Keywords
valve-sparing root reimplantation - David procedure - root aneurysm - upper - hemisternotomy
Traditionally, root aneurysms have been treated with valve-replacing root replacement
(VRRR) procedures, in which the aortic valve and root are replaced with a composite
valve–graft conduit.[1] Valve-sparing root replacement (VSRR) procedures, first described by David and colleagues
with the root reimplantation technique, are safe alternatives to VRRR procedures in
carefully selected patients with normal or repairable cusp morphology.[2]
[3]
[4]
[5] Unlike VRRR procedures, VSRR procedures allow for the avoidance of anticoagulation
and theoretically have decreased risk of thromboembolic events often associated with
mechanical valves and decreased risk of structural valve degeneration often associated
with bioprosthetic valves.[4]
[5]
[6]
Modifications to the original reimplantation technique have improved valve dynamics
and allowed for a more natural restoration of the structure and function of the aortic
root while also increasing reproducibility.[7]
[8]
[9]
[10]
[11] Minimally invasive adaptations to perform VSRR through an upper hemisternotomy (UHS)
have also been reported, with good midterm outcomes.[12]
[13]
[14]
[15] In this paper, we aim to share our experience and mid-term outcomes with VSRR for
diverse indications using the reimplantation technique.
Methods
Study Design
Between August 2005 and January 2023, 84 patients underwent VSRR procedures by a single
surgeon (K.A.P.). Of these 84 patients, 3 patients who utilized the remodeling technique
were excluded from analysis. All patients, including those undergoing reoperations
with a history of prior cardiac or aortic surgery, those with concomitant procedures,
and those requiring emergency repair of acute type A aortic dissections were included
in the study. Mid-term outcomes include freedom from mortality, freedom from recurrence
of aortic insufficiency (AI) (greater than mild), and freedom from aortic valve-related
reoperation.
Patients were captured from a prospectively compiled Aortic Wellness Database containing
preoperative, intraoperative, and postoperative data, and their data were analyzed,
retrospectively. Institutional review board approval for use of the deidentified database
for analysis (number: 20D.802) and the study (iRISID-2023-1860) was obtained on January
31, 2023. Informed consent was waived due to the retrospective nature of the study.
Selection
Patients with aortic root aneurysms (>5.0 cm in diameter) and normal or repairable
(small fenestrations, lack of calcifications, pliable leaflets with or without prolapse)
cusp morphology with or without AI were considered for surgery. For patients with
connective tissue disorders, careful case-specific considerations are made between
the patient and a multidisciplinary aortic team to operate at lower thresholds.
Preoperative Evaluation
Echocardiography is used to assess valve anatomy (bicuspid aortic valve (BAV) vs.
tricuspid (TAV)), severity and mechanism of AI, size of the aortic annulus, presence
of cusp calcification, stenosis, prolapse, perforation, and ventricular size and function.
Any degree of stenosis or cusp calcification impairing leaflet pliability noted on
echocardiography are contraindications to VSRR procedures. Contrast-enhanced computed
tomography scan with 3-dimensional reconstruction is used to assess root size, morphology,
and the need for concomitant aortic procedures. Preoperative pulmonary function testing
and coronary angiography are performed in all patients.
Surgical Technique
We have previously written in great detail about our technique ([Video 1], available in the online version).[14] In short, median sternotomy is performed and after cannulation, cardiopulmonary
bypass (CPB) is initiated. The aorta is cross-clamped and the heart is arrested with
cold cardioplegia. The aorta is transected 1 cm above the sinotubular junction (STJ)
and three commissural sutures are placed. The aortic annulus, cusps, and sinuses of
Valsalva are inspected. The leaflets are evaluated for pliability, prolapse, fenestrations,
and calcifications. Any degree of calcification and multiple, large fenestrations
are contraindications to VSRR procedures.
Video 1 How we perform valve-sparing root reimplantation.
If prolapse is identified, it is repaired using free-edge plication prior to valve
reimplantation. The right coronary (RC) and left coronary (LC) buttons are mobilized
and the noncoronary (NC) sinus is excised. The aortic root is dissected and mobilized
circumferentially as deep as possible. The membranous and muscular septum may limit
the external dissection. The Valsalva graft (Vascutek Ltd, Terumo Aortic, Renfrewshire,
UK) is then sized using the height of the LC–NC commissure.[11] The proximal aspect of the graft is trimmed corresponding to the anatomical limits
of dissection at the base of the aortic root at the RC/LC and NC/RC commissures.
