Biological versus mechanical aortic valve replacement in non-elderly patients: a single-centre analysis of clinical outcomes and quality of life

Abstract

OBJECTIVES

The aim of this study was to evaluate short- and long-term clinical outcomes, including the perceived health-related quality of life, in patients younger than 65 years having undergone aortic valve replacement either with biological or mechanical valve prostheses.

METHODS

Between 2002 and 2013, 242 consecutive patients <65 years of age underwent isolated aortic valve replacement at our institution, either with biological (n = 134, 55.4%) or mechanical (n = 108, 44.6%) prostheses. Survival, health-related quality of life, short- and long-term clinical outcomes and echocardiographic data were analysed with a retrospective, single-centre study. Propensity matching was performed.

RESULTS

No significant difference in survival was found between the 2 groups (mechanical versus biological: 100% vs 96.6% at 1 year, 98.2% vs 93.1% at 5 years and 92.3% vs 83.4% at 10 years after surgery, P = 0.091). For all the interviewed patients (n = 161, 66.5%), perceived quality of life at the latest follow-up was excellent. Need for reoperation was higher in the bioprosthetic group (8% vs 0%, P = 0.995), whereas the rate of major bleedings was higher in the mechanical valve group (3% vs 20%, P = 0.094). The mean and maximum transvalvular pressure gradients were 20.5 ± 9.7 and 37.4 ± 17.5 mmHg in the biological group and 14.8 ± 4.8 and 26.6 ± 9.2 mmHg in the mechanical group (P = 0.014).

CONCLUSIONS

No significant differences were found between biological and mechanical valves in terms of patients’ survival, clinical outcomes and quality of life. Mean and maximum transvalvular pressure gradients were significantly higher in the biological group. The majority of patients would opt for the same prosthesis type, if asked to choose again.

INTRODUCTION

Replacement with a biological prosthesis is the most common adopted strategy to treat aortic valve defects in elders [1, 2]. In young patients, the discussion about the ideal device to implant during aortic valve replacement (AVR) falls into the unending dilemma about pros and cons of mechanical versus biological devices. To date, an unequivocal answer has not been found [3]. Mechanical valves offer the advantage of a lifetime durability but require lifelong anticoagulation, resulting in an increased risk of bleeding and of complications during pregnancy. Biological valves, on the other hand, do not require lifelong anticoagulation unless another indication is present. However, they are subjected to structural valve deterioration (SVD) over time resulting in the necessity of reoperation, especially in younger patients [3]. Nevertheless, recent randomized data suggest that long-term survival after biological and mechanical AVR is equivalent [4]. Consequently, the rate of valve-related complications becomes prominent in the choice of the prosthesis type. The number of implanted bioprostheses has steadily increased in the last decade, due to patients’ willingness to an anticoagulation-free life [5] and new data demonstrating that the risk of reintervention in young patients receiving a biological stented valve is less frequent than what predicted in the past [2]. Furthermore, during the last decade, the use of valve-in-valve transcatheter aortic valve implantation (ViV-TAVI) has complemented surgical aortic valve replacement (sAVR) in the treatment of high-risk patients with degenerated aortic bioprostheses [6].

With this study, we aimed to evaluate short- and long-term outcomes after AVR with mechanical or biological prostheses in non-elderly patients, at our centre. Furthermore, we analysed the impact that surgery had on their perceived quality of life.

PATIENTS AND METHODS

Ethical statement

Informed consent has been waived due to the retrospective nature of the study. At our institution, all patients, at the time of admission, are asked to sign a consent form for data collection for research purposes. Verbal consent for the interview and for data collection for scientific purposes was obtained from all the patients included in the study.

Methods

In this retrospective study, we included all consecutive patients (n = 242) younger than 65 years (non-elderly) undergoing sAVR with either mechanical [mechanical heart valve (MHV)] or biological [biological heart valve (BHV)] prosthetic valves, between January 2002 and December 2013, at our institution. All procedures were performed under general anaesthesia through median sternotomy. Surgery was performed on cardiopulmonary bypass and after aortic cross-clamping. Myocardial protection was achieved by the infusion of cold cardioplegia. A mechanical device was implanted in 108 patients (44.6%, SJM Regent™ valve in 103 patients and Masters HP™ in 5). A bioprosthetic implant was performed in 134 patients (55.4%, Pericardial Bovine Carpentier Edwards in 106 patients, Porcine Biocor in 25 patients and stentless valve in 3). Inclusion criteria were age <65 years and an isolated AVR procedure. Therefore, all patients undergoing associated cardiac procedures were excluded. The mean follow-up length for the overall population was 8.5 ± 3.5 years.

