- Email: [email protected]
Morphological findings in explanted Toronto stentless porcine valves
Cardiovascular Pathology 15 (2006) 41 – 48
Original Article
Morphological findings in explanted Toronto stentless porcine valvesB Jagdish Butanya,f,T, Michael J. Collinsa, Vidhya Naira, Richard L. Leaskc, Hugh E. Scullyb,e, William G. Williamsd,e, Tirone E. Davidb,e a Department of Pathology, Toronto Medical Laboratories, Toronto General Hospital/University Health Network, Toronto, Canada Department of Cardiac Surgery, Toronto Medical Laboratories, Toronto General Hospital/University Health Network, Toronto, Canada c Department of Chemical Engineering, McGill University, Montreal, Canada d Department of Cardiac Surgery, Hospital for Sick Children, Toronto, Canada e Department of Surgery, University of Toronto, Toronto, Canada f Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Canada
b
Received 19 July 2004; accepted 9 August 2005
Abstract Background: Stentless porcine valves have many documented advantages over stented valves. Since its introduction in 1991, the Toronto stentless porcine valve (T-SPV) has shown excellent hemodynamic performance. Methods: A total of 332 T-SPVs have been implanted at our institution up to December 2003, 25 of which have been explanted at surgery. Herein, we report a study of 30 explanted T-SPVs seen at our institution over a 5-year period. Results: The mean patient age at explant was 61.2F11.8 years with a mean implant duration of 100.7F 27.8 months (after excluding one valve removed early postoperatively for infective endocarditis). Twenty-seven of 30 valves (90%) showed structural deterioration characterized by tissue degeneration, cusp tears, calcification, and lipid insudation. Eight valves (26.7%) showed evidence of calcification on gross inspection and a total of 23 valves (76.7%) showed at least one microscopic focus of calcification, located primarily in the basal and commissural regions of the cusp. Twenty valves (66.6%) showed cusp tears. Pannus was visible grossly on the surface of 27 of 30 valves (90%), while histologically, at least some degree of pannus was observed on both the inflow and outflow surfaces of all but two valves. Twelve T-SPV (40.0%) showed calcification in the porcine aortic tissue, four of which involved calcification of the porcine muscle shelf in the right coronary cusp. Two T-SPV showed no significant structural deterioration. Their clinical reason for explantation was incompetence or infective endocarditis. Conclusion: With a freedom from reoperation of about 87.0% at up to 10 years, the T-SPV shows excellent durability. The majority of explanted valves show structural valve deterioration similar to that seen in other porcine heart valves. D 2006 Elsevier Inc. All rights reserved. Keywords: Toronto Stentless porcine valve; Bioprosthesis; Calcification; Mode of failure; Tissue failure
1. Introduction Valvular heart disease is an increasingly common medical problem in the 20th century. While rheumatic heart disease as a cause of valvular disease in developed countries B
The editor in charge of this article was chosen from the editorial board. The two regular Editors-in-Chief played no role in its peer review process. The editor for this paper was Dr G. Thiene, University of Padua Medical School, Padova, Italy. T Corresponding author. Department of Pathology, E4-322, Toronto General Hospital, Toronto, ON, Canada M5G 2C4. Tel.: +1 416 340 3003; fax: +1 416 340 4213. E-mail address: [email protected] (J. Butany). 1054-8807/06/$ – see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.carpath.2005.08.010
has decreased over the past 50 years, an increasing number of patients are undergoing surgery to replace valves affected by senile degeneration, myxomatous degeneration, congenital disease, and infective endocarditis [1]. Biological (bioprosthetic) heart valves account for over one third of all implanted prosthetic heart valves. These valves are composed of synthetic stents and supports with leaflets fashioned usually from porcine aortic valves (AVs) or bovine pericardium. Homografts, from human cadavers and autografts, the patient’s own pulmonary valve, are increasingly used in the correction of AV replacement. Contemporary stented bioprosthetic valves (as well as mechanical valves) all have a valve ring and sewing cuff,
42
J. Butany et al. / Cardiovascular Pathology 15 (2006) 41 – 48
which make them easy to implant. The prostheses by themselves are associated with some narrowing of the orifice and with a residual transvalvular gradient [2]. Stentless valve prostheses minimize the transvalvular gradient by eliminating the stent and maximizing the orifice. The Toronto stentless porcine valve (T-SPV, St. Jude Medical, St. Paul, MN) has been used at our institution since 1991 (clinical trial) and received FDA approval in 1997. It has a synthetic fabric covering the adventitial surface of the porcine aorta and folds over the proximal end of the valve conduit to increase ease of implantation. The valve is fixed in glutaraldehyde at a pressure of 1.5 mm Hg without antimineralization treatments and is sewn directly into the patients’ aorta and thus relies on the geometry of the patient’s aortic root. The T-SPV has shown excellent clinical results and durability compared with other bioprosthesis [3–5]. However, there have been no previous reports of any significantly sized series of explanted T-SPVs. Herein, we report a detailed analysis of 30 explanted T-SPV seen at our institution over a 5-year period.
2. Materials and methods This study examined 30 sequentially explanted T-SPVs, seen in the Department of Pathology at the Toronto General Hospital over a 5-year period by one pathologist (JB). All valves had been implanted in 1991 and subsequently, when the valve received initial FDA approval for clinical studies. All explanted valves were fixed in 10% formalin, photographed on the inflow and outflow surfaces, and radiographed using a Faxitron at 78 kV0.25 min (Hewlett Packard, McMinnville, OR). The gross features were described and recorded by the same pathologist (JB). Multiple histological sections were taken from the cusps as well as the porcine commissural regions along with the tissue and Dacron covering. Sections from the body of the cusp were taken perpendicular to the base when a longitudinal tear was absent, and transverse sections parallel to the free margin were taken when a longitudinal tear was present. All tissues were processed for light microscopy, embedded in paraffin, and sectioned at 5 Am. All sections were stained with hematoxylin and eosin and Movat pentachrome or elastic trichrome. Gram’s stain for bacteria was performed when indicated. Immunohistochemical stains (DAKO Scientific, Mississauga, ON, Canada) were undertaken; CD45 was used to determine the presence of leukocytes, CD68 for staining macrophages and CD31 for endothelial cells. For all patients, a detailed chart review was conducted and any relevant data including patient demographics, reason for native valve replacement, duration of function of the protheses, period between implantation and explantation, and concomitant surgery or associated cardiovascular pathology were noted. Significant events such as infection, thrombotic complications and hemorrhage were also recorded.
Cusp tears were graded according to the criteria of Ishihara et al. [6] as follows: Type 1, tears involving the free edges of the cusp; Type 2, tears that extended along the base of the valve cusps parallel to the sewing ring; Type 3, large round or oval perforations occupying the central regions of the cusps; Type 4, small pinhole-like perforations. The presence of host tissue overgrowth (pannus) and calcification, along with their respective locations were determined on both gross and histological examination [7]. Calcification was graded as absent or present in the cusps, porcine aortic tissue and pannus.
3. Results 3.1. Patients Between 1991 and 2003, over 332 T-SPVs have been implanted at our institution. Of the 30 valves examined in this study, 25 were from patients undergoing surgical replacement of their T-SPV at our institution, one was explanted at autopsy, and four were referred from other institutions for pathological examinations. Clinical data on these four referred valves were somewhat limited and these are discussed separately. There were 24 males and 6 females with the patient’s age at the time of valve implantation ranging from 36 to 74 years (meanFS.D., 54.0F11.4 years). Nine patients had calcified three cuspid native AVs and 12 had congenitally bicuspid AVs (BAVs). Four patients received a T-SPV to replace a failed bioprosthetic valve,
Table 1 Demographics and patient data for T-SPV explant cases Total cases Number of males Number of females Referred cases Meant age at implant Mean age at explant Native valve type Three cusps Bicuspid Previous prosthesis Unable to determine Reason for T-SPV implantation Aortic incompetence Aortic stenosis Both AI and AS Previous prosthesis Unable to determine Reason for explant Aortic incompetence Incompetence and stenosis Aortic stenosis Infective endocarditis Autopsy Mean implant duration Mean implant duration (excluding early case of valve failure)
30 24 7 4 54.0F11.8 years 61.2F11.4 years 9 12 4 1 4 16 1 4
19 1 1 4 2 97.1F33.4 months 100.7F27.8 months
J. Butany et al. / Cardiovascular Pathology 15 (2006) 41 – 48
43
3.2. Pathology
Fig. 1. Explants by postoperative year.
and we were unable to determine the status of one patient’s native valve. The patients’ age at the time of valve explantation ranged from 42 to 82 years (61.2F11.8 years). The explanted valves had been in place for an average of 97.1F33.4 months. One valve was explanted after only 1 month, for infective endocarditis, while all the remaining valves lasted for more than 2.5 years. If this valve is excluded as an early failure (less than 30 days postoperatively), the mean implant duration was 100.7F27.8 months. The demographic and preoperative data are listed in Table 1 and Fig. 1.
