The introduction of magnetic resonance imaging (MRI) to the field of spine trauma has vastly improved the clinical diagnosis of spinal cord injury (SCI). MRI has advantages over conventional X-ray and computer tomography (CT) as it more precisely details posttraumatic compression of the spinal cord, either due to soft tissue (i.e., traumatic disc herniation or bleeding within the spinal canal) or spinal canal encroachment (e.g., following a burst fracture or vertebral misalignment).
Also in cases where an SCI is clinically suspected, but without obvious vertebral column fracture or discoligamentary injury (SCIWORA ¼ spinal cord injury without obvious radiological abnormality), MRI is essential for the diagnosis of injury and planning of appropriate surgical interventions. In addition, MRI studies in acute SCI are applied in order to estimate long-term outcomes (i.e., prognosis) and as a postsurgical outcome measure in selected cases to confirm that the spinal cord is indeed sufficiently decompressed.
With the advent of novel treatment options aiming at the repair of the injured spinal cord, insights into disease mechanisms paralleling functional recovery are essential in order to distinguish treatment-induced changes from spontaneous recovery (i.e., pattern and extent of repair/regeneration beyond spontaneous recovery). Regardless if applied during spontaneous or therapeutically derived recovery, changes in anatomical substrates in the spinal cord may be subthreshold to detection compared to other clinical measurement instruments.
NEUROPHYSIOLOGY AND MRI: COMPLEMENTARY APPROACHES
At present, most clinical applications of MRI to assess the injured spinal cord are rather qualitative (i.e., defining the localization, size, and type of damage, such as bleeding and compression). Therefore, MRI findings of the injured spinal cord need to be complemented by clinically meaningful readouts as well as quantitative neurophysiological measures that provide additional insights into spinal cord function.
Neurophysiological approaches commonly adopted after SCI include those that objectively examine spinal conduction in specific ascending and descending pathways. Ascending fibers in the dorsal columns, which convey light touch sensation and proprioception, are most frequently studied by measuring somatosensory evoked potentials (SSEPs) using electroencephalography (EEG) techniques following surface electrical stimulation of mixed nerves in the periphery (e.g., tibial nerves).
The other major ascending sensory pathway, conveying pain and temperature sensation (i.e., spinothalamic tract), can likewise be investigated monitoring evoked EEG responses to contact or radiant heat stimulation (contact heat evoked potentials, CHEPs, or laser evoked potentials, LEPs, respectively). Descending pathways (e.g., corticospinal tract) are routinely examined by measuring motor evoked potentials (MEPs) following noninvasive stimulation of the motor cortex (i.e., transcranial magnetic stimulation or TMS) and recording electromyography (EMG) in the periphery.
MORPHOLOGICAL CHANGES IN THE SPINAL CORD AFTER SCI
The primary strength of MRI in the acute stages of SCI is to objectively assess the localization and extent of morphological damage in the spinal cord, from which surgical interventions can be planned. During the transition from the acute to chronic stage of SCI, the role of MRI changes, with the emphasis to disclose secondary morphological changes ongoing in the spinal cord.
Therefore, it is important to understand the typical evolution of spinal cord damage according to clinical MRI findings and the extent to which observed changes can be related to clinical outcome. From the perspective of a clinical trial, an understanding of the dynamic temporal and spatial pattern of the changes in the gross damage and eventual formation of the posttraumatic (i.e., consolidated) lesion area is required to distinguish potential effects of interventions on the lesion area.
While the acquisition of MRI after SCI is routinely performed in the acute stages of injury, there are no studies that have systematically tracked (prospective longitudinal follow-up) changes in the spinal cord during the course of recovery. This is likely due, in part, to some of the challenges of spinal cord MRI. However, based on clinical observations of T1- and T2-weighted anatomical images collected serially in the first year after SCI in patients demonstrating representative patterns of spontaneous recovery, and where surgical instrumentation was either not implanted or not resulting in significant MRI artifacts, SCI are often characterized by three prominent morphological stages.
RELATIONSHIP BETWEEN NEUROLOGICAL RECOVERY AND MORPHOLOGICAL CHANGES
Some changes in clinical MRI findings may correspond with obvious features of spontaneous neurological recovery. For example, the greatest period of spontaneous neurological recovery according to clinical sensory and motor testing outcomes is expected in the initial months after injury, begins to plateau at approximately 4–6 months, and remains relatively stable thereafter.
Therefore, spontaneous recovery roughly occurs in parallel with some of the morphological changes described before. However, the relationship between morphological changes and spontaneous recovery is poorly understood. Even in patients where gross morphological changes are observed (e.g., enlargement of the posttraumatic cyst), the severity of injury may not change and the patient remains clinically stable.
While this represents an apparent discrepancy between MRI and functional outcomes in longitudinal pathways traversing the lesion epicenter, this underscores the importance of coupling MRI with neurophysiological outcomes to assess function. Indeed, neurophysiological outcomes assessing conduction in ascending and descending pathways traversing through the lesion level generally remain unchanged during recovery.
