Keywords
aortic valve and root - surgery - complications - outcomes (includes mortality - morbidity)
- minimally invasive surgery (includes port access - mini-thoracotomy)
Introduction
Degenerative aortic valve stenosis is the most frequent valvular heart disease in
western countries, and valve replacement is the gold standard for symptomatic severe
cases.[1] In addition to the two established procedures (i.e., surgical aortic valve replacement
[SAVR] and transcatheter aortic valve implantation [TAVI]), the last decade showed
the rising of a third way, which is based on the concept of the surgical replacement,
but taking some advantages of the TAVI prosthesis's design, such as the faster and
simpler anchoring mechanism. This category includes the interventions performed with
the sutureless (Perceval; LivaNova, Italy) and the rapid-deployment (Intuity; Edwards
Lifesciences, California, United States) prostheses. Sutureless and rapid-deployment
(SuRD)-AVR has been applied to high- and intermediate-risk populations with good clinical
results.[2] However, concerns were raised because of the incidence of postprocedural permanent
pacemaker implantation (PPI), which was higher than in conventional SAVR[3] and pretty similar to that observed following TAVI.[4] Possible factors associated with this complication have been assessed only in single-center
studies with small sample populations. Moreover, recent analyses reported a reduction
of this complication over time, suggesting the possible role of a time bias or learning
curve effect.[2]
[5]
[6] For these reasons, we aimed to investigate the need for PPI following SuRD-AVR in
a large cohort using a progress report approach.
Methods
We performed a retrospective cohort study of patients enrolled in the “Sutureless
and Rapid Deployment International Registry” (SURD-IR), a multicentric international
registry that includes patients undergoing SuRD-AVR, irrespective of the surgical
access, in 19 large research centers. The rationale and methods of SURD-IR have been
previously published.[7] Inclusion criteria were age older than 18 years and severe valve disease of the
native aortic valve according to international guidelines referred for isolated or
combined cardiac surgery. Exclusion criteria were prior pacemaker implantation, use
of the off-market sutureless Enable 3F (Medtronic Inc., Minnesota, United States),
lack of in-hospital data, and prior intervention on the aortic valve. For the present
study, 4,166 patients who underwent SuRD-AVR between January 2008 and April 2019 were
suitable for the analysis. Surgical techniques, the choice to perform a full-sternotomy
or minimally invasive AVR, and the postoperative management were performed according
to the specific standards at each center. Ethics approval was obtained at each of
the participating centers.
Although they have been designed according to the same principle of reducing surgical
times, the sutureless and rapid-deployment prostheses are deeply different in design,
technology, and implantation techniques. Moreover, they have been introduced in different
years (2008 for Perceval and 2010 for Intuity), and several centers included in the
registry used only one type of prosthesis. Thus, to minimize inclusion and treatment
bias, we decided to divide the study population according to the implanted prosthesis
(Su cohort and RD- cohort) and to perform the statistical analysis separately.
According to a previous analysis of our register, a constant trend in reduction of
the PPI rate starting from 2016 (from 8.1 to 5.9% on the whole registry's population)
was reported.[2] This prior evidence was used as reference to divide the study population into two
groups: the first including all patients who underwent surgery until December 2016
(early group, “EG”) and the second with all patients between January 2017 and April
2019 (late group, “LG”). The time point division was agreed by all authors. The following
data have been retrieved from the central database and compared between groups: patient
baseline characteristics and comorbidities, functional status, surgical data, postoperative
course, and clinical and hemodynamic outcomes.
The end point for the specific purpose of this analysis was the need for PPI during
the hospital stay following the index procedure.
Statistical Analysis
Categorical variables are reported as absolute values and percentages. Percentages
were calculated with the available data as the denominator. Continuous variables are
expressed as mean ± standard deviation or median and interquartile range when continuous
variables did not follow a normal distribution (tested using the Kolmogorov–Smirnov
test for normality and Q–Q plots). Categorical variables were compared using a two-sided
χ2 analysis or Fisher's exact test, where appropriate.
Differences among patient groups for continuous variables were determined using unpaired
t-test or one-way analysis of variance, as appropriate. A p-value < 0.05 was considered statistically significant. SPSS (version 25.0, IBM Corp.;
New York, United States) was used for all analyses.
