Head-to-Head Trial Finds Two Tau PET Scans Do Not Always Agree

A multicentre trial that scanned the same 682 people with two tau PET tracers found the scans agreed on tau status most of the time but disagreed often enough to change who would count as tau-positive, a distinction that can decide trial eligibility and treatment access. Researchers say the finding does not tell them which tracer is right, since no independent test of tau in a living brain exists to check either one against.

A prospective, multicentre trial that scanned 682 people with two different tau PET tracers within a 45-day window found the tracers agreed on tau-positive status only 87 percent of the time in the medial temporal lobe and 94 percent of the time across the wider cortex, according to results published in The Lancet on May 30, 2026. The study, run by researchers at the University of Pittsburgh and collaborating sites across North America as part of the HEAD study, compared flortaucipir (marketed as Tauvid, the tracer approved by the US Food and Drug Administration and used clinically in the United States and Europe) with MK6240, an investigational tracer used mainly in research and clinical trials. Neither tracer was tested against tissue samples in this study.

The two scans did not just disagree at random. Among cognitively unimpaired participants who were amyloid-positive, MK6240 identified roughly twice as many people as tau-positive in the medial temporal lobe as flortaucipir did (15 percent versus 6 percent). Among people with cognitive impairment, MK6240 flagged 28 percent as tau-positive versus 16 percent for flortaucipir. Of the 84 people whose two scans disagreed on medial temporal lobe status, 75 (89 percent) were positive on MK6240 only, and this group had significantly higher amyloid burden, more APOE epsilon4 gene carriers, and worse cognitive scores than people who tested negative on both tracers, the authors reported. MK6240 also showed higher accuracy than flortaucipir at distinguishing Alzheimer's disease from other causes of cognitive impairment (area under the curve 0.93 versus 0.86, p<0.0001).

The authors frame this as evidence that tracer choice, not just underlying biology, can determine whether a person is classified as having tau pathology, with direct consequences for who is offered anti-amyloid drug therapy or enrolled in a trial. But the study cannot say which tracer's readout is closer to the truth. There is no living-brain reference standard for regional tau burden that either scan can be checked against, so a discordant case being 'MK6240-positive only' cannot be read as MK6240 catching disease that flortaucipir missed, only as the two agents producing different verdicts in the same brain at the same time.

The findings add to a broader push within the HEAD study and related work to compare tau PET tracers directly in the same participants, rather than relying on separate cohorts scanned with different agents. The authors say the disagreement matters most in early and borderline cases, such as mild cognitive impairment, where a positive or negative tau PET reading can determine trial eligibility.