Comparison of ductus stent versus surgical systemic-to-pulmonary shunt as initial palliation in patients with univentricular heart

Abstract

Open in new tabDownload slide

OBJECTIVES

In this study, we aimed to compare infants with univentricular hearts who underwent an initial ductus stenting to those receiving a surgical systemic-to-pulmonary shunt (SPS).

METHODS

All infants with univentricular heart and ductal-dependent pulmonary blood flow who underwent initial palliation with either a ductus stenting or a surgical SPS between 2009 and 2022 were reviewed. Outcomes were compared after ductus stenting or SPS including survival, probability of re-interventions and the probability to reach stage II palliations.

RESULTS

A total of 130 patients were evaluated, including 49 ductus stenting and 81 SPSs. The most frequent primary diagnosis was tricuspid atresia in 27, followed by pulmonary atresia with intact ventricular septum in 19 patients. There was comparable hospital mortality (2.0% stent vs 3.7% surgery, P = 0.91) between the groups, but shorter intensive care unit stay (median 1 vs 7 days, P < 0.01) and shorter hospital stay (median 7 vs 17 days, P < 0.01) were observed in patients with initial ductus stenting, compared to those with SPS. However, acute procedure-related complications were more frequently observed in patients with ductus stenting, compared with those with SPS (20.4 vs 6.2%, P = 0.01), and 10 patients needed a shunt procedure after the initial ductus stent. The cumulative incidence of reaching stage II was similar between ductus stenting and SPS (88.0 vs 90.6% at 12 months, P = 0.735). Pulmonary artery (PA) index (median 194 vs 219 mm²/m², P = 0.93) at stage II was similar between patients with ductus stenting and SPS. However, the ratio of the left to the right PA index [0.69 (0.45–0.95) vs 0.86 (0.51–0.84), P = 0.015] was higher in patients who reached stage II with surgical shunt physiology, compared with patients with ductus stent physiology.

CONCLUSIONS

After initial ductus stenting in infants with univentricular heart, survival is comparable and post-procedural recovery shorter, but more acute stent dysfunctions and lower development of left PA are observed, compared to acute shunt dysfunctions. The less invasive procedure and shorter hospital stay are at the expense of more stent reinterventions.

INTRODUCTION

Patients with univentricular heart and reduced ante-grade pulmonary blood flow (PBF) or ductus-dependent PBF need augmentation of PBF in the neonatal period [1]. Historically, this was achieved by a surgical systemic-to-pulmonary shunt (SPS) [2, 3]. SPS improves PBF, promotes pulmonary artery (PA) growth and thus allows for stage II palliation to be performed several months later. However, there remains a certain risk of morbidity and mortality with this surgical procedure [4–6]. It is well known that patients with single ventricle and pulmonary atresia are at the highest risk of death and re-interventions, compared to those with other anatomical types [7–9]. Over the last decade, stenting for patent ductus arteriosus (PDA) has emerged as a viable alternative to SPS in patients with reduced ante-grade PBF and/or ductus-dependent PBF [10, 11]. Interventional ductal stenting (DS) is a less invasive approach than SPS [12–17]. Recent studies comparing the results after DS and SPS demonstrated earlier recovery after DS than SPS [18–21]. Nevertheless, DS also carries the risk of procedural-related complications, such as dislocation and dysfunction of the stent and an increased need for reintervention including a switch to the emergent SPS [22]. Adequate PA growth after DS is controversial [23, 24]. Previous studies comparing DA and SPS have been mainly performed in patients with both uni- and biventricular heart and few studies have been performed focusing on the univentricular patients. Outcomes including stage II and stage III palliation have not been fully studied.

Thus, this study aimed to compare the survival after DS and SPS in infants with univentricular hearts with reduced PBF and/or ductus-dependent PBF. Our other objective was to compare DS with SPS regarding hospital recovery, complications and PA development. The comparison of outcomes for stage II and stage III palliation was also performed.

METHODS

Ethical statement

This study was approved by the Institutional Review Board of the Technical University of Munich (approved number of 422/2023 S-SR on 14 August 2023). Because of the retrospective nature of the study, the need for individual patient consent was waived.

