RESULTS
Preoperative characteristics of patients were shown in Table 1. past
history of cerebral infarction was observed in 16 (8%). Preoperative
malperfusion syndrome was observed in 54 (26%) and cardiogenic shock
was also observed in 55 (27%). They had no significant difference
between the two groups. Preoperative neurological symptoms were observed
in 14 (7%), significantly more frequent in the stroke group. Hemiplegia
was observed in 10 (5%) (including insufficiency), coma in 1 (0.5%),
delirium in 1 (0.5%), and syncope in 2 (1%). Preoperative hemiplegia,
coma, and delirium were significantly more frequently observed in the
stroke group.
Preoperative CT findings were shown in Table 2. Aortic dissection with a
patent false lumen was found in 73% (n=148), and 27% (n=54) were
intramural hematomas. Dissection of any supra-aortic vessels was
observed in 71% (n=142) of the patients, significantly more frequent in
the stroke group (P = .028). 64% (n=129) of patients had
dissection in the BCA, 38% (n=77) in the LCCA, and 35% (n=71) in the
LSCA. Only the BCA dissection was significantly higher in the stroke
group (P = .004). Of the 22 patients of BCA dissection with
postoperative stroke, 6 patients had a patent false lumen of BCA and 16
had a thrombotic false lumen of BCA.
Operative characteristics were shown in Table 3. Cardiopulmonary bypass
time in the stroke group was significantly longer than in no stroke
group. Lower body circulatory arrest time and lowest temperature did not
show a significant difference in two groups. Cerebral perfusion was
antegrade in 99%, and there was no significant difference in cerebral
perfusion time. Aortic range of replacement was also not different
significantly. Coronary artery bypass grafting was administered in 6%
(n=12), and aortic valve surgery or valve sparing was observed in 5%
(n=11), and there was no significant difference between the two groups.
Postoperative stroke was observed in 12% (n=25), operative mortality
was 6% (n=13) and the hospital mortality was 8% (n=16). In the 25
patients who developed stroke, brain CT and/or MRI showed that 12
patients had unilateral cerebral infarction (7 patients on the right and
5 patients on the left) and 13 patients had bilateral cerebral
infarction. In 7 patients in whom dissection did not reach LCCA,
cerebral infarction of left hemisphere was observed. In addition, Right
hemisphere cerebral infarction was observed in 3 patients without BCA
dissection. Of these 13 patients with bilateral cerebral infarction,
preoperative cardiac shock was observed in 8 patients, and preoperative
cardiac tamponade was observed in 8 patients. The mechanism of cerebral
infarction seemed to be an embolic infarction in 19 patients and a
watershed damage in 6 patients, judging by the distribution of
infarction. Other early results showed in Table 4.
In this study, we evaluated the changes of false lumen of the BCA in 21
of the 25 patients who developed postoperative stroke after ATAAD
repair. We could not perform postoperative CT in four patients. Of the
13 patients who had preoperatively thrombosed false lumen of the BCA,
false lumen remained thrombosed in 7 patients, healed in 2, and became
patent in 4 postoperatively. Other changes were described in Figure 2.
Combination of BCA and LCCA dissection was analyzed in Table 5. The
incidence of stroke in patients without BCA dissection was 4%(n=3),
while the incidence of stroke in patients with BCA dissection
accompanied with severe stenosis of true lumen was as high as 24%(n=5).
However, even in cases of BCA dissection without severe stenosis of true
lumen, the incidence of stroke was as high as 16%(n=17). On the other
hand, in cases where there was a dissection in LCCA but not in BCA,
there was no occurrence of stroke after ATAAD repair. Stroke occurred in
3 of 11 LCCA dissection cases with severe stenosis of true lumen, and
two of the three patients also had BCA dissection with severe stenosis
of true lumen.
On univariable analysis, BCA dissection, preoperative neurological
dysfunction, and longer CPB time were significant risk factors of stroke
after ATAAD repair. Only BCA dissection was an important risk factor for
stroke after ATAAD repair on multivariable analysis (Odds ratio 3.89,
95% CI: 1.104-13.780, P = .035). (Table 6)
Mid-term results of patients with or without stroke after ATAAD repair
were shown in Figure 3. The median follow-up period was 35 months (IQR,
9-53 months), and follow-up was completed in 95.5% of the patients.
Overall mortality was 11.9% (n=24). Overall survival was compared with
or without stroke after ATAAD repair. Overall survival was 80.0 ± 8.0%
vs 91.8 ± 2.1%, 72.7 ± 10.1% vs 91.8 ± 2.1%, 72.7 ± 10.1% vs 89.1 ±
2.6% after 1, 2, 3 years, respectively. Survival rate was significantly
lower in postoperative stroke group in mid-term survival (Hazard Ratio
2.95, 95% CI: 1.159-7.509, log-rank, P = .015).