Keywords
aorta/aortic - aneurysm - surgery - complications - cardiac
Introduction
Total arch replacement with the frozen elephant trunk (FET) surgery has changed the
landscape of treatment for complex aortic arch diseases. FET not only improves aneurysm
exclusion and intimal tear coverage, both of which promote distal aortic remodeling,
but it also facilitates vascular reintervention when required.[1]
There is no clear consensus on the optimal FET landing zone. FET was traditionally
anastomosed in zone 3 of the Ishimaru zones.[2] Many studies have recently compared the outcomes of patients whose FET device landed
in zone 2 instead of zone 3[3]
[4]
[5]; moreover, a zone 0 repair has been advocated.[6] The term “proximalization” refers to the process of moving the FET landing zone
forward. FET proximalization reduces surgical trauma, visceral ischemic time, and
descending thoracic aorta coverage, all of which may be associated with better clinical
outcomes.[3]
[4] However, there is a possibility of suboptimal aortic remodeling if this approach
is used.[5] This study aimed to compare the clinical outcomes of proximal (zone 0 and 1) and
distal (zone 2 and 3) FET anastomosis and to report our early and midterm results
of patients who received total arch replacement with FET procedure.
Patients and Methods
Study Design
This was a single-center retrospective observational study. One hundred consecutive
patients who underwent total arch replacement using the FET technique for acute and
chronic aortic arch pathologies were identified from the hospital electronic database
between November 2014 and August 2021. The location of FET anastomosis over the aortic
arch was used to classify patients into two groups. FET was performed on 67 patients
in zone 0/1 and 33 patients in zone 2/3. To ensure comparability, variables including
age 70 or above, gender, concomitant procedure, and previous sternotomy were matched.
Thirty-one matched pairs were obtained. Early outcomes, including in-hospital mortality
and major morbidities, reassessment computed tomography (CT) findings, and aorta-related
events were examined. Our clinic regularly follows up patients. The most recent follow-up
for this analysis occurred in December 2021. After discharge, two patients continued
to receive follow-up care in their home country, their follow-up data were not available.
Therefore, complete follow-up was conducted in 97.8% of patients (90/92). A schematic
illustration of the study design and patient selection criteria is shown in [Supplementary Fig. S1] (available in the online version). The local institutional review board approved
this study (Reference number: UW 21–647) with patients' informed consent.
Definitions
Major morbidities included stroke with persistent neurological deficit, bleeding requiring
reexploration for hemostasis, spinal cord injury (SCI) with transient or permanent
paraplegia, hoarse voice with documented vocal cord palsy, prolonged ventilation with
a tracheostomy, acute kidney injury requiring dialysis, thromboembolic event other
than stroke (myocardial infarction, pulmonary embolism, and “trash foot” phenomenon),
and major infection (mediastinitis and deep sternal wound infection). Composite endpoint
was defined as resternotomy for bleeding, requirement for dialysis, and SCI. Acute
malperfusion syndrome referred to major organ or limb ischemia in patients with acute
aortic dissection. At least two of the three criteria (clinical, biochemical, or radiological)
had to be considered malperfusion, including the following conditions: renal (increased
creatinine, kidney hypoenhancement, and renal artery disruption/dissection), mesenteric
(abdominal pain, hyperlactatemia, metabolic acidosis, and mesenteric artery disruption/dissection),
limb (signs of acute limb ischemia and peripheral artery disruption/dissection), cerebral
(neurological deficit and head and neck vessel disruption/dissection), and coronary
(raised cardiac enzymes, electrocardiographic changes, and coronary artery disruption/dissection).
Aorta-associated mortality, unknown sudden death, death related to an aortic procedure,
and aortic reintervention were all examples of aorta-related events. Distal stent
graft-induced new entry (dSINE) was defined as new communications between true and
false lumens via an intimal tear adjacent to the distal end of the FET that was absent
before FET implantation.
