Perforation of the pharynx or oesophagus during tracheal intubation is an uncommon complication that may occur when intubation is difficult. The consequences of pharyngo-oesophageal perforations include retropharyngeal abscess, mediastinitis, respiratory distress, subcutaneous emphysema, septic shock and death.
Retropharyngeal abscess occurs in the retropharyngeal space, which lies between the pharynx and the posterior layer of the deep fascia that bound the prevertebral region. The space extends from the base of the skull to the posterior mediastinum at the level of the tracheal bifurcation. In children, this space contains lymph nodes, which can become infected during an upper respiratory tract infection with abscess formation. In adults, these lymph nodes atrophy , and an abscess may be either acute or chronic.
The acute abscess usually follows trauma to the posterior pharyngeal wall such as a penetrating foreign body, intubation or oesophagoscopy. It may also result from spread of infection from the ears, the sinuses, throat or teeth and, therefore, an underlying illness should be sought .
Chronic retropharyngeal abscess is usually caused by tuberculosis of the cervical spine and is a rare condition.
Lateral X-ray of the neck is usually diagnostic, but widening of the retropharyngeal space can also be caused by retropharyngeal cellulitis or oedema and, for a differential diagnosis, a computerized tomography (CT) scan is often very helpful .
A 42-year-old man was admitted to the Ear, Nose and Throat department as an emergency with sore throat, tightness of the neck and complete dysphagia, having undergone an elective umbilical hernia repair 3 weeks previously at another hospital. The patient complained of a sore throat from the very first post-operative day, which was treated with lozenges and gargles. There had been no mention of difficulty in intubation. The patient was discharged with sore throat on the fifth post-operative day. A course of antibiotics from his general practitioner improved the symptoms but, on discontinuation of the treatment, the symptoms flared up leading to complete dysphagia. On admission, although the patient exhibited complete dysphagia and was unable to swallow saliva, no sign of respiratory compromise was present. Indirect laryngoscopy revealed normal larynx and bulging of the posterior pharyngeal wall. The soft-tissue, X-ray film of the neck (see Fig. 1) showed a widening of the retropharyngeal space confirming the diagnosis of retropharyngeal abscess. The abscess was drained intraorally and a nasogastric tube inserted under general anaesthesia for enteral nutrition. The post-operative course was uneventful, and the patient was discharged on the 6th post-operative day.
A preoperative examination revealed an anxious and moderately obese 91-kg patient with no respiratory distress. Assessment of the airway showed a full range of neck movement, Mallampati class I, and mouth opening to three fingers. However, indirect laryngoscopy had shown a normal larynx, but there was bulging of the posterior pharyngeal wall, indicating possible difficulty in airway management. The patient was haemodynamically stable with a slight tachycardia and a leucocytosis of 19.9 × 109 L−1.
No premedication was given. After establishing standard monitoring and venous access, the patient was placed in the left lateral and head down position. After preoxygenation, atropine 0.6 mg was administered to reduce salivary secretion. Inhalation induction was carried out with a mixture of oxygen, nitrous oxide and halothane. After confirming the ability to ventilate, suxamethonium was administered to permit laryngoscopy. Laryngoscopy was grade I, and intubation with an endotracheal tube 9.0 mm RAE cuffed was straightforward. The correct placement of the tube was confirmed by auscultation over both lung fields and capnograph. Anaesthesia was continued with fentanyl, atracurium, halothane, oxygen and nitrous oxide. The ETCO2 was maintained at 4.7 kPa with a haemoglobin oxygen saturation of 95%. Dexamethasone was administered to reduce post-operative oedema. At the end of the procedure, the patient was reversed with glycopyrrolate and neostigmine. Awake extubation in the head down position was accomplished. The patient was discharged to the ward with humidified oxygen per mask. The post-operative course was uneventful, and the patient was discharged on the 6th post-operative day.
It is probable that the patient suffered a pharyngeal perforation during attempts at tracheal intubation 3 weeks previously for the elective procedure of umbilical hernia repair. This gave rise to a retropharyngeal abscess leading to complete dysphagia.
There are no cases reported in the medical literature, as far as we are aware (since 1966), about the anaesthetic management of retropharyngeal abscess. Anticipated problems associated with retropharyngeal abscess include difficult airway management, oedema of the airways, rupture of the abscess, regurgitation and aspiration.
Options available include tracheostomy under local anaesthesia, awake intubation or intubation under general anaesthesia. Our patient had a normal preoperative airway assessment and, therefore, intubation under general anaesthesia was our method of choice. Inhalation anaesthesia with the patient breathing spontaneously is perceived to be safer than intravenous induction, because the anaesthesia can be more readily discontinued if airway obstruction should occur.
