Introduction

Several surgical techniques including hemilaminectomy, pediculectomy, foraminotomy and dorsal laminectomy have been advocated for decompression of the thoracolumbar spinal cord affected by disc disease. Decision for one of those is based on: lesion localization along the spine, the location of compressing material with regard to the spinal cord, and the surgeon's personal preference. Decompression of mainly ventrally located lesions is technically more demanding, since compression cannot be resolved without significant spinal cord manipulation if approached by one of the surgeries listed above. In those cases the surgeon has to walk a fine line between complete decompression and minimizing iatrogenic spinal cord trauma.

The ideal surgical approach for ventrally or ventrolaterally located material would involve complete removal of disc material, with minimal manipulation of the spinal cord, causing minimal spinal instability, as well as preventing recurrence of symptoms at the surgery site. The thoracolumbar lateral partial corpectomy (TLLPC) as described by Moissonnier et al. 2004 fulfils most of those requirements. The authors of that original paper defined TLLPC as "creation of a lateral slot through the vertebral epiphysis of 2 adjacent vertebral bodies and the intervertebral disc".

At the Department of Small Animal Medicine of the University of Leipzig, we have investigated the feasibility of the TLLPC in intervertebral disc disease and have studied in our hospital setting patient's prognosis and possible complications.

Surgical Procedure Description

For surgery, the patient is placed in lateral recumbency. The side depended on the degree of lateralization of disc protrusion/extrusion or it was arbitrarily chosen due to the surgeon's preference in cases with pure medial ventral compression. Care was taken to achieve exact lateral recumbency and the position was checked using fluoroscopy, since orientation during the procedure seemed to be more difficult than in other types of thoracolumbar spinal surgery. After surgical preparation a paramedian skin incision was made lateral of the palpated transverse processes in the lumbar spine or at the level of the suspected spinal rib attachments in the thoracic spine.

The lateral aspect of the annulus fibrosus and of the adjacent vertebral bodies reaching dorsally from the accessory process to the ventral border of the vertebral body was exposed. The spinal nerve and associated blood vessels were separated and gently retracted cranially using a silicon vessel loop. For disc disease at the intervertebral spaces Th11-12 and Th12-13, sometimes rib luxation and resection of about 2 to 4 cm of the proximal rib (depending on patient size) was required in order to allow horizontal drilling. This was invariably the case in disc disease cranial to Th11. The final size of the created slot largely depended on the extension of the spinal cord compression as assessed on a CT myelography. The aim was complete decompression of the spinal cord. The dorsal margin of the slot did correspond with the level of the floor of the spinal canal, whereas the ventral margin was chosen to create a slot large enough for insertion of small hooks for removal of disc material. However, care was taken not to extend to the following margins whenever possible: craniocaudally: maximum 1/3 of the vertebral body length into each vertebra; dorsoventrally: maximum ½ of the vertebral body height; depth: maximum: 2/3 of the vertebral body width.

Drilling was started at the ventral margin and then extended dorsally until the spinal canal was reached. The final shelf of bone was removed using a diamond burr and the longitudinal ligament was resected. The protruded type II disc material was gently extracted using rongeurs, whereas extruded type I disc material was removed using a blunt hook and forceps. Decompression was considered complete when no further material could be removed and when the spinal cord could be visualized crossing above the corpectomy slot without being elevated from the floor of the vertebral canal.

Results

In a case series of 51 dogs with 60 corpectomies, surgical success, slot morphology and complications were evaluated. Decompression was considered complete or good in 90% of cases based on post-surgical CT myelography. Slot morphology was as follows: mean height: 43% of vertebral body height; mean depth: 64.1% of vertebral body width; mean cranial extension: 29.5% of the vertebral body length; median caudal extension: 22.0% of the caudal vertebral body length.

The following complications were observed with the following frequency: pneumothorax: 33% (did not require intervention); significant venous sinus bleeding: 25%; inadvertent laceration of a spinal nerve: 12%; and dural laceration: 5%. Vertebral venous sinus haemorrhage is the most significant complication. Such bleeding seems to be unavoidable considering the approach, but was encountered only in about one quarter of those dogs. In most cases, it can be stopped by padding the lesion with a gelatine sponge. Alternatively, drilling has to progress under continuous suction, but care has to be taken not to underestimate the amount of blood removed, since uncontrollable haemorrhage can become life threatening.

In a smaller case series of 36 dogs, neurological outcome was evaluated. The post-surgical neurological examination at 24-48 hours revealed an improved neurological status in 12/36 dogs, an unchanged status in 18 dogs, whereas initial post surgical neurological deterioration was seen in 6 dogs. At a recheck examination one month later, 19 dogs were neurologically normal, 14 were improved at least one grade, 2 were unchanged and 1 dog was still worse than when initially presented.

Conclusions

In conclusion, the thoracolumbar lateral partial corpectomy is a way to approach ventral spinal cord compression with minimal cord manipulation. However, this technique is more demanding than the standard hemilaminectomy and therefore it has a much shallower learning curve for the surgeon.

References

1.  Moissonnier, et al. Vet Surg 2004; 33: 620.