Author + information
- Evaldas Girdauskas, MD, PhD∗ (, )
- Mina Rouman, MD,
- Kushtrim Disha, MD,
- Andres Espinoza, MD,
- Martin Misfeld, MD, PhD,
- Michael A. Borger, MD, PhD and
- Thomas Kuntze, MD
- ↵∗Department of Cardiac Surgery, Central Clinic Bad Berka, Robert-Koch-Allee 9, 99437, Bad Berka, Germany
The effect of bicuspid aortic valve (BAV) phenotype on the future risk of type A aortic dissection is unknown. On the basis of previous data (1), we hypothesized that the risk of post–aortic valve replacement (AVR) aortic dissection might be different in BAV insufficiency versus BAV stenosis.
A meta-analysis was performed in accordance with published guidelines (2). We conducted a systematic search on PubMed, Embase, Ovid, and Google Scholar by using the terms “aortic dissection,” “bicuspid aortic valve,” “aortic valve replacement,” “previous cardiac surgery,” and “bicuspid aortopathy.” The search was limited to original adult human studies of type A aortic dissection occurring after previous isolated AVR surgery in patients with BAV disease and known functional phenotype (i.e., insufficiency vs. stenosis). Reports on post-AVR dissection in patients with tricuspid aortic valve disease and after combined procedures were excluded. Papers reporting aortic dissection within 14 days of AVR were also excluded. The primary endpoint was the risk of post-AVR dissection in BAV insufficiency versus BAV stenosis.
All analyses were conducted using Comprehensive Meta-Analysis version 2 (Biostat, Englewood, New Jersey). Events were compared as odds ratios by using a 95% confidence interval. A random effects model was used to derive pooled estimates. Study-specific estimates were calculated on the assumption that type A dissection develops in 0.6% of patients after AVR (3) and that patients undergoing surgery for BAV disease are 6 to 8 times more likely to present with aortic stenosis than with insufficiency (4). Heterogeneity was evaluated, and a sensitivity analysis was performed. Fixed-effect meta-regression was performed to determine the modulating effect of arterial hypertension/aortic diameters. Publication bias was evaluated using a funnel plot and Begg and Mazumdar’s rank correlation test.
The keyword-based search revealed 17,068 potential publications. After removal of 3,056 duplicate studies, 14,012 papers were screened using the abstract. Abstract screening yielded 296 relevant papers for full-text review, which revealed 44 eligible papers. Corresponding authors of 30 publications, who were blinded to our study endpoint, were contacted to obtain missing data on BAV phenotype. Thirteen studies with 20 patients were further excluded due to no answer from authors (n = 8), no access to database (n = 4), or no database entries on BAV phenotype (n = 1). A total of 31 observational studies with 79 patients undergoing aortic dissection were thus included in the final analysis.
Fifty-six patients (71%) underwent previous AVR for BAV insufficiency, whereas 23 (29%) underwent AVR for BAV stenosis (p < 0.01). The pooled estimate of dissection rate was 2.8% in the BAV insufficiency versus 0.2% in the BAV stenosis cohort (p < 0.01). Proximal aortic diameter at the time of AVR was available in 65 patients (82%). All but 1 patient had an aortic diameter ≥40 mm. There was a tendency toward larger aortic diameters in BAV stenosis versus BAV insufficiency (52 ± 5 mm vs. 46 ± 5 mm; p = 0.06). One patient (5%) in the BAV stenosis group had an aortic diameter ≤45 mm compared with 20 patients with BAV insufficiency (44%; p = 0.03).
BAV insufficiency was associated with a 10-fold higher risk of post-AVR aortic dissection compared with BAV stenosis (odds ratio: 10.0; 95% confidence interval: 6.2 to 16.2; p < 0.001) (Figure 1). No evidence of between-study heterogeneity (I2 = 0%; p = 0.9) or publication bias (p = 0.4) could be shown. Exclusion of any single trial did not alter the pooled effect result. Meta-regression coefficients were not significant for hypertension (p = 0.8) or aortic diameter (p = 0.8).
Most reports of post-AVR aortic dissection include a limited number of patients and do not differentiate between BAV phenotypes (1). One systematic review with 14% of BAV patients found that aortic insufficiency and fragility/thinning of the aortic wall were predictive of post-AVR dissection (4).
Study limitations include the restricted focus on BAV function only (i.e., excluding other potentially important features) and the relevant rate of missing data (i.e., regarding BAV morphotype and aortic diameters).
Our meta-analysis revealed a 10-fold higher risk of aortic dissection in patients who undergo AVR for BAV insufficiency compared with BAV stenosis. Moreover, the smaller aortic diameters in patients with BAV insufficiency indicate an increased risk of dissection at smaller diameters in this BAV cohort. In contrast, BAV stenosis–associated aortopathy seems to follow a more benign course post-AVR. Such information may be helpful when deciding on management of the aorta in BAV patients undergoing AVR surgery.
Please note: The authors have reported that they have no relationships relevant to the contents of this paper to disclose. These data were originally presented at the 44th Annual Meeting of the German Society for Thoracic and Cardiovascular Surgery, February 8–11, 2015, Freiburg, Germany.
- American College of Cardiology Foundation