Twenty-seven of 30 valves (90%) showed structural deterioration characterized by tissue degeneration, cusp tears, cusp calcification, lipid insudation and significant pannus growth (Table 2, Fig. 2). Of the three remaining valves, two were explanted for infective endocarditis (at 1 and 91 months) and the third showed surprisingly little structural deterioration despite being in place for almost 10 years (Fig. 3), the clinical reason for explantation was incompetence. Eleven of the 12 valves explanted from patients with native BAV showed SVD. Four of these 12 valves showed significant SVD with gross evidence of calcification and cusp tears. The other seven valves showed microfoci of calcification and a few small tears close to commissural regions while one T-SPV, explanted at 111 months, showed no calcification and no tears. Fifteen valves showed lipid insudation on gross inspection, while microscopically, five cases showed significant fat cell and fluid infiltration at the base of the cusp. Seven cases showed areas of recent hemorrhage into the cusp (most likely related to the explantation process) and two showed loss of elastic tissue with notable thinning of the valve leaflets. In
Table 2 Morphological findings explanted T-SPVs Gross calcification (out of 30 valves) Present on gross None Location of calcification (out of 8) Commissural region Body of leaflets Both Pannus (out of 30 valves) Present Absent Location of Pannus (out of 27) On flow surface On nonflow surface Both Cannot determine Lipid insudation Absent Present Cusp tears (out of 30 valves) None Single tears Multiple tears Unable to determine Type 1 tears Type 2 tears Type 3 tears Type 4 tears Types 1 and 2 Type 1, 3 Type 2, 3
8 22 1 5 2 27 3 13 12 12 2 15 15 8 8 12 2 11 2 1 1 2 2 1
Fig. 2. Inflow (A) and outflow (B) surfaces of a T-SPV from a 56-year-old male at 9 years. (Panel A) Inflow surface shows pannus present all around the circumference of the valve (white arrows). One cusp shows a large tear with calcification around the margins. (Panel B) The outflow surface shows a ridge of calcification (arrowheads) close to one commissure.
44
J. Butany et al. / Cardiovascular Pathology 15 (2006) 41 – 48
3.3. Calcification Eight of the explanted valves (26.7%) showed evidence of calcification on gross inspection and radiologic examination. The mean age of the grossly calcified valves was 117.4F24.5 months while the valves that grossly showed no calcification were in situ for a significantly shorter duration 88.5F34.4 months ( P=.027, two-tailed Mann–Whitney test). Calcific nodules were noted in the body of the cusp in five cases (Figs. 1 and 4), predominantly at the base of the cups. One showed nodular calcification in the commissural region only, while two valves showed gross calcification involving both the body and base of the cusps. Three of the valves showed gross calcification of the cusps directly associated with cusp tears. On histological examination, a total of 23 valves (76.6%) showed at least one microfocus of calcification in cusp tissue. Microfoci of calcification were seen at the base of the cusps in almost all cases; in addition, six showed calcification of cusp tissue in the commissural region while one case showed a single small microfocus of calcification at the tip of a cusp. Eight cases (26.7%) showed calcification in the porcine aortic tissue (Fig. 4), while another four valves showed calcification of the muscle shelf in the porcine right coronary cusp. One
Fig. 3. A T-SPV from a 72-year-old male explanted for bioprosthesis incompetence at 119 months. (Panel A) The inflow surface shows intact cusps and pannus, which extends onto the base of the cusps and areas where the pannus was lost during surgical explantation (white arrows). (Panel B) The outflow surface shows intact cusps and mild pannus in the commissural regions. The commissural regions are mildly splayed outward. (Panel C) Lateral view of the bioprosthesis shows the intact fabric and cusps. Pannus is seen. This belies the feeling that explantation of these prostheses would be difficult to impossible.
addition, two valves showed areas of inflammatory cell infiltration into the leaflets though neither one of these cases were associated with infectious microorganisms.
Fig. 4. (Panel A) Histological sections of a T-SPV explanted from a 46-yearold male at 103 months showing well-preserved tissue with microfocal areas of calcification. (Panel B) Transverse section at a commissure shows microfoci of calcification of the cusps (arrows) and of the porcine aortic cuff (asterisk).
J. Butany et al. / Cardiovascular Pathology 15 (2006) 41 – 48
45
Fig. 6. A T-SPV explanted from a 72-year-old male at 112 months shows cusp tears (white arrows).
significant difference in the ages of patients who showed cusp tears and the patients with no cusp tears. Six valves that showed the presence of calcification on gross examination also showed the presence of cusp tears, while 19 of 23 valves showing microfoci of calcification also showed cusp tears. 3.5. Pannus
Fig. 5. Histological sections of a T-SPV explanted from a 72-year-old male at 125 months. Panel A shows inflammatory infiltrate around the porcine aortic tissue. (Panel B) Higher magnification of the boxed area from panel A shows mixed inflammatory infiltrate including multinucleated giant cells [hematoxylin and eosin, original magnification 25 (A), 200 (B)].
Pannus was visible grossly on the surface of 27 of 30 valves (90.0%). Pannus was present grossly on inflow (ventricular) surface alone in 13 valves, while 12 valves showed the gross presence of pannus on both the inflow and outflow (aortic) surfaces (Fig. 7). In 2 cases, the pannus was received detached from the valve, and from its shape it is most likely that it had been detached from the valve ring, that is the inflow surface of the bioprosthesis. Histologically, at least some degree of pannus was observed on both the inflow and outflow (aortic) surface of all but two valves. Surprisingly, one of the valves that did not show pannus had been in
valve showed an inflammatory reaction to the porcine aortic tissues (Fig. 5). Ten of the 15 valves showing lipid insudation also showed at least one focus of calcification. 3.4. Tears and perforations Twenty valves (66.6%) showed cusp tears. These valves had been in place for an average of 107.2F24.8 months. Eight valves showed only one tear while 12 showed multiple tears affecting multiple cusps. Two valves were explanted and received in several pieces, making it difficult to determine the presence or absence of tears and perforations. Of the 20 valves, 11 showed Type 1 tears located predominantly in the commissural region (Fig. 6). Two valves showed Type 2 tears, one valve showed a Type 3 tear and one a Type 4 tear. Two valves showed both a commissural region Type 1 tear and a Type 2 tear, two showed a Type 1 tear and a Type 3 tear (perforation) in the same valve, while one showed the presence of a Type 2 and Type 3 tear in the same valve. In contrast, the eight valves (33.3%) that did not show any tears had been in place for between 1 and 120 months with a mean implant duration of 77.8F46.3 months. There was no
Fig. 7. T-SPV explanted at 104 months from a 58-year-old male. Longitudinal sections of the cusp and porcine aortic tissues show significant pannus on the inflow (black arrow) and outflow (white arrow) surfaces. Focal calcification is seen close to the cusp junction with the pannus (arrowhead).