Results
Su Cohort: Patients Characteristics
During the study period, 2,604 patients underwent Su-AVR as isolated (1,790) or combined
procedure (814). Of those, 1,934 belong to the EG and 670 to the LG. Patients' baseline
characteristics are listed in [Table 1]. Patients in LG were significantly younger, has a slightly higher incidence of male,
and had a lower incidence of obesity and chronic atrial fibrillation. Accordingly,
both logistic EuroSCORE and EuroSCORE II were significantly lower. Interestingly,
the pathological characteristics were also significantly different, with a lower incidence
of bicuspid aortic valve (BAV) but more frequent relevant baseline aortic valve insufficiency.
Table 1
Patients' demographics
|
Sutureless cohort
|
Rapid-deployment cohort
|
|
Total
(n = 2,604)
|
EG
2008–2016
(n = 1,934)
|
LG
2017–2019
(n = 670)
|
p-Value
|
Total
(n = 1,562)
|
EG
2010–2016
(n = 935)
|
LG
2017–2019
(n = 627)
|
p-Value
|
|
Female
|
1,604 (61.6)
|
1,217 (62.9)
|
387
|
57.8
|
0.02
|
737 (47.2)
|
431 (46.1)
|
306 (48.9)
|
0.3
|
|
Age (y)
|
77 (±6.6)
|
77.6 (±6.3)
|
75.8 (±6.2)
|
<0.001
|
73.9 (±7.8)
|
74.3 (±7.7)
|
73.4 (±7.8)
|
0.02
|
|
NYHA class
|
|
I
|
189 (7.5)
|
115 (6.2)
|
74 (11.3)
|
0.06
|
67 (4.3)
|
48 (5.2)
|
19 (3.1)
|
0.01
|
|
II
|
927 (37)
|
618 (33.3)
|
309 (47.4)
|
496 (32.2)
|
324 (35.2)
|
172 (27.7)
|
|
III
|
1,270 (50.6)
|
1,005 (54.1)
|
265 (40.6)
|
903 (58.6)
|
506 (54.9)
|
397 (64)
|
|
IV
|
122 (4.9)
|
118 (6.4)
|
4 (0.6)
|
75 (4.9)
|
43 (4.7)
|
32 (5.2)
|
|
Hypertension
|
1,918 (86.2)
|
1,330 (85.3)
|
588 (88.4)
|
0.06
|
1,213 (80.8)
|
689 (78.4)
|
524 (84.2)
|
0.005
|
|
Obesity
|
700 (27.4)
|
545 (28.7)
|
155 (23.5)
|
0.01
|
431 (27.9)
|
259 (28)
|
172 (27.8)
|
1
|
|
Diabetes
|
740 (30.5)
|
535 (30.4)
|
205 (30.8)
|
0.8
|
415 (27.8)
|
253 (28.8)
|
162 (26.2)
|
0.3
|
|
AF
|
343 (18.4)
|
253 (20.1)
|
90 (14.7)
|
0.004
|
249 (18.7)
|
136 (18.5)
|
113 (19)
|
0.9
|
|
BAV
|
92 (4.1)
|
79 (5)
|
13 (2)
|
0.001
|
112 (8.4)
|
62 (7.4)
|
50 (10.3)
|
0.09
|
|
Aortic valve disease
|
|
Stenosis
|
1,806 (72)
|
1,348 (73.1)
|
458 (69.1)
|
<0.001
|
1,181 (78.9)
|
708 (79.4)
|
473 (78.3)
|
0.5
|
|
Regurgitation
|
45 (1.8)
|
19 (1)
|
26 (3.9)
|
38 (2.5)
|
25 (2.8)
|
13 (2.2)
|
|
Mixed disease
|
655 (26.1)
|
476 (25.8)
|
179 (27)
|
277 (18.5)
|
159 (17.8)
|
118 (19.5)
|
|
Other
|
1 (0.04)
|
1 (0.1)
|
–
|
–
|
–
|
–
|
|
Cerebrovascular disease
|
239 (11.2)
|
201 (11.7)
|
38 (9.3)
|
0.2
|
256 (16.5)
|
143 (15.4)
|
113 (18.2)
|
0.2
|
|
Renal insufficiency
|
1,236 (55.2)
|
887 (56)
|
349 (53.4)
|
0.3
|
810 (52.9)
|
448 (49.3)
|
362 (58.1)
|
0.03
|
|
Dialysis
|
23 (1.3)
|
15 (1.3)
|
8 (1.3)
|
1
|
13 (0.9)
|
7 (0.9)
|
6 (1)
|
1
|
|
Chronic lung disease
|
325 (14.4)
