Results
Preoperative patient characteristics of all 314 patients are displayed in Table 1. In 70% of cases, dissection involved all major segments of the aorta. Malperfusion of one or more end organs was observed in 107 (34%) patients, of which 45 (14%) had coronary malperfusion. Sixty-five patients (21%) required preoperative inotropic support and cardiopulmonary resuscitation was performed in 27 cases (9%). There were 24 patients (8%) with confirmed connective tissue disorders and 46 (15%) with bicuspid aortic valve.
The total number of Bentall procedures performed per year rose from 4 - 7 cases in 1996 – 2000 to 17 - 26 interventions in 2014 – 2018, due to overall increase of ATAAD cases undergoing surgical treatment at our institution. At the same time, a decrease in operative mortality was observed: from 22% mean in-hospital mortality in the earlier years (1996 – 2002), to 19% in 2003 – 2010 and 14% in 2011 – 2018.
Initially, Bentall procedures were more commonly performed using mechanical valve conduits (1996 – 2000), followed by introduction of xeno- and homograft roots (with the peak in 2008 – 2009) and biological valve conduits (used in the majority of cases since 2015) (Figure 1). For mechanical Bentall procedure we used prefabricated conduits in most of the cases: ATS Aortic Valves Graft Model 502 AG (ATS Medical, Minneapolis, MN, USA) was implanted in 80% patients and St. Jude Medical Aortic Valved Graft (St Jude Medical, Inc, St Paul, MN, USA) was used as a second option (18%) (Figure 2). BioBentall procedures were performed using intraoperative creation of a valved conduit consisting of a stented biological valve (Carpentier-Edwards Perimount Aortic Model, Edwards Lifesciences LLC, Irvine, CA, USA in the 90% of cases) and a Dacron aortic graft (Hemashield, Maquet, Wayne, NJ, USA). In several cases a prefabricated Vascutek Biovalsalva Conduit (Vascutek Terumo, Inchinnan, Scotland, UK) was used. Xenograft aortic root conduits included Medtronic Freestyle in 93% (Medtronic Inc, Minneapolis, MN, USA) and St. Jude Medical Toronto Root (St Jude Medical, Inc, St Paul, MN, USA) in 6% cases. In one patient an aortic valve conduit homograft was used.
Intraoperative characteristics are presented in Table 2. Indications for Bentall procedure included extensive dissection involving the aortic root in the majority of cases, aortic valve calcification, and failure of supracoronary aortic replacement or valve-sparing procedure (Table 2). In 20 cases root replacement strategy was chosen due to patient’s age and comorbidities or based on the surgeon’s experience.
The use of tissue adhesive in order to compress the separated layers of the dissected coronary buttons (n = 34) was completely abandoned in our institution in 2015. Coronary artery patch plasty or bypass grafting was required in 52 cases (17%) due to dissection or rupture of the artery; concomitant bypass grafting for coronary artery disease was performed in 11 patients.
The most commonly performed distal aortic intervention was hemiarch replacement. In more recent years, the elephant trunk (n = 55) and frozen elephant trunk (n = 22) procedures were performed more commonly in DeBakey type I aortic dissection repair. In 4 cases, open implantation of uncovered aortic stents was performed in order to open the collapsed true lumen. Among other concomitant procedures were septal myectomy (n = 3), mitral or tricuspid valve repair (n = 3) and caesarian section (n = 1).
The Table 3 summarizes the early postoperative clinical outcomes after surgical treatment. The overall 30-day mortality in this group of patients was 26%. There were 53 (17%) hospital deaths, with 13 patients who died intraoperatively. The causes of hospital death included intractable low cardiac output syndrome in half of the cases, major brain injury in 16 patients, multiorgan failure and sepsis in 6 and 2 cases, respectively. Twelve percent of patients underwent reexploration for bleeding. There were 47 (15%) cases of postoperative low cardiac output syndrome and 18 (6%) cases of myocardial infarction. A total of 69 (22%) of patients had permanent neurologic deficit: focal deficit was diagnosed in 25 and non-focal – in 30 cases; in 13 (4%) patients a newly developed permanent paraplegia or paraparesis were observed, one patient was discharged with monoparesis.
Multivariate logistic regression model revealed the following independent in-hospital death risk factors: critical preoperative state (OR, 5.6; p < 0.001), coronary malperfusion (OR, 3.6; p = 0.002), coronary artery disease (OR, 2.6; p = 0.033) and prior cerebrovascular accident (OR, 5.6; p = 0.002) (Table 4). The area under the ROC curve for this model was 79.5 (95% CI, 72.4 to 86.7) and the Hosmer-Lemeshow test was nonsignificant (p = 0.979) (Figure 3).