INVESTIGATING THE CLINICAL IMPACT OF [ 18 F]FLUTEMETAMOL PET IN A TERTIARY MEMORY CLINIC SETTING IN PATIENTS WITH UNCERTAIN DIAGNOSIS

Since the first application of carbon-11 Pittsburgh Compound-B ([11C]PIB) PET more than a decade ago, amyloid PET has been instrumental in advancing the research agenda for Alzheimer's disease (AD). By detecting a core feature of AD pathology, amyloid PET holds potential in clinical settings, particularly given the high rate of misdiagnosis. Clinical studies using recently approved amyloid PET tracers, however, have far been few, with these mainly in highly selected research cohorts. The aim of the present study was thus to investigate the added clinical value of [18F]flutemetamol PET in memory clinic patients whose diagnosis remained uncertain following routine clinical work up. 135 patients were included from the Department of Geriatric Medicine, Karolinska University Hospital, Huddinge, Sweden, following referral from primary care physicians. Following standard diagnostic workup, including medical and neurological examination, clinical chemistry (including CSF Aβ1–42), rating batteries for depression and neuropsychiatric symptoms, neuropsychological assessment, and structural imaging, the clinical picture remained unclear. [18F]flutemetamol PET was thus performed, using a Biograph mCT PET/CT (Siemens/CTI, Knoxville, TN), with the acquisition protocol consisting of a static 20-min scan, 90-min post-injection of 185 MBq. In addition to visual assessment by an experienced nuclear medicine physician, [18F]flutemetamol uptake was quantified on a region of interest basis using a fully automated software (Hermes Hybrid BRASS). Diagnoses before and after [18F]flutemetamol investigations were reached using a multidisciplinary consensus based approach. Based on preliminary results from 61 subjects (38 MCI, 13 AD, 6 dementia NOS, 2 SCI, one FTD, and one DLB), [18F]flutemetamol investigations led to a change in diagnosis in 68% of patients. Agreement between visual and quantitative assessment of [18F]flutemetamol images was high (89%). Concordance between CSF Aβ1–42 (<550 pg/mL) and [18F]flutemetamol was 57% and 53%, using visual and quantitative approaches, respectively. Among discordant cases, most showed isolated [18F]flutemetamol positivity (75% using visual, 77% using quantification). While further analyses are ongoing for the remaining 74 patients, preliminary findings highlight the added value of [18F]flutemetamol over standard diagnostic work-up. Discordance between CSF Aβ1–42 and [18F]flutemetamol PET highlights the issue of biomarker interchangeability in clinical settings.