Keywords
ascending aortic surgery - acute Type A dissection - surgeon specialization
Introduction
Acute Type A aortic dissection (ATAD) is a devastating vascular event with a high
pre-hospital and perioperative mortality ranging from 16 to 28%,[1]
[2] with a 1 to 2% hourly mortality rate.[3] Surgical mortality varies among different hospitals according to institutional volume
and surgeon experience.[4]
[5] Dissection events have been shown to peak during early morning (6 a.m.–12 p.m.)
and late afternoon hours.[6] Consequently, the majority of ATAD patients present at an unconventional time of
the day, having no prior diagnosis of aortic disease, and usually require emergent,
complex aortic replacement by an on-call cardiac surgeon. The complexity of ATAD surgery
further diverges from a standard ascending aortic procedure as coexisting complications
(e.g., malperfusion syndrome) demand adaptation and appropriate improvisation.
Some centers have instituted specialized aortic surgery teams that manage all emergent
aortic cases.[7] In the context of temporal improvement in outcomes of aortic surgery owing to technological
advancements (e.g., use of better grafts, availability of Bioglue [Cryolife, Kennasaw,
GA]),[8] a contemporary benchmark for aortic surgical outcomes incorporating surgeon experience
is needed.
We aimed to evaluate perioperative and 2-year outcomes associated with ATAD surgery
performed by aortic specialists (AS) compared with cardiac surgeons without extensive
aortic surgery background (non-AS).
Materials and Methods
We conducted a retrospective review of consecutive patients who were operated for
ATAD between 2008 and 2017 at Yale New Haven Hospital (YNHH). Out of 223 patients
presenting with acute aortic emergencies, 128 patients underwent ascending aortic
surgical intervention and had a complete record of admission-intervention timings
and outside medical reports (transferred patients). Further exclusions were made for
patients who were found to have chronic, rather than acute, ascending aortic dissection
(15 cases), traumatic aortic dissection (7 cases), or incomplete data (4 cases). The
final cohort consisted of 102 ATAD cases that were surgically managed at YNHH. The
cohort was divided into those operated by AS (3 surgeons) and non-AS (5 surgeons)
cardiac surgeons. An aortic surgery specialist was defined as a cardiac surgeon who
consistently performed ≥30 ascending aortic replacement cases annually during the
most recent 3 years of practice. To risk stratify the patients, multiple baseline
characteristics, presence of preoperative comorbidities, and complications due to
dissection were included in the analysis. Hypotension was defined as a systolic blood
pressure less than 90 mm Hg. Malperfusion was confirmed according to the clinical
ruling of the emergency care provider and cardiac surgeon according to clinical symptoms
and examination (e.g., pulselessness, pallor, sudden onset of abdominal pain, diarrhea,
neurological deficits) at the YNHH. For patients who were directly transferred to
the operating room from outside medical facilities, their diagnosis was confirmed
through respective hospital record or by surgeon examination before the procedure
was performed. The Yale Institutional Review Board approved this study (HIC # 2000021950)
and individual patient consent was waived.
Surgeon Experience
Surgeons in the AS and non-AS groups had a mean age of 52 ± 13.1 and 49 ± 2.5 years,
respectively. The mean annual ascending aortic case volume for the AS group was 32.3 ± 2.1,
contrasting to 6 ± 1.8 cases performed by the non-AS group. Surgeons in the AS group
had a mean experience of 23.6 ± 10.3 years in cardiac surgery practice compared with
20.3 ± 6.0 years for the non-AS group.
