Page 50 - Read Online

Page 4 of 11 Perezgrovas-Olaria et al. Vessel Plus 2023;7:10 https://dx.doi.org/10.20517/2574-1209.2022.54

Multivariable analysis was performed to identify associations with follow-up mortality and reoperation;
results were reported as hazard ratios (HR) with their 95% confidence intervals (CI). Included variables
were age, sex, chronic obstructive pulmonary disease (COPD), diabetes, urgent or emergent operation,
preoperative renal dysfunction, the New York Heart Association (NYHA) class, underlying connective
tissue disorders, and type of composite valve graft used. A 2-tailed p-value threshold of 0.05 was used to
determine statistical significance without multiplicity adjustment.

Because of the heterogeneity in patient characteristics among the mCVG and bCVG groups, propensity
score matching (PSM) was used to adjust for baseline differences and reduce confounding. The probability
of being assigned to different surgical treatments was calculated from demographic and preoperative
patients’ characteristics; the most clinically important variables were then entered into the PSM model.
Selected variables were age, family history of aortic disease, hypertension, diabetes, urgent/emergency
procedure, previous open-heart surgery, dissection at presentation, aneurysm size, connective tissue
disorders, and the New York Heart Association (NYHA) class. These covariates were used to compare both
surgical techniques by logistic regression algorithm in 1-1 PSM. The nearest neighbor matching algorithm
without replacement and a caliper size of 0.10 was used. Propensity matching models were assessed using
balance diagnostics and standardized mean differences (SMD), with SMD < 0.10 reflecting a proper balance
between groups.

Data were stored using Microsoft Access 2010 (Microsoft) and analyzed using R version 3.6.2 (R
Foundation for Statistical Computing) within RStudio. Tableone, Survival, Survminer, and MatchIt were
used.

RESULTS
From May 1997 to December 2019, 1,210 patients underwent aortic valve and root replacement with a
Bentall-De Bono operation at our institution. Of these, 798 (65.9%) received a bCVG, while 412 (34.1%)
received a mCVG. Among those who received a bCVG, 759 (95.1%) had a bovine valve, while 39 (4.9%) had
a porcine prosthesis. The mean follow-up time was 6.64 ± 0.21 years.

Baseline characteristics
Compared to patients with bCVG, those with mCVG were younger (P < 0.001), with a positive family
history of aortic disease (P = 0.007). Patients with mCVG had a higher prevalence of connective tissue
disorders (P < 0.001) and were more likely to have undergone previous open-heart surgery (P < 0.001). They
were also more likely to undergo urgent or emergent procedure (P < 0.001). Patients with bCVG were older,
with a higher prevalence of hypertension (P < 0.001) and diabetes (P = 0.03). They were also more likely to
present with a worse NYHA functional class (P < 0.001). Details of the preoperative variables are
summarized in Table 1.

Intraoperative characteristics
At surgery, deep hypothermic circulatory arrest was more frequently used in patients with mCVG
(P < 0.001). This group also had longer cardiopulmonary bypass time (P < 0.001) and cardio-ischemic
(circulatory arrest + aortic cross-clamp) time (P < 0.001). Details of the intraoperative data are available in
Table 2.

In-hospital outcomes
Overall, operative mortality was 0.7%, with no significant difference between groups (0.4% for bCVG vs.
1.2% for mCVG, P = 0.18). MAEs occurred in 6.2% of patients and no significant difference was found