Reproducibility of test-retest with [¹¹C]-PK11195 using different input function approaches

Objectives Tracer Kinetic modeling can add value to clinical studies, but these models often require blood sampling, which has some technical and humanistic limitations, particularly in longitudinal studies within subjects. The image-derived input function (IDIF) has been developed as an alternative. This study compares the reproducibility of the kinetic modeling for a PET brain study using IDIF and AIF for [¹¹C]-PK11195, a marker of microglial activation.

Methods A test-retest protocol was executed in 7 control subjects using [¹¹C]-PK11195. The subjects were given two injections of the tracer on the same day with a two hour interval with no pharmacological intervention. The carotid artery region was used as ROI for the IDIF. Recovery factors and spill-in factors were determined using a static phantom system. These correction factors were applied to the IDIF data to compensate for recovery and spill-in effects. Regions of interest (ROI) were placed with Free Surfer on contemporaneously acquired MRI scans. To compare test-retest reproducibility, Logan-distribution volume (V_T) values were calculated three different ways: a partial volume corrected IDIF, an uncorrected IDIF, and the AIF in 22 different regions in the brain.

Results The static phantom gave results that were entirely consistent with the known resolution of the scanner and the sizes of the syringes used. The uncorrected and corrected IDIFs were directly compared to the AIFs. The reproducibility was excellent in all regions. The average difference between test and retest with the uncorrected IDIF was 11.1±5%, with the corrected IDIF, this was reduced to 7.5±4.6%, while for AIF the average was 5.5±3.6%.

Conclusions Corrected IDIF can be used as an alternative for AIFs to perform [¹¹C]-PK11195 kinetic modeling in brain analyses. Test-retest reproducibility using the corrected IDIF is similar to the AIF. These conclusions support the feasibility of using a corrected IDIF in longitudinal studies that require multiple measurements within subjects.