My initial interest in spontaneous intracranial hypotension (SIH) and localization of CSF leaks began over 10 years ago. The early work of my colleagues Drs. Gary Miller, Bahram Mokri, and David Piepgras served to increase clinical awareness and MRI recognition of SIH. We began to accrue a cohort of patients who had chronic debilitating symptoms of SIH, were unresponsive to multiple large-volume blood patches, and required CSF leak localization for targeted therapy. Many of these patients had rapid spinal CSF leaks that could not be adequately localized with conventional CT myelography. The advent of multidetector CT allowed us to develop a dynamic CT myelogram technique that we reported in 2003.¹ However, because multiple CT acquisitions are performed, dynamic CT myelography is associated with a higher radiation dose and performed without the benefit of a tilting table. Therefore, we initially advocated conventional CT myelography before considering a dynamic study. This approach resulted in the need for all patients with fast CSF leaks to undergo a second myelogram using the dynamic technique.

To mitigate the need for repeat studies, we evaluated potential predictors of fast versus slow leaks, including presence of “brain sag” and dural enhancement on cranial MRI, extradural fluid on spinal MRI, and CSF opening pressure. While there was a significant association between presence of fast leak and extradural fluid on spinal MRI, the other findings were not predictive of fast leak.

We now obtain MR imaging of the cervical, thoracic, and lumbar spine in all patients with refractory SIH who require CSF leak localization. We use the MR results to predict CSF leak rate, which then determines whether we perform a conventional or a dynamic myelogram. In a future effort, we plan to assess the impact of spinal MRI in reducing repeat myelographic studies for fast CSF leak localization in our practice.

Continued advancements in CT technology and an increase in understanding of CSF leak pathophysiology have led to refinements in our dynamic CT myelographic technique.² Recent technological advances in rapid processing and display of images, as well as CT radiation dose reduction techniques, have reduced the number of necessary CT acquisitions and the associated radiation exposure during the dynamic technique.

References

1. Luetmer PH, Mokri B. Dynamic CT myelography: a technique for localizing high-flow spinal cerebrospinal fluid leaks. AJNR Am J Neuroradiol 2003;24:1711–14.
2. Thielen KR, Sillery J, Morris JM, et al. Ultrafast dynamic CT myelography for the precise identification of high-flow cerebrospinal fluid leaks caused by spiculated osteophytes of the spine. Annual Meeting of the American Society of Neuroradiology, Seattle, Washington, June 6, 2011.

 

Read this article at AJNR.org . . .