland * Corresponding author: Tel.: +41-31-6322375; fax: +41-31-6324443. thierry.carrel{at}insel.ch
Aortic valve replacement using the homograft valve has a special place in the cardiac surgical practice although it has never been widely used, in part due to the lack of tissue donors but also due to the perceived difficulty of the procedure compared with aortic valve replacement using prosthetic devices and concerns regarding homograft valve failure. The principal indication for aortic valve replacement using a homograft aortic valve is for active aortic valve endocarditis (native or prosthetic) with or without perivalvular tissue destruction (abscess cavity, fistula, detachment of the anterior mitral valve leaflet from the aortic annulus). Since the homograft tissue is pliable and adaptable, it can be used to repair defects in complex cases with root destruction. A second interesting application of homograft aortic valve is in the treatment of small aortic root (in which replacement with a small prosthetic valve would produce an unacceptable orifice ratio and ultimately affect long-term outcome) and left ventricular outflow tract obstruction when the homograft aortic valve can be combined to the Konno procedure. Homograft aortic valve can be used in these cases without sacrifying the pulmonary valve to be used as aortic valve substitute (for instance in adolescents and young adults who do not want to undergo a ‘two-valve’ procedure like the Ross–Konno procedure). The most frequent operation technique is the cylindrical aortic root replacement, performed in a similar way than the classical Bentall procedure. This technique is the most easiest one and performed more frequently than the subcoronary implantation, which is substantially more demanding. Results from centres that have significant experience with homograft valve surgery report equivalent survival data. The University of Alabama had an 87% survival at 5 years in a 10-year period from 1981 to 1991. Of those who underwent isolated aortic valve replacement with a homograft, there was 99% survival at 30 days and 94% survival at 8 years. In the Mayo Clinic series, 82% were alive at 8 years. Like all tissue valves, homograft aortic valves may fail. An understanding of the mechanisms of homograft valve failure and the way in which these mechanisms interact has important surgical implications. Homograft aortic valves may develop progressive regurgitation as a result of a change in the mechanical properties of the leaflets over time.
Key Words: Aortic valve surgery • Full root replacement • Homograft • Subcoronary implantation • Surgical technique
Aortic valve replacement using the homograft valve has a special place in the cardiac surgical practice although it has never been widely used, in part due to the lack of tissue donors but also due to the perceived difficulty of the procedure compared with aortic valve replacement using prosthetic devices and concerns regarding homograft valve failure. Although vascular homografts were used in animal experimentation as early as in the beginning of the 20th century by Alexis Carrel, their clinical use occurred some decades later when Gross performed aortic coarctation repair with an arterial homograft and Dubost repaired an infrarenal abdominal aortic aneurysm with a homograft [1,2,3]. Lam implanted a homograft aortic valve into the descending aorta in a canine experiment in 1952 [4]. Later, homograft aortic valves had a short investigational and clinical application by insertion into the descending thoracic aorta by Murray and Beall [5, 6]. The first subcoronary implantation was performed by Donald Ross (single suture line technique) in 1962 and shortly after Barratt-Boyes and Duran (double suturing line method for the subcoronary technique) were among the first to use an aortic valve homograft to treat aortic valve pathology [7,8,9]. Later, O'Brien developed the technique of cryopreservation and reported in 1987 about excellent results after implantation of cryopreserved human valves [10].
