Partial replacement of the tricuspid valve using cryopreserved tricuspid homograft: 20-year outcomes Authors: Samad Raza, Bishwo Shrestha, Fiona Doig, Peter Pohlner, Homayoun Jalali, Rishendran Naidoo Published: 2025 ABSTRACT AbstractOBJECTIVESThere is limited consensus on the optimal strategy for surgical treatment of severe tricuspid valve pathology. At our institution, we have implemented a unique technique of using a tricuspid homograft with its supporting apparatus for partial replacement of the native tricuspid valve. We now present the long-term outcomes of these patients.METHODSWe analysed a cohort of patients who underwent partial tricuspid valve replacement using tricuspid homograft. Yearly clinical and echocardiographic follow-up was performed. Fine–Gray methods were used to estimate freedom from death and reoperation and reverse Kaplan–Meier methods were used to calculate follow-up.RESULTSFourteen patients were included (age range 15 days to 73 years). Indications included congenital anomalies (n = 9, 64%) and infective endocarditis (n = 5, 36%). The median follow-up was 17 years (95% confidence interval (CI) 10–21 years). Two patients (14%) died due to causes unrelated to the primary tricuspid valve surgery, and three (21%) underwent redo tricuspid valve operations. In the remaining cohort, seven (50%) were asymptomatic and two (14%) reported class II dyspnoea, while none had severe tricuspid regurgitation on echocardiogram. Estimated freedom from death was 93% at 10 years and 83% at 15 and 20 years, while estimated freedom from reoperation was 77% at 10, 15 and 20 years.CONCLUSIONSPartial replacement of the tricuspid valve using tricuspid homograft tissue effectively restores the anatomical conformity of the native tricuspid valve and has durable long-term survival and freedom from severe tricuspid regurgitation in patients with congenital anomalies and infective endocarditis. INTRODUCTION Tricuspid valve replacement has several limitations. Biological valves have limited durability and growth potential, while mechanical valves have numerous issues such as the requirement for anticoagulation and high rate of conduction system injury [1–3]. There is therefore a need to explore alternative management strategies for severe tricuspid valve disease. A novel technique of partial replacement of the tricuspid valve using homograft tissue for severely deformed or damaged native tricuspid valves was applied by two surgeons at our institution between 1997 and 2010. While the use of mitral valve homograft tissue for tricuspid valve replacement has been described in several studies [4–9]. The use of tricuspid homograft tissue has been limited to sporadic cases [10–12]. Our cohort of 14 cases is the largest case series reported so far. This paper presents the surgical technique of tricuspid homograft implantation and its long-term results with over 20 years of follow-up, serving as a complement to the mid-term outcomes published in 2009 [13]. Long-term follow-up of these patients included assessment of clinical status, valve function and requirement for reintervention. We hypothesized that partial replacement of the tricuspid valve with tricuspid homograft tissue would result in long-term preservation of valvular function while limiting the rate of adverse complications such as pacemaker insertion and reinfection. PATIENTS AND METHODS Study design This study was a cohort study involving a prospectively collected database which was retrospectively reviewed for patients who underwent partial tricuspid valve replacement using tricuspid valve homograft tissue between 1992 and 2010 at The Prince Charles Hospital in Brisbane, Australia. Follow-up was closed on 19 February 2024. Surgical technique Tricuspid homograft tissue (Fig. 1) is available from donor hearts harvested for aortic or pulmonary homografts. Compared to mitral homografts, tricuspid homografts have three leaflets that correspond closely to the tissue characteristics of the native recipient valve, along with subvalvular chordae and papillary muscles that provide intrinsic leaflet support. Although the harvesting and cryopreservation technique is not significantly different from that of aortic and mitral homografts, valve bank staff require extra training to process tricuspid homograft tissue. Partial replacement of tricuspid valve using homograft. (A) Tricuspid valve homograft includes all its leaflets and associated papillary muscles which are prepared as per standard cryopreservation technique. (B) The homograft papillary muscle was attached to free wall and/or native papillary muscle of the right ventricle using polypropylene monofilament mattress