Failure of the Laryngeal Mask Airway

The laryngeal mask airway (LMA) is useful in the management of the difficult airway [1]. There are no previous reports of failure to correctly insert or ventilate through the LMA in a "cannot ventilate, cannot intubate" scenario [2]. We report the failure of the LMA to ventilate a patient with an unrecognized difficult airway presenting for emergency surgery. Case Report A 67-yr-old, 72-kg man presented for completion of a traumatic partial amputation of the arm. He was 175 cm tall and appeared his stated age. His medical history was significant for heavy smoking and a colectomy with a resultant colostomy. Preoperatively, he was conversant, with no vocal abnormalities, orthopnea, or positional effects on spontaneous ventilation, and he was hemodynamically stable. Physical examination revealed a Mallampati class I airway, normal head and neck, and good range of motion at the temporomandibular joints and cervical spine. Routine laboratory tests and electrocardiogram were within normal limits. The patient refused regional anesthesia; therefore, a general anesthetic was planned. A rapid-sequence induction using etomidate (0.2 mg/kg) and succinylcholine (2 mg/kg) IV was performed. Cricoid pressure was applied. Laryngoscopy with a Macintosh 3 blade performed by a third-year anesthesiology resident revealed a vascular supraglottic mass 1-2 cm in diameter below a normal epiglottis, obscuring the vocal cords. Two unsuccessful attempts were made to pass 8-mm and 7-mm endotracheal tubes (ETT) using a malleable stylet. Bag-mask ventilation was also unsuccessful. The patient's SpO (2) decreased to 85% and a No. 3 LMA was inserted using the conventional technique, but ventilation was still impossible. The SpO2 decreased further to 50%. A No. 4 LMA was similarly inserted, but ventilation was still unsuccessful. Further attempts at bag-mask ventilation were unsuccessful, and needle cricothyroid puncture was performed with a 16-gauge IV catheter to allow high-pressure jet ventilation, with no significant improvement in SpO2. An emergency tracheostomy was then performed via the initial insertion site. The SpO2 decreased to <50% with the patient remaining hemodynamically stable. A 6-mm ETT was inserted via the tracheostomy approximately 4 min after the induction of anesthesia. However, no expired CO2 was detected by capnography, despite an adequate blood pressure (measured via a radial arterial catheter). At this time, the SpO2 increased to 50% and the patient made some spontaneous respiratory efforts. Breath sounds were not detectable with continued bag ventilation via the ETT, and the colostomy bag was inflating. A chest radiograph was obtained but proved unremarkable. We suspected that the ETT was incorrectly situated and had been passed supraglottically through the tracheostomy into the esophagus. A second tracheostomy was immediately performed. At this point, the patient became profoundly hypotensive with no palpable pulses; cardiopulmonary resuscitation was initiated. Over the next 2 min, 3 mg of epinephrine and 2 mg of atropine were administered. A second 6-mm ETT was placed via the second tracheostomy with immediate detection of expired CO (2). Shortly after the establishment of a secure airway, the arterial blood pressure returned to normal and cardiopul-monary resuscitation was discontinued. At this point, it was decided to proceed with the scheduled surgery, despite our being unaware of the patient's current neurological status. Anesthesia was maintained with isoflurane in 100% O2, rocuronium, and fentanyl. Mild hyperventilation was initiated, and the maintenance anesthetic was titrated to achieve a mean arterial blood pressure greater than 70 mm Hg. After the originally scheduled orthopedic procedure, the ear, nose, and throat service formalized the tracheotomy site and placed a No. 6 Shiley tracheostomy tube. On endoscopy, a large supraglottic mass was seen. The larynx was in an inferior position, with the cricoid cartilage close to the level of the sternal notch. There was no evidence of a tracheoesophageal fistula or esophageal pathology. The total operative time was 4 h. Biopsy samples of the mass showed an invasive squamous cell carcinoma arising from the left vocal cord. Postoperative examination detected no neurological deficits. Subsequent computed tomography scans of the neck (Figure 1) showed an extensive supraglottic tumor invading the anterior neck musculature. Two days after his initial presentation, the patient underwent a total laryngectomy and bilateral radical neck dissection. Three days later, he underwent an uneventful regional anesthetic for debridement and definitive wound closure of the amputation site. He made an uneventful recovery and was discharged home 4 days later.Figure 1: Computed tomography scan at the level of the fifth to sixth cervical vertebrae indicates extensive tumor invasion of the glottis and anterior neck. The arrow indicates the narrow, deviated airway.Discussion Our patient presented with no symptoms of his laryngeal tumor, which placed him at significant risk of difficulty in the establishment of an airway. We believe that this case demonstrates three primary issues: 1) The inability to ventilate via the LMA when a supraglottic mass causes a "ball-valve" obstruction in an anesthetized patient. Although the LMA is contraindicated or may be unsuccessful with airway obstruction (e.g., tumor, stenosis, mucus plug, bronchospasm) distal to the laryngeal opening, in our case, it was reasonable to try the LMA briefly in the failed airway drill because anatomical airway obstruction was not considered a possibility. One previous report demonstrated successful ventilation with the LMA as a temporary measure in a patient with laryngeal papillomas after failure of bag-mask ventilation [3]. This case emphasizes the need for multiple options in airway management as described in the ASA failed airway algorithm. 2) The failure to detect CO2 via capnography after the surgeons had assured us that they had successfully obtained an airway points to the importance of careful follow-up when clinical difficulties arise in emergency situations. 3) The use of succinylcholine for the initial neuromuscular blockade as part of the rapid sequence induction may have contributed to the favorable outcome in this case, as the patient was able to breathe through a partially obstructed airway in the interval between the first and second tracheostomies. Recommendations that succinylcholine should not be routinely administered as part of a formal rapid sequence induction should not be discouraged until a fast-onset, short-acting, nondepolarizing muscle relaxant is available. The authors recognize the efforts of Paul Morrow, MD, for his assistance in the resuscitation and clinical management of this patient.