Abstract
272
Objectives Therapeutic drugs targeting receptor systems may cause heterogeneous occupancy among brain regions potentially due to different affinities for systems/subsystems and direct actions and interactions with other systems. This study introduces a method that allows identification of heterogeneous occupancy patterns that may contribute to our understanding of mechanisms of therapeutic and adverse effects of drugs.
Methods Alcohol dependent subjects (n=23) were studied with PET and [11C]carfentanil, a mu-opioid receptor ligand before and during treatment with naltrexone (50 mg/day for 5 days). Maps of binding potential (BPND) were constructed using reference tissue graphical analysis (Logan et al., 1996). Regression lines of scatter plots of BPND values of 24 brain regions were obtained (x:baseline BPND; y: post-treatment BPND) for individual subjects. Model-predicted post-Tx BPND maps (i.e., under an assumption of homogenous occupancy across regions) were constructed by applying the regression lines to baseline BPND maps. Observed and model-predicted post-Tx BPND maps were submitted to SPM to identify clusters that showed more or less occupancy than the homogenous occupancy assumption.
Results The scatter plots were described by linear regression in all subjects (R2>0.84). SPM showed bilateral thalamus and ventral putamen with greater occupancy (p<0.05, corrected) than other brain regions (i.e., than expected by the homogenous displacement assumption) while bilateral sub-callosal anterior cingulate and temporal base showed less occupancy (p<0.05, corrected).
Conclusions This fluctuations along regression lines of scatter plots of BPND (baseline vs. post-Tx) occur in a region specific manner and can be captured with voxel-wise-statistical method. This proposed method may add a new tool for understanding the mechanisms of therapeutic and adverse effects of drugs when results of different drugs for a neuropsychological disease are compared