Circulating Biomarkers Linked to Stroke

Pauline Anderson

August 25, 2016

Elevated levels of four biomarkers —– C-reactive protein (CRP), tumor necrosis factor receptor 2 (TNFR2), homocysteine, and vascular endothelial growth factor (VEGF) — increase the risk for incident ischemic stroke (ISS), according to a new analysis.

The study showed that adding these markers to those already included in an existing model (the Framingham Stroke Risk Profile) increases the ability to predict stroke.

"What's unique about this study is that it shows that these markers are predictive of having a first stroke — so an incident stroke — above and beyond the other classical risk factors," including high blood pressure and diabetes, said lead study author Ashkan Shoamanesh, MD, assistant professor of medicine (neurology) and director, Stroke Fellowship Program, McMaster University, Hamilton, Ontario, Canada.

The study was published online August 24 in Neurology.

The analysis included 3224 participants in the Framingham Offspring Study attending the cycle 7 examination (1998 to 2001), which included measurement of 15 biomarkers of inflammation and endothelial dysfunction. The mean age of participants was 61 years, and 54% were women.

Of these participants, 98 had an ischemic stroke during a mean follow-up of 9.3 years.

To facilitate comparison, researchers standardized all biomarkers to a mean of 0 and a standard deviation (SD) of 1. They calculated hazard ratios (HRs) for ischemic stroke per 1-SD increment of each biomarker.

In a model that controlled for age and sex, they found that per SD increment, CRP (HR, 1.28; 95% confidence interval [CI], 1.04 - 1.56; P = .02), TNFR2 (HR, 1.33; 95% CI, 1.09 - 1.63; P = .005), total homocysteine (HR, 1.32; 95% CI, 1.11 - 1.58; P = .002), and VEGF (HR, 1.25; 95% CI, 1.07 - 1.46; P = .005) were all associated with ischemic stroke risk.

A model that also controlled for the other variables of the Framingham Stroke Risk Profile — systolic blood pressure, hypertension treatment, smoking, diabetes, cardiovascular disease, and atrial fibrillation — found that all biomarkers remained significant, except for CRP.

The HRs for the biomarkers were similar across individual ischemic stroke subtypes.

Predictive Ability

Adding these four biomarkers to the Framingham Stroke Risk Profile improved its ability to predict ischemic stroke risk, with a continuous net reclassification improvement (NRI) of 0.34 (95% CI, 0.12 - 0.57; P < .05).

That doesn't mean that the prediction is improved by 34%, noted Dr Shoamanesh. He explained that the NRI "doesn't provide a proportional figure, so it's difficult to properly quantify the degree of improvement," but it does provide "an estimate of the degree to which the strength of the score improves."

VEGF had the greatest individual degree of discrimination for future ischemic stroke. But the story with VEGF is somewhat complicated.

Dr Shoamanesh pointed out that some research has shown that increased VEGF levels can improve cerebral circulation and protect neurons. On the other hand, evidence suggests that high levels of VEGF increase the chance of developing atherosclerotic disease. A study of patients with macular degeneration showed those who took a VEGF inhibitor did not have a lower risk for stroke.

"That study was observational and had limited power to detect a meaningfully lower level of stroke, but there was no signal to suggest that being on a VEGF inhibitor for macular degeneration also provided additional protection for ischemic stroke."

That observation, the authors write, "would suggest an indirect rather than causal link between VEGF and future stroke."

Total homocysteine, which accelerates atherosclerotic disease, is perhaps a more established marker of increased stroke risk, as is CRP, which flags systemic inflammation and plaque instability.

TNFR2 appears to be the new kid on the biomarker block. "This is the first time that I'm aware of that any study has shown that it's a risk factor for first ischemic stroke," said Dr Shoamanesh.

He pointed out that patients with rheumatoid arthritis receiving a TNF-α inhibitor have been found to have lower vascular events.

Dr Shoamanesh stressed that the new study is observational, so none of the markers can be said to cause stroke. Also, he said, the study needs to be validated.

In the meantime, it would be "premature" to use the data from the study in clinical practice, he said.

"To really begin using these makers in a clinical setting, you would need to show that you could impact outcomes by some sort of intervention, and we aren't nearly there yet."

However, that's the ultimate aim, and this new study is a step in that direction.

"The more we are able to identify patients who are at greatest risk for future stroke, the more we would be able to test those particular individuals for the benefit of existing as well as emerging stroke preventive therapies," said Dr Shoamanesh.

He and his coauthors listed several limitations to their research. For example, the sample is almost entirely of European descent, and researchers didn't account for factors that could have affected biomarker measurement, such as infections, renal impairment, inflammatory or rheumatologic disease, and malignancies.

In addition, the study didn't consider preventive medications, such as statins and antithrombotic therapy, that might have affected stroke risk. Researchers also measured biomarkers only once, so there's no information on changes in concentration over time.

Pierre Fayad, MD, professor, Department of Neurological Sciences, director, Nebraska Stroke Center, University of Nebraska Medical Center, Omaha, and member of the American Academy of Neurology, noted additional limitations. The number of patients with stroke was relatively small, and, as well as not considering treatments that may lower stroke risk, the authors also didn't factor in how tightly diabetes was controlled.

"We don't know even what the blood pressure was lowered to," said Dr Fayad, adding that the ideal blood pressure level is "still a topic of discussion" when it comes to stroke risk.

It's very important to consider these limitations, he said. "The information we have is still not strong enough to tell us that we should start getting these lab tests for all our patients."

That, he said, would be costly for the healthcare system and would entail significant discomfort for patients. "And then, we don't yet know what to do with that information."

Dr Fayad praised the research design, saying that the Framingham study "has certainly given us a lot of information" on cardiovascular risks. And he described the attempt by the authors to improve the current stroke prediction model as "valiant."

But he stressed the importance of determining what level of these biomarkers indicates high risk. "Once we know what the threshold is, then we need to see whether or not by doing an intervention we can lower that risk."

The results are "starting to show some trends" but need to be replicated, said Dr Fayad.

This need was echoed in an accompanying editorial by Stephen R. Williams, PhD, Department of Neurology, University of Virginia, Charlottesville, and Svetlana Lorenzano, MD, PhD, Department of Neurology and Psychiatry, Sapienza University, Rome, Italy.

The study "provides a platform to formulate future studies utilizing the full potential of circulating biomarkers, particularly in the related but independent area of primary prevention," write the editorialists.

For example, they say, "further prospective studies should validate sensitivity and specificity and replicate IIS-risk predictive performance of these biomarkers in the general population."

Large nonobservational studies assessing the causal relationship of biomarker levels and stroke occurrence are needed, they added, "to advance our knowledge and identify new molecular targets for translational research and development of novel specific management and therapeutic strategies."

It would be "of interest" if the authors used genetic data to investigate not only the influence on the levels of these 15 biomarkers but also how genetic variants that are associated with biomarker levels directly influence risk for IIS, the editorialists write.

"This could be a powerful approach, which in combination with the Framingham Stroke Risk Profile score, will be increasingly applicable as genomic information is transitioned into clinical care."

The study was conducted by researchers at the Boston University School of Medicine and Public Health. The study was supported by the Framingham Heart Study's National Heart, Lung, and Blood Institute and by grants from the National Institute of Neurologic Disorders and Stroke, the National Institute on Aging, and the National Institutes of Health. Dr Shoamanesh is supported by the Marta and Boris Char in Stroke Research and Care. Dr Fayad and the editorial writers have disclosed no relevant financial relationships.

Neurology. Published online August 24, 2016. Abstract Editorial

For more Medscape Neurology news, join us on Facebook and Twitter


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.