46
J. Butany et al. / Cardiovascular Pathology 15 (2006) 41 – 48
place for 111 months (it is possible that the pannus had gotten detached during explantation and was not submitted for pathological examination), while the other had been in place for less than 1 month. Four cases (13.3%) showed pannus on the base of the cusp, which extended onto the cusp tissue for a significant distance. Fourteen cases showed calcification of pannus. In these cases, the pannus would likely have contributed to valve dysfunction. In none of the cases was pannus the only clinical indication for explantation. In the one case where the indication for T-SPV explantation was stenosis, there were extensive cusp calcification, tears and pannus seen at 135 months postimplantation in a 44-year-old male. In addition, perivalvular (surrounding the fabric) fibrosis and pannus were present in 21 (70.0%) valves. Seven cases showed the deposition of significant fibrous tissue between the porcine aorta and the Dacron graft covering the external surface of the prosthesis. Nineteen cases showed host tissue growth around the porcine aorta, in the area between the Dacron fabric and the host aorta. In one case, the presence of pannus on the nonflow surfaces was associated with small amounts of thrombus, suggesting the possibility of a perivalvular leak. Interestingly, six cases showed host tissue infiltration and fibrous tissue deposition within the interstices of the synthetic material with 3 of these cases showing calcification within the Dacron fabric and the fibrous tissue. All cases examined showed some element of host inflammatory reaction to the synthetic material. This reaction was relatively mild in 9 (30%) of cases. Immunohistochemical stains confirmed that this host reaction was chronic in nature and was composed primarily of macrophages, multinucleate giant cells and lymphocytes. Small areas of thrombus were visible on 11 (36.7%) of the valves, mainly at the base of the cusps. In 10 of these cases, the thrombus had formed on the sinus surface of the cusps, while in the remaining, thrombus had formed in the region of the commissures superimposed on existing pannus. 3.6. Infective endocarditis Four patients developed infective endocarditis requiring explantation of their T-SPV. The mean patient age at explantation was 59.3F15.4 years and these valves had been in place for a mean of 47.5F39.9 months, but this included the one patient who developed IE within 1 month of implantation. One patient showed the presence of Gram-positive microorganisms, another showed a coagulase-negative streptococcus, a third case showed small microorganisms, possibly Bordatella while in the remaining case no microorganisms could be demonstrated on histology. The explanted valves showed minor structural valve deterioration (SVD), none showed any gross evidence of calcification and only one showed a microfocus of calcification histologically. However, three of the four cases showed extensive pannus with extension onto the cusps resulting in cusp thickening and stiffening, which made it difficult for the cusps to move.
4. Discussion Since 1991, the T-SPV has been widely used in a number of institutions with studies documenting excellent hemodynamic performance and durability [3,4]. Since then, it has been used worldwide with equally good results. The T-SPV shows a low transvalvular gradient, large effective orifice area, decreased left ventricular mass index in patients receiving the valve and possibly improved patient survival, compared with stented valves [5]. A follow-up study of 332 patients by David et al. showed freedom from SVD at 10 years to be 99% in patients greater than 65 years of age and 77% in patients under 65 years of age [8]. In the current study, the mean patient age at the time of T-SPV implant was only 54.0F11.4 years, which likely accounted for the high incidence of SVD. The T-SPV has been quite durable in comparison to other prosthetic valves such as the Hancock II stented porcine and the Carpentier–Edwards stented porcine [5,9,10]. Of the explanted valves examined in our study, 27 (90%) showed SVD. SVD has been observed as a major cause of failure in other prosthetic valves. In a similar study, Butany et al. [7,11] observed an incidence of SVD of approximately 71% for the Ionescu–Shiley low-profile pericardial prosthesis, which had been in place for 2–5 years, and 56% of the Hancock II bioprosthesis, which had been in place for a mean of 5.1 years. In contrast, the T-SPV examined in this study had a mean implant duration of 96 months (8.0 years), including one valve explanted after only 1 month due to infective endocarditis. Our results therefore support earlier studies that found only minimal tissue degeneration in explanted T-SPV at up to six years [12]. Since the T-SPV is implanted in the subcoronary position, proper valve function is dependent on the native aorta. Previous investigations of the T-SPV describe a strong correlation between dilation of the sinotubular junction (STJ) and the occurrence of aortic incompetence [13,14] T-SPV failure has been related to dilation of the ascending aorta in the absence of SVD [15]. Although no measure of aortic dilation was collected in our study, 12 patients had native bicuspid AV disease, a known cause of aortic root dilation [16]. Eleven of the 12 T-SPVs explanted from patients with native BAV valves showed some evidence of SVD with four valves showing significant SVD with gross evidence of calcification and multiple cusp tears. In the remaining valves, SVD was mild and the cusp tears may well have been secondary to STJ dilation. Only eight valves in our study showed gross evidence of calcification. These valves had been in place for a mean of 9.75 years (117 months), which is higher than the mean duration of implantation for other valves in reported studies [7,11,17,18]. A majority of the valves did not show gross calcification, which supports the findings from previous studies at up to 6 years postimplantation [5,12]. However, earlier case reports had suggested that calcification of the T-SPV can occasionally begin early. Butany and David [19]
J. Butany et al. / Cardiovascular Pathology 15 (2006) 41 – 48
reported a case of a T-SPV, which showed evidence of calcification 32 months postimplant. In addition, Tansley et al. [20] reported a case of calcific stenosis in a T-SPV 6 years postimplant. Recent studies using subdermal implants in animal models showed high levels of calcium (165 mg/g dry tissue weight in leaflets and 94.6 mg/g dry weight in aorta) in the T-SPV compared with other stented and stentless prosthesis [21]. Evidence suggests that the valve fixation process may contribute to the process of calcification. The extent of calcification is reported to be directly dependent on the amount of glutaraldehyde present in the tissue [22]. Ultrastructurally, T-SPV implanted subdermally showed a clear alteration in their surface superstructure, with loss of endothelium and disruption of the helical structure of collagen bundles [21]. Fixation under pressure can lead to loss of endothelial cells and alterations in 3D structure of collagen, which can lead to disruption of fibers under strain [23]. These disrupted collagen fibers can act as nucleation sites for calcification [21]. Interestingly, while only eight T-SPVs in our study showed gross evidence of calcification, 23 valves (76.6%) showed at least one microscopic focus of calcification in cuspal tissue. Almost all cases showed calcification in the base of the cusps. Of the 23 valves that showed microscopic calcification, 19 showed tears. In a similar study of explanted Hancock II bioprosthesis, it was observed that almost 80% of aortic prostheses with evidence of calcification also showed tears [7]. These observations suggest that even small foci of calcification may alter the mechanical stress on the leaflet tissue and predispose the cusp to tears. It has been suggested that extracellular lipid droplets can act as nucleation sites for calcification as lipid droplets and areas of leaflet calcification show a similar pattern of distribution [24]. It is interesting to note that 10 of the 15 valves in this study, which showed gross evidence of lipid insudation, also showed microscopic evidence of calcification. Two thirds of T-SPVs examined in this study showed evidence of cusp tear(s). Type 1 tears were the most common and usually affected the commisural regions. Our results are similar to other studies where examination of the Hancock II bioprosthesis showed tears in 53% of valves, with Type 1 tears being the most common [7]. Examination of the porcine aorta showed calcification in the porcine aortic tissue of 12 valves. Interestingly, 10 valves showed the shelf of porcine muscle at the right coronary cusp base, and in four of these valves, the muscle shelf showed evidence of calcification while one showed an inflammatory reaction to the porcine tissue. Calcification of the porcine wall, though relatively rare, can have significant effects on hemodynamics and stiffening of the aortic root [25]. In addition to calcification, significant pannus was observed on the sinus surface of the valve in all but 2 cases. Pannus is considered to be the deposition of exuberant host tissue on the surface of grafts. It is composed primarily of fibroblasts and collagen. In moderate amounts, pannus can be beneficial as it covers the suture lines allowing a nonthrombogenic surface to form
47
between the graft and the host aorta [19]. However, excessive pannus can result in stenosis, cusp thickening and stiffening, which can hinder the mobility of the leaflets [1,12]. The valves in the series that showed significant pannus were mainly removed for AI or IE. These valves also showed calcification and/or cusp tears. Hence, while the significant degree of pannus would have contributed to valve dysfunction, it is not possible to say whether valve failure was due to pannus alone. In the present study, one valve showed significant pannus on the outflow surface of the aorta extending onto the cusp such that the flexion point would likely have changed. Such a change would have contributed to the significant tears observed in that valve. Pannus was also observed outside the T-SPV, both between the porcine aorta and the Dacron graft and between the Dacron graft and the host tissues. Interestingly, 14 cases showed calcification of the pannus on the sinus surface of the porcine aorta. In most cases, the surgeon was able to explant the bioprosthesis, without significant damage to it, using the plane around the Dacron fabric covering the bioprosthesis. The above changes in the porcine aorta are significant for the design of stentless valves as current antimineralization pretreatments, while effective at reducing cusp calcification, do not reduce calcification of the porcine aorta [26]. In fact, pretreatments that remove cells and collagen from the aorta do not reduce calcification of the porcine aorta, whereas degradation and fragmentation of elastin appear to cause increased calcification [26]. Matrix metalloproteinases are a family of enzymes that can be secreted by a number of different cell types, including fibroblasts [27,28], and can degrade components of the extracellular matrix. Studies have shown that MMP-mediated collagen degradation is an initial step in elastin calcification [29]. From these data it appears possible that fibroblasts present in pannus may play a role in promoting calcification of the porcine aorta, possibly by secreting proteolytic enzymes which damage elastin in the porcine aorta.
