The presence of air in the epidural space as a result of multiorgan trauma, was first described by Gordon and colleagues in 1977 [1]. Ten years later, the term “pneumorrhachis” was introduced by Newbold [2].

At present, the causes of air presence in the extradural space are divided into iatrogenic, non-traumatic and traumatic. The iatrogenic causes include spinal surgery and extradural anaesthesia, particularly when the space is identified using the loss of resistance technique with a syringe filled with air. The non-traumatic causes are strong cough fits during bronchial asthma, spastic bronchitis (13 cases described) or aspiration of a foreign body to the airway (1 case) [3]. Extradural pneumatocele was also described as a complication of resuscitation procedures (1 case), persistent vomiting (3 cases) or as a result of inhalation intoxication with narcotics (“Ecstasy, hemp). The third group of causes is constituted by trauma to the head, neck, thorax, abdomen and pelvis. 

Except for the mechanism of direct trauma, the prerequisite of air penetration to the vertebral canal is the increased-pressure space formed in the vicinity of the spine. The sources of gas are the organs usually filled with it: the lungs, paranasal sinuses, gastrointestinal tract [4]. The most common causes in this group include pneumothorax [5], mediastinal emphysema [6], and trauma to the pharynx, oesophagus or intestines combined with pleural, peritoneal or spinal meningeal trauma. Gas may also permeate from the degenerated nucleus polposus of the intervertebral disc, which was described by Magnusson in 1937 [4].

The aim of the present report was to describe the case of pneumorrhachis resulting from injuries sustained during a road traffic accident.

CASE REPORT

A 24-year-old male driver of a passenger car sustained multiorgan trauma due to crashing into the tree. According to the emergency medical team, the patient was conscious, excited, intoxicated with alcohol; the respiratory and circulatory systems were efficient. At the accident site, spinal trauma and paresis of lower limbs were found.

On admission, the patient was still conscious, although occasionally excited, without features of impaired respiratory function; cardiovascular parameters were within normal limits. The CT scan demonstrated generalized moderate cerebral oedema, contusion of both lungs, rib VII fracture on the left side and haematoma in both pleural cavities. Moreover, the comminuted fracture of Th7, fractures of the Th8 diathesis and spinous process as well as  Th9 right transverse process were detected. The bony lesions were accompanied by features of massive trauma and disruption of the spinal cord in the region corresponding to bony injuries.  The neurological examination confirmed flaccid paralysis of lower limbs.

The standard therapy with methylprednisolone was instituted and the patient qualified for urgent surgical stabilization of the thoracic spine and possible decompression of the spinal cord. Before anaesthesia, the drain was introduced to the right pleural cavity and the haematoma evacuated. The thoracic spine was repositioned and stabilized using the transpedicular fixation method with the XIA set under combined general anaesthesia. Intraoperatively, extensive spinal injuries with dural mater disruption and liquorrhoea were confirmed. Moreover, massive injuries to the paraspinal tissues were observed. The course of anaesthesia and surgery was uneventful.

Immediately after anaesthesia, the patient was admitted to the ITU, where SIMV was initiated with the pressure of 20 cm H2O (2 kPa), at FIO2 0.6. Under CVP guidance, the intravascular volume was supplemented and, due to persisting hypotension (SAP/DAP 85/45 mmHg) and impaired tissue function, the infusion of dobutamine included. Steroids were continued according to the protocol of spinal trauma management; antibiotic (cephamandol) and ulcer prophylaxis were administered. Additionally, the gastric probe was inserted to deliver food; no features of upper gastrointestinal bleeding were observed.

On hospitalization day 2, abnormalities of lung gas exchange increased; FIO2 was increased to 0.8. Temperature was elevated, on day 5 > 38°C; periodic chills were observed. The blood was sampled for microbiological tests. Moreover, WBC – 16.5 G L-1 and procalcitonine level – 2.2 ng mL-1 were found increased.

On day 5, the pleural drain and obstructed wound drains were removed. On the following day, the inflammatory infiltration was found within the surgical wound and surrounding tissues. The fluid escaping though the skin sutures suggested persisting liquorrhoea. On day 7, the results of microbiological tests of blood and wound swab were obtained demonstrating the presence of Acinetobacter baumannii. Antibiotic therapy was widened with meropenem and ampicillin with sulbactam. Despite this, the patient`s general condition deteriorated: features of generalized inflammatory reaction and severe imbalance of tissues perfusion fulfilling the criteria of septic shock were observed. To maintain vascular peripheral resistance, noradrenaline was started. The concentration of procalcitonine was 151.8 ng mL-1. Additionally, meningeal symptoms were found in the form of nuchal rigidity.

