Prognostic models based on several biomarkers can help predict whether an individual patient with mild cognitive impairment (MCI) may develop Alzheimer’s disease (AD) dementia, researchers report.
They note these “practical models