Rates and Predictors of Structural Valve Degeneration and Failure of Trifecta Bioprosthetic Valve Over a 5-year Follow-up Period: A Single-center Experience

Abstract

Introduction:

The Trifecta bioprosthetic valve has been commonly used for surgical aortic valve replacement (SAVR). Multiple studies have been done to define the rate of structural valve degeneration (SVD) and failure (SVF), but the outcomes are still debatable. Therefore, we aim to conduct this single-center study to estimate the rate and predictors of SVD/SVF.

Methodology:

This retrospective observational cohort single-center study was conducted between 2014 and 2019 among Trifecta SAVR patients. Data were patient’s characteristics collected from electronic medical records at baseline and follow-up (3–5 years). Statistical analysis was performed with a significance level of P ≤ 0.05.

Results:

A total of 271 eligible patients were identified. Most of our sample were males (57.9%), with a mean age of 71.1 ± 10.6 years. The mean baseline preoperative ejection fraction (EF) was 53.0%, with no change (P = 0.88) in the immediate postoperative EF (53.6%). A most recent follow-up EF revealed a significant increase of EF (55.2%), P = 0.01. Furthermore, there was a significant increase from peak velocity to PV_MRE (mean difference [MD] ± standard error of mean (SEM) [0.15 ± 0.04], P < 0.01), an increase in pressure gradient (PG_IPE) to PG_MRE (MD ± SEM [1.70 ± 0.49], P < 0.01), and a decrease in Doppler velocity index (DVI_IPE) to DVI_MRE (MD ± SEM [−0.037 ± 0.01], P = 0.01). Regarding the SVF rate, 13 (4.8%) patients had failed valves requiring replacement throughout the study period.

Conclusions:

Over a 5-year follow-up period, 4.8% had SVF with an SVD of 23.2%, with the majority of SVD not being clinically significant except in six patients. These results corroborate with a previously published study suggesting a bad clinical outcome of Trifecta valve placement.

INTRODUCTION

The use of bioprosthetic aortic heart valves has increased steadily over the past two decades. The Society of Thoracic Surgeons database shows that the use of biomechanical aortic valves has increased from 43.6% in 1997 to 78.4%, whereas the use of mechanical valves has decreased from 49.9% in 1997 to 20.5% in 2006.[1] Mechanical valves are more durable but require lifelong anticoagulation to prevent thromboembolic disease, hence increasing the risk of bleeding complications, especially in elderly patients with multiple comorbidities.[2,3,4]

The Trifecta aortic pericardial valve (St Jude Medical, St Paul, Minn) was introduced into commercial use in 2010, with favorable results on the safety and performance of the valve with excellent safety profile and hemodynamic performance maintained through a 6-year follow-up period.[5] However, since its introduction, multiple studies have been published describing premature structural failure in Trifecta bioprosthetic valves.[6] A recently published single-center retrospective study of 2807 bioprosthetic aortic valves showed more valve failures in Trifecta valves compared to Perimount or Perimount Magna Ease.[7] Another retrospective review of 1058 patients who underwent aortic valve replacement showed that the rates of structural valve degeneration (SVD) of the Trifecta valve were more than expected compared with other bioprostheses, particularly in younger patients, with a cumulative incidence of SVD of 27.9%.[8] Furthermore, the Trifecta valve was associated with higher reintervention rates compared with Perimount valves in patients who underwent a surgical aortic valve replacement (SAVR).[9]

In this study, we conducted a single-center retrospective study as we aim to evaluate the echocardiographic changes of the aortic valve replacement using the Trifecta valve, rates of structural valve failure (SVF) at our center, and the possible contributing baseline characteristics to its prognosis.

METHODOLOGY

Study design and setting

This is a retrospective observational cohort single-center clinical study conducted from 2014 to 2019 among patients who had aortic valve replacement using the Trifecta valve prosthesis. A total number of 300 patients of either sex between the ages of 18 and 90 years who underwent SAVR utilizing the Trifecta valve in our tertiary center were initially included for screening purposes. We defined and classified variables and possible values before the data collection into four main categories: baseline characteristics, operative interventions, immediate postoperative echocardiographic findings, and finally the most recent follow-up outcomes. We excluded patients who were lost to follow-up, did not have a recent baseline echocardiogram or a follow-up within 3 years of the surgical intervention, or who died in the outpatient setting.