|
278 (15.1)
|
47 (11.5)
|
0.7
|
225 (15.7)
|
153 (18.9)
|
72 (11.6)
|
<0.001
|
|
Re-intervention
|
172 (6.6)
|
131 (6.8)
|
41 (6.1)
|
0.6
|
76 (4.9)
|
51 (5.5)
|
25 (4)
|
0.2
|
|
LVEF (%)
|
58.7 (±10.8)
|
58.9 (±10.9)
|
58.2 (±9.5)
|
0.2
|
57.4 (±10.8)
|
56.2 (±10.5)
|
59 (±11.1)
|
0.05
|
|
Peak AVG (mm Hg)
|
77 (±24.6)
|
76.7 (±25.3)
|
77.7 (±22.7)
|
0.4
|
81.2 (±27)
|
82.5 (±28.3)
|
79.1 (±24.8)
|
0.04
|
|
Mean AVG (mm Hg)
|
47.2 (±16.1)
|
47.2 (±16.7)
|
47.1 (±14.3)
|
0.8
|
50.1 (±17.6)
|
51.5 (±18.3)
|
49.6 (±16.4)
|
0.05
|
|
Logistic EuroSCORE (%)
|
8.1 [5.6–12]
|
8.6 [6.2–13.1]
|
6.9 [4.7–10.9]
|
<0.001
|
6.7 [4–11.7]
|
7 [4.3–11.7]
|
6.1 [3.3–11.6]
|
0.02
|
|
EuroSCORE II (%)
|
2.7 [1.6–5.2]
|
3.2 [1.8–6.4]
|
2.2 [1.3–3.7]
|
<0.001
|
2.3 [1.4–4.2]
|
2.4 [1.4–4.6]
|
2.1 [1.3–3.8]
|
0.02
|
Abbreviations: AF, atrial fibrillation; AVG, aortic valve gradient; BAV, bicuspid
aortic valve; EG, early group; LG, late group; LVEF, left ventricular ejection fraction;
NYHA, New York Heart Association.
Note: Values are presented as mean (±standard deviation), n (%), or median [interquartile range].
Su Cohort: Operative Data
Operative data are shown in [Table 2]. Over time, a reduction in both full- and mini-sternotomy in favor of a more frequent
use of anterior mini-thoracotomy was observed. Moreover, the incidence of combined
procedures, mostly due to reduction of concomitant myocardial revascularization and
septal myectomy, and consequently the cardiopulmonary bypass time, dropped significantly.
A redistribution of sizing, with a significant increasing use of size “S,” is worthy
of attention.
Table 2
Operative data
|
Sutureless cohort
|
Rapid-deployment cohort
|
|
Total
(n = 2,604)
|
EG
2008–2016
(n = 1,934)
|
LG
2017–2019
(n = 670)
|
p-Value
|
Total
(n = 1,562)
|
EG
2010–2016
(n = 935)
|
LG
2017–2019
(n = 627)
|
p-Value
|
|
Full sternotomy
|
1,252 (48.2)
|
1,034 (53.5)
|
218 (32.6)
|
<0.001
|
777 (49.7)
|
472 (50.5)
|
305 (48.6)
|
0.5
|
|
Mini-sternotomy
|
943 (36.3)
|
747 (38.7)
|
196 (29.3)
|
<0.001
|
572 (36.6)
|
349 (37.3)
|
223 (35.6)
|
0.5
|
|
ART
|
404 (15.5)
|
149 (7.7)
|
255 (38.1)
|
<0.001
|
204 (13.1)
|
113 (12.1)
|
91 (14.5)
|
0.2
|
|
Valve label size[a]
|
|
(−)/19 mm
|
–
|
–
|
–
|
<0.001
|
157 (10.1)
|
97 (10.4)
|
60 (9.6)
|
0.2
|
|
S/21 mm
|
396 (15.5)
|
259 (13.7)
|
137 (20.5)
|
412 (26.5)
|
251 (26.9)
|
161 (25.8)
|
|
M/23 mm
|
949 (37.2)
|
723 (38.3)
|
226 (33.8)
|
496 (31.9)
|
312 (33.5)
|
184 (29.5)
|
|
L/25 mm
|
880 (34.5)
|
665 (35.3)
|
215 (32.2)
|
355 (22.8)
|
198 (21.2)
|
157 (25.2)
|
|
XL/27 mm
|
329 (12.9)
|
239 (12.7)
|
90 (13.5)
|
136 (8.8)
|
74 (7.9)
|
62 (9.9)
|
|
Associated procedures
|
814 (31.3)
|
652 (33.7)
|
162 (24.2)
|
<0.001
