Stentless Aortic Valve Replacement: Autograft/Homograft

	INTRODUCTION

Top Next

	HISTORICAL PERSPECTIVE

Top Previous Next

Since these initial efforts to utilize homografts in the aortic position, the procurement and preservation of these biologic valves have changed significantly. Initially, aortic valves were implanted shortly after collection.⁹ This technique fell out of favor and was rapidly supplanted by techniques to sterilize and preserve the valve for later use—a fundamental strategy of contemporary tissue banking. Valves were collected cleanly or sterilely and then were sterilized with beta-propiolactone^8,10 or 0.02% chlorhexidine,¹¹ followed by ethylene oxide¹¹ or radiation exposure.¹² After chemical sterilization, valves were placed in Hanks balanced salt solution at 4°C for up to 4 weeks, followed by freeze-drying.^8,13 Recognizing that the incidence of valve rupture was high in chemically treated valves, Barratt-Boyes introduced antibiotic sterilization of homografts in 1968.¹⁴ Cryopreservation of allografts was introduced in 1975 by O’Brien in an attempt to increase the cell viability of preserved allografts.¹⁵ Cryopreservation continues to be the most commonly used method for aortic allografts.

Recognizing that homografts may incite alloreactivity, it was suggested that autologous biologic valves would diminish this risk while maintaining the superior hemodynamic profile. In accord, use of the pulmonic valve to replace another valve was first reported in 1961 when Lower and colleagues of Stanford transposed the pulmonic valve to the mitral position in dogs.¹⁶ Shortly thereafter, Pillsbury and Shumway, also of Stanford, experimentally transposed the autologous pulmonic valve to replace a diseased aortic valve.¹⁷ Donald Ross extended this work to humans, reporting in 1967 a series of 14 patients in whom he replaced a diseased aortic valve (AV) with an autologous pulmonic valve.¹⁸ Since that time, this procedure has come to bear his name—the Ross procedure—also described as a pulmonary autograft. Widespread early fervor about this procedure was quickly tempered when surgeons appreciated its technical demands as well as considering it to be a "double valve" replacement for a "single valve" problem. While there was a surge of renewed interest in the Ross procedure in the 1990s, this has diminished in recent years. Currently, over 240 surgeons worldwide have performed the Ross procedure, as reported in The Ross Procedure International Registry, which was established in the early 1990s to catalogue these procedures and follow the outcomes.

	ALLOGRAFTS

Top Previous Next

 
Procurement and Preservation

In the United States, the majority of allograft valves are obtained from heart-beating organ donors whose hearts are not suitable for transplantation. Allografts obtained from fresh cadavers less than 24 hours old comprise the second main source of valves. The processing of cadaveric tissue is increasingly performed by regional tissue centers that specialize in the procurement and preservation of human tissues for ultimate allotransplantation. The increasing prevalence of cardiac allotransplantation has allowed the use of fresh "homovital" allografts,¹⁹ which may have enhanced preservation of cellular viability. Ideally, fresh valves would be implanted within 24 hours, but even at centers with significant experience, the interval between harvesting and implantation is up to 60 days, with an average interval of 3.9 days.¹⁹ Accordingly, cryopreservation is the most common technique used, which optimizes cellular viability¹⁵ and prolongs shelf life, which is an obvious advantage given the shortage of organ donors.

For cryopreservation, the heart is procured under sterile (multi-organ donor) or clean (cadaveric donor) conditions and gently rinsed with cold isotonic salt solution (e.g., Ringer lactate) to remove the blood and its elements from the cardiac chambers. The heart is then placed in a bag containing iceslush solution and kept cold until further processing. Warm ischemia time does not exceed 12 hours, unless the donor is in an environment with temperature 8°C within 6 hours of death, which can extend the "warm" ischemia time to 24 hours. Donor blood is obtained for culture and serologic testing for common infectious agents (e.g., hepatitis B and C, human immunodeficiency virus, human T-cell lymphoma virus, and Treponema pallidum). The following details are based on procedural protocols of The Northwest Tissue Center, Seattle, WA,²⁰ and are similar to those of other institutions.²¹ Once at the tissue center, an aortic block is dissected in a controlled environment (a class 100 clean room environment). Donor tissue and the transport solution are cultured for aerobic and anaerobic organisms, fungi, and acid-fast bacilli. Proximally the dissection includes the aortic ring and the anterior leaflet of the mitral valve with a variable amount of ventricular muscle, and extends distally to the left subclavian artery, including the branches of the thoracic aorta. The coronary ostia are ligated, allowing a subsequent interposition "free-root" allograft if needed. The base of the graft contains a variable amount of ventricular muscle that may be trimmed at the time of graft implantation. The valve is inspected for leaflet fenestrations, atheroma, or damage that may make it unsuitable for implantation. Of these, leaflet atheroma absolutely contraindicates use of the allograft, while fenestrations or other damage are relative contraindications, and depend on their severity and magnitude. Obturators are then used to size both the valve and the aorta (Fig. 35-1).

View larger version (110K): [in this window] [in a new window]   Figure 35-1 Aortic valve allograft after harvesting from the donor. The block includes a variable amount of ventricular muscle and the anterior leaflet of the mitral valve. Additional trimming for replacement is performed at the time of implantation. (Reproduced with permission from The North-west Tissue Center, Puget Sound Blood Center.)

The sine qua non of cryopreservation is controlled rate freezing, as introduced by O’Brien and colleagues in 1975.¹⁵ Prior to packaging, the air in the packaging room is sampled for viable particles, as are the gloves of the technologist, and three designated locations in the sterile packaging field. The allograft is transferred from the antibiotics to a sterile storage pouch containing culture medium (e.g., RPMI-1640), 10% fetal calf serum, and 7.5% dimethyl sulfoxide DMSO (a cryoprotectant). These substrates are designed to provide nutritive support to the allograft and minimize crystal formation and tissue damage during the freezing process. A final sample is taken of the packaging solution for aerobic and anaerobic organisms. Within 2 hours of exposure to DMSO, the allograft is frozen at –1°C per minute down to –40°C and then placed in vapor-phase liquid nitrogen storage (about –195°C) until it is used. After tissue cultures and serology results are available (about 4 to 6 weeks) and negative, the valve may be used for implantation; if there are any positive culture or serology results, the valve is discarded.