Twelve Ethicon 2–0 pledgeted sutures are passed underneath the annulus and passed
through the proximal aspect of the graft following which the prosthesis is tied down.
The aortoventricular junction is stabilized to 23 mm in men and 21 mm in women over
a Hegar dilator. The commissures are reimplanted inside the graft 120-degree apart
in patients with TAV and 180-degree apart in patients with BAV establishing a neo-STJ
where the skirt of the Valsalva graft ends. A secondary continuous 4–0 polypropylene
suture line is placed, securing the valve inside the graft. At this point, the valve
is reassessed for prolapse by inspection and the use of the Schafers caliber and any
residual prolapse is repaired with free edge plication. The coronary buttons are reattached
and the distal anastomosis is constructed 1 cm proximal to the aortic cross-clamp
(ACC), the heart is deaired, and the patient is taken off CPB.
In recent years, we have implemented several key adaptations to facilitate performing
VSRR through an UHS. These include the use of Custodial–histidine–tryptophan–ketoglutarate
and Del Nido cardioplegia, a modified cannulation strategy using the right femoral
vein for venous inflow, and the use of Cor-Knot among other minimally invasive instruments.[14]
[16]
Follow-up
Patients were followed yearly after surgery with clinical visits either at our office
or with their primary care physician or cardiologist. Echocardiography was performed
yearly to assess valve function. Follow-up was obtained through a combination of institutional
records, including outpatient clinical notes and echocardiograms, direct patient outreach,
the use of the social security death index, and public obituary records.
Statistical Analysis
Primary outcomes considered were all-cause mortality, aortic valve-related reoperation,
and freedom from recurrence of AI. Kaplan–Meier estimates for freedom from mortality
probability were constructed with study departure treated as a censoring event. A
Kaplan–Meier estimate looking at a composite of reoperation, recurrence of AI, or
death with study departure treated as a censoring event was constructed. Due to the
low number of events in the UHS group, outcomes between the FS and UHS groups were
not directly compared in formal hypothesis tests. Data were analyzed using R version
4.3.1.
Results
Patient Characteristics
The preoperative baseline characteristics are shown in [Table 1]. A total of 81 patients undergoing VSRR using the reimplantation technique were
included in this study. The patients were primarily male (89.3%) with mean age of
50.3 ± 1.4 years. Forty-nine patients had hypertension (60.5%), 6 had diabetes (7.4%),
8 had chronic obstructive pulmonary disease (9.9%), and 2 had prior stroke (2.5%).
Six patients (7.4%) had previous cardiac operations and 2 patients (2.5%) had previous
proximal aortic procedures. Four patients (4.9%) underwent redo sternotomy. Of the
10 patients (12.3%) who required VSRR for urgent or emergent indications, 8 (9.9%)
had VSRR for acute type A aortic dissection and were performed exclusively through
