Spinal Cord Infarction Following Central-Line Insertion

Abstract

Hemothorax is a recognized complication of central line insertion into the jugular or subclavian vein. We describe a case of hemothorax consequent upon acute dialysis catheter insertion, which resulted in spinal cord infarction and quadriplegia. We postulate that the extensive mediastinal shift induced after insertion of the catheter resulted in stretching of the veins draining the cord with a resultant drop in perfusion pressure and infarction. This case highlights a hitherto unreported complication of this procedure.

• Hemodialysis access
• Complication
• Paraplegia
• Hemothorax
• Spinal cord infarction

Introduction

The central venous route is frequently used as a means of hemodialysis access. Although the internal jugular vein is the site of first choice, complication rates of 6.2% have been reported in the acute setting.[[1]] The major complication is pneumothorax, although hemothorax, brachial plexus or thoracic duct damage, laryngeal hematoma and air embolus are also well recognized.[[2]] Ruptured superior thyroid artery[[3]] and cardiac tamponade have also been reported.[[4]] We describe here a case of spinal cord infarction as a consequence of dialysis catheter insertion, a hitherto unrecognized complication.

Case Report

A 50-year-old Chinese man with end-stage renal failure of unknown cause was admitted with peritoneal dialysis-associated peritonitis. There was a past history of low infectivity Hepatitis B carriage and he had no known macrovascular disease. It was necessary to remove the Tenchkoff catheter so temporary hemodialysis access was required. Left internal jugular dual-lumen dialysis catheter insertion was terminated because the patient developed chest pain and shortness of breath. A left-sided hemothorax was confirmed on chest X-ray, which was drained percutaneously. A temporary femoral dialysis catheter was placed and the patient made good clinical progress over the ensuing days.

However, ten days after the initial hemothorax, the patient became acutely distressed whilst being dialyzed. The dialysis was terminated and, on returning to the ward, he became hypotensive and unrousable. There was clinical evidence of a recurrent left hemothorax. Following colloid infusion, the blood pressure improved from 60/40 to 80/50 and he was transferred to intensive care. A chest X-ray confirmed that the entire left lung field was opaque with marked shift of the mediastinum to the right. As the patient was hemodynamically stable, after insertion of an intercostal drain, he was observed overnight. The following day an apical chest drain was also inserted on the left. However, there was minimal drainage of sero-sanguinous fluid from these drains and plans were made for formal evacuation at thoracotomy.

Later that evening, 26 h after the hypotensive event, the patient developed sudden onset acute flaccid paralysis and neck pain associated with sensory loss below the level of T4. Magnetic resonance imaging demonstrated that, below the level of T4/5, there was increased signal within the cord, which was ill defined on the T2 weighted image. The lower thoracic and lumbar cord was swollen and occupied the majority of the canal. No extra-axial collections were seen and there was no disk prolapse or other cause of cord compression. In addition, there was no evidence of aortic dissection. On the axial images, there were regions of increased signal within the anterior horns of the cord. It was concluded that the appearances were consistent with spinal cord infarction in an anterior spinal artery distribution. The patient subsequently underwent thoracotomy, evacuation of hematoma and repair of left subclavian vein.

Flaccid paraplegia persisted with little sensation below the chest. He has recovered some movements in his left foot. However, prognosis for further recovery is poor.[[5]]

Discussion

The anterior and paired posterior spinal arteries, primarily derived from the vertebral arteries, are fed by segmental vessels at different levels. A limited number supplies the anterior spinal artery. Usually, there are three or more vessels in the cervical and upper thoracic region, one in the mid-thoracic region and, caudal to this, a single large vessel, the artery of Adamkiewicz. The anterior and posterior spinal arteries give off branches that form a fine network around the spinal cord, which supply much of the white matter and the posterior horns of the grey matter. The venous drainage of the cord is similarly organized into interconnecting anterior and posterior systems that drain into the vena cava via perforating veins.

The anterior spinal artery syndrome is characterized by an abrupt onset of flaccid paraplegia or tetraplegia below the level of the lesion due to bilateral corticospinal tract damage. The etiology is usually related to impaired blood supply. Infarctions most commonly occur in the ‘watershed’ areas or boundary zones, where the distal arterial branches anastomose, at the levels of T1–T4 and at L1.[[6]]

We believe that two specific events led to the outcome in this case. Firstly, there was a period of hypotension on dialysis, a well-recognized cause of anterior spinal artery syndrome.[[5]] However, this was not the direct cause of the spinal cord infarction as there was a marked delay before the acute paraplegia occurred. It is probable that the vascular supply was further compromised on the venous side by marked shift of the mediastinum induced by the hemothorax. We postulate that, in a spinal cord already compromised by an episode of hypotension, stretching of the draining veins led to a critical drop in perfusion pressure with consequent infarction. There was no evidence of embolism (air or thrombus), arterial occlusion, aortic dissection or any of the other potential causes of spinal-cord ischemia. To our knowledge, this is the first reported case of such a complication following central insertion or hemothorax of any cause and it reinforces the potential for harm associated with placement of a catheter in one of the great veins.