The sources of homograft aortic valves include organ donors where the heart is not being used for transplantation, autopsies (within 24 h of death) and excised cardiac transplant recipient hearts. There is obviously a need for sterilization and storage of homograft aortic valves and an extensive discussion of these aspects is beyond the scope of this brief review. Method of homograft valve preparation has been described in detail elsewhere [11, 12]. The aortic valve, the ascending aorta, the aortic arch and in some cases the complete descending aorta (for vascular homograft) as well as the pulmonary valve and the common pulmonary artery with its branches along with the heart, are removed through a median sternotomy from subjects who serve as tissue donors. The procedure is done aseptically in the operating theatre. Heart and the great vessels are removed in the same manner as for cardiac transplantation. Rinsing with saline solution removes blood from the cardiac cavities and the aorta. The heart is placed in three different plastic bags (the first containing saline solution and the second ice) to maintain the temperature at 4 °C during the transfer to the tissue processing laboratory. Valves are accepted from donors up to age 60. Exceptions are death from endocarditis, septicaemia, hepatitis and HIV. Carcinoma is usually not a contraindication. The valve cusps must be free of atheroma and any calcification, normally formed and tricuspid, and free of major fenestrations that may weaken the commissural cusp attachment. The valve is stored at 4 °C and transferred to fresh antibiotic solution after 24 h. Thereafter, the valve is prepared under sterile conditions for freezing, by first placing it in tissue culture medium containing 10% dimethyl-sulfoxide (DMSO) and transferring it to a plastic bag. This bag is then placed in two additional bags of increasing size. The homograft is then frozen in a freezing chamber while freezing is controlled at 1 °C per min to –40 °C. At this temperature, the homograft is transferred to its permanent storage in liquid nitrogen. The valve is packed in a polystyrene container and suspended in the vapour phase of a liquid nitrogen refrigerator at about –180 °C.
When needed for insertion, the frozen homograft in its container is removed from its storage position and placed in a container containing
Indications for insertion of a homograft aortic valve The principal indication for aortic valve replacement using a homograft aortic valve is for active aortic valve endocarditis (native or prosthetic) with or without perivalvular tissue destruction (abscess cavity, fistula, detachment of the anterior mitral valve leaflet from the aortic annulus). Since the homograft tissue is pliable and adaptable, it can be used to repair defects in complex cases with root destruction. There is strong evidence that homograft aortic valves may be very resistant to infection even if they are implanted in a situation with active purulent process. A second interesting application of homograft aortic valve is in the treatment of small aortic root (in which replacement with a small prosthetic valve would produce an unacceptable orifice ratio and ultimately affect long-term outcome) and left ventricular outflow tract obstruction when the homograft aortic valve can be combined to the Konno procedure. Homograft aortic valve can be used in these cases without sacrifying the pulmonary valve to be used as aortic valve substitute (for instance in adolescents and young adults who do not want to undergo a ‘two-valve’ procedure like the Ross–Konno procedure). In order to obtain the best possible long-term durability, the selected homograft should be one with the donor and recipient ages within 10 years of one another and if possible, from the same blood group. The descriptive report on the homograft from the tissue bank should be available (length, diameter, quality of the valve, preserved anterior leaflet of the mitral valve) before implantation.
The case presented in the video sequences is that of a 16-year-old Marfan patient who presented with a moderate to severe aortic insufficiency and a dilatation of the aortic root of  5.5 cm. End-diastolic dimension of the left ventricle was 6.5 cm but the left ventricular ejection fraction was normal. The operation was performed in 1996. Because of the annulus size slightly >32 mm and multiple fenestrations of the aortic valve close to the commissures, a valve-sparing root repair was not considered. An aortic homograft from the European Homograft Bank, size 27 mm was implanted. The intraoperative course was uneventful with a perfusion time of 65 min and a cross-clamp time of 51 min. Postoperative recovery was uneventful. In 2007, the aortic valve of the homograft had to be replaced because of combined aortic stenosis and regurgitation. A 25-mm mechanical valve was implanted. In 2008, the patient underwent replacement of the thoraco-abdominal aorta because of dilatation secondary to type B dissection. After insertion of the usual monitoring devices including a transoesophageal echocardiography probe, either a median sternotomy or a limited access mini-sternotomy is performed. Cardiopulmonary bypass is established with a single cannula in the ascending aorta and a two-stage cannula through the right atrial appendage. The ascending aorta is cannulated as high as possible (close to the origin of the innominate artery) for arterial return to allow sufficient exposure of the ascending aorta when the latter should be replaced (Video 1). Alternative cannulation site (e.g. right subclavian artery) is performed only in case of large aortic aneurysm extending into the aortic arch, aortic dissection or complex ascending aortic/aortic root redo-procedure with expected close relationship between the ascending aorta and the sternum. Venting of the left heart cavities is achieved via the right superior pulmonary vein or, alternatively, through the main pulmonary artery.