After release of the allograft from the tissue center, it is shipped in a container validated to maintain temperatures below –100°C for 10 days. Two temperature-sensitive indicators, which turn red if temperatures exceed –100°C, are included with each shipment to ensure maintenance of shipping temperatures. Upon arrival at the institution where it will be used, the storage pouch is removed from the liquid nitrogen and placed in warm saline (37° to 42°). The allograft size and number are confirmed. After removal from the storage pouch, the allograft then undergoes a series of gentle rinses and thawing, in solutions that have increasingly dilute DMSO, followed by a final rinse in pure nutrient media prior to use. The allograft is then ready for final trimming and implantation.

Cellular and Immunologic Aspects of Allografts

Normal valves are ultrastructurally comprised of viable cellular components, including endothelium, fibroblasts, and smooth muscle cells, and an amorphous and fibrillar extracellular matrix, derived primarily from fibroblasts and smooth muscle cells.²⁴ In nonpathologic states, there is a steady state between destruction of these elements and remodeling, which underlies an overall structure and function. In accordance with the paradigm that structural integrity, and thus function, depends on cellular viability of an allograft, preservation techniques have attempted to optimize the preservation of viable cellular elements to improve function and durability.

As mentioned, earlier methods of chemical sterilization and irradiation had a prohibitive incidence of cusp rupture¹⁴ and histologic analysis of these allograft valves reveals nonviable cellular elements.²⁵ Antibiotic sterilization of allografts and storage in a balanced salt solution or nutrient medium at 4°C does not maintain cellular viability beyond a few days.^26,27 The durability of these valves is improved over chemically sterilized valves with a freedom from reoperation for valve degeneration at 10 years of 89%.¹⁵ Gentle procurement and cryopreservation of allografts has been shown to maintain donor fibroblast viability up to 9¹/₂ years after implantation in one patient, and viable fibroblasts have been consistently demonstrated in a small number of other patients.¹⁵ More recently, persistence of viable, functional donor fibroblasts in allografts harvested up to 70 months earlier has been reported.²⁹ Also, using in situ hybridization technology, the viable fibroblasts in the explanted allografts were demonstrated to be of both recipient and donor origin. Other investigators have demonstrated no cellular viability of allograft valves after antibiotic treatment²⁶ or after explantation.²⁸ Of note, however, the methods of valvular preservation in the latter investigation²⁸ were not reported, and these valves were explanted primarily for deterioration, contributing to significant selection bias. In the end, the extent to which viable cellular elements persist in allografts after cryopreservation is not clear. Given the occasional findings of viable cells, particularly fibroblasts, up to 9 years after implantation,¹⁵ it is likely that some viable cells persist in allografts, at least some of the time. Moreover, when viable cells are present, they are likely to be of both donor and recipient origin.^15,29 The discrepant findings with respect to the persistence of viable cells in the allograft may relate to warm ischemia times or differences in procurement and preservation techniques. More importantly, when present, the extent to which these viable fibroblasts remain functional and contribute to the structural integrity of the allograft components is not known.

While modifications of valve preservation have attempted to maintain cellular viability of the allograft¹⁵ to improve long-term structural integrity and function,³⁰ the presence of viable donor cells may be detrimental by inciting an allograft rejection reaction.³¹ Multiple studies have demonstrated the generation of donor-specific alloantibodies directed against human leukocyte antigens (HLA) class I (A and B antigens) and II (DR antigens) in allograft valve recipients.^32,33 The development of panel reactive antibodies seems to increase with time,³³ approximating 82% at 6 years postgraft implantation,³⁴ occurs in both adults and children, and is irrespective of the method of cryopreservation.³⁵ Despite consistent evidence of antibody formation directed against the HLA antigens of the allograft, the clinical significance is not clear.

Dignan and associates reported a significant association between HLA class II antigen mismatch and postoperative fever and homograft dysfunction in recipients of cryopreserved allografts.³⁶ In recipients of homovital allografts, Smith and colleagues reported no significant association between HLA class I or II antigens and long-term (6 years) valve function,³⁴ although there was an increased prevalence of valve degeneration in those patients with HLA antibodies: in HLA antibody–negative patients the actuarial freedom from valve degeneration at 1, 5, and 10 years was 100%. In patients with panel reactive antibodies <50% freedom from valve degeneration at 1, 5, and 10 years was 100, 97, and 92%, and 98, 94, and 88% in those patients who were highly sensitized.³⁴ Taken together, these data suggest that antibody-mediated alloreactivity may play a causal role in allograft valve dysfunction over time, but larger studies are certainly needed with adequate power to detect small (i.e., 5% at 10 years) differences between allograft recipients who are HLA matched and mismatched. In the meantime, until such data are available, it has been suggested that prospective matching of HLA antigens may be warranted.³⁴ Recognizing that investigation of allografts in humans is limited to preimplanation analysis (i.e., during procurement or preservation) or valves that require explantation due to valve failure, heart transplantation of the recipient, or recipient death, investigators have developed alternative models for this inquiry.

Accordingly, to better understand the immunologic aspects of allograft valve dysfunction, numerous investigators have used animal models of allograft implantation. Multiple studies of allograft implantation across major histocompatibility complex barriers in rats demonstrate significant cellular infiltration into thickened valve leaflets over the first 28 days, which is temporally followed by valve degeneration and failure.^37–40 The cellular infiltration may be phasic, characterized by early monocyte infiltration,³⁷ followed by progressive monocyte/macrophage and T-lymphocyte infiltration, which is maximal by 7 days postimplantation.^37,40 There is a coincidental decline in allograft donor cell viability over this time that is paralleled by declining valve structure and function.³⁷ Importantly, in T-cell–deficient rats, these cellular events do not occur and allograft function is preserved, providing additional support for immune-mediated valvular destruction.³⁸ Moreover, immune modulation with cyclosporine and antiadhesion molecule (anti-4/β2 integrin) therapy attenuates leaflet cellular infiltration and prevents allograft structural failure.⁴¹ Taken together, these data strongly support a donor-specific, cell-mediated (primarily T lymphocytes) immune reaction directed against the donor alloantigens that is followed by structural valve failure—a cellular cascade typical of solid organ rejection.

