Measuring the brain lags the biology it tries to stage

Two tracers, two stages

A staging tool that gives different answers depending on which tracer is used cannot be treated as a neutral ruler, and a preprint on tau PET imaging makes that problem concrete. The authors find that two tracers, Flortaucipir and MK6240, are not interchangeable for Braak staging across the Alzheimer disease continuum. MK6240 systematically assigns higher stages than Flortaucipir in over 90 percent of discordant cases. For clinical trials, the consequence is direct: a patient staged with one tracer may not be comparable to a patient staged with the other, and pooling them without harmonization risks comparing groups that were never matched. m:…9003

The gap widens when imaging is compared with neuropathology. The preprint reports that PET-based Braak staging lags behind neuropathological staging by approximately 2 to 3 stages in early Alzheimer disease, suggesting current PET imaging systematically underestimates tau pathology at the stage when intervention is most often contemplated. The honest reading of a PET stage is therefore as a tracer-specific estimate, not as a direct readout of pathology, and trial designs should either standardize the tracer or account for the discordance explicitly. m:…9003

A small stimulation signal on cognition

On the intervention side, a preprint on Parkinson disease reports a small but carefully described result. Combining continuous stimulation of the subthalamic nucleus with intermittent stimulation of the nucleus basalis of Meynert produced clinically meaningful cognitive improvements in all five Parkinson patients studied. Two surgical trajectories for placing the nucleus basalis leads, vertical and lateral, were implemented without adverse effects, and automated programming using a three-dimensional algorithm targeted the nucleus based on modeled volumes of tissue activated. The sample is tiny and there is no control group, so the result cannot establish efficacy, and the durability of the cognitive benefit remains unknown. What it does offer is a testable protocol: a defined combination, defined trajectories, and automated targeting that a larger controlled study could replicate or refute. m:…6754

Two Parkinson mechanisms converge on damage control

Two mechanistic preprints point at different ways dopamine neurons lose function, and both name a handle for intervention. In dopamine neurons derived from stem cells carrying a triplication of the alpha-synuclein gene, elevated alpha-synuclein increases basal mitophagy and mitochondrial mass loss. The authors find this pathological mitophagy does not depend on the PINK1-Parkin pathway, and that increased fission driven by the protein DRP1 contributes to mitochondrial dysfunction. Pharmacological inhibition of DRP1 rescues neuronal firing and cellular bioenergetics in these neurons. The result is confined to a cell model without clinical validation, and no specific DRP1 inhibitor compound for clinical use is named, so it should be read as a mechanism and a target class, not as a treatment. m:…8121

The second mechanism concerns LRRK2, a protein whose variants are linked to Parkinson disease. The preprint describes autoinhibition in which leucine-rich repeats block the kinase active site, and finds that disease-associated variants preferentially unlatch the kinase domain, priming it for activity. Unlatched LRRK2 localizes more to membranes where its Rab substrates reside, and decreased phosphorylation at S935 serves as a biomarker for the unlatched, active conformation. The biomarker claim still lacks clinical validation in patient-derived tissue, and not all rare variants have been mapped to the model. Even so, the two Parkinson preprints share a useful shape: a specific molecular event, excess fission in one case and unlatching in the other, with a measurable readout that a future study could test. The field's next step is to close the same gap the Alzheimer preprint exposes, by validating measurement tools in patients at the same pace as mechanisms in models. m:…9829