Cardiopulmonary bypass is conducted in moderate hypothermia (32–34 °C) and cardioplegia is administered through an aortic root cannula or directly into the coronary ostia in case of significant aortic regurgitation (Video 2). We usually start with 100 ml of crystalloid cardioplegia (Cardioplexol®, Bichsel Laboratory, Interlaken, Switzerland) and repeat every 30 min with cold blood cardioplegia either antegrade or retrograde.
Aortotomy is performed in a conventional way and stay sutures are placed on the edges of the aortic incision or at the top of the commissures of the native aortic valve. Before starting with the homograft implantation, the surgeon is required to analyze the morphology of the host aortic root carefully, especially the geometry of the native aortic valve, the quality of the annulus and the relationships of the commissures to each other and to the coronary arteries. It is well known that the relationship between annulus diameter and size of the sinotubular junction changes with increasing age. In younger patients, the annulus is generally larger than the sinotubular junction but in adults, the sinotubular junction is usually larger. If there is any asymmetry of the host aortic root with abnormally placed coronary ostia, then this needs to be taken into account during insertion of the homograft valve. The aortic valve is excised and if necessary the aortic annulus decalcified. The aortic annulus is usually sized preoperatively to select the appropriate homograft. The homograft represents in general a normal aortic valve and root from a younger healthy individual. The homograft is a 3-dimensional structure looking more or less like a symmetric cylinder. The distance between the annulus and the coronary ostia of the homograft is a further important point to analyze so that subsequent implantation of the coronary buttons can be performed without overstretching, kinking or any other unfavourable angulations. Distortion of the homograft during construction of the proximal suture line should be avoided because this would alter leaflet coaptation and lead to insufficiency of the homograft valve. Finally, the diameter of the homograft and the native ascending aorta at the level of the distal anastomosis should be carefully assessed. Immediately after thawing, the homograft is inspected carefully to ensure adequate valve quality and morphology, absence of injuries due to procurement or the thawing process itself. Usually, samples of the freezing solution and of the washing fluids as well as the remnant tissue from final trimming are sent for bacteriological examinations. Trimming of the homograft is an important step. At the proximal (cardiac) end of the homograft, a 5-mm cuff of myocardium with the corresponding endo- and epicardial layers is left and if required for closure of endocarditis abscess cavities, the anterior leaflet of the mitral valve are retained (Video 3). The coronary arteries of the homograft are cut at the level of the adventitial layer of the aorta and either oversewn with 6/0 polypropylene running sutures or enlarged if the coronary ostium of the homograft corresponds to the site of the native coronary button to be reimplanted.
The distal (aortic) end of the homograft is left intact when replacement of the ascending aorta is required. If there is insufficient length of the homograft, extension with a prosthetic graft is easily possible.
Cylindrical aortic root replacement The aortic homograft may be used to replace the aortic root completely when gross deformity is caused by infection, congenital anomaly, or it may be used to enlarge the root. Many surgeons use this technique routinely because aortic valve competence is virtually assured due to retention of the valve relationships within the intact aortic root of the homograft. If the native annulus is >30 mm, some reduction annuloplasty may be mandatory to ensure optimal matching with the annulus size of the homograft. Reduction of the annulus size may be performed using a purse-string suture of polypropylene. Some surgeons add a flexible ring to stabilize the annulus size in the long-term. In case of small native aortic annulus and pronounced left ventricular hypertrophy (especially located asymmetrically in the muscular septum), myectomy of the left ventricular outflow tract may be necessary. Additional enlargement technique (Manougian or Cooley procedure or even Konno's septal enlargement plasty) are rarely necessary.
The ascending aorta is transected
The coronary buttons are thereafter prepared close to the aortic wall, leaving some aortic tissue which facilitates the construction of the anastomoses and improves haemostasis (Video 4). Kinking, overstretching and other malpositioning of the coronary buttons is avoided when the surgeon keeps in mind the distance that the host coronary ostium was from its own annulus and by reproducing this distance with the homograft. Distension of the homograft aortic root during cardioplegia instillation into the aortic homograft (which is distally occluded using a Crawford clamp) may be helpful to accurately decide the optimal site of the right coronary button before reimplantation.
