Keywords
complete atrioventricular septal defects - surgical repair - long-term results
Introduction
Complete atrioventricular septal defects (CAVSDs) account for approximately 2.5% of
congenital heart anomalies.[1] In patients with Down syndrome, nearly 50% have an atrioventricular (AV) septal
defect, usually the complete variant.[2] Since the first successful repair of a CAVSD by Lillehei in 1955,[3] three different techniques have been developed to accomplish repair. The single-patch
technique was introduced by Maloney in 1962,[4] followed by the two-patch technique by Trusler in 1975,[5] and finally the modified single-patch technique by Wilcox and Nunn[6]
[7] in the late 1990s. Today all techniques are in use at discretion of the surgeon.
In contrast to the two-patch and modified single-patch techniques, the original single-patch
technique requires dividing of the bridging leaflets and thus, this technique is of
particular interest for the long-term AV valve (AVV) function. Patients with repaired
CAVSD resemble a growing population in grown-ups with congenital heart disease (GUCHs).
With respect to this, we reviewed our long-term results in patients with CAVSD who
had undergone single-patch technique repair up to 40 years ago.
Patients and Methods
Between 1978 and 2001, 100 consecutive patients with isolated CAVSD underwent single-patch
repair at our institution. This technique was fully standardized (B.J.M.) and applied
in unison.
Patient Selection
In this historic series, hospital mortality was 11%. The 89 hospital survivors were
subjected to a survey in 2020 using a standardized postal questionnaire or telephone
interview with the patients and their cardiologists. If the patient and/or cardiologist
was not detectable after decades, the most recent follow-up was considered.
Variables
Demographic Variables
Prior to CAVSD Repair
At the time of CAVSD Repair
-
Rastelli type of CAVSD, presence of a double orifice “mitral” valve, left-sided papillary
muscle anatomy (single, narrow spaced, normally spaced), cleft management (open, partially
closed, completely closed), and presence of third degree AV block.
At the Time of Follow Up
-
Death, duration of follow-up, age, QoL (self- or caregiver's estimation), congestive
heart failure (New York Heart Association [NYHA] classification), cardiac medication,
electrocardiography (rhythm, conduction, tachycardia), 2-day echocardiography (residual
shunts, left-ventricular systolic function, AVV's function, subaortic stenosis (SAS),
aortic valve function), and CAVSD-related reoperations.
Surgical Technique
In brief, repair of CAVSD was done via median sternotomy in deep hypothermic circulatory
arrest at 20 °C rectal temperature. If repair could not be completed within approximately
60 minutes, hypothermic low flow bypass was commenced. The right atrium was opened
and cold crystalloid cardioplegic solution was administered. After determining the
line of apposition, the anterior and posterior bridging leaflets were divided in great
favor of the left side leaving only a small rim of tissue for reconstruction of the
“septal” part of the right-sided AVV (R-AVV). A Dacron patch was cut to fit into the
inlet ventricular septal defect (VSD) far on the right side to prevent from SAS as
well as to fit into the primum atrial septal defect (ASD). The patch was sutured to
the right part of the crest using continuous 5–0 Prolene. Next, the left-sided AVV
(L-AVV) components of the bridging leaflets were resuspended to the patch using 6–0
Ethibond mattress sutures without pledgets and were tied on the right side of the
patch. The cleft was closed with several interrupted 6–0 Prolene sutures. The L-AVV
was tested with cold normal saline. If necessary, additional commissuroplasty was
done. The remainder of the patch was sewn in place to close the primum ASD. Finally,
the bridging leaflet components of the R-AVV were resuspended to the patch using 6–0
continuous Prolene. The right atrium was closed and the patient rewarmed and discontinued
from bypass and the chest was closed in standard fashion. At that time, intraoperative
transesophageal echocardiography was not performed routinely and since echocardiography
in the late 1970s and early 1980s was not given or is not comparable to today's standards,
echo findings were focused on the status quo at follow-up.