In summary, despite intensive investigation over the past three decades, the relative contribution of the immune response, preservation techniques, and warm ischemia time to ultimate valve degeneration (i.e., sclerosis or calcification) is not clear. More importantly, after consideration of the structural benefits and the immune-reaction risks, the net advantage of maintaining cellular (particularly fibroblasts) viability in the allograft is not well defined.³¹ Future investigations should endeavor to further clarify the relative contributions of these factors to allograft valve antigenicity, immunogenicity, and durability. Furthermore, the impact of immune modulation on long-term allograft function warrants further study.

Indications

Aortic valve replacement with an allograft has a number of advantages including excellent hemodynamic profile with low transvalvular gradients and possibly enhanced regression of left ventricular mass (LVM),⁴² low risk of thromboembolism without the need for systemic anticoagulation, and low risk of prosthetic valve infection. Allograft durability is limited, however, with a freedom from reoperation at 20 years of 38 to 50% and a freedom from structural valve failure at 20 years of 18 to 32%.^43,44 The incidence of structural failure is dependent on the ages of the recipient and the donor. Allograft failure increases with decreasing recipient age, and durability is improved in older recipients. Also, allograft failure increases as the donor age increases. Considering these data underscoring poor allograft durability, we believe the current indications for an aortic allograft are limited. The primary indication in adults is for treatment of active AV endocarditis. However, we are not aware of any data indicating the superiority of allografts compared to other valve options in this setting, and the ideal valve choice for replacement of an infected AV has not been defined. Aortic allografts can also be considered for patients requiring composite valve/root aortic replacement who cannot be anticoagulated.

Preoperative Evaluation

Preoperative preparation for placement of an aortic allograft is similar to that for other AV operations. Transthoracic echocardiography (TTE) is an invaluable diagnostic tool for evaluation of the AV and associated anatomic structures, including the leaflets, the annulus, the sinuses, the sinotubular junction, the subvalvular left ventricular outflow tract (LVOT), and the ascending aorta. In this regard, the preoperative TTE can accurately predict aortic annulus diameter within a few millimeters, and thus the size of the homograft required.^45–47 Transesophageal echocardiography (TEE) should routinely be used for intraoperative confirmation of the anatomy and to assist with sizing of the homograft, and to assess postrepair function. The accuracy of TEE in assessing annular dimensions has been demonstrated.^48,49 While TTE and TEE can accurately approximate the annular dimensions and homograft size, direct surgical measurement is imperative.

Operative Technique

General preparation

A full median sternotomy is used with standard techniques of cardiopulmonary bypass (CPB). Aortic cannulation is obtained as far distally as possible, near the innominate artery, and a two-stage cannula is placed into the right atrial appendage. A ventricular vent can be placed into the right superior pulmonary vein. Both antegrade and retrograde blood cardioplegia are delivered. The aortotomy incision can be transverse or an obliquely oriented reverse "lazy-S" that begins above the right coronary ostia (Fig. 35-2). After retraction of the aorta, facilitated with silk stay sutures, the valve is excised. The aortic root and annulus are then closely examined for geometric morphology and symmetry, indicating feasibility of a good result with a homograft. Next, the aortic ring is sized with standardized obturators or Hagar dilators. Since this represents an external diameter of the recipient, and allografts are sized based on internal diameters, an allograft 2 to 4 mm smaller than the recipient measurement is obtained. The appropriately sized allograft can then be thawed. If the allograft was selected based on accurate TTE or TEE measurements, it should already have been thawed. The allograft is then trimmed.

View larger version (43K): [in this window] [in a new window]   Figure 35-2 Standard cardiopulmonary bypass is performed. A "lazy-S" or oblique aortotomy is performed beginning above the right coronary cusp with the transverse portion just cephalad to the noncoronary cusp. This incision provides excellent exposure for all techniques when replacing the aortic valve and root. (Reproduced with permission from Schaff HV, Cable DG: Aortic valve replacement with homograft, in Kaiser LR, Kron IL, Spray TL (eds): Mastery of Cardiothoracic Surgery. Philadelphia, Lippincott-Raven, 1998.)

View larger version (31K): [in this window] [in a new window]   Figure 35-3 Preparation of an aortic homograft from the aortic block. The ventricular muscle and mitral leaflet are removed. (A) The freehand subcoronary insertion technique involves removal of sinus aorta within 5 mm of the cusp attachments down to within 3 mm of the cusp bases, thus removing all three sinuses. (B) The intact noncoronary sinus technique involves preservation of the noncoronary sinus. (C) For aortic root replacement, the entire aortic wall can be retained, and then be implanted as an "inclusion cylinder" or "mini-root" insertion. (Reproduced with permission from Schaff HV, Cable DG: Aortic valve replacement with homograft, in Kaiser LR, Kron IL, Spray TL (eds): Mastery of Cardiothoracic Surgery. Philadelphia, Lippincott-Raven, 1998.)

There are multiple techniques for placing the allograft in the aortic position, and the strategies have continued to evolve over time. Ross⁷ and Barratt-Boyes⁵⁰ originally described the 120° rotation scalloped freehand technique. Numerous groups later modified this technique, whereby the right and left coronary sinuses were scalloped, but the noncoronary sinus was left intact—the intact noncoronary sinus technique. Later this technique was further altered to preserve all the sinuses on the donor and insert the allograft as a cylinder within the recipient aortic root, with reimplantation of the coronary ostia as needed. More recently, the allograft has been implanted as a mini-root interposition graft when the severely diseased aortic root must be excised. Thus, there are essentially four techniques for placing an aortic allograft: (1) 120° rotation scalloped implant, (2) intact noncoronary sinus scalloped technique, (3) aortic root inclusion cylinder technique, with reimplantation of the coronary ostia as needed, and (4) aortic mini-root replacement with interposition allograft.