The classical approach described by Tunstall and Sheik  with the patient in the left lateral and head down position was thought to be an ideal position for induction, because the left lateral position allows the anaesthetist to confirm that the mask is correctly placed, and it is an ideal recovery position in case of a failed intubation. Head down tilt also limits the possibility of aspiration in the case of abscess rupture allowing the contents to drain into the mask or onto the floor; likewise regurgitated matter from the stomach.
This approach has been questioned recently, and a proposed modification is to leave the patient in the supine position without head down tilt. The major concern is that the anaesthetists are not normally taught to maintain an airway and to oxygenate a patient in the left lateral position.
Halothane was the volatile agent of choice rather than sevoflurane, because of its established efficacy and being less respiratory depressant than sevoflurane. Halothane provides a gradual onset of anaesthesia such that, if the airway begins to deteriorate, the volatile agent can be discontinued and the patient allowed to awaken. If a fairly deep anaesthesia has been obtained before the airway deteriorates, awaking is then caused primarily by the redistribution of the agent from the blood to the tissues. In both instances, the lower blood/gas and tissue/blood solubilities of sevoflurane make the agent less suitable than halothane [5,6].
Dexamethasone was administered in order to reduce the oedema of the pharynx and the laryngobronchial tree, together with the post-operative administration of humidified oxygen to prevent delayed post-extubation complications.
Alternative strategies considered were tracheostomy and fibre optic intubation. Both procedures require to be performed under local anaesthesia and are more distressing for the patient.
Tracheostomy was considered in the case of failed intubation, the ENT surgeon being available. It bypasses the abscess in the upper airway and, once the tracheostomy tube is in situ, the abscess can be drained under general anaesthesia. It is a separate procedure with long- and short-term complications. The complications are  haemorrhage, tracheal injury, surgical emphysema, tracheitis, pneumothorax and laryngeal oedema in children.
Fibre optic intubation was our next strategy. This requires local anaesthesia, and one of the major problems of introducing local anaesthetic solutions into the respiratory tract is that foreign material, for example gastric contents, can enter the lungs. However, Ovassapian and colleagues , following a small series, recorded their experiences. They argue that, if the patient is awake and the upper airway anaesthetized, the bronchial reflexes are still retained, together with their ability to expel foreign materials.
Our patient was not an optimal candidate for this procedure because, as he had complete dysphagia and was unable to swallow his saliva, this may have led to fogging of the lens and an obscured view. It demands expertise on the part of the operator, and there is also a real danger of rupturing the abscess at endoscopy and completely losing control of the airway. Other disadvantages are that subsequent passage of the tube over the fibre optic scope can be difficult and, even in the most experienced hands, an intubation using this technique can be prolonged compared with conventional methods [9,10].
The management of this case demonstrates that well-established techniques such as inhalational induction still have an important role in the armamentorium of the anaesthetist.
We acknowledge Dr J. Aslam, consultant anaesthetist, for his help in providing guidance and support.
1 Adams DA, Cinnamond MJ. Paediatric Otolaryngology. Scott-Brown's Otolaryngology,
6th edn. Bath: Butterworth-Heinemann International Editors, 1997.
2 Barratt GE, Koopman CF, Couthard SW. Retropharyngeal abscess
. A 10 year experience. Laryngoscope
3 Goldenberg D, Golz A, Jochims HZ. Retropharyngeal abscess
; a clinical review. J Laryngol Otol
4 Tunstall MF, Sheik A. Failed intubation protocol: oxygenation without aspiration. In: Ostheimer GW, ed. Clinics in Anaesthesiology, Obstetric Analgesia and Anaesthesiology I,
Vol. 4. Philadelphia: WB Saunders, 1986: 171-187.
5 Davies MW. Sevoflurane: a note of precaution (letter). Anaesthesia
6 Fenlon S, Pearce A. Sevoflurane induction and difficult airway management
7 Atkinson RS, Respiratory Failure. In: Atkinson, RS, Rushman GB, Davies NJH, eds. Lees Synopsis of Anaesthesia.
Oxford: Butterworth-Heinemann, 1993: 807-810.
8 Ovassapian A, Krejcie TC, Yelich SJ, Dykes MHM. Awake fibreoptic intubation of the patient at high risk for aspiration. Br J Anaesth
9 Smith JE, Mackenzie AA, Scott-Knight VCE. Comparison of two methods of fibrescope-guided tracheal intubation. Br J Anaesth
10 Schaefer HG, Marsch SCU. Comparison of orthodox with fibreoptic orotracheal intubation under total intra-venous anaesthesia. Br J Anaesth