|
656 (42)
|
377 (40.3)
|
279 (44.5)
|
0.1
|
|
CABG
|
622 (24)
|
503 (26.1)
|
119 (17.8)
|
<0.001
|
459 (29.4)
|
263 (28.1)
|
196 (31.3)
|
0.2
|
|
Mitral surgery
|
105 (4.2)
|
68 (3.7)
|
37 (5.5)
|
0.06
|
101 (6.5)
|
54 (5.8)
|
47 (7.5)
|
0.2
|
|
Tricuspid surgery
|
29 (1.2)
|
22 (1.2)
|
7 (1)
|
0.8
|
54 (3.5)
|
29 (3.1)
|
25 (4)
|
0.4
|
|
AF surgery
|
52 (2.1)
|
39 (2.2)
|
13 (1.9)
|
0.8
|
72 (4.6)
|
41 (4.4)
|
31 (4.9)
|
0.6
|
|
Thoracic aorta surgery
|
23 (0.9)
|
19 (1)
|
4 (0.6)
|
0.4
|
72 (4.6)
|
42 (4.5)
|
30 (4.8)
|
0.8
|
|
Septal myectomy
|
66 (2.7)
|
62 (3.4)
|
4 (0.6)
|
<0.001
|
42 (2.7)
|
18 (1.9)
|
24 (3.8)
|
0.03
|
|
CPB time (min)
|
67 [51–89]
|
77 [53–90]
|
61 [47–84]
|
<0.001
|
97 [74–123]
|
99 [75–124]
|
96 [73–123]
|
0.8
|
|
X-clamp time (min)
|
43 [32–60]
|
44 [32–60]
|
43 [32–62]
|
0.6
|
65 [49–85]
|
65 [48–84]
|
65 [50–87]
|
0.3
|
|
Valve malpositioning
|
22 (1.1)
|
13 (0.9)
|
9 (1.6)
|
0.3
|
24 (1.7)
|
16 (2.1)
|
8 (1.3)
|
0.3
|
Abbreviations: AF, atrial fibrillation; ART, anterior right thoracotomy; CABG, coronary
artery bypass grafting; CPB, cardiopulmonary bypass; EG, early group; LG, late group.
Note: Values are presented as n (%) or median [interquartile range].
a Valve label sizes are different between sutureless and rapid-deployment cohorts and
cannot be directly compared. The sutureless prosthesis is available in the following
sizes: “S,” which should cover a range of annulus between 19 and 21 mm; “M,” between
21 and 23 mm; “L,” between 23 and 25 m; and “XL” between 25 and 27 mm. The rapid-deployment
is available in five sizes (ranging from 19 to 27 mm).
Su Cohort: PPI and Other Outcomes
A marked reduction of PPI was observed between the groups (EG = 209/1,934 [10.8%]
vs LG = 42/670 [6.3%], p < 0.001) ([Table 3] and [Fig. 1]). The same reduction was also present in the subgroup of isolated procedures (EG = 129/1,282
[10.1%] vs LG = 33/508 [6.5%], p = 0.02). [Fig. 2] shows the analysis of PPI incidence according to implanted prosthesis's size. Beside
the above-mentioned trend to undersizing, a significant reduction of PPI was found
in those patients who received a prosthesis “L” (from 12.5 to 7.4%) and “XL” (from
12.6 to 4.4%).
Fig. 1 Bar chart showing the incidence of PPI in the sutureless and rapid-deployment groups
in the early and late period of time. A significant reduction of PPI following sutureless
AVR has been observed in last years, together with changes in patients' characteristics.
The same phenomenon has not been registered for rapid-deployment AVR, but any substantial
difference in strategy has been recorded.
Fig. 2 Bar chart showing the use and distribution of PPI according to prosthesis's size
in sutureless cohort.
