Emergency tracheostomy for failed intubation due to glottic stenosis

Abstract

Glottic stenosis can be an unexpected finding during an intubation, causing difficulties that may result in a ‘can’t intubate, can’t ventilate’ situation. We present a case of a patient who required an emergency tracheostomy, in the setting of a failed intubation secondary to glottic stenosis. The patient underwent open laryngotracheal reconstruction, followed by tracheostomy decannulation 2 months post-surgery. This paper highlights the importance of awareness of laryngeal pathology masquerading as respiratory conditions. It also outlines the critical approach to managing ‘can’t intubate, can’t ventilate’ situations.

Background

Glottic stenosis can cause difficulty with intubations, however, it can go unrecognised during pre-anaesthetic assessment, especially if there is no history to suggest this or if they have been diagnosed in the past with another condition that mimics the symptoms. Glottic stenosis is easily misdiagnosed as asthma or chronic obstructive pulmonary disease, as the patients often have dyspnoea, wheeze or stridor.^{1 2} This misdiagnosis can be potentially life-threatening, as the inability to intubate after anaesthetic induction can lead to the inability to ventilate the patient.

Case presentation

A 36-year-old Indigenous woman was admitted for an elective nasal bone manipulation under anaesthesia and tooth extraction under the maxillofacial team. She was noted to have a history of poorly controlled asthma, with an intensive care unit admission aged 11 years and admission under the general medical team aged 22 years for acute exacerbation of asthma. There was no documentation of prior procedures under general anaesthetic.

Following anaesthetic induction, the anaesthetist noted that bag-mask ventilation (BMV) was easy with small volumes around 150 mL. Laryngoscopy was performed with a C-MAC video laryngoscope both directly and indirectly. Although the bougie passed through the vocal cords easily, a size 7 endotracheal tube (ETT) was unable to be passed. A second attempt was repeated after a period of BMV, this time with a size 6 ETT over the bougie, also unsuccessfully. At this time, it was noted that the cords appeared to be almost fully opposed with only a small opening. A third attempt after BMV with a size 5 microlaryngoscopy tube over a bougie was also unsuccessful.

At this time, the anaesthetist decided to place a laryngeal mask airway (LMA) instead. With the LMA in situ, the patient was able to be ventilated with 150 mL volumes, however, with significant leakage, resulting in inadequate ventilation. Bronchoscopy down the LMA showed opposed cords with a small slit of space between.

The otolaryngologist on call was contacted, and on her arrival, the decision was made to attempt waking the patient. The patient failed to wake, with hypercapnia (end-tidal carbon dioxide 80 mmHg) suspected to be the cause. Flumazenil and naloxone were given. Two additional attempts to intubate by the anaesthetist and by the otolaryngologist failed.

With the LMA replaced, the otolaryngologist performed a tracheostomy. Interestingly, the patient already had a horizontal scar across the midline of the anterior neck at the level of the cricoid cartilage that appeared to be from a previous tracheostomy, though there was no mention of a tracheostomy in her available medical records. A direct laryngoscopy was performed following the tracheostomy. This revealed rigidity of the arytenoids bilaterally and inability to spread the true vocal cords laterally (see figure 1).

Figure 1.

Endoscopic photographs of the patient’s glottic stenosis (A) and the subglottic view after instrument separation of the vocal cords (B).

Post-tracheostomy review of the patient’s medical and developmental history highlighted the concern of previous airway surgery as an infant. The patient had been born prematurely at 25 weeks’ gestation by spontaneous vaginal delivery. She had a long admission, reported by the family to be 11 months’ duration, during which she was diagnosed with congenital subglottic stenosis, requiring intubation to secure her airway. It is unclear whether she also had tracheostomy.

Throughout the rest of her life, the patient had frequent ‘noisy breathing’ and dyspnoea, diagnosed as asthma. She would become dyspnoeic with light physical activity, such as sweeping the floor or climbing stairs, occasionally with presyncopal symptoms. These symptoms were greatly improved post-tracheostomy, suggesting that the glottic stenosis was the culprit, rather than asthma.