Patients and data collection

We reviewed the medical records of infants with univentricular heart and reduced PBF and/or ductus-dependent PBF who underwent SPS or DS as initial palliation between 2009 and 2022. A preoperative conference between paediatric cardiologists and paediatric cardiac surgeons made the choice of initial SPS or DS. This was mainly due to the feasibility of DS depending on the size and anatomy of the PDA and also due to the technical issues of DS. Infants with hypoplastic left heart syndrome and its variants who underwent the Norwood procedure were excluded from this study. Among non-Norwood patients, 15 SPS patients who underwent concomitant surgeries (8 total anomalous pulmonary venous drainage repairs, 2 Starnes procedures, 2 aortic arch repair and 3 pacemaker implantations) were excluded. Medical records included baseline morphology and demographics, as well as pre-, intra- and post-procedural data using electronic and paper chart reviews of each patient. Follow-up data were collected using the institutional single ventricle database regarding the outcomes of stage II and stage III palliations. As for the measurements of PA size, all angiographic measurements were obtained by 1 investigator (Dimitrij Grozdanov). Then, the PA indices were calculated as described by Nakata et al. [25]. The right PA- and left PA indices were calculated by dividing the cross-sectional area of each PA at the bifurcation of the first branch by the body surface area. To evaluate the symmetric PA development, PA symmetry index was also calculated as described by Glatz and colleagues [14]. Shortly, PA symmetry index was a ratio of the cross-sectional area of the smaller PA to the larger PA. The symmetry index is always ≤1, with values closer to 1 reflecting more symmetrical vessel size. The ratio of the left PA index to the right PA index was also calculated.

As for the comparison of clinical outcomes and PA development between the groups, clinical outcomes were compared between patients who underwent initial DS and those who underwent initial SPS. As for the PA development before bidirectional cavopulmonary shunt (BCPS), the comparison was performed between patients who reached BCPS with DS physiology and those who reached BCPS with SPS physiology.

Operative techniques

SPS was performed through a median sternotomy and standard cardiopulmonary bypass (CPB) was used in most of the patients [6]. According to the anatomy of the patients, either modified Blalock–Taussig shunt (MBTS) or central aortopulmonary shunt (CS) was performed. As for the patients undergoing MBTS, a Gore‐Tex tube (W. L. Gore and Associates) was anastomosed between the subclavian artery and the respective branch PA. The choice of the tube size was dependent on the patients' weight. A tube size of 3.5 mm was most frequently used. As for the patients undergoing CS, a Gore‐Tex tube was anastomosed between the ascending aorta and the main PA.

Interventional techniques

Before the intervention, the PGE infusion was titrated to achieve a PDA with an echocardiographic visible stenosis or an arterial saturation below 80%. Depending on the intracardiac anatomy and the angiographic anatomy of the PDA, the approach for DS was chosen between transvenous or transarterial (either femoral or axillary artery). Usually, a telescope technique with a 5-F guiding catheter, a 4-F diagnostic catheter and a soft coronary guidewire was used. After probing the duct with the coronary wire, the diagnostic catheter was removed and a premounted coronary stent was brought into position and implanted. The length of the stent was chosen according to the angiographic length of the PDA with the wire in position. Care was taken that the whole length of the duct was covered and, if needed, additional stents were implanted. No specific anatomy of the PDA was considered a contraindication for this treatment. All patients were treated with Clopidogrel 0.2 mg until the next surgical stage.

Statistical analysis

Categorical variables are presented as absolute numbers and percentages. A chi-squared test was used for categorical data. Continuous variables are expressed as medians with interquartile ranges. An independent sample Student’s t-test was used to compare normally distributed variables. The Mann–Whitney U-test was used for variables that were not normally distributed. Follow-up was closed on 1 April 2023 and was completed. To analyse survival in the follow-up, censoring occurs for those patients who did not die at the time of the last follow-up. Survival after the initial procedure and the cumulative incidence of reaching stage II and stage III palliation were calculated using the Kaplan–Meier method, and the differences between the groups were determined using a log-rank test. Listwise deletion was used for missing data. Data analysis was performed using SPSS version 28.0 for Windows (IBM, Ehningen, Germany).