5. Conclusion In our study of 30 explanted T-SPVs, 90% showed structural deterioration at a mean implant duration of 96.2 months. Eight valves showed gross evidence of calcification, though 23 showed microscopic foci of calcification in the base of the leaflets, in the commissural regions in the pannus and in the porcine aortic tissue. All implanted prostheses showed pannus. The higher incidence of pannus on the T-SPV as compared with stented valves [7,11] is likely related to the absence of the large sewing cuff, which separates the valve itself from the host tissue. A new mode of failure was seen in two cases, wherein the cuspal tissue was largely intact, though the bioprosthesis had become functionally incompetent due to dilatation of the sinotubular junction. This is likely related to the dilatation of the native ascending aortic tissue in a patient with a congenitally
48
J. Butany et al. / Cardiovascular Pathology 15 (2006) 41 – 48
abnormal AV. The T-SPV, an aldehyde-fixed bioprostheses, provides excellent long-term clinical results; however, it appears to have the deficiencies of all other glutaraldehydefixed porcine valves. Further analyses of larger series of explanted bioprostheses are essential.
References [1] Butany J, Leask R. The failure modes of biological prosthetic heart valves. J Long-Term Eff Med Implants 2001;11:115 – 35. [2] Park SZ, Reardon MJ. Current status of stentless aortic xenografts. Curr Opin Cardiol 2000;15:74 – 81. [3] David TE, Feindel CM, Scully HE, Bos J, Rakowski H. Aortic valve replacement with stentless porcine aortic valves: a ten-year experience. J Heart Valve Dis 1998;7:250 – 4. [4] Goldman B, Christakis G, David T, Rakowski H, Bach D, Wood J, Goldman S, Pepper J, Yacoub M, Verrier E, Petracek M. Will stentless valves be durable? The Toronto valve (TSPV) at 5 to 6 years. Semin Thorac Cardiovasc Surg 1999;11:42 – 9. [5] Bach DS, Goldman B, Verrier E, Petracek M, Wood J, Goldman S, David TE. Durability and prevalence of aortic regurgitation nine years after aortic valve replacement with the Toronto SPV stentless bioprosthesis. J Heart Valve Dis 2004;13:64 – 72 [discussion 72]. [6] Ishihara T, Ferrans VJ, Boyce SW, Jones M, Roberts WC. Structure and classification of cuspal tears and perforations in porcine bioprosthetic cardiac valves implanted in patients. Am J Cardiol 1981;48:665 – 78. [7] Butany J, Yu W, Silver MD, David TE. Morphologic findings in explanted Hancock II porcine bioprostheses. J Heart Valve Dis 1999;8: 4 – 15. [8] David TE, Feindel CM, Bos J, Ivanov J, Armstrong S. Ten-year results of aortic valve replacement with Toronto SPV bioprosthesis. Presented at the 84th Annual Meeting of the American Association for Thoracic Surgery, Toronto, Canada, April 25–28, 2004. [9] Banbury MK, Cosgrove DM, Lytle BW, Smedira NG, Sabik JF, Saunders CR. Long-term results of the Carpentier–Edwards pericardial aortic valve: a 12-year follow-up. Ann Thorac Surg 1998;66:S73 – 6. [10] David TE, Ivanov J, Armstrong S, Feindel CM, Cohen G. Late results of heart valve replacement with the Hancock II bioprosthesis. J Thorac Cardiovasc Surg 2001;121:268 – 77. [11] Butany J, Vanlerberghe K, Silver MD. Morphologic findings and causes of failure in 24 explanted Ionescu–Shiley low-profile pericardial heart valves. Hum Pathol 1992;23:1224 – 33. [12] Butany J, de Sa M, Feindel CM, David TE. The Toronto SPV bioprosthesis: review of morphological findings in eight valves. Semin Thorac Cardiovasc Surg 1999;11:157 – 62. [13] David TE, Ivanov J, Eriksson MJ, Bos J, Feindel CM, Rakowski H. Dilation of the sinotubular junction causes aortic insufficiency after aortic valve replacement with the Toronto SPV bioprosthesis. J Thorac Cardiovasc Surg 2001;122:929 – 34. [14] Shargall Y, Goldman B, Christakis G, David T. Analysis of explants and causes of mortality during long-term follow-up of the Toronto stentless porcine valve. Semin Thorac Cardiovasc Surg 2001;13:106 – 12.
[15] Butany J, Ahn C, Leask RL, Graba J, David T, Ahluwalia MS. Stentless porcine valves: new mode of failure. Can J Cardiol 2003;19: 185 – 6. [16] Keane MG, Wiegers SE, Plappert T, Pochettino A, Bavaria JE, Sutton MG. Bicuspid aortic valves are associated with aortic dilatation out of proportion to coexistent valvular lesions. Circulation 2000;102: III35 – 9. [17] Bortolotti U, Milano A, Mazzaro E, Thiene G, Talenti E, Casarotto D. Hancock II porcine bioprosthesis: excellent durability at intermediateterm follow-up. J Am Coll Cardiol 1994;24:676 – 82. [18] Bottio T, Thiene G, Pettenazzo E, Ius P, Bortolotti U, Rizzoli G, Valfre C, Casarotto D, Valente M. Hancock II bioprosthesis: a glance at the microscope in mid-long-term explants. J Thorac Cardiovasc Surg 2003;126:99 – 105. [19] Butany J, David TE. Pathology of explanted stentless porcine valves (human experience). In: Gabbay S, editor. New horizons and the future of heart valve prosthesis. Austin7 Silent Partners Inc, 1994 pp. 183 – 9. [20] Tansley PD, Sheppard MN, Pepper J. Symptomatic calcific stenosis of a Toronto stentless porcine valve. Eur J Cardiothorac Surg 2000;17: 763 – 5. [21] Mirzaie M, Meyer T, Schorn B, Schwartz P, Baryalei M, Rastan A, Lotfi S, Dalichau H. Calcification tendency of various biological aortic valves in an experimental animal model. Cardiovasc Surg 1999; 7:735 – 41. [22] Golomb G, Schoen FJ, Smith MS, Linden J, Dixon M, Levy RJ. The role of glutaraldehyde-induced cross-links in calcification of bovine pericardium used in cardiac valve bioprostheses. Am J Pathol 1987; 127:122 – 30. [23] Broom ND, Thomson FJ. Influence of fixation conditions on the performance of glutaraldehyde-treated porcine aortic valves: towards a more scientific basis. Thorax 1979;34:166 – 76. [24] Dunmore-Buyze J, Boughner DR, Macris N, Vesely I. A comparison of macroscopic lipid content within porcine pulmonary and aortic valves. Implications for bioprosthetic valves. J Thorac Cardiovasc Surg 1995;110:1756 – 61. [25] Schoen FJ, Levy RJ. Founder’s Award, 25th Annual Meeting of the Society for Biomaterials, perspectives. Providence, RI, April 28–May 2, 1999. Tissue heart valves: current challenges and future research perspectives. J Biomed Mater Res 1999;47:439 – 65. [26] Bailey MT, Pillarisetti S, Xiao H, Vyavahare NR. Role of elastin in pathologic calcification of xenograft heart valves. J Biomed Mater Res 2003;66A:93 – 102. [27] Shi Y, Patel S, Niculescu R, Chung W, Desrochers P, Zalewski A. Role of matrix metalloproteinases and their tissue inhibitors in the regulation of coronary cell migration. Arterioscler Thromb Vasc Biol 1999;19:1150 – 5. [28] Rabkin E, Aikawa M, Stone JR, Fukumoto Y, Libby P, Schoen FJ. Activated interstitial myofibroblasts express catabolic enzymes and mediate matrix remodeling in myxomatous heart valves. Circulation 2001;104:2525 – 32. [29] Vyavahare N, Jones PL, Tallapragada S, Levy RJ. Inhibition of matrix metalloproteinase activity attenuates tenascin-C production and calcification of implanted purified elastin in rats. Am J Pathol 2000;157: 885 – 93.