Correct alignment of the homograft (site of the commissures of the homograft must correspond to the location of the commissures of the native valve) may be facilitated by appropriate placement of situation stitches. When the diameters of the host and homograft annuli correspond closely to each other and the native aortic valve was tricuspid, the implantation will be straightforward (Video 5).
Construction of the proximal suture line may be performed either using continuous 3/0 or 4/0 polypropylene or interrupted U-stitches of 2/0 Ethibond with or without Teflon felt pledgets (Video 6). Special attention should be paid to place the sutures through the endo-/epicardial tissue of the homograft since muscular stitches only may lead to weak attachment and subsequent bleeding which may be difficult to control. Good bites of both host annulus and homograft tissue allow stable fixation and sufficient haemostasis. Another important point is that the sutures through the native annulus should follow an imaginary horizontal line passing across the base of each commissure, rising anteriorly to avoid the conduction bundle. Constant tension on the suture line by the assisting surgeon ensures optimal haemostasis. Adequate tension on the suture line may be improved by pulling it with a nerve hook before tying the suture (Schematic 2).
The coronary button anastomoses are performed according to the standard technique used for radical replacement of the aortic root using a composite graft prosthesis. Sometimes, the host coronary buttons do not align ideally with the homograft coronary ostia. If such is the case, a new 5–6 mm hole is made using an aortic punch at the appropriate site or just by enlarging the coronary ostia of the homograft. If the ostia of the homograft cannot be used, they are closed with a 6/0 polypropylene suture (Schematic 3). Before construction of the right coronary anastomosis, the correct positioning may be verified. If sufficient length of the homograft ascending aorta is available, a cardioplegic infusion line is inserted through the distal end of the homograft and the latter is occluded with a Crawford clamp. At the same time, haemostasis of the left coronary button reimplantation site can be verified (Videos 7 and 8).
When the native aortic valve is truly bicuspid (with or without a third commissure), the coronary ostia are 180° opposite one another and this makes the implantation of a tri-leaflet valve (homograft or stentless valve for instance) more difficult. In this situation, the coronary ostia cannot be transferred to the middle of the corresponding sinus of Valsalva of the homograft, but have to be reimplanted close to the commissure. Care must be taken not to damage the valve leaflet. Following adequate tailoring of the length of the homograft, the distal aortic anastomosis is performed in a running suture fashion using 4/0 or 5/0 polypropylene (Schematic 4). Tension on the anastomosis should be avoided and if there is any frailty of the native aortic tissue or the homograft, the suture line may be supported by a small strip of xeno-pericardium (Video 9). If the ascending aorta of the patient is dilated, replacement using a prosthetic graft is necessary; the distal anastomosis between the native aortic stump and the graft is performed first and thereafter, the anastomosis between the graft and the homograft. If the pathology extends into the proximal aortic arch, the distal anastomosis should be performed during a short period of moderate hypothermic (26–28 °C) circulatory arrest and selective antegrade cerebral perfusion.
Subcoronary implantation Because with this technique the homograft is to be sewn in ‘freehand’ using the native aorta for support, a clear understanding of the anatomical and spatial relationships of the aortic root is essential. It is critical that the aortic root of the host is symmetrical and that the annulus diameter is very similar to the sino-tubular diameter so that the pillar alignment is as symmetrical as possible. Important deformity of the aortic sinuses should be appreciated and corrected or the procedure abandoned in favour of conventional aortic valve replacement or aortic root replacement techniques. The homograft valve is then prepared by trimming the muscle and anterior leaflet of the mitral valve resected with a finger placed inside the aorta to stabilize the homograft, gauge the thickness of the trimmed graft. Excess aortic tissue is trimmed from the valve cusps, leaving a 3–4 mm rim of aorta beyond attachment of the cusps (Schematic 5). The graft is implanted in anatomical position. Three stitches are used to attach it to the outflow tract. The first suture is placed through the host aortic outflow tract below the medial commissure between the right and left coronary sinuses and through the septal myocardium below the corresponding commissure in the homograft. The stitch is placed below the annulus and the host aortic valve. The other two stitches are simply stay sutures placed to assist in aligning the allograft to the aortic root: they are placed beneath the appropriate commissure of the homograft and directly below the anterior and posterior commissures of the host aorta (Schematic 6).
