Surgical repair of Tetralogy of Fallot – Long-term outcomes into the second decade after transannular patch repair and annulus-sparing patch repair

Abstract

Objectives

Different techniques are used to perform repair of Tetralogy of Fallot (TOF), depending on the development of the pulmonary valve and the degree of the right ventricular outflow tract obstruction.

Methods

All patients who underwent TOF repair with either an annulus-sparing patch or a transannular patch (TAP) between January 1995 and April 2021 were included in this retrospective chart review. Mortality was cross-checked with the national health insurance. Survival and competing-risks analyses were conducted to report on long-term outcomes.

Results

One-hundred-fifty patients (50.7% male; median age at surgery 5.8 months; 28.7% staged repair with primary shunt palliation) underwent TOF repair with annulus-sparing patch (43.3%; 65/150) or TAP (56.7%; 85/150) respectively. Median follow-up was 5.4 years (IQR 0.1-16.4 years) with a maximum follow-up of 26.7 years. Six deaths (4 early deaths and 2 late deaths) occurred and Kaplan-Meier estimated survival was 96.3% at 5 years, 94.3% at 15 years and 20 years. The cumulative incidence of pulmonary valve replacement was significantly higher (p = 0.011) in the TAP group than in the annulus-sparing patch group with 58.9% (95% CI 37.8-75) and 40.2% (95% CI 21.3-58.5) at 20 years respectively. Also, the cumulative incidence of any Fallot-related reintervention (percutaneous and surgical) was significantly higher (p = 0.007) in the TAP group.

Conclusion

Early survival rates and long-term survival up to the second postoperative decade after TOF repair show good results. There was no difference in survival between patients with TAP and annulus-sparing patch. TAP repair was associated with an increased need for Fallot-related reintervention and for pulmonary valve replacement.

Graphical abstract

Highlights

• •Tetralogy of Fallot repairs from 1995 until 2021 were included.
• •Annulus-sparing repair and transannular patch repair were compared.
• •Survival up to the second decade after Tetralogy of Fallot repair is excellent.
• •Transannular patch repair was associated with higher reintervention rates.
• •Annulus-sparing repair has shown higher freedom from pulmonary valve replacement.

Abbreviations

CI –

confidence interval

HR –

hazard ratio

IQR –

interquartile range

RVOT –

right ventricular outflow tract

TAP –

transannular patch

TOF –

Tetralogy of Fallot

1. Introduction

Tetralogy of Fallot (TOF) is characterized by a ventricular septal defect, an obstruction in the right ventricular outflow tract (RVOT), right ventricular hypertrophy and an aorta overriding the ventricular septal defect. TOF is the most common cyanotic congenital heart defect and the presentation can vary from less symptomatic forms to severe cyanosis [1]. Depending on the degree and level of RVOT obstruction, the size of the pulmonary valve annulus and the presence of pulmonary artery stenosis surgical repair is performed using different techniques [2,3]. The study included patients with annulus-sparing patch and transannular patch (TAP) repair and analysis of long-term outcomes was performed to assess survival and freedom from reintervention in the second decade after TOF repair.

An interdisciplinary approach that begins at a pediatric heart center and continues through the crucial transition to a dedicated adult congenital heart disease clinic is necessary to sustain follow-up at a specialized center over a lifetime in this complex patient cohort.

2. Patients and methods

2.1. Patients

A retrospective chart review was conducted to include patients undergoing TOF repair with either an annulus-sparing patch or a TAP from January 1st^, 1995 until April 30th^, 2021. A positive vote from the local ethics committee board was obtained (Ethics committee submission number: 1545/2021) and patient consent was waived due to the retrospective study design. Patients, who required pulmonary valve replacement and/or additional corrective surgery in the setting of TOF diagnosis in complex congenital heart disease were excluded. Also, patients, who underwent TOF repair without the use of a patch were excluded. Follow-up data after TOF repair was included until December 31st^, 2021. Median follow-up was 5.4 years (IQR 0.1-16.4 years) with a maximum follow-up of 26.7 years. Mortality was cross-checked with the national health insurance. The cross-check was not feasible in 32 patients (21.3%; 32/150), who had been transferred from foreign centers for TOF repair with the majority being treated as part of a humanitarian assistance program. These patients were censored for analysis at their last available follow-up at the study center and have a shorter follow-up time (p < 0.001) with a median follow-up of 0.06 years (IQR 0.03-0.08 years; range 0.02-5.9 years) compared to the remaining cohort with a median follow-up of 8.2 years (IQR 3.5-17.7 years). The follow-up time between the annulus-sparing patch repair group and the TAP repair group did not differ (p = 0.193) with a median follow-up of 7.3 years (IQR 0.4-17.8 years) and 4.7 years (IQR 0.1-12.3 years) respectively.