Structural valve degeneration definition

As there are various definitions for SVD and SVF in the literature, we utilized and modified the definition proposed by Fukuhara et al.[8] based on the available reported echocardiographic findings in our center. We classified SVD into two main categories, echocardiographic SVD (eSVD) and clinical SVD (cSVD). eSVD is defined as an increase in the mean pressure gradient (PG) across the valve by more than 20 mmHg with or without a concomitant decrease in the DVI >0.15 in the follow-up echocardiography without any clinical symptoms suggestive of AS or AR. cSVD is defined as the presence of eSVD along with symptomatic AS or AR. SVF was defined as any valvular degeneration due to infection (endocarditis), thrombosis, or mismatch necessitating valve replacement, as per the Valve Academic Research Consortium 3 criteria.[10]

Data collection and ethical considerations

Clinical data and outcomes were manually retrieved from the electronic medical record at Rochester General Hospital. Following data collection and before data analysis, patients’ information was deidentified to prevent any breach of confidentiality. The data collection was done by two pairs of researchers independently and matched for consistency; any discordance was resolved by a senior author. This research adhered to the ethical guidelines after obtaining approval by the Institutional Review Board (IRB) at Rochester Regional Health; IRB 2024 A.

Operative details

Preoperative assessment

Patients who underwent aortic valve replacement were evaluated for symptoms, aortic stenosis or regurgitation severity, and comorbid conditions. An echocardiogram and low-dose dobutamine stress test, or computed tomography, were done as indicated at the discretion of the cardiologists and cardiothoracic surgeons. Preoperative assessment included a left heart catheterization to evaluate the need for coronary artery bypass surgery (CABG) with the aortic valve replacement procedure. A bedside spirometry and/or a 5-m walk test were also completed to evaluate the risk of the surgery. Routine blood work included a complete blood count, a basic metabolic panel, and a coagulation panel.

Surgical intervention

The patient went to the operating suite and underwent general anesthesia and endotracheal intubation in the usual fashion. The patient was cooled to 30°. An aortic cross-clamp was placed. Antegrade cardioplegia was used. The aorta was opened, the valve was removed, and a new valve was sewn in place using individual Dacron sutures. The patient was rewarmed, and the cross-clamp was removed. A transesophageal echocardiogram (TTE) was done pre- and postaortic valve replacement to confirm the correct position. The valves used were 19, 21, 23, or 25 mm St. Jude Trifecta tissue valve. Other cardiac procedures such as CABG, myomectomy, mitral valve repair or replacement, and/or maze procedures were performed concomitantly as indicated at the discretion of the surgeon and interdisciplinary team.

Postprocedural care and follow-up period

Routine postoperative care involved admission to the intensive care unit and close observation. They were discharged home once medically stable and cleared by physical therapy. Outpatient clinic follow-up of patients involved visits with a cardiologist and a repeat TTE within the first 3–6 months, and then again as indicated.

Study outcomes (primary and secondary)

Our primary outcome is to evaluate the rate of SVD and SVF in patients following Trifecta valve implantation. Our secondary outcomes include postoperative pacemaker placement and to study the effect of comorbidities and implanted valve size on the rate of SVD across the study period.

Statistical analysis

Data were compiled into Microsoft Excel for Office 365 MSO spreadsheet. The statistical analysis of the data was conducted using JMP Pro version 14 software (SAS Institute, Cary, NC, USA). Graphs were designed using GraphPad Prism (GraphPad Software, Inc., CA, US). Descriptive statistics, including frequencies, means, and percentages, were used to summarize the data; these included the prevalence and characteristics of headaches. Categorical variables like sex were expressed as numbers (%), whereas continuous variables such as age, ejection fraction (EF), and Doppler velocity index (DVI) will be summarized as mean ± standard deviation. Comparisons for continuous variables will be made using paired or unpaired Student’s-t-tests, or analysis of variance test as appropriate. Valid percentages were reported to account for missing data. Categorical variables, such as hypertension and sex, will be compared using the Chi-square or Fisher’s exact test as appropriate. Multivariate linear/logistic regression models were used to examine, for example, the relationship between the incidence of SVF and various patients’ baseline characteristics (e.g., sex, age at surgical intervention, weight, and baseline echocardiogram findings). Correlation rules were applied to measure the strength of relations. P < 0.05 was considered statistically significant. All the results were obtained and compared to those of previous articles to validate and conclude.