Investigations

A CT of the neck performed post-tracheostomy confirmed structural abnormalities to the glottis. These included abnormal morphology of the cricoid, which was ossified, with narrowing of the immediate subglottic space to 3 mm ventrally and 2 mm dorsally. There was ankylosis of the arytenoids bilaterally with medialisation of the vocal cords. This narrowed the lumen to 1.5 mm at the level of the vocal cords. There was mild prominence of the laryngeal ventricles. No mucosal lesion was identified. Similarly, there was calcification of the tracheal cartilage with abnormal morphology to the level of the tracheostomy.

The autoimmune screens (rheumatoid factor, C3/C4, antinuclear antibodies, antineutrophil cytoplasmic antibody) were negative.

Treatment

Although her airway was stable with the tracheostomy, the patient wanted her tracheostomy decannulated. This required definitive surgery on her narrowed glottis. Posterior laryngeal narrowing can be managed either through an endoscopic or open approach. The decision in this case was for open laryngotracheal reconstruction with a posterior cricoid rib graft.

Intraoperative findings included a heavily scarred and calcified cricoid cartilage, supporting the concern regarding previous airway intervention as an infant.

The posterior cricoid was split and rib cartilage graft spreader was inserted. This significantly improved the laryngeal aperture as shown below (see figure 2). A laryngeal stent was secured in place and kept in situ for 4 weeks with the tracheostomy.

Figure 2.

Intraoperative C-MAC video laryngoscope view of the larynx at the end of the laryngotracheal reconstruction procedure (A), and flexible nasoendoscopic view 6 months post-operatively (B) Increased space in the posterior glottis can be seen.

Outcome and follow-up

Repeat microlaryngoscopy and bronchoscopy (MLB) and removal of laryngeal stent at 4 weeks showed significant glottic opening. There was almost complete mucosalisation across the posterior laryngeal wall. Repeat MLB another 4 weeks later showed complete mucosalisation of the posterior graft. Capping and decannulation of the tracheostomy were successful over the next few days.

Discussion

A search through the available literature on PubMed revealed that although glottic stenosis as a result of endotracheal intubation is well documented, there is very little in the literature describing difficult intubation secondary to glottic stenosis and only one previous report of emergency tracheostomy for glottic stenosis in a 4-year-old boy.³

Glottic stenosis is either congenital or acquired.^{4 5} Congenital glottic stenosis is very uncommon^{4 5} and was not present for this patient, according to her medical records. Of the acquired cases, the majority are secondary to endotracheal intubation-induced injuries, as a reported 6%–19% of long-term intubated patients develop some form of laryngotracheal stenosis.3 4 6–9 Other causes include infective, traumatic (other than intubation injuries), neoplastic, and inflammatory causes, as well as vocal cord paralysis.^{5 6} In the case of this patient, it is difficult to determine the exact cause without a clear developmental history. It is most likely that this patient had congenital laryngeal stenosis. Surgical intervention as an infant did not correct her stenosis and the prolonged period of intubation was very likely a contributing factor.

The mechanism by which intubation causes glottic stenosis is thought to be through the pressure effect of the ETT on the posterior glottis, which causes ischaemia. The cricoid cartilage is a continuous ring and prevents outward release of oedema. This area is the narrowest in the paediatric airway. Mucosal necrosis and ulceration occur as a result, and perichondritis of the posterior laryngeal cartilages follows. Occasionally there is superimposed infection. The resulting fibrosis from the inflammation leads to contraction and scarring.4 5 8–10 Duration of intubation and size of ETT are the most important factors in the development of laryngeal stenosis. The period of intubation is less significant than the degree of ischaemic necrosis, as necrosis and ulceration can occur within 48–72 hours of intubation,¹⁰ and there are reports of patients developing glottic stenosis after only 3–5 days of intubation.^{3 8}