2.2. Surgical technique

Depending on the development of the pulmonary arteries and the grade of cyanosis a staged repair with primary shunt palliation (Blalock-Taussig Shunt and/or aortopulmonary shunt) was performed more commonly in the earlier study period. In the later study period primary TOF repair is performed regularly and the preferred approach at our center. All included patients underwent TOF repair with a transatrial-transpulmonary approach. If the subvalvular stenosis cannot be reduced by an extended transatrial myectomy and separation of the infundibular muscle ring, TOF repair is performed with the use of a single patch widening the RVOT. If the pulmonary valve annulus is sufficiently developed (>−2 standard deviations) necessary widening of the RVOT is performed with an annulus-sparing patch technique. In case of an annulus smaller than −2 standard deviations, a transannular incision is necessary, followed by an insertion of a TAP. Additionally, the valvular stenosis is relieved by performing a commissurotomy and/or leaflet shaving to allow for better leaflet mobility and valve opening. The TAP should not be fashioned too large in order to achieve a suitable diameter of the pulmonary valve annulus and to avoid obstruction of the coronary arteries. In the case of a supravalvular stenosis, an additional patch or in the setting of TAP an extension of the patch into the main pulmonary trunk above the stenotic area is performed. Care must be taken to remove the stenosis completely in order to avoid possible restenosis or vessel kinking.

2.3. Definitions

Death within the first 30 days postoperatively or during the postoperative index-hospitalization was defined as early death. Classification of coronary anatomy was performed according to the modified Leiden Convention [4]. Outcomes (survival, pulmonary valve replacement, any Fallot-related reintervention including surgical and percutaneous interventions) were compared between patients, who received an annulus-sparing patch and patients, who received TAP. Additionally, the study endpoints were compared between surgical eras: 1995 to 2008 (44%; 66/150) and 2009 to 2021 (56%; 84/150).

2.4. Statistical analysis

Continuous variables were compared between groups using the Mann-Whitney U test and categorial variables were compared with Chi-square test or Fisher's exact test, in case of expected cell frequencies less than 5. Survival probabilities were determined by Kaplan-Meier estimation. To report on potential associations of independent variables with survival univariable Cox proportional hazards regression models were calculated. Competing risk analysis was performed calculating cumulative incidence functions to report on the probability of pulmonary valve replacement or any Fallot-related reintervention with death considered as a competing event. Differences between cumulative incidence functions were examined with the Gray's test. Univariable and multivariable Fine and Gray proportional subdistribution hazards models [5] were performed to evaluate potential associations of independent variables with reinterventions (pulmonary valve replacement, and any Fallot-related reintervention, respectively). Death was considered as a competing event within the Fine and Gray models. Variables with a p-value <0.05 or considered clinically relevant were included in univariable and multivariable Cox proportional hazards models or, where appropriate, univariable and multivariable Fine–Gray proportional subdistribution hazards models. The number of events prevented further adjustment for co-founders. Statistical significance was set at p < 0.05. Data was analyzed using the software package SPSS® 29 (IBM Corp., Chicago, Illinois, USA) and SAS 9.4 (SAS Institute Inc., Cary, NC, USA).