FET Technique and Prosthesis Types
The E-VITA open plus/open neo prosthesis (Jotec Inc., Hechingen, Germany) or the Thoraflex
hybrid graft (Vascutek, Inchinnan, United Kingdom) were implanted in patients who
underwent total arch replacement and FET procedure. The operating surgeon chose the
stent graft at his or her discretion. The extent of arch resection and choice of FET
distal landing zone was determined by the presence of an intimal tear in the arch.
Otherwise, the FET distal landing zone was decided according to the surgeon's preference.
In cases of acute aortic dissection, graft oversizing was avoided. We oversized the
FET by 20% relative to the mean diameters of the true and false lumen over the proximal
thoracic descending aorta for chronic aortic dissection. To achieve a tight distal
seal for aortic aneurysms, the FET diameter was 20% greater than the aorta distal
to the aneurysm. For the length of the FET, the general rule of avoiding extensive
coverage of the descending thoracic aorta was followed. The length of FET was determined
by the length of the aneurysm to achieve complete exclusion of the aneurysmal sac
in aortic aneurysm.
Surgical Procedure
All cases were performed in the operating theater with on-request portable fluoroscopic
support. Perioperative cerebral spinal fluid drainage was not routinely performed.
Depending on the etiology, different cannulation strategies were adopted after systemic
heparinization. Axillary artery cannulation via an 8-mm Hemashield graft was preferred.
Routine carbon dioxide field flooding was used to reduce the risk of air embolism.
The left heart was vented by inserting a Medtronic DLP 16 Fr cannula through the right
superior pulmonary vein. The distal ascending aorta was cross-clamped when the heart
started to fibrillate during systemic hypothermia. After aortotomy, the heart was
arrested by infusing Custodiol HTK solution through the coronary ostia. The proximal
aorta was trimmed at the level of the sino-tubular junction and prepared using the
“sandwich” technique with Teflon felt strips or the adventitial inversion technique.
Upon reaching 22°C to 25°C, circulatory arrest and selective antegrade cerebral perfusion
(SACP) to all three head and neck vessels were started. Pharmacological cerebral protection
with steroids and mannitol was administered in selected patients. Head and neck vessels
were disconnected from the arch of the aorta. If the left subclavian artery origin
was located deep, a Hemashield graft extension was adopted to facilitate subsequent
anastomosis. Extra-anatomical aorta-left subclavian artery bypass was performed only
in the case of left upper limb malperfusion. The FET was deployed in the true lumen
and the collared skirt was fixed to the arch of the aorta using multiple interrupted
felt sutures. When a visual inspection was inadequate to distinguish between the true
and false lumen, a guidewire was advanced through the femoral artery for rail guidance.
Direct examination using an angioscope was rarely performed. Cardiopulmonary bypass
through the side branch of the graft and rewarming were initiated after reapplying
the cross-clamp to the proximal prosthetic graft. The aortic cross-clamp was removed
after the proximal end of the FET was anastomosed to the ascending aorta. Head and
neck vessels were anastomosed to the side branches of the FET and SACP was stopped.
Imaging
Contrast-enhanced CT images were evaluated before and after surgery. Within 3 to 6
months of discharge, the initial postoperative CT reassessment was arranged. An earlier
scan was considered within the same admission if clinically indicated. Using multiplanar
or three-dimensional reconstruction images, the extent of FET coverage was determined
with reference to the vertebral level ([Fig. 1]). The degree of aneurysm and false lumen thrombosis was assessed. At the distal
end of the FET stent, complete false lumen thrombosis was evaluated.
Fig. 1 A Patients' postoperative contrasted computed tomography (CT) after total arch replacement
and the frozen elephant trunk (FET) procedure. (A) Three-dimensional (3D) volume rendering of intrathoracic structures showing the
T8 coverage of the descending thoracic aorta by the stented portion of the FET in
a patient with zone 2 FET. (B) Sagittal image of a contrasted CT scan of a patient with zone 0 FET and a 100-mm
graft. The FET extended to the T4 level of the descending thoracic aorta.
