Optimizing long-term outcomes in pediatric mitral valve repair: A focus on posterior leaflet augmentation

Abstract

Objective

To evaluate mid- and long-term outcomes of posterior leaflet augmentation with fresh autologous pericardium for congenital or acquired mitral regurgitation in pediatric populations, and to identify predictors of valve-related reintervention.

Methods

A retrospective review was conducted on all patients who underwent mitral valve repair, including posterior leaflet augmentation, at Bambino Gesù Children's Hospital between 2011 and 2024. Patients with atrioventricular septal defects or single-ventricle physiology were excluded. Demographic, operative, and echocardiographic data were analyzed. Freedom from reintervention was estimated using Kaplan-Meier analysis and risk factors were identified by Cox proportional hazards modeling.

Results

Twenty-eight patients (median age, 3.8 years; interquartile range, 0.6-9.2 years; 57% men) underwent posterior leaflet augmentation using fresh autologous pericardium. Congenital mitral dysplasia was the most common etiology (50%), followed by rheumatic disease (29%). Mean preoperative mitral regurgitation grade was 3.9 ± 0.6, decreasing to 1.4 ± 0.5 postoperatively, and remained stable during follow-up (mean residual mitral regurgitation, 1.7 ± 0.7 mm Hg). Mean transmitral gradient increased from 2.5 ± 1.8 mm Hg to 4.2 ± 2.1 mm Hg, remaining stable over time (4.9 ± 2.2 mm Hg at last follow-up). During a median follow-up of 7.7 years (interquartile range, 6.6-10.5 years), 9 patients (32%) required reintervention, all consisting of mitral valve replacement. Freedom from reintervention was 84.9% at 1 year, 75.9% at 5 years, and 54.2% at 10 years. A postoperative mean gradient >5 mm Hg was the only independent predictor of reintervention (hazard ratio, 7.81; 95% CI, 1.9-31.4; P = .004).

Conclusions

Posterior leaflet augmentation with fresh autologous pericardium provides a durable, growth-preserving, and physiologic repair for pediatric mitral valve disease predominantly presenting with severe regurgitation. A postoperative transmitral gradient >5 mm Hg identifies patients at higher risk for late mitral valve replacement, emphasizing the importance of accurate intraoperative tailoring and early echocardiographic follow-up.

Graphical Abstract

Freedom from mitral valve reintervention after posterior leaflet augmentation.

Central Message.

Posterior leaflet augmentation ensures long-term valve growth and function in pediatric mitral repair; an intraoperative gradient >5 mm Hg predicts reintervention.

Perspective.

Posterior leaflet augmentation with fresh autologous pericardium offers a growth-preserving and durable repair for pediatric mitral valve disease predominantly presenting with severe regurgitation. Future investigations should focus on optimizing patch geometry and confirming its role as a bridge to definitive valve replacement in growing patients.

Congenital mitral valve disease (MVD) encompasses a wide spectrum of structural abnormalities that severely influence valve mobility and competence.¹ Among these, dysplastic forms characterized by restricted leaflet motion are particularly frequent in pediatric populations.^2,3 In this setting, the posterior leaflet is often hypoplastic, thickened, or adherent to the posterior wall of the left ventricle, resulting in limited excursion and reduced coaptation surface.

Such anatomical conditions make conventional repair strategies, derived from adult experience, unsuitable for pediatric use.3, 4, 5, 6 Posterior leaflet augmentation (PLA) using fresh autologous pericardium offers an alternative, physiologic approach. By enlarging the posterior leaflet, this technique improves the coaptation area and restores mobility while maintaining the potential for valve growth and remodeling during follow-up.7, 8, 9 However, long-term data on outcomes and predictors of reintervention after PLA in congenital mitral dysplasia remain limited. The present study aims to evaluate mid- and long-term results of PLA in children and adolescents with congenital MVD and to identify risk factors for valve-related adverse events, focusing on postoperative residual stenosis.

Materials and Methods

Ethical Statement

The study conforms to the ethical principles of the Declaration of Helsinki and is in compliance with the current Italian regulations. The institutional review board was notified February 5, 2024. Individual consent waived due to retrospective design and anonymity.

