Quantification of P-glycoprotein function at the human blood-brain barrier using [F-18]MC225 and PET

Introduction P-glycoprotein (P-gp) is one of the most studied efflux transporters at the blood-brain barrier. It plays an important role in brain homeostasis by protecting the brain from a variety of endogenous and exogeneous substances. Changes in P-gp function are associated both with the onset of neuropsychiatric diseases, including Alzheimer's disease and Parkinson's disease, and with drug-resistance, for example in treatment-resistant depression. The most widely used approach to measure P-gp function in vivo is (R)-[C-11]verapamil PET. (R)-[C-11]verapamil is, however, an avid P-gp substrate, which complicates the use of this tracer to measure an increase in P-gp function as its baseline uptake is already very low. [F-18]MC225 was developed to measure both increases and decreases in P-gp function. Aim The aim of this study was (1) to identify the pharmacokinetic model that best describes [F-18]MC225 kinetics in the human brain and (2) to determine test-retest variability. Methods Five (2 male, 3 female) of fourteen healthy subjects (8 male, 6 female, age 67 +/- 5 years) were scanned twice (injected dose 201 +/- 47 MBq) with a minimum interval of 2 weeks between scans. Each scanning session consisted of a 60-min dynamic [F-18]MC225 scan with continuous arterial sampling. Whole brain grey matter data were fitted to a single tissue compartment model, and to reversible and irreversible two tissue-compartment models to obtain various outcome parameters (in particular the volume of distribution (V-T), K-i, and the rate constants K-1 and k(2)). In addition, a reversible two-tissue compartment model with fixed k(3)/k(4) was included. The preferred model was selected based on the weighted Akaike Information Criterion (AIC) score. Test-retest variability (TRTV) was determined to assess reproducibility. Results Sixty minutes post-injection, the parent fraction was 63.8 +/- 4.0%. The reversible two tissue compartment model corrected for plasma metabolites with an estimated blood volume (V-B) showed the highest AIC weight score of 34.3 +/- 17.6%. The TRVT of the V-T for [F-18]MC225 PET scans was 28.3 +/- 20.4% for the whole brain grey matter region using this preferred model. Conclusion [F-18]MC225 V-T, derived using a reversible two-tissue compartment model, is the preferred parameter to describe P-gp function in the human BBB. This outcome parameter has an average test-retest variability of 28%.

Automated summary

The aim of this study was to determine the best pharmacokinetic model to describe [F-18]MC225 kinetics in the human brain and assess test-retest variability. The reversible two tissue compartment model corrected for plasma metabolites with an estimated blood volume (VB) showed the highest AIC weight score. The preferred parameter to describe P-gp function in the human BBB using this model was [F-18]MC225 VT, which had an average test-retest variability of 28%.