an FS.
Table 1
Preoperative baseline characteristics
Characteristics
|
Full (n = 59)
|
Mini (n = 22)
|
Total (n = 81)
|
Patients (n)
|
59
|
22
|
81
|
Age (y) (mean ± SE)
|
49.6 ± 1.7
|
52.1 ± 2.2
|
50.3 ± 1.4
|
Gender
|
53 (89.8%)
|
20 (90.9%)
|
73 (90.1%)
|
Race (Caucasian)
|
37 (62.7%)
|
18 (90%)
|
55 (69.6%)
|
Symptoms at presentation
|
33 (55.9%)
|
14 (63.6%)
|
47 (58%)
|
Pain
|
22 (37.3%)
|
3 (13.6%)
|
25 (30.9%)
|
Rupture
|
0 (0%)
|
0 (0%)
|
0 (0%)
|
Congestive heart failure
|
0 (0%)
|
0 (0%)
|
0 (0%)
|
NYHA dyspnea
|
|
|
|
1 (n, %)
|
2 (3.4%)
|
0 (0%)
|
2 (2.5%)
|
2
|
46 (78%)
|
19 (86.4%)
|
65 (80.2%)
|
3
|
6 (10.2%)
|
0 (0%)
|
6 (7.5%)
|
4
|
5 (8.5%)
|
2 (9.1%)
|
7 (8.6%)
|
NYHA angina
|
|
|
|
1 (n, %)
|
42 (77.8%)
|
21 (95.5%)
|
46 (78%)
|
2
|
6 (11.1%)
|
0 (0%)
|
6 (10.2%)
|
3
|
2 (3.7%)
|
1 (4.5%)
|
2 (3.4%)
|
4
|
2 (3.7%)
|
0 (0%)
|
3 (5.1%)
|
BMI (kg/m2)
|
28.3 ± 0.63
|
28.5 ± 1.1
|
28.37 ± 0.54
|
Associated disease
|
|
|
|
Smoking
|
20 (33.9%)
|
8 (36.3%)
|
29 (34.5%)
|
Ascending aneurysm
|
58 (98.3%)
|
21 (95.5%)
|
79 (97.5%)
|
Acute dissection
|
8 (13.6%)
|
0 (0%)
|
8 (9.9%)
|
Chronic dissection
|
1 (1.7%)
|
1 (4.5%)
|
2 (2.5%)
|
Medial degeneration
|
36 (61%)
|
19 (86.4%)
|
55 (67.9%)
|
Marfan syndrome
|
7 (11.9%)
|
3 (13.6%)
|
10 (12.3%)
|
Bicuspid aortic valve
|
5 (8.5%)
|
3 (13.6%)
|
8 (9.9%)
|
Hypertension
|
34 (57.6%)
|
15 (68.2%)
|
49 (60.5%)
|
COPD
|
6 (10.2%)
|
2 (9.1%)
|
8 (9.9%)
|
Diabetes
|
5 (8.5%)
|
1 (4.5%)
|
6 (7.4%)
|
Cerebrovascular disease
|
4 (6.8%)
|
0 (0%)
|
4 (4.9%)
|
Prior stroke
|
2 (3.4%)
|
0 (0%)
|
2 (2.5%)
|
Prior transient ischemic attack
|
3 (5.1%)
|
0 (0%)
|
3 (3.7%)
|
Renal insufficiency
|
1 (1.7%)
|
0 (0%)
|
1 (1.2%)
|
Creatinine > 2.5
|
1 (1.7%)
|
0 (0%)
|
1 (1.2%)
|
Peripheral vascular disease
|
0 (0%)
|
1 (4.5%)
|
1 (1.2%)
|
Preoperative findings
|
|
|
|
Root size (cm)
|
4.9 ± 0.09
|
5.2 ± 0.12
|
5.0 ± 0.07
|
Ascending size (cm)
|
4.4 ± 0.11
|
4.4 ± 0.13
|
4.4 ± 0.09
|
Annulus size (cm)
|
2.7 ± 0.09
|
2.7 ± 0.25
|
2.7 ± 0.08
|
Sinotubular junction size (cm)
|
4.1 ± 0.1
|
4.0 ± 0.29
|
4.0 ± 0.09
|
Arch size (cm)
|
3.2 ± 0.1
|
3.7 ± 0.18
|
3.3 ± 0.09
|
Urgent or emergent
|
9 (13.3%)
|
1 (4.5%)
|
10 (12.3%)
|
Previous proximal aortic surgery
|
2 (3.4%)
|
0 (0%)
|
2 (2.5%)
|
Previous distal aortic surgery
|
0 (0%)
|
0 (0%)
|
0 (0%)
|
Previous cardiac operations
|
4 (6.8%)
|
2 (9.1%)
|
6 (7.4%)
|
Ejection fraction
|
59 ± 1
|
58 ± 1.5
|
59 ± 0.8
|
Aortic insufficiency
|
|
|
|
None (n, %)
|
19 (32.2%)
|
12 (54.5%)
|
31 (38.3%)
|
Mild
|
20 (33.9%)
|
5 (22.7%)
|
25 (30.9%)
|
Moderate
|
13 (22%)
|
3 (13.6%)
|
16 (19.8%)
|
Severe
|
7 (11.9%)
|
2 (9.1%)
|
9 (11.1%)
|
Central (CM)
|
16 (27.1%)
|
3 (13.6%)
|
19 (23.5%)
|
Mitral regurgitation
|
|
|
|
None (n, %)
|
38 (64.4%)
|
15 (68.2%)
|
53 (65.4%)
|
Mild
|
16 (27.1%)
|
7 (31.8%)
|
23 (28.4%)
|
Moderate
|
4 (6.8%)
|
0 (0%)
|
4 (4.9%)
|
Severe
|
1 (1.7%)
|
0 (0%)
|
1 (1.2%)
|
Redo
|
3 (5.1%)
|
1 (4.5%)
|
4 (4.9%)
|
Abbreviations: BMI, body mass index; COPD, chronic obstructive pulmonary disease;
NYHA, New York Heart Association; SE, standard error.
