Flexible Bronchoscopic Thrombus Cryoextraction in a Neonate on Extracorporeal Membrane Oxygenation

To the Editor:

Airway hemorrhage and subsequent thrombosis are complications of extracorporeal membrane oxygenation (ECMO) support that are difficult to manage in neonates. Recanalization of the airway is necessary for ventilation and decannulation from ECMO but is limited owing to bleeding risks and available tools. Rigid bronchoscopy with manual recanalization is possible but may be difficult in small patients because of equipment and positioning requirements (1). Although flexible bronchoscopy may be used to clear thrombus, suction and tools are limited because only the smallest bronchoscopes can be used in neonates.

The use of cryoprobes to remove thrombi has been reported in adults, including those on ECMO (2, 3). Modern systems use single-use cryoprobes and carbon dioxide electronically controlled with a footswitch to freeze endobronchial material to −50°C. Although cryoprobes suitable for pediatric patients have recently become available, there is limited evidence supporting this practice in children (2, 4–6). We report using a 1.1-mm flexible cryoprobe (Erbe) to safely and effectively extract obstructive thrombi from a neonate on ECMO.

Patient Case

A 3.6-kg female was delivered at 37 5/7 weeks with a severe left congenital diaphragmatic hernia, with intestine, stomach, spleen, and left hepatic lobe in the left hemithorax. She was intubated with a 3.5-mm endotracheal tube but experienced respiratory failure refractory to surfactant and mechanical ventilation. Ultimately, she was placed on venoarterial ECMO on Day of Life (DOL) 1.

The patient developed radiographic opacification of the right hemithorax (Figure 1A) despite multiple ventilation modalities. Flexible bronchoscopy on DOL 7 revealed copious mucus plugging, which was removed with saline lavage and suction (BF-XP160F, working channel 1.2 mm; Olympus). The right hemithorax remained opacified, and the patient developed bloody secretions from the endotracheal tube. Repeat flexible bronchoscopy on DOL 9 revealed a large plug that appeared to be an arborized thrombus filling the right bronchial tree (Figure 2A), which was unable to be evacuated.

Figure 1.

Chest radiograph revealing (A) complete opacification of the right hemithorax and absence of air bronchograms and (B) improved aeration following cryoextraction of an endobronchial thrombus.

Figure 2.

(A and D) Bronchoscopic image of the RMB and RLL with an obstructive endobronchial thrombus. (B and E) A 1.1-mm flexible cryoprobe (Erbe) was passed through the working channel with (C and F) successful extraction. RLL = right lower lobe; RMB = right main bronchus.

Repeat flexible bronchoscopy on DOL 10 and 12 were unsuccessful in removing this thrombus despite exchanging the patient’s endotracheal tube for a laryngeal mask airway to allow passage of a bronchoscope with a larger working channel (BF-P190, 2.0 mm; Olympus). Biopsy forceps were unable to open in the patient’s airways owing to size limitations. The patient was placed on high-frequency oscillatory ventilation, saline nebulization, and N-acetylcysteine nebulization, without radiographic change. On DOL 11, she underwent congenital diaphragmatic hernia repair, which resulted in partial expansion of the left lung but persistent right hemithorax opacification.

On DOL 14, flexible bronchoscopy was performed, using a disposable 1.1-mm flexible cryoprobe (Erbe) through the Olympus BF-XP160F. The clot was partially removed from the right main bronchus in three pieces using cryoextraction with ∼10-second freezing cycles (Figures 2B and 2C). Visibility of the lobar bronchi was impaired by secretions and inability to clear the field of view with the probe in place. The procedure was repeated on DOL 15 and DOL 16, this time using a laryngeal mask airway and the Olympus BF-MP190F (working channel 1.7 mm), which improved instrument flexibility and suction with the cryoprobe in place. Using this technique, all residual thrombus was successfully extracted in two pieces (Figures 2D–2F). With support from the pulmonary, neonatology, and surgical teams, the thrombus was extracted without significant hemorrhage or complication and with marked improvement of right lung aeration (Figure 1B). The patient tolerated these procedures well with procedural sedation requirements similar to her previous bronchoscopies.

Discussion

Pediatric flexible bronchoscopy is frequently performed for diagnostic purposes and is increasingly used for therapeutic and interventional indications, such as removal of mucus plugs and thrombi in mechanically ventilated patients, including those on ECMO. In neonates, clearance of the airways can be difficult owing to the small working channel of pediatric bronchoscopes and because many endoscopic tools are too large to deploy in the neonatal airway.

Use of cryotherapy to treat airway diseases has been described for some time and has become prevalent in adult centers (2, 3, 7). Data on cryotherapy in children are limited (4). Reports have established the safety of cryotherapy in pediatric school-age patients for treatment of atelectasis, endobronchial tuberculosis, and foreign body extraction (5, 6, 8, 9). To our knowledge, only one instance of cryoextraction in a child on ECMO has been reported (6). This patient was older and larger than our patient (7 yr and 27 kg) and was managed on heparin-free ECMO. In the present case, we were able to deploy a newly developed 1.1-mm flexible cryoprobe (Erbe) to extract a thrombus from a 3.6-kg anticoagulated neonate with severe pulmonary hypertension on ECMO support without complication. The advent of the 1.1-mm cryoprobe permits the use of this technique in any pediatric patient that can accommodate a 2.8-mm bronchoscope and will permit cryoextraction to the level of the segmental bronchi.

Patient-specific risks are essential in consideration of cryoextraction. In patients on ECMO, anticoagulation may significantly worsen any airway hemorrhage. This may be further exacerbated in patients with pulmonary hypertension. As cryoextraction itself may cause mucosal trauma, the size of the airway relative to the cryoprobe is important; for this reason, the 1.1-mm cryoprobe was used in place of the 2.0-mm probe. Finally, the relative risks and benefits of other available therapies such as fibrinolytics, rigid bronchoscopy, or repeat flexible bronchoscopy to slowly remove clot over time must be considered against the risk of prolonged ECMO support (e.g., intracranial hemorrhage). In our patient, repeated failure of thrombus extraction using traditional bronchoscopy tools, coupled with urgency to improve lung expansion and wean ECMO support, prompted a trial of cryoextraction, which was ultimately successful.

This example of flexible bronchoscopic cryoextraction in a critically ill, anticoagulated neonate on ECMO provides proof of concept for children of almost any age and size, although risks must be weighed before such an undertaking. This should only be considered in facilities with the experienced, multidisciplinary team necessary to support cryoextraction in high-risk situations. The ongoing development of advanced tools able to deploy via small-working-channel bronchoscopes will empower novel therapeutic and diagnostic approaches in pediatric bronchoscopy.