Abstract
1768
Objectives Administration of therapeutics to the cerebrospinal fluid (CSF) is a promising method for the therapy of diseases involving CNS and meninges. Investigation of drug pharmacokinetics in CSF is difficult and generally requires quantitative real-time imaging, such as PET. However, in the PET image, drug associated with arachnoid structures cannot be resolved from the liquid phase fraction dissolved in the CSF. Sampling of CSF can supplement the PET data, but removal of CSF can distort the kinetics. We report an alternative method based on in-catheter “sampling” of CSF by PET.
Methods Macacas Fascicularis with implanted subcutaneous lumbar ports equipped with catheters leading to CSF were imaged with the MicroPET Focus 220 and CereTom NL 3000. The model drugs labeled with I-124 were administered through the ports and the latter were flushed. To obtain liquid phase activity values, a small volume of CSF (<0.1 ml) was pooled back into the connecting catheter. Drug concentration in the catheter segment filled with CSF was determined from the total radioactivity as measured by PET and the known volume of the segment. CSF was then flushed back into the leptomeningeal space (LMS). Accuracy validation was carried out in vitro, using analogous injection port.
Results In vitro, the accuracy of the concentration measured in catheter by PET was shown to deviate from the true value by no more than ±10%. In contrast, the values obtained by CSF collection through the injection port were not accurate unless an unacceptably large volume of CSF was taken (>1 ml). Animal data demonstrated the utility of the proposed approach for analysis of drug distribution between the liquid and solid phases in the LMS at multiple time points.
Conclusions PET imaging enables non-invasive real-time selective quantification of the radiolabeled drug concentration in the liquid phase of the LMS.
Research Support This work was supported by NIH grant R21 CA152384, DoD grant BC100684 and internal funds of the laboratory.