Outcome Definition
Perioperative outcomes evaluated were mortality and composite adverse events, which
consisted of mortality and stroke. The primary outcome of interest in the time-dependent
model postdischarge was death. The long-term follow-up and mortality data were acquired
through patient follow-up records according to the methodology described by Peterss
et al.[9]
Statistical Analysis
Categorical variables were compared using chi-square test to analyze significance
of proportions in the AS and non-AS group. Continuous variables were represented as
mean ± standard deviations and compared using two-tailed t-test. Multivariable logistic regression models were fitted to evaluate associations
between the surgeon training experience and perioperative mortality. A Cox proportional
hazard model was fitted to evaluate associations between the surgeon training experience
and 2-year survival. Proportional hazard assumption was evaluated with log–log curve,
and covariates not satisfying the assumption were stratified in the Cox model. Kaplan–Meier
and log rank analysis was performed to compare 2-year mortality rates between the
two groups. SAS software Version 9.4 (SAS Institute Inc., Cary, NC) was used for the
analysis. Statistical significance was set at p < 0.05.
Results
The study population consisted of 102 patients, an aggregate of 70 (68.9%) male and
32 (31.1%) female patients. The overall mean age of subjects at the time of surgery
was 59.2 ± 14.3 years and did not differ significantly between the AS and non-AS groups
([Table 1]). Out of 102 patients, 60 patients were operated by AS and 42 by non-AS. Preoperative
characteristics ([Table 1]) were similar between patients operated by AS and non-AS, except for a higher frequency
of chronic obstructive pulmonary disease (COPD) in the non-AS group (26.2% vs. 6.7%,
p = 0.009). Severe aortic insufficiency was demonstrated in 19.7% of AS and 33.3% of
non-AS on the preoperative echocardiogram (p = 0.12).
Table 1
Preoperative/baseline patient characteristics for the aortic specialist and non-aortic
specialist group
|
Variables
|
Non-AS (n = 42)
|
AS (n = 60)
|
p-Value
|
|
Mean (SD) age, y
|
59.2 ± 14.6
|
59.2 ± 14.3
|
1.0
|
|
Male
|
30 (71.4%)
|
40 (66.7%)
|
0.7
|
|
Race:
|
|
|
|
|
Caucasian
|
32 (76.2%)
|
40 (66.7%)
|
0.3
|
|
African-American
|
8 (19.1%)
|
11 (18.3%)
|
|
Other
|
2 (4.8%)
|
9 (15.0%)
|
|
Mean (SD) BMI (kg/m2)
|
30.4 ± 5.5
|
28.5 ± 6.4
|
0.13
|
|
Transfer from outside facility
|
22 (52.4%)
|
35 (58.3%)
|
0.7
|
|
Time from admission to CT (min)
|
113.72 ± 107.1
|
123 ± 186.09
|
0.77
|
|
Admission-incision interval (h)
|
5.68 ± 3.5
|
5.88 ± 4.7
|
0.81
|
|
Baseline comorbidities:
|
|
|
|
|
Hypertension
|
38 (90.5%)
|
49 (81.7%)
|
0.3
|
|
Mean (SD) eGFR
|
39.0 ± 33.2
|
37.3 ± 31.2
|
0.8
|
|
COPD
|
11 (26.2%)
|
4 (6.7%)
|
0.009
|
|
Coronary artery disease
|
9 (21.4%)
|
5 (8.3%)
|
0.08
|
|
Previous PCI
|
1
|
0
|
1.0
|
|
Previous CABG
|
3
|
1
|
1.0
|
|
Medication only
|
5
|
4
|
1.0
|
|
History of stroke
|
4 (9.5%)
|
1 (1.7%)
|
0.2
|
|
Previous cardiac surgery
|
3 (7.1%)
|
2 (3.3%)
|
0.4
|
|
Prior AVR
|
0
|
1
|
1.0
|
|
Prior CABG
|
3
|
0
|
|
Mean (SD) hematocrit (%)
|
37.5 ± 8.1
|
38.6 ± 5.9
|
0.5
|
|
Mean (SD) platelets
|
222.2 ± 103.7
|
205.9 ± 75.8
|
0.4
|
|
Dissection-related variables:
|
|
|
|
|
