In sub study one, we observed no significant difference in the increase of ISP following TSCI compared to the sham operated group. The absolute mean difference in ISP between TSCI and sham was not higher than the mean difference in ISP following repositioning. In sub study two, we observed no effect of increasing impact nor duration of the compression compared to the regime of sub study one and the sham operated group. The animals were kept physiologically stable throughout the course of the respective sub studies.
We observed an increase in ISP for both sham and TSCI animals. The increase was higher in the TSCI groups. However, the absolute difference between the highest ISP in both groups was 0.8 mmHg which is lower than considered clinically significant.
Following these findings, a subsequent study applying increased TSCI impact severity and duration was conducted (sub study two). Neither of these regimes produced significant ISP rises compared to sham or inter- injury regimes. We decided upon a large falling height (125 mm) and prolonged compression (4 hours). The large falling height produced substantial macroscopic injury, in which hematoma below dura was evident. For one animal subjected to the 125 mm falling height, we observed an ISP in the range of 20–25 mmHg, but generally, the increased falling height and longer compression failed to produce further elevation of ISP compared to lower falling heights and short duration of compression.
The mixed model of sub study two, found a positive and significant interaction effect of the 75 mm 240 min group. This means that the 75 mm 240 min group had ISP significantly increased compared with sham over time. It is imperative to state that it did not find the ISP of the 75 mm 240 min group to be higher than the sham. Since this group started out at a markedly lower ISP than the sham, the finding merely suggests, that the group increases towards the sham; not above. The reason for this could be the removal of the rod after 240 min. It was evident that the cord was completely compressed, in turn reducing the subdural tissue volume. The dura would quickly expand, leaving the pressure at zero or even negative.
Previous TSCI studies investigating ISP have shown divergent results. Human studies have shown increased ISP in human TSCI20,27. In one study, the ISP increase was evaluated against a control group (12 mmHg), consisting of patients suspected of normal pressure hydrocephalus having their CSF pressures measured, using a lumbar catheter8. The authors found a significant difference in mean ISP (approximately 10 mmHg). Two porcine studies found elevated ISP after onset of TSCI16,17. The first study applied intraparenchymal probes and found a 220% increase in ISP compared to baseline (10.5 mmHg vs 23 mmHg). This study did not have a control group. However, in another study applying the same model to evaluate the effect of duroplasty, the control group (TSCI minus duroplasty) mean ISP was only just above 10 mmHg16. This is comparable to ISP observed in our study.
The pathophysiological processes associated with TSCI may differ between pigs and humans, and both our results and previous results suggest that our TSCI porcine model does not reach the anticipated compartment syndrome16. Perhaps the development of high ISP is a multifactorial process including contributions from both the subdural and extradural compartments, and other factors, that we do not have control of in this study. This argument would explain the observed pressures in both ends of the spectrum in two different studies 16,17, and why we observed one animal with high ISP (20–25 mmHg). Further, a multifactorial mechanism may explain the human findings, where not all patients seem to develop high ISP8.
Another explanation of our model not producing a subdural compartment syndrome, could be the anatomical differences between cervical and thoracic regions of the spine. We opted for a thoracic injury to reduce the risk of an unstable model due to autonomic complications. The cervical spinal cord is larger in the traverse plane and might fill the dura more readily. It cannot be excluded as one of the factors driving a compartment syndrome and may explain why our model found different ISP increases compared to those reported in human studies with cervical patients8,22.
Another difference between our study and earlier studies is the use of general anesthesia throughout the study. In TBI, anesthesia, using propofol and midazolam will lower cerebral metabolism, in turn lowering intracranial pressure (ICP)28. Since our animals were anesthetized for the entire study, we cannot exclude a diminished ISP rise in TSCI compared to sham using anesthesia. However, a general and close to similar increase in ISP was found in both sham and TSCI. Following, anesthesia should selectively diminish the trauma effects to introduce bias.
Considering our knowledge from TBI, it is known that hyperventilation decreases ICP through vasoconstriction and hence a decreased intracranial volume of blood28. In our study, animals were kept normo-ventilated. One way of mitigating ICP increase in TBI is through plasma sodium concentration increase. Despite meticulous corrections, two TSCI animals had sodium levels above reference at the late stage of the experiments. The group mean plasma sodium concentration was within the reference range. We find it less likely, that minor divergence in plasma sodium could account for the lack of ISP increase in the TSCI group alone.
Despite not observing differences in ISP between groups, we did observe an overall ISP rise from baseline for all groups. These finding suggest that the ISP rise, or at least a part of it, is not dependent on TSCI, but is rather the result of the surgical procedure and/or probe placement. First, the surgical procedure itself produces concomitant edema and minor hemorrhage in the soft tissue above the spinal column could cause external dural pressure. This may explain the observed slow pressure buildup during the first 12–24 hours. To evaluate this mechanism, supplementary studies measuring the epidural pressure should be conducted. Second, the probe is located intimately against the spinal cord and could theoretically produce spinal cord edema following placement by inflicting a minimal trauma. Disregard it would be minimal and we find this mechanism to be likely. Third, the subarachnoid space is sparse, and probe volume may contribute to ISP increase by occupying part of the subarachnoid space. If this mechanism was true, one would expect an abrupt pressure increase to a steady state during seconds. This is not observed in our data; contrary, a gradual buildup of pressure is observed during the first 12 hours. Finally, the observed rise may represent the reconstitution of the normal CSF pressure. During probe placement, CSF leaks through the incision in the dura is unavoidable. After placement, a watertight dural closure was performed. This could allow for a gradual reconstitution of CSF. Following, a pressure buildup could be observed. Despite different reports on the reference range and upper normal level for CSF pressure in humans, there is consensus on the pressure not being zero and probably in the range of 4–30 mmHg29,30. This range includes the ISPs we have observed in the present study.