Statistical Analysis
Statistical analysis was done with SPSS version 22 (SPSS Inc., Chicago, Illinois,
United States). Missing values were not replaced and were considered as empty cells
in all calculations. Central tendency is expressed by mean and median, dispersion
by standard deviation and minimum–maximum. Normal distribution was ruled out by the
Kolmogorov–Smirnov test for most parameters. Consequently, differences among groups
were assessed using the two-tailed chi-square or Fisher's exact test for binary and
categorical variables and the Mann–Whitney U-test for continuous variables. A logistic
regression model was applied, if appropriate. For actuarial survival and for freedom
from reoperation, Kaplan–Meier estimates were calculated; in case of more than one
reoperation in a single patient, a new case was created for each reoperation. Primary
endpoints were QoL and clinical status, AVV function, and freedom from reoperation
in long term. A value of p <0.05 was considered statistically significant.
Results
Follow-Up
Eighty-nine patients were subjected to follow-up. Of these, six patients (6.7%) were
identified to be survivors, but no additional information could be acquired. These
patients were included for survival estimates, but excluded for further analysis.
In the remaining 83 patients (study group), age at follow-up was 22.9 ± 9.3 (22; 2.5–44.9)
years; 65 patients (78%) were older than 15 years, 52 (63%) older than 20, and 18
(22%) older than 30 years; 42 patients (51%) were male and 65 (78%) had trisomy 21.
Follow-up after surgical repair was 21.0 ± 8.7 (21.5; 2.1–40) years (1,744 patient-years).
[Fig. 1] summarizes the status quo at the time of follow-up.
Fig. 1 Status quo at follow-up. ACE, angiotensin converting enzyme; AV, atrioventricular;
LA, left atrial; LV, left ventricular; LVOT(O), left ventricular outflow tract (obstruction);
NYHA, New York Heart Association; SVES, supraventricular extrasystole; SVT, supraventricular
tachycardia; VES, ventricular extrasystole.
Long-Term Survival
Late mortality (n = 3) was related to pulmonary hypertension, unknown reason, and pacemaker malfunction
after trauma 2.1, 8.3, and 8.5 years after CAVSD repair, respectively. Thus, long-term
actual mortality was 3.4% and actuarial long-term survival was 98.8, 96.3, and 96.3%
at 5, 10, and 20 years (number of patients at risk: 78, 73, and 49, respectively).
Quality of Life and Clinical Status
Quality of Life
QoL estimation (self- or caregiver reported in patients with Down syndrome) was obtained
from 67/80 patients (3 nonsurvivors and 13 nonresponders to the question regarding
their QoL were excluded). Considering all 80 eligible patients, 65 (81%) considered
their QoL excellent or good. In two patients, one with Down syndrome, QoL was estimated
as impaired. The only pathological finding in these 28-year-old patients was a moderate
L-AVV regurgitation and mild congestive heart failure. There was no statistically
significant difference in estimated QoL between patients with or without trisomy (p = 0.247).
Clinical Status
Sixteen patients (19%) had congestive heart failure, mostly mild: of these, three
patients were in NYHA stage III; one died 2.1 years after CAVSD repair and pacemaker
implantation. This patient developed SAS, moderate L-AVV regurgitation, and pulmonary
hypertension. The other two patients, age 28 and 38, both with trisomy 21, had an
unremarkable follow-up, except for mild L-AVV regurgitation.
Rhythm
Detailed information is displayed in [Fig. 1]. A surgical complete AV-block has implications for lifetime. Thus, it accounts for
the majority of reoperations. In this series there were no major complications related
to pacemaker therapy. A notable finding is the occurrence of a late-onset sick sinus
syndrome (SSS) in five patients 8.6 to 16.3 years after CAVSD repair necessitating
pacemaker implantation. The occurrence of SSS was not related to any other investigated
variable. All patients with a history of supraventricular tachycardia or ventricular
extrasystole were treated with β-blockers resulting in symptom and tachyarrhythmia
free regular follow-up.