DeBakey Type I
|
35 (83.3%)
|
47 (78.3%)
|
0.6
|
|
Hypotension
|
5 (11.9%)
|
5 (8.3%)
|
0.7
|
|
Bicuspid aortic valve
|
3 (7.1%)
|
3 (5.0%)
|
0.7
|
|
Aortic arch involvement
|
39 (92.9%)
|
51 (85.0%)
|
0.4
|
|
Abdominal aorta involvement
|
25 (59.5%)
|
34 (56.7%)
|
0.8
|
|
Iliac artery involvement
|
17 (40.5%)
|
23 (38.3%)
|
0.8
|
|
Bovine arch
|
5 (11.9%)
|
10 (16.7%)
|
0.6
|
|
Malperfusion present
|
12 (28.6%)
|
17 (28.3%)
|
1.0
|
|
Cardiac tamponade
|
9 (21.4%)
|
17 (28.3%)
|
0.5
|
|
Aortic valve involvement
|
12 (28.5%)
|
12 (20%)
|
0.29
|
|
Aortic root involvement
|
28 (66.6%)
|
34 (56.6%)
|
0.26
|
Abbreviations: AS, aortic specialist; AVR, aortic valve replacement; BMI, body mass
index; CABG, coronary artery bypass grafting; COPD, chronic obstructive pulmonary
disease; CT, computed tomography; eGFR, estimated glomerular filtration rate; PCI,
percutaneous coronary intervention; SD, standard deviation.
Note: Boldfaced values represent p < 0.05.
The AS group performed more root replacements (41.6% vs. 23.8%, respectively, p = 0.049) with higher rates of composite (Bentall procedure) root replacements than
the non-AS group (Bentall procedure, 26.7% vs. 2.4%, p < 0.001). The AS group also employed frequent adjunct cerebral perfusion in cases
with circulatory arrest (58.8%) compared with the non-AS group (14.3%) (p < 0.001) ([Table 2]).
Table 2
Operative management performed by the aortic specialist and non-aortic specialist
group
|
Variables
|
Non-AS (n = 42)
|
AS (n = 60)
|
p-Value
|
|
Root replacement
Bentall is a different variable
|
10 (23.8%)
1 (2.4%)
|
25 (41.6%)
16 (26.7%)
|
0.049
< 0.01
|
|
Valve-sparing
|
0
|
3
|
0.3
|
|
Hemiarch replacement
|
37 (88.1%)
|
50 (83.3%)
|
0.6
|
|
Total arch replacement
|
1 (2.4%)
|
6 (10.0%)
|
0.2
|
|
Descending procedure
|
0 (0%)
|
4 (6.7%)
|
0.1
|
|
Concomitant CABG
|
4 (9.5%)
|
3 (5.0%)
|
0.4
|
|
DHCA use
|
42 (100%)
|
59 (98.3%)
|
1.0
|
|
Mean (SD) DHCA time, min
|
27.5 ± 12.5
|
28.9 ± 13.0
|
0.6
|
|
Adjunct cerebral perfusion
|
6 (14.3%)
|
39 (58.8%)
|
< 0.001
|
|
Antegrade
|
1
|
14
|
|
|
Retrograde
|
5
|
25
|
|
|
Mean (SD) Cx time, min
|
103.7 ± 45.2
|
107.0 ± 40.7
|
0.7
|
|
Mean (SD) CPB time, min
|
195.6 ± 53.5
|
185.8 ± 46.6
|
0.3
|
Abbreviations: AS, aortic specialist; CABG, coronary artery bypass grafting; CPB,
cardiopulmonary bypass; Cx, aortic cross-clamp; DHCA, deep hypothermic circulatory
arrest; SD, standard deviation.
Note: Boldfaced values represent p < 0.05.
In the entire cohort, mortality within 30 days of procedure occurred in 11 patients
(10.8%), with 4 (6.6%) deaths in the AS group and 7 (16.6%) among the non-AS group
(p = 0.20). Logistic regression adjusting for 16 preoperative variables demonstrated
that AS-status was not significantly associated with any difference in 30-day mortality
(odds ratio [OR]: 0.32, 95% confidence interval [CI]: 0.07–1.30, p = 0.11). Four (9.5%) patients in the non-AS group and 7 (11.7%) patients in the AS
group suffered from postoperative stroke (p = 1.0). Improving renal function (estimated glomerular filtration rate) was protective
of 30-day mortality (OR: 0.95, 95% CI: 0.91–0.99, p = 0.03) and concomitant coronary artery bypass grafting operation was associated
with increased risk of postoperative early mortality (OR: 8.62, 95% CI: 1.23–60.53,
p = 0.03).
