Amanda Rose Nguyen stands at the intersection of brain imaging and precision medicine, armed with a discovery that could spare thousands from an ineffective treatment. At the University of California, Los Angeles, the clinical fellow in nuclear medicine has demonstrated something crucial: not everyone diagnosed with Alzheimer's disease actually has Alzheimer's disease—and their brain scans can prove it.
The problem is both simple and profound. Two new anti-amyloid therapies targeting the hallmark plaques of Alzheimer's disease were recently approved by the FDA, and they work—but only for some patients. Many people meeting the clinical criteria for Alzheimer's actually harbor different pathologies underneath: Lewy body disease, limbic-predominant age-related TDP-43 encephalopathy, or frontotemporal lobar degeneration. When you treat the wrong disease with the wrong drug, outcomes plummet. Patients get sicker, face unnecessary side effects, and drain resources on futile interventions.
Nguyen's team examined 124 consecutive patients evaluated for amyloid immunotherapy at UCLA, analyzing their brain metabolic patterns using 18F-FDG PET scans—a nuclear imaging technique that lights up how actively different brain regions are working. They categorized each patient's metabolism as consistent or inconsistent with Alzheimer's disease, then tracked what happened over at least one year of treatment. The results were striking. Patients whose brain metabolism patterns actually matched Alzheimer's disease saw their cognitive performance scores improve or stabilize. Everyone else—those whose scans revealed Lewy body disease, TDP-43 pathology, or other conditions—experienced significant cognitive decline despite receiving anti-amyloid therapy.
The implication is transformative. This wasn't a laboratory curiosity presented at a small conference. The Society of Nuclear Medicine and Molecular Imaging chose Nguyen's work as the Abstract of the Year from nearly 1,500 submissions to their 2026 Annual Meeting—recognition reserved for research that exemplifies the field's most promising advances.
"This work demonstrates that 18F-FDG PET is an important tool in the diagnosis of dementia," Nguyen said, articulating what the data makes clear: brain imaging matters. Armed with metabolic data from these scans, physicians can make treatment decisions with unprecedented precision. Those most likely to benefit from expensive, complex anti-amyloid therapy get it. Those who won't benefit are spared from it—avoiding adverse effects, unnecessary expense, and the demoralization of watching their condition worsen under ineffective treatment.
The path forward is already visible. Nguyen and her team are conducting larger, higher-powered analyses with an expanded patient sample, expected by year-end, that will further define how reliably brain metabolism predicts treatment success. In the interim, she recommends that physicians gather comprehensive neuroimaging data on an individual patient basis before prescribing these treatments.
What makes this discovery hopeful isn't just the science—it's the recognition that precision medicine in dementia care is no longer theoretical. It's happening now, in hospitals, with tools we already have. The barrier isn't technology; it's integration. When doctors use what they can see on a PET scan to inform what they prescribe, patients get the right treatment, at the right time, with far better odds of actually getting better.