suture buttressed with pledgets. (C) Subsequently, the homograft leaflet was sewn into the TV annulus by continuous polypropylene monofilament suture. Note—image in A taken by author BS, Image B and C reproduced under license from Beth Croce Due to the limited availability of tricuspid homograft, the decision to use tricuspid homograft tissue was made on a case-by-case basis by the operating surgeon and was based on the degree of valve destruction whereby standard repair techniques were unlikely to restore the functionality of the valve. Figure 2 provides a schema for this decision-making process. Decision-making schema for selecting tricuspid homograft for partial valve replacement The following are principles utilized by the authors for partial tricuspid valve replacement using homograft tissue: Identify the pathology of the tricuspid valve and aim to preserve as much native leaflet tissue as possible. The surgical approach is tailored to the individual pathology. Identify the pathology of the tricuspid valve and aim to preserve as much native leaflet tissue as possible. The surgical approach is tailored to the individual pathology. Measure the size and location of the tricuspid valve annulus. In adults, the aim is to bring the annulus size to 28–30 mm and to its normal anatomical position; there might be a need for ring or De Vega annuloplasty in the case of a dilated annulus or replacement of more than two leaflets with homograft tissue. Delamination and mobilization of leaflet tissue may be possible in the case of an Ebstein tricuspid valve, particularly both anterior and posterior tricuspid leaflets. Measure the size and location of the tricuspid valve annulus. In adults, the aim is to bring the annulus size to 28–30 mm and to its normal anatomical position; there might be a need for ring or De Vega annuloplasty in the case of a dilated annulus or replacement of more than two leaflets with homograft tissue. Delamination and mobilization of leaflet tissue may be possible in the case of an Ebstein tricuspid valve, particularly both anterior and posterior tricuspid leaflets. In the event of tricuspid leaflet deficiency, select available homograft leaflet tissue, based upon the deficiency of the native leaflets. In the event of tricuspid leaflet deficiency, select available homograft leaflet tissue, based upon the deficiency of the native leaflets. The anterior leaflet of the tricuspid homograft is most commonly used, due to its large size with broad area of coaptation. The posterior leaflet is the second most commonly used homograft leaflet, whereas the septal leaflet was never used due to its small and rudimentary size. The donor is typically older than the recipient and therefore there is ample leaflet tissue available for use; it is normally cut to size. The anterior leaflet of the tricuspid homograft is most commonly used, due to its large size with broad area of coaptation. The posterior leaflet is the second most commonly used homograft leaflet, whereas the septal leaflet was never used due to its small and rudimentary size. The donor is typically older than the recipient and therefore there is ample leaflet tissue available for use; it is normally cut to size. Trim the homograft leaflet attachment at the right atrial leaflet hinge tissue, to provide an annulus of 3–4 mm at the leaflet hinge. Trim the homograft leaflet attachment at the right atrial leaflet hinge tissue, to provide an annulus of 3–4 mm at the leaflet hinge. Arrange the corresponding papillary muscle and chordal apparatus of the donor homograft leaflet to be used and identify a fixation point in the right ventricle depending upon the length and orientation of the donor chordal apparatus. A chordae suture ruler can be used to measure the length of the chordal apparatus. Attach the papillary muscle to the right ventricle using a 6-0 prolene suture with 3 mm Teflon pledgets (Fig. 1). In some patients, the donor papillary muscle is inserted directly into the recipient papillary muscle; however, if there is a discordance in length of the subvalvular apparatus, the homograft should be secured to the right ventricular free wall or septum to preserve the leaflet’s natural shape and function. Arrange the corresponding papillary muscle and chordal apparatus of the donor homograft leaflet to be used and identify a fixation point in the right ventricle depending upon the length and orientation of the donor chordal apparatus. A chordae suture ruler can be used to measure the length of the chordal apparatus. Attach the papillary muscle to the right ventricle using a 6-0 prolene suture with 3 mm Teflon pledgets (Fig. 1). In some patients, the