J Neurol Surg B Skull Base 2014; 75 - A081
DOI: 10.1055/s-0034-1370487

Detection of Skull Base CSF Leaks Using High Resolution CT-MRI Fusion Imaging: A New Algorithm in Management and Treatment

Raj K. Shrivastava 1, Satish Govindaraj 1, Abib Agbetoba 1
  • 1New York City, USA

Introduction: The detection of CSF leaks in the spontaneous or post-operative setting can be difficult and complex to investigate. This process often involves multiple studies and invasive procedures. Often delineating the anatomic pathway can lead to multiple duplicative testing which increase risk and cost to the patient. We investigated a new non-invasive high-sensitivity algorithm to detect CSF leaks. The objective of this study was to evaluate a novel technique of fusing High Resolution CT (HRCT) and MRI images for the detection and localization of CSF leaks noninvasively.

Methods: We developed a method to enhance the CSF pulsation with MRI windowing to track the path of CSF on fused CT/MRI images. The protocol for HRCT imaging included non-contrast helical acquisition with reconstructed images of the paranasal sinuses and temporal bones. MR imaging was then performed using combined sinus and temporal bone protocol, both before and after intravenous gadolinium administration. Fused images were subsequently displayed with T2 MR sequences programmed to render CSF in color, and the MR sequences were superimposed over CT- grayscale bone windows. Analysis was then performed to determine the accuracy of HRCT-MRI fusion imaging modality in detection of CSF leaks. Twenty-four patients who underwent HRCT-MRI fusion imaging for evaluation of suspected CSF leaks were retrospectively reviewed.

Results: Twenty two patients underwent surgical exploration, and 19 were found to have CSF leaks and surgical repair. HRCT-MRI fusion imaging accurately identified the CSF leak and its location in each of the subjects yielding a sensitivity of 100% and a specificity of 95%. Four of the 24 patients were found to have no signs of active leak or skull base defect on intra-operative nasal endoscopy, three of which correlated with negative findings on fusion imaging.

Conclusion: HRCT-MRI fusion is a useful method of localizing CSF leaks with high sensitivity and specificity, and also provides useful anatomical information that is useful during surgical localization and repair. This study was both time and cost efficient with minimal risk to the patients. Most important this protocol provides a high accuracy anatomical pathway with which to do treatment and /or surgical planning. Its use should be incorporated into the diagnostic algorithm for the management of CSF leaks.