ORIGINAL ARTICLE Quantification and discriminative power of 18 F-FE-PE2I PET in patients with Parkinson’s disease Aline Delva 1,2 & Donatienne Van Weehaeghe 3,4 & June van Aalst 4 & Jenny Ceccarini 4 & Michel Koole 4 & Kristof Baete 3,4 & Johan Nuyts 4 & Wim Vandenberghe 1,2 & Koen Van Laere 3,4 Received: 15 July 2019 /Accepted: 17 October 2019 # Springer-Verlag GmbH Germany, part of Springer Nature 2019 Abstract Rationale Dopamine transporter (DAT) imaging is an important adjunct in the diagnostic workup of patients with Parkinsonism. 18 F-FE-PE2I is a suitable PET radioligand for DAT quantification and imaging with good pharmacokinetics. The aim of this study was to determine a clinical optimal simplified reference tissue-based image acquisition protocol and to compare the discriminatory value and effect size for 18 F-FE-PE2I to that for 123 I-FP-CIT scan currently used in clinical practice. Methods Nine patients with early Parkinson’ s disease (PD, 64.3 ± 6.8 years, 3M), who had previously undergone a 123 I-FP-CIT scan as part of their diagnostic workup, and 34 healthy volunteers (HV, 47.7 ± 16.8 years, 13M) underwent a 60-min dynamic 18 F-FE-PE2I PET-MR scan on a GE Signa 3T PET-MR. Based on dynamic data and MR-based VOI delineation, BP ND , semi- quantitative uptake ratio and SUVR [t1–t2] images were calculated using either occipital cortex or cerebellum as reference region. For start-and-end time of the SUVR interval, three time frames [t 1 –t 2 ] were investigated: [15–40] min, [40–60] min, and [50–60] min postinjection. Data for putamen (PUT) and caudate nucleus-putamen ratio (CPR) were compared in terms of quantification bias versus BP ND and discriminative power. Results Using occipital cortex as reference region resulted in smaller bias of SUVR with respect to BP ND + 1 and higher correlation between SUVR and BP ND + 1 compared with using cerebellum, irrespective of SUVR [t 1 –t 2 ] interval. Smallest bias was observed with the [15–40]-min time window, in accordance with previous literature. The correlation between BP ND + 1 and SUVR was slightly better for the late time windows. Discriminant analysis between PD and HV using both PUT and CPR SUVRs showed an accuracy of ≥ 90%, for both reference regions and all studied time windows. Semi-quantitative 123 I-FP-CIT and 18 F-FE-PE2I values and relative decrease in the striatum for patients were highly correlated, with a higher effect size for 18 F- FE-PE2I for PUT and CPR SUVR. Conclusion 18 F-FE-PE2I is a suitable radioligand for in vivo DAT imaging with high discriminative power between early PD and healthy controls. Whereas a [15–40]-min window has lowest bias with respect to BP ND , a [50–60]-min time window at pseudoequilibrium can be advocated in terms of clinical feasibility with optimal discriminative power. The occipital cortex may be slightly preferable as reference region because of the higher time stability, stronger correlation of SUVR with BP ND + 1, and lower bias. Moreover, the data suggest that the diagnostic accuracy of a 10-min static 18 F-FE-PE2I scan is non-inferior compared with 123 I-FP-CIT scan used in standard clinical practice. Keywords Parkinson’ s disease . DAT (dopamine transporter) . PET (positron emission tomography) . 18 F-FE-PE2I . 123 I-FP-CIT Aline Delva and Donatienne Van Weehaeghe contributed equally to this work. This article is part of the Topical Collection on Neurology * Aline Delva aline.delva@kuleuven.be 1 Department of Neurosciences, KU Leuven, Leuven, Belgium 2 Department of Neurology, University Hospitals Leuven, Herestraat 49, Leuven 3000, Belgium 3 Division of Nuclear Medicine, University Hospitals Leuven, Herestraat 49, Leuven 3000, Belgium 4 Nuclear Medicine and Molecular Imaging, Department of Imaging and Pathology, KU Leuven, Leuven, Belgium European Journal of Nuclear Medicine and Molecular Imaging https://doi.org/10.1007/s00259-019-04587-y