donor papillary muscle is inserted directly into the recipient papillary muscle; however, if there is a discordance in length of the subvalvular apparatus, the homograft should be secured to the right ventricular free wall or septum to preserve the leaflet’s natural shape and function. The homograft leaflet is then sutured to the recipient’s annulus with a continuous 6-0 prolene suture except in the region of the conduction system where interrupted sutures are used. Depending on the observed coaptation surfaces of the leaflet tissue, it can be useful to conjoin the native and homograft leaflet tissue at the basal portion with one or two interrupted 6/0 prolene sutures. The homograft leaflet is then sutured to the recipient’s annulus with a continuous 6-0 prolene suture except in the region of the conduction system where interrupted sutures are used. Depending on the observed coaptation surfaces of the leaflet tissue, it can be useful to conjoin the native and homograft leaflet tissue at the basal portion with one or two interrupted 6/0 prolene sutures. Every attempt should be made to achieve a 5–8-mm coaptation zone at the free margins of the leaflets. Every attempt should be made to achieve a 5–8-mm coaptation zone at the free margins of the leaflets. After completion of the tricuspid valve homograft replacement, the valve is then tested with water with the heart beating. Trivial to mild TR was accepted. After completion of the tricuspid valve homograft replacement, the valve is then tested with water with the heart beating. Trivial to mild TR was accepted. Data collection and statistical methods Patients’ medical records were reviewed, and yearly clinical and echocardiographic follow-up was performed. Outcomes analysed included New York Heart Association (NYHA) functional status, echocardiographic valve function, reoperation, and mortality. Fine–Gray competing risk analyses were used to estimate freedom from reoperation and overall survival, while reverse Kaplan–Meier methods were used to calculate median follow-up in the entire cohort. There was no missing data at baseline. Ethics The study was approved by the institutional ethics committee on 29 August 2008 (approval number EC2873), prior to data collection for the mid-term outcomes, and utilized anonymized data that was stored in a secure database. Individual informed consent was waived. RESULTS Fourteen patients have undergone partial replacement of the tricuspid valve using tricuspid homograft at TPCH. Since the publication of our series of mid-term outcomes of the technique in 2009 [13], one further patient underwent this procedure. This 20-year-old female had a history of Ebstein malformation and pulmonary stenosis, having previously undergone pulmonary valve replacement with a pulmonary homograft, tricuspid valve repair and closure of an atrial septal defect as a child. She re-presented with severe tricuspid and pulmonary regurgitation and underwent a partial tricuspid valve replacement utilizing tricuspid homograft tissue and a 36 mm Physio annuloplasty ring and right ventricular outflow tract reconstruction with pulmonary homograft (Table 1). Table 1:Baseline characteristics and follow-upPatientAge (at time of surgery)SexPathologyAssociated pathologySurgeryFollow-up144 yearsFActive IESeptic pulmonary emboliAnterior and posterior TV leaflet replacement with TV HG21 years274 yearsMActive IE (recurrent)Previous vegetectomy and TV repair, moderate PAH, coronary artery diseaseSeptal leaflet repair with TV HG and De Vega annuloplasty. + CABG X 210 years until death361 yearsMActive IESeptic PE poor LV functionAnterior and posterior TV leaflet replacement with TV HG2 years until lost to follow-up467 yearsFActive IESeptic pulmonary emboli, MV vegetation, moderate PAHSeptal leaflet repair with TV HG and De Vega annuloplasty + tissue MVR2 years until death541 yearsMActive IESeptic pulmonary emboli, right heart failure with ascitesAnterior leaflet repair with TV HG and 35 mm Duran ring13 years659 yearsFEbsteinNoneSeptal leaflet repair with TV HG and 27 mm Duran ring2 years until redo TVR78 monthsFEbsteinSevere RV dysfunctionSeptal leaflet replacement with TV HG, RV plication and bidirectional Glenn shunt23 years830 yearsMEbsteinNoneSeptal leaflet replacement with TV HG and 28-mm Duran ring20 years924 yearsMEbsteinLeft main coronary artery diseaseSeptal leaflet replacement with TV HG and Edwards MC3 30 mm band + CABG x 11 year until lost to FU102 weeksMAV septal defectSevere mitral regurgitationSeptal leaflet replacement with TV HG, MV repair and VSD closure17 years114 monthsMAV septal defectTrisomy 21, previous surgery for AV septal defectSeptal leaflet