SCALLOPED 120° ROTATION FREEHAND TECHNIQUE: For the scalloped 120° freehand rotation technique, the sinus aorta is trimmed within 5 mm of the cusp attachments down to within 3 mm of the cusp bases, effectively removing all three sinuses.⁵⁰ The valve is then rotated 120° in the counterclockwise direction, so that the donor right sinus lies below the recipient left coronary sinus (Fig. 35-4). This critical maneuver brings the weaker muscular portion of the allograft posteriorly adjacent to the fibrous trigone and anterior leaflet of the mitral valve.^50,51 Two suture lines are required for this technique. The lower suture line can either be a continuous running suture or simple interrupted sutures can be used, usually of 4-0 or 5-0 polypropylene (Figs. 35-5 through 35-7). For the running suture, the homograft can be turned inside out to facilitate the lower suture line. In the area of the membranous septum, sutures are placed more superficially to avoid the conduction system.

View larger version (20K): [in this window] [in a new window]   Figure 35-4 For the scalloped 120° rotation freehand technique, the allograft is rotated 120° counterclockwise so that the donor right sinus lies below the recipient left coronary sinus. Three orientation stay sutures are placed below the nadir of each cusp.

View larger version (50K): [in this window] [in a new window]   Figure 35-5 The inferior suture line with simple interrupted sutures. (Reproduced with permission from Schaff HV, Cable DG: Aortic valve replacement with homograft, in Kaiser LR, Kron IL, Spray TL (eds): Mastery of Cardiothoracic Surgery. Philadelphia, Lippincott-Raven, 1998.)

View larger version (47K): [in this window] [in a new window]   Figure 35-7 The downstream continuous suture line. (A) The allograft is everted after completion of the inferior suture line, by applying traction to the commissural posts. (B) A continuous suture completes the implantation. (C) Completed suture line. (Reproduced with permission from Schaff HV, Cable DG: Aortic valve replacement with homograft, in Kaiser LR, Kron IL, Spray TL (eds): Mastery of Cardiothoracic Surgery. Philadelphia, Lippincott-Raven, 1998.)

View larger version (56K): [in this window] [in a new window]   Figure 35-6 The inferior suture line with continuous suture. (A) The double-arm orientation stay sutures facilitate placement.(B) The allograft is inverted into the ventricle and the stay sutures are tied. (C) The stay sutures are then "run" in a clockwise manner to complete the lower suture line. (Reproduced with permission from Schaff HV, Cable DG: Aortic valve replacement with homograft, in Kaiser LR, Kron IL, Spray TL (eds): Mastery of Cardiothoracic Surgery. Philadelphia, Lippincott-Raven, 1998.)

FREEHAND INTACT NONCORONARY SINUS TECHNIQUE: Scalloping of the right and left coronary sinuses while preserving the noncoronary sinus is a technical extension of the scalloped 120° subcoronary implant technique (Fig. 35-8). This modification increases stability of the homograft and maintains symmetry more easily. Furthermore, the risk for noncoronary cusp prolapse is reduced in patients with a dilated or abnormal sinotubular junction.⁵⁶ Accordingly, it is a reasonable choice for patients with mildly dilated or asymmetric aortic roots and those with AI. For the intact noncoronary sinus technique, the allograft is prepared as above, except the noncoronary sinus is preserved. The allograft is then inserted into the aortic root maintaining anatomic alignment without any rotation and is sutured as described above. Additionally, mattress sutures are placed through the native aorta and the noncoronary sinus of the allograft to obliterate the space between the noncoronary sinus and the native aorta.

View larger version (30K): [in this window] [in a new window]   Figure 35-8 The intact noncoronary sinus technique. (A) This technique is nearly identical to the scalloped freehand technique, except the anatomic alignment of the donor allograft is maintained. (B) The noncoronary sinus is sutured to the corresponding aortic wall after partial closure of the aortotomy. This ensures minimal tension on this suture line. (C) The space behind the noncoronary sinus is obliterated with a U-stitch and the aortotomy is closed.

AORTIC ROOT REPLACEMENT WITH INCLUSION CYLINDER TECHNIQUE: Ross and others modified the aforementioned techniques to place the allograft as a sleeve or cylinder within the aortic root. The technique of implantation is very similar to that outlined above. For the inclusion cylinder technique, the sinuses are retained, and the cusp-sinus relationships are preserved (Fig. 35-9). Depending on the length of the cylinder, the recipient coronary ostia may need reimplantation into a buttonhole in the side of the allograft or directly into the donor coronary ostia. Conversely, if the distal aspect of the allograft is caudal to, or near the recipient coronary ostia, then minimal scalloping of the allograft wall will suffice to ensure coronary flow. We create the proximal suture line with everting, pledgeted horizontal mattress 2-0 Ticron sutures, and use a running 4-0 polypropylene suture for the distal suture line. The coronary buttons are reimplanted with a running 5-0 polypropylene suture.

View larger version (34K): [in this window] [in a new window]   Figure 35-9 The inclusion cylinder technique for aortic root replacement. The cylinder retains the aortic allograft anatomic relationship. Depending on the length of the cylinder, the recipient coronary ostia may require reimplantation.

Postreplacement assessment

Intraoperative TEE with Doppler color-flow measurement is the most valuable postreplacement assessment tool, after the patient has been separated from CPB. The accuracy of the TEE assessment can be further enhanced with volume loading and administration of phenylephrine to effect vasocon-striction. Moreover, the transvalvular gradient, orifice area, subvalvular structures, and presence of regurgitation may all be accurately determined with TEE. With appropriate loading conditions, moderate to severe AI warrants reinstitution of CPB with inspection and revision of the allograft as needed. Mild AI is usually tolerated well and does not warrant re-exploration.

Postoperative management

Postoperative management following allograft placement is similar to that of other aortic valve replacements, and is dictated by the antecedent physiology that resulted from the aortic stenosis (AS) or regurgitation that required surgery. Aortic stenosis produces a hypertrophied, noncompliant ventricle that requires adequate preload and maintenance of sinus rhythm for adequate cardiac output.