Original Article
Morphological findings in explanted Toronto stentless porcine valvesB Jagdish Butanya,f,T, Michael J. Collinsa, Vidhya Naira, Richard L. Leaskc, Hugh E. Scullyb,e, William G. Williamsd,e, Tirone E. Davidb,e a Department of Pathology, Toronto Medical Laboratories, Toronto General Hospital/University Health Network, Toronto, Canada Department of Cardiac Surgery, Toronto Medical Laboratories, Toronto General Hospital/University Health Network, Toronto, Canada c Department of Chemical Engineering, McGill University, Montreal, Canada d Department of Cardiac Surgery, Hospital for Sick Children, Toronto, Canada e Department of Surgery, University of Toronto, Toronto, Canada f Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Canada
b
Received 19 July 2004; accepted 9 August 2005
Abstract Background: Stentless porcine valves have many documented advantages over stented valves. Since its introduction in 1991, the Toronto stentless porcine valve (T-SPV) has shown excellent hemodynamic performance. Methods: A total of 332 T-SPVs have been implanted at our institution up to December 2003, 25 of which have been explanted at surgery. Herein, we report a study of 30 explanted T-SPVs seen at our institution over a 5-year period. Results: The mean patient age at explant was 61.2F11.8 years with a mean implant duration of 100.7F 27.8 months (after excluding one valve removed early postoperatively for infective endocarditis). Twenty-seven of 30 valves (90%) showed structural deterioration characterized by tissue degeneration, cusp tears, calcification, and lipid insudation. Eight valves (26.7%) showed evidence of calcification on gross inspection and a total of 23 valves (76.7%) showed at least one microscopic focus of calcification, located primarily in the basal and commissural regions of the cusp. Twenty valves (66.6%) showed cusp tears. Pannus was visible grossly on the surface of 27 of 30 valves (90%), while histologically, at least some degree of pannus was observed on both the inflow and outflow surfaces of all but two valves. Twelve T-SPV (40.0%) showed calcification in the porcine aortic tissue, four of which involved calcification of the porcine muscle shelf in the right coronary cusp. Two T-SPV showed no significant structural deterioration. Their clinical reason for explantation was incompetence or infective endocarditis. Conclusion: With a freedom from reoperation of about 87.0% at up to 10 years, the T-SPV shows excellent durability. The majority of explanted valves show structural valve deterioration similar to that seen in other porcine heart valves. D 2006 Elsevier Inc. All rights reserved. Keywords: Toronto Stentless porcine valve; Bioprosthesis; Calcification; Mode of failure; Tissue failure
1. Introduction Valvular heart disease is an increasingly common medical problem in the 20th century. While rheumatic heart disease as a cause of valvular disease in developed countries B
The editor in charge of this article was chosen from the editorial board. The two regular Editors-in-Chief played no role in its peer review process. The editor for this paper was Dr G. Thiene, University of Padua Medical School, Padova, Italy. T Corresponding author. Department of Pathology, E4-322, Toronto General Hospital, Toronto, ON, Canada M5G 2C4. Tel.: +1 416 340 3003; fax: +1 416 340 4213. E-mail address: [email protected] (J. Butany). 1054-8807/06/$ – see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.carpath.2005.08.010
has decreased over the past 50 years, an increasing number of patients are undergoing surgery to replace valves affected by senile degeneration, myxomatous degeneration, congenital disease, and infective endocarditis [1]. Biological (bioprosthetic) heart valves account for over one third of all implanted prosthetic heart valves. These valves are composed of synthetic stents and supports with leaflets fashioned usually from porcine aortic valves (AVs) or bovine pericardium. Homografts, from human cadavers and autografts, the patient’s own pulmonary valve, are increasingly used in the correction of AV replacement. Contemporary stented bioprosthetic valves (as well as mechanical valves) all have a valve ring and sewing cuff,
42
J. Butany et al. / Cardiovascular Pathology 15 (2006) 41 – 48
which make them easy to implant. The prostheses by themselves are associated with some narrowing of the orifice and with a residual transvalvular gradient [2]. Stentless valve prostheses minimize the transvalvular gradient by eliminating the stent and maximizing the orifice. The Toronto stentless porcine valve (T-SPV, St. Jude Medical, St. Paul, MN) has been used at our institution since 1991 (clinical trial) and received FDA approval in 1997. It has a synthetic fabric covering the adventitial surface of the porcine aorta and folds over the proximal end of the valve conduit to increase ease of implantation. The valve is fixed in glutaraldehyde at a pressure of 1.5 mm Hg without antimineralization treatments and is sewn directly into the patients’ aorta and thus relies on the geometry of the patient’s aortic root. The T-SPV has shown excellent clinical results and durability compared with other bioprosthesis [3–5]. However, there have been no previous reports of any significantly sized series of explanted T-SPVs. Herein, we report a detailed analysis of 30 explanted T-SPV seen at our institution over a 5-year period.
2. Materials and methods This study examined 30 sequentially explanted T-SPVs, seen in the Department of Pathology at the Toronto General Hospital over a 5-year period by one pathologist (JB). All valves had been implanted in 1991 and subsequently, when the valve received initial FDA approval for clinical studies. All explanted valves were fixed in 10% formalin, photographed on the inflow and outflow surfaces, and radiographed using a Faxitron at 78 kV0.25 min (Hewlett Packard, McMinnville, OR). The gross features were described and recorded by the same pathologist (JB). Multiple histological sections were taken from the cusps as well as the porcine commissural regions along with the tissue and Dacron covering. Sections from the body of the cusp were taken perpendicular to the base when a longitudinal tear was absent, and transverse sections parallel to the free margin were taken when a longitudinal tear was present. All tissues were processed for light microscopy, embedded in paraffin, and sectioned at 5 Am. All sections were stained with hematoxylin and eosin and Movat pentachrome or elastic trichrome. Gram’s stain for bacteria was performed when indicated. Immunohistochemical stains (DAKO Scientific, Mississauga, ON, Canada) were undertaken; CD45 was used to determine the presence of leukocytes, CD68 for staining macrophages and CD31 for endothelial cells. For all patients, a detailed chart review was conducted and any relevant data including patient demographics, reason for native valve replacement, duration of function of the protheses, period between implantation and explantation, and concomitant surgery or associated cardiovascular pathology were noted. Significant events such as infection, thrombotic complications and hemorrhage were also recorded.
Cusp tears were graded according to the criteria of Ishihara et al. [6] as follows: Type 1, tears involving the free edges of the cusp; Type 2, tears that extended along the base of the valve cusps parallel to the sewing ring; Type 3, large round or oval perforations occupying the central regions of the cusps; Type 4, small pinhole-like perforations. The presence of host tissue overgrowth (pannus) and calcification, along with their respective locations were determined on both gross and histological examination [7]. Calcification was graded as absent or present in the cusps, porcine aortic tissue and pannus.
3. Results 3.1. Patients Between 1991 and 2003, over 332 T-SPVs have been implanted at our institution. Of the 30 valves examined in this study, 25 were from patients undergoing surgical replacement of their T-SPV at our institution, one was explanted at autopsy, and four were referred from other institutions for pathological examinations. Clinical data on these four referred valves were somewhat limited and these are discussed separately. There were 24 males and 6 females with the patient’s age at the time of valve implantation ranging from 36 to 74 years (meanFS.D., 54.0F11.4 years). Nine patients had calcified three cuspid native AVs and 12 had congenitally bicuspid AVs (BAVs). Four patients received a T-SPV to replace a failed bioprosthetic valve,
Table 1 Demographics and patient data for T-SPV explant cases Total cases Number of males Number of females Referred cases Meant age at implant Mean age at explant Native valve type Three cusps Bicuspid Previous prosthesis Unable to determine Reason for T-SPV implantation Aortic incompetence Aortic stenosis Both AI and AS Previous prosthesis Unable to determine Reason for explant Aortic incompetence Incompetence and stenosis Aortic stenosis Infective endocarditis Autopsy Mean implant duration Mean implant duration (excluding early case of valve failure)
30 24 7 4 54.0F11.8 years 61.2F11.4 years 9 12 4 1 4 16 1 4
19 1 1 4 2 97.1F33.4 months 100.7F27.8 months
J. Butany et al. / Cardiovascular Pathology 15 (2006) 41 – 48
43
3.2. Pathology
Fig. 1. Explants by postoperative year.
and we were unable to determine the status of one patient’s native valve. The patients’ age at the time of valve explantation ranged from 42 to 82 years (61.2F11.8 years). The explanted valves had been in place for an average of 97.1F33.4 months. One valve was explanted after only 1 month, for infective endocarditis, while all the remaining valves lasted for more than 2.5 years. If this valve is excluded as an early failure (less than 30 days postoperatively), the mean implant duration was 100.7F27.8 months. The demographic and preoperative data are listed in Table 1 and Fig. 1.