Clinical charts, operative reports, follow-up visits and pre- and postoperative imaging (including transthoracic echocardiography) were reviewed. Echocardiographic data were collected from 147 patients (60.7%, 79 MHV and 68 BHV) at latest follow-up. In addition, health-related quality of life (HRQoL) was assessed by phone interview in 161 patients (66.5%, 85 MHV and 76 BHV) at the latest follow-up. As a primary outcome, survival was evaluated at 30 days (short term) and at 1, 5 and 10 years follow-up (long term). At the same time points, the need for reoperation and any adverse events, including overt bleedings and cerebral ischaemic/haemorrhagic events (secondary outcomes), were evaluated. As additional primary outcome, HRQoL was assessed through the administration of the SF36 questionnaire at the latest follow-up. SF36 consists of 36 health-related questions grouped into 8 main domains, as follows: physical functioning, role limitations due to physical health problems, bodily pain, general health perceptions, vitality, social functioning, role limitations due to emotional problems and general mental health. For each domain, the answers received a score from 0 to 100, where higher scores reflect better patient’s health perception [7]. In addition to SF36, a qualitative and self-administered analysis on valve-specific questions was completed, aimed at collecting information in regard of patients’ experience with the implanted device [3].

Statistical analysis

Descriptive statistics was reported as I quartile/median/III quartile for continuous variables (for echocardiographic variables and SF36 scores as mean ± SD) and percentages (absolute numbers) for categorical variables.

A propensity score matching approach was employed to account for potential confounding related to the non-random allocation of the patients to the 2 groups. Propensity scores were estimated using covariate balancing propensity scores algorithm [8]. A matching approach was employed using generic algorithm with a 1:1 ratio.

To analyse postoperative outcomes in the matched-pair cohort, a generalized linear model approach was used for accounting overdispersion and correlation in data.

The survival distribution in the 2 groups was evaluated using the Kaplan–Meier approach. Outcomes at follow-up were assessed using Cox proportional hazard models when the date of the event of interest was available.

P-values from the quality of life, echocardiographic and short- and long-term outcome analyses underwent Benjamini–Hochberg [9] correction to account for the multiplicity of testing. Both unadjusted and adjusted P-values are reported.

The analyses were performed using R software (version 3.6.2) with the packages covariate balancing propensity scores, MatchIt, Survival and RMS.

RESULTS

Patient characteristics

The preoperative characteristics of the overall patient population before and after propensity matching are described in Table 1. The unadjusted analysis showed that the patients in the BHV group were older than those in the MHV group. BHV patients presented more frequently hypertension and dyspnoea, whereas MHV patients had higher incidence of previous stroke. Indication to AVR was most frequently valve stenosis in the BHV group and regurgitation or mixed lesions in the MHV group. After propensity matching, no statistically significant differences were observed.

Table 1:

Preoperative characteristics before and after propensity matching

	Overall cohort (n = 242)			Propensity matched (n = 116)
	Biological (N = 134)	Mechanical (N = 108)	P-value	Biological (N = 58)	Mechanical (N = 58)	P-value
Male	73 (98)	74 (80)	0.869	81 (47)	72 (42)	0.272
Age (at operation)	52.0/57.5/61.0	45.0/54.5/59.0	0.004*	50.5/56.0/60.0	49.0/56.0/61.0	0.947
Prosthesis diameter (mm)	21/23/25	21/23/25	0.579	21/23/25	21/23/25	0.406
Endocarditis	10 (13)	9 (10)	0.907	12 (7)	10 (6)	0.769
LVEF <40%	7 (9)	4 (4)	0.301	3 (2)	5 (3)	0.648
Hypertension	69 (92)	50 (54)	0.003*	59 (34)	59 (34)	1
Renal impairment	6 (8)	3 (3)	0.236	3 (2)	3 (2)	1
NYHA III–IV	50 (67)	32 (35)	0.006*	40 (23)	38 (22)	0.849
COPD	14 (19)	18 (19)	0.468	14 (8)	14 (8)	1
Severe peripheral vasculopathy	2 (3)	0 (0)	0.118	5 (3)	0 (0)	0.079
Previous stroke	2 (3)	8 (9)	0.03*	3 (2)	3 (2)	1
Diabetes	4 (5)	1 (1)	0.166	2 (1)	2 (1)	1
Aortic valve insufficiencya	26 (35)	32 (35)	0.037*	31 (18)	28 (16)	0.915
Aortic valve stenosisa	66 (89)	52 (56)	–	55 (32)	59 (34)	–