replacement with TV HG17.5 years127 yearsFTOFModerate pulmonary regurgitation after previous TOF repairAnterior leaflet replacement with TV HG and Cosgrove-Edwards 28 mm ring annuloplasty + PV repair7 years until TVR + pulmonary homograft replacement1315 monthsMSevere TR pulmonary atresia with intact ventricular septumSevere PAH after previous palliative surgeryAnterior leaflet replacement with TV HG + ASD closure, removal of central shunt5 years until redo TV repair + pulmonary homograft replacement1420 yearsFEbsteinSevere PS repaired with pulmonary allograft, previous TV repair and ASD repairRedo TV anterior leaflet replacement with TV HG and 36 mm Physio annuloplasty ring, RVOT reconstruction with pulmonary allograft11 years until lost to follow-upASD, atrial septal defect; AV, atrioventricular; CABG, coronary bypass graft; FU, follow-up; HG, homograft; IE, infective endocarditis; LV, left ventricle; MV, mitral valve; PAH, pulmonary arterial hypertension; PE, pulmonary embolism; PS, pulmonary stenosis; PV, pulmonary valve; TOF, Tetralogy of Fallot; TV, tricuspid valve; MVR mitral valve replacement; RVOT, right ventricular outflow tract; TVR, tricuspid valve replacement; VSD, ventricular septal defect. Baseline characteristics and follow-up ASD, atrial septal defect; AV, atrioventricular; CABG, coronary bypass graft; FU, follow-up; HG, homograft; IE, infective endocarditis; LV, left ventricle; MV, mitral valve; PAH, pulmonary arterial hypertension; PE, pulmonary embolism; PS, pulmonary stenosis; PV, pulmonary valve; TOF, Tetralogy of Fallot; TV, tricuspid valve; MVR mitral valve replacement; RVOT, right ventricular outflow tract; TVR, tricuspid valve replacement; VSD, ventricular septal defect. The indications for surgery were as follows: five patients (36%) with infective endocarditis, five (36%) with Ebstein’s anomaly, two (14%) with AV septal defect, one (7%) with Tetralogy of Fallot and one (7%) with pulmonary atresia with intact ventricular septum. There were six females (43%) and eight males (57%) in our series. There were five paediatric patients (36%), age range from 2 weeks to 7 years (mean 22.3 months) and nine adult patients (64%), age ranged from 20 to 67 years (mean 45.8 years). Median follow-up was 17 years (95% confidence interval (CI) 10–21 years, maximum 23 years) and three patients (21%) were lost to follow-up, one in the endocarditis group at 2 years and two in the Ebstein group at 1 and 11 years. Of the 11 patients who completed follow-up, there were two deaths (18%). One patient (patient 2) died at 10 years due to aspiration pneumonia while the other (patient 4) died from multiorgan failure due to infective endocarditis of the concurrently implanted mitral valve prosthesis, not involving the tricuspid valve (Tables 2 and 3). Ten patients (91%) were in NYHA class I. Recurrence of greater than moderate tricuspid regurgitation occurred in two patients (18%): one with endocarditis (patient 3) and the other with Ebstein’s anomaly (patient 7). Both patients tolerated this well with NYHA class I and II symptoms, respectively. Patient 3 had been treated with Lisinopril 20 mg daily and patient 7 with Bisoprolol 5 mg daily. Table 2:Outcomes by pathologyPathologyMedian FU (years)Mortality (%)Reoperation on TV (%)TR > grade II (moderate)NYHA (pts)Endocarditis (5)102 (40)01 I (4)II (1) Ebstein (5)1101 (20)1 I (4)II (1) AV septal defect (2)17000I (2)Tetralogy of Fallot (1)701 (100)0III (1)PAIVS (1)501 (100)0I (1)Total10.52 (14)3 (21)2 I (11)II (2)III (I) AV, atrioventricular; FU, follow-up; NYHA, New York Heart Association; PAIVS, pulmonary atresia with intact ventricle septum; pts patients; TR, tricuspid regurgitation; TV, tricuspid valve. Outcomes by pathology I (4) I (4) II (1) II (1) I (4) I (4) II (1) II (1) I (11) I (11) II (2) II (2) III (I) III (I) AV, atrioventricular; FU, follow-up; NYHA, New York Heart Association; PAIVS, pulmonary atresia with intact ventricle septum; pts patients; TR, tricuspid regurgitation; TV, tricuspid valve. Table 3:Cause of death and redo surgeryPatient IDDiagnosisSurgical procedureAge at redo surgery/deathInterval period for redo surgery/deathCause of redo surgery/deathDeath 2Active IESeptal leaflet repair with TV HG and De Vega annuloplasty. + CABG X 277 years10 yearsAspiration pneumonia associated with gastric cancer 4Active IESeptal leaflet repair with TV HG and De Vega annuloplasty + MVR (tissue)69 years2 yearsIE of prosthetic MV, sepsis and multiorgan failureRedo surgery 6Ebstein anomalySeptal leaflet repair with TV HG and 27 mm Duran ring61 years2 yearsDehiscence of septal leaflet with severe TR 12TOFAnterior leaflet replacement with TV HG and Cosgrove-Edwards 