Study Population and Follow-up

We retrospectively reviewed all consecutive patients who underwent MV repair with PLA at Bambino Gesù Children's Hospital between February 2011 and December 2024. All procedures were primary MV repairs; no patient had previously undergone MV repair. Patients with complete or partial atrioventricular septal defects and single-ventricle physiology were excluded from analysis. A total of 28 patients were identified. Clinical, operative, and echocardiographic data were obtained from institutional databases. Echocardiographic follow-up was performed before discharge, at 6, 12, and 24 months, and annually thereafter. End points of the study were survival after MV repair and cumulative incidence of reoperation.

Echocardiographic Evaluation

Transthoracic and intraoperative transesophageal echocardiography were used to assess the integrity of MV repair. Quantitative assessment of MV stenosis was estimated based on the mean continuous-wave Doppler gradient across the valve and stratified in the following manner: grade I with mean gradient <5 mm Hg, grade II with mean gradient from 5 to 10 mm Hg, and grade III with a mean gradient of 10 mm Hg or more. The MV regurgitation was semiquantitatively assessed based on the jet regurgitation and was classified as: grade I = mild, grade II = mild-moderate, grade III = moderate, and grade IV = severe.

Timing of Surgery

Indications for surgery varied according to the etiology, anatomy, patient age, size of the MV annulus, need for intravenous drugs and the clinical status.^5,6 Neonates and infants with severe mitral regurgitation underwent surgery only if they were symptomatic despite maximal medical therapy. Patients with isolated mitral cleft underwent operation based on the grade of regurgitation, regardless of symptoms. Surgical indications for reoperation were dictated by symptoms, the amount of medications required, or the need for mechanical ventilation. No specific thresholds for either pulmonary arterial pressure or transmitral gradient triggered a surgical indication, if few or no symptoms were present. According to our institutional policy, posterior leaflet augmentation with fresh autologous pericardium is reserved for patients in whom restricted motion of the posterior leaflet is the main mechanism of mitral regurgitation. Typical substrates include congenital MV dysplasia with short, thick, and partially muscularized chordae predominantly involving the posterior leaflet and its subvalvular apparatus, selected postischemic lesions with posterior papillary muscle involvement, and small children with severe mitral regurgitation in whom increasing the global coaptation surface of the valve is preferred to isolated annuloplasty.

Surgical Technique

After median sternotomy, cardiopulmonary bypass was instituted with aortic and bicaval cannulation under moderate hypothermia (28 °C). Myocardial protection was achieved using intermittent antegrade cold blood cardioplegia. Access to the MV was gained through a left atriotomy in the interatrial Waterston's groove in children and adolescents, or by a transseptal approach extending into the dome of the left atrium in neonates and infants.¹⁰ Beginning in 2011, PLA with fresh autologous pericardium was introduced as the preferred technique to restore leaflet mobility and improve coaptation in cases of posterior leaflet restriction.7, 8, 9 A rectangular or oval patch of untreated autologous pericardium was harvested and carefully trimmed of all adipose and extrapleural tissue. The patch was tailored to extend from 1 commissure to the other, with the atrial surface facing the left atrium. To reproduce the native curvature of the posterior leaflet and prevent excessive tissue redundancy, the free-edge side of the patch was made slightly longer than the annular side, thereby avoiding the aortic cusp effect.⁸

In neonates and infants, the height of the patch (H-patch) was determined according to both preoperative echocardiographic measurements and intraoperative inspection, using the following formula:

$$\mathrm{H}-\text{patch}=(\text{APL}-\text{PHL})+(\text{CL}-\text{target}–\text{CL}-\text{current})$$

where APL = anterior leaflet free-edge length (measured in mid-systole), PHL = posterior leaflet height, CL-target = desired coaptation length (7 mm), and CL-current = measured coaptation length at baseline.

This quantitative method ensured adequate augmentation to achieve physiological coaptation without introducing excessive tissue or iatrogenic stenosis. In older children and adolescents, the same principle was applied qualitatively, using direct intraoperative measurements to achieve a smooth coaptation line with a target zone of 6 to 8 mm.