3. Results

3.1. Patient and perioperative characteristics

From January 1995 to April 2021, 150 patients (50.7%; 76/150 male) underwent TOF repair at a median age at time of surgery of 5.8 months (IQR 3.2-18.3 months). Repair was performed with annulus-sparing patch in 65 patients (43.3%; 65/150) and with TAP in 85 patients (56.7%; 85/150). Patient and perioperative characteristics regarding the two groups are summarized in Table 1. Patients undergoing TAP repair exhibited a significantly higher prevalence of hypoplastic pulmonary arteries (p = 0.001) and valvular stenosis requiring balloon pulmonary valvuloplasty (p = 0.008) than patients undergoing annulus-sparing patch repair. In addition, preoperative cardiac decompensation occurred more frequently in the TAP cohort (p = 0.012). In the overall TOF repair cohort, 19.3% of patients (19.3.%; 29/150) had undergone staged repair with primary shunt palliation (Blalock-Taussig Shunt and/or aortopulmonary shunt), which was performed more frequently (p < 0.001) in the earlier surgical era from 1995 to 2008 (34.8%; 23/66) than in the surgical era from 2009 to 2021 (7.1%; 6/84).

Table 1.

Patient and perioperative characteristics.

Variable	Annulus-sparing patch (n = 65)	Transannular patch (n = 85)	p-value
Surgical era
1995-2008	34 (52.3)	32 (37.6)	0.073
2009-2021	31 (47.7)	53 (62.4)
Male	37 (56.9)	39 (45.9)	0.180
Age at surgery (months)	7.2 (3.5-19.4)	5.1 (2.5-16.8)	0.157
Cyanosis	27 (41.5)	45 (52.9)	0.166
Hypoxic crisis	21 (32.3)	35 (41.2)	0.266
Major aortopulmonary collateral arteries	3 (4.6)	11 (12.9)	0.082
Atrial septal defect	44 (67.7)	64 (75.3)	0.304
Coronary anomaly	7 (10.8)	10 (11.8)	0.849
Hypoplastic pulmonary arteries	16 (24.6)	43 (50.6)	0.001
Weight at surgery (kg)	6.7 (5.4-9.5)	6.3 (4.5-8.4)	0.152
Height at surgery (cm)	66 (60-74)	62 (54-77)	0.059
BSAHaycock at surgery	0.35 (0.3-0.42)	0.34 (0.3-0.42)	0.174
Previous cardiac intervention (percutaneous and surgical)	14 (21.5)	29 (34.1)	0.091
Staged repair with primary shunt palliation	12 (18.5)	17 (20.0)	0.813
Balloon pulmonary valvuloplasty	2 (3.1)	14 (16.5)	0.008
Preoperative cardiac decompensation	8 (12.3)	25 (29.4)	0.012
Cardiopulmonary bypass time	129 (105-154)	133 (110-175)	0.223
Aortic cross clamp time	80 (63-98)	82 (65-113)	0.283
Concomitant LPA/RPA patch	14 (21.5)	27 (31.8)	0.164

Values are presented as n, n (%), median (interquartile range). Continuous variables were compared using the independent-samples Mann-Whitney U test and categorial variables were compared with Chi-square test or Fisher's exact test, in case of expected cell frequencies less than 5. BSA, body surface area; LPA, left pulmonary artery; RPA, right pulmonary artery.

3.2. Early outcome

Early postoperative outcomes are seen in Table 2. Overall early mortality was 2.7% (4/150) and did not differ (p = 0.633) between the two groups, but differed (p = 0.036) between surgical eras 1995 to 2008 (6.1%; 4/66) and 2009 to 2021 (0%; 0/84). Causes of deaths are given in Table 3. The rate of postoperative pacemaker implantation was significantly higher (p = 0.036) in the TAP group with a rate of 7.1% (6/85; indication was complete atrioventricular block in all cases) compared to no pacemaker implantation in the annulus-sparing patch group. Intensive care unit stay (p = 0.002) and overall hospital stay (p < 0.001) were longer in the TAP cohort (Table 2). Though the need for delayed sternal closure (p = 0.090) and requirement of inhalative nitric oxide administration (p = 0.560) did not differ in the postoperative phase. Also, junctional ectopic tachycardia occurred without a difference (p = 0.658) in both groups with 13.8% of patients in the anulus-sparing patch group and 16.5% of patients in the TAP group presenting with junctional ectopic tachycardia in the early postoperative phase.

Table 2.

Early outcomes.