Statistical Analysis
SPSS software version 28 was used to analyze all of the data (IBM Corporation, Armonk,
New York, United States). The Shapiro–Wilk test was used to determine whether the
distribution of continuous variables was normal. The mean and standard deviation of
normally distributed continuous data are reported, and between-group differences were
examined using the t-test. The median and interquartile range (IQR) are used to show nonnormally distributed
continuous variables, and the Mann–Whitney U test was used to analyze between-group differences. Numbers and proportions are used
to represent categorical variables. If the minimum predicted count was at least one,
the chi-square test was employed to compare categorical variables. Fisher's exact
test was used in the other cases. Kaplan–Meier survival estimates were used to analyze
the freedom from aorta-related events, and the log-rank test was performed to investigate
between-group differences. To identify independent risk factors for the composite
endpoint, binary logistic regression was used. Univariable testing was used as a prescreening
method, and variables with a p-value of < 0.25 were chosen for multivariable analysis. The enter method was used
to perform logistic regression. Statistical significance was defined as a p-value of < 0.05.
Results
Patient characteristics are summarized in [Table 1]. Operative and postoperative details are shown in [Table 2]. Before matching, when compared with zone 2/3 FET, patients with zone 0/1 FET had
a higher mean age (59.2 ± 12.7 vs. 56.0 ± 8.8, p = 0.035) and shorter circulatory arrest time (57 minutes [IQR, 48–75 minutes] vs.
73 minutes [IQR, 51–89.5 minutes], p = 0.032). Majority of patients had DeBakey I aortic dissection and more than half
of them had an intimal tear identified in the arch. After matching, the circulatory
arrest time was 12 minutes longer in the zone 2/3 group. Slightly more patients in
the zone 0/1 group received branched FET. However, the differences were not statistically
significant. There was no difference in the use of stent graft lengths greater than
150 mm between the two groups. The in-hospital mortality rates in patients with zone
0/1 FET and zone 2/3 FET were 3.2% versus 9.7%, respectively (p = 0.301). Zone 2/3 FET was associated with significantly higher rates of resternotomy
for bleeding (3.2% vs. 19.4%, p = 0.045) and SCI (0% vs. 12.9%, p = 0.039).
Table 1
Preoperative patient characteristics
|
Unmatched
Overall
(n = 100)
|
Matched pairs
|
p-Value
|
|
Zone 0/1
(n = 31)
|
Zone 2/3
(n = 31)
|
|
Mean age, y (SD)
|
58.2 (11.6)
|
53.6 (11.1)
|
56.5 (8.9)
|
0.260
|
|
Male, n (%)
|
85 (85)
|
29 (93.5)
|
29 (93.5)
|
> 0.99
|
|
Resternotomy, n (%)
|
27 (27)
|
10 (32.3)
|
10 (32.3)
|
> 0.99
|
|
MFS/LDS, n (%)
|
6 (6)
|
2 (6.5)
|
3 (9.7)
|
0.641