Most patients had a preoperative New York Heart Association Dyspnea class of II (80%).
No AI, mild, moderate, and severe AI were present in 31 (38.3%), 25 (30.9%), 16 (19.8%),
and 9 (11.1%) patients, respectively. Five patients (6.2%) had greater than mild mitral
regurgitation. Mean aortic root size was 5.0 ± 0.07 cm and ascending aortic size was
4.4 ± 0.09 cm. Mean preoperative annulus size was 2.7 ± 0.08 cm and preoperative STJ
size was 4.0 ± 0.09 cm. Eight patients (9.9%) had BAV and 10 patients (12.3%) had
Marfan syndrome.
Operative
Operative characteristics are displayed in [Table 2]. Fifty-nine patients (72.8%) had VSRR performed through an FS and 22 patients (27.2%)
had VSRR performed through an UHS. The most common Valsalva graft size was 32 mm (60.5%).
Mean CPB time was 223 ± 5.0 minutes and mean ACC time was 193 ± 4.2 minutes.
Table 2
Operative characteristics
Characteristics
|
Full (n = 59)
|
Mini (n = 22)
|
Total (n = 81)
|
Cardiopulmonary bypass time (min)
|
229.6 ± 6.1
|
204.9 ± 7.0
|
223.1 ± 5.0
|
Clamp time (min)
|
199.7 ± 5.0
|
174.5 ± 6.3
|
193.1 ± 4.2
|
Circulatory arrest time (min)
|
14 (10.5–20.5)
|
16.5 (8.75–24.25)
|
14 (10–21)
|
Incision type (full sternotomy)
|
59 (72.8%)
|
22 (27.2%)
|
81
|
Deep hypothermic circulatory arrest
|
18 (30.5%)
|
2 (9.1%)
|
20 (24.7%)
|
Blood loss (mL)
|
253 (0–555)
|
0 (0–89.5)
|
125 (0–500)
|
PRBCs (units)
|
0 (0–2)
|
0 (0–0)
|
0 (0–2)
|
Fresh frozen plasma (units)
|
2 (0–2.5)
|
0 (0–2)
|
0 (0–2)
|
Platelets (units)
|
1 (0–2)
|
1 (0–1.75)
|
1 (0–2)
|
Cryoprecipitate (units)
|
0 (0–2)
|
2 (0–4)
|
0 (0–4)
|
Simultaneous procedures
|
29 (49.2%)
|
19 (86.4%)
|
48 (59.3%)
|
CABG
|
1 (1.7%)
|
0 (0%)
|
1 (1.2%)
|
Mitral valve repair
|
3 (5.1%)
|
0 (0%)
|
3 (3.7%)
|
Patent foramen ovale closure
|
1 (1.7%)
|
0 (0%)
|
1 (1.2%)
|
Distal wrap
|
21 (35.6%)
|
15 (68.2%)
|
36 (44.4%)
|
Arch
|
|
|
|
Hemiarch
|
5 (8.5%)
|
2 (9.1%)
|
7 (8.6%)
|
Trifurcation
|
4 (6.8%)
|
0 (0%)
|
4 (4.9%)
|
Elephant trunk + trifurcation
|
1 (1.7%)
|
0 (0%)
|
1 (1.2%)
|
Annulus postoperative (cm)
|
2.2 ± 0.02
|
2.2 ± 0.14
|
2.2 ± 0.03
|
Sinotubular junction postoperative size (cm)
|
3.0 ± 0.07
|
2.9 ± 0.13
|
3.0 ± 0.06
|
Abbreviations: CABG, coronary artery bypass grafting; PRBC, packed red blood cell.
Right coronary cusp (RCC) prolapse was most common (n = 17, 21.0%), followed by Non-coronary cusp (NCC) prolapse (n = 11, 13.6%), and left coronary cusp (LCC) prolapse (n = 7, 8.6%). There were 24 (29.6%) patients who had leaflet repair. Leaflet repair
characteristics are displayed in [Table 3]. RCC, LCC, and NCC plication were performed in 21 (25.9%), 13 (16%), and 4 (4.9%)
patients, respectively. There were 11 (13.6%) patients, 11 (13.6%) patients, and 2
(2.5%) patients that had 1, 2, and 3 leaflet plication respectively. Subcommissural
annuloplasty between the NCC and LCC was performed in 1 (1.2%) patient. The LCC was
the most common location of fenestrations (n = 3, 3.7%). Twelve (14.8%) patients had residual prolapse after reimplantation that
required additional leaflet plication after evaluation of the leaflets with the Schafers
caliper. Immediately postrepair, there were no cases of moderate or severe central
AI or eccentric AI, with only 10 patients (12.3%) having trivial to mild central AI.