RESULTS

Patient characteristics and periprocedural data

A total of 130 patients were identified during the study period, including 49 (37.7%) patients who initially underwent DS, and 81 (62.3%) patients who underwent SPS. Among 81 SPS patients, 5 (6.2%) were intended to receive DS, but this could not be achieved due to technical difficulties. Because no DS was placed in these 5 patients, they must be grouped with the SPS group. Patient characteristics are presented in Table 1. Patients with pulmonary atresia and intact ventricular septum were more frequently palliated with DS than SPS (P = 0.049). No significant difference was observed in other characteristics.

Perioperative data are shown in Table 2. In 49 patients with DS, a 3.5-mm stent was most frequently placed. In 81 patients with SPS, MBTS was performed in 62 (76.5%) patients and CS in 19 (23.5%). Postoperatively, early deaths within 30 days [0 (0%) vs 3 (3.7%), P = 0.173] and hospital mortality [2 (4.1%) and 3 (3.7%), P = 0.914] were similar among patients with DS and SPS. There were 2 hospital deaths in more than 30 days in the DS group. Both patients need cardiopulmonary resuscitation in the intensive care unit (ICU) and extracorporeal membrane oxygenation (ECMO) support. One patient died on the 32^nd post-procedural day in the ICU and the other was discharged from ICU on the 88^th post-procedural day, but died in the ward on 202nd post-procedural day. The median length of ICU stay [1 (0–5) vs 7 (5–13) days, P < 0.001] and median hospital stay [7 (5–13) vs 17 (9–26) days, P = 0.004] were shorter in patients with DS, compared to those with SPS. Whereas the prevalence of acute stent/shunt dysfunction/dislocation [10 (20.4%) vs 5 (6.2%), P = 0.014] and the probability of need for interventional balloon dilatation [10 (20.8%) vs 4 (4.9%), P = 0.005] were higher in patients with DS, compared to those with SPS. The need for surgical shunt procedure after the initial procedure [10 (20.4%) vs 5 (6.2%), P = 0.014] was higher in patients with DS, compared to those with SPS. The details of the 10 patients who needed an SPS after the initial DS are shown in Supplementary Material, Table S1. As for the postoperative events, the probability of reintubation [6 (12.5%) vs 11 (13.6%), P = 0.861], need for peritoneal dialysis [3 (6.3%) vs 2 (2.5%), P = 0.282] and probability of necrotizing enteropathy [4 (8.3%) vs 6 (7.4%), P = 0.849] were similar between the groups. The probability of postoperative ECMO support [5 (10.2%) vs 7 (8.6%), P = 0.766] was also similar between the groups. As for the 5 ECMO supports in DS patients, ECMO was implanted in the catheter laboratory in 1 patient, in the ICU in 2 patients and in the operating room after switching to SPS in 2 patients.

Second- and third-stage palliation

The follow-up are shown in Table 3. There was no difference in the cumulative incidence of reaching stage II palliation (90.0 vs 89.7% at 12 months, P = 0.766) or stage III palliation (86.4 vs 85.3% at 3 years, P = 0.599) between the groups. The median age at stage II [4.5 (3.7–7.1) vs 4.3 (3.7–5.7) months, P = 0.403] and stage III [2.0 (1.7–2.4) vs 1.8 (1.7–2.4) years, P = 0.903] palliation was also similar. The Kaplan–Meier survival curve is shown in Fig. 1. Because no patient underwent heart transplantation, survival is the same as transplant-free survival. Overall survival after the initial procedure at 2 years was 91.3% in DS patients and 87.0% in SPS patients (P = 0.774).

Comparison between patients with ductal stenting physiology and systemic-to-pulmonary shunt physiology

Because 10 patients after DS underwent surgical SPS, we re-analysed the post-discharge variables between 39 patients who stayed DS physiology and 91 patients with SPS physiology (81 initial SPS patients and 10 patients who underwent SPS after the initial DS). The results are shown in Table 4. There was no difference in the cumulative incidence of reaching stage II palliation (88.0 vs 90.6% at 12 months, P = 0.735) or stage III palliation (83.7 vs 87.0% at 3 years, P = 0.418) between the groups.

