ULTRASONOGRAPHIC ANATOMY OF THE NORMAL CANINE SPINAL-CORD AND CORRELATION WITH HISTOPATHOLOGY AFTER INDUCED SPINAL-CORD TRAUMA

Prior to trauma, intraoperative ultrasound of the spinal canal in 31 n ormal dogs was performed through a hemilaminectomy in the left pedicle of L2. A ventral compressive model of spinal cord injury was performe d as part of a clinical drug trial. Maximum ultrasonographic spinal co rd diameter ranged from 4.9-7.2 mm (5.7 +/- 0.6). Significant positive correlation (p = 0.023, r = 0.49) was found between age and spinal co rd diameter. The dura mater was a separate, well-defined, echogenic ho rizontal line in 28 (90%) dogs, dorsally, and in 29 (94%) dogs, ventra lly. Cerebrospinal fluid was anechoic. Eighteen (58%) dogs had a well- defined anechoic dorsal subarachnoid space, whereas 22 (71%) had a wel l-defined ventral space. Pia mater was thin but strongly echogenic and covered spinal cord. Central canal was a double hyperechoic line in 1 7 (55%) dogs and a single-line in 14 (45%) dogs. A difference in the u ltrasonographic appearance between gray and white matter was not seen. Epidural fat and connective tissue was a lobular echogenic material i n the ventral epidural space. The periosteal-vertebral body interface was seen as a bright curvilinear echo with distal acoustic shadowing. Spinal cord parenchyma could be classified subjectively into four grou ps based upon ultrasonographic appearance. Spinal cord parenchyma had a uniform hypoechogenicity in 8 (27%) dogs (Group 1), subtle low level echoes in 7 (23%) dogs (Group 2), multiple clusters of defined echoge nic foci in 12 (37%) dogs (Group 3), and multiple sharply-defined line ar echoes in 4 (13%) dogs (Group 4). There was a significant relations hip between pre-trauma ultrasonographic appearance of the spinal cord and histopathology 21 days after trauma. One (13%) dog in Group 1, 4 ( 57%) dogs in Group 2, 10 (91%) dogs in Group 3, and 3 (75%) dogs in Gr oup 4 had malacia on histological evaluation. Therefore, dogs with ech ogenic spinal cords or linear echoes within cord parenchyma were signi ficantly more likely to develop malacia rather than Wallerian degenera tion after induced spinal cord trauma (p = 0.002). Spinal cord echogen icity may indicate vascularity in a segment of spinal cord and might b e prognostic following spinal cord trauma. No complications were found related to intraoperative ultrasound. Hematoma or fibrous tissue form ation appeared to impede percutaneous ultrasound of the spinal cord in dogs re-evaluated forty-eight hours and one week after surgery.