Data expressed as I quartile/median/III quartile or percentages (absolute numbers).

^aPredominant valvular defect.

^*Statistically significant data.

COPD: chronic obstructive pulmonary disease; LVEF: left ventricle ejection fraction; NYHA: New York Heart Association.

Primary end points

Survival

No intraoperative and 30-day mortality was recorded in the overall population (n = 242, 100%). In the long term, death occurred in 24 patients (10%) equally distributed between groups (12 patients per group, 11% in MHV vs 9% in BHV). Compared to the MHV group, in BHV, death more frequently occurred for non-cardiac causes (n = 6, 50%). In the same group, other causes were cardiogenic shock secondary to endocarditis in 2 patients (16.7%), heart failure in 1 patient (8.3%) and major arrhythmic event in another case (8.3%). Unknown were the causes of death for the 2 remaining patients (16.7%). In the MHV group, unknown causes and major arrhythmic events were responsible for 2 deaths (16.7%); further 2 (16.7%) patients died for cerebral haemorrhage. Heart failure (n = 1, 8.3%), myocardial infarction (n = 1, 8.3%) and prostheses dysfunction (n = 1, 8.3%) were also found as causes of death. After Kaplan–Meier analysis, the survival rate of the 2 groups (MHV versus BHV, Fig. 1 and Supplementary Material S5) was 100% vs 96.6% at 1 year, 98.2% vs 93.1% at 5 years and 92.3% vs 83.4% at 10 years after surgery (P = 0.091).

Figure 1:

Kaplan–Meier curves for survival for the overall population and for the 2 separate groups.

Quality of life

For all the interviewed patients (n = 161, 66.5%), perceived quality of life at the latest follow-up was excellent. No statistical difference was shown between groups in any of the 8 main domains of the SF36 (Fig. 2 and Supplementary Material S1). As mentioned, patients were also asked to answer valve-specific questions (Table 2) and to describe their experience with their involvement in the decision-making process leading to the prosthetic valve selection (Table 3). After the survey, 89.5% of the MHV patients and 88.2% of BHV ones, at the question ‘If I had to do it over again, would I make the same decision to have surgery?’ answered ‘yes’ or ‘probably’ (Table 2). As expected, MHV patients were more bothered (at least rarely) by the valve-closing sound (23.7% vs 0%), the frequency of blood tests (36.8% vs 0%) and more concerned (at least rarely) about bleedings (21.1% vs 2.4%, Table 2). On the other hand, BHV patients were more concerned about valve-related complications (41.2% vs 5.2%), valve failure (31.8% vs 6.5%) and reoperation (48.2% vs 3.9%) (Table 2).

Figure 2:

SF36 Short Form Health Survey scores after propensity matching.

Table 2:

Valve-specific questionnaire results

	All (N = 161) (%)	Biological (N = 85) (%)	Mechanical (N = 76) (%)
1—If I had to do it over again, would I make the same decision to have surgery?
Yes	82.6	82.3	82.9
Probably	6.2	5.9	6.6
I do not know	6.2	8.2	3.9
Probably not	1.3	1.2	1.3
No	3.7	2.4	5.3
2—Is there a valve sound that bothers me?
Never	88.8	100	76.3
Rarely	3.1	0	6.6
Occasionally	6.8	0	14.5
Frequently	0	0	0
Always	1.3	0	2.6
3—Following my surgery, the frequency of doctor visits and blood tests bothers me
Never	82.5	100	63.2
Rarely	1.3	0	2.6
Occasionally	6.2	0	13.2
Frequently	7.5	0	15.8
Always	2.5	0	5.2
4—The possibility of complications due to my implanted valve concerns me
Never	75.8	58.8	94.8
Rarely	3.1	4.7	1.3
Occasionally	11.2	18.8	2.6
Frequently	5.6	9.5	1.3
Always	4.3	8.2	0
5—I am concerned about possible bleeding caused by my anticoagulant medication
Never	88.8	97.6	78.9
Rarely	0.6	0	1.3
Occasionally	5.6	1.2	10.6
Frequently	3.7	1.2	6.6
Always	1.3	0	2.6
6—I am afraid that my valve may fail
Never	80.2	68.2	93.5
Rarely	3.7	5.9	1.3
Occasionally	9.9	16.5	2.6
Frequently	3.7	5.9	1.3
Always	2.5	3.5	1.3
7—I am afraid that I may need another valve operation
Never	72.7	51.8	96.1
Rarely	3.7	7	0
Occasionally	13.1	22.4	2.6
Frequently	6.8	11.8	1.3
Always	3.7	7	0

Table 3:

Patient experience with involvement in prosthetic valve selection

	All (N = 161) (%)	Biological (N = 85) (%)	Mechanical (N = 76) (%)
1—The doctor has involved me in prosthetic valve selection
Totally agree	75.8	89.4	60.5
Not agree/disagree	9.9	7.1	13.2
Totally disagree	14.3	3.5	26.3
Not applicable	0	0	0
2—I have received enough information to make a deliberate choice
Totally agree	97.5	96.4	98.7
Not agree/disagree	1.3	1.2	1.3
Totally disagree	0.6	1.2	0
Not applicable	0.6	1.2	0
3—I think it is important to be involved in prosthetic valve selection
Totally agree	98.1	96.4	100
Not agree/disagree	1.3	2.4	0
Totally disagree	0.6	1.2	0
Not applicable	0	0	0
4—I am satisfied with my prosthetic aortic valve
Totally agree	95.7	94.1	97.4
Not agree/disagree	3.7	4.7	2.6
Totally disagree	0.6	1.2	0
Not applicable	0	0	0

A complete satisfaction with the implanted device was found in 97.4% of the MHV patients and in 94.1% of the BHV ones. However, patients’ involvement in the decision-making process that led to the prosthesis type selection was reported in 90% in the BHV group and in 60.5% in the MHV one (Table 3).

Secondary end points

Need for reoperation on the aortic valve

In the short term (30 days), no redo procedures were recorded. At long-term follow-up, the unadjusted analysis showed a statistically significant difference in the need for reoperation between the 2 groups (N = 7, 8% in BHV vs N = 0, 0% in MHV, P = 0.027, Supplementary Material S2). After matching, the difference observed lost statistical significance (P = 0.995). Among the patients requiring reoperation, 5 underwent conventional sAVR whereas 2 received ViV-TAVI. The mean time elapsed from the first operation was 8 ± 2.5 years. In 2 of the 7 patients requiring reoperation, endocarditis led to a redo procedure. Both patients underwent open surgery. In the remaining 5 patients, reoperation was due to SVD, without superimposition of acute events.

Major bleedings

No differences were noticed between the 2 groups in terms of major bleedings at 30 days. Before discharge, 11 (4.5%) patients needed surgical revision for pericardial effusion, 5(4%) in the BHV group and 6 (8%) in the MHV group (P = 0.521) (Supplementary Material S3). At long-term follow-up, the rate of recorded overt bleeding events was significantly higher in the MHV group in the unadjusted analysis (propensity matched: N = 9, 20% vs N = 1, 3%, P = 0.047, Supplementary Material S2) but lost its statistical significance after the P-value underwent correction (P = 0.094).

Other adverse events

No differences were shown for all the other considered adverse events (acute renal impairment, myocardial infarction, infections, strokes, respiratory insufficiency, AV block requiring pacemaker implantation, need for postoperative circulatory support, such as intra-aortic balloon) in the short term (Supplementary Material S3) and in the long term (Supplementary Material S2).

Echocardiographic data

At follow-up, after matching, mean and maximum transvalvular gradients were significantly different between the 2 groups (mean 20.5 ± 9.7 vs 14.0 ± 4.8 mmHg, P = 0.014; maximum 37.4 ± 17.5 vs 26.6 ± 9.2 mmHg, P = 0.014, in the BHV and MHV groups, respectively). Only mild perivalvular leak was observed (Supplementary Material S4).