Variable	Annulus-sparing patch (n = 65)	Transannular patch (n = 85)	p-value
Readmission to intensive care unit	6 (9.2)	13 (15.3)	0.269
Re-intubation	6 (9.2)	15 (17.6)	0.141
Ventilation time (days)a	2 (1-6)	3 (2-7)	0.138
Intensive care unit stay (days)a	4 (3-8)	7 (4-13)	0.002
Hospital stay (days)a	16 (11-23)	23 (14-32)	<0.001
Pacemaker implantation	0 (0)	6 (7.1)	0.036
Revision for bleeding	4 (6.2)	0 (0)	0.033
Subxiphoidal pericardial effusion drainage	0 (0)	5 (5.9)	0.069
Delayed sternal closure	7 (10.8)	18 (21.2)	0.090
Junctional ectopic tachycardia	9 (13.8)	14 (16.5)	0.658
Nitric oxide inhalation	12 (18.5)	19 (22.4)	0.560
Dialysis/hemofiltration	10 (15.4)	12 (14.1)	0.828
Extracorporeal membrane oxygenation	1 (1.5)	3 (3.5)	0.633
Early death	1 (1.5)	3 (3.5)	0.633

Values are presented as n, n (%), median (interquartile range). Continuous variables were compared using the independent-samples Mann-Whitney U test and categorial variables were compared with Chi-square test or Fisher's exact test, in case of expected cell frequencies less than 5.

^aTimes of patients, who died during intensive care unit/hospital stay were set to a maximum of 100 days.

Table 3.

Early and late deaths.

No. (sex)	Surgery (year)	Age at time of surgery	Diagnoses	Previous interventions	Death (postoperative days/years)	ECMO	Cause of death
Early deaths
1 (f)	TAP (1995)	17 months	TOF, hypoplastic pulmonary arteries, major aortopulmonary collaterals, coronary anomaly	Aortopulmonary shunt	12 days	yes	Sepsis and disseminated intravascular coagulopathy with inner bleeding and necrosis
2 (f)	Annulus-sparing patch repair (2000)	13 months	TOF	Modified Blalock-Taussig Shunt	1 day	no	Pulmonary hemorrhage and cardiorespiratory failure in dilated cardiomyopathy and myocardial ischemia
3 (m)	TAP (2001)	5 months	TOF, situs inversus totalis, major aortopulmonary collaterals, coronary anomaly	Modified Blalock-Taussig Shunt	1 month	yes	Sepsis and cardiorespiratory failure
4 (f)	TAP (2001)	4 months	TOF with pulmonary atresia, coronary anomaly, duodenal/rectal atresia with colovesical fistula	Modified Blalock-Taussig Shunt	16 days	yes	Sepsis and cardiorespiratory failure
Late deaths
5 (f)	TAP (2006)	16 years	TOF, partial agenesis of the Corpus Callosum		13 years	no	Right heart failure
6 (m)	TAP (2018)	6 months	TOF, hypoplastic pulmonary arteries, suspected mitochondrial disease, muscle hypotonia		6 months	no	Unknown aetiology

ECMO, extracorporeal membrane oxygenation; TAP, transannular patch; TOF, Tetralogy of Fallot.

3.3. Long-term follow-up

3.3.1. Survival

Six deaths (4 early deaths and 2 late deaths) occurred and Kaplan-Meier estimated survival was 96.3% (95% CI 91.3-98.5) at 5 years, 94.3% (95% CI 86.4-97.6) at 15 years and 20 years (Fig. 1A). Kaplan-Meier estimated survival did not differ (p = 0.168) between the annulus-sparing patch and the TAP group (Fig. 1B). The two late deaths were observed in patients who had undergone TAP repair. One patient, who had undergone TAP repair at the age of 16 years for previously unoperated TOF, died 13 years after surgical repair due to progressive right heart failure. The second patient, who had undergone TAP repair at six months of age died six months postoperatively from an unknown cause in the context of a suspected mitochondrial disease. Detailed causes of late deaths are seen Table 3. At univariable Cox proportional hazards analysis (Table 4) coronary anomaly (HR 8.9, 95% CI 1.8-44.3; p = 0.008) and a staged repair with primary shunt palliation (HR 7.1, 95% CI 1.3-39.5; p = 0.025) were associated with mortality.

Fig. 1.