Illustrative Case (Patient 11)

A 6-month-old infant (weight, 4.8 kg) with severe congenital MV dysplasia and restricted posterior leaflet motion was referred for surgery due to persistent heart failure and inability to wean from mechanical ventilation. The mitral annulus measured 13 mm. To determine the appropriate patch size, the maximal height of the anterior leaflet (15 mm) and the posterior leaflet (11 mm) were measured echocardiographically in mid-systole at the expected coaptation point. Applying the formula above, we obtained a coaptation surface of at least 7 mm:¹¹

$\mathrm{H}-\text{patch}=(15\text{mm}–11\text{mm})+(7\text{mm}–3\text{mm})=4\text{mm}+4\text{mm}=8\text{mm}$

A pericardial patch with a maximal height of 8 mm (+2 mm for suture line) was fashioned and gently tapered toward both commissures (Figure 1). As shown in Figure 1, the pericardial patch assumes a sail-like configuration that creates a new, extended line of leaflet closure and coaptation. The functional coaptation surface is provided by the patch itself, whereas the native posterior leaflet free margin falls into the left ventricle, forming a smooth and continuous transition between the native and augmented tissues. Postoperative transesophageal echocardiography demonstrated satisfactory leaflet mobility and coaptation, with a mean transmitral gradient of 4 mm Hg. The pericardial patch was sutured to the annulus and native leaflet edge using a continuous 6-0 polypropylene suture reinforced by interrupted stitches at both commissures to prevent dehiscence.

Figure 1.

A, Diagram showing measurement of the anterior leaflet free-edge length (APL) (15 mm) and posterior leaflet height (PHL) (11 mm) to determine patch height (H-patch) using the formula H-patch = (APL − PHL) + (desired coaptation length − measured coaptation at baseline), yielding a final patch height of 8 mm (+2 mm for suture line). The pericardial patch is tailored and tapered toward both commissures. B, Postoperative transesophageal echocardiography image demonstrating the sail-like (height = 0.8 cm) configuration of the augmented posterior leaflet with good coaptation and a mean transmitral gradient of 4 mm Hg.

The result of the repair was evaluated by saline injection through the valve and confirmed by intraoperative transesophageal and/or epicardial echocardiography. Surgical revision was considered in case of a suboptimal immediate postoperative result—defined as mitral regurgitation grade ≥III and/or mean continuous wave Doppler gradient >10 mm Hg—although decision making was individualized to each case. Associated repair procedures are summarized in Table 1; artificial expanded polytetrafluoroethylene chordae were used exclusively in patients with associated anterior leaflet prolapse.

Table 1.

Baseline patient characteristics

Characteristic	Result
No of patients	28
Age (y)	3.5 (0-17)
Weight (kg)	16 (2.3-54)
BSA (m2)	0.74 (0.19-1.60)
Sex
Female	16
Male	12
ECC (min)	156 (94-320)
XCL (min)	102 (66-253)
Annulus (mm)	19 (16 ± 24)
Diagnosis
Congenital dysplasia	11 (39.2)
RAA	7 (25)
Endocarditis	4 (14.3)
Ischemia	3 (10.7)
Shone syndrome	3 (10.7)
Carpentier classification
I	5 (17.8)
II	4 (14.2)
III	19 (67.9)
Associated procedure
Artificial PTFE chordae	5
Triangular resection, anterior leaflet	2
Papillary splitting	1

Values are presented as median (range), n (%), or n, unless otherwise indicated. BSA, Body surface area; ECC, extracorporeal circulation time; XCL, crossclamp time; RAA, rheumatic disease; PFTE, polytetrafluoroethylene.

Follow-up

All patients underwent postoperative examinations once during the first month, every 6 months during the first year, and at least yearly thereafter, including a clinical assessment, an electrocardiogram, and a 2-dimensional echocardiogram with color Doppler study performed by the referring cardiologist. Clinical and echocardiographic mitral functional status and the number of reoperations at the most recent follow-up were obtained from local records. The end of the follow-up period was set at December 31, 2024. Median follow-up was 7.7 years (interquartile range [IQR], 6.6-10.5 years).

