The elective use of percutaneous transtracheal oxygen insufflation for laryngeal surgery in a patient with a known difficult airway

Abstract

We report the case of successful elective percutaneous transtracheal oxygen insufflation in a patient with high-grade laryngeal stenosis, requiring repeat surgical laryngeal dilation, in the setting of multiple previous failed attempts at intubation and ventilation. This case report highlights the role of this technique as an initial management plan to provide general anaesthesia in a safe and simple way to patients with a known difficult airway. We also describe the use of an intravenous extension kit which allowed end-tidal carbon dioxide to be measured during transtracheal oxygen insufflation.

Case history

We report the case of successful tubeless anaesthesia with elective percutaneous transtracheal oxygen insufflation in a 56-year-old man with high-grade laryngeal stenosis. The patient required repeat surgical laryngeal dilation, in the setting of multiple previous failed attempts at oral intubation, supraglottic airway insertion and bag-mask ventilation. We also describe the use of an intravenous (IV) extension kit which allowed end-tidal carbon dioxide to be measured during expiration.

In 2019 the patient was diagnosed with a T4N2 squamous cell carcinoma of the right base of tongue. This was treated with a three-month course of chemoradiotherapy, which was complicated by severe, circumferential, supraglottic stenosis, resulting in a pinhole posterior glottic opening and an epiglottis that was adherent to the aryepiglottic folds. This had resulted in two previous emergency admissions with airway compromise. On both occasions awake tracheostomy was required (after the failure of awake fibreoptic intubation), followed by balloon dilatation of the glottic opening. After the second presentation of airway compromise, a subtotal supraglottic anterior laryngectomy and partial epiglottectomy was performed with the removal of the majority of anterior supraglottic scar tissue. The patient was then discharged with a tracheostomy in situ.

Two months later he returned for elective laryngeal balloon dilatation. With the tracheostomy still in situ following the previous admission, general anaesthesia was provided with no complications. Owing to the recent partial epiglottectomy a percutaneous gastrostomy was then inserted to reduce the risk of aspiration. The patient was successfully decannulated on the second postoperative day and then discharged home. Surgical closure was not required.

Two months later the patient returned for repeat elective laryngeal balloon dilatation for the first time since being decannulated. A relaxant general anaesthetic was performed. Although the glottic opening could be visualised with a hyperangulated videolaryngoscope, it was too narrow for a Hunsacker Mon-Jet transglottic catheter (Medtronic, Minneapolis, MN, USA), of outer diameter 4.3 mm, to be passed. Ventilation was attempted using a bag and mask, and an Ambu® AuraGain™ supraglottic device (Ambu A/S, Ballerup, Denmark); both were also unsuccessful. A can’t intubate, can’t oxygenate (CICO) situation was declared and a 14-gauge IV cannula was inserted through the healed tracheostomy site in an emergency fashion and jet ventilation was successfully performed through this. He was then discharged home the following day.

A further two months later the patient presented again for a second elective laryngeal dilatation, which is the subject of this case report. On this occasion, the patient had biphasic stridor at rest and a soft, high-pitched voice. He was unable to lie supine due to respiratory distress and therefore slept on his side.

On examination, he was comfortable at rest when sitting up at a 60° angle. His oxygen saturation was 98% on room air with a respiratory rate of 14/min. His mouth opening was 2 cm with poor mandibular protrusion and a Mallampati score of 4. Radiotherapy-induced scarring of the anterior neck severely reduced neck flexion and extension. The site of the recent tracheostomy had healed but was still identifiable.

A 20-gauge IV cannula and a 20-gauge radial arterial line were inserted. Given the previous problems encountered with supraglottic airway access, and the previous success with needle cricothyroidotomy in a CICO situation, our initial airway management plan was elective percutaneous transtracheal access. Hence, the patient was sat up at an angle of 60° with routine anaesthetic monitoring and pre-oxygenated at 30 l/min using the Optiflow™ Transnasal Humidified Rapid Insufflation Ventilatory Exchange system (THRIVE™, Fisher & Paykel Healthcare, Auckland, New Zealand). Anaesthesia was then induced using targeted controlled infusions of remifentanil and propofol, titrated to a bispectral index of 40–60 (BIS, Covidien, Boulder, CO, USA), while maintaining spontaneous ventilation. The THRIVE oxygen flow was increased to 70 l/min and 1% lidocaine with adrenaline 1:200,000 5 ml was infiltrated into the site of the previous tracheostomy. A 14-gauge IV cannula (Introcan Safety IV Catheter, B. Braun Medical Inc., Melsungen, Germany) was then inserted through this site. Adequate placement in the trachea was achieved on the first attempt and was confirmed by the aspiration of air.