Echocardiographic Findings
In terms of surgical technique, the function of the AVVs is of major interest. Functional
parameters of the L-AVV were analyzed in relation to pre- and intraoperative findings
as well as to follow-up data ([Table 1]).
Table 1
Long-term left AV-valve incompetence related to potential risk factors
|
Variable
|
LAVVI > grade 1 (n = 23)
|
Prob.
|
|
Gender
Male (n = 42)/female
|
26%/32%
|
0.629
|
|
Trisomy 21
Yes (n = 65)/no
|
29%/29%
|
1.0
|
|
Previous PAB
Yes (n = 10)/no
|
26%/50%
|
0.141
|
|
Rp/Rs > 0.20
Yes (n = 39)/no
|
16%/37%
|
0.092
|
|
Rastelli type
A (n = 63)/B(2)/C(18)
|
29%/100%/19%
|
0.057
|
|
Papillary muscle
Normally spaced (n = 72)/narrow (2)/single (5)
|
29%/0%/40%
|
0.466
|
|
Double orifice
Yes (n = 5)/no
|
31%/0%
|
0.314
|
|
Cleft LAVV
Open (n = 6)/partially closed (27)/closed (47)
|
17%/26%/32%
|
0.683
|
|
NYHA
Stage I (n = 67)/II (13)/III (3)
|
25%/50%/33%
|
0.200
|
|
Pacemaker
Yes (n = 12)/no
|
55%/25%
|
0.069
|
|
LAVVS (any)
Yes (n = 13)/no
|
31%/28%
|
1.0
|
|
LVOTO (any)
Yes (n = 10)/no
|
40%/27%
|
0.462
|
|
Aortic valve reg.
Yes (n = 15)/no
|
47%/25%
|
0.116
|
Abbreviation: AV, atrioventricular; LAVV, left atrioventricular valve; LVOTO, left
ventricular outflow tract (obstruction); NYHA, New York Heart Association; PAB, pulmonary
artery band; Rp/Rs, pulmonary to systemic resistance.
Note: Statistical tendency toward significance (p < 0.10) is highlighted in bold and cursive.
Right AV-Valve
A more than mildly incompetent R-AVV was found only in four patients (4.8%), thus
analytic statistics were not performed. All had Rastelli type A. The only patient
with a moderate R-AVV regurgitation had additional moderate L-AVV regurgitation, mild
SAS, mild aortic insufficiency, and history of PAB. Regurgitation did not allow for
right-ventricular systolic pressure estimation in the majority of patients, thus,
a correlation between regurgitation and pressure could not be evaluated.
Left AV-Valve
In contrast to the R-AVV, there were more patients with more than mild regurgitation
(28%), which was not related to the duration of and age at follow-up (p = 0.114 and 0.133, respectively). There were no statistically significant relations
of L-AVV regurgitation to any investigated variables; however, a more than mild L-AVV
regurgitation tends to be related to an increased Rp/Rs before CAVSD repair, Rastelli
type B, and the presence of a pacemaker ([Table 1]).
There were 14 patients with L-AVV stenosis, mostly mild (n = 12). The stenosis was statistically related to the presence of a double-orifice
L-AVV (3/5 (60%) vs. 10/75 (13%); p = 0.028). However, L-AVV stenosis was mild in these cases. Double orifice was genuine
anatomy not related to an approach to repair an incompetent L-AVV. Stenosis was not
related to the Rastelli type of anatomy (p = 0.721), presence of a single- or narrow-spaced papillary muscle (p = 0.610), or cleft management (p = 0.405). However, not from the statistical point of view, but in practical terms,
the anatomy and the surgical technique most likely alter the AVV function ([Fig. 2]).
Fig. 2 21-year-old CAVSD repair (with partial closure of the cleft). 3D echocardiography
patient post (view from left ventricle) showing full opening of the left-sided AV-valve
(A), the partial closure of the left (B), and the thickness cleft line in ventricular systole (C). Restriction of the anterior leaflet of the reconstructed left-sided AV valve and
partial closure of the cleft led to mild to moderate insufficiency; 2D echocardiography
seen from four-chamber view (D). CAVSD, complete atrioventricular septal defect.
