Traumatic central cord syndrome is the commonest subtype of spinal cord injury now seen in our society. In this talk, I will present an analysis of a 211 prospectively accrued patients in which we have developed a novel clinical prediction tool to evaluate motor recovery. And in this work, we also demonstrate the effectiveness of early surgical intervention within 24 hours in enhancing motor recovery. I wish to acknowledge my co-authors and collaborators who are illustrated on this slide. We have no relevant disclosures. The epidemiology of acute spinal cord injury is changing with an aging population. This enhances the global burden of traumatic brain and spinal cord injury. While the overall incidence of traumatic spinal cord injury remains stable, there are a number of challenges that we face. Patients are getting older, there are more fall mechanisms, and we're seeing more cervical injuries. And as a result, central cord syndrome is becoming a public health priority. The classical description of traumatic central cord injury by Richard Schneider invoked a hyperextension mechanism with preexisting spinal stenosis being a common entity. And patients had disproportionate weakness of the upper and lower limbs. These factors remain valid. However, there are a number of factors that require revisitation. And in this slide, I would indicate that four key assumptions require revisitation. Firstly, it was assumed that the pathophysiology of traumatic central cord injury involved a central hemorrhage that picked off the medial fibers of a somatotopically organized corticospinal tract. In fact, this is not the case. Marche degeneration and neuroanatomical tracer studies reveal the absence of a somatotopic organization of the corticospinal tract. And in fact, autopsy and MRI studies in humans have shown that central hemorrhage in traumatic central cord injury is rare, and in fact, there is diffuse axonal damage to the lateral columns. And because the corticospinal tract is most important for hand and upper extremity function, this accounts for the disproportionate weakness seen in the hands and upper extremities in patients with traumatic central cord syndrome. It was also assumed that the clinical course in traumatic central cord injury was usually favorable. In fact, universally so. But indeed, this is not the case. In fact, the outcomes for patients with central cord syndrome are not that different from any patient with an incomplete cervical cord injury. And here we present novel data from our prospective cohort of patients in 652 patients with traumatic central cord syndrome, as well as non-traumatic central cord syndrome. And here we illustrate that all patients with an incomplete cervical cord injury have a similar trajectory of outcome, and that there really aren't significant differences between the non-central cord syndrome and the central cord syndrome patients. And indeed, the phenotype of central cord syndrome accounts for only 0.3% of the variance in motor scores. If we now graph the data differently, and we show the clinical outcomes of patients with an AIS-C versus an AIS-D, we see that the ASIA grade, in fact, accounts for 40% of the variance in the motor scores. It was also previously assumed that surgery was unsafe, and accordingly, conservative measures were recommended. But we now recognize from Stascus and other prospective studies that early surgery can be done safely with improved motor outcomes. Here we present the data from over 1,500 patients from the NACD and spinal cord injury registry, the NASCS-3 trial, the SciGen trial, and the prospective Stascus cohort study, in which we demonstrate that total motor scores, sensory scores, and AIS grade change are all improved with early surgical intervention done within 24 hours. And finally, it's assumed that central cord syndrome captures a relatively homogeneous population of patients. This is also incorrect. In fact, we see a wide heterogeneity in age, mechanism of injury, and so on. And this is captured here when we actually present the neurological outcomes of every patient in our study with traumatic central cord syndrome. And we see a very heterogeneous trajectory of recovery in these patients. So because of these points, we have argued in the literature that there's a need to revisit the whole concept of central cord syndrome. And this has really prompted us to examine the following aims in this study. One was to develop a clinical prediction model for neurological recovery in traumatic central cord syndrome. And secondly, we sought to evaluate the impact of timing of surgical decompression. We merged the data from four large prospective multi-center data sets, and we specifically examined the population of central cord syndrome, defining this as patients with an AIS grade C or D, where there was a five or more motor point difference between the upper and lower extremities, and patients had a neurological level from C1 to C8. The primary endpoint were the ASIO motor score outcomes. We undertook clinical prediction modeling using multiple linear regression. The variables included in the model were defined a priori based on supporting literature and specific hypotheses. An a priori systematic review found that age and severity of injury were the strongest predictors of neurological recovery in patients with traumatic central cord syndrome. Based on the systematic review of the literature, we included age as a continuous variable, baseline ASIO motor score as a continuous variable, and baseline ASIO impairment scale grade as an economist variable. And based on formulated hypotheses, we examined the timing of surgery as well as the interaction between ASIO grade and timing of surgery. We used multiple imputation techniques, and we internally validated the model using bootstrapping approaches. The CONSOR diagram demonstrates the flow of patients, and we ultimately analyzed 264 patients with traumatic central cord syndrome who underwent surgical intervention. Here we see the baseline characteristics in this patient population. Age of around 50, 21% of females, predominant mechanisms of injury were falls and motor vehicle collisions, just under 75% had ASIO grade D, median time of surgery was 36 hours, and just under 60% of patients received corticosteroids. There was no evidence of non-linearity and no evidence of co-linearity in the key predictive variables. And this table illustrates the parameter estimates in the model, which I will show in the next slide. This is the ultimate equation for the clinical prediction model that we developed, and we found that the following four factors were very important in predicting outcomes from traumatic central cord injury. These included age, the baseline ASIO motor score, the baseline ASIO impairment scale grade, and specifically whether the patient had a C versus a D injury, as well as the timing of surgery. And here we model the outcomes of patients, and we demonstrate that there is, in ASIO C patients, a clear effect of early surgery, which resulted in a 12.7 point additional motor recovery versus late surgery. And there was less of a difference in the ASIO D patients, likely because of a ceiling effect in this population. So in conclusion, the outcomes after central cord syndrome are heterogeneous, age and baseline severity of neurological injury are important predictors of recovery. These are non-modifiable factors. The timing to surgical decompression is modifiable, and critically importantly, this is associated with improved motor recovery, especially in patients with an AIS grade C injury. There is likely a ceiling effect in AIS grade D patients, and this presents the need and the opportunity to potentially look at different ways to model outcomes in these patients. We trust that this clinical prediction model will facilitate patient counseling and clinical decision making. And finally, I wish to acknowledge all of the contributing centers and authors involved in NASCSIS-3 study, the STASCIS Prospective Study, the NACD and Spinal Cord Injury Registry, as well as in the CYGEN trial. Thank you.

traumatic central cord syndrome

spinal cord injury

motor recovery

early surgical intervention

epidemiology of spinal cord injury