An armoured triple lumen IV extension line (triple lumen valve V set, Go Medical Industries Pty Ltd, Subiaco, Western Australia) was then attached to the transtracheal cannula via a Luer Lock connection (Figure 1). This IV extension kit is commonly used for total intravenous anaesthesia (TIVA) and consists of a large bore armoured reinforced main lumen and two smaller side lumens. The anti-reflux valves of the IV set were removed. A Rapid-O₂ device (Meditech Systems Ltd, Dorset, UK) was attached to the large bore armoured reinforced lumen to allow oxygen insufflation. A capnography sampling line was attached to one of the smaller diameter side lumens to measure expired carbon dioxide and confirm the egress of air. Tracheal placement of the airway cannula was confirmed with visualisation and measurement of end-tidal carbon dioxide (ETCO₂). To prevent dislodgement, sutures were used to secure the cannula to the skin.

Figure 1.

14-Gauge cannula inserted into the trachea with an intravenous extension. A Rapid-O₂ (Meditech Systems Ltd, Dorset, UK) oxygen insufflation device is attached to the main lumen of the intravenous extension and a capnography sampling line is attached to one of the side lumens. Sutures were used to secure the equipment at two points (the site of insertion and over the left clavicle).

The patient was paralysed with rocuronium 50 mg, then oxygen insufflation was carried out using the Rapid-O₂ device. This was connected to the anaesthetic machine’s auxiliary oxygen flow rotameter and oxygen was delivered at 15 l/min. Oxygen was delivered at a frequency of 10–15 cycles/min, with each inhalation lasting 1–2 s and each exhalation lasting 3–4 s. To avoid barotrauma and ensure adequate exhalation, we observed the chest rise and fall, and monitored expired carbon dioxide using the capnography line. Laryngoscopy was performed by the surgical team using a C-MAC videolaryngoscope (Karl Storz, Tuttlingen, Germany) with a hyperangulated D blade. Initially the vocal cords could not be identified; however, successful oxygen insufflation aided the identification of the pinhole larynx by the presence of retrograde bubbles (Figure 2).

Figure 2.

Laryngoscopic view before dilatation, showing retrograde bubbles through pinhole larynx.

The laryngeal inlet was then dilated using a wire-guided balloon dilatation catheter and, following this, was easily identifiable and of greater calibre (Figure 3). With the procedure complete, arterial blood gas was measured. The arterial partial pressure of carbon dioxide was 73 mmHg and the arterial partial pressure of oxygen was 488 mmHg. The ETCO₂ at this time was 78 mmHg. Hypercapnia was tolerated because this was seen as a lesser risk compared with the potential barotrauma that could result from more aggressive ventilation to achieve a normal ETCO₂ in the context of almost complete laryngeal stenosis and the use of a device designed for oxygen insufflation rather than ventilation. The oxygen saturations measured by pulse oximetry remained above 98% throughout.

Figure 3.

Laryngoscopic view after dilatation showing identifiable laryngeal inlet and greater calibre.

In order to reduce the ETCO₂ to normal levels the patient was sat up at an angle of 60°, the muscle relaxant was reversed with 200 mg of sugammadex and the patient was allowed to spontaneously ventilate. The hypercapnia was not perceived to be severe in this case as the patient was observed to be breathing adequately. The ETCO₂ level was closely monitored and subsequently seen to return to normal levels within 10 min without further assistance. The propofol and remifentanil infusions were ceased and, once awake and alert, the patient was transferred to the post-anaesthesia care unit. The transtracheal cannula was left in place in the immediate postoperative period, and subsequently removed when the patient was ready to return to the ward. The patient’s informed consent was obtained for publication of this case report and associated de-identified photographs.

Discussion

Patients presenting for ear, nose and throat (ENT) surgery frequently have anatomical abnormalities which make airway management challenging. Common pathologies include tumours, and scarring and stenosis of the oropharynx, glottis and subglottis. Several options are available to manage the airway in these situations. These include awake intubation, fibreoptic guided intubation, and jet ventilation. These may be administered by a supraglottic, transglottic or transtracheal approach. Recent years have seen the emergence of apnoeic oxygenation techniques in general anaesthesia. These include THRIVE and transtracheal oxygen insufflation using a flow-regulated oxygen insufflator such as the Rapid-O₂. Severe glottic narrowing, as in this case, can make supraglottic and transglottic techniques difficult or even impossible. Hence, tracheostomy under local anaesthesia has traditionally been the gold standard in such patients; however, this is associated with complications in up to 30%. ¹ Conversely, non-surgical options that do not require supraglottic access to the trachea include percutaneous transtracheal techniques that make use of jet ventilation or oxygen insufflation. While predominantly known as a technique used as a last resort in CICO situations, percutaneous transtracheal jet ventilation (PTJV) has been reported in elective surgery for many years. ²

