In this study, we established severe SCI models to investigate the feasibility of DTI in evaluating the protective effects of HBO-PC. Our findings demonstrate that HBO-PC mitigated pathological alterations in the injured spinal cord and enhanced hindlimb motor function in rats following SCI. Furthermore, DTI was shown to be a valuable tool for assessing functional recovery in the hind limbs of rats.
To date, there is no standardized HBO-PC protocol for SCI. In several rodent experimental studies on SCI, HBO-PC has been conducted at pressures ranging from 2.0 to 2.5 ATA and durations lasting from 60 to 120 minutes per session over 5–10 consecutive days[18, 19, 25]. Considering that the common clinical pressure for HBOT is 2.0 ATA[26], the HBO-PC regimen used in this experiment consisted of 2.0 ATA, two 60-minute sessions per day for 5 consecutive days.
The neuroprotective mechanisms of HBO-PC include anti-oxidative stress[27–30], anti-inflammation[12, 19], suppression of apoptosis[17–19, 25, 30], and promotion of autophagy[31, 32]. Chen et al.[19] demonstrated that HBO-PC intervention reduced the expression of inflammatory cytokines (TNF α, IL-6, and IL-1β) in SCI rats. Cell apoptosis, a critical component of secondary injury, is a major contributor to the reduction in neuron numbers after SCI[33]. It initiates shortly after the primary injury and peaks around the 7th day[34]. This timeframe guided the choice of our 7-day research duration. Pan et al.[18] confirmed that HBO-PC inhibited the expression of Caspase-3/7/8/12 by blocking the endoplasmic reticulum apoptotic pathway in neurons of SCI rats. In our study, we observed that the HBO-PC group exhibited more intact cell structures, less edema, and fewer inflammatory cells compared to the SCI group at the same time point. Additionally, the HBO-PC group showed relatively regular neuronal morphology and a higher number of granules compared to the SCI group in Nissl staining. These findings suggest that HBO-PC can promote the preservation of spinal cord neurons in rats with SCI.
To date, most studies investigating the neuroprotective effects of hyperbaric oxygen preconditioning (HBO-PC) have relied on invasive methods such as histopathology, electrophysiological examination, behavioral scales, or signaling pathway analysis. Only one animal study utilizing conventional MRI sequences found no significant spinal cavitation or atrophy in the preconditioning group[19]. However, conventional MRI is limited in its ability to assess the integrity of white matter tracts, which are crucial for identifying spinal cord conditions[35]. In this study, DTI, a noninvasive and quantifiable MRI sequence, was employed to assess the effects of HBO-PC. The results indicated a significant decrease in FA values after SCI, consistent with our previous findings[22, 23]. FA values reflect the integrity of white matter tracts, with lower values indicating more severe injury[36]. Following surgery, the FA values of the preconditioning group were consistently higher than those of the SCI group, suggesting that the preconditioning intervention effectively mitigated the loss of white matter bundles. Furthermore, differences in AD and RD values between the HBO-PC and SCI groups were investigated. The results showed that the SCI group had higher AD values at 24 hours and 3 days, as well as higher RD values from 24 hours to 7 days, compared to the time-matched HBO-PC group. Previous studies have suggested that a decrease in AD values is indicative of spinal cord axonal damage[25], while an increase in RD values reflects demyelination and axonal damage[37]. However, these changes are not always consistent following a spinal cord injury[37], which may explain the higher AD and RD values observed in the SCI group compared to the HBO-PC group. Additionally, D’souza et al.[38] found that patients with acute cervical trauma exhibited higher MD values at the injury levels compared to healthy controls. Similarly, in our study, MD values significantly increased in the SCI group compared to the HBO-PC group, except at the 6-hour time point.
DTT, an important extension of DTI, enables direct observation of changes in white matter fibers in three-dimensional space. Cao et al.[39] proposed that the severity of motor function impairment correlates with the extent of injury to the ventral and lateral white matter of the spinal cord. In our study, the SCI group at 6 hours exhibited significant interruption of nerve fiber bundles in the spinal cord injury site, along with varying degrees of ventral fiber bundle loss over time, indicating ongoing secondary injury. The HBO-PC group showed less ventral fiber bundle loss, likely attributed to HBO-PC’s mitigation of secondary spinal cord injury. Moreover, the presence of more residual fiber bundles in the HBO-PC group may contribute to improved motor function in later stages of recovery in rats.
Regarding the BBB locomotor rating scale, we observed a significant decrease in hind limb BBB scores in SCI rats at 0 hours post-surgery, followed by a gradual increase over time, consistent with our previous findings[22]. Additionally, we investigated the effect of HBO-PC on SCI rats over one week. The data indicated that the BBB score of the HBO-PC group was higher than that of the SCI group only at day 7, suggesting that the BBB behavioral rating scale may be less sensitive in detecting the protective effects of preconditioning, which may only become apparent in the late subacute stage of SCI.
In this study, we conducted a correlation analysis between DTI parameters and BBB scores. Unlike previous studies, our analysis revealed a weak or moderate correlation between DTI parameters and BBB scores in the SCI or HBO-PC groups. This finding may be attributed to the fact that BBB scores in the SCI and HBO-PC groups were predominantly 0 (indicating no significant hindlimb movement) at the first three time points. However, in the correlation analysis of the two groups, FA values exhibited a moderate correlation with BBB scores, indicating that changes in DTI parameter values can reflect the functional recovery of rats with SCI after HBO-PC to some extent.
This study has several limitations that should be acknowledged. First, due to the small size of the rat spinal cord and the quality of MRI images acquired from a relatively low magnetic field intensity scanner, further segmentation and analysis of the spinal cord structure were not feasible. Further verification will be conducted using magnetic resonance with higher field strength. Second, there may be a potential bias introduced by the presence of cerebrospinal fluid and hemorrhagic foci when delineating the regions of interest. Efforts were made to minimize partial volume effects during the delineation process. Lastly, the findings are based on rodent experiments and require validation in high-risk populations for SCI.