Survival after repair of Tetralogy of Fallot. 1A) Kaplan-Meier estimated survival curve with 95% confidence interval (CI). 1B) Kaplan-Meier estimated survival curves with 95% CI comparing annulus-sparing patch and transannular patch.

Table 4.

Associations of independent variables with mortality (Cox proportional hazards regression analysis).

Variable	Univariable analysis
	Hazard ratio	Confidence interval (95%)	p-value
Agea	1.3	0.9-2	0.145
Surgical repair (TAP vs annulus-sparing patch)	4.1	0.5-35	0.203
Surgical era (2009-2020 vs 1995-2008)	0.2	0.02-1.7	0.140
Hypoxic crisis	0.7	0.1-4.1	0.731
Major aortopulmonary collateral arteries	5.7	1-32.3	0.050
Coronary anomaly	8.9	1.8-44.3	0.008
Hypoplastic pulmonary arteries	0.7	0.1-3.9	0.697
Previous cardiac intervention (surgical and percutaneous)	4.2	0.8-23.4	0.096
Staged repair with primary shunt palliation	7.1	1.3-39.5	0.025

TAP, transannular patch.

^alog2-transformed.

3.3.2. Pulmonary valve replacement

Thirty-three patients (22%; 33/150) underwent at least one pulmonary valve replacement with pulmonary valve regurgitation being the predominant indication for surgery (regurgitation: 78.8%; 26/33; stenosis: 9.1%; 3/33; combined valve disease 12.1%; 4/33). Patients in the TAP group were younger (p = 0.016) at time of surgical pulmonary valve replacement with a mean age of 14.2 ± 7.1 years in the annulus-sparing patch cohort and a mean age of 7.7 ± 5.1 years in the TAP cohort. Pulmonary valve replacement was performed with a bioprosthetic valve in 17 patients (51.5%; 17/33) and with a homograft in 16 patients (48.5%; 16/33). The use of a bioprosthesis valve replacement or a homograft did not differ (p = 0.805) between the two TOF repair cohorts. The cumulative incidence of pulmonary valve replacement was 9.5% (95% CI 4.8-16.2) at 5 years, 43.5% (95% CI 30.6-55.6) at 15 years and 50.6% (95% CI 36.5-63.1) at 20 years (Fig. 2A). The cumulative incidence of pulmonary valve replacement was significantly higher (Gray test: p = 0.011) in the TAP group than in the annulus-sparing patch group with 58.9% (95% CI 37.8-75) and 40.2% (95% CI 21.3-58.5) at 20 years respectively (Fig. 2B). The Fine and Gray regression analysis for independent variables associated with pulmonary valve replacement are seen in Table 5. Four patients (12.1%; 4/33) underwent a second pulmonary valve replacement after a mean time of 6.2 ± 5.7 years.

Fig. 2.

Cumulative incidence curves of pulmonary valve replacement after repair of Tetralogy of Fallot. 2A) Cumulative incidence curve with 95% confidence interval (CI) of pulmonary valve replacement after repair of Tetralogy of Fallot. 2B) Cumulative incidence curves with 95% CI of pulmonary valve replacement after repair of Tetralogy of Fallot comparing annulus-sparing patch and transannular patch.

Table 5.

Fine and Gray subdistribution hazards regression analyses for independent variables associated with pulmonary valve replacement.

Variable	Univariable models			Multivariable model
	Hazard ratio	Confidence interval (95%)	p-value	Hazard ratio	Confidence interval (95%)	p–value
Agea	1	0.9-1.2	0.632	-	-	-
Surgical repair (TAP vs annulus-sparing patch)	2.5	1.3-4.9	0.009	2.0	1.0-4.1	0.059
Surgical era (2009-2020 vs 1995-2008)	2.7	1.2-5.9	0.017	2.4	1.0-5.7	0.049
Previous balloon pulmonary valvuloplasty	2.3	0.9-5.5	0.067	-	-	-
Major aortopulmonary collateral arteries	1.1	0.2-5.1	0.942	-	-	-
Coronary anomaly	0.4	0.1-1.9	0.271	-	-	-
Hypoplastic pulmonary arteries	2.2	1.1-4.4	0.022	1.9	0.9-3.9	0.091
Previous cardiac intervention (surgical and percutaneous)	1.8	0.9-3.4	0.101	-	-	-
Staged repair with primary shunt palliation	1.4	0.7-2.7	0.362	-	-	-
Concomitant LPA/RPA patch	1.6	0.8-3.2	0.158	-	-	-

LPA, left pulmonary artery; RPA, right pulmonary artery; TAP, transannular patch.