Mean leaflet coaptation height was 10.3 ± 0.5 mm.
Table 3
Valve repair techniques
Characteristics
|
Full (n = 59)
|
Mini (n = 22)
|
Total (n = 81)
|
Coaptation height (mm)
|
10.0 ± 0.5
|
12.0 ± 0.9
|
10.3 ± 0.5
|
Graft size
|
|
|
|
30
|
17 (30.9%)
|
4 (19%)
|
21 (27.6%)
|
32
|
32 (58.2%)
|
14 (66.7%)
|
46 (60.5%)
|
34
|
3 (5.5%)
|
2 (9.5%)
|
5 (6.6%)
|
RCC prolapse
|
12 (20.3%)
|
5 (22.7%)
|
17 (21.0%)
|
RCC fenestration
|
1 (1.7%)
|
0 (0%)
|
1 (1.2%)
|
RCC plication
|
16 (27.1%)
|
5 (22.7%)
|
21 (25.9%)
|
RCC edge suspension
|
2 (2.4%)
|
1 (4.5%)
|
3 (3.7%)
|
LCC prolapse
|
6 (10.2%)
|
1 (4.5%)
|
7 (8.6%)
|
LCC fenestration
|
3 (5.1%)
|
0 (0%)
|
3 (3.7%)
|
LCC plication
|
11 (18.6%)
|
2 (9.1%)
|
13 (16%)
|
LCC edge suspension
|
2 (3.4%)
|
0 (0%)
|
2 (2.5%)
|
NCC prolapse
|
7 (11.9%)
|
4 (18.2%)
|
11 (13.6%)
|
NCC fenestration
|
0 (0%)
|
0 (0%)
|
0 (0%)
|
NCC plication
|
0 (0%)
|
1 (4.5%)
|
4 (4.9%)
|
NCC edge suspension
|
0 (0%)
|
0 (0%)
|
0 (0%)
|
1 leaflet prolapse
|
8 (13.6%)
|
3 (13.6%)
|
11 (13.6%)
|
2 leaflet prolapse
|
4 (6.8%)
|
2 (9.1%)
|
6 (7.4%)
|
3 leaflet prolapse
|
3 (5.1%)
|
1 (4.5%)
|
4 (4.9%)
|
1 leaflet plication
|
9 (15.3%)
|
2 (9.1%)
|
11 (13.6%)
|
2 leaflet plication
|
8 (13.6%)
|
3 (13.6%)
|
11 (13.6%)
|
3 leaflet plication
|
2 (2.5%)
|
0 (0%)
|
2 (2.5%)
|
NC-LC annuloplasty
|
0 (0%)
|
1 (4.5%)
|
1 (1.2%)
|
LC-RC annuloplasty
|
0 (0%)
|
0 (0%)
|
0 (0%)
|
RC-NC annuloplasty
|
0 (0%)
|
0 (0%)
|
0 (0%)
|
Correction of residual prolapse
|
11 (18.6%)
|
1 (4.5%)
|
12 (14.8%)
|
Abbreviations: LC, left coronary; NC, noncoronary; RC, right coronary.
Major concomitant procedures include coronary artery bypass grafting (n = 1, 1.2%), mitral valve surgery (n = 3, 3.7%), and arch surgery (n = 12, 14.8%). Overall usage of blood products, including packed red blood cells (0
[interquartile range, IQR: 0–2]), fresh frozen plasma (0 [IQR: 0–2]), platelets (1
[IQR: 0–2]), and cryoprecipitate (0 [IQR: 0–4]) was minimal.
Early Outcomes
There was no 30-day or in-hospital mortality ([Table 4]). Additionally, median intensive care unit stay was 3 (2–5) days and median hospital
stay was 7 (6–9) days. One (1.2%) patient had a postoperative transient ischemic attack
and 1 patient (1.2%) had a stroke. Ten patients (12.3%) required prolonged ventilation
support (>48 hours), 7 patients (8.6%) developed atrial fibrillation, and 6 patients
(7.4%) required reoperation for bleeding. One patient (1.2%) developed new renal insufficiency.
