Durability of bioprosthetic surgical aortic valve replacement: valve reintervention is only the tip of the iceberg

Bioprosthetic valves are the most frequently used valve substitute for surgical aortic valve replacement (SAVR). Long-term durability of the bioprosthetic valve is a major issue, which is becoming even more important, nowadays, as SAVR with a bioprosthesis (bio-SAVR) is expanding to lower-risk and younger populations with longer life-expectancy. In this issue of the journal, Schmiegelow et al. reports the results of the nationwide Danish Register of Surgical Procedures, which included 11 251 patients ≥40 years with isolated bio-SAVR with or without concomitant coronary artery bypass graft surgery between 2000 and 2016. Of these patients, the authors excluded 733 patients who died or underwent early valve reintervention within 90 days following SAVR, therefore, resulting in the final study cohort including 10 518 patients. In this cohort, the authors observed low valve reintervention rate <1.7% (3.7/1000 person-years) but high mortality rate of 21.7% at 5 years following SAVR with 37% of the reinterventions being related to endocarditis. After accounting for competitive risk of death, younger age remained associated with the higher risk of valve reintervention, whereas sex, ethnicity or education level did not. Re-intervention consisted in redo-SAVR in ∼90% of the cases, and transcatheter valve-in-valve procedure was rarely used in this study period. The 30-day (10.1%) and 1-year (21.6%) mortality rates following reintervention were approximately two-fold higher than those following initial SAVR (4.4% and ∼10%, respectively in the whole registry cohort). The results of this large nationwide study are consistent with those reported in the SAVR arms of recent randomized trials comparing transcatheter aortic valve replacement versus SAVR. In the CoreValve U.S. pivotal High-Risk Trial, the 5-year rate of reintervention was 1.1% following SAVR [1]. In the PARTNER 2A Trial SAVR arm, the 5-year rate of reintervention was 1.0% with 67% of these re-interventions being related to endocarditis, and 30-day mortality following reintervention was very high (50%). The main conclusion of this important and elegant study by Schmiegelow et al. is that valve reintervention is rare during the first 5 years following bio-SAVR and is predominantly related to endocarditis. Although these data are compelling and reassuring, this study has several limitations that merit to be discussed. First, this study primarily reports mid-term (5 years) results, whereas structural valve deterioration (SVD) and ensuing valve failure and reintervention generally occur beyond 7 years. Second, this registry captured medico-administrative data, which are not necessarily accurate and may lack information, granularity and completeness with regard to baseline and outcome data. Third, the occurrence and aetiology of events (reintervention and mortality) were not adjudicated. The primary end point of this study [2] was valve reintervention, which is often used as a marker for bioprosthetic valve failure (BVF). However, several studies suggested that the sole consideration of valve reintervention grossly underestimates the true incidence and clinical impact of structural or non-structural valve dysfunction and failure following SAVR [3, 4]. Indeed, reintervention is only the tip of the valve dysfunction/failure iceberg. First, this end point does not capture the deaths, strokes or cardiac re-hospitalizations caused, directly or indirectly, by bioprosthetic valve dysfunction (BVD). Second, BVD related to SVD, non-SVD (i.e. paravalvular regurgitation, prosthesis–patient mismatch), valve thrombosis or endocarditis may have significant detrimental impact on patient’s longevity and quality of life, even if no reintervention is required nor performed. In particular, elderly patients with BVF are often at high or prohibitive surgical risk and may thus not undergo reintervention. This may have been an important bias in the present study [2] given that the study period was in large part prior the introduction of transcatheter valve in valve, which nowadays provides a less-invasive alternative to redo SAVR in patients with BVF. To overcome these limitation, the European standardized definitions [5] and Valve Academic Research Consortium-3 [4] proposed a more comprehensive and rigorous definition of BVF, which not only includes valve reintervention but also valve-related death as well as any BVD with clinically expressive criteria (new-onset or worsening symptoms, LV dilation/dysfunction or pulmonary hypertension). Furthermore, these expert consensus statements recommended to report and stage the deterioration of bioprosthetic valve structure and function following AVR, i.e. stage 1: morphological valve deterioration with no haemodynamic changes, stage 2: stage 1 associated with moderate haemodynamic valve deterioration (stenosis and/or regurgitation), and stage 3: stage 1 associated with severe haemodynamic valve deterioration [4]. We previously reported that the incidence of valve reintervention at 10 years following bio-SAVR was 3.5%, which is similar to the rate estimated in the present study [3]. However, in this previous study, the 10-year rates of stage 2 or 3 structural/haemodynamic valve deterioration (41%) was >10-fold higher than the rate of reintervention and was independently associated with 2.2-fold increase in subsequent mortality. The present study [2] did not report the rates of valve haemodynamic deterioration as defined in VARC-3 but it is likely that these rates would be much higher than the rate of reintervention. In the PARTNER 2A SAVR arm, the 5-year rate of stage 2–3 valve deterioration was 3.5% (vs 1.3% for BVF and 1.0% for reintervention) [6]. The European guidelines recommend annual transthoracic echocardiographic follow-up after bio-SAVR, whereas American guidelines recommend annual follow-up only beyond 5 years. Both guidelines consider that annual echocardiographic follow-up is reasonable for patients with transcatheter AVR. Based on their results and, in particular, on the high ratio of the number echocardiograms per reintervention (i.e. 248), the authors advocate for not performing routine annual echocardiographic follow-up during the first 5 years following bio-SAVR. This suggestion should however be considered with caution given that regular echocardiographic follow-up also allows early detection of structural and non-structural BVD, which may have negative impact on outcomes, or may require rapid onset of specific treatment (e.g. anticoagulation in case of valve thrombosis) or early valve re-interventions. Early diagnosis of BVD may help to optimize decision-making process with regard to timing and type of therapeutic management and follow-up and may therefore contribute to reduce unplanned hospitalizations and adverse events. In conclusion, the rate of reintervention is low during the first 5 years following Bio-SAVR and predominantly related to endocarditis. However, reintervention reflects only the tip of the iceberg of bioprosthetic valve dysfunction and complications. Hence, regular clinical and echocardiographic surveillance remain relevant even during the first 5 years following both surgical and transcatheter bio-AVR. Conflict of interest: Philippe Pibarot received research grants from Edwards Lifesciences and Medtronic for echocardiography core laboratory services, for which he receives no direct industry compensation. The other authors have no disclosure.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.