The comparisons of PA development are shown in Table 4. The median PA index [202 (147–292) vs 215 (134–254) mm²/m², P = 0.727], median right PA index [122 (77–155) vs 100 (59–146) mm²/m², P = 0.322] and median left PA (LPA) index [79 (47–128) vs 91 (58–135) mm²/m², P = 0.663] were similar between the groups (Fig. 2). However, the ratio of the left PA to the right PA index [0.69 (0.45–0.95) vs 0.86 (0.51–0.84), P = 0.015] was higher in patients who reached BCPS with SPS physiology, than in those with DS physiology (Fig. 3). Median PA symmetry index [0.67 (0.45–0.83) vs 0.68 (0.45–0.84), P = 0.829] was similar between the groups (Fig. 3).

DISCUSSION

The present study demonstrated that patients recovered faster after DS than after SPS. However, the procedure-related complications were more frequent. The probability of reaching stage II and stage III palliation and overall survival were similar between patients with DS and SPS. When compared between patients with DS and SPS physiology at the time of BCPS, the ratio of left PA to right PA index was lower in DS patients than in SPS patients.

Patent ductus arteriosus stent

Since its introduction in 1992, percutaneous DS has been used by a number of centres to provide initial palliation for duct-dependent lesions [10, 11]. As most of the patients with univentricular heart receive a prenatal diagnosis, they are referred to our centre just after birth, the ductus arteriosus being still patent. DS has become a primary procedure in such patients. The technical feasibility is improving and nowadays, DS is indicated for almost all patients. One reason might be the high mortality after SPS in these patients. Previous studies demonstrated that the operative mortality after SPS in single ventricle patients ranged from 6% to 18% [4, 5, 7–9]. It must be said that the initial stage of single ventricle palliation is in fact a very dangerous one. The indication for surgical SPS has become limited and it is performed when there are lesions to be addressed concomitantly with shunt placement, or when the interventional approach is deemed technically too risky. In our experience, patients with small or closing ductus have typically high-risk anatomy. Usually, the PDA is visualized and a probe-wire is positioned in the left PA. However, the PDA is narrowest at the distal end. Therefore, the stent catheter cannot be pushed forward to the narrowest point. Some patients experience saturation drop due to repeated manipulation and the attempt to place a PDA stent must be abandoned. Furthermore, patients with large ductus (>4 mm diameter) and patients with strongly meandered ductus also have technically risky anatomy. These patients are at risk for stent dislocation or dysfunction even if the initial stenting was successful. In small patients below 2.5 kg who have an increased risk for SPS, DS is more likely to be performed. The low body weight does not increase the risk for DS.

The great advantage of DS over SPS is that DS requires shorter hospital stays. Our data showed a median 1-day stay in the ICU and 7 days of hospital stay after DS, compared to 7 days of ICU stay and 17 days of hospital stay after SPS. The use of DS eliminates the risks associated with the use of CPB and reduces the risk of open chest procedure-related complications. In this study, a relatively high rate of patients received SPS using CPB. The reasons for using CPB are haemodynamic instability during the test-clumping of PA, anatomical difficulties and the need for concomitant atrioseptectomy. It is our institutional policy to perform early atrioseptectomy at the time of stage I palliation. Previous studies have demonstrated a shorter hospital stay in DS patients, compared to SPS patients [16, 18, 19]. Consistent with previous studies, there was no difference in mortality between DS and SPS in this study [13, 14, 18, 19].

Re-interventions after ductal stenting

The incidence of acute dysfunction/dislocation was higher in DS patients than SPS patients. There remains a technical challenge in DS, although the tortuosity of the PDA was not considered a contraindication to stent deployment. Stent dislocation or dysfunction was observed in some difficult cases, and a switch to surgical SPS was needed. In this study, 10 patients needed surgical SPS after the initial DS. Six patients had stent dysfunction, 3 patients stent dislocation into the right (n = 1) or left (n = 2) PA, and 1 patient had cardiopulmonary resuscitation during the attempt for second DS and needed emergency SPS. Technical improvement could be achieved in the future, and this problem might be solved by further development in devices and techniques.