In contrast, AI results in a dilated left ventricle that may be coincidentally hypertrophied. Again, ensuring adequate preload and aggressively treating arrhythmias is imperative. Since patients with AI are chronically vasodilated to maintain systemic perfusion, vasoconstricting agents may be required after the patient is normothermic in the intensive care unit.

In both of these groups, systolic hypertension should be treated aggressively to protect the aortic suture line. Other than atrial arrhythmias, which are treated aggressively to restore atrial-ventricular synchrony, these patients are also susceptible to heart block, since the atrioventricular node and left bundle lie in the membranous septum underneath the right coronary annulus. When this occurs, epicardial pacing is employed as needed, and when persistent beyond a few days, a permanent pacer may be placed.

Long-term anticoagulation is not required and once-daily aspirin suffices.

Results

Perioperative complications

In patients without endocarditis at the time of allograft placement, operative mortality is 1 to 5%.^43,44,57 Notably, numerous experienced and talented groups have reported that the root replacement technique does not impact early mortality.^44,57 In contrast, patients with endocarditis at the time of allograft valve placement have a much higher early mortality, from 8 to 16%.^{19,44,59–61} In these patients, early mortality was higher in patients with cardiogenic shock,⁶¹ or prosthetic valve endocarditis (18.8%) compared to native valve endocarditis (10%).⁵⁹

Early postoperative AI occurs infrequently and most often results from technical factors, like inaccurate sizing of the allograft, or valve distortion during placement, particularly with the scalloped subcoronary implant technique. This complication should be appreciated intraoperatively with loading maneuvers and intraoperative TEE, as previously mentioned.

Hemorrhage, heart block, stroke, myocardial infarction, and infectious complications occur with similar frequency to those of other AVRs, and are not unique to homografts.

Long-term outcomes

As mentioned, 30-day mortality following homograft placement is less than 5% in patients without endocarditis. Crude survival at 10 and 20 years is reported to be 67 and 35%, respectively,⁴⁴ while in other studies actuarial survival at 10, 20, and 25 years is 81,⁴³ 58,⁴³ and 19%⁵⁷ (Table 35-2).

Importantly, when structural valve deterioration occurs, we have found reoperation and allograft valve/root excision to be technically quite demanding. Grossly, the allografts become severely calcified, which makes circumferential dissection very challenging, particularly at the coronary button ostia. While the coronary arteries themselves are usually normal-quality tissue, the proximal rim of the button is often calcified, making removal and reimplantation more challenging. Furthermore, because the base of the allograft also becomes heavily calcified, we have often found the annulus to be heavily scarred and calcified, and subsequently narrowed. Bearing in mind the concept of patient-prosthesis mismatch, and the goal of providing an effective orifice area index (>0.75 cm²/m²), we have had a low threshold for performing an annular enlargement procedure (both posterior and anterior). Taken together with the limited durability, relative to other currently available options with comparable versatility, we almost never use allografts for replacement of diseased AVs.

Hemodynamic characteristics of allografts are excellent at short- and medium-term follow-up, both at rest and during exercise.^62,63 However, progressive allograft dysfunction develops over time that coincides with deteriorating hemodynamic performance of a progressively abnormal valve.

Infectibility of allografts is low, with freedom from endocarditis at 10 years of 93 to 98%,^43,44 and at 20 years of 89 to 95%.^43,44,57 Similarly, freedom from thromboembolism at 15 and 20 years in patients undergoing AVR plus coronary artery bypass graft (CABG) is 92 and 83%, respectively.⁵⁷ O’Brien and colleagues found that neither preservation methods or implantation techniques affected overall 20-year rates of thromboembolism, endocarditis, or structural valve deterioration.⁵⁷

In patients with active endocarditis requiring AVR, results are much poorer. Operative mortality is nearly twice that of patients without endocarditis, from 8 to 17%,^59–61 and is higher in patients with prosthetic valve endocarditis.⁶⁰ Late survival ranges from 58% at 5 years⁵⁹ to 91% at 10 years,⁶¹ and is significantly lower in patients with prosthetic valve endocarditis (PVE).⁶⁰ Importantly, the risk of recurrent endocarditis is <4% up to 4 years postoperatively.^51,60,61 As a result of these acceptable outcomes in a high-risk group of patients, many consider allografts the preferred valve for aortic replacement in patients with active endocarditis.

Conclusions

Allograft replacement of an aortic valve has become less common with the increased availability of bioprosthetic valve alternatives. The aortic allograft is a versatile valve allowing placement as a valve or as an aortic root replacement, which is the most common insertion technique. Aortic allografts have many favorable attributes including low operative mortality rate, excellent early and midterm hemodynamics across the allograft, low valve-related noise, low thrombogenicity, and low infectibility. The primary, and significant, shortcoming of aortic allografts is progressive deterioration of valve structure and function over time, which limits its use in younger patients with long life expectancy. Limited availability and technical expertise for insertion are also limiting factors for more widespread utilization of aortic allografts. For patients with active endocarditis of either the native or prosthetic AV, allograft replacement is a reasonable option. Furthermore, with respect to biologic replacement of the aortic root, longer follow-up is needed for the currently available xenografts to determine the relative durabilities between aortic allograft root replacement and xenograft root replacement.

	PULMONARY AUTOGRAFT

Top Previous Next

 
Theoretical Considerations

Replacement of a diseased aortic valve with a pulmonary autograft has a number of advantages including: (1) freedom from thromboembolism without the need for anticoagulation; (2) improved hemodynamics through the valve orifice without obstruction or turbulence; (3) growth of the autograft with time, particularly beneficial for young patients who continue to grow after receiving the aortic autograft;⁶⁴ and (4) the assumption that replacement of the AV with living autologous tissue is preferential to prosthetic or xenogeneic materials.

