Baseline characteristics and outcomes of cardiac BS patients
There were 66 cardiac BS patients [aged 43.0 (31.5-48.8) years, 84.85% male] selected from 1950 patients in the Shanghai Behçet’s syndrome database. After deduplication, 1467 BS patients remained in the database. The proportion of cardiac BS in total BS was 4.50% (66/1467) from Shanghai Behçet’s syndrome database. The median follow-up time was 4.0 (2.0-7.0) years, ranging from 1.0 to 9.5 years. Baseline characteristics and outcomes of the 66 cardiac patients are summarized in Table 1. The number of male cardiac BS patients was 5.6 times (56/10) the number of female patients. In the 66 patients, 47 patients were definite cardiac BS patients, and 19 patients were suspected cardiac BS patients.
In the follow-up duration, 54 (81.82%) patients were in stable condition. Death occurred in 7 (10.61%) patients, among whom, cardiac death occurred in 5 (7.58%) patients, 1 (1.52%) patient died of severe pneumonia, 1 patient died of ruptured aneurysm of ascending aorta. Five (7.58%) patients underwent rehospitalization owing to worsened cardiac lesions or postoperative complications (heart failure in non-operated patients with severe AR: 4 patients, postoperative paravalvular leakage: 1 patient) (Table 1, Figure 1a).
Twenty-two (33.33%) patients experienced two or more surgeries, of whom 2 patients underwent 4 surgeries and 4 patients had 3 surgeries. Notably, the outcomes of each surgery were different, even in the same patient. To find the influencing factors of the prognoses of cardiac BS patients, we observed the outcome of every surgery in each patient. And the 66 patients were collected as 96 cases. In the 96 cases, 35 cases were rehospitalized due to worsened cardiac lesions (4 cases) or postoperative complications (31 cases) (Table 1, Figure 1b).
Preoperative clinical characteristics of the 96 cases of cardiac BS patients
In the 96 cases, favourable prognoses were observed in 54 cases, and 42 cases ended with poor prognosis. No difference in prognoses was found in definite cardiac BS patients and suspected cardiac BS patients. The smoking status was different in patients with favourable or unfavourable prognoses. The case numbers of current and former smokers with unfavourable prognoses were significantly larger than cases with favourable prognoses (P = 0.007 and P = 0.004, Table 2). Compared to cases with unfavourable prognoses, more cases with palpitation before surgery achieved favourable prognoses (P = 0.043, Table 2). And basic diseases, including hypertension, diabetes, renal insufficiency and hyperlipidemia, did not influence the outcomes of cardiac BS patients in the current study (all P > 0.05, Table 2). NO statistical difference was found in the levels of WBC, Hb, PLT, NLR and ESR between groups with opposite outcomes (Table 3). The CRP levels in cases with unfavourable prognoses were significantly higher than that in cases with favourable prognoses (P = 0.007, Table 3).
Surgical procedures
The surgical procedures significantly influenced the prognoses of cardiac BS patients (P = 0.005, Table 4). The outcomes of cases undergoing the Bentall procedure were significantly better than those who did not (P = 0.003). Cases that performed cardiac valve replacement or repairment operations suffered worse outcomes than cases treated by other surgical procedures (P = 0.000). Besides, it showed that the operation frequency did not influence the outcomes of cardiac BS patients (P = 0.655, Table 4).
Immunosuppressive therapies
Both preoperative and postoperative immunosuppressive therapies could promote favourable outcomes for cardiac BS patients (both P = 0.000, Table 5). And the treatment of preoperative and postoperative JAK inhibitor or biological agent seemed to be conducive to a good outcome (P = 0.044 and P = 0.000, Table 5). Regularly used postoperative biologics included: tocilizumab (n=1) and golimumab (n=1) ending with unfavourable prognoses, infliximab (n=1), adalimumab (n=3), etanercept (n=2) and other anti-TNF fusion protein (n=10) achieving good outcomes. The median observation time of postoperative biologics was 24.0 (15.5-38.0) months. The effective rate of postoperative biological agents was 88.89% (16/18, Figure 2a). Biological agents failed in two patients. One patient redid operation after using tocilizumab for the sake of postoperative PVL after aortic valve replacement (4 months later). The other patient using golimumab underwent reoperation due to postoperative cardiac vascular pseudoaneurysm after aortic arch aneurysm embolization (9 months later). No adverse reaction was observed in cardiac BS patients treated by biologics.
The effective rate of postoperative JAK inhibitor in cardiac BS patients was 81.25% (13/16, Figure 2b). JAK inhibitors included tofacitinib (11 patients with good outcome, 3 patients with poor outcome) and baricitinib (2 patients with good outcome). The median observation time was 16.5 (12.5-22.5) months. No patient underwent secondary surgery. But, after using tofacitinib for one year, one patient died from severe pneumonia 2 months after the Bentall procedure and one patient died of sudden cardiac death 1 year after the Cabrol procedure. Another patient with poor outcome after postoperative tofacitinib treatment occurred pseudoaneurysm 8 months after the Bentall procedure.
Predictors of unfavourable prognoses in 96 cases of cardiac BS
By multivariate logistic regression analysis, former smoking (OR: 8.17, 95% CI: 1.84-36.32, P = 0.006), valve replacement or repairment operation (OR: 13.49, 95% CI: 2.64-68.97, P = 0.002) and poor compliance (OR: 7.50, 95% CI: 1.42-39.62, P = 0.002) were found the independent risk factors of unfavourable prognoses in cardiac BS. Preoperative immunosuppressive therapies (OR: 0.23, 95% CI: 0.06-0.90, P = 0.035) and postoperative immunosuppressive therapies (OR: 0.22, 95% CI: 0.05-0.95, P = 0.043) were independent protective factors for unfavourable prognoses in cardiac BS (Table 6). The AUC of the predicting model of unfavourable prognoses in cardiac BS was 0.92 (95% CI: 0.86-0.98, P = 0.000), indicating a good accuracy of the model.
Predictors of unfavourable prognoses in 65 cardiac BS cases who underwent surgery for severe AR.
Former smoking (OR: 12.49, 95% CI: 1.04-149.94, P = 0.046) was an independent risk factor of unfavourable prognoses in cardiac BS patients undergoing surgeries for severe AR. Bentall procedure (OR: 0.07, 95% CI: 0.01-0.78, P = 0.031), postoperative DMARDs (OR: 0.02, 95% CI: 0.00-0.26, P = 0.002) and postoperative JAK inhibitor or biological agent (OR: 0.16, 95% CI: 0.03-0.94, P = 0.043) were independent protective factors for unfavourable prognoses in BS patients after aortic valve surgeries (Table 7). The predicting model of unfavourable prognoses in cardiac BS patients undergoing surgeries for severe AR had a satisfactory accuracy with AUC equalling to 0.93 (95% CI: 0.88-0.99, P = 0.000).