The homograft valve is lowered down into position in the native aortic root. Commissures of the homograft are inverted through its annulus into the left ventricle of the host so as to expose the subvalvular edge of the homograft. A knot is placed in the primary suture and the stay sutures are tightened to align the homograft with the LV outflow tract. Stitches are placed between the graft and the LV outflow tract at or below the level of the annulus, attempting to make a level suture line. Because the aortic annulus is not circular but crescent shaped, the stitches are well below the fibrous annulus in the subcommissural region. It is advantageous to start with the suture within the left coronary sinus. The commissures of the aortic homograft are pulled out of the left ventricle so that the homograft assumes its normal position and configuration. The commissures of the homograft are attached to the aorta of the host using continuous polypropylene material (Schematic 7). Separate sutures are used for each aortic sinus. The stitches are passed deep in the aortic sinus slightly above the aortic annulus and then passed through the graft. Suturing proceeds in each aortic sinus until the graft is completely attached. Then the aortotomy is closed in a usual fashion.
Terminating the procedure De-airing of the aorta and left heart cavities is performed as usual following a warm blood reperfusion cardioplegia (Video 10). Epicardial electrodes are placed on the right atrium and ventricle and weaning from cardiopulmonary bypass is performed under AAI pacing rate of 90/min. Haemostasis may be supported by fibrin sealant or any other biologic glue. Optimizing coagulation helps for a sufficient haemostasis but the latter is generally excellent after homograft aortic root replacement. Drainage of the operative site and closure of the chest are performed in a standard fashion.
During the last 10 years we have used aortic homograft in the cylindrical configuration only in a small number of patients (18 patients). Main indication was aortic root destruction because of native (n=4) and prosthetic (n=13) endocarditis and in one adolescent patient who strictly refused a David procedure in case of annuloaortic ectasia in Marfan syndrome. In our institution, aortic homograft has been largely replacement by stentless xenograft valves and xenograft valved conduits which are available from the industry in all sizes. Mortality was 0% and no significant complications appeared in this small group of patients. Results from centres that have significant experience with homograft valve surgery report equivalent survival data. The University of Alabama had an 87% survival at 5 years in a 10-year period from 1981 to 1991 [13]. Of those who underwent isolated aortic valve replacement with a homograft, there was 99% survival at 30 days and 94% survival at 8 years. In the Mayo Clinic series, 82% were alive at 8 years [14]. A more recent and the most complete follow-up published of 1022 patients by O'Brien over a 29-year period showed an actuarial late survival at 25 years of 19% [15]. This drop was not related to cardiac or valve reasons. The total experience of this group accounts for patients operated between 1969 and 1994 and includes homografts used in the very early experience with now obsolete techniques of preservation and storage. Table 1 summarizes the results of the most important series reported in the literature.
Like all tissue valves, homograft aortic valves may fail. An understanding of the mechanisms of homograft valve failure and the way in which these mechanisms interact has important surgical implications. Homograft aortic valves may develop progressive regurgitation as a result of a change in the mechanical properties of the leaflets over time. The loss of leaflet radial extensibility is more pronounced than the normal changes of native leaflet tissue due to aging. This results in a progressive decrease in the area of leaflet coaptation. Homograft valves may fail because of leaflet degeneration characterized by thinning, tearing and perforation but also leaflet calcification, all of which may in part be due to an immunological response. Homograft valves may also fail due to geometric distortion related either to the insertion technique used or other factors such as aortic root geometry, technical factors and surgical experience. In the Brisbane series, reoperation was primarily for valve leaflet degeneration, endocarditis and technical reasons [16]. Progressive dilatation of the host aortic root may also result in regurgitation of the homograft aortic valve. When thinking about homograft valve failure, it is important to understand that these mechanisms may be influenced by a number of known risk factors (Schematic 8).