Reoperations
Thirty-four reoperations deemed necessary in 21 patients (25%) ([Fig. 3]). Among those, surgical complete AV-block and late SSS were the most common etiologies
for reoperations and for multiple surgical interventions in single patients. Reoperations
for L-AVV pathology was the second most common, necessitating up to three reoperations
in a single patient ([Table 2]). Reoperation for SAS deemed necessary in four patients in whom there is no residuum
21 to 30 years after SAS repair.
Table 2
Details in patients with reoperation of the left-sided AV-valve
|
T 21
|
Age at CAVSD repair (months)
|
Rast. type
|
DOMV
|
Left-sided pap. muscle
|
Cleft management
|
Indication for (initial) valve reoperation
|
Age/type of first redo
|
Age/type of second redo
|
Age/type of third redo
|
Age at follow-up (years)
|
Result at follow-up
|
|
# 1
|
Yes
|
14
|
A
|
No
|
Single
|
Open
|
Severely malformed AV-valve; regurgitation
|
14 months (at CAVSD repair)/replaced by turned SJM Nr. 21 standard aortic mechanical
prosthesis
|
10 years/upsizing to SJM Nr. 25 HP mitral mechanical prosthesis
|
None
|
35
|
AV-block at first valve replacement consecutive PM implant, further course unremarkable
|
|
# 2
|
Yes
|
9
|
C
|
No
|
Normal
|
Closed
|
Fungal endocarditis; regurgitation
|
10 months/valve repair
|
14 months/stenosis. Replaced by turned SJM Nr. 19 standard aortic bileaflet valve
prosthesis
|
15 years/upsizing to SJM Nr. 25 HP mitral mechanical prosthesis
|
28
|
Unremarkable
|
|
# 3
|
No
|
13
|
C
|
No
|
Single
|
Partially closed
|
Stenosis
|
10.6 years/replaced by SJM Nr. 25 HP mitral mechanical prosthesis
|
None
|
None
|
41
|
Mild/moderate stenosis at mitral valve prosthesis
|
|
# 4
|
Yes
|
11
|
C
|
No
|
Normal
|
Partially closed
|
Regurgitation
|
3.5 years/valve repair
|
None
|
None
|
28
|
Mild incompetence, mild stenosis, mild–moderate SAS
|
|
# 5
|
No
|
2
|
A
|
Yes
|
Normal
|
Open
|
Regurgitation
|
19.3 years/valve repair
|
None
|
None
|
22
|
Mild incompetence, mild stenosis
|
Abbreviations: AV, atrioventricular; CAVSD, complete atrioventricular septal defect;
DOMV, double-orifice mitral valve; SAS, subaortic stenosis.
Fig. 3 Freedom from CAVSD repair related reoperations.
Discussion
Since its introduction in the early 1960s,[4] the single-patch technique has become an established technique for repair of CAVSDs.
However, this technique requires division of the bridging leaflets, and thus is of
concern, particularly in the long term. We applied this technique solely and unmodified
from 1978 until 2001 resulting today in a large group of GUCH patients repaired uniformly
and providing a long follow-up.
Long-term actual mortality was 3.4% and actuarial long-term survival was 98.8, 96.3,
and 96.3% at 5, 10, and 20 years, which is in line with other large and recent series.[8]
[9]
[10] Reasons for death were unknown in one and disease-related in two cases.
Quality of Life and Clinical Presentation
Up to 40 years after CAVSD repair, in the majority of patients a self- or proxy-reported
(caregivers of patients with Down syndrome) excellent or good QoL was estimated, with
no difference for patients with trisomy 21. The possible impact of procedural-related
factors (e.g., deep hypothermic circulatory arrest) on long-term (e.g., neurologic)
outcome is a matter of concern. However, a differentiated statement based on our data
is not possible. Application of established standardized tests to evaluate health-related
QoL or objective exercise capacity in patients with Down syndrome is not possible.