Twenty-seven of 30 valves (90%) showed structural deterioration characterized by tissue degeneration, cusp tears, cusp calcification, lipid insudation and significant pannus growth (Table 2, Fig. 2). Of the three remaining valves, two were explanted for infective endocarditis (at 1 and 91 months) and the third showed surprisingly little structural deterioration despite being in place for almost 10 years (Fig. 3), the clinical reason for explantation was incompetence. Eleven of the 12 valves explanted from patients with native BAV showed SVD. Four of these 12 valves showed significant SVD with gross evidence of calcification and cusp tears. The other seven valves showed microfoci of calcification and a few small tears close to commissural regions while one T-SPV, explanted at 111 months, showed no calcification and no tears. Fifteen valves showed lipid insudation on gross inspection, while microscopically, five cases showed significant fat cell and fluid infiltration at the base of the cusp. Seven cases showed areas of recent hemorrhage into the cusp (most likely related to the explantation process) and two showed loss of elastic tissue with notable thinning of the valve leaflets. In
Table 2 Morphological findings explanted T-SPVs Gross calcification (out of 30 valves) Present on gross None Location of calcification (out of 8) Commissural region Body of leaflets Both Pannus (out of 30 valves) Present Absent Location of Pannus (out of 27) On flow surface On nonflow surface Both Cannot determine Lipid insudation Absent Present Cusp tears (out of 30 valves) None Single tears Multiple tears Unable to determine Type 1 tears Type 2 tears Type 3 tears Type 4 tears Types 1 and 2 Type 1, 3 Type 2, 3
8 22 1 5 2 27 3 13 12 12 2 15 15 8 8 12 2 11 2 1 1 2 2 1
Fig. 2. Inflow (A) and outflow (B) surfaces of a T-SPV from a 56-year-old male at 9 years. (Panel A) Inflow surface shows pannus present all around the circumference of the valve (white arrows). One cusp shows a large tear with calcification around the margins. (Panel B) The outflow surface shows a ridge of calcification (arrowheads) close to one commissure.
44
J. Butany et al. / Cardiovascular Pathology 15 (2006) 41 – 48
3.3. Calcification Eight of the explanted valves (26.7%) showed evidence of calcification on gross inspection and radiologic examination. The mean age of the grossly calcified valves was 117.4F24.5 months while the valves that grossly showed no calcification were in situ for a significantly shorter duration 88.5F34.4 months ( P=.027, two-tailed Mann–Whitney test). Calcific nodules were noted in the body of the cusp in five cases (Figs. 1 and 4), predominantly at the base of the cups. One showed nodular calcification in the commissural region only, while two valves showed gross calcification involving both the body and base of the cusps. Three of the valves showed gross calcification of the cusps directly associated with cusp tears. On histological examination, a total of 23 valves (76.6%) showed at least one microfocus of calcification in cusp tissue. Microfoci of calcification were seen at the base of the cusps in almost all cases; in addition, six showed calcification of cusp tissue in the commissural region while one case showed a single small microfocus of calcification at the tip of a cusp. Eight cases (26.7%) showed calcification in the porcine aortic tissue (Fig. 4), while another four valves showed calcification of the muscle shelf in the porcine right coronary cusp. One
Fig. 3. A T-SPV from a 72-year-old male explanted for bioprosthesis incompetence at 119 months. (Panel A) The inflow surface shows intact cusps and pannus, which extends onto the base of the cusps and areas where the pannus was lost during surgical explantation (white arrows). (Panel B) The outflow surface shows intact cusps and mild pannus in the commissural regions. The commissural regions are mildly splayed outward. (Panel C) Lateral view of the bioprosthesis shows the intact fabric and cusps. Pannus is seen. This belies the feeling that explantation of these prostheses would be difficult to impossible.
addition, two valves showed areas of inflammatory cell infiltration into the leaflets though neither one of these cases were associated with infectious microorganisms.
Fig. 4. (Panel A) Histological sections of a T-SPV explanted from a 46-yearold male at 103 months showing well-preserved tissue with microfocal areas of calcification. (Panel B) Transverse section at a commissure shows microfoci of calcification of the cusps (arrows) and of the porcine aortic cuff (asterisk).
J. Butany et al. / Cardiovascular Pathology 15 (2006) 41 – 48
45
Fig. 6. A T-SPV explanted from a 72-year-old male at 112 months shows cusp tears (white arrows).
significant difference in the ages of patients who showed cusp tears and the patients with no cusp tears. Six valves that showed the presence of calcification on gross examination also showed the presence of cusp tears, while 19 of 23 valves showing microfoci of calcification also showed cusp tears. 3.5. Pannus
Fig. 5. Histological sections of a T-SPV explanted from a 72-year-old male at 125 months. Panel A shows inflammatory infiltrate around the porcine aortic tissue. (Panel B) Higher magnification of the boxed area from panel A shows mixed inflammatory infiltrate including multinucleated giant cells [hematoxylin and eosin, original magnification 25 (A), 200 (B)].
Pannus was visible grossly on the surface of 27 of 30 valves (90.0%). Pannus was present grossly on inflow (ventricular) surface alone in 13 valves, while 12 valves showed the gross presence of pannus on both the inflow and outflow (aortic) surfaces (Fig. 7). In 2 cases, the pannus was received detached from the valve, and from its shape it is most likely that it had been detached from the valve ring, that is the inflow surface of the bioprosthesis. Histologically, at least some degree of pannus was observed on both the inflow and outflow (aortic) surface of all but two valves. Surprisingly, one of the valves that did not show pannus had been in
valve showed an inflammatory reaction to the porcine aortic tissues (Fig. 5). Ten of the 15 valves showing lipid insudation also showed at least one focus of calcification. 3.4. Tears and perforations Twenty valves (66.6%) showed cusp tears. These valves had been in place for an average of 107.2F24.8 months. Eight valves showed only one tear while 12 showed multiple tears affecting multiple cusps. Two valves were explanted and received in several pieces, making it difficult to determine the presence or absence of tears and perforations. Of the 20 valves, 11 showed Type 1 tears located predominantly in the commissural region (Fig. 6). Two valves showed Type 2 tears, one valve showed a Type 3 tear and one a Type 4 tear. Two valves showed both a commissural region Type 1 tear and a Type 2 tear, two showed a Type 1 tear and a Type 3 tear (perforation) in the same valve, while one showed the presence of a Type 2 and Type 3 tear in the same valve. In contrast, the eight valves (33.3%) that did not show any tears had been in place for between 1 and 120 months with a mean implant duration of 77.8F46.3 months. There was no
Fig. 7. T-SPV explanted at 104 months from a 58-year-old male. Longitudinal sections of the cusp and porcine aortic tissues show significant pannus on the inflow (black arrow) and outflow (white arrow) surfaces. Focal calcification is seen close to the cusp junction with the pannus (arrowhead).