Development of pulmonary arteries

Recent evidence has shown that DS might provide a more evenly distributed PBF and promote more balanced growth of the pulmonary arteries [14]. Glatz et al. [14] reported that balanced PA growth favoured PDA stenting. Santoro et al. [21, 23, 24] demonstrated that DS is highly effective in promoting a significant and balanced catch-up growth of PA. Whereas Helal et al. [19] reported that DS patients had a higher rate of stent thrombosis and interstage reintervention. Our data had no significant difference in symmetricity between DS and SPS. However, the ratio of left PA index to right PA index was lower in DS patients compared to SPS patients. We observed frequently that the DS was placed from the PDA to the right PA and the blood flow of the left PA was supplied through the stent. This might cause a stenosis of the LPA after DS. Further studies are mandatory to determine the superiority of DS over SPS in the development of PA in patients with univentricular heart.

Future prospective

DS is now a default procedure at our institute. There was no favourable indication for primary SPS. This is partially due to the evolving techniques and newly developed kinds of stent products. Another reason is the first track recovery and cost-benefit of DS. As this study showed, DS patients need only 1 day or even some hours in the ICU. A cost comparison of these 2 approaches is beyond the scope of this study but is achievable with the data presented. Shorter ICU stay and shorter hospital stay need to be balanced against a greater need for reintervention in the DS group. Until now, DS is not always successful, and some patients need surgical SPS following DS or an attempt to DS. Switching to the surgical SPS procedures still plays an important role. DS and SPS remain complementary procedures until we have data to show that the DS group performs better up to the Fontan stage and beyond, compared to the SPS group. We could not reach this conclusion from our presented data.

Limitations

This study has clear drawbacks of retrospective nature and single-centre analysis with its inherent biases. A small number of patients does not have the power to detect clinically relevant differences between the groups. The decision-making process has changed during the study period, adding to possible biases. The long-term outcomes including survival and neurological outcomes were not evaluated. Missing data are likely linked to the outcome and could introduce bias. Patient selection bias was an obvious limitation, although our pragmatic approach has generally been to perform DS in patients at increased risk of surgical complications.

CONCLUSIONS

DS is associated with shorter hospital stays and avoids the surgical complications of SPS. However, it is associated with a higher probability of acute stent dysfunction/dislocation necessitating reintervention. Survival and the number of patients who reached stage II and stage III palliation are comparable between the groups. DS appears to be a safe and effective alternative to SPS in patients with univentricular heart because of its cost-effectiveness and better survival, if successful.

Presented at the 37th Annual Meeting of the European Association for Cardio-Thoracic Surgery, Vienna, Austria, 4–7 October 2023.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

FUNDING

This study was not supported by any grants.

Conflict of interest: none declared.

DATA AVAILABILITY

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Author contributions

Dimitrij Grozdanov: Conceptualization; Data curation; Investigation; Writing – original draft. Takuya Osawa: Conceptualization; Data curation; Investigation; Methodology; Writing – original draft. Kristina Borgmann: Data curation; Investigation; Methodology; Writing – review & editing. Thibault Schaeffer: Validation; Writing – review & editing. Helena Staehler: Conceptualization; Data curation. Chiara Di Padua: Conceptualization; Investigation; Methodology; Supervision; Validation. Paul Philipp Heinisch: Conceptualization; Data curation; Validation; Writing – review & editing. Nicole Piber: Conceptualization; Data curation. Stanimir Georgiev: Conceptualization; Supervision; Writing – review & editing. Alfred Hager: Conceptualization; Investigation; Methodology; Supervision; Validation; Writing – review & editing. Peter Ewert: Conceptualization; Investigation; Methodology; Project administration; Supervision; Validation; Writing – review & editing. Jürgen Hörer: Conceptualization; Investigation; Methodology; Supervision; Validation; Writing – review & editing. Masamichi Ono: Conceptualization; Data curation; Formal analysis; Validation; Visualization; Writing – original draft.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Yoshihiro Oshima and the other anonymous reviewers for their contribution to the peer review process of this article.

REFERENCES

ABBREVIATIONS

ABBREVIATIONS
 
• BCPS

  Bidirectional cavopulmonary shunt

 
• CPB

  Cardiopulmonary bypass

 
• CS

  Central aortopulmonary shunt

 
• DS

  Ductal stenting

 
• ICU

  Intensive care unit

 
• MBTS

  Modified Blalock–Taussig shunt

 
• PA

  Pulmonary artery

 
• PBF

  Pulmonary blood flow

 
• PDA

  Patent ductus arteriosus

 
• SPS

  Systemic-to-pulmonary shunt

Author notes