Syringomyelia denotes a fluid-filled cyst, termed a syrinx, that forms in the spinal cord. Affected patients suffer from progressive paralysis. Syringomyelia is most frequently associated with Chiari I malformation in which cerebellar tonsils ectopically protrude through the foramen magnum into the spinal canal. Subarachnoid cerebrospinal fluid (CSF) outflow from the spinal canal is constricted such that, during systole, heightened pressure waves are generated in the partially enclosed spinal subarachnoid space and transmitted compressively to the spinal cord. One hypothesis for the unknown mechanism underlying the development and progression of syringomyelia is that the cyst fluid accumulates because of an imbalance of CSF movement through the spinal cord extracellular space during the cardiac-derived pressure cycles. This hypothesis might be testable by following a marker that tracks the movement of CSF. The computed tomography (CT) contrast agent, iopamidol was tested as a surrogate marker. Serial CT iopamidol myelography was performed in 15 Chiari I patients before and after corrective surgery for the Chiari malformation. The surgery consisted of craniocervical decompression and duraplasty to improve CSF flow at the foramen magnum. For the myelography, iopamidol was injected into the CSF through lumbar punctures and CT scans were obtained at the level of the C4-C7 vertebrae at nominal post-injection times of 2, 4, 6, 8, 10, 22 and, in a few cases, 30 and 48 hrs. The scan images were segmented into subarachnoid space (SAS), spinal cord (SC) and syrinx regions. The gray-scale values within each region were spatially averaged and recorded in Hounsfield units. Three scanners were used and each was individually calibrated to permit conversion from Hounsfield units to iopamidol concentrations. The half-life of iopamidol in the SAS was shorter after surgery as expected for enhanced CSF flow. The time profiles of iopamidol concentration in the SAS and syrinx were fit to simulations from an empirical two-compartment pharmacokinetic model to quantitatively characterize the kinetics of iopamidol exchange between these two regions. There were no statistically significant differences either between the rate constants for iopamidol influx to and efflux from the syrinx or between the preoperative and postoperative mean rate constants. A mechanistic mathematical model was developed to describe the exchange based on the assumption that it occurs as a result of iopamidol diffusion through the extracellular space of the intervening spinal cord tissue. The measured SAS, spinal cord and syrinx concentration profiles were compared to profiles predicted from this mathematical model to assess whether the observations could be explained without invoking other mechanisms for exchange besides extracellular diffusion. The comparison suggested that diffusion could be the predominant mechanism, provided that the intervening spinal cord is edematous with an extracellular volume fraction that is enlarged to the order of 0.4 from a value in normal tissue of about 0.2. The result suggests that a less diffusive marker would be needed for assessing whether CSF convection through the extracellular space is appreciable.