Patient Selection

In addition to individual surgeon experience, a number of patient factors influence consideration of the Ross procedure for replacement of a diseased AV. Table 35-3 summarizes the important patient factors to bear in mind when considering the Ross procedure. The only absolute contraindications are significant pulmonary valve disease, congenitally abnormal pulmonary valves (e.g., bicuspid or quadricuspid), Marfan syndrome, unusual coronary artery anatomy, and probably severe coexisting autoimmune disease, particularly if it is the cause of the aortic valve disease. Of note, bacterial endocarditis is not a contraindication for the Ross procedure, though when present, it usually dictates that the root replacement technique be used. Additional minor considerations often come into play including patient age, associated medical conditions, physiologic reserve, suitability for anticoagulation, and underlying ventricular function, as the time on CPB is potentially long.

View larger version (28K): [in this window] [in a new window]   Figure 35-10 Ross procedures performed according to the Ross Registry 1988–2000. (Adapted from the Ross Procedure International Registry; http://www.rossregistry.com.)

Since the initial description by Ross of the scalloped sub-coronary implant, a number of modifications have been described, including the inclusion cylinder technique and the root replacement technique. These techniques are performed identically to those outlined above for aortic homografts. According to the Ross Procedure International Registry, the root replacement technique is the most commonly performed variation owing to its superior versatility and possible decreased incidence of early and late graft failure.⁶⁵

The conduct of the operation is similar to placement of an aortic homograft. Full CPB is utilized with arterial cannulation of the distal aorta. Again, the ventricular vent is placed through the right superior pulmonary vein. Cardioplegia is delivered antegrade and retrograde, via the coronary sinus catheter. A transverse or oblique reverse "lazy-S" aortotomy is performed similarly to placing an allograft. After retraction of the aorta, facilitated with stay sutures, the AV and root are inspected and the suitability of the pulmonary autograft for repair is confirmed. The AV is then excised, along with the coronary ostia as buttons. When the root replacement technique is used, the aortic root is excised as previously described for the allograft. Attention is then turned to the pulmonary artery (PA) and pulmonic valve.

The PA is mobilized to its bifurcation. The PA is then sharply divided transversely just proximal to the bifurcation (Fig. 35-11). This allows visual inspection of the pulmonic valve endoluminally, which is normally tricuspid, without fenestrations or atheroma (Fig. 35-12). The discovery of a bicuspid or quadricuspid valve, or the presence of large fenestrations or atheroma precludes use of the valve as an autograft. The mobilization of the PA then begins distally, near the transverse arteriotomy, and continues proximally toward the valve. The dissection is initiated posteriorly staying very close to the PA, taking care not to buttonhole the wall (Fig. 35-13). The left main coronary artery and its bifurcation into the left anterior descending (LAD) artery with the septal perforators, and circumflex arteries should be identified and avoided. When these are not easily appreciated, a probe may be placed into the left coronary os, through the aortotomy, to facilitate identification of the left main coronary artery and its branches. The dissection continues proximally until septal musculature is reached, taking care to avoid injury to the conal branch of the right coronary artery. A point 3 to 4 mm below the pulmonic annulus is identified, often facilitated by passing a probe through the pulmonic valve, and the right ventricular outflow tract (RVOT) is divided (Fig. 35-14). Where the RVOT meets the septum, the dissection must be kept superficial (i.e., on the right ventricular side) to avoid injury to the septal perforators of the LAD artery. The adventitia of the autograft is preserved. After complete division, the autograft is then accurately sized at its base, as well as the RVOT, and an appropriately sized allograft is obtained.

View larger version (61K): [in this window] [in a new window]   Figure 35-11 The distal pulmonary artery is incised at the origin of the right pulmonary artery. A transverse arteriotomy is made to allow careful inspection of the pulmonary artery. Shown from the surgeon’s perspective, as though standing on the patient’s right side. (Reproduced with permission from Elkins RC: Aortic valve: Ross procedure, in Kaiser LR, Kron IL, Spray TL (eds): Mastery of Cardiothoracic Surgery. Philadelphia, Lippincott-Raven, 1998.)

View larger version (71K): [in this window] [in a new window]   Figure 35-12 The normal trileaflet pulmonic valve with three equal sinuses and no fenestrations or other abnormalities. (Reproduced with permission from Schaff HV, Cable DG: Aortic valve replacement with homograft, in Kaiser LR, Kron IL, Spray TL (eds): Mastery of Cardiothoracic Surgery. Philadelphia, Lippincott-Raven, 1998.)

View larger version (100K): [in this window] [in a new window]   Figure 35-13 Dissection of the pulmonary autograft is initiated on the posterior aspect of the proximal pulmonary artery. Dissection is continued in this plane, adjacent to the pulmonary artery, until the septal myocardium is encountered. The left main coronary artery and left anterior descending artery are protected. (Reproduced with permission from Schaff HV, Cable DG: Aortic valve replacement with homograft, in Kaiser LR, Kron IL, Spray TL (eds): Mastery of Cardiothoracic Surgery. Philadelphia, Lippincott-Raven, 1998.)

View larger version (112K): [in this window] [in a new window]   Figure 35-14 Identification of the anterior right ventriculotomy is facilitated by placement of a right-angled clamp through the pulmonary valve and indenting the myocardium 3 to 4 mm below the pulmonary valve annulus. (Reproduced with permission from Schaff HV, Cable DG: Aortic valve replacement with homograft, in Kaiser LR, Kron IL, Spray TL (eds): Mastery of Cardiothoracic Surgery. Philadelphia, Lippincott-Raven, 1998.)

The autograft is oriented so that the posterior pulmonary sinus becomes the noncoronary sinus. The proximal suture line is performed first. In adults, this is performed with either everting pledgeted horizontal mattress 2-0 Ticron or a running 4-0 or 5-0 polypropylene. In children and adolescents, an absorbable monofilament suture (e.g., Maxon, Davis & Geck, Manati, Puerto Rico) is used, recognizing future growth of the autograft with the patient.⁶⁴ The left coronary artery is then implanted at the midpoint of the left coronary sinus, as previously described. Again, 5-0 polypropylene is used in adults and 6-0 Maxon is used in children. The distal suture line is then performed, after the autograft is trimmed 4 to 5 mm beyond the sinotubular junction of the autograft. A running 4-0 or 5-0 polypropylene suture is used in adults, while Maxon is used in children. The importance of similarly sized aortic and pulmonary valves has recently been suggested.⁶⁶ If a size or geometric mismatch between valves occurs, the diameter of the aortic annulus and/or sinotubular junction may be surgically reduced, with a reduction aortoplasty, as performed by Elkins.⁶⁷ Also, if the ascending aorta is aneurysmal, it may be replaced with an interposition Dacron graft bridging between the distal autograft and the ascending aorta just proximal to the innominate artery. This is followed by reimplantation of the right coronary artery, similar to that described for the left coronary artery. After the suture lines are inspected, the cross-clamp is removed and the operation is completed during rewarming.