THE MRI PARADOX: WHEN ANATOMICAL CHANGES DO NOT CORRESPOND TO FUNCTIONAL CHANGES
The “MRI paradox”, whereby gross morphological changes are observed according to MRI in the absence of any or severe functional deterioration in the spinal cord or vice versa, continues to puzzle clinicians and researchers.Some of the most complex and interesting cases that highlight the MRI paradox involve patients with posttraumatic holocord syringomyelia. In the example shown in Figure 1.3B.2, the neurophysiological readouts indicate that spinal conduction is, in fact, almost completely normal in the upper limbs.
Thus, despite a gross morphological change in the spinal cord affecting cervical segments, function in ascending and descending pathways remains nearly completely intact. The reason for such gross morphological changes unaccompanied by severe functional changes or neurological deterioration is not well understood.
However, this paradox is not unique to MRI; histological studies of the spinal cord also illustrate that function may not accurately correspond to the anatomical neuropathology after injury. Indeed, the spinal cord is rarely completely transected even in cases with complete sensory and motor loss, and spared white matter traverses the injury epicenter.
In part, the discrepancy between anatomical and functional changes may be related to the low resolution and/or low specificity/sensitivity of clinical MRI techniques to distinguish differences in pathology (e.g., demyelination). Implementing a neurophysiological approach is therefore important in order to determine if subtle changes in spinal conduction may be present in the white matter, as changes in latency and amplitude of evoked potentials (sensory and motor) may also precede the onset of measurable clinical deficits.
CONVENTIONAL MRI APPROACHES IN SCI CLINICAL TRIAL MANAGEMENT
At present, there is no effective treatment option to resolve the sensory loss and motor paralysis associated with SCI. However, there are a number of potential strategies that might translate to patient benefits, and implementing valid and reliable outcome measures to assess safety and efficacy of interventional strategies is paramount. Since small treatment effects may accompany initial approaches to resolve sensorimotor deficits, it is important that outcomes are sensitive to detect subtle changes (i.e., responsive).
Therapeutic options in SCI generally fall into one of two categories: neuroprotective or neurorestorative. Whereas neuroprotection aims to prevent secondary damage caused by the cascade of biochemical events in the central nervous system triggered by SCI, including hemorrhagic events and edema, neurorestorative strategies intend to reconstitute partial or complete loss of function in neural circuitry affected by SCI. The latter category includes regeneration and remyelination of axons across or around a lesion site in the spinal cord.
Regardless of the treatment strategy employed, the changes in neurological structures are expected to occur over time, often during spontaneous recovery, and therefore require longitudinal investigation. The capability of conventional MRI to noninvasively examine gross morphological changes in the spinal cord represents a potentially powerful outcome measure on which to determine therapeutic safety and efficacy of a given treatment. The use of MRI in this way goes beyond most current clinical trial applications, which have employed MRI criteria for the purposes of reducing subject heterogeneity (i.e., inclusion/exclusion criteria).
FUTURE DIRECTIONS OF MRI AND SCI
The use of MRI to serially evaluate pathological events within the spinal cord in neurological conditions, such as multiple sclerosis (MS), is well established. Despite recent and numerous advances in MRI techniques, it remains technically challenging to obtain meaningful MRI results at the level of the injured spinal cord.
As previously discussed, this is partially due to artifacts from fractured disks and fixative vertebral implants near the level of injury. A promising strategy to overcome the issues related to surgical instrumentation is to focus on areas rostral to the injury site initially unaffected by the lesion area. A valid and established MRI tool is the assessment of cross-sectional spinal cord area.
Usually, the crosssectional spinal cord area is estimated on reformatted axial slices at cervical level C2/C3 on 3D T1-weighted anatomical scans. In individuals with SCI and instrumentation fixating the spinal column, this is generally well above the injury level and thus artifact free. In MS, cross-sectional spinal cord area measurement has proven to be sensitive toward changes in disease state15 and has been employed to track potential treatment effects associated with a therapeutic intervention.
Diffusion Weighted and Magnetization Transfer Imaging
Characterizing microstructural alteration of spinal white matter integrity with diffusion tensor imaging (DTI) can further improve our understanding between central tract specific changes to clinical impairment. During acute SCI, DTI has been shown to be predictive of long-term functional recovery In chronic SCI, DTI is correlated with clinical measures of injury severity.
The reproducibility of DTI has also recently been demonstrated in children. DTI indexes are altered at the lesion site and rostral to it suggesting trauma-induced degenerative processes in ascending and descending central pathways. Importantly, reduced white matter integrity of specific central pathways was linked to clinical disability.
Thus, DTI holds promise to quantify the degree of white matter integrity, to predict recovery, and to monitor the effects of therapeutic interventions. Similarly, magnetization transfer imaging has shown potential for detecting pathology in the white matter associated with demyelination and could be a potential avenue for probing the effect of remyelinating therapies in SCI.
Author: John Kramer, Patrick Freund, Armin Curt