A multivariate model assessing perioperative composite adverse events (30-day death
and stroke) did not show significant association with surgeon specialty. The presence
of preoperative malperfusion (OR: 3.19, 95% CI: 1.10–9.24, p = 0.03) and kidney injury (each unit increase in eGFR [OR: 0.97, 95% CI: 0.94–0.99,
p = 0.02]) were associated with increased risk of perioperative composite adverse event.
Overall early postoperative outcomes ([Table 3]) were comparable between the two groups. Median follow-up duration of patients in
the non-AS and AS group was 20.1 (interquartile [IQ] range: 1.3–58.2) and 34.9 (IQ
range: 5–58.2) months, respectively. The 6-month and 1-year survival for the AS group
was 93.3 ± 3.3% and 88.7 ± 4.4%, respectively. For the non-AS group, 6-month and 1-year
survival rate was 78.1 ± 6.5% and 78.1 ± 6.5%, respectively.
Table 3
Postoperative outcomes defined according to the aortic specialist and non-aortic specialist
group
|
Variables
|
Non-AS (n = 42)
|
AS (n = 60)
|
p-Value
|
|
30-d mortality
|
7 (16.7%)
|
4 (6.7%)
|
0.2
|
|
Death < 24 h of surgery
|
0 (0%)
|
1 (1.7%)
|
1.0
|
|
Revision for bleeding
|
7 (16.7%)
|
8 (13.3%)
|
0.8
|
|
Stroke
|
4 (9.5%)
|
7 (11.7%)
|
1.0
|
|
Dialysis need
|
1 (2.3%)
|
1 (1.7%)
|
1.0
|
|
Mechanical ventilation > 48 h
|
17 (40.5%)
|
18 (30.0%)
|
0.3
|
|
Sepsis
|
3 (7.1%)
|
2 (3.3%)
|
0.4
|
|
Late aortic reoperation
|
4 (9.5%)
|
5 (8.3%)
|
1.0
|
|
Root/arch reoperation
|
4 (9.5%)
|
2 (3.3%)
|
0.22
|
|
Descending reoperation
|
0
|
3
|
−
|
|
Cause of death:
|
|
|
|
|
Cardiogenic shock
|
4 (9.5%)
|
0 (0%)
|
0.03
|
|
Heart failure
|
1 (2.3%)
|
1 (1.7%)
|
|
Malperfusion
|
0
|
1 (1.7%)
|
1
|
|
Multiorgan failure
|
2 (4.7%)
|
0
|
0.16
|
|
Stroke
|
0 (0%)
|
2 (3.3%)
|
Abbreviation: AS, aortic specialist.
Log rank analysis revealed a statistically significantly superior 2-year survival
associated with surgery by the AS group (p = 0.026) ([Fig. 1]). Cox proportional hazard analysis also demonstrated AS-status as an independent
predictor of improved survival at 2 years (hazard ratio: 0.37, 95% CI: 0.15–0.92,
p = 0.03).
Fig. 1 Kaplan–Meier curve demonstrating difference in 2-year survival among patients operated
on by the aortic specialist (AS) group compared with the non-AS group (p = 0.026).
Discussion
The overall 30-day mortality for ATAD in this study was 10.8% with 6.7% perioperative
deaths in the group operated by the AS group. The stroke rate was comparable in the
AS and non-AS groups even though the AS group adopted a higher rate of adjunct neuroprotective
perfusion. Literature on cerebral protection is variable and adjunct perfusion may
not be superior than isolated deep hypothermic circulatory arrest in ATAD cases; however,
this observation is beyond the scope of this article. Concisely, patients operated
on by the AS group had a significantly lower adjusted risk of 2-year mortality.