This may reflect the sparse information on these issues in the literature as well.
However, Graves et al have reported on self-reported QoL assessment in a group of
adults with Down syndrome, irrespective of the diagnosis of a cardiac defect: they
had average scores about one negative standard deviation below normative samples for
healthy persons. Especially their physical role scores were reduced due to the impact
of diverse health problems on typical accomplishments.[11]
Following the NYHA classification of congestive heart failure, no or only mild limitation
of the subjective physical exercise capacity was estimated for 96% of the patients.
This is underlined by the excellent clinical status and the very limited need for
cardiac medication which is mainly administered for anticoagulation in patients with
mechanical heart valve prostheses and for atrial or ventricular arrhythmias. Systolic
left ventricular function was within the normal range, additionally supporting the
estimation in terms of good physical exercise capability and low prevalence of congestive
heart failure.
Atrioventricular Valve Function
The repaired AVVs, particularly the left one, have always been a matter of concern,
particularly after single-patch repair. In our series, a more than mild R-AVV regurgitation
was only found in 5% of patients and has not been an indication for reoperation so
far. There was no stenosis, although at CAVSD repair, tissue of the common AVV, has
been separated generously in favor toward the “left side.” The L-AVV is hemodynamically
the most loaded valve of the heart. In CAVSD, this valve is, in addition, structurally
different from a normal mitral valve; furthermore, it is subject to subtle surgical
repair. Malfunction eventually will have impact on exercise capability, QoL, and life
expectancy.[12]
[13]
Nevertheless, a moderate regurgitation was found in only 15% of patients with no indication
for reoperation up to the time of the study. There was a trend (0.1 > p > 0.05) toward more than mild regurgitation for patients with higher pulmonary artery
resistance prior CAVSD repair, Rastelli type B, and the presence of a pacemaker. So
far, there are no long-term data available confirming these observations. However,
from a pathoanatomical standpoint, these findings may be related to the severity of
regurgitation. In contrast, we found that a moderate regurgitation of the L-AVV was
not related to the presence of trisomy 21, history of pulmonary artery banding, cleft
management, and duration of or age at follow-up. Particularly, pulmonary artery banding
has been of concern altering AVV function. In our series, 10 patients had prior pulmonary
artery banding; however, it was not related to L-AVV regurgitation in the long term.
In contrast to our finding, Buratto and colleagues report freedom from L-AVV reoperation
of 60% after 20 years in a series of 40 patients with prior pulmonary artery banding.[14]
Given a normal systolic left ventricular function and no residual shunts in all patients
in our series, evaluation and comparison of bilateral AV function should be reliable.
Both AVVs were statistically evaluated taking preoperative, operative, and postoperative
variables and follow-up time into account. In general, stenosis was not found on the
R-AVV, where it was found to a mild to moderate extent in 2% in the L-AVV. More than
mild regurgitation was more common on the “mitral” side as well (28 vs. 5%). This
result is probably related to the higher pressure load on the systemic side and reflects
why reoperations on the “tricuspid” side are rare after CAVSD repair.
Reoperations
Atrioventricular-Valve-Related Reoperations
In general, L-AVV regurgitation represents the most common indication for reoperation
in 7 to 20% of patients.[8]
[15]
[16]
[17]
[18] Freedom from reoperation for L-AVV malfunction has been reported as low as 62% after
10 years.[16]
In our series, reoperation for L-AVV pathology deemed necessary in five patients (6%),
which is comparable with other series of single-patch repair.[19] Taking into account the reoperations for valve upsizing, actuarial freedom from
“mitral” reoperation was 93% after 10 and 89% after 20 years, this again nicely fits
to the experience of other groups.[19] Comparing these results with those of the Australian CAVSD study group[10] having included more than 800 patients with two-patch or modified single-patch repair,
freedom from L-AVV index (L-AVVI)-related reoperation at 10 and 20 years was 82.7
and 77%, respectively. In addition, Fong and colleagues found that patients without
Down syndrome and moderate L-AVV regurgitation on postoperative echocardiogram were
found to be independent risk factors for reoperation in the Australian series. We
did not find a relation between the presence of trisomy and L-AVV malfunction. There
was no obvious reason for these very promising results, nevertheless, as surgical
results are undoubtedly related to small technical details, comparing the “same” technique
among different surgical teams does not mean that they are necessarily fully comparable.
