What is the optimal sequence of decompression for multilevel noncontinuous spinal cord compression injuries in rabbits?

Abstract

Background: In recent years, multilevel spinal cord injuries (SCIs) have gained a substantial amount of attention from clinicians and researchers. Multilevel noncontinuous SCI patients cannot undergo the multiple steps of a one-stage operation because of a poor general condition or a lack of proper surgical approaches. The surgeon subsequently faces the decision of whether to initially relieve the rostral or caudal compression. In this study, we established a spinal cord compression model involving two noncontinuous segments in rabbits to evaluate the effects of differences in decompression order on the functional recovery of the spinal cord.

Methods: A Fogarty catheter was inserted into the epidural space through a hole in T6-7 and advanced 3 cm rostrally or caudally. Following successful model establishment, which was demonstrated by an evaluation of evoked potentials, balloons of different volumes (40 μl or 50 μl) were inflated in the experimental groups, whereas no balloons were inflated in the control group. The experimental groups underwent the first decompression in the rostral or caudal area at 1 week post-injury; the second decompression was performed at 2 weeks post-injury. For 6 weeks post-injury, the animals were tested to determine behavioral scores, somatosensory evoked potentials (SEPs) and radiographic imaging changes; histological and apoptosis assay results were subsequently analyzed.

Results: The behavioral test results and onset latency of the SEPs indicated that there were significant differences between priority rostral decompression (PRD) and priority caudal decompression (PCD) in the 50-μl compression group at 6 weeks post-injury; however, there were no significant differences between the two procedures in the 40-μl group at the same time point. Moreover, there were no significant peak-to-peak amplitude differences between the two procedures in the 50-μl compression group.

Conclusions: The findings of this study suggested that preferential rostral decompression was more beneficial than priority caudal decompression with respect to facilitating spinal cord functional recovery in rabbits with severe paraplegia and may provide clinicians with a reference for the clinical treatment of multiple-segment spinal cord compression injuries.

Keywords: Balloon compression; Decompression surgery; Multilevel spine injuries; Somatosensory evoked potentials.

Fig. 1

Photographs showing the anatomy of the thoracic vertebrae and the surgical approach used in the model. a Intact rabbit thoracic vertebrae and the corresponding spinal cord segments (T1, T3, T6, T8, T10, and T12). b Locations of T4 and T11 in the spinal cord. c A 2-French Fogarty catheter was inflated with 40 μl (left) or 50 μl (right) of saline solution. d Two small holes were drilled in the vertebral arches of T6 and T7 (arrows indicate the sites of drilling). e A 2-French Fogarty catheter was inserted into the epidural space and advanced 3 cm cranially or caudally. f The catheter was fixed to the skin

Fig. 2

Graphs showing the behavioral outcomes of the different sequences of decompression for double-segmental noncontinuous spinal cord compression injury. The control group exhibited only slight functional changes on the first day after the operation. In the 40-μl compression group, incomplete paralysis occurred in most of the animals after compression. The animals exhibited gradual spontaneous improvement by 1 week post-injury compared with their initial post-injury presentation (**p < 0.01). By the fourth week after decompression, there was a significant difference in behavioral outcomes in the 50-μl PRD group compared with the 50-μl PCD group (^$ p < 0.05); however, there was no difference in behavioral outcomes between the 40-μl PRD and 40-μl PCD groups. Panel a Results of the Reuter test. Panel b Results of Rivlin’s test (^** p < 0.01, ^$ p < 0.05, ^$$ p < 0.01; 1st dec: first decompression, 2nd dec: second decompression)

Fig. 3

Mean SEP waveforms recorded during hindlimb stimulation in the five groups over 6 weeks from the Cz-Fz. The baseline SEP waveforms recorded prior to compression injury are presented in the first row. Each column shows representative SEP waveforms that were recorded in the control, 40-μl PRD, 40-μl PCD, 50-μl PRD, and 50-μl PCD groups (from left to right). B: Baseline. D: Day. W: Week

Fig. 4

SEP onset latencies (a) and peak-to-peak amplitudes (b) of hindlimb stimulation in the five groups of rabbits at scheduled times. ** indicates significant differences between the 50-μl PRD and 50-μl PCD groups (p < 0.01). $$ and $$$$ indicate significant differences in the onset latency at 3 weeks and 2 weeks after SCI (p < 0.01 and p < 0.0001, respectively). # and #### indicate significant differences in the onset latency and amplitudes at 1 week and 1 day post-compression (p < 0.05 and p < 0.0001, respectively). D: Day. W: Week. 1st dec: first decompression, 2nd dec: second decompression

Fig. 5

X-ray images of rabbit models of spinal cord compression injury involving two noncontinuous segments in the control (a) and 40- (b) and 50-μl (c) groups. Representative transverse CT images at the T11 level in the 40-μl group (d) and 50 μl group (e), which show invasion ratios of the spinal canal of 44.74 ± 2.08% and 72.36 ± 3.02% in the 40- and 50 μl groups, respectively. Sagittal T2-weighted MRI images of the spinal cord at 2 weeks post-injury in the 50-μl group. Images an intramedullary high-intensity signal and a dilated central canal (f). The arrows indicate the lesion epicenter

Fig. 6

Histological features of each group, as demonstrated by HE (a-e) and LFB (p-y) staining. Stained cross-sections of representative samples located rostral (a-j, p-t) and caudal (k-o, u-y) to the injury epicenter are presented. The short arrow indicates the VHMNs (magnification a-e and k-y 50×; f-j: 200×)

Fig. 7

Bar graphs indicate differences in the number of VHMNs (a), the size of the cavity area (b) and the degree of white matter sparring (c) in the spinal cord in the control, 40-μl PRD, 40-μl PCD, 50-μl PRD and 50-μl PCD groups. There were significant differences in the number of VHMNs, the sizes of the cavitations and the degree of white matter sparring in the rostral lesion centers between the 50-μl PRD and 50-μl PCD groups. Data are presented as the mean ± SD. ^**** p < 0.0001

Fig. 8

TUNEL immunofluorescence staining for apoptosis in the rostral spinal cord tissue of the control (a), 40-μl PRD (b), 40-μl PCD (c), 50-μl PRD (d) and 50-μl PCD (e) groups. The nuclei were stained with DAPI (blue). The control group did not exhibit apoptosis-positive cells (green). The numbers of apoptotic cells were significantly increased in the 50-μl compression group compared with the 40-μl compression group. Scale bar = 100 μm

Fig. 9

Graphs illustrating the TUNEL assay results at 6 weeks post-injury in the control, 40-μl PRD, 40-μl PCD, 50-μl PRD and 50-μl PCD groups. The numbers of apoptosis-positive cells were significantly decreased in the rabbits subjected to the 50-μl compression injury when they initially underwent rostral compression release compared with the rabbits that underwent PCD. ^**** p < 0.0001, ^$$$$ p < 0.0001