Combining Biomarkers May Help to Predict Cognitive Impairment in MS
Both blood and imaging biomarkers used in study of early MS
Combining blood and imaging biomarkers might help clinicians better predict cognitive impairment in people with early multiple sclerosis (MS) than using either one alone, a new study suggests.
Researchers found that using the two together worked better to predict information processing speed than did either blood or MRI biomarkers individually. The new prediction model accounted for both blood levels of the neurofilament light chain (NfL) protein and MRI measurements of lesions and brain volumes.
“Prognostic models based on combinations of biomarkers such as the one introduced here may improve the identification of multiple sclerosis patients who are likely to experience cognitive decline,” the researchers wrote, adding, “Our findings demonstrate that combining blood and imaging measures improves the accuracy of predicting cognitive impairment.”
The study, “Improved prediction of early cognitive impairment in multiple sclerosis combining blood and imaging biomarkers,” was published in Brain Communications.
The degree of a patient’s MS-associated disability is often determined based on motor symptoms. But cognitive impairments, including slowed information processing and attention and memory problems, are common in MS, and begin to emerge in early disease stages.
It is important, therefore, to monitor cognition in MS patients so as to determine the best course of treatment for each individual, and to avoid significant impacts on quality of life.
Predicting cognitive impairment in MS
Usually, this monitoring occurs through a neuropsychological evaluation involving a range of cognitive tests. But these tests are time consuming and may not be generally available in all clinics, making it important to identify biomarkers of cognitive impairment that can be readily measured.
“However, a substantial gap still exists between candidate biomarkers, validated biomarkers and clinically useful biomarkers in multiple sclerosis,” the researchers wrote.
Levels of NfL in the serum — a liquid component of blood — also have received recent attention as a biomarker of nerve cell injury across a range of neurodegenerative conditions. In MS, serum NfL, or sNfL, is increased during relapse, and its levels correlate with new brain lesions. But less is known about the association between sNfL and cognition.
For a disease as complex as MS, a combination of imaging and blood-based measures may be necessary to best predict cognitive problems, the researchers hypothesized.
Thus, the team — based at the University Medical Center of the Johannes Gutenberg University Mainz, in Germany — aimed to further examine the usefulness of blood sNfL alone and in combination with MRI measures for predicting cognitive impairment in people with early MS.
The mean age of all patients was 33 years, and 70.4% were female. Most (71.1%) were using a disease-modifying treatment (DMT) at the time they were included in the study. Their median score on the Expanded Disability Status Scale (EDSS) — an assessment of disability progression — was 1, reflecting minimal symptoms only.
Participants underwent a battery of tests to evaluate cognitive performance and mood, and blood samples were collected to measure sNfL.
Results showed that sNfL levels were correlated with worse information processing speed after adjusting for potentially influential factors, including age, sex, disease duration, DMT use, and EDSS scores. Information processing speed was measured with the Symbol Digit Modalities Test (SDMT).
However, levels of NfL were not significantly associated with other measures of cognitive performance or self-reported fatigue, depression, or anxiety.
SDMT performance also was significantly correlated with MRI measures, including lesion volume and gray matter volume — two markers of neurodegeneration. In particular, better SDMT performance was linked with a higher gray matter volume and lower lesion volume.
“Taken together, sNfL, [lesion volume] and [gray matter volume] in patients with early multiple sclerosis significantly correlate with information processing speed (SDMT) test performance, whereas tests for memory, learning, affective parameters and fatigue did not show an association with sNfL levels,” the researchers wrote.
Combining multiple biomarkers
To find out if a combination of all three markers could predict SDMT performance, researchers generated an artificial intelligence-based prediction model based on sNfL, lesion volume, and gray matter volume. Alone, sNfL was 75.6% accurate at predicting SDMT score, whereas lesion volume had a 68.6% accuracy and gray matter volume had 69.1% accuracy.
Combining all three measures led to the highest accuracy, achieving a prediction that was 88.7% accurate.
To confirm these findings, the scientists examined these measures in an independent group of 101 early MS patients, 15 of whom had CIS and 86 RRMS. Again, combining all three methods led to better accuracy at predicting SDMT performance (90.8%) than any of the markers alone.
“By utilizing three biomarkers we obtained a robust prediction model for cognitive impairment in early multiple sclerosis,” the researchers wrote.
The researchers noted, however, that the study is limited by taking only a single sNfL measurement rather than assessing changes over time. Such longer-term changes should be assessed in future studies, they said.
Nevertheless, “our model might provide a surrogate screening tool for cognitive decline, whereby multiple sclerosis patients at risk for cognitive impairment may be selected for in-depth neuropsychological testing,” the team wrote.