A pulmonary homograft, oversized by 4 to 6 mm, is then obtained. The homograft is trimmed, and the proximal anastomosis is performed to the RVOT with running 4-0 polypropylene. Again, the left coronary artery and its branches lie close to the posterior aspect of the anastomosis, and sutures must be precisely placed in the endocardium to avoid kinking or injuring these structures. The distal suture line is then completed with a running 4-0 or 5-0 polypropylene suture. De-airing and weaning of CPB is performed in a standard fashion.

Inclusion cylinder technique

The inclusion cylinder technique is nearly identical to that previously described for the allograft (Figs. 35-15 and 35-16).

View larger version (67K): [in this window] [in a new window]   Figure 35-15 Inclusion cylinder technique. (A) Placement of three polypropylene sutures to orient the pulmonary autograft. The posterior sinus of the pulmonary autograft becomes the new left coronary sinus. (B) The autograft is inverted into the left ventricle and the proximal sutures are tied and divided. (C) The pulmonary autograft is reinverted. Horizontal mattress sutures are placed to secure the height and position of the autograft (but not tied until the right and left coronary arteries are implanted). An aortic punch (4 or 5 mm) is used to create an opening in the autograft to allow attachment of the coronary artery ostia. (D) Completion of coronary artery anastomosis. Commissural stay sutures are tied and divided, and the distal suture line is initiated at the commissure between the left and right coronary artery. This is continued to the aortotomy extension into the noncoronary sinus. This portion of the aortotomy is closed with a running suture line with the suture including a full-thickness bite of the noncoronary sinus of the pulmonary autograft. (Reproduced with permission from Schaff HV, Cable DG: Aortic valve replacement with homograft, in Kaiser LR, Kron IL, Spray TL (eds): Mastery of Cardiothoracic Surgery. Philadelphia, Lippincott-Raven, 1998.)

View larger version (133K): [in this window] [in a new window]   Figure 35-16 Completion of the closure of the aortotomy for the inclusion cylinder technique. The aortic cross-clamp is removed. The pulmonary homograft reconstruction of the outflow tract is accomplished with two continuous sutures of polypropylene. The proximal suture line is completed first. (Reproduced with permission from Schaff HV, Cable DG: Aortic valve replacement with homograft, in Kaiser LR, Kron IL, Spray TL (eds): Mastery of Cardiothoracic Surgery. Philadelphia, Lippincott-Raven, 1998.)

Ross initially reported his results in 1991, setting the standard for outcomes after the Ross procedure.⁶⁸ He reported follow-up in 339 patients up to 24 years with an 80% survival and an 85% freedom from reoperation. More recently, a longer follow-up report from this initial series included 131 patients (long-term survivors) with a mean follow-up of 20 years (range 9 to 26 years).⁶⁹ In this report, freedom from reoperation was 76 and 62% at 10 and 20 years, respectively. Freedom from autograft replacement was 88 and 75% at 10 and 20 years, respectively. The main indication prompting autograft replacement was severe regurgitation in 28/30 patients. At 25 years, the pulmonary homograft was free of replacement in 69% of patients. Indeed, these outcome results have set the benchmark for AV repairs using autologous pulmonic valve. It is worth noting that most of the patients in Ross’ series were repaired by the scalloped subcoronary implant technique, whereas the root replacement technique is the most commonly performed technique today.

Current data from the Ross Registry resemble Ross’ initial results with a 68% survival, 84% freedom from RVOT repair/replacement, and an 82% freedom from autograft explant over 25 years.

Operative risk

With increasing experience and improved perioperative care, operative risk has declined since Ross first described this procedure. According to the International Ross Registry, overall perioperative mortality (i.e., <30 days) is now 4.1% (129 deaths in 3922 patients). Although 4.1% perioperative mortality is acceptable for some cardiac operations, many believe it is unacceptably high for the younger patients who are most often subjected to the Ross procedure. This controversial point highlights the need for individualized therapy with respect to valve replacement through consideration of all the risks and benefits associated with the available options. We do not believe that the risks of the Ross procedure are prohibitive in appropriately selected patients. Furthermore, as more experience is gained with this procedure, a volume-outcome relationship may be apparent, supporting regional specialization and referrals to "centers of excellence."

Pulmonary autograft dysfunction

Early autograft dysfunction occurs infrequently. Elkins reported early autograft dysfunction (<6 months) in 3/195 patients (1.5%).⁶⁴

Late autograft dysfunction is an increasingly recognized phenomenon following the Ross procedure, although few studies have followed patients longitudinally with routine evaluations. A recent report of midterm follow-up (mean 2.47 years) in 132 consecutive patients who underwent routine echocardiographic evaluations of the pulmonary autograft revealed mild aortic insufficiency (graded 1/4) in 39.2 to 53.6% of patients, depending on follow-up interval.⁷⁰ Three percent of patients had moderate insufficiency early after surgery, increasing to 14.3% at 5 years. The mean transvalvular gradient across the AV was minimal (3 mm Hg) early and remained stable during follow-up.

In Elkins’ series of 289 patients, 6% of the patients (16 patients) required autograft reoperation.⁶⁴ In the patients who received root replacement implants, 97% had no change in autograft function during follow-up, while only 1% had severe insufficiency (3+). In contrast, of those who received scalloped subcoronary or inclusion cylinder implants, 86% had no change in autograft function, while 8% had progressed to severe insufficiency (3+).