Table 3
In-hospital outcomes
|
Sutureless cohort
|
Rapid-deployment cohort
|
|
Total
(n = 2,604)
|
EG
2008–2016
(n = 1,934)
|
LG
2017–2019
(n = 670)
|
p-Value
|
Total
(n = 1,562)
n %
|
EG
2010–2016
(n = 935)
|
LG
2017–2019
(n = 627)
|
p-Value
|
|
Hospital mortality
|
57 (2.3)
|
49 (2.5)
|
8 (1.4)
|
0.2
|
29 (1.9)
|
17 (1.8)
|
12 (1.9)
|
1
|
|
Stroke
|
70 (3.1)
|
50 (3.2)
|
20 (3.1)
|
1
|
40 (2.8)
|
28 (3.4)
|
12 (1.9)
|
0.1
|
|
Ventilatory support > 72 h
|
113 (4.8)
|
89 (4.6)
|
24 (5.8)
|
0.3
|
70 (4.5)
|
40 (4.3)
|
30 (4.8)
|
0.7
|
|
New-onset AF
|
631 (29)
|
530 (29.8)
|
101 (25.2)
|
0.07
|
374 (26.5)
|
227 (28.9)
|
147 (23.4)
|
0.02
|
|
PPI
|
251 (9.6)
|
209 (10.8)
|
42 (6.3)
|
<0.001
|
140 (9)
|
82 (8.8)
|
58 (9.3)
|
0.8
|
|
Bleeding requiring revision
|
74 (4.4)
|
51 (4.2)
|
23 (5.1)
|
0.4
|
68 (4.4)
|
45 (4.9)
|
23 (3.7)
|
0.4
|
|
AKI > stage 1
|
101 (5.9)
|
81 (5.9)
|
20 (5.8)
|
0.9
|
71 (4.7)
|
40 (4.3)
|
31 (5.4)
|
0.4
|
|
Temporary dialysis
|
54 (3.1)
|
40 (3)
|
14 (3.5)
|
0.6
|
37 (2.4)
|
18 (2)
|
19 (3)
|
0.2
|
|
ICU stay (d)
|
2 [1–3]
|
2 [1–3]
|
2 [1–3]
|
0.7
|
2 [1–3]
|
2 [1–3]
|
2 [1–3]
|
0.9
|
|
Hospital stay (days)
|
9 [7–13]
|
9 [7–14]
|
8 [7–12]
|
0.02
|
11 [8–17]
|
11 [8–17]
|
11 [8–18]
|
0.7
|
|
Aortic regurgitation
|
276 (11.9)
|
235 (14.2)
|
41 (6.3)
|
<0.001
|
88 (6)
|
58 (6.7)
|
30 (5.1)
|
0.5
|
|
Mild
|
241 (10.4)
|
211 (12.7)
|
30 (4.6)
|
|
61 (4.2)
|
40 (4.6)
|
21 (3.6)
|
|
|
Moderate
|
32 (1.4)
|
21 (1.3)
|
11 (1.7)
|
|
18 (1.2)
|
13 (1.5)
|
5 (0.8)
|
|
|
Severe
|
3 (0.1)
|
3 (0.2)
|
–
|
|
9 (0.6)
|
5 (0.6)
|
4 (0.7)
|
|
|
Peak AVG (mm Hg)
|
26.3 (±10.1)
|
26.5 (±10.4)
|
25.7 (±9.3)
|
0.08
|
20.3 (±8.8)
|
20.6 (±8.9)
|
19.9 (±8.6)
|
0.3
|
|
Mean AVG (mm Hg)
|
14.2 (±5.6)
|
14.2 (±5.8)
|
14.2 (±5.2)
|
0.8
|
11.1 (±5)
|
11.1 (±5)
|
11.1 (±5)
|
1
|
Abbreviations: AF, atrial fibrillation; AKI, aortic kidney injury; AVG, aortic valve
gradient; EG, early group; ICU, intensive care unit; LG, late group; PPI, permanent
pacemaker implantation.
Note: Values are presented as mean (± standard deviation) or n (%), or median [interquartile range].
RD Cohort: Patients Characteristics
The RD cohort consists of 1,562 patients operated between 2010 and April 2019 (EG = 935;
LG = 627). Patients' baseline characteristics are shown in [Table 1]. Even in this cohort, a significant reduction of logistic EuroSCORE and EuroSCORE
II has been observed over time, mainly due to a younger age in LG.
