Prone position enhances oxygenation and facilitates weaning from percutaneous extracorporeal membrane oxygenation in infants after the bidirectional Glenn procedure

Abstract

The prone position has been shown to improve oxygenation in patients with acute respiratory distress syndrome (ARDS) and is commonly used in mechanically ventilated patients. However, the short-term efficacy and outcomes of prone positioning in infants requiring veno-venous extracorporeal membrane oxygenation (V-V ECMO) after a bidirectional Glenn procedure have not been reported. A 6-month-old female infant with pulmonary atresia and intact ventricular septum underwent a bidirectional Glenn procedure. On postoperative day 6, the patient was extubated and showed favorable progress. On postoperative day 12, the patient developed ARDS and Glenn failure, leading to the initiation of V-V ECMO through the right subclavian vein (drainage site) and femoral vein (infusion site) for severe hypoxemia. After initial management in the supine and lateral positions failed to improve oxygenation, the prone position was introduced from postoperative day 25 onward, for 8–12 h daily. Oxygenation gradually improved over 12 days, and the prone positioning was discontinued on postoperative day 38. The patient was successfully weaned from V-V ECMO on postoperative day 42, weaned from the mechanical ventilation on day 69, and discharged from the pediatric intensive care unit on day 82 without significant neurological complications. Prone position can be safely implemented in infants undergoing V-V ECMO after a bidirectional Glenn procedure, which might contribute to improving oxygenation and weaning.

Introduction

The prone position has been shown to facilitate clearance of secretions, improve oxygenation, reduce ventilator-associated pneumonia, and play a significant role in early mobilization for patients with acute respiratory distress syndrome (ARDS) [1]. However, in patients requiring extracorporeal membrane oxygenation (ECMO), studies have found no significant difference in ECMO weaning time between the supine and prone positions [2].

The bidirectional Glenn procedure is a palliative procedure for patients with a single ventricle [3]. If Glenn failure occurs due to factors such as increased pulmonary vascular resistance or impaired ventricular function, ECMO may be required. However, despite careful initiation and management, these patients often experience low survival rates and a high incidence of neurological complications [3, 4].

To date, no studies have reported the short-term efficacy of prone positioning in patients requiring ECMO after a bidirectional Glenn procedure. This case report highlights the potential benefits of prone positioning in improving oxygenation and facilitating clinical management in these patients.

Case presentation

A 6-month-old female infant, measuring 61 cm in length and weighing 4.7 kg, had an estimated weight-for-age Z-score (WAZ) of approximately − 3.5, indicating a severe underweight status. The patient underwent a Blalock-Taussig shunt at 24 days of age due to pulmonary atresia with an intact ventricular septum. At 5 months of age, catheterization revealed a pulmonary artery resistance of 1.8 u-m² and a pulmonary artery area index of 203 mm²/m². The patient then underwent the Glenn procedure at 6 months of age. The patient was extubated on postoperative day 6 and exhibited a favorable clinical course. However, on postoperative day 10, hypoxemia and hypercapnia were observed despite the administration of inhaled nitric oxide therapy at 20 ppm using AINOFLO® (Mallinckrodt Pharmaceuticals, Tokyo, Japan), with fraction of inspired oxygen (FiO₂) of 1.0, partial pressure of oxygen (PaO₂) of 38 mmHg, and partial pressure of carbon dioxide (PaCO₂) of 27 mmHg.

Chest radiography revealed ground-glass opacities throughout the entire lung field and consolidation in the right upper lobe (Fig. 1A). Bedside ultrasonography was employed to monitor diaphragmatic motion, revealing no significant impairment. Furthermore, chest radiography showed no evidence of phrenic nerve paralysis, such as diaphragmatic elevation. Laboratory studies revealed elevated white blood cell count (25,500 WBCs/μL), neutrophils (21,800 NEUTs/μL), and C-reactive protein (11.15 mg/dL). To evaluate the potential for infection, serial sputum cultures and laboratory investigations were performed throughout the course of treatment. However, no evidence suggestive of bacterial pneumonia or other infectious etiologies was identified. Consequently, antibiotic therapy was considered unwarranted and was therefore not initiated.