David and associates have provided additional insight into late autograft dysfunction following the Ross procedure, by assessing dilatation of the pulmonary autograft after the Ross procedure.⁷¹ From 1990 to 1997, 118 patients with a mean age of 34 (17 to 57) underwent the Ross procedure. Of note, if there was 2 mm size mismatch between the aortic and pulmonic sinotubular junction or annuli, they were surgically reduced prior to implantation. The root replacement technique was most commonly used (71/118, 60%), followed by the root inclusion technique (45/118, 38%), and the subcoronary implant least commonly (2/118, 1.7%). Follow-up was 12 to 96 months (mean 44 months) including annual echocardiography. Over the observation period the diameter of the sinuses of Valsalva significantly increased from 31.4 to 33.7 mm. Furthermore, with respect to operative techniques, aortic root replacement was significantly positively correlated with increased risk of dilatation. No interpretable changes were seen in the aortic annulus over time. However, the diameter of the sinotubular junction increased in patients who had aortic root replacement and decreased in those subjected to root inclusion technique. During the observation period, only 5.9% (7/118) of patients developed moderate AI. However, all the patients with AI had dilatation of the aortic annulus and/or the sinotubular junction.

Taken together, these data indicate that the autologous pulmonic valve in the aortic position is quite durable and able to withstand the increased stresses at this position. Furthermore, although the valve effective orifice area (EOA) does not seem to change, up to 50% of patients may develop mild AI, which seems to increase with time, likely resulting from dilatation of the pulmonary autograft sinotubular junction and/or sinuses of Valsalva. Although it is suggested that this may be affected by surgical technique, it is unclear whether the use of the root replacement technique or root inclusion technique impacts the development of late pulmonary autograft dysfunction. Additional investigation is needed to elucidate risk factors for this complication and surgical interventions that may attenuate the frequency (e.g., reduction aortoplasty or Dacron banding of the aortic annulus and sinotubular junction).

Homograft dysfunction

Although the cryopreserved homograft has many advantages for a right ventricle to pulmonary artery conduit, it has become increasingly clear that it is susceptible to stenosis, degeneration, and calcification. Since current cryopreservation techniques result in varying degrees of donor cell viability, these effects may be immunologically mediated, although this remains controversial.⁷² Again, the precise incidence of this phenomenon is likely underappreciated due to lack of routine echocardiographic screening.

Midterm echocardiographic follow-up results of 132 patients (mean follow-up 2.47 years) after the Ross procedure demonstrated early minimal transvalvular gradients (3 ± 4 mm Hg) with significant worsening (+6 ± 8 mm Hg) over time.⁷⁰ Furthermore, there was decreasing valve EOA (–0.74 ± 0.82 cm²), mostly within the first 6 months. This resulted in 19.3% of patients having an EOA index of <0.85 cm²/m² at 1-year follow-up. Put another way, after 2 years the pulmonary valve EOAs were, on average, 31% less than immediately following surgery. Through multivariate analysis, only small homograft size and hypertension were found to be significant negative predictors of EOA at 1 year follow-up.

Raanani and colleagues followed 109 consecutive patients after the Ross procedure to identify the incidence and risk factors for homograft stenosis.⁷³ Echocardiographic follow-up (mean 39 ± 20 months) was available in 105 (97%) patients. The primary abnormality identified was homograft stenosis. In support, they identified a peak systolic transvalvular gradient of >20 mm Hg in 28.5% (30/105) of patients and >40 mm Hg peak gradient in 3.8% (4/105) of patients. This obstruction occurred at all levels of the homograft, as opposed to just at anastomotic sites, and was associated with homograft thickening. Moderate or severe homograft insufficiency was identified in 9.5% (10/105) of patients. Through multivariate analysis, the only independent significant predictors of homograft stenosis were cryopreservation duration <20 months, donor age <30 years, and small homograft size, which approached statistical significance.

Taken together, these data indicate that pulmonary homograft stenosis may develop in up to 30% of patients following the Ross operation. More importantly, the clinical significance of these data is not clear. For example, in the context of congenital pulmonary stenosis, it appears that peak transvalvular gradients of <50 mm Hg are usually well tolerated.⁷⁴ However, it would be erroneous to extrapolate these data to the demographic of the Ross procedure, as many of these patients are into adulthood when they undergo this procedure. Furthermore, similar to the pulmonic autograft in the aortic position, the incidence and severity of pulmonic insufficiency appears to increase with time. Clearly, these data suggest that future endeavors to improve the Ross procedure should be directed to elucidating the mechanisms underlying the pulmonary homograft stenosis and toward therapy to minimize these untoward effects.

	SUMMARY

Top Previous Next

 
The search for the ideal valve to replace the diseased AV is ongoing and available techniques are imperfect. Nonetheless, the Ross operation, whereby the pulmonic valve is transposed to the aortic position and the RVOT is replaced with cryopreserved allograft (most commonly), has proven to be a durable solution for complex congenital abnormalities and disease isolated to the AV, particularly in children and young adults.

Observations of the long-term durability of the pulmonic autograft in the aortic position affirm its ability to withstand the increased physical stresses at this location and maintain near normal hemodynamics over time, making it an excellent choice for replacement of a diseased AV. The incidence of AI is low and increases with time, likely resulting from dilatation of the pulmonary autograft sinotubular junction and/or the sinuses of Valsalva. Further studies are needed to discern the risk factors for this occurrence, but utilizing the root replacement technique compared to the root inclusion technique may affect it.

The allograft in the pulmonic position is more susceptible to complications, namely stenosis, which occurs primarily within the first year. Up to 30% of patients may have a hemodynamically significant stenosis at this location that is likely immunologically mediated, though the clinical significance of these findings remains unclear. Small homograft size is consistently a risk factor for stenosis, thereby supporting the current practice of oversizing the homograft by 2 to 3 mm. Future attempts to improve outcome after the Ross procedure should be directed toward reducing homograft stenosis.

The Ross procedure remains a reasonable option for replacement of the diseased AV, particularly in children and young adults. For adults, given the alternatives for AV and root replacement, the utility of the pulmonary autograft is more limited. As longer follow-up continues to accumulate, the risks and benefits of this procedure relative to other treatment options for replacement of a diseased AV will be better characterized.