28 mm ring annuloplasty + PV repair15 years7 yearsSevere PR, moderate TS with TR 2–3/4 13PAIVS and severe TRAnterior leaflet replacement with TV HG + ASD closure, removal of central shunt6 years5 yearsSevere TR and PRCABG, coronary artery bypass surgery; HG, homograft; ID, identification number; IE, infective endocarditis; MV, mitral valve; PAIVS, pulmonary atresia with intact ventricular septum; PR, pulmonary regurgitation; PV, pulmonary valve; TOF, Tetralogy of Fallot; TR, tricuspid regurgitation; TS, tricuspid stenosis; TV, tricuspid valve. Cause of death and redo surgery CABG, coronary artery bypass surgery; HG, homograft; ID, identification number; IE, infective endocarditis; MV, mitral valve; PAIVS, pulmonary atresia with intact ventricular septum; PR, pulmonary regurgitation; PV, pulmonary valve; TOF, Tetralogy of Fallot; TR, tricuspid regurgitation; TS, tricuspid stenosis; TV, tricuspid valve. Reoperations were required in three patients (27%). These occurred exclusively in patients with congenital pathology and the decision to re-operate on the tricuspid valve was influenced by the presence of pulmonary incompetence in two of these patients. Dehiscence of the implanted homograft occurred in one patient (patient 6), 2 years following the primary surgery. At reoperation, the previously inserted septal homograft leaflet had become detached at the base. Repair of the defect was attempted with bovine pericardium but competence of the valve could not be restored, and therefore the valve was replaced with a 29 mm Mosaic bioprosthesis. The excised homograft tissue in this patient demonstrated surface fibrin, fibrosis and neovascularization on histological analysis (Fig. 3). Aside from one patient requiring reoperation for mixed tricuspid stenosis and regurgitation (patient 12), no other patient had significant tricuspid stenosis on latest echocardiographic follow-up. None of the patients in our series required permanent pacemaker insertion. Histology of dehisced tricuspid homograft tissue. Microscopy of the 2-year-old implanted tricuspid homograft revealed surface fibrin, fibrosis and neo-vascularization Estimated freedom from death was 93% (95% CI 73–100%) at 10 years and 83% (95% CI 58–97%) at 15 and 20 years, while estimated freedom from reoperation was 77% (95% CI 53–94%) at 10, 15 and 20 years (Fig. 4). Cumulative subhazard for death (A) and reoperation (B) in overall cohort using the Fine–Gray method. The number at risk is shown in below the x-axis Three-dimensional echocardiography was performed to assess the tricuspid valve on one patient at 21 years of follow-up (Supplementary Videos 1 and 2). This patient had severe staphylococcal endocarditis involving the anterior and posterior leaflets and subvalvular apparatus. These were excised and replaced with tricuspid homograft tissue, while the native septal leaflet was preserved. Echocardiography showed mild tricuspid regurgitation. While there was mild tethering of the posterior leaflet, there was no calcification or tricuspid stenosis, with a mean transvalvular gradient of 1 mmHg. DISCUSSION The optimal surgical approach to tricuspid valve pathology remains controversial. While tricuspid valve replacement is less technically demanding than surgical repair, it has numerous disadvantages. Both mechanical and biological prostheses perform poorly in adults, with a 15-year reoperation-free survival of 62% and 53%, respectively [14], due to complications such as valve thrombosis and degeneration. Furthermore, valve prostheses are susceptible to reinfection, particularly in intravenous drug users [3], and pose a high risk of conduction system injury leading to heart block and permanent pacemaker insertion, which can be as high as 12.5% [15]. Given their lack of growth potential, tricuspid prostheses are poorly tolerated in the paediatric population, with an average time to redo surgery of 2.7 ± 2.6 years [16]. While simple tricuspid valvular pathologies can be repaired with primary techniques or reconstruction with pericardial patches [1], in complex valvular pathologies such as endocarditis resulting in total leaflet destruction or significant leaflet deficiency due to congenital malformation, traditional repair techniques may not restore the anatomical conformity of the valve. Our study has demonstrated that the use of homograft tissue is a valuable alternative in such patients. The long-term freedom from reoperation in our cohort is comparable to published literature on pericardial patch repair techniques [17]. Unlike pericardial patches, homograft tissue has superior tissue characteristics and is less susceptible to foreign body response and