Fig. 1.

Chest X-ray progress. A Postoperative day 11, diffuse ground-glass opacity was observed throughout the lung fields. Lung consolidation and an air bronchogram were observed in the right upper lobe. B Postoperative day 16, V-V ECMO treatment was initiated. The air bronchogram of the right upper lobe was reduced. C Postoperative day 19 before the coil embolization. Extensive consolidation and atelectasis associated with pulmonary hemorrhage were also observed. D Postoperative day 37 after V-V ECMO device detachment. Consolidation in the right and left lung fields improved after prone position ended. V-V ECMO, veno-venous extracorporeal membrane oxygenation

The cardiothoracic ratio was 53.9%, with no enlargement of the cardiac shadow, and the serum brain natriuretic peptide level was 60 pg/mL. Cardiogenic pulmonary edema was excluded due to the absence of systolic or diastolic ventricular dysfunction on echocardiography. The patient was diagnosed with ARDS and Glenn failure, requiring mechanical ventilation.

On postoperative day 12, the blood gas analysis (BGA) showed slight improvement, with FiO₂ of 1.0, PaO₂ of 29 mmHg, and PaCO₂ of 30 mmHg. Veno-venous extracorporeal membrane oxygenation (V-V ECMO) was initiated percutaneously with a Bio-Medicus NextGen 8 Fr (Medtronic Pic, Dublin, Ireland) via the right subclavian vein (drainage site) and femoral vein (infusion site) to address the hypoxemia. The ECMO circuit was equipped with a Mera centrifugal pump, NSH-R (HCF-MP23H, Izumi Kogyou, Tokyo, Japan), and a Nipro BIOCUBE C Z000P artificial lung (Nipro, Osaka, Japan). Oxygenation and gas exchange stabilized at 270 mL/min with ECMO. Thrombo-elastography (TEG® 6 s: Haemonetics Corp, Boston, USA) was used to adjust heparinization, fibrinogen transfusion, and platelet transfusion as part of the anticoagulation therapy. Chest radiography revealed consolidation throughout the lung fields, although the pulmonary vascular shadows were reduced due to ECMO blood drainage (Fig. 1B). Despite correction of hypercapnia after postoperative day 14, the hypoxemia persisted. The BGA showed FiO₂ of 0.8, PaO₂ of 28 mmHg, and PaCO₂ of 36 mmHg.

On postoperative day 19, significant bloody secretions were aspirated from the endotracheal tube, and extensive consolidation was noted on chest radiography (Fig. 1C). Pulmonary hemorrhage from the collateral vessels to the right pulmonary and right bronchial arteries led to the bloody secretions, and coil embolization was performed on the same day.

On postoperative day 24, the pulmonary hemorrhage was controlled with catheter embolization to the atrio-pulmonary collateral, but the patient still showed severe hypoxemia, with FiO₂ of 0.8 and PaO₂ of 28 mmHg. Oxygenation did not improve in the supine or lateral positions. On postoperative day 25, the patient was placed in the prone position for 8–12 consecutive hours daily. Made-to-order fixation cushions were used to stabilize the ECMO cannula and the body when transitioning from the prone to the supine position (Fig. 2). Oxygenation gradually improved over 12 days in the prone position compared with the supine position (Fig. 3). On postoperative day 37 (12 days after beginning prone positioning), the BGA showed improved oxygenation and gas exchange, with PaO₂ of 43 mmHg and PaCO₂ of 39 mmHg under FiO₂ of 0.6, and atelectasis in both lung fields decreased on chest radiography (Fig. 1D). The prone positioning was discontinued on postoperative day 38. The patient was weaned from V-V ECMO on postoperative day 42, weaned from mechanical ventilation on postoperative day 69, and discharged from the pediatric intensive care unit (PICU) on postoperative day 82. The patient was discharged at the age of 1 year without any serious neurological complications.

