Brain Amyloid Scans Reveal Surprising Clues About Severe Cognitive Impairment in Treatment-Resistant Schizophrenia

For roughly three in ten people living with schizophrenia, standard antipsychotic treatments simply fail. This stubborn condition, known as treatment-resistant schizophrenia, has long puzzled clinicians and researchers alike. Beyond the hallmark hallucinations and delusions that resist medication, many of these patients carry an additional burden: profound cognitive deficits affecting memory, attention, and executive function. Now, a team of Italian researchers has taken an unprecedented look inside the living brains of these patients, using a molecular imaging technique borrowed from Alzheimer’s disease research to ask a deceptively simple question: could the cognitive decline seen in treatment-resistant schizophrenia share biological roots with early dementia?

The study, led by Mariateresa Ciccarelli and colleagues at the University of Naples Federico II, represents one of the first attempts to directly visualize amyloid-beta protein deposition in the brains of patients with treatment-resistant schizophrenia who also show cognitive impairment resembling mild cognitive impairment, or MCI. This clinical profile, characterized by cognitive difficulties that exceed normal aging but fall short of dementia, has long been recognized in Alzheimer’s disease as a potential precursor stage. Yet the overlap between schizophrenia’s cognitive symptoms and MCI has remained largely unexplored at the molecular level, leaving researchers uncertain whether the two conditions might converge on shared pathological mechanisms.

To probe this question, the research team recruited twelve patients diagnosed with treatment-resistant schizophrenia who exhibited MCI-like cognitive impairment, a group they designated TRS-MCI. Each participant underwent a positron emission tomography scan using a radioactive tracer called 18F-Flutemetamol, a compound specifically designed to bind to amyloid-beta plaques in the brain. This tracer, already approved for clinical use in dementia evaluation, emits a detectable signal when it attaches to amyloid deposits, allowing researchers to map their distribution across cortical regions with remarkable precision. The patients also completed comprehensive clinical assessments, including the Positive and Negative Syndrome Scale to measure symptom severity and the Brief Assessment of Cognition in Schizophrenia to quantify cognitive performance across multiple domains.

The comparison group consisted of twelve patients with cognitive impairment but no psychiatric disorders, all of whom had previously been confirmed amyloid-negative through the same imaging approach. This control group proved essential for interpreting the schizophrenia findings, as it provided a benchmark for what amyloid-negative cognitive impairment actually looks like on a Flutemetamol PET scan. By contrasting the two groups, the researchers could determine whether treatment-resistant schizophrenia produces any detectable amyloid signal above and beyond what occurs in non-psychiatric cognitive decline.

The imaging analysis employed sophisticated software called Diagnosis ORIented ANalysis, which generates multiple semiquantitative indices of regional cortical amyloid signal. Rather than relying on a single global measure, this approach allowed the team to examine amyloid distribution across specific brain regions, potentially revealing patterns invisible to conventional whole-brain summaries. Both expert visual assessment by trained nuclear medicine physicians and semi-automated quantification were used, providing complementary perspectives on the same data.

The central finding was striking in its clarity: no significant differences emerged between the treatment-resistant schizophrenia group and the amyloid-negative comparison group. Neither the expert visual reads nor the quantitative measures detected elevated amyloid deposition in the schizophrenia patients. This null result carries substantial weight, effectively ruling out the hypothesis that severe cognitive impairment in treatment-resistant schizophrenia reflects occult Alzheimer’s-type amyloid pathology. For clinicians, this suggests that cognitive deficits in these patients arise from fundamentally different biological processes than those driving neurodegenerative dementia.

Yet the study was not without its surprises. When the researchers explored correlations within the schizophrenia group itself, they uncovered an unexpected pattern: regional amyloid composite scores in the right parietal cortex showed a negative association with Positive and Negative Syndrome Scale positive scores, with statistical significance reaching a threshold of P less than or equal to 0.01. In other words, patients with more severe positive symptoms, such as hallucinations and delusions, tended to show lower Flutemetamol signal in this specific cortical region, even though all values remained within the amyloid-negative range.

This inverse relationship raises intriguing questions about the neurobiology of treatment-resistant schizophrenia. The right parietal cortex plays critical roles in spatial attention, body awareness, and the integration of sensory information, functions that are frequently disrupted in psychotic states. The finding that Flutemetamol signal varies inversely with positive symptom severity within the amyloid-negative range suggests that the tracer may be picking up on non-amyloid molecular features of parietal cortex biology, possibly reflecting differences in synaptic density, neuroinflammation, or protein turnover that correlate with psychotic symptom intensity. Alternatively, the pattern could reflect compensatory mechanisms or developmental differences that shape both parietal function and vulnerability to positive symptoms.

The researchers are careful to frame these findings as preliminary. With only twelve patients in each group, the study lacks the statistical power to detect subtle group differences, and the correlational nature of the parietal finding means causality cannot be established. The cross-sectional design also prevents any conclusion about whether the observed signal variations change over time or predict clinical trajectory. Nevertheless, the proof-of-concept value is considerable: the study demonstrates that amyloid PET imaging is feasible in treatment-resistant schizophrenia populations and that regional signal variations, even within the amyloid-negative range, may carry meaningful clinical information.

Looking forward, the research team emphasizes the need for larger, longitudinal cohorts to confirm and extend these preliminary observations. If the parietal cortex signal association with positive symptoms holds up in bigger samples, it could eventually inform a new generation of imaging biomarkers for treatment-resistant schizophrenia, helping clinicians identify which patients are most likely to experience severe cognitive decline and guiding the development of targeted interventions. The convergence between schizophrenia research and dementia imaging techniques also opens a broader avenue for cross-pollination between these historically separate fields, suggesting that molecular tools developed for neurodegenerative disease may illuminate psychiatric illness in unexpected ways.

Subject of Research: Amyloid-beta PET imaging of cognitive impairment in treatment-resistant schizophrenia

Article Title: Probing severe cognitive impairment in treatment resistant schizophrenia: a proof-of-concept in vivo study with (18F)-Flutemetamol PET

Article References: Ciccarelli, M., Gaudieri, V., Mazza, B., Barone, A., Iasevoli, F., Vellucci, L., De Simone, G., Cuocolo, A., Pappatà, S., & de Bartolomeis, A. (2026). Probing severe cognitive impairment in treatment resistant schizophrenia: a proof-of-concept in vivo study with (18F)-Flutemetamol PET. Schizophrenia. https://doi.org/10.1038/s41537-026-00801-7

Image Credits: AI Generated

DOI: 10.1038/s41537-026-00801-7

Keywords: treatment-resistant schizophrenia, amyloid-beta, 18F-Flutemetamol, PET imaging, mild cognitive impairment, parietal cortex, PANSS, cognitive deficits, biomarkers, neuroimaging, psychosis, Alzheimer's disease