degenerative calcification [18–20]. It is particularly useful in young patients who are poor candidates for valve replacement due to the risk of long-term prosthesis-related complications. Furthermore, in comparison to pericardial patches, the presence of intrinsic subvalvular support enables the restoration of native valve geometry and optimizes haemodynamic performance by maintaining annuloventricular continuity. Additionally, homograft tissue is less thrombogenic and less susceptible to infection than prosthetic valves [21, 22], and there is no requirement for anticoagulation. The lack of postoperative heart block in our cohort is another advantage of this technique—a rim of native leaflet tissue adjacent to the triangle of Koch is preserved, enabling safe suturing of the homograft and avoiding direct instrumentation of the annulus. Most of the pre-existing literature on the use of homograft tissue for tricuspid valve replacement involves the use of mitral homografts. The technique was pioneered by Pomar et al. in 1993 in a series of three patients [4]. Later studies have demonstrated good functional outcomes and long-term survival; however, the follow-up duration has been limited in most studies [5–8], with only one case report citing follow-up >10 years [9]. There have only been three other reported cases on the use of tricuspid homograft [10–12], with a maximum follow-up of 9 months [10]. We have previously reported the largest series in the literature on the use of tricuspid homograft tissue rather than mitral homografts [13]. In our experience, the use of tricuspid valve leaflets, which are thinner and more pliable than mitral valve leaflets, enables us to more closely replicate the structure and function of the native valve. Since the publication of our mid-term results in 2010 [13], two patients in the cohort have died, due to causes unrelated to the tricuspid valve. In the patients requiring reintervention, we identified a technical failure resulting in dehiscence of homograft tissue (in an Ebstein’s anomaly case) and the presence of associated congenital anomalies and pulmonary regurgitation as potential contributors (TOF and PAIVS). This highlights a possible limitation of this technique in patients with associated pulmonary valve pathology. In our series, we noted the best results were achieved in Ebstein’s disease and infective endocarditis. An alternative technique for tricuspid valve restoration involves the use of pericardial and extracellular matrix cylinders to reconstruct the valve [18, 23]. While these techniques avoid the use of prosthetic material and are more easily accessible than homograft tissue, current data suggests a high rate of complications. In a series of 19 patients who underwent reconstruction with CorMatrix ECM cylinder, three patients experienced early mechanical dehiscence of the papillary muscle attachments or ECM cylinder seam [18]. In another series of seven patients who underwent reconstruction with a pericardial cylinder, three patients experienced degeneration and stenosis of the cylinder within six months of the index operation [23]. Our study has some limitations. We lost three patients (21.4% of the cohort) to follow-up in an otherwise small series. Additionally, this cohort study did not enrol a control group of patients undergoing repair of tricuspid valves using standard techniques or replacement with prosthetic valves. The uptake of tricuspid homograft replacement has been limited by several factors, including the small case numbers in previously published reports, limited availability of tricuspid homograft tissue in valve banks, and the technical complexity of the operation, with different techniques published by different authors [4–8]. Overall, the technique requires a high degree of surgical expertise and must be individualized to the patient’s pathology. It was only performed by two surgeons in our institution, each with several years of specialist congenital cardiac surgical experience and would require significant training to reproduce. Since 2010 the valve bank in our institution moved to a different location and the paediatric cardiac surgical department moved to a different hospital, limiting further use of tricuspid homografts at our institution. CONCLUSION The long-term results of partial tricuspid valve replacement with tricuspid homograft, while limited to a small number of patients, indicate that this is a viable technique in complex valvular pathology, especially infective endocarditis and Ebstein’s anomaly, and demonstrates good long-term survival and freedom from reoperation, comparable with standard repair techniques. Supplementary Material