^alog2-transformed.

3.3.3. Any Fallot-related reintervention

Forty-two patients (28%; 42/150) underwent at least one Fallot-related reintervention (surgical and percutaneous). The cumulative incidence of any Fallot-related reintervention was 21.4% (95% CI 14.1-29.6) at 5 years, 48.5% (95% CI 35.8-60.1) at 15 years and 53.4% (95% CI 39.7-65.2) at 20 years (Fig. 3A). The cumulative incidence of any Fallot-related reintervention was significantly higher (Gray test: p = 0.007) in the TAP group than in the annulus-sparing patch group with 63.7% (95% CI 43.5-78.4) and 40.8% (95% CI 22.2-58.7) at 20 years respectively (Fig. 3B). In the multivariable Fine and Gray regression analysis (Table 6) only the surgical repair type (TAP vs annulus-sparing patch) remained significantly associated (p = 0.017) with any Fallot-related reintervention.

Fig. 3.

Cumulative incidence curves of any Fallot-related reintervention after repair of Tetralogy of Fallot. 3A) Cumulative incidence curve with 95% confidence interval (CI) of any Fallot-related reintervention (percutaneous and surgical) after repair of Tetralogy of Fallot. 3B) Cumulative incidence curves with 95% CI of any Fallot-related reintervention (percutaneous and surgical) after repair of Tetralogy of Fallot comparing annulus-sparing patch and transannular patch.

Table 6.

Fine and Gray subdistribution hazards regression analyses for independent variables associated with any Fallot related reintervention (surgical and percutaneous).

Variable	Univariable models			Multivariable model
	Hazard ratio	Confidence interval (95%)	p-value	Hazard ratio	Confidence interval (95%)	p–value
Agea	1.1	1-1.3	0.121	-	-	-
Surgical repair (TAP vs annulus-sparing patch)	2.4	1.3-4.5	0.006	2.3	1.2-4.4	0.017
Surgical era (2009-2020 vs 1995-2008)	1.4	0.7-2.8	0.351	-	-	-
Previous balloon pulmonary valvuloplasty	1.4	0.6-3	0.401	-	-	-
Major aortopulmonary collateral arteries	2.1	0.7-6.8	0.209	-	-	-
Coronary anomaly	0.5	0.2-1.8	0.319	-	-	-
Hypoplastic pulmonary arteries	1.7	1.0-3.2	0.069	1.2	0.6-2.3	0.602
Previous cardiac intervention (surgical and percutaneous)	2.1	1.2-3.9	0.014	1.6	0.6-4.3	0.350
Staged repair with primary shunt palliation	1.9	1.0-3.7	0.050	1.3	0.4-3.7	0.652
Concomitant LPA/RPA patch	1.2	0.7-2.2	0.550	-	-	-

LPA, left pulmonary artery; RPA, right pulmonary artery; RVOT, right ventricular outflow tract; TAP, transannular patch.

^alog2-transformed.

Fourteen patients (9.3%; 14/150) underwent at least one percutaneous dilatation and/or stenting of the pulmonary arteries with a cumulative incidence of 11.1% (95% CI 6-18) at 5 years, 14.4% (95% CI 8-22.7) at 15 and 20 years. The cumulative incidence of dilatation and/or stenting of the pulmonary arteries was significantly higher (Gray test: p = 0.049) in the TAP group than in the annulus-sparing patch group with 19% (95% CI 9.7-30.6) and 8.2% (95% CI 1.9-20.4) at 20 years respectively.

3.4. Endocarditis and arrhythmias

No endocarditis occurred after TOF repair. Two patients experienced an endocarditis of their first implanted pulmonary valve replacement (a homograft 12.6 years after implantation and of a bovine jugular vein conduit 0.4 years after implantation respectively) and underwent re-replacement. Two patients underwent late pacemaker implantation due to atrioventricular block 56 days after TAP and due to atrioventricular block after electrical cardioversion for atrial flutter 10.3 years after annulus-sparing patch.