On day 8, in the morning, the endotracheal tube dislocated to the throat. Re-intubation was difficult due to substantial obesity and oedema of the oral and pharyngeal soft tissues. Some part of the respiratory volume delivered by the ventilator leaked to the oesophagus and stomach, from where it was evaluated through the gastric probe. Finally, the endotracheal tube was reintroduced and proper lung ventilation restored. The lowest percutaneous blood saturation at that time was 88%.

Due to persisting symptoms of septic shock, poor healing of the wound and positive culture from the wound material, CT was repeated, which evidenced numerous gas collections with blurred surrounding structures within the paravertebral, thoracic and retromediastinal soft tissues (Fig. 1, 2). Gas collections were also found in the spinal canal, intracerebrally and intracranially (Fig. 3, 4).

The patient was urgently qualified for wound revision during which the kit stabilizing the spinal fracture and necrotic tissue lesions was removed. The tissue material was collected and sent for microbiological tests, which 36 h after surgery demonstrated massive increase in Enterobacter cloace. Ciprofloxacin was added. Twenty hours after surgery, repeated CT of the head, cervical and thoracic spine disclosed total absorption of gas collections within the cranium and spinal canal.

Based on the clinical picture, CT and microbiological results (wound infection with the gastrointestinal flora), the oesophageal fistula was suspected. According to gastroenterologist’s suggestions, esophagoscopy was abandoned. The diagnosis was based on the clinical features and conservative treatment instituted with proton pump inhibitors, targeted antibiotic therapy and total parenteral feeding.

During the following days, the general condition of the patient gradually improved. After successive 36 h, percutaneous tracheostomy was performed. Artificial lung ventilation was continued for the further 48 h. Setons filling the wound were changed every 12 h, and the wound rinsed with 10% polyvidon solution and 10% NaCl solution, which resulted in wound healing by granulation. During hospitalization in the ITU, oral feeding was also started, which was tolerated well.

On day 28, the patient was transferred to the unit of orthopaedics and traumatology for further treatment; he was conscious, movements of both upper limbs were dexterous; however, flaccid paralysis of lower limbs persisted.  

DISCUSSION

It is believed that the presence of gas within the spinal canal per se is not rare yet does not require any therapeutic actions. For instance, such cases are observed during extradural anaesthesia using the loss of resistance technique when the space is identified with the syringe filled with air. However, in traumatic cases the presence of gas within the spinal canal is an absolute indication for determining the pathomechanism of the phenomenon [7]. 

Depending on the location of gas collections, pneumorrhachis may be divided into epidural or subarachnoid when the visceral lamina of the dura mater is injured. Cephalad shift of gas within the epidural space is limited by the anatomy of the dural sac. In general, gas from the epidural space is spontaneously absorbed and does not induce any significant clinical symptoms [8]. However, once the dural barrier loses its tightness gas may freely permeate the entire CNS fluid space, including intracranial, leading to pneumocephalus. The most common causes of the presence of gas within the cranium are traumas resulting in bony tectum fractures, usually involving the paranasal sinuses or air cells of the mastoid processes. In the available literature there are single descriptions of cases in which the amount of gas in the cranium was substantial resulting in brain compression and more or less pronounced neurologic symptoms [9].

In our case, the presence of gas in the spinal canal, cranium and CNS fluid spaces was demonstrated on the 8th post-traumatic day; therefore, trauma, as the direct cause of this pathology, may be excluded. Moreover, in the period immediately preceding its development pneumothorax was not observed although the patient’s lungs were ventilated for 8 days. Due to the shift of the endotracheal tube from the trachea to the throat during the morning hygiene, part of respiratory volume was distributed to the lungs (thanks to which relatively proper respiratory exchange was maintained); the remaining volume was likely to penetrate the throat and oesophagus. It was assumed that the respiratory gas mixture leaked through the injured oesophageal wall to the posterior mediastinum and subsequently to the spinal canal. Disruption of the visceral layer of dura mater enabled the penetration of gases to the subarachnoid space. The pressure of gas delivered with every inspiration generated by the ventilator was constantly higher than that in the surrounding tissues and cerebrospinal fluid, which continuously increased pneumocephalus. The cephalad shift of gas bubbles was favoured by the patient`s position with the headrest elevated by 30°.

