A new approach for simple radioisotope cisternography examination in cerebrospinal fluid leakage detection.

OBJECTIVE

We developed a new quantitative interpretation technique of radioisotope cisternography (RIC) for the diagnosis of spontaneous cerebrospinal fluid hypovolemia (SCH).

METHODS

RIC studies performed for suspected SCH were evaluated. (111)In-DTPA RIC images were taken at 0, 1, 3, 6, and 24-h after radioisotope injection following the current protocol. Regions of interest (ROI) were selected on 3-h images to include brain, spine, bladder or the whole body. The accumulative radioactivity counts were calculated for quantitative analysis. Final diagnoses of SCH were established based on the diagnostic criteria recently proposed by Schievink and colleagues.

RESULTS

Thirty-five patients were focused on. Twenty-one (60.0%) patients were diagnosed as having SCH according to the Schievink criteria. On the 3-h images, direct cerebrospinal fluid leakage sign was detected in nine of 21 SCH patients (42.9%), as well as three patients with suspected iatrogenic leakage. Compared to non-SCH patients, SCH patients showed higher bladder accumulation at 3-h images (P = 0.0002), and higher brain clearance between the 6- and 24-h images (P < 0.0001). In particular, the 24-h brain clearance was more conclusive for the diagnosis than 24-h whole cistern clearance. The combination of direct sign and 24-h brain accumulation resulted in 100% of accuracy in the 32 patients in whom iatrogenic leakage was not observed. 1- and 6-h images did not provide any additional information in any patients.

CONCLUSIONS

A new simple ROI setting method, in which only the 3-h whole body and 24-h brain images were necessary, was sufficient to diagnose SCH.