DISCUSSION

The ongoing dilemma regarding the choice of the best aortic prosthetic device in young patients has always been centred in the premise that bioprosthetic devices encounter a structural deterioration over time and require 1 or more other cardiac operations, while the mechanical devices guarantee a lifetime durability in exchange of the distress and risks related to a lifelong anticoagulation [2]. The 2006 ACC/AHA guidelines for the management of patients with valvular heart disease recommended the implantation of a mechanical device in patients younger than 65 years [10]. The 65 years old cut-off was considered the point in which the risk of reoperation due to SVD of the bioprosthesis was low enough to grant a higher advantage, if compared to the risks related to the implantation of a mechanical device [11].

In our study, 20 (15%) of the patients belonging to the BHV group were under 45 years at the time of the operation. For all of them, surgery was performed after 2006. Around that time, indeed, physicians and surgeons started to consider reasonable to re-evaluate the 65-year old cut-off value, setting it at lower age, based on data reporting longer durability of the new generation bioprosthetic valves [12]. Patients’ decline to lifetime anticoagulation therapy and the concomitant availability of novel technologies (e.g. TAVI) for less-invasive aortic valve implantation played a significant role in changing the common scientific attitude. As a consequence, the implantation of bioprosthetic devices has steadily increased over the last 10 years in all the age groups [5].

With regard to long-term mortality, recent evidence shows contrasting opinions. For some authors, survival is not different between BHV and MHV [13, 14]; for others, mechanical valves offer a higher survival rate [15]. Stassano et al. [4] in the only existing randomized trial on the matter demonstrated that there is no difference in long-term survival, at a follow-up of 13 years.

Schnittman et al. [16] in a recent observational study concluded that 15-year mortality in patients younger than 50 years who underwent an AVR using biological aortic valves is similar to those using mechanical prosthesis.

Our study demonstrated that biological valves are not inferior to mechanical prostheses with regard to long-term survival. No statistically significant difference was found in mortality at 1, 5 and 10 years follow-up after surgery.

Several studies thoroughly highlighted the long-term risk of major bleedings due to the anticoagulant therapy in MHV patients and the risk of reoperation due to SVD in the BHV patients [11, 13–15]. The average lifespan of a bioprosthetic valve is estimated at 15 years in elderly patients. However, the risk of SVD is accelerated in younger patients due to more pronounced immunological response to the valve and enhanced calcification of the valve [17]. Our analysis of long-term outcomes confirmed the presence of the above well-known risks, as reported in the literature.

Because of a recently observed considerable shift towards the implantation of bioprostheses, it is expected that patients will increasingly present with degenerated devices [18]. Conventional redo sAVR is regarded as the gold standard for patients with degenerated aortic bioprostheses and is associated with an acceptable in-hospital mortality rate of up to 5.1% [18]. In patients with degenerated aortic valve bioprostheses, especially elderly or high-risk patients, ViV-TAVI could be a safe, feasible alternative to redo sAVR [6]. ViV-TAVI is associated with a significantly lower rate of permanent pacemaker implantations and shorter intensive care unit and hospital stays. In contrast, redo sAVR offers lower incidence of patient–prosthesis mismatch, fewer paravalvular leaks and lower mean postoperative aortic valve gradients [18]. The ViV-TAVI approach could be a safe and feasible alternative to re-sAVR that may offer an effective, less-invasive treatment for patients with failed surgical aortic valve bioprostheses who are inoperable or at high risk [18]. In our study, only 2 of 7 patients requiring reoperation underwent ViV-TAVI procedure.

With regard to HRQoL, several studies demonstrated that MHV patients are generally more bothered by valve sound and frequency of blood tests and are also afraid of possible bleeding events. Compared to BHV patients, they are substantially less worried about SVD and reoperation [3, 19]. In the current clinical practice, many AVR patients experience decisional conflict and suboptimal involvement in the prosthetic valve selection [19]. Korteland et al. found that only 64% of patients were actively involved in the decision-making process leading to the choice of the prosthetic valve. A better patient experience with involvement in prosthetic valve selection was associated with better mental health [3].