At that time, the essential clinical problem was sepsis resulting from infection of  surgical wound or/and the stabilizing kit. The initial hypothesis regarding etiopathogenesis of CT-detected lesions was massive infection of the tissues surrounding the injured spine whereas gas collections were associated with the bacterial process [10]. The detection of bacterial flora in all tissue fragments and fluids collected from the wound (Enterobacter cloacae) was conclusive for oesophageal fistula. The solution was complete isolation of airways by endotracheal intubation combined with targeted antibiotic therapy.

Posttraumatic oesophageal injuries are difficult to diagnose [11]; the only reliable method is esophagoscopy. In the case described, this examination was abandoned due to the risk of additional iatrogenic oesophageal damage and high probability of establishing proper diagnosis based on the clinical symptoms. In the initial hospitalization period no symptoms indicating oesophageal damage were observed. The conclusive symptom was the infection of the mediastinum, paravertebral tissues and vertebra [12, 13]. Despite delayed diagnosis [14] and unfavourable initial clinical course (spread of infection in the posterior mediastinum and surgical site, which ultimately led to septic shock), the general condition of our patient improved, multiorgan dysfunction gradually subsided and the wound healed. The patient was discharged from ITU in good general condition for further therapy. Based on the retrospective analysis, we can assume that conservative treatment used in this case [15] proved effective. 

..............................................................................................................................................................

REFERENCES

1. Gordon IJ, Hardman DR: The traumatic pneumomyelogram. A previously undescribed entity. Neuroradiology 1977; 13: 107-108.

2. Newbold RG, Wiener MD, Vogler JB: Traumatic pneumorrhachis. AJR Am J Roentgenol 1987; 148: 615-616.

3. Oertel MF, Korinth MC, Reines MHT, Krings T, Terbeck S, Gilsbach JM: Pathogenesis, diagnosis and management of pneumorrhachis. Eur Spine J 2006; 15: S636-S643.

4. McIntosch SE, Salcedo-Dovi H, Cortrs V: Air in the canal associated with trauma. J Emerg Med 2006; 31: 33-35.

5. Arslan G, Yarar G, Sindel T: Epidural emphysema associated with spontaneous pneumothorax. Eur J Radiol  2005; 54: 11-13.

6. Koelliker P, D., Brannam L, A: Epidural pneumatosis associated with spontaneous pneumomediastinum: case report and review of the literature. J Emerg Med 1999; 17: 247-250.

7. Sahin M, Cobanpglu M, Kilicaslan I: Entire pneumorrhachis due to isolated head trauma. Am J Emerg Med 2009; 27: 902.e3-902.e6.

8. Akay S, Bayram B: Traumatic pneumorrhachis: a rare entity of trauma. Int J Emerg Med 2008; 1: 53.

9. Kapoor T, Shetty P: Pneumocephalus. J Emerg Med 2008; 35: 453-454.

10. Thompson G, R, Crawford G, E: Pneumorachis caused by metastatic gas gangrene. Diagn Microbiol Infect Dis 2009; 63: 108-110.

11. Akman C, Kantarci F, Cetinkaya S: Imaging in mediastinitis: a systematic review based on aetiology. Clin Radiol 2004; 59: 573-585.

12. Metcalfe S, Morgan-Hough C: Cervical epidural abscess and vertebral osteomyelitis following non-traumatic oesophageal rupture: a case report and discussion. Eur Spine J 2009; 18: S224-S227.

13. Bernard A,W., Ben-David K, Pritts T: Delayed presentation of thoracic esophageal perforation after blunt trauma. J Emerg Med 2008; 34: 49-53.

14. Muir AD, White J, McGuigan JA, McManus KG, Graham AN: Treatment and outcomes of oesophageal perforation in a tertiary referral Centre. Eur J Cardiothoracic Surg 2003; 23: 799-804.

15. Hinojar AG, Rodrigues Castejon MA, Honojar AA: Conservative management of  a case of cervical  esophagus perforation with mediastinal abscess and bilateral pleural effusion. Auris Nasus Larynx 2002; 29: 199-201.

..............................................................................................................................................................

address:

*Jarosław Wośko
Klinika Anestezjologii i Intensywnej Terapii
Uniwersytet Medyczny w Lublinie
ul. Jaczewskiego 8, 20-950 Lublin
tel.: 81 724 43 32
e-mail: anest@umlub.pl

received: 12.07.2010
accepted: 21.11.2010