Statistical Analysis

Continuous variables are expressed as mean ± SD or median (IQR), as appropriate. Categorical variables are presented as frequencies and percentages. Freedom from reintervention was estimated using the Kaplan-Meier method with 95% CI, and curves (with their corresponding numbers at risk) were truncated when the number of patients at risk dropped below predefined thresholds, in accordance with recommendations. Differences between groups (postoperative gradient ≤5 vs >5 mm Hg) were evaluated using the log-rank test. A Cox proportional hazards model was used to identify independent predictors of reintervention, and results are reported as hazard ratios with 95% CI. All analyses were performed with SPSS Statistics version 21 (IBM Corp).

Results

Between 1995 and 2024, 112 pediatric MV repair procedures were performed at our institution; 28 (25%) of these patients underwent PLA with fresh autologous pericardium for congenital or acquired mitral regurgitation. The median age at surgery was 3.8 years (IQR, 0.6-9.2 years), and the median body surface area was 0.65 m² (IQR, 0.35-1.10 m²). Overall, 9 of 28 patients (32.1%) were infants younger than age 1 year, including 1 neonate who underwent repair at 14 days of life. Of the 28 patients, 20 (71.4%) had nonrheumatic lesions (predominantly congenital mitral dysplasia) and 8 (28.6%) had rheumatic MVD. During follow-up, 3 of 8 rheumatic patients (37.5%) and 6 of 20 nonrheumatic patients (30.0%) required reintervention; given the small numbers, no formal statistical comparison was performed. A Carpentier type III restriction was present in 71% of cases, whereas genetic syndromes were documented in 3 patients (11%). None of the patients in the present series underwent concomitant mitral annuloplasty. The general characteristics of the population are summarized in Table 1.

Survival and Reintervention Rates

All patients survived the operation. During the entire follow-up period, 1 patient (3.6%) died due to progressive ventricular dysfunction unrelated to valve function. There were no intraoperative failures requiring conversion to MV replacement. One patient experienced very early failure due to partial detachment of the pericardial patch on postoperative day 2 and underwent re-repair with resuturing of the dehiscent portion of the patch. A second cardiopulmonary bypass run during the index procedure was required in 1 patient to extend the posterior annular incision from commissure to commissure and further enlarge the patch. During follow-up, 9 patients (32.1%) required at least 1 MV reintervention for a total of 14 procedures, all consisting of MV replacement. The mechanism of failure was progression to significant mitral stenosis, in some cases associated with residual or recurrent regurgitation (ie, mixed steno-insufficiency). Isolated recurrent pure regurgitation was not observed. At reoperation, the pericardial patches appeared globally thickened but showed no aneurysmal transformation or bulging toward the left atrium. The estimated freedom from reintervention was 84.9% (95% CI, 68.3%-92.8%) at 1 year, 80.6% (95% CI, 63.1%-89.9%) at 3 years, 75.9% (95% CI, 56.8%-87.3%) at 5 years, and 54.2% (95% CI, 33.5%-70.8%) at 10 years. The cumulative hazard curve demonstrated that most events occurred within the first decade after repair, followed by a plateau thereafter.

Evolution of Valve Function Over Time

Mean mitral regurgitation grade decreased from 3.9 ± 0.6 preoperatively to 1.4 ± 0.5 immediately after repair and remained stable during follow-up. At discharge and at 6, 12, and 24 months, mitral regurgitation grade persisted around 1.4 to 1.6, confirming durable valve competence. At the last follow-up, residual mitral regurgitation averaged 1. 7 ± 0.7, with only minor late progression in a few patients. The mean transmitral gradient increased from 2.5 ± 2.7 mm Hg preoperatively to 4.2 ± 2.1 mm Hg immediately after repair, reflecting the intrinsic tendency of this technique to create a mild functional narrowing of the mitral orifice while restoring valve competence. Thereafter, gradients remained stable (≈4.1 mm Hg up to 12 months, 4.5 mm Hg at 24 months, and 4.9 mm Hg at the latest follow-up) indicating long-term hemodynamic stability without evidence of progressive obstruction.