4. Discussion

Perioperative mortality is low and late survival shows good results with Kaplan-Meier estimated survival of 94.3% at 20 years. Patients requiring TAP repair present with a more severe anatomical substrate and correspondingly with a more pronounced clinical manifestation. The surgical strategy deciding between TAP repair or the feasibility of an annulus-sparing patch repair was determined by the size of the pulmonary valve annulus and the degree of RVOT obstruction. When the pulmonary valve annulus was sufficiently developed (Z-score > −2 standard deviations), RVOT enlargement can be achieved with an annulus-sparing patch technique. In cases with a pulmonary valve annulus smaller than Z-score −2 standard deviations, a TAP repair was performed. Staged repair was rarely necessary in the later study period (2009-2021). Pulmonary valve regurgitation after TOF repair is the predominant factor for pulmonary valve replacement, which is more often required after TAP than after annulus-sparing patch.

Early mortality was low with 2.7% and was comparable to other recent retrospective chart review studies with early mortality ranging from 1% to 7.3% [3,[6], [7], [8], [9], [10]]. In our analysis early mortality did not differ between the TAP and annulus-sparing patch group. In this cohort, which included patients since 1995, we have observed an early mortality difference between surgical eras 1995 to 2008 (6.1%) and 2009 to 2021 (0%). Ylitalo et al. [3], who included patients undergoing TOF repair from 1962 to 2007 have seen an early mortality of 7% with an early mortality rate of 1.5% for the last two decades and no early deaths were observed after the year 2000. Saygi et al. [8], who included patients undergoing TOF repair from 2010 to 2013 found low preoperative oxygen saturation, high right ventricular/aortic pressure ratio immediately after surgery, presence of coronary anomaly, requirement of postoperative ECMO and pacemaker implantation as parameters associated with increased perioperative mortality. We found major aortopulmonary collateral arteries, coronary anomalies and staged repair with primary shunt palliation as factors associated with mortality. The frequency of staged repair with primary shunt palliation did not differ between the TAP and annulus-sparing patch group, but did differ between the surgical eras 1995 to 2008 and 2009 to 2021. Primary TOF repair is the preferred approach at our center. In the cohort of Ylitalo et al. [3] primary palliation was performed in 25% of the patients, which showed a significantly inferior late survival compared to patients with primary TOF repair (40-year survival rate 75% vs. 84%). Additionally, TAP was more often performed as corrective surgery in patients with primary palliation compared to primary TOF repair (41% vs 29%). Smith et al. [7] also found an increased risk of early mortality with staged repair with shunt as well as with non–valve-sparing operation.

Longstanding pulmonary valve regurgitation or RVOT stenosis pose detrimental long-term consequences after TOF repair [11,12]. Chronic pulmonary valve regurgitation resulting in progressive right ventricular dysfunction due to chronic volume overload, increases the risk of arrhythmias posing a risk for sudden cardiac death and results in a reduced exercise capacity [11,[13], [14], [15]]. Right ventricular dysfunction occurs in the setting of long-term pulmonary valve regurgitation and/or stenosis and requires pulmonary valve replacement and/or relief of RVOT obstruction at other levels before right ventricular dysfunction becomes irreversible [16]. With 78.8% pulmonary valve regurgitation was the predominant indication for pulmonary valve replacement in our cohort. The cumulative incidence of valve replacement was 50.6% at 20 years. In concordance with several studies focusing on outcomes after TOF repair [3,6,7,[17], [18], [19]], we have found that the use of TAP is associated with higher reintervention rates. Luijten et al. [6] report a higher event-free survival (combined endpoint of pulmonary valve replacement, reoperation, balloon dilatation of the pulmonary valve, and pacemaker/internal cardiac defibrillator implantation) for patients with no transannular patch than for patients with TAP with 78.5% and 27.9% at 25 years. We have seen a significantly higher cumulative incidence of any Fallot-related reintervention (surgical and percutaneous) in patients with TAP than in patients who underwent annulus-sparing patch repair with 58.9% and 40.2% at 20 years respectively. Additionally, the surgical correction type (TAP vs annulus-sparing patch repair) has shown to be an associated risk factor in the multivariable analysis. Additionally, patients in the TAP group were younger at time of surgical pulmonary valve replacement, but the use of a bioprosthetic valve replacement or a homograft valve replacement did not differ between the two TOF repair cohorts.