RD Cohort: Operative Data
Intraoperative variables are shown in [Table 2]. No significance differences were found in intraoperative variables, but the incidence
of septal myectomy was twice higher in LG (3.8 vs 1.9%, p = 0.03). The use of different prosthesis sizes was not significantly different between
EG and LG.
RD Cohort: PPI and Other Outcomes
The incidence of PPI was higher in LG, but not statistically significant (EG = 8.8%
vs LG = 9.3%, p > 0.05; [Fig. 1]). In the subgroup of isolated AVR, no differences were observed (EG = 7.5% vs LG = 7.5%,
p > 0.05). As shown in [Fig. 3], no significant differences were observed in the analysis according to prosthesis's
size. The only difference in clinical outcomes was a reduction in the incidence of
new onset of atrial fibrillation ([Table 3]).
Fig. 3 Bar chart showing the use and distribution of the PPI according to prosthesis's size
in rapid-deployment cohort.
Discussion
Concerns about a higher rate of PPI following SuRD-AVR limited the routine use of
these prostheses in the low-risk population so far. Indeed, the incidence of PPI after
SAVR with a standard prosthesis is reported to be between 2.6 and 3.9%.[8]
[9] Previous analyses, from our group as well as individual experiences, contributed
to the hypothesis that the learning curve is one of the reasons for this outcome.[2]
[5]
[6] A learning curve effect is usually difficult to investigate as a variable. Indeed,
it influences not only the implantation technique and the surgical performance, but
also the patient's selection. Moreover, as the learning curve could be different between
centers, researchers usually perform single center's analysis or enroll centers with
similar characteristics to avoid this bias, making the collection of a large sample
population difficult. The scope of our progress report was to address this problem
without waiving a large sample size. We aimed to use this approach to shed light on
the issue, to report a fair “state of the art” of the problem, and possibly to shorten
the learning effect for the surgeons who are not yet confident with SuRD-AVR.
The main findings of our report are (1) the incidence of PPI following Su-AVR has
significantly decreased over time and (2) PPI after RD-AVR has remained stable over
time.
The reduction of PPI after Su-AVR has been already reported not only from our international
register, but also from other single-center studies.[5]
[9] In the present report, several clinical variables, such as age, gender, obesity,
history of atrial fibrillation, and risk profile, were found significantly different
between the two temporal groups and possibly may have influenced the observed outcome.
Interestingly, also anatomical and functional characteristics of the aortic valve,
namely, the presence of a congenital BAV and the prevalence of pure aortic valve stenosis,
were significantly different between the groups. The association of BAV with increased
risk of PPI following aortic valve interventions has been already investigated in
prior studies with conflicting results. Haunschild and colleagues reported no differences
in PPI between patients with BAV and tricuspid valves in a matched population, but
they found a higher rate of third-degree atrioventricular block in BAV patients.[10] Biswas and colleagues recently found a higher His-to-ventricular interval conduction
in BAV patients, with an increased requirement for pacemaker therapy over a 10-year
follow-up in comparison to a matched population with tricuspid aortic valve.[11] As previously reported, the overall incidence of PPI in BAV patients was 7.9% in
our registry.[12] The lower incidence of aortic valve stenosis could also have contributed to the
improvement of the outcome. Indeed, the main predominant etiology of aortic stenosis
is the calcific degeneration, which may extend from the aortic ring to the bundle,
provoking right bundle and left anterior hemiblock as well.[13]
In addition to the baseline variables, even changes in intraoperative strategies could
be responsible for the reduction in PPI in Su cohort. We found a clear shift over
time toward a more frequent use of anterior right thoracotomy and a reduction of combined
procedures, mainly concomitant revascularization and septal myectomy. In other words,
patients underwent a significantly simpler procedure, in a shorter time (as observed
from the shorter cardiopulmonary bypass time). Less need for myocardial revascularization
could be correlated to less conduction disturbances, as the conduction system is sensitive
to ischemic conditions.[14]
However, the most interesting finding of our study is the significant reduction of
PPI in two specific sizes of the Su prosthesis (the “L,” covering annuli between 23
and 25 mm, and the “XL,” between 25 and 27 mm). Remarkably, the use of these prostheses
was not significant different in the LG, suggesting that the surgeons moved to a different
sizing strategy in the most recent period. Indeed, the period of time of our LG coincides
with the publications of several studies focused on the risks (e.g., stent recoil,
higher gradients) correlated to the oversizing in Su-AVR.[15]
[16] The avoidance of oversizing could be correlated to a reduction of conduction injury,
in accordance with the study of Geršak and colleagues.[17]
A possible alternative explanation of the reduction of PPI in patients receiving the
“XL” prosthesis could be related to the introduction on the market in mid-2018 of
an updated model called “Perceval Plus.” Both Su and RD prostheses, despite the considerable
structural differences, are based on the intra-annular position with an anchoring
system based on radial forces. This can result in a damage of the left bundle branch
fibers of His bundle, located close to the membranous septum, just beneath the commissures
of the right and noncoronary cusps. This may explain the higher incidence of PPI with
these new prostheses in comparison to conventional bioprostheses (which typically
are placed in a supra-annular position). The use of the “Perceval Plus XL” characterized
by a thinner annular ring and a different distribution of radial forces could have
contributed to a reduction of PPI, at least in the XL group.
