A study led by researchers at the Johns Hopkins University School of Medicine in collaboration with the National Institutes of Health used a new imaging technique to show disruption of the blood-brain barrier (BBB) and predict progression of white matter hyperintensities (WMH) linked to vascular cognitive decline.

The study, published Oct. 21 in Annals of Neurology, followed patients who had experienced a stroke and underwent MRI scanning at baseline and one year later. By adapting a contrast-based MRI method originally used for perfusion imaging, the team created a way to measure BBB leakage using scans already standard in stroke care. The findings point to a promising imaging biomarker that could help identify patients at risk of vascular cognitive decline, and support more targeted clinical trials.

“After a stroke, a significant number of patients go on to develop cognitive issues in the months and years that follow,” says corresponding author Richard Leigh, M.D., associate professor of neurology at Johns Hopkins. “That risk is tied to ongoing vascular injury in the brain, and it’s something we urgently need better tools to detect and address early.”

WMHs are a common radiographic marker of cerebral small vessel disease, and strongly correlate with cognitive outcomes. However, the mechanisms driving lesion progression have remained uncertain. Because BBB disruption has long been suspected to play a role in the cascade leading from vascular injury to neurodegeneration, researchers used this adapted MRI technique to quantify leakage, rather than relying on new or specialized sequences.

“It’s a unique method that we developed at Hopkins,” Leigh says. “Because the images are contrast-based, if contrast leaks through the vessels, the picture changes. We found a way to isolate BBB disruption and quantify it.”

The team examined two regions of interest: existing WMHs and the surrounding normal-appearing white matter, known as the penumbra. BBB disruption within WMH at baseline was the strongest predictor of how much those lesions would expand, while disruption in the penumbra best predicted which patients would progress. Most participants showed some WMH progression over the year, while a smaller group remained stable or, in some cases, regressed.

This predictive ability underscores the growing role of imaging biomarkers in clinical research and patient care. Because the Johns Hopkins-developed method builds on a widely available scan, it could be adopted across diverse research settings.

Researchers say the approach could also accelerate the development of future therapies. By identifying high-risk patients, researchers can enrich trial populations and reduce the size and duration of studies.

“With a biomarker of disease activity, we can pinpoint those at greatest risk, giving us a way to measure how well potential treatments may work,” says Leigh. “If we can identify those patients earlier, we can design smarter trials, smaller studies, faster answers and a clearer path to finding effective therapies.”

While not yet validated for routine use, BBB disruption measures could eventually inform clinical decision-making and patient counseling. Researchers say the scan results may help strengthen patient conversations around prevention.

“I see a lot of stroke patients who are starting to notice cognitive changes,” Leigh says. “If I can point to something measurable on their scan, it helps underline why prevention and risk-factor control matter.”

The research team is now focused on translating these findings into clinical trials. The next step is to use the biomarker to identify high-risk patients and test candidate therapies aimed at slowing or preventing disease progression. Because anti-inflammatory therapies already exist, including monoclonal antibodies and other agents that target pathways implicated in BBB dysfunction, the team plans to incorporate the biomarker into small, short-term studies to screen these drugs before larger trials.

“We want to use this biomarker to make trials smarter and faster,” Leigh says. “If we can find the right patients and the right drugs, we can move much more quickly toward real treatments.”

Other contributors to the study include Kyle Kern, Nae-Yuh Wang, Rebecca Gottesman and Clinton Wright.

No authors reported conflicts of interest.