Another important factor in this case is that the symptoms the patient experienced pre-diagnosis were accepted as being asthma related. The available literature on misdiagnosis of laryngeal pathology as obstructive airway disease is scarce, with one other case report published by Krishna and Rosen in 2006 describing glottic stenosis misdiagnosed as chronic obstructive pulmonary disease.¹ Abnormal breathing noises and dyspnoea are common features of both laryngotracheal stenosis and obstructive lung disease; however, a thorough history and examination can indicate that laryngotracheal stenosis (including glottic stenosis) is the more likely culprit for the symptoms.¹ A history of upper airway conditions, trauma, surgery, intubations, or congenital abnormalities may hint at upper airway pathology. Poor response to standard treatment for obstructive lung disease can also suggest that there is another cause.¹ It appears that aspects of our patient’s medical history were lost as time passed, as more recent medical records had no mention of her neonatal history, which unfortunately meant that there was no suspicion of upper airway pathology during her pre-anaesthetic assessment.

Examination findings can also differentiate between obstructive lung disease and upper airway obstruction. Obstructive lung diseases cause lower airway obstruction, which results in wheezing and expiratory noises, opposed to the stridor of glottic stenosis and other laryngotracheal stenoses.⁵ Inspiratory stridor is typical of supraglottic or glottic obstruction, whereas biphasic stridor is more common at subglottic or tracheal levels of obstruction.⁵ Flexible fibre-optic endoscopy can confirm if there is upper airway obstruction and is the gold standard for glottic stenosis diagnosis.^{4 5}

Another learning point highlighted in this case is the importance of early recognition of a ‘can’t intubate, can’t ventilate’ situation and the appropriate response. Catastrophic anaesthetic airway events predominantly occur when airway difficulty is not realised before anaesthetic induction.¹¹ Fortunately these events are rare, but when they do occur and the attempts to intubate, to place a supraglottic airway device such as an LMA, or to ventilate with face mask have failed, brain damage and death are the results unless a rapid solution is found.¹² The Difficult Airway Society (DAS) revised the guidelines in 2015 for management of unanticipated difficult intubation based on findings from the Fourth National Audit Project in 2011.¹³ The DAS guidelines recommend preparation prior to every intubation for potential failed intubation, with four plans in place to follow in sequence if failure occurs.¹³

Plan A includes optimising the chances of a successful first attempt, and if it fails, to set a maximum number of three attempts (with a fourth allowed by a senior) to reduce airway trauma and the likelihood of a ‘can’t intubate, can’t ventilate’ situation.¹³ It is easy in this time of high stress to lose track of the time passing while attempting to secure the patient’s airway. Plan B involves maintaining oxygenation with a supraglottic airway device (such as an LMA) to provide time to reassess the situation and make a choice from the following options: wake the patient, reattempt intubation, continue anaesthesia without an ETT, or proceed straight to a surgical airway.¹³ If unable to ventilate via the supraglottic airway device, plan C is to ventilate via face mask and wake up the patient. If unable to ventilate via face mask, then plan D is to perform a scalpel cricothyroidotomy.¹³

Needle cricothyroidotomy, although preferred by many anaesthetists, has been shown to have a low success rate.¹³ If a surgeon experienced in tracheostomy is present and the situation is appropriate, then a tracheostomy may be performed instead. It is important that anaesthetic and surgical teams are aware of these recommendations, as often the delay before the decision to perform a surgical airway results in poor outcomes. In the USA, most cricothyroidotomies and tracheostomies were performed at periarrest or at death, or they were performed too late to prevent complications.¹³

Fortunately for the patient, her airway was secured before complications occurred. In hindsight, the attempt to wake the patient could have occurred sooner, and similarly, the otolaryngologist could have been contacted earlier, but otherwise, the anaesthetist followed the DAS guidelines closely.

There were many interesting points to this case, including a ‘can’t intubate, can’t ventilate’ situation due to glottic stenosis, as well as having to piece together the medical history from old records, collateral history, and clinical findings.

Learning points.

• Acquired glottic stenosis although rare can occur post-long-term intubation and should be considered in patients with upper respiratory symptoms.

• Causes other than obstructive lung diseases should be considered if the symptoms are not responsive to bronchodilator treatment. A thorough history and examination may reveal upper rather than lower airway pathology.

• It is critical to recognise ‘can’t intubate, can’t ventilate’ situations early and get appropriate help (i.e. early otolaryngology involvement if available).