46
J. Butany et al. / Cardiovascular Pathology 15 (2006) 41 – 48
place for 111 months (it is possible that the pannus had gotten detached during explantation and was not submitted for pathological examination), while the other had been in place for less than 1 month. Four cases (13.3%) showed pannus on the base of the cusp, which extended onto the cusp tissue for a significant distance. Fourteen cases showed calcification of pannus. In these cases, the pannus would likely have contributed to valve dysfunction. In none of the cases was pannus the only clinical indication for explantation. In the one case where the indication for T-SPV explantation was stenosis, there were extensive cusp calcification, tears and pannus seen at 135 months postimplantation in a 44-year-old male. In addition, perivalvular (surrounding the fabric) fibrosis and pannus were present in 21 (70.0%) valves. Seven cases showed the deposition of significant fibrous tissue between the porcine aorta and the Dacron graft covering the external surface of the prosthesis. Nineteen cases showed host tissue growth around the porcine aorta, in the area between the Dacron fabric and the host aorta. In one case, the presence of pannus on the nonflow surfaces was associated with small amounts of thrombus, suggesting the possibility of a perivalvular leak. Interestingly, six cases showed host tissue infiltration and fibrous tissue deposition within the interstices of the synthetic material with 3 of these cases showing calcification within the Dacron fabric and the fibrous tissue. All cases examined showed some element of host inflammatory reaction to the synthetic material. This reaction was relatively mild in 9 (30%) of cases. Immunohistochemical stains confirmed that this host reaction was chronic in nature and was composed primarily of macrophages, multinucleate giant cells and lymphocytes. Small areas of thrombus were visible on 11 (36.7%) of the valves, mainly at the base of the cusps. In 10 of these cases, the thrombus had formed on the sinus surface of the cusps, while in the remaining, thrombus had formed in the region of the commissures superimposed on existing pannus. 3.6. Infective endocarditis Four patients developed infective endocarditis requiring explantation of their T-SPV. The mean patient age at explantation was 59.3F15.4 years and these valves had been in place for a mean of 47.5F39.9 months, but this included the one patient who developed IE within 1 month of implantation. One patient showed the presence of Gram-positive microorganisms, another showed a coagulase-negative streptococcus, a third case showed small microorganisms, possibly Bordatella while in the remaining case no microorganisms could be demonstrated on histology. The explanted valves showed minor structural valve deterioration (SVD), none showed any gross evidence of calcification and only one showed a microfocus of calcification histologically. However, three of the four cases showed extensive pannus with extension onto the cusps resulting in cusp thickening and stiffening, which made it difficult for the cusps to move.
4. Discussion Since 1991, the T-SPV has been widely used in a number of institutions with studies documenting excellent hemodynamic performance and durability [3,4]. Since then, it has been used worldwide with equally good results. The T-SPV shows a low transvalvular gradient, large effective orifice area, decreased left ventricular mass index in patients receiving the valve and possibly improved patient survival, compared with stented valves [5]. A follow-up study of 332 patients by David et al. showed freedom from SVD at 10 years to be 99% in patients greater than 65 years of age and 77% in patients under 65 years of age [8]. In the current study, the mean patient age at the time of T-SPV implant was only 54.0F11.4 years, which likely accounted for the high incidence of SVD. The T-SPV has been quite durable in comparison to other prosthetic valves such as the Hancock II stented porcine and the Carpentier–Edwards stented porcine [5,9,10]. Of the explanted valves examined in our study, 27 (90%) showed SVD. SVD has been observed as a major cause of failure in other prosthetic valves. In a similar study, Butany et al. [7,11] observed an incidence of SVD of approximately 71% for the Ionescu–Shiley low-profile pericardial prosthesis, which had been in place for 2–5 years, and 56% of the Hancock II bioprosthesis, which had been in place for a mean of 5.1 years. In contrast, the T-SPV examined in this study had a mean implant duration of 96 months (8.0 years), including one valve explanted after only 1 month due to infective endocarditis. Our results therefore support earlier studies that found only minimal tissue degeneration in explanted T-SPV at up to six years [12]. Since the T-SPV is implanted in the subcoronary position, proper valve function is dependent on the native aorta. Previous investigations of the T-SPV describe a strong correlation between dilation of the sinotubular junction (STJ) and the occurrence of aortic incompetence [13,14] T-SPV failure has been related to dilation of the ascending aorta in the absence of SVD [15]. Although no measure of aortic dilation was collected in our study, 12 patients had native bicuspid AV disease, a known cause of aortic root dilation [16]. Eleven of the 12 T-SPVs explanted from patients with native BAV valves showed some evidence of SVD with four valves showing significant SVD with gross evidence of calcification and multiple cusp tears. In the remaining valves, SVD was mild and the cusp tears may well have been secondary to STJ dilation. Only eight valves in our study showed gross evidence of calcification. These valves had been in place for a mean of 9.75 years (117 months), which is higher than the mean duration of implantation for other valves in reported studies [7,11,17,18]. A majority of the valves did not show gross calcification, which supports the findings from previous studies at up to 6 years postimplantation [5,12]. However, earlier case reports had suggested that calcification of the T-SPV can occasionally begin early. Butany and David [19]
J. Butany et al. / Cardiovascular Pathology 15 (2006) 41 – 48
reported a case of a T-SPV, which showed evidence of calcification 32 months postimplant. In addition, Tansley et al. [20] reported a case of calcific stenosis in a T-SPV 6 years postimplant. Recent studies using subdermal implants in animal models showed high levels of calcium (165 mg/g dry tissue weight in leaflets and 94.6 mg/g dry weight in aorta) in the T-SPV compared with other stented and stentless prosthesis [21]. Evidence suggests that the valve fixation process may contribute to the process of calcification. The extent of calcification is reported to be directly dependent on the amount of glutaraldehyde present in the tissue [22]. Ultrastructurally, T-SPV implanted subdermally showed a clear alteration in their surface superstructure, with loss of endothelium and disruption of the helical structure of collagen bundles [21]. Fixation under pressure can lead to loss of endothelial cells and alterations in 3D structure of collagen, which can lead to disruption of fibers under strain [23]. These disrupted collagen fibers can act as nucleation sites for calcification [21]. Interestingly, while only eight T-SPVs in our study showed gross evidence of calcification, 23 valves (76.6%) showed at least one microscopic focus of calcification in cuspal tissue. Almost all cases showed calcification in the base of the cusps. Of the 23 valves that showed microscopic calcification, 19 showed tears. In a similar study of explanted Hancock II bioprosthesis, it was observed that almost 80% of aortic prostheses with evidence of calcification also showed tears [7]. These observations suggest that even small foci of calcification may alter the mechanical stress on the leaflet tissue and predispose the cusp to tears. It has been suggested that extracellular lipid droplets can act as nucleation sites for calcification as lipid droplets and areas of leaflet calcification show a similar pattern of distribution [24]. It is interesting to note that 10 of the 15 valves in this study, which showed gross evidence of lipid insudation, also showed microscopic evidence of calcification. Two thirds of T-SPVs examined in this study showed evidence of cusp tear(s). Type 1 tears were the most common and usually affected the commisural regions. Our results are similar to other studies where examination of the Hancock II bioprosthesis showed tears in 53% of valves, with Type 1 tears being the most common [7]. Examination of the porcine aorta showed calcification in the porcine aortic tissue of 12 valves. Interestingly, 10 valves showed the shelf of porcine muscle at the right coronary cusp base, and in four of these valves, the muscle shelf showed evidence of calcification while one showed an inflammatory reaction to the porcine tissue. Calcification of the porcine wall, though relatively rare, can have significant effects on hemodynamics and stiffening of the aortic root [25]. In addition to calcification, significant pannus was observed on the sinus surface of the valve in all but 2 cases. Pannus is considered to be the deposition of exuberant host tissue on the surface of grafts. It is composed primarily of fibroblasts and collagen. In moderate amounts, pannus can be beneficial as it covers the suture lines allowing a nonthrombogenic surface to form
47
between the graft and the host aorta [19]. However, excessive pannus can result in stenosis, cusp thickening and stiffening, which can hinder the mobility of the leaflets [1,12]. The valves in the series that showed significant pannus were mainly removed for AI or IE. These valves also showed calcification and/or cusp tears. Hence, while the significant degree of pannus would have contributed to valve dysfunction, it is not possible to say whether valve failure was due to pannus alone. In the present study, one valve showed significant pannus on the outflow surface of the aorta extending onto the cusp such that the flexion point would likely have changed. Such a change would have contributed to the significant tears observed in that valve. Pannus was also observed outside the T-SPV, both between the porcine aorta and the Dacron graft and between the Dacron graft and the host tissues. Interestingly, 14 cases showed calcification of the pannus on the sinus surface of the porcine aorta. In most cases, the surgeon was able to explant the bioprosthesis, without significant damage to it, using the plane around the Dacron fabric covering the bioprosthesis. The above changes in the porcine aorta are significant for the design of stentless valves as current antimineralization pretreatments, while effective at reducing cusp calcification, do not reduce calcification of the porcine aorta [26]. In fact, pretreatments that remove cells and collagen from the aorta do not reduce calcification of the porcine aorta, whereas degradation and fragmentation of elastin appear to cause increased calcification [26]. Matrix metalloproteinases are a family of enzymes that can be secreted by a number of different cell types, including fibroblasts [27,28], and can degrade components of the extracellular matrix. Studies have shown that MMP-mediated collagen degradation is an initial step in elastin calcification [29]. From these data it appears possible that fibroblasts present in pannus may play a role in promoting calcification of the porcine aorta, possibly by secreting proteolytic enzymes which damage elastin in the porcine aorta.