Risk Factors for Reintervention

On univariate analysis, an immediate postoperative mean transmitral gradient >5 mm Hg emerged as the only factor significantly associated with the risk of reintervention (P = .004). Patients discharged with a gradient >5 mm Hg had a hazard ratio of 7.81 (95% CI, 1.9-31.4) for subsequent reoperation compared with those with gradients ≤5 mm Hg (Figure 2). No other variable, including age, annulus size indexed to body surface area, or postoperative mitral regurgitation grade, was significantly associated with valve-related reintervention. In multivariate Cox analysis, the postoperative gradient remained the sole independent predictor of late repair failure.

Figure 2.

Postoperative transmitral gradient by reintervention status. Box-and-whisker plots show the distribution of postoperative transmitral gradients in patients with and without reintervention. The lower and upper borders of each box represent the 25th and 75th percentiles, and the horizontal line inside the box represents the median. The whiskers indicate the minimum and maximum nonoutlier values. Individual dots represent single patient values (including outliers when present); for the smaller group (reintervention), all individual values are plotted.

Kaplan-Meier analysis stratified by postoperative gradient (Figure 3) confirmed a markedly reduced freedom from reintervention in patients with gradients >5 mm Hg. The 2 survival curves diverged early, within the first years after repair, and remained separated throughout follow-up.

Figure 3.

Kaplan-Meier curves showing freedom from reintervention after posterior leaflet augmentation, stratified by postoperative (post-op) transmitral mean gradient (≤5 mm Hg vs >5 mm Hg). The y-axis shows freedom from reintervention in percent, and the x-axis shows time after surgery in years. Solid lines indicate Kaplan-Meier estimates, with shaded bands representing 95% CI. The table below the graph reports the number of patients at risk at 0, 2, 4, 6, and 8 years for each group. Patients with post-op gradients >5 mm Hg had significantly lower freedom from reintervention compared with those ≤5 mm Hg (P = .004; hazard ratio, 7.81; 95% CI, 1.9-31.4).

The median time to reintervention was approximately 5 years, indicating that most events occurred within the first decade after repair. After an initial mild postoperative increase, the mean transmitral gradient remained stable during follow-up in a majority of patients. Only a few showed a progressive rise above 5 mm Hg, which correlated with later reintervention. Overall, gradients remained low and steady over time, confirming the durability and physiological adaptation of the repair (Figure 4).

Figure 4.

Individual patient transmitral gradients (thin lines) and overall mean trend after posterior leaflet augmentation. The dark line with dots represents the mean transmitral gradient at each predefined follow-up (FU) timepoint (preoperation [pre-op], postoperation [post-op], discharge, 6 months, 12 months, 24 months, and last FU); the shaded area indicates the 95% CI. After a mild postoperative increase, gradients remained low and stable in most patients, confirming durable valve function and absence of progressive obstruction.

Discussion

Surgical repair of congenital MVD in children remains a major technical and physiological challenge due to the heterogeneity of lesions and the need to ensure valve growth over time.1, 2, 3 Among these conditions, restricted motion of posterior leaflet represents among the most common morphologic substrates, often leading to incomplete coaptation and residual regurgitation after conventional repair.2, 3, 4, 5, 6, 7, 8 Our experience demonstrates that PLA using fresh autologous pericardium provides durable valve competence and stable hemodynamics, confirming its reproducibility and long-term effectiveness in pediatric settings.⁶

In our experience, PLA represents the optimal surgical option for Carpentier type III lesions, characterized by restricted motion of the posterior leaflet with preserved excursion of the anterior leaflet.^1,2 These anatomic conditions are common in congenital mitral dysplasia, where posterior leaflet hypoplasia or adherence to the left ventricular wall results in inadequate coaptation despite a normally mobile anterior leaflet. Although the technique was originally developed for rheumatic mitral disease, its application in the setting of congenital dysplasia with posterior restriction has proven particularly effective, restoring leaflet motion and achieving a physiological closure line.^7,8 This observation is consistent with the lesion profile of our series, in which the majority of patients (71%) were classified as Carpentier type III, and highlights the importance of tailoring the augmentation strategy to the leaflet imbalance typical of congenital forms.

