PT - JOURNAL ARTICLE AU - Alexandra Gogola AU - Davneet Singh Minhas AU - Victor L Villemagne AU - Ann D Cohen AU - James M. Mountz AU - Tharick Ali Pascoal AU - Charles M Laymon AU - N. Scott Mason AU - Milos D. Ikonomovic AU - Chester A Mathis AU - Beth E. Snitz AU - Oscar L. Lopez AU - William E Klunk AU - Brian J Lopresti TI - Direct comparison of the tau PET tracers [<sup>18</sup>F]flortaucipir and [<sup>18</sup>F]MK-6240 in human subjects AID - 10.2967/jnumed.120.254961 DP - 2021 Apr 01 TA - Journal of Nuclear Medicine PG - jnumed.120.254961 4099 - http://jnm.snmjournals.org/content/early/2021/04/15/jnumed.120.254961.short 4100 - http://jnm.snmjournals.org/content/early/2021/04/15/jnumed.120.254961.full AB - Tau PET tracers exhibit varying levels of specific signal and distinct off-target binding patterns that are more diverse than amyloid PET tracers. This study compares two frequently used tau PET tracers, [18F]flortaucipir (FTP) and [18F]MK-6240, in the same subjects. METHODS: [18F]flortaucipir and [18F]MK-6240 scans were collected within two months in 15 elderly subjects varying in terms of clinical diagnosis and cognition. FreeSurfer v5.3 was applied to 3T MR images to segment Braak pathologic regions (I-VI) for PET analyses. Off-target binding was assessed in choroid plexus, meninges, and striatum. SUVR outcomes were determined over 80-100 min ([18F]flortaucipir) or 70-90 min ([18F]MK-6240) normalized to cerebellar grey matter. Blinded visual interpretation of images was performed by five raters for both medial temporal lobe (MTL) and neocortex (NEO) and an overall (majority) rating determined. RESULTS: Overall visual ratings showed complete concordance between radiotracers for both MTL and NEO. SUVR outcomes were highly correlated (r2&gt;0.92; P &lt;&lt; 0.001) for all Braak regions except Braak II. The dynamic range of SUVR values in target regions was approximately two-fold higher for [18F]MK-6240 compared to [18F]flortaucipir. Cerebellar SUV values were similar for [18F]MK-6240 and [18F]flortaucipir, suggesting that differences in SUVR values are driven by specific signal. Apparent off-target binding in striatum and choroid plexus was often observed with [18F]flortaucipir, and most often in meninges with [18F]MK-6240. CONCLUSION: Both [18F]MK-6240 and [18F]flortaucipir are capable of quantifying signal in a common set of brain regions that develop tau pathology in AD and perform equally well in visual interpretations. Each also shows distinct patterns of apparent off-target binding. [18F]MK-6240 showed greater dynamic range in SUVR estimates, which may be an advantage for detecting very early signal or in longitudinal studies designed to detect small interval changes.