Imaging of Mitochondrial Complex 1 with ¹⁸F-BCPP-EF in the Healthy Human Brain

Objectives: Mitochondrial Complex 1 (MC1) availability has been quantified in the living brain using the novel radioligand [¹⁸F]BCPP-EF, in pre-clinical models of Alzheimer’s Disease (AD), Parkinson’s Disease(PD), ischemic insult and aging (Tsukada, Nishiyama, Ohba et al., 2014; Kanazawa et al., 2017; Tsukada, Ohba, Kanazawa, et al., 2014; Tsukada, Ohba, Nishiyama, et al., 2014). The successful translation of [¹⁸F]BCPP-EF as a tool for quantifying MC1 expression in the human brain, requires the determination of an appropriate kinetic modelling approach. In this study we have evaluated a range of modelling approaches to quantify MC1 in the healthy human brain.

Methods: Eight healthy volunteers (ages 22-75) underwent a 90 minute dynamic [¹⁸F]BCPP-EF scan with arterial blood samples acquired during the scan to enable full quantification. A structural T1 MRI scan of the brain was acquired to allow delineation of anatomical regions of interest. A metabolite corrected arterial plasma input function and regional time activity curves (TACs) were derived from the arterial blood data and the [¹⁸F]BCPP-EF PET emission data respectively. Both the two-tissue compartmental model (2TCM) and the multilinear analysis 1 model (MA1) were applied to the data and the resultant model fits were evaluated in the cerebral white matter (CWM), frontal cortex (FCTX), striatum (STR), hippocampus (HIP) and substantia nigra (SN). Analyses of the PET data were performed using MIAKAT^TM software (version 4.3.1). A variety of outcome parameters including V_T (volume of distribution), V_T/f_p (where f_p is the free fraction of parent metabolite in plasma) and DVR (distribution volume ratio) were examined. DVR values were calculated using the region with the lowest V_T values (CWM) as a pseudo-reference region. The relationship between these outcome parameters and subjects age were also evaluated. Results: [¹⁸F]BCPP-EF uptake was rapid in all regions and peaked between 10-20 minutes(Fig 1a,b), consistent with published preclinical data (Tsukada., 2014), followed by a gradual washout from all regions. Both 2TCM and MA1 produced good fits when applied to the TAC data. V_T values estimated using both the 2TCM and MA1 models were lowest in the CWM and highest in the STR (Fig 1c), with V_T^2TCM values highly correlating with V_T^MA1 values (r = 0.99). V_{T /} f_p results mirrored the V_T estimates with the highest values observed for the STR and lowest for the CWM (Fig 1d). DVR and V_T showed a slight negative trend with increasing age in all regions analysed (STR: -0.05 V_T/year, -0.003 DVR/year, p=0.77). Conclusion: [¹⁸F]BCPP-EF binding in the human brain is consistent with previously published non-human primate data (Tsukada et al, 2014). Kinetic modelling of the data using MA1 appears to be suitable for the quantification of [¹⁸F]BCPP-EF in the human brain. Effects of age on [¹⁸F]BCPP-EF binding appear to be modest. The low uptake and V_T observed for the CWM region warrants further investigation into its potential use as a pseudo-reference region.

• Download figure
• Open in new tab
• Download powerpoint