|
|
Hypertension, n (%)
|
66 (66)
|
8 (25.8)
|
11 (35.5)
|
0.409
|
|
Diabetes mellitus, n (%)
|
8 (8)
|
3 (9.7)
|
2 (6.5)
|
0.641
|
|
Chronic lung disease, n (%)
|
4 (4)
|
1 (3.2)
|
2 (6.5)
|
0.554
|
|
Creatinine > 115 μ mol/L, n (%)
|
33 (33)
|
9 (29.0)
|
11 (35.5)
|
0.587
|
|
Preserved LVEF, n (%)
|
94 (94)
|
3 (9.7)
|
1 (3.2)
|
0.301
|
|
Indication for FET, n (%)
|
|
- Acute aortic dissection
|
46 (46)
|
15 (48.4)
|
14 (45.2)
|
0.799
|
|
- Chronic aortic dissection
|
38 (38)
|
13 (41.9)
|
14 (45.2)
|
0.798
|
|
- Aortic aneurysm
|
16 (16)
|
3 (9.7)
|
3 (9.7)
|
> 0.99
|
|
DeBakey I, n (%)
- Intimal tear at arch
|
76 (90.5)
43 (56.6)
|
27 (87.1)
15 (55.6)
|
25 (80.6)
13 (52.0)
|
0.490
0.797
|
|
Presentation, n (%)
|
|
- Incidental finding
|
13 (13)
|
2 (6.5)
|
5 (16.1)
|
0.229
|
|
- Chronic pain
|
4 (4)
|
2 (6.5)
|
1 (3.2)
|
0.554
|
|
- Progression
|
33 (33)
|
12 (38.7)
|
10 (32.3)
|
0.596
|
|
- Acute pain
|
27 (27)
|
6 (19.4)
|
9 (29.0)
|
0.374
|
|
- Malperfusion
|
18 (18)
|
8 (25.8)
|
5 (16.1)
|
0.349
|
|
- Rupture/tamponade
|
5 (5)
|
1 (3.2)
|
1 (3.2)
|
> 0.99
|
Abbreviations: FET, frozen elephant trunk; LDS, Loeys–Dietz syndrome; LVEF, left ventricular
ejection fraction; MFS, Marfan syndrome; SD, standard deviation.
Note: Values are presented as mean (standard deviation) or n (%).
Table 2
Operative and postoperative data
|
Unmatched
|
Matched pairs
|
p-Value
|
|
Overall
(n = 100)
|
Zone 0/1
(n = 31)
|
Zone 2/3
(n = 31)
|
|
Emergency, n (%)
|
51 (51)
|
16 (51.6)
|
15 (48.4)
|
0.799
|
|
Axillary cannulation, n (%)
|
72 (72)
|
25 (80.6)
|
22 (71.0)
|
0.374
|
|
Bypass time, min (SD)
|
236 (57)
|
233 (51.3)
|
245 (60.1)
|
0.397
|
|
Aortic cross-clamp time, min (SD)
|
140 (50)
|
136 (40.9)
|
152 (38.8)
|
0.132
|
|
Circulatory arrest time, min (SD)
|
60 (28)
|
63 (18.8)
|
75 (30.4)
|
0.068
|
|
Branched FET, n (%)
|
91 (91)
|
30 (96.8)
|
26 (83.9)
|
0.086
|
|
Stent graft length ≥ 150 mm, n (%)
|
32 (32)
|
8 (25.8)
|
12 (38.7)
|
0.277
|
|
Concomitant procedure, n (%)
|
42 (42)
|
10 (32.3)
|
10 (32.3)
|
> 0.99
|
|
Postoperative ECMO, n (%)
|
2 (2)
|
1 (3.2)
|
1 (3.2)
|
> 0.99
|
|
Major complications, n (%)
|
32 (32)
|
4 (12.9)
|
15 (48.4)
|
0.002
|
|
- Stroke
|
7 (7)
|
1 (3.2)
|
1 (3.2)
|
> 0.99
|
|
- Resternotomy for bleeding
|
10 (10)
|
1 (3.2)
|
6 (19.4)
|
0.045
|
|
- Spinal cord injury
|
5 (5)
|
0
|
4 (12.9)
|
0.039
|
|
- Vocal cord palsy
|
7 (7)
|
1 (3.2)
|
4 (12.9)
|
0.162
|
|
- Tracheostomy
|
6 (6)
|
1 (3.2)
|
2 (6.5)
|
0.554
|
|
- Dialysis
|
7 (7)
|
1 (3.2)
|
4 (12.9)
|
0.162
|
|
- Thromboembolic event
|
3 (3)
|
0
|
2 (6.5)
|
0.151
|
|
- Major infection
|
7 (7)
|
1 (3.2)
|
4 (12.9)
|
0.162
|
|
In-hospital mortality, n (%)
|
8 (8)
|
1 (3.2)
|
3 (9.7)
|
0.301
|
Abbreviation: ECMO, extracorporeal membrane oxygenation; FET, frozen elephant trunk;
SD, standard deviation.