The argument for the use of homografts in the presence of active valvular infection is the lack of synthetic or prosthetic material left at the site of infection. When replacing an infected native valve, the incidence of subsequent prosthetic valve endocarditis is relatively low, with 95–97% of patients free from this complication at 5 years [17], there is a high early incidence being highest 6 weeks after surgery. This then regresses to a low constant risk after 9 months. These results have been confirmed by other centres [18, 19]. Homografts in contrary have a low constant risk [20].
The evolution of surgical techniques for homograft aortic valve implantation to deal with all types of aortic valve and aortic root pathology is interesting to examine the background of the mechanisms of homograft valve failure. Failure of homograft aortic valves inserted in the subcoronary technique is probably due in part to the geometric distortion of the pillars during insertion. The pillars containing the commissures of the valve must be aligned just as they were in the donor (although the homograft may have some natural asymmetry), and this perfect alignment may not be possible due to distortion of the host aortic root, surgical inexperience and technical factors, for example, allowing the pillars on each side of the aortotomy to approximate each other during aortotomy closure. For this reason, some surgeons use a pericardial patch to be inserted at the lower end of the aortotomy to maintain the geometric relationship of the two commissures on each side of the aortotomy. This problem may become even more important when the homograft valve is inserted into an aortic root of increasing size. Realizing that proper alignment of the commissures of the homograft was critical to homograft valve competence stimulated an increasing use of the cylindrical (or miniroot) technique in which the homograft is inserted in the aortic root (or in place of the aortic root) as a unit. Despite all variables involved in homograft valve failure, there is some evidence that the long-term competence of the homograft valve inserted by the cylindrical technique is superior to that of the subcoronary technique. Reoperation following homograft implantation is primarily for valve leaflet degeneration, endocarditis and technical reasons. The results published in the literature are significantly different with the initial valves that were stored at 4 °C and non-viable (58% at 15 years) and cryopreserved valves (78% at 15 years). In the latest follow-up of O'Brien, freedom from re-operation from all causes was 50% at 20 years and was independent of valve preservation. Durability is still controversial and in the Alabama series, overall freedom from reoperation was 92% at 5 years and 77% at 8 years [21]. These results have been confirmed by other centres [22,23,24,25,26,27,28,29]. However, it is important to note that freedom from reoperation does not accurately mean freedom from structural deterioration, as there are patients who will develop aortic incompetence but remain asymptomatic, or failure that does not require reoperation. There is no doubt from data in the early experience with homografts that non-viability of the homograft affects long-term durability. Donor age affects the pliability of the leaflet tissue and influences most probably its ability to withstand the constant stresses placed on them. Prolonged warm ischaemia, delayed sterilization and subsequent time to cryopreservation may also affect homograft viability. Finally, the immune response to what is in reality a transplanted tissue, is still a debated topic; but the implantation of viable homograft has been demonstrated to evoke both humoral and cellular responses from the host [30, 31]. More recently, some technologies have been developed which rival with aortic homograft valves [32]. The stentless biological valve (either from bovine pericardium or porcine valve tissue) has been available for more than 15 years [33, 34]. The concept of stentless biological prosthesis has been developed mainly to overcome the limited lifespan predominantly due to calcification of stented prostheses which was thought to be due to the presence of a stent and a sewing ring which both reduce the effective orifice area resulting in suboptimal haemodynamics, especially in small annulus. The stentless prosthesis should allow for a flexible prosthesis that should provide better haemodynamics and better durability due to decreased stiffness of the ‘integrated biological’ commissures which are fixed to the aortic wall. In fact, controversial results have been demonstrated in terms of haemodynamic and long-term durability between stented and stentless valves by repetitive prospective randomized trials so far [35, 36]. Tissue engineering is another technology which has emerged years ago with the aim to create the ideal valve. Two approaches are currently followed: (1) a heart valve is decellularized and autologous cells seeded on this biological scaffold and (2) seeding of cells on to a synthetically created biodegradable scaffold which takes the shape of native valve leaflets [37, 38]. It remains to be seen if this new technology will find its place in the armamentarium of the cardiac surgeon soon or not.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | SEARCH |
Top
Summary
Introduction and history