PTJV was first described in the 1950s to maintain oxygenation in patients with airway obstruction undergoing anaesthesia.^3,4 Moreover, its successful use has been reported for airway management in elective general and endolaryngeal procedures lasting up to 2 h.^5,6

PTJV has also been utilised in the emergency setting, notably as a rescue technique when conventional methods to secure the airway have failed.^{7 –11} For instance, the United Kingdom’s prospective 4th National Audit Project (NAP4) reported 19 attempts at cannula cricothyroidotomy with jet ventilation over a one-year period. ¹² Of these, 12 attempts (63%) failed and six patients (32%) suffered barotrauma. This high failure rate can be explained in context, such that emergency use of this procedure commonly occurs in situations where unfamiliar equipment is used with infrequent clinical training, with the associated impact of human factors and resultant crisis resource management. ¹³ As a consequence, proceduralists may favour scalpel cricothyroidotomy over cannula techniques in CICO scenarios. ¹⁴

In contrast, multiple reports have demonstrated that PTJV, when employed in a non-emergency clinical context, is a safe and effective technique for the management of a difficult airway in patients undergoing ENT surgery, and has a lower complication rate compared with scalpel techniques such as awake tracheostomy. ¹⁵ A recent systematic review compared the complication rate of PTJV in elective surgery with that seen in CICO emergencies. ² In the 18 studies describing PTJV in 296 elective surgical procedures, device failure was recorded in one (0.3%) and barotrauma in 23 (8%). By comparison, in the 23 studies describing 90 PTJV procedures in CICO emergencies, device failure was recorded in 38 (42%), barotrauma in 29 (32%) and one or more complications found in 46 (51%).

In addition, percutaneous transtracheal techniques have many advantages that include access to the airway before induction of general anaesthesia, thus avoiding the need for laryngoscopy and awake tracheostomy (and the associated complications), as well as providing an unobstructed view of the larynx and leaving the cannula in situ postoperatively in order to supply respiratory support if needed.^{6,16 –18} Interestingly, when the larynx is severely distorted or even unidentifiable, as in our clinical case, the retrograde escape of air through the vocal cords during passive exhalation can aid in location and identification of the glottis. ¹⁵

Possible complications of percutaneous transtracheal techniques include kinking and obstruction of the catheter and barotrauma-related injuries such as subcutaneous emphysema, pneumomediastinum and pneumothorax. ¹⁶ Bleeding and infection at the site of cannula placement have also been described but these are less commonly encountered. ¹⁹ Although the risk of barotrauma with the use of PTJV is low, it can seriously impact patient outcome. Outflow obstruction, high ventilation pressures and the presence of obstructive lung disease have all been described as major risk factors.^20,21 Such complications may be reduced with careful patient selection, the use of an automated jet ventilator (with airway pressure monitoring and control), and an experienced clinician familiar with the technique.

This case report adds to existing knowledge in two ways. First, we describe the employment of an infrequently used technique that is usually restricted to emergency CICO scenarios, to provide general anaesthesia in a safe and simple way in an elective setting, to a patient with distorted airway anatomy and a known difficult airway. Consequently it avoided the repeated use of alternative difficult airway techniques, most of which had been unsuccessful in this patient in the past. Second, this case demonstrates how percutaneous transtracheal insufflation can be made safer and easier by the use of an IV extension kit (such as those commonly used for TIVA). A capnography sampling line can be attached to one of the smaller diameter side ports to measure ETCO_2. The increase in ETCO₂ with expiration confirms the egress of air and in theory may help avoid barotrauma. The numerical ETCO₂ measurements can help to assess the adequacy of ventilation and, in this case report, showed close agreement to a single paired measurement of arterial partial pressure of carbon dioxide (at 78 mmHg and 73 mmHg respectively). It should be noted, though, that using this IV extension kit to measure ETCO₂ has not been approved by the manufacturer and technically should be classified as an off-label use.

In summary, we report the successful use of elective percutaneous transtracheal oxygen insufflation in a patient with severe laryngeal stenosis. The use of an IV extension kit allowed ETCO₂ measurement during expiration and enabled oxygen insufflation during inspiration through a single tracheal cannula. This reinforces the previous reports of this technique as a safe and effective method in establishing ventilation in a patient with a difficult airway undergoing elective ENT surgery. Moreover, securing the airway before induction of anaesthesia in this simple fashion avoids the complications of tracheostomy and laryngoscopy, and provides a tubeless operative field for surgery.