Biomarkers in spinal fluid may flag frontotemporal dementia before symptoms emerge

A new study has revealed a set of proteins in the cerebrospinal fluid that could help identify frontotemporal dementia (FTD) before symptoms appear and track its progression. The findings, published in Nature Aging, come from scientists who analyzed over 4,000 proteins in people with inherited forms of FTD. The research points to disrupted gene expression and damaged brain connections as key features of the disease, offering potential new targets for diagnosis and treatment.

Frontotemporal dementia is a progressive brain disorder that primarily affects the frontal and temporal lobes, which are involved in behavior, personality, language, and movement. Unlike Alzheimer’s disease, which typically begins with memory loss, FTD often starts with changes in behavior, decision-making, or speech.

People with FTD may become socially inappropriate, emotionally detached, or struggle to speak and understand language. In some cases, the condition also causes movement problems similar to those seen in Parkinson’s disease. FTD can be caused by inherited genetic mutations, although it also occurs in people without a family history. There are currently no approved treatments to stop or slow the disease.

“FTD is the most common cause of dementia in people under 60, and it can have devastating effects on personality, behavior, language, and movement,” said study author Rowan Saloner, an assistant professor of neurology at the University of California, San Francisco.

“Despite its severity, FTD still has no approved treatments and lacks reliable biomarkers for diagnosis or monitoring. We focused on studying inherited forms of FTD, where we can determine the underlying brain pathology with high confidence even before symptoms begin. This provides a powerful model for discovering early biological changes in the brain by measuring molecules in cerebrospinal fluid.”

The study included 116 adults who carried mutations in one of three genes known to cause FTD—C9orf72, GRN, and MAPT. These participants were compared to 39 of their relatives who did not carry the mutations. Because these individuals came from the same families, they shared similar genetic backgrounds apart from the disease-causing mutations.

Cerebrospinal fluid samples were collected from all participants through a spinal tap. Researchers then used a highly sensitive technique known as aptamer-based proteomics to measure the levels of over 4,000 proteins in each sample. To make sense of this large dataset, they used a method called network analysis, grouping proteins into “modules” based on how their levels rose and fell together. This approach identified 31 distinct modules of co-expressed proteins, each representing a different biological process.

Several modules showed strong relationships with disease severity and progression. One key finding was that proteins involved in RNA splicing—the process by which genetic information is edited before being used to make proteins—were abnormally elevated in people with FTD, especially those with C9orf72 and GRN mutations. These changes were seen even before symptoms began in some individuals.

“We were struck by how early some of these protein changes appeared in FTD mutation carriers,” Saloner told PsyPost. “For example, certain proteins involved in RNA metabolism and synaptic ion transport were dysregulated in people who carried a disease-causing mutation but were still cognitively normal. Some spinal fluid protein signatures could even predict future cognitive decline in these cognitively normal individuals. This suggests that detectable molecular changes may begin years before clinical onset and could be used to track progression or even guide future preventive treatments.”

Other modules that stood out were related to the brain’s structural support system, known as the extracellular matrix, and to proteins involved in synaptic signaling and autophagy—the cellular process for clearing damaged components.

The researchers found that these protein modules were not only associated with current disease status but also with how quickly participants’ cognitive function declined over time. For example, lower levels of synaptic proteins predicted faster cognitive deterioration, while higher levels of RNA splicing proteins were linked to worse outcomes. This suggests that these proteins are not just markers of disease but may be part of the underlying mechanisms driving FTD.

To confirm that these findings were not limited to their initial sample, the team tested the same protein modules in two independent groups. One group included patients with a related neurodegenerative disease called progressive supranuclear palsy, and the other included people diagnosed with FTD, Alzheimer’s disease, or neither. Despite differences in disease type and the tools used to measure proteins, many of the same modules appeared in these new samples. This supports the idea that the protein changes seen in genetic FTD are relevant to broader groups of patients.

One particularly promising protein was NPTX2, a known marker of synaptic function. Its levels in cerebrospinal fluid strongly predicted cognitive decline—more so than any other individual protein in the study. Other proteins that stood out were involved in maintaining the structure and plasticity of brain cells or regulating gene expression from within the cell’s nucleus.

“By analyzing spinal fluid from people with genetic forms of FTD, we identified specific patterns of protein changes related to RNA processing, immune response, and synaptic health that were strongly associated with disease severity,” Saloner told PsyPost. “These protein ‘signatures’ help us understand in real time what’s going wrong in the brains of patients with FTD and could become the foundation for future biomarkers. Importantly, many of the same patterns were seen in people with sporadic (non-inherited) forms of FTD, suggesting our findings are relevant to a large portion of those affected by the disease.”

Despite its strengths, the study has some limitations. Collecting cerebrospinal fluid through a spinal tap is invasive and not suitable for large-scale screening. “To address this, we are actively screening blood samples, which are less invasive to collect, for biomarkers that perform similarly to those observed in spinal fluid,” Saloner said.

Another limitation is that the study primarily focused on genetic forms of FTD, which account for about 20 to 30 percent of cases. “We are actively working to identify protein changes that are specific to brain pathologies that uniquely cause sporadic forms of FTD,” Saloner added.

“Our ultimate goal is to develop reliable biomarkers that can identify different forms of FTD, track disease progression, and assess response to therapy. These tools are essential for designing better clinical trials and for moving toward precision medicine in neurodegenerative disease.”

“Unlike Alzheimer’s disease, which now has reliable fluid biomarkers and emerging treatments, FTD has neither,” Saloner noted. “Making progress toward biomarkers—especially ones that reflect the underlying biology of the disease—is a critical step toward changing clinical care for patients with FTD while also providing information about shared versus unique features across all neurodegenerative diseases.”

The study, “Large-scale network analysis of the cerebrospinal fluid proteome identifies molecular signatures of frontotemporal lobar degeneration,” was authored by Rowan Saloner, Adam M. Staffaroni, Eric B. Dammer, Erik C. B. Johnson, Emily W. Paolillo, Amy Wise, Hilary W. Heuer, Leah K. Forsberg, Argentina Lario-Lago, Julia D. Webb, Jacob W. Vogel, Alexander F. Santillo, Oskar Hansson, Joel H. Kramer, Bruce L. Miller, Jingyao Li, Joseph Loureiro, Rajeev Sivasankaran, Kathleen A. Worringer, Nicholas T. Seyfried, Jennifer S. Yokoyama, Salvatore Spina, Lea T. Grinberg, and William W. Seeley, on behalf of the ALLFTD Consortium.