The use of the stomach has become a well established method of reconstruction of the alimentary tract after total laryngopharyngoesophagectomy (TLPE). The method of extrathoracic extraction of the esophagus by transhiatal blunt finger dissection has become the technique commonly employed for performing TLPE with gastric pull-up. The resection part of this procedure has been blamed as likely to do more damage than the reconstruction, ¹ especially while doing blunt finger dissection in the mediastinum to extract the thoracic esophagus. Blunt finger dissection in the mediastinum to mobilize and extract the esophagus has been reported to produce serious complications varying from pleural effusion, pneumothorax to catastrophic tracheal tears.²⁻⁶ This adds to the morbidity of the procedure, as it is very often performed in patients who are in poor general health.

Though pneumothorax in these procedures has previously been reported,the development of intraoperative pneumomediastinum has not been reported before.

A 40-year-old, 50 kg, ASA (American Society of Anesthesiologists) I man was  scheduled  for  transhiatal  esophagectomy.  His  preoperative  blood  in-vestigations and physical examination were within normal limits. Anesthesia was  induced  with  fentanyl  and  propofol  and  the  trachea  was  intubated  after muscle relaxation was achieved with vecuronium. Surgery proceeded uneventfully. For extrathoracic extraction of the esophagus, blunt dissection of the esophagus in the mediastinum was employed, which lasted for about 4 minutes. At this time, the patient’s lungs were ventilated manually with 100% oxygen. At the completion of blunt dissection, the invasive arterial blood pressure tracing was noted to have decreased to about 0−30 mmHg. The electrocardiogram (ECG) tracing was of low amplitude but normal mor-phology  with  few  premature  atrial  ectopics.  Radial  arterial  pulse  was  hardly palpable with only a few occasional palpable beats (1−2/min). Peak airway pressure continued to remain at 15−17 cmH2O, and there was no drop in oxygen saturation.

The surgeons were informed and bilateral chest tubes were inserted. This resulted in an immediate increase in the invasive blood pressure with a nor-mal tracing and a palpable radial pulse. The ECG tracing also became normal. The rest of the opera-tion  proceeded  uneventfully.  Postoperative  chest  X-ray was normal.

The clinical findings of pneumomediastinum are subcutaneous air (not pathognomonic), and subcu-taneous  crepitations  suggestive  of  free  air  in  the  thoracic  cavity.  Hamman’s  sign  means  precordial  systolic crepitations and diminution of heart sounds. Associated pneumothorax is suspected in individu-als  with  a  sudden  increase  in  peak  airway  pres-sure, asymmetry of breath sounds, and hypoxemia. Diagnosis is confirmed by chest radiography, which reveals air within the mediastinal space and which may coexist with pneumothorax or pneumopericar-dium. Chest computed tomography is more sensi-tive than X-ray. Contrast radiography is mandatory in suspected esophageal perforation. ECG is of de-creased  voltage,  with  ST  segment  depression  and  nonspecific T-wave changes.

Treatment  of  pneumomediastinum  is  conser-vative,  with  supplementation  of  100%  oxygen.  A  minimal quantity of air is absorbed spontaneously. Mechanical ventilation is indicated if there is asso-ciated respiratory compromise. Mediastinoscopy is performed in life-threatening pneumomediastinum and percutaneous mediastinal drain is indicated in tension pneumomediastinum.

The most probable diagnosis in our patient was pneumomediastinum as his ECG was of low ampli-tude with impalpable pulses. Also, there was no in-crease in airway pressure and no drop in saturation, which were suggestive of pneumothorax. The likely explanation for the development of pneumomedi-astinum in our patient is that the blunt dissection during surgery caused hiatal opening, and putting he hand in the thoracic cavity caused alveoli rup-ture. Alveoli rupture in turn could have caused air dissection  along  the  bronchovascular  sheath,  al-lowing free air to reach the mediastinum. Air may also leak via tissue plane through the sternocostal diaphragm  attachment,  causing  pneumopericar-dium, pneumothorax or subcutaneous emphysema. Although the thoracic cavity is open to the atmo-sphere, its exposure seldom results in tension pneu-mothorax or pneumomediastinum. However, if air is not able to escape, it may cause pneumomediasti-num or pneumothorax.⁷

We could not confirm our diagnosis of pneumo-mediastinum  with  other  modalities  like  magnetic  resonance  imaging,  ultrasonography  and  contrast  radiography; however, the clinical features of the case point towards this as being the most likely di-agnosis. Prompt diagnosis and immediate correction are the keys to avoiding catastrophic sequelae.