Note: Values are presented as mean (standard deviation) or n (%).
Risk Factor Analysis
We first performed univariable analysis for the in-hospital mortality, resternotomy
for bleeding, SCI, and requirement for dialysis ([Supplementary Tables S1]–[S4], available in the online version). Age ≥ 70 years was associated with higher in-hospital
mortality (odds ratio: 4.418, 95% confidence interval: 0.925–21.105, p = 0.046). We only performed the multivariable binary logistic regression analysis
on the composite endpoint due to the small number of events. Zone 2/3 landing (odds
ratio: 6.491, 95% confidence interval: 1.930–21.835, p = 0.003) was identified as independent predictors of the early composite endpoint
of complications (resternotomy for bleeding, dialysis, and SCI). The detailed results
are shown in [Table 3].
Table 3
Risk factor analysis for composite endpoint (resternotomy for bleeding, dialysis,
and spinal cord injury)
|
Univariable analysis
|
Multivariable analysis
|
|
p-Value
|
OR
|
95% CI
|
p-Value
|
OR
|
95% CI
|
|
Age ≥ 70 y
|
0.719
|
1.291
|
0.321–5.197
|
|
|
|
|
Male
|
0.827
|
0.857
|
0.215–3.416
|
|
|
|
|
Hypertension
|
0.033
|
0.331
|
0.117–0.941
|
0.082
|
0.343
|
0.103–1.144
|
|
Creatinine > 115 μ mol/L
|
0.090
|
2.417
|
0.855–6.832
|
0.194
|
2.292
|
0.655–8.024
|
|
Previous sternotomy
|
0.614
|
0.733
|
0.218–2.461
|
|
|
|
|
Malperfusion/rupture/tamponade
|
0.077
|
2.625
|
0.879–7.837
|
0.198
|
2.462
|
0.625–9.710
|
|
Acute aortic dissection
|
0.884
|
0.926
|
0.332–2.584
|
|
|
|
|
Concomitant procedure
|
0.448
|
1.485
|
0.533–4.134
|
|
|
|
|
Axillary cannulation
|
0.981
|
1.014
|
0.325–3.164
|
|
|
|
|
Zone 2/3 landing
|
< 0.001
|
5.810
|
1.937–17.423
|
0.003
|
6.491
|
1.930–21.835
|
|
Stent graft length ≥ 150 mm
|
0.489
|
1.451
|
0.504–4.179
|
|
|
|
|
Branched FET
|
0.573
|
1.838
|
0.215–15.693
|
|
|
|
Abbreviations: CI, confidence interval; FET, frozen elephant trunk; OR, odds ratio.
Midterm Outcomes
Among the 92 survivors in the entire cohort, two defaulted our follow-up. Of these,
the mean follow-up duration was 32.3 ± 22.4 months. A total of six (6.5%) late postdischarge
deaths were noted. Four of these were due to aorta-related event. Twenty patients
(22.2%) required reintervention. Seven patients (9.2%) who suffered from either acute
or chronic aortic dissection developed dSINE. The overall 3-year survival was 92.7%.
In the matched cohort, the reintervention rates were 22.6 and 16.1% in zone 0/1 and
zone 2/3 groups, respectively (p = 0.520). More dSINE was observed after zone 0/1 FET in patients with aortic dissection
(14.8% vs. 4.0%, p = 0.186). There was no significant difference in the number of aorta-related events
between patients who underwent zone 0/1 and zone 2/3 FET (30.0% vs. 21.4%, p = 0.456). On Kaplan–Meier analysis, the 3-year freedom from aorta-related events
in all patients, and in patients with zone 0/1 and zone 2/3 FET were 73.0, 70.2, and
75.0%, respectively. In this, there were no significant between-group differences
(log-rank test, p = 0.500) ([Fig. 2]).