Table 4
Early postoperative outcomes
Characteristics
|
Full (n = 59)
|
Mini (n = 22)
|
Total (n = 81)
|
ICU stay (d)
|
4 (2–5)
|
3 (2–4)
|
3 (2–5)
|
Hospital stay (d)
|
7 (6–11)
|
6 (6–7)
|
7 (6–9)
|
Prolonged ventilatory support
|
8 (13.6%)
|
2 (9.1%)
|
10 (12.3%)
|
Chest left open
|
2 (3.4%)
|
1 (4.5%)
|
3 (3.7%)
|
Stroke
|
1 (1.7%)
|
0 (0%)
|
1 (1.2%)
|
Transient ischemic attack
|
1 (1.7%)
|
0 (0%)
|
1 (1.2%)
|
New renal insufficiency
|
1 (1.7%)
|
0 (0%)
|
1 (1.2%)
|
Postoperative bleeding
|
5 (8.5%)
|
1 (4.5%)
|
6 (7.4%)
|
Vocal cord paralysis
|
2 (3.4%)
|
0 (0%)
|
2 (2.5%)
|
Temporary neurological dysfunction
|
2 (3.4%)
|
0 (0%)
|
2 (2.5%)
|
Myocardial infarction
|
0 (0%)
|
0 (0%)
|
0 (0%)
|
Congestive heart failure
|
0 (0%)
|
0 (0%)
|
0 (0%)
|
Arrhythmia
|
7 (11.9%)
|
0 (0%)
|
7 (8.6%)
|
Pacemaker
|
0 (0%)
|
1 (4.5%)
|
1 (1.2%)
|
Effusion
|
6 (10.2%)
|
1 (4.5%)
|
7 (8.6%)
|
Atelectasis
|
1 (1.7%)
|
0 (0%)
|
1 (1.2%)
|
Tracheostomy
|
1 (1.7%)
|
0 (0%)
|
1 (1.2%)
|
Reintubation
|
2 (3.4%)
|
0 (0%)
|
2 (2.5%)
|
Mortality
|
0 (0%)
|
0 (0%)
|
0 (0%)
|
Postoperative aortic insufficiency
|
|
|
|
None
|
52 (88.1%)
|
19 (86.4%)
|
71 (87.7%)
|
Mild
|
7 (11.9%)
|
3 (13.6%)
|
10 (12.3%)
|
Moderate
|
0 (0%)
|
0 (0%)
|
0 (0%)
|
Severe
|
0 (0%)
|
0 (0%)
|
0 (0%)
|
Late Outcomes
Late outcomes are displayed in [Table 5]. The mean follow-up length was 7.6 ± 0.5 years with 29 (35.8%) patients being followed
for >10 years and 5 (6.2%) patients being followed for >15 years. Clinical follow-up
was complete and echocardiographic follow-up was 89% complete. On follow-up echocardiography,
average mean and peak gradient were 5 (3–8) and 10 (6–16) mm Hg, respectively. Freedom
from all-cause mortality at 1, 5, and 10 years was 100, 96.6 and 94.4%, respectively
([Fig. 1]). There were five late deaths, of which four were confirmed not to be cardiac-related
(coronavirus disease, meningioma, stroke, unspecified neurological condition). Cause
of death could not be determined for one patient.
Table 5
Mid-term outcomes
Characteristics
|
Full (n = 59)
|
Mini (n = 22)
|
Total (n = 81)
|
Follow-up length (y)
|
8.7 ± 0.64
|
4.6 ± 0.4
|
7.6 ± 0.5
|
Freedom from mortality
|
|
|
|
1 y
|
100%
|
100%
|
100%
|
5 y
|
97.70%
|
94.10%
|
96.60%
|
10 y
|
95.20%
|
|
94.40%
|
Freedom from recurrence of aortic insufficiency
|
|
|
|
1 y
|
|
|
100%
|
5 y
|
|
|
100%
|
10 y
|
|
|
97.3%
|
Freedom from reoperation
|
|
|
|
1 y
|
|
|
98.8%
|
5 y
|
|
|
95.4%
|
10 y
|
|
|
95.4%
|
Ejection fraction long-term
|
60.32 ± 1.22
|
58.4 ± 2.0
|
59.72 ± 1.1
|
LVEDD (cm)
|
5.1 ± 0.14
|
4.26 ± 0.23
|
4.91 ± 0.14
|
LVESD (cm)
|
3.41 ± 0.13
|
2.89 ± 0.19
|
3.3 ± 0.11
|
Aortic root size (cm)
|
3.64 ± 0.08
|
3.92 ± 0.12
|
3.75 ± 0.06
|
Mean gradient (mm Hg)
|
5 (3.1–7)
|
7.5 (4.5–10.25)
|
5.3 (3.1–8)
|
Peak gradient (mm Hg)
|
10 (6.4–16)
|
10 (6–15)
|
10 (6.1–15.75)
|
Fig. 1 Freedom from mortality.