Fig. 2.

Fixation of ECMO cannulas using made-to-order cushions. To prevent fluctuations in blood flow due to ECMO cannula movement, made-to-order fixation cushions were applied and further secured with elastic bandages. (A) Image showing the overall condition of an infant under V-V ECMO management, with the cannula inserted into the right subclavian vein (drainage site) and the femoral vein (infusion site). (B-I) The upper limb and ECMO cannula were stabilized with a made-to-order cushion, which was further fixed using elastic bandages. (B-II) The thigh and ECMO cannula were similarly stabilized. (C) The infant was placed in the prone position, ensuring proper ECMO cannula alignment to ensure stability. Black dashed lines indicate the ECMO cannulas. ECMO, extracorporeal membrane oxygenation; V-V ECMO, veno-venous extracorporeal membrane oxygenation

Fig. 3.

Time course of the PaO₂/FiO₂ ratio in relation to V-V ECMO and prone position. The solid line represents the PaO₂/FiO₂ ratio in the supine position, while the dashed line represents the PaO₂/FiO₂ ratio in the prone position. V-V ECMO was initiated on postoperative day 12 and discontinued on day 42. Prone position was introduced on day 25 and continued for 12 days. Oxygenation gradually improved in the prone position, leading to successful weaning from V-V ECMO. The duration of mechanical ventilation is also shown. V-V ECMO, veno-venous extracorporeal membrane oxygenation; PaO₂, partial arterial pressure; FiO₂, fraction of inspiratory oxygen

Discussion

This case involved an infant who underwent a bidirectional Glenn procedure and subsequently experienced severe hypoxia and Glenn failure due to ARDS. During the treatment course, the prone position appeared to improve oxygenation and may have facilitated ECMO weaning. This report discusses the prone position and treatment course in infants requiring ECMO following the Glenn procedure. Additionally, three key treatment features enabled the safe implementation of the prone position after the bidirectional Glenn procedure.

In patients with ARDS, the prone position could be important for early mobilization by helping clear secretions, improving oxygenation, and reducing ventilator-associated pneumonia [1, 2]. The role of prone positioning for early mobilization in patients requiring ECMO remains unclear [2]. The prone position improves oxygenation in adults with ARDS by dilating the alveoli, improving the ventilation-perfusion rate, limiting the movement of the anterior chest wall and ventral diaphragm, and releasing the restrictive movement of the posterior chest wall [5]. Although anatomical and functional differences exist between adult and pediatric respiratory systems, the prone position may also improve oxygenation in pediatric patients with ARDS. One study reported that the prone position could be effective in pediatric patients with severe respiratory failure requiring ECMO, but no reports exist on its effectiveness in children after a bidirectional Glenn procedure [6]. When used in pediatric patients requiring ECMO, concerns arise about the potential for inadvertent removal of medical devices such as central venous catheters, endotracheal tubes, and ECMO cannulas [6, 7]. In this case, prone positioning was safely performed in an infant requiring ECMO, with the following two features: First, ECMO cannulation via the internal jugular vein can lead to hemorrhagic complications due to limited neck motion [8]. A drainage cannula for ECMO was inserted via the subclavian vein to mitigate this risk. Compared to using the internal jugular vein for drainage cannulation, this technique may prevent hemorrhagic complications by reducing the movement of the cannula that results from neck motion. Second, when the patient is moved from one position to another, the position change of the ECMO cannula may affect ECMO blood flow and increase bleeding complications [7]. A study described a method for repositioning an ECMO patient from supine to prone [2]. However, no reports have outlined the optimal technique for position change from supine to prone during ECMO, especially in pediatric patients. In this case, made-to-order fixation cushions were used to fix the ECMO cannula and the patient’s body for position changes, which may have helped prevent changes in ECMO blood flow and reduced the risk of bleeding complications.