Fig. 2 Kaplan–Meier plot of freedom from aortic events after receiving total arch replacement
and frozen elephant trunk operation (FET). Three-year freedom from aortic event in
all patients, patients with FET landed at zone 0/1, and patients with FET landed at
zone 2/3 were 73.0, 70.2, and 75.0%, respectively. There were no significant differences
in this respect (log-rank test, p = 0.500).
Radiological Follow-Up
In the matched cohort, after an average of 73.4 ± 85.9 days, 56 patients (90.3%) underwent
contrast CT scans. Patients who underwent zone 2/3 FET had more coverage of the descending
thoracic aorta T8 or beyond than those who received zone 0/1 FET (21.4% vs. 32.1%,
p = 0.365); however, the between-group difference was not statistically significant.
All individuals with an aortic aneurysm had complete thrombosis of the aneurysmal
sac. Complete false lumen thrombosis around the end of the FET stent was similar in
both groups (64.3% vs. 71.4%, p = 0.567) among the patients with aortic dissection.
Discussion
Extensive evidence supporting the FET technique is available worldwide, especially
in the Western countries.[7]
[8]
[9] However, there is a paucity of reports from the Asian regions.[10]
[11]
[12] In this single-center review, the in-hospital mortality rate was 8% and the overall
3-year survival was 92.7%. In a recent systematic review of 37 studies (combined n = 4,178), the pooled in-hospital mortality rate and 3-year survival after the FET
procedure was 10.2% (range 0–20%) and 85.2%, respectively.[13] Our results are comparable to those reported from other countries.
FET is a more complex surgical procedure that requires a higher level of surgical
expertise as well as a longer operative time. Despite advances in surgical techniques,
the FET procedure has still been linked to a high rate of postoperative mortality
and morbidity. Conventionally, the FET device was implanted in zone 3 of the thoracic
aorta. Our research has found that anastomosing the FET to zone 2 or beyond (zone
2/3 FET) increased the risk of resternotomy for bleeding, acute kidney injury requiring
dialysis, and SCI by six times. However, previous studies have found conflicting evidence
of the benefits of proximalization of FET. Zone 3 FET exhibited a greater 30-day mortality
rate than zone 2 FET in research by Detter et al (3.3% vs. 17.7%), but the difference
was not statistically significant.[3] Two other studies showed comparable in-hospital/30-day mortality rates between zone
2 FET and zone 3 FET.[4]
[5] There were no significant between-group differences in the rates of SCI and dialysis
in the three abovementioned studies. These studies, however, only compared zone 3
and zone 2 FET. As shown in our study, the benefits of proximalization become clearer
as the anastomosis moves closer to zone 0 or zone 1. In their cohort of 108 patients,
Yamamoto et al reported outstanding outcomes with the “zone 0 repair” strategy. Their
cohort had a 30-day death rate of 2.8%, and none of the patients developed SCI.[6]
Better outcomes with proximalization of FET may be related to the reduced visceral
ischemic time. In our study, zone 0/1 FET required a shorter circulatory arrest time,
although the difference was not statistically significant, and it could be due to
the more frequent use of branched FET. Nevertheless, this phenomenon has been observed
in studies comparing zone 2 FET and zone 3 FET.[3]
[4]
[14]
According to Preventza et al's meta-analysis, a stent graft length of 15 cm or longer
was associated with a greater SCI rate.[15] This was also demonstrated in our univariable analysis. However, zone 0/1 FET did
not reliably reduce distal thoracic aorta coverage. Hence, the extent of coverage
is multifactorial. Other important determinants are the length, tortuosity, and elasticity
of the thoracic aorta, in addition to the location of the FET anastomosis and the
length of the FET graft. Therefore, the 100-mm graft is becoming more popular for
aortic dissection.[16] The use of transesophageal echocardiography as intraoperative guidance during FET