Recently, Bashir et al demonstrated improved outcomes of aortic surgery for ATAD associated
with higher volume aortic surgeons based on the national U.K. data.[10] Their study demonstrated that surgeons performing on average ≥ 4 cases/year had
lower in-hospital mortality than surgeons performing < 4 cases annually (12.6 and
19.3%, respectively, p = 0.015).[10] Similarly, Andersen et al[ 7] demonstrated a considerable decrease in perioperative mortality by around 25% after
initiation of an exclusive aortic team in a single-center setting. This program restricted
all thoracic aortic surgery to be performed by cardiac surgeons specialized in aortic
surgery, leading to an improvement in operative mortality.[7] Another study looking at the volume outcome relations in “all-comer” (elective + emergent)
ascending aorta/arch surgery highlighted lower midterm mortality risk for patients
operated on by a high volume operator (HR: 0.67, 95% CI: 0.51–0.88).[11] Analysis of the national (U.S.) patient population by Hughes et al[12] revealed that patients operated at high volume aortic centers (30–100 cases/year)
have considerably lower adjusted risk of mortality than low volume centers (< 6 cases/year)
for elective ascending aortic surgery (OR: 0.42, 95% CI: 0.31–0.58). Notwithstanding
this correlation (for all aortic surgery cases) of short-term (30-day) mortality with
surgeon status, some results were inconsistent for emergent procedures and need further
exploration in this specific group.[11] Increasing evidence supports experience–outcome relationships in high-risk operations,
and this relationship is also expected to be reflected in emergent aortic surgery.[7]
[10]
[12]
A potential solution advocated by some centers is for referral of all aortic cases
to centers specialized for aortic surgery. Volume centralization is expected to improve
outcomes but this may be difficult given the emergent nature of aortic dissection
and difficulty in transferring/triaging these patients in a timely fashion. Case specialization
within a given center may be a feasible alternative that improves the operative outcomes.
We have implemented and also advocated the development of a designated “on-call aortic
surgery” team. This involves an exclusive on-call AS in addition to the regular cardiac
surgery attending staff. A protocol declaring the precise algorithm for rapid engagement
of various additional members (nurses, physician assistants, perfusionists) on the
team is also required for an efficient coordination. Our own protocol has proven to
be very efficient for rapid triage and management of emergent aortic cases (dissection
and rupture).[13]
Although there were no major differences in the baseline characteristics or preoperative
comorbid conditions between the two groups, there were some striking distinctions
in the way these patients were managed. The AS group performed a significantly higher
number of Bentall procedures than the non-AS group and also made relatively liberal
use of adjunct cerebral perfusion. These choices could be partly explained in terms
of greater surgeon comfort to adopt a more complex technique according to experience.
It was beyond the scope of this article to measure the exact factors leading to improved
outcomes in the AS group. However, the results from this study argue for establishing
thoracic aortic emergency on-call teams staffed by AS. This can be an efficient approach
to manage all ATAD cases optimally; however, this methodology may only be practical
at large centers that can afford a focused aortic team.
Limitations
This study harbors limitations inherent to the single-center and retrospective design.
The cohort size is relatively small due to the stringent exclusion criteria, and this
likely limited a robust multivariable adjustment in the models, although statistical
significance was reached in several of the variables outlined above. Selectivity of
cases by each surgeon category could not be assessed, as the cohort consisted of only
those who underwent the operation, although this single-center setting likely limited
this heterogeneity. Furthermore, higher rates of baseline COPD and aortic insufficiency
in the non-AS group likely prejudiced this sample for a dismal outcome.
Conclusion
AS status of surgeons performing ATAD surgery defined improved 2-year survival for
patients presenting with acute thoracic aortic emergencies. This argues for aortic
surgery centers to form dedicated thoracic aortic emergency teams so that surgeons
with aortic expertise are available to manage emergent cases at all times.