Patients requiring ECMO after a bidirectional Glenn procedure often have a lower survival rate and a higher incidence of neurological complications [3, 4]. In the bidirectional Glenn procedure, the superior vena cava is connected to the pulmonary artery, while the inferior vena cava is connected separately to the right atrium. Poor blood drainage can lead to increased cerebral venous pressure, decreased cerebral perfusion, and reduced systemic perfusion [9]. Drainage cannulation is often performed via placement of the femoral vein, superior vena cava under the open chest, or the right internal jugular vein to the superior vena cava [9]. However, there are concerns that ECMO via cannulation of the superior vena cava under an open chest may result in stenosis and thrombus in the Glenn circuit [4]. To date, no reports have identified the optimal insertion site for ECMO in patients after the bidirectional Glenn procedure [9]. In this case, ECMO cannulation was performed from the right subclavian vein into the superior vena cava, avoiding the internal jugular vein. This approach successfully reported the patient’s recovery without neurological complications. ECMO cannulation through the subclavian vein into the superior vena cava does not directly obstruct the internal jugular vein, which may prevent a decrease in cerebral perfusion. However, this technique does not eliminate the risk of stenosis or thrombus formation in the Glenn circuit.

Three types of V-V ECMO circuit systems can be considered for respiratory support in patients post-Glenn procedure. The first circuit involves ECMO cannulation via the femoral vein (drainage site) and femoral vein (infusion site), which may result in recirculation. The second circuit system uses the femoral vein or inferior vena cava (drainage site) and the Glenn circuit (infusion site), which could affect ECMO blood flow to the Glenn circuit. Increased blood flow to the Glenn circuit may lead to pulmonary edema and increase Glenn pressure [9]. The third circuit system employs ECMO cannulation via the Glenn circuit (drainage site) and the femoral vein or inferior vena cava (infusion site), which may help reduce Glenn pressure by decreasing pulmonary blood flow. However, this system could obstruct the Glenn circuit due to the ECMO cannula insertion. In patients with elevated Glenn pressure and hypoxia, a V-V ECMO system may still be necessary, as previously discussed.

This study has several limitations. First, percutaneous ECMO may not be feasible in smaller infants who cannot accommodate 8 Fr cannulas, thereby complicating the application of prone positioning. Second, as this is a single case report, the findings have limited generalizability. While infection was considered and ruled out based on consistently negative test results, a subclinical infectious process could not be entirely excluded, raising the possibility of systemic inflammatory response syndrome (SIRS). Additionally, the patient’s low body weight and presumed preoperative malnutrition may have contributed to delayed respiratory recovery [10], emphasizing the importance of carefully evaluating nutritional status when interpreting the clinical course.

Conclusion

Prone positioning was safely applied in an infant requiring percutaneous V-V ECMO after the bidirectional Glenn procedure. It contributed to improved oxygenation and successful ECMO weaning without complications. Although no neurological complications were observed at discharge, the long-term outcomes remain uncertain and warrant further follow-up. This case suggests that prone positioning may be a feasible adjunctive strategy in selected pediatric ECMO patients.

Author contribution

Naoki Kaneda conceived the study, collected patient data, and drafted the manuscript. Tomohiro Nawa reviewed the patient information. Wataru Sakai contributed to the study conception and provided direct supervision during manuscript preparation. Tomohiro Chaki was involved in manuscript supervision and critical revision. All authors read and approved the final version of the manuscript for publication.

Funding

None.

Data availability

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

Declarations

Ethical approval

Not required as per our hospital protocol.

Informed consent

The parents of the patient were fully informed of the purpose of this case report, and consent was obtained to describe the physical therapy evaluation and progress.

Statement of human and animal rights

This case report was conducted in accordance with the ethical standards of the 1964 Declaration of Helsinki and its subsequent amendments. No studies involving animal subjects were included in this article.

Conflict of interest

The authors declare that there are no conflicts of interest.