Cumulative freedom from mortality, freedom from reoperation, and freedom from recurrence
was 98.8, 92.1, and 87.3% at 1, 5, and 10 years ([Fig. 2]). Four patients required aortic valve-related late reoperation at 0.25, 2.2, 4.9,
and 16.0 years. There were no late reoperations for recurrence of greater than mild
AI. The first patient developed an aortic root abscess a few months following surgery
ultimately requiring a Bentall procedure with a bioprosthetic valve. The second patient
developed endocarditis and ultimately received a transcatheter aortic valve replacement
(TAVR). The third patient developed ventricular tachyarrhythmia following surgery
and had an iatrogenic aortic leaflet perforation following a transcatheter intervention
requiring an aortic valve replacement. The final patient developed severe aortic stenosis
and received a TAVR. Four patients developed late recurrence of new mild-to-moderate
central AI at 7.5, 13.0, 13.1, and 16.9 years. All four patients were asymptomatic.
There was no recurrence of moderate or severe AI in the follow-up period.
Fig. 2 Composite freedom from mortality, aortic valve-related reoperation, and recurrence
of AI.
Discussion
Outcomes
Since the first series of VSRR described by David and colleagues, multiple series,
including David's own 25-year experience, have demonstrated excellent outcomes, with
<1% operative mortality and 10-year valve-related freedom from reintervention > 90%.[17]
[18]
[19]
[20]
[21] Our results are consistent with the literature with no operative mortality and an
overall survival of 94.4% at 10 years, which is excellent considering the inclusion
of patients with aortic dissection and redo procedures. Moreover, none of the deaths
were confirmed to be cardiac-related. One (1.2%) patient who underwent VSRR for acute
type A aortic dissection developed a stroke and had near resolution of symptoms by
postoperative day 8 and no patients developed serious cardiac complications. The low
perioperative morbidity can be partially attributed to the fact that we are very selective
in performing this operation primarily in younger individuals with low comorbid conditions
and adhere to strict technical details to prevent perioperative bleeding or coronary
malperfusion. In addition, we maintain tight blood pressure control (systolic blood
pressure < 120 mm Hg and diastolic blood pressure < 80 mm Hg) during the hospital
course.
Due to the limited size of the series, a composite of mortality, reoperation, and
recurrence of AI was calculated in lieu of individual analysis of reoperation and
recurrence of AI and found to be 87.3% at 10 years, which compares favorably to the
established >90% freedom from reoperation at 10 years surgical quality standards set
by the National Marfan Foundation. David and colleagues have reported freedom from
AI of 93.2 and 78% at 10 and 18 years and freedom from aortic valve reoperation of
97.0 and 94.8% at 10 and 18 years.[19] The largest series of VSRR procedures reported an 8.2% recurrence of AI across the
follow-up period, with 85% of those patients subsequently necessitating reintervention
at 15 years.[20] Notably, only 4 (4.9%) of our patients developed recurrence of mild-to-moderate
AI during the follow-up period with no recurrence of moderate AI or greater. Three
out of the four recurrences occurred after 10 years, and there were no late reinterventions
for recurrence of AI. Of the four late reinterventions, the aortic valve replacement
for iatrogenic leaflet perforation from a transcatheter procedure was potentially
avoidable.
In our opinion, our cohort has had excellent valve durability for several reasons.