Our analysis of HRQoL confirmed and reinforced the findings from the cited available literature. Further analysis demonstrated that none of the MHV patients was already on total anticoagulation therapy and received a mechanical valve as a consequence. Ten women with child-bearing potential were included in our study (4.1%). Among them, 7 received a metallic valve and 3 a biological valve. These numbers confirm that the choice of the prosthesis type was not forced by their fertility. The marked difference observed in patients’ involvement in the decision-making process (89.4% in BHV vs 60.5% in MHV) led to further investigations on the underlying reasons. Deeper analysis showed that all the MHV patients stating that they were not involved, were operated before 2006, when the decision of implanting a biological device in patients under 65 years of age was not supported by guidelines. From our perspective, patients’ active involvement in the decision-making process might positively affect their life perception and their level of compliance to postoperative therapies and management strategies.

This study is one of the very few available in the literature combining the analysis of the clinical outcomes to the perceived quality of life for non-elderly patients after AVR. The different types of aortic prosthesis can impact heavily on the patient’s quality of life, if not chosen wisely [3]. For some professionals, the quality of life is not an element that can be compared to more objective end-points such as the disease characteristics or the outcome data [20]. Nevertheless, most of the patients are happy to report to someone the way in which their disease and all the associated treatments impact on their daily routine [21].

Over the past few years, technology has made huge advances. For mechanical valves, the hypothesis of introducing non-vitamin K antagonist anticoagulants (NOACs) is under evaluation. However, no large randomized trials are currently ongoing. Some randomized trials evaluating lower international normalized ratio (INR) targets have shown that, although lower INR ranges are beneficial in terms of bleeding events, it is unclear to what extent the INR target can be lowered to be considered safe for the prevention of valve thrombosis and stroke [17].

On the bioprosthesis’ side, ViV-TAVI has been developed as a safe and feasible alternative to redo sAVR for inoperable or high-risk patients [18]. However, experience with the new technologies is very limited and long-term follow-up is absent or scarce. New studies and randomized trials data are therefore needed.

Limitations

Our study should be viewed in light of its limitations. We performed a retrospective observational, single-centre study with limited long-term follow-up data. Propensity–matched analysis was performed to account for potential confounding related to the non-random allocation of the patients.

Only a fraction of the patients included in the study agreed to complete the HRQoL and valve-related quality of life questionnaires; therefore, a potential selection bias might ultimately be present.

CONCLUSIONS

Our data show that no significant differences exist between biological and mechanical devices in terms of patients’ survival and HRQoL in the short term and the long term. Bioprostheses undergo deterioration and require reoperation, while mechanical prostheses require lifelong anticoagulation and expose the patient to the risk of bleedings. Mean and maximum transvalvular pressure gradients were found to be significantly higher in the biological group. Despite experiencing different levels of involvement in the decision-making process, the majority of the patients would opt for the same prosthetic valve type, if asked to choose again.

SUPPLEMENTARY MATERIAL

Supplementary material is available at ICVTS online.

Funding

This work did not require the support of any funding agency.

Conflict of interest: none declared.

Author contributions

Fabio Stocco: Conceptualization, Resources, Formal Analysis, Investigation, Project administration, Methodology, Writing—original draft. Assunta Fabozzo: Validation, Visualization, Writing—review & editing. Lorenzo Bagozzi: Formal analysis. Chiara Cavalli: Formal Analysis, Validation. Vincenzo Tarzia: Conceptualization, Investigation, Data curation, Validation. Augusto D’Onofrio: Validation, Writing—review & editing. Giulia Lorenzoni: Formal analysis. Valentina Chiminazzo: Formal analysis. Dario Gregori: Formal analysis. Gino Gerosa: Conceptualization, Validation, Supervision.

Reviewer information

Interactive CardioVascular and Thoracic Surgery thanks Pepijn Grashuis, David C. Reineke and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.

ABBREVIATIONS

AVR

Aortic valve replacement

BHV

Biological heart valve

HRQoL

Health-related quality of life

INR

International normalized ratio

MHV

Mechanical heart valve

sAVR

Surgical aortic valve replacement

SVD

Structural valve deterioration

ViV-TAVI

Valve-in-valve transcatheter aortic valve implantation