5. Conclusion In our study of 30 explanted T-SPVs, 90% showed structural deterioration at a mean implant duration of 96.2 months. Eight valves showed gross evidence of calcification, though 23 showed microscopic foci of calcification in the base of the leaflets, in the commissural regions in the pannus and in the porcine aortic tissue. All implanted prostheses showed pannus. The higher incidence of pannus on the T-SPV as compared with stented valves [7,11] is likely related to the absence of the large sewing cuff, which separates the valve itself from the host tissue. A new mode of failure was seen in two cases, wherein the cuspal tissue was largely intact, though the bioprosthesis had become functionally incompetent due to dilatation of the sinotubular junction. This is likely related to the dilatation of the native ascending aortic tissue in a patient with a congenitally
48
J. Butany et al. / Cardiovascular Pathology 15 (2006) 41 – 48
abnormal AV. The T-SPV, an aldehyde-fixed bioprostheses, provides excellent long-term clinical results; however, it appears to have the deficiencies of all other glutaraldehydefixed porcine valves. Further analyses of larger series of explanted bioprostheses are essential.
References [1] Butany J, Leask R. The failure modes of biological prosthetic heart valves. J Long-Term Eff Med Implants 2001;11:115 – 35. [2] Park SZ, Reardon MJ. Current status of stentless aortic xenografts. Curr Opin Cardiol 2000;15:74 – 81. [3] David TE, Feindel CM, Scully HE, Bos J, Rakowski H. Aortic valve replacement with stentless porcine aortic valves: a ten-year experience. J Heart Valve Dis 1998;7:250 – 4. [4] Goldman B, Christakis G, David T, Rakowski H, Bach D, Wood J, Goldman S, Pepper J, Yacoub M, Verrier E, Petracek M. Will stentless valves be durable? The Toronto valve (TSPV) at 5 to 6 years. Semin Thorac Cardiovasc Surg 1999;11:42 – 9. [5] Bach DS, Goldman B, Verrier E, Petracek M, Wood J, Goldman S, David TE. Durability and prevalence of aortic regurgitation nine years after aortic valve replacement with the Toronto SPV stentless bioprosthesis. J Heart Valve Dis 2004;13:64 – 72 [discussion 72]. [6] Ishihara T, Ferrans VJ, Boyce SW, Jones M, Roberts WC. Structure and classification of cuspal tears and perforations in porcine bioprosthetic cardiac valves implanted in patients. Am J Cardiol 1981;48:665 – 78. [7] Butany J, Yu W, Silver MD, David TE. Morphologic findings in explanted Hancock II porcine bioprostheses. J Heart Valve Dis 1999;8: 4 – 15. [8] David TE, Feindel CM, Bos J, Ivanov J, Armstrong S. Ten-year results of aortic valve replacement with Toronto SPV bioprosthesis. Presented at the 84th Annual Meeting of the American Association for Thoracic Surgery, Toronto, Canada, April 25–28, 2004. [9] Banbury MK, Cosgrove DM, Lytle BW, Smedira NG, Sabik JF, Saunders CR. Long-term results of the Carpentier–Edwards pericardial aortic valve: a 12-year follow-up. Ann Thorac Surg 1998;66:S73 – 6. [10] David TE, Ivanov J, Armstrong S, Feindel CM, Cohen G. Late results of heart valve replacement with the Hancock II bioprosthesis. J Thorac Cardiovasc Surg 2001;121:268 – 77. [11] Butany J, Vanlerberghe K, Silver MD. Morphologic findings and causes of failure in 24 explanted Ionescu–Shiley low-profile pericardial heart valves. Hum Pathol 1992;23:1224 – 33. [12] Butany J, de Sa M, Feindel CM, David TE. The Toronto SPV bioprosthesis: review of morphological findings in eight valves. Semin Thorac Cardiovasc Surg 1999;11:157 – 62. [13] David TE, Ivanov J, Eriksson MJ, Bos J, Feindel CM, Rakowski H. Dilation of the sinotubular junction causes aortic insufficiency after aortic valve replacement with the Toronto SPV bioprosthesis. J Thorac Cardiovasc Surg 2001;122:929 – 34. [14] Shargall Y, Goldman B, Christakis G, David T. Analysis of explants and causes of mortality during long-term follow-up of the Toronto stentless porcine valve. Semin Thorac Cardiovasc Surg 2001;13:106 – 12.
[15] Butany J, Ahn C, Leask RL, Graba J, David T, Ahluwalia MS. Stentless porcine valves: new mode of failure. Can J Cardiol 2003;19: 185 – 6. [16] Keane MG, Wiegers SE, Plappert T, Pochettino A, Bavaria JE, Sutton MG. Bicuspid aortic valves are associated with aortic dilatation out of proportion to coexistent valvular lesions. Circulation 2000;102: III35 – 9. [17] Bortolotti U, Milano A, Mazzaro E, Thiene G, Talenti E, Casarotto D. Hancock II porcine bioprosthesis: excellent durability at intermediateterm follow-up. J Am Coll Cardiol 1994;24:676 – 82. [18] Bottio T, Thiene G, Pettenazzo E, Ius P, Bortolotti U, Rizzoli G, Valfre C, Casarotto D, Valente M. Hancock II bioprosthesis: a glance at the microscope in mid-long-term explants. J Thorac Cardiovasc Surg 2003;126:99 – 105. [19] Butany J, David TE. Pathology of explanted stentless porcine valves (human experience). In: Gabbay S, editor. New horizons and the future of heart valve prosthesis. Austin7 Silent Partners Inc, 1994 pp. 183 – 9. [20] Tansley PD, Sheppard MN, Pepper J. Symptomatic calcific stenosis of a Toronto stentless porcine valve. Eur J Cardiothorac Surg 2000;17: 763 – 5. [21] Mirzaie M, Meyer T, Schorn B, Schwartz P, Baryalei M, Rastan A, Lotfi S, Dalichau H. Calcification tendency of various biological aortic valves in an experimental animal model. Cardiovasc Surg 1999; 7:735 – 41. [22] Golomb G, Schoen FJ, Smith MS, Linden J, Dixon M, Levy RJ. The role of glutaraldehyde-induced cross-links in calcification of bovine pericardium used in cardiac valve bioprostheses. Am J Pathol 1987; 127:122 – 30. [23] Broom ND, Thomson FJ. Influence of fixation conditions on the performance of glutaraldehyde-treated porcine aortic valves: towards a more scientific basis. Thorax 1979;34:166 – 76. [24] Dunmore-Buyze J, Boughner DR, Macris N, Vesely I. A comparison of macroscopic lipid content within porcine pulmonary and aortic valves. Implications for bioprosthetic valves. J Thorac Cardiovasc Surg 1995;110:1756 – 61. [25] Schoen FJ, Levy RJ. Founder’s Award, 25th Annual Meeting of the Society for Biomaterials, perspectives. Providence, RI, April 28–May 2, 1999. Tissue heart valves: current challenges and future research perspectives. J Biomed Mater Res 1999;47:439 – 65. [26] Bailey MT, Pillarisetti S, Xiao H, Vyavahare NR. Role of elastin in pathologic calcification of xenograft heart valves. J Biomed Mater Res 2003;66A:93 – 102. [27] Shi Y, Patel S, Niculescu R, Chung W, Desrochers P, Zalewski A. Role of matrix metalloproteinases and their tissue inhibitors in the regulation of coronary cell migration. Arterioscler Thromb Vasc Biol 1999;19:1150 – 5. [28] Rabkin E, Aikawa M, Stone JR, Fukumoto Y, Libby P, Schoen FJ. Activated interstitial myofibroblasts express catabolic enzymes and mediate matrix remodeling in myxomatous heart valves. Circulation 2001;104:2525 – 32. [29] Vyavahare N, Jones PL, Tallapragada S, Levy RJ. Inhibition of matrix metalloproteinase activity attenuates tenascin-C production and calcification of implanted purified elastin in rats. Am J Pathol 2000;157: 885 – 93.