Backer and colleagues performed a meta-analysis of the three repair techniques and
found that the rate of reoperation for L-AVV regurgitation was 2% in the modified
single-patch versus 7% in the two-patch group and 9.7% in the classic single-patch
group, respectively.[20] In a large recent series, Schleiger and colleagues report on an indication for reoperation
for L-AVVI after repair of different anatomical variants of AVSD of 13.8%.[21] Other series of CAVSD repair using the two-patch technique proved a 10-year freedom
from LAVV reoperation of 94%.[15]
However, we did not provide a comparative study of different repair techniques; nevertheless,
we report on a considerable number of GUCH patients in whom the need for reoperation
for AVV malfunction was low.
In case of irreparable L-AVV, as we have experienced in three patients, mechanical
valves are the usual option in pediatric patients.[22] If implanted early in life, there is a need for upsizing the prosthesis due to outgrowth
which may be achieved with low morbidity and mortality.[22]
[23] Even after three decades on mechanical valves in mitral position, there was an unremarkable
history in terms of thromboembolism or bleeding events in all, underlining the experience
in the literature.[22]
[23]
[24] It is probably noteworthy that long-term self-management of anticoagulation may
play an important role in this setting.[24]
Pacemaker-Related Reoperations
In contrast to the majority of series, L-AVV pathology was not the most common indication
for reoperation. We found that pacemaker-elated reoperations were more common and
that the number of patients with immediate damage to the conduction system during
CAVSD repair was increased by a considerable number of patients with SSS in the long
term. This is a known finding and varies between 2.3 and 12%.[16]
[25] In our series, the incidence of SSS-related pacemaker implantation was 6%. However,
as we have observed that SSS may develop up to 16 years post-CAVSD repair, life-long
observation deemed necessary. Taking both etiologies into account, there is a low
freedom from pacemaker-related freedom from reoperations, i.e., 70% after 20 and only
52% after 30 years. Although, there were no major issues related to pacemaker therapy
in general, one of our long-term deaths was most likely related to pacemaker malfunction.
Left Ventricular Outflow Tract Obstruction
SAS is a well-known mid- to long-term complication after CAVSD repair.[26]
[27]
[28] In a large series from Mayo Clinic, Stulak et al report that SAS was responsible
for approximately 10% of all patients requiring reoperations.[29] In our series, 4.8% of all patients required SAS resection, which is in line with
other series of single-patch repair.[19] Ginde et al report on a freedom from reoperation of 97% after 10 and 94% after 20
years, however, with necessity of a re-reoperation in about one-third of these patients.[8] We observed a comparable freedom from reoperation, with no need for re-repair after
up to 30 years. However, the complexity of the SAS will most likely influence the
recurrence of the obstruction.
Conclusion
Up to 40 years after single-patch repair for CAVSD in infancy or childhood, the functional
outcome is excellent and the need for L-AVV reoperation is very promising. Pacemaker-related
reoperations are common and related to surgical and late-onset etiology as well.
Limitations
This study is limited by its retrospective character and, due to the long timeframe,
by the inability to obtain a recent clinical follow-up examination from every patient.
In addition, the study is based on the clinical estimation, electrocardiographic and
in particular echocardiographic interpretation of many further care cardiologists,
rather than obtaining findings exclusively by the authors of the present study. Nevertheless,
this study is not aiming at meticulous analysis, which is, based on the methodology,
not possible, but it provides much useful information of long-term results of a surgical
technique applied uniformly in a large cohort.