RESULTS

Baseline demographics and characteristics

We included 271 patients in our analysis. The mean age of our sample was 71.1 ± 10.6 years, and 57.9% (n = 157) of the patients were males. Further information regarding our sample’s characteristics is in Table 1.

Table 1.

The demographics and clinical characteristics of study participants at baseline with their body weight and caloric intake over different dietary interventions

Variables	Value
Number of subjects (n)	271
Gender (male:female)	157:114
Age (mean±SD)	71.1±10.6
Weight (kg)	85.6±20.5
Height (m)	166.4±15.3
BMI (kg/m2)	30.5±6.7
BSA (m2)	2.0±0.3
Comorbidities
Hypertension	232 (85.6)
Hyperlipidemia	209 (77.1)
Type 2 diabetes mellitus	97 (35.8)
Ischemic heart disease	166 (61.3)
Atrial fibrillation	94 (34.7)
ESRD	14 (5.2)
Primary aortic valve lesion
AS	207 (76.4)
AR	38 (14.0)
ASR	26 (9.6)
Bicuspid valve	38 (14.0)

Continuous variables are expressed as mean±SD. Frequencies are expressed as n (%), whereas continuous variables are expressed as mean±SD. BSA=Body surface area, ESRD=End-stage renal disease, AS=Aortic stenosis, AR=Aortic regurgitation, ASR=Aortic stenosis and regurgitation, SD=Standard deviation, BMI=Body mass index

Operative characteristics

The most common primary aortic valve lesion requiring surgical intervention was aortic stenosis in 207 (76.4%) patients. Aortic regurgitation (14.0%, n = 38) and combined aortic disease (9.6%, n = 26) were less prevalent. Based on valve morphology, we had 38 (14.0%) patients with bicuspid aortic valves. We had 109 patients who underwent isolated Trifecta surgery without any other concomitant procedure. Meanwhile, 99 (36.7%) patients had concomitant CABG, 31 (11.5%) underwent MAZE procedure, 16 (5.9%) had mitral valve procedure, and 22 (8.1%) had myomectomy, [Figure 1]. Different valve sizes were used for the SAVR procedure; the most used was 23 mm (35.6%, n = 96), followed by 21 mm (32.6%, n = 88), 25 mm (16.3%, n = 44), 19 mm (12.6%, n = 34), 27 (n = 1) and 29 (n = 1), [Figure 2]. Postoperatively, 93 (34.3%) patients required pacemaker placement before discharge.

Figure 1.

The concomitant procedures performed with the Trifecta valve placement

Figure 2.

The distribution of patients based on the Tifecta valve size

Echocardiographic evaluation

Echocardiography was performed just before the surgery, immediately after postoperative (IPE), and the most recent echocardiogram (MRE). The baseline preoperative echocardiographic assessment showed the mean left ventricular EF to be 53.0 ± 11.3%. The IPE findings revealed a mean EF_IPE of 53.6 ± 10.1%, with a peak velocity (PV_IPE) of 2.34 ± 0.53 m/s, PG_IPE of 12.58 ± 6.26 mmHg, and DVI_IPE of 0.56 ± 0.87. The mean duration of the MRE was 5 ± 0.7 years, it showed a mean EF_MRE of 55.2 ± 10.4 mmHg, with a PV_MRE of 2.47 ± 0.69 mmHg, PG_MRE of 14.09 ± 8.74 mmHg, and DVI_MRE of 0.45 ± 0.18.

Paired t-test analysis of the EF revealed no significant change from the preoperative to IPE, P = 0.88. However, there was a significant increase between the EF_IPE to EF_MRE (P < 0.01, r = 0.7) and EF_preop to EF_MRE (P < 0.01, r = 0.5). When we assessed other echocardiographic characteristics, there was a significant increase from PV_IPE to PV_MRE (mean difference [MD] ± SEM [0.15 ± 0.04], P < 0.01), an increase in PG_IPE to PG_MRE (MD ± SEM [1.70 ± 0.49], P < 0.01), and a decrease in DVI_IPE to DVI_MRE (MD ±SEM [−0.037 ± 0.01], P = 0.01). Figure 3 shows the changes in the echocardiographic findings.

Figure 3.