TOF repair evolved over the decades with an increasing experience and follow-up of repair strategies, which has shown that TAP repair is prone for pulmonary regurgitation leading to right ventricular dilatation and the need of repeated reinterventions. Lastly the delamination technique by Vida et al. [[20], [21], [22]] became part of the armamentarium of TOF repair and has also been recently adopted at our center whenever technically feasible. Valvular tethering is defused and delamination plasty is performed with an ophthalmic scalpel and fine scissors at the base of the cusps to the hinge point allowing for more leaflet tissue to be gained for coadaptation by opening the area directly underneath the cusp for lengthening. Creation of commissures is reached by resuspension of the extended cusps. Analysis of long-term outcomes must be awaited, but short-term studies show promising results [[23], [24], [25]].

Also, a lowering of the reoperation burden over lifetime might be achieved by postponing surgical re-replacement by percutaneous valve replacement. Evaluation of surgical techniques regarding enablement of an optimal landing zone in native and patch-augmented annuli and a wider selection of available transcatheter pulmonary valves (self-expanding percutaneous pulmonary valves, RVOT pre-stents, balloon-expandable pulmonary valves) has been seen in the last years.

4.1. Study limitations

The use of the compared different surgical techniques with either anulus-sparing patch or TAP is decided by the individual congenital anatomy of the patient. It should also be acknowledged that the more sever anatomical substrate necessitating TAP is inherently associated with differences in clinical manifestation and therefore periprocedural presentation. Therefore, it has to be taken into account that the statistical comparison is performed to allow for identification of clinically relevant differences in the long-term care of these patients to guide management. In the setting of a retrospective analysis the evolution of perioperative and surgical practice may not be completely accounted for, though the analysis of outcomes was additionally stratified for surgical eras. Follow-up time of the cohort (960 patient-years) is long and mortality cross-check was performed, though it is necessary to state that in this cohort 21.3% (32/150) of patients could not be cross-checked as the majority of these had been transferred for surgery as part of a humanitarian assistance program. Their censoring at the last clinical follow-up at the center has to be taken into account regarding the outcomes of the long-term survival analysis.

5. Conclusion

Early mortality was low and long-term survival up to the second postoperative decade after TOF repair was good. There was no difference in survival between patients with TAP repair and annulus-sparing patch repair. TOF repair with TAP was associated with an increased risk for pulmonary valve replacement and Fallot-related reintervention compared to TOF repair with annulus-sparing patch.

CRediT authorship contribution statement

Johanna Schlein: Writing – original draft, Visualization, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Martina Tava: Writing – review & editing, Investigation, Data curation, Conceptualization. Felix Wollmann: Writing – review & editing, Methodology, Investigation, Data curation, Conceptualization. Alexandra Kaider: Writing – review & editing, Visualization, Validation, Software, Methodology, Formal analysis. Barbara Karner: Writing – review & editing. Clemens Atteneder: Writing – review & editing. Paul Werner: Writing – review & editing. Thomas Wasserscheid: Writing – review & editing. Bea Goessinger: Writing – review & editing. Erhan Urganci: Writing – review & editing. Suriya Prausmüller: Writing – review & editing. Christiane Pees: Writing – review & editing, Validation, Supervision, Conceptualization. Ina Michel-Behnke: Writing – review & editing, Validation, Supervision. Eva Base: Writing – review & editing, Validation, Supervision. Lore Schrutka: Writing – review & editing, Validation, Supervision, Resources, Conceptualization. Daniel Zimpfer: Writing – review & editing, Validation, Supervision, Resources, Conceptualization. Peter Murin: Writing – review & editing, Validation, Supervision, Conceptualization.

Informed consent statement

Patient consent was waived in the retrospective study setting.

Institutional review board approval

Ethics committee submission number: 1545/2021 with approval on June 22nd, 2021.

Funding

This research received no external funding.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.