In our cohort, mean mitral regurgitation grade decreased from 3.9 ± 0.6 preoperatively to 1.4 ± 0.5 after surgery and remained stable throughout follow-up, whereas the mean transmitral gradient increased modestly to 4 to 5 mm Hg and showed no progressive rise. These findings are consistent with those of Wierup and colleagues,¹² who reported stable gradients (4.4 ± 1.7 mm Hg) and mild residual mitral regurgitation in 82% of patients after leaflet expansion with fresh pericardium and subpartial annuloplasty. Together, these data confirm that fresh autologous pericardium ensures flexibility, adaptation to growth, and long-term coaptation without calcification or stiffening.^9,13

In our series, severe mitral regurgitation was the predominant lesion; in mixed cases, increased Doppler gradients often reflected high mitral inflow with complete valve opening rather than true anatomical stenosis. Conversely, patients with predominant anatomical mitral stenosis represent poor candidates for posterior leaflet augmentation, in whom patch size is critical and alternative repair strategies are usually preferred.

A residual mean transmitral gradient >5 mm Hg detected intraoperatively identifies patients at risk for late reintervention. This value reflects early residual obstruction rather than progressive late stenosis and likely depends on patch oversizing or inadequate contouring. In most patients, the gradient remained stable at follow-up, whereas only those with an early high gradient developed late dysfunction. This finding emphasizes the importance of precise intraoperative evaluation of valve opening and gradient immediately after weaning from bypass. In contrast to other approaches that combine leaflet augmentation with partial or complete annuloplasty, our technique deliberately avoids any form of annuloplasty.^12,14, 15, 16, 17, 18, 19

This strategy aims to preserve annular growth potential and minimize the risk of postoperative mitral stenosis, which could otherwise increase with annular support. Because the PLA procedure itself tends to mildly reduce the valvular orifice, the addition of annuloplasty would further enhance the risk of obstruction, particularly in small children. The key to reducing obstruction lies in accurate sizing and shaping of the pericardial patch. The patch is designed so that its free-edge side is slightly longer than the annular side to prevent the aortic cusp effect and maintain a smooth leaflet motion and wide coaptation area.⁸ Functionally, the sail-like configuration of the pericardial patch defines a new coaptation plane that replaces the native posterior leaflet margin as the effective closure surface. This geometry promotes laminar flow and prevents residual prolapse or turbulence along the coaptation line, explaining the stable gradients observed during long-term follow-up.

An important refinement in our strategy has been the abandonment of glutaraldehyde-treated pericardium. In children, chemical fixation promotes calcification and may impair annular and leaflet growth, leading to late restrictive physiology. The use of fresh autologous pericardium preserves tissue pliability, allows integration with native structures, and supports physiological remodeling during growth.^9,13 Clinically, PLA should be regarded as a growth-preserving and temporizing repair, particularly valuable in small or symptomatic patients. When reintervention becomes necessary, the procedure serves as a bridge to definitive valve replacement, allowing time for somatic and annular growth and reducing the risks associated with early prosthetic implantation.^20,21

Limitations

The main limitations of this study are its retrospective single-center design and the relatively small cohort size. However, the long and complete follow-up and the homogeneity of surgical technique strengthen the reliability of the findings. Future multicenter studies and biomechanical analyses may better define optimal patch geometry and its adaptation to growth.

Conclusions

PLA with fresh autologous pericardium provides a durable, growth-preserving, and physiologic repair for pediatric MVD predominantly presenting with severe regurgitation. Avoiding annuloplasty minimizes the risk of iatrogenic stenosis, whereas meticulous patch tailoring reduces obstruction and preserves coaptation. A residual mean transmitral gradient >5 mm Hg detected intraoperatively identifies patients at risk for late reintervention. In this context, PLA should be regarded as a safe and effective bridge to definitive valve replacement, allowing somatic and annular growth and optimizing long-term outcomes.

Webcast

You can watch a Webcast of this AATS meeting presentation by going to: https://www.aats.org/resources/optimizing-long-term-outcomes--11371.

Conflict of Interest Statement

The authors reported no conflicts of interest.

The Journal policy requires editors and reviewers to disclose conflicts of interest and to decline handling or reviewing manuscripts for which they may have a conflict of interest. The editors and reviewers of this article have no conflicts of interest.