deployment has also been proven to be beneficial.[17]
The FET proximalization method was associated with lower surgical trauma. In our investigation,
the incidence of left recurrent laryngeal nerve injury was lower in the zone 0/1 group;
however, the difference was insignificant. This could be due to our preference for
direct end-to-end anastomosis of the left subclavian artery, as indicated by the fact
that only 7% of our cohort received extra-anatomical axillary artery bypass. Zone
2 FET was associated with a decreased frequency of recurrent laryngeal nerve injury
in a study by Leone et al.[4] They used a hybrid vascular graft instead to allow for a sutureless anastomosis
when the left subclavian artery was difficult to reach. In another study, when extra-anatomical
bypass of limb vessels was used exclusively throughout the FET surgery, none of the
patients developed recurrent laryngeal nerve palsy.[18]
It is difficult to access the distal anastomosis of the aortic arch and the head and
neck vessels to guarantee hemostasis.[19] In our group, zone 0/1 FET was also associated with fewer episodes of resternotomy
for bleeding, despite inconsistent results in prior studies comparing zone 2 and zone
3 FET.[5]
[20] In the zone 0/1 FET group, less frequent resternotomy for hemostasis may be related
to less blood product transfusion, and hence contribute to a lower rate of dialysis
necessity.
FET has been demonstrated to improve the coverage of the distal descending thoracic
aorta, resulting in improved false lumen or aneurysm exclusion. Residual false lumen
patency after aortic dissection repair was linked to more aorta-related events, including
late deaths, according to a meta-analysis by Li et al.[21] Positive aortic remodeling with false lumen thrombosis and aorta shrinkage were
found after the FET procedure.[22]
[23] In a study by Panfilov et al, however, the rate of complete false lumen thrombosis
in zone 2 FET was lower than in zone 3 FET.[5] Therefore, zone 3 FET had a higher 2-year freedom from reintervention (74.1% vs.
91.7%), but the difference was not statistically significant. In our study, the rate
of reintervention and early complete false lumen thrombosis around the level of FET
were comparable in the zone 0/1 FET and zone 2/3 FET groups. Aorta-related events
were also comparable in incidence. The potential negative impact of proximalization
necessitates more research and a longer follow-up period.
Proximalization of FET improves surgical outcomes even more. To get satisfactory results,
however, careful patient selection and an understanding of the risk–benefit profile
are still required. Although patients of any age should be able to undergo the FET
procedure, the increased surgical risk of FET in this demographic deserves attention.
According to a recent study, in-hospital mortality after FET was approximately 2.5
times higher in patients ≥ 70 years old.[24] In our study, age ≥ 70 years was found to be a predictor of in-hospital mortality
in univariable analysis as well. Physiological changes due to aging, including frailty
and multiple comorbidities, are likely the underlying causes that hinder recovery
and increase the likelihood of complications in older patients. We believe that there
should not be an age cutoff for FET; older patients, if chosen appropriately, can
benefit from the procedure with an acceptable risk level.
Limitations
The heterogeneity in our cohort limited our initial assessment of the FET procedure.
Further comparisons were also hindered by the small number of patients. As the operating
surgeon decided whether they should use FET, the results are prone to selection bias.
Despite this, we could portray the real-world situation of FET use at our institution.
Finally, our study has a retrospective design, which would be a limitation.
Conclusion
Zone 0/1 FET was associated with a lower composite endpoint of resternotomy for bleeding,
dialysis-dependent acute kidney injury, and SCI. At 3 years, similar aorta-related
events were observed. Further research into to the strategy of more aggressive proximalization
of FET is required.