The RD cohort showed a stable rate of PPI. A recent multivariate analysis of Coti
and colleagues performed on 700 patients showed that a baseline right bundle branch
block, concomitant procedures in general and in particular concomitant myocardial
revascularization are correlated with the atrioventricular block after RD-AVR.[18] Our findings—with the limitation of the missing electrocardiographic (ECG) data—are
in line with those results, as the only significant differences over time were the
younger age of patients in LG and a higher rate of septal myectomy. The latter finding
must receive particular attention, as atrioventricular blocks belong to the typically
consequences of subvalvular myectomy and this could explain the observed trend of
higher rate over time (in the overall population, while the isolated AVR showed no
changes).
Although the scope of the present analysis was not a direct comparison of the two
prosthesis, it is worthy to highlight the differences in the Su and RD groups, not
only in the outcomes (no reduction in PPI rate over time and consequently longer hospital
stay: 11 vs 8 days), but also in the baseline and procedural variables (e.g., higher
incidence of associated procedures in the RD group, 42 vs 31%). The choice of the
prosthesis in the present study was at the discretion of the surgeon. As the two prostheses
are structurally different, they could be more suitable in different scenario, thus
influencing the prosthesis choice by the surgeon. For example, due to the smaller
profile, some surgeon could prefer the RD prosthesis in cases requiring revascularization
with multiple proximal anastomosis.
Summing up these observations, as well as the findings from the Su cohort, we may
conclude that patient selection plays a crucial role in the strategy aiming to reduce
the risk of PPI. Our findings do not exclude the possibility to obtain a similar reduction
also in the RD cohort in the future, as we recently hypothesized.[19] Specific risk factors for pacemaker implantation (mainly right bundle branch block),
which could be avoided by patient selection, may improve the pacemaker rate in the
RD group in the future. However, the rapid-deployment and the sutureless prostheses
are structurally different: the Su prosthesis does not protrude in the left ventricular
outflow tract at all, unlike the RD prosthesis. If this aspect is relevant for the
risk of developing atrioventricular blocks, especially when severe subvalvular calcifications
are present,[4] it should be cleared in future studies.
Our data are drawn from a population in the seventh decade of life with a life expectancy
of around 10 years. If the improvements of PPI rate, as observed in our study, will
be confirmed in further studies with long-term follow-ups, the use of SuRD prostheses
could be expanded to a younger and low-risk population, for which PPI could be undesirable
due to the increased risks of all-cause mortality and heart failure hospitalization
in a very long-term follow-up.[20]
Limitation of the Study
Limitations of the present study included its retrospective nature. Moreover, potentially
important variables for the need for PPI, such as pathological preoperative ECGs,
or preexisting conduction abnormalities, were not available in the dataset. The differences
between Su and RD cohorts did not allow a direct comparison between these two groups.
Our analysis is focused on the in-hospital outcome and does not investigate the mid-
or long-term outcome.
Conclusion
The present analysis showed a significant decrease in in-hospital PPI requirement
rate in patients who underwent Su-AVR over time. Patient selection as well as surgical
modifications and a more accurate sizing are probably the main reasons for this phenomenon.
The RD cohort revealed no differences in PPI between the two time periods so far,
but there were also less differences in patient selection.