First, we are highly selective with the patients we choose to perform VSRR in. We
do not perform VSRR in patients with any calcification in TAV valves or more than
mild raphe calcification in bicuspid valves, or any aortic stenosis. Also, patients
with large fenestrations particularly associated with prolapse are excluded. Small
fenestrations are not a contraindication. To address leaflet prolapse, we perform
mostly central free-edge plication and rarely leaflet resuspension. In patients with
pliable leaflets with a thicker free edge, we perform leaflet shaving with an 11-blade
followed by central free edge plication if needed. In patients requiring extensive
patch repair, which is associated with early repair failure, we will elect to replace
the valve.[22]
[23] On preoperative echocardiographic evaluation, 21 (25.9%) patients had 1 or more
prolapsed leaflets. Intraoperatively, 24 (29.6%) patients required leaflet plication,
with 11 (13.6%) patients, 11 (13.6%) patients, and 2 (2.5%) patients needing 1, 2,
and 3 leaflet plication, respectively. Twelve (14.8%) patients required additional
plication for residual prolapse following valve reimplantation. One (1.2%) patient
was placed back on bypass before leaving the operating room for additional leaflet
plication after weaning because of the development of moderate AI. Nearly half of
our patients that required leaflet repair required additional leaflet repair after
reimplantation. These results suggest that valves need to be carefully inspected for
residual prolapse even after reimplantation as there can be induced prolapse. Without
aggressively addressing leaflet prolapse both before and after reimplantation, patients
may be more prone to developing early repair failure and recurrence of AI. Thus, we
never leave the operating room with greater than mild central AI.
Modifications and Reproducibility
Over the years, various modifications have been made to not only better mimic the
physiology of the native aortic root, but also to increase reproducibility. The original
“David I” utilized a straight tube graft, which increased aortic cusp closing velocities
causing increased stress and theoretically, quicker degeneration of the leaflets.
The current “David V” procedures utilizes a larger and smaller graft create a “pseudosinus,”
which has more physiological leaflet mechanics.[24]
[25]
We believe that the “El Khoury” technique, which utilizes Valsalva grafts, allows
for the most reproducible procedure without sacrificing outcomes. We always remove
the collar of the Valsalva graft. The transition line between the skirt and the body
of the Valsalva graft effectively serves as a well-demarcated neo-STJ, which allows
for reproducible reimplantation of the commissures every time. To account for anatomical
differences in the height of the commissures, the proximal aspect of the skirt is
trimmed in the areas of the LC/NC and NC/RC commissures, if needed to respect the
natural anatomical limits of dissection of the aortic root. There are also markings
at 120 degrees, which allow for the most consistent commissure positioning at reimplantation.
Although the David V and Stanford medication allows the surgeon to better customize
the pseudosinuses to each patient, improved long-term durability has not been demonstrated
compared with the use of commercially available Valsalva grafts. Routine utilization
of Schafers caliper to look for an effective height > 8 mm has also resulted in a
more systematic, reproducible assessment for prolapse compared with visual inspection.
Minimally Invasive Experience
Since 2015, we have adopted an UHS first approach for all patients undergoing elective
or urgent VSRR. In the follow-up period, in our 22 patients, there were no cardiac-related
deaths and one reoperation at 3 months due to a mediastinal abscess. Monsefi et al,
in 26 UHS patients with mean follow-up time of 3 ± 2 years, reported no operative
mortalities and 1 reintervention for recurrence of severe AI in the follow-up period.[12] Similarly, Shrestha et al, in 42 UHS patients with mean follow-up time of 4.2 ± 2.1
years, reported an operative mortality of 2.4% with 1 late mortality and 3 late reinterventions.[13] Our results are consistent with the previous conclusions, that in the midterm, VSRR
through an UHS is an acceptable operation, although long-term durability is still
to be determined.
Limitations
This was a retrospective review of a large prospectively maintained aortic registry
and is limited by smaller sample size. With the low number of observed events, direct
comparison was also not made directly between FS and UHS groups due to limited statistical
power. Given this is a single-surgeon series, selection bias may be present and results
may not be generalizable. Also, we were unable to obtain the latest echocardiographic
data on a small group of patients, although at the latest clinical follow-up these
patients were asymptomatic. In addition, longer follow-up is still needed to determine
long-term valve durability in patients undergoing VSRR through an UHS.
Conclusion
VSRR using the reimplantation technique is a durable procedure that when performed
properly and judiciously, carries a low mid-term mortality. It can be performed safely
in high-risk patients, including those with acute type A aortic dissection, BAV, Marfan
syndrome, or requiring redo operation. Efforts to standardize aspects of the procedure,
including graft sizing, commissure placement, and assessment of prolapse have increased
the reproducibility of the procedure. Although minimally invasive approaches are promising,
long-term follow-up is necessary to determine durability.
Central Message
This study examined 81 patients undergoing David procedure using the reimplantation
technique for diverse indications. With careful preoperative selection, VSRR is a
durable procedure for patients with aortic root aneurysm.