Shows the difference in the preoperative and postoperative echocardiography findings based on the ejection fraction (a), peak velocity (b), mean pressure gradient (c) and Doppler velocity index (d)

Structural valve degeneration and SVF rates and predictors

After an average of a 5-year follow-up period, a total of 59 (21.8%) patients had SVD (ΔPG >20 mmHg: 5 patients, ΔDVI <−0.15: 47 patients, and both: 7 patients), whereas only 6 (1.5%) had cSVD with a total SVD of 63 (23.2%). A comparison was performed between those with SVD and those without in terms of ΔPG and ΔDVI. ΔPG was significantly higher in patients with SVD (7.87 ± 11.3 vs. −0.77 ± 5.52; difference: −8.6 ± 1.17 mmHg, 95%CI [−10.94–(−6.3)], P < 0.01), and ΔDVI was significantly lower in patients with SVD (−0.51 ± 1.62 vs. 0.04 ± 0.15; difference: 0.55 ± 0.13, 95%CI [0.13–0.98], P < 0.01).

Regarding the SVF rate, 13 (4.8%) patients had failed valves requiring replacement throughout the study period. The mean duration of the valve failure from the implant is 38.8 ± 24.4 months, with a range of 3–75 months. The indications for the replacement were AR (n = 4), AS (n = 4), endocarditis (n = 3), severe paravalvular leakage (n = 1), and mismatch (n = 1). Echocardiographic findings of the failed valves are listed in Table 2. To determine the predictors for valvular degeneration, we performed multivariate logistic regression, and none of the included variables predicted SVD, as detailed in Figure 4.

Table 2.

The echocardiographic characteristics of failed valves

Variables	Value
LVEF (%)	52.5±14.2
Transvalvular PG (mmHg)	34.6±19
LVOT (mm)	27.1±6.0
LVOT VTI	87.8±23.8
DVI	0.33±0.1
Peak velocity (m/s)	3.8±1
DPG (mmHg)	27.6±15.7
DDVI	−0.12±0.13

Continuous variables are expressed as mean±SD. LVEF=Left ventricular ejection fraction, LVOT=Left ventricular outflow tract, VTI=Velocity time integral, DVI=Doppler velocity index, PG=Pressure gradient, SD=Standard deviation

Figure 4.

Predictors for the development of structural valve degeneration

A summary of the study outcomes is summarized in the central illustration [Figure 5].

Figure 5.

Central illustration

DISCUSSION

This retrospective, single-center study shows the 5-year follow-up outcomes of 270 patients who underwent aortic valve replacement procedures using the Trifecta Valve between 2014 and 2019. Our study demonstrated that the incidence of echocardiographic SVD was 21.8%, clinical SVD had an incidence of 1.5%, and 4.8% of patients had SVF requiring replacement. This rate of valve failure was similar to the rate reported in a larger prospective study that included 710 patients (4.0%, [95% CI 2.4–6.6]).[5] Higher rates of reoperation were seen in another study published in 2020, with 9.3% of patients with Trifecta valves requiring reoperation over a mean follow-up period of 4.45 ± 1.63 years.[8] This is significantly higher than the rate of reoperation seen with other types of bioprosthetic aortic valves, with a systematic review and meta-analysis from 2021 comparing Trifecta valves to Perimount valves, including 11,135 patients across six observational studies, showing that Trifecta valves have higher reintervention rates with an HR of 3.16 (95% CI 1.83–5.46, P < 0.0001, I² = 40%).[9]

The current literature regarding the rates of SVD in Trifecta valves is conflicting, with some studies showing excellent outcomes with the use of Trifecta valves[11,12,13,14] and low rates of SVD, with some studies reporting freedom from SVF of 98% at 5 years.[15,16] However, these studies must be interpreted cautiously, as they often used unclear or different definitions for “SVD.” For example, a study by Anselmi et al. in 2017 reported freedom from SVD rate 98% ± 0.9% (n = 6) for recipients of Trifecta valves at 5 years but provided no clear definition for SVD other than “changes intrinsic to the device causing dysfunction, evident at echocardiography, reoperation, or autopsy.”[15,17] The issue of standardizing these definitions was addressed by Dvir et al. in 2018, where they highlighted that a lack of clear definition presented a barrier to comparing the durability of different types of implants and proposed a definition built in stages to describe different degrees of valve degeneration ranging from morphological leaflet abnormality without significant hemodynamic change (Stage 1) to severe stenosis or regurgitation (Stage 3).[18] To date, there is no consensus on a definition for SVD. The more recent studies, especially those with a control group, showed higher rates of SVD with Trifecta valve recipients. A study by Fukuhara et al., published in 2020, used the same definition for SVD as we used in our study and showed that the cumulative incidence of SVD was higher in the Trifecta group compared to the non-Trifecta group (13.3% vs. 4.6%; P = 0.010).[8] However, it is notable that patients who received the Trifecta valve were older and received smaller valves (23.0 mm vs. 25.0 mm; P < 0.001) than patients in the non-Trifecta group, limiting the interpretability of these results.[8]

There was a similar study to ours conducted by Wakami et al. evaluating the rates of SVD and risk factors for early SVD. Out of 110 studied cases, seven cases encountered SVF. Their reported SVD rate was lower than ours, 4.8% at 5 years and 6.6% at 7 years. The main focus of their study was to evaluate the impact of patient–prosthesis mismatch (PPM) as their SVD rate was significantly higher 2.8% at 5 years and 20.0% at 7 years (vs. 2.8% and 12.6% for non-PPM).[19] Surprisingly, in our study, we had only one case of mismatch which can be explained by the short follow-up period of 5 years, as compared with their study of 7-year follow-up. Looking at their SVD rate, it was low at 5 years (only 2.8%).

In our study, using multivariable logistic regression, we could not identify any possible predictor factors associated with SVD. Our findings were similar to Fukuhara et al., who used multivariable competing risks regression to identify risk factors associated with SVD; however, they were only able to identify younger age as the sole factor associated with SVD, with a hazard ratio of 0.56 per 10-point increase in age (95% CI [0.44–0.72], P < 0.001).[8] Regarding concomitant cardiac surgeries with Trifecta valve implantation, more than a quarter of our patients (n = 99, 36.7%) had concomitant CABG procedure, whereas 31 patients had a concomitant MAZE procedure, and another 22 patients had a concomitant myomectomy. Recent studies have shown that combined CABG with SAVR is associated with increased in-hospital mortality and complications[20,21] meanwhile, long-term survival was comparable between isolated SAVR versus combined CABG and SAVR.[22] A recent meta-analysis which included 74,560 from 44 retrospective studies concluded that isolated SAVR is associated with a lower rate of early mortality compared to SAVR with concomitant CABG, with OR = 0.70 (95% CI 0.66–0.75; P < 0.001).[22] In addition, concomitant CABG was associated with a higher incidence of renal failure, the need for renal replacement therapy, and prolonged mechanical ventilation. This is an area of debate, however, with no randomized control trials available to date.

Strengths and limitations

Our study results are best interpreted within its limitations, which include the fact that it is a retrospective, observational study. Our sample size was limited by some loss to follow-up, with only 271 patients being included in the final analysis, out of 300 patients who underwent aortic valve replacement using Trifecta during our study period. Our follow-up period ranged from 3 to 5 years, with the short follow-up period in some patients limiting the interpretability of the results. Due to the limited patient population, our study was underpowered to adequately detect all the risk factors/predictors associated with valve failure.

CONCLUSIONS

Over a 5-year follow-up period, 4.8% had SVF with an SVD of 23.2%, with the majority of SVD not being clinically significant except in six patients. These results corroborate with a previously published study suggesting a bad clinical outcome of Trifecta valve placement.

Ethics review board approval

Ethical approval was obtained from the Institutional Review Board of Rochester Regional Health on September 6, 2022. IRB approval number: 2024 A.

CRediT author statement

Anas Hashem: Data collection and Formal Analysis, Methodology, Software, Conceptualization, Visualization and Writing – Original Draft. Ahmed Elkhapery: Conceptualization, Validation, Formal analysis, and Data collection. Rupinder Buttar: Data collection, Writing – Original Draft. Amr Mohamed: Conceptualization, and Methodology, Data Curation Writing – Reviewing and Editing. Mohamed Salah Mohamed: Writing – Original Draft, and Validation. Amani Khalouf: Investigation, Data Curation and Writing – Original Draft. Faiz Abbas: Data Collection, Formal analysis, Writing-Original Draft and Resources. Saif ur Rehman: Writing-Original Draft, Resources and Validation. Mahmoud Eisa: Data Collection – Writing – Revision and Editing. Bipul Baibhav: Writing – Review and Editing, Supervision and Project administration. Devesh Rai: Data Curation, Writing – Review and Editing, and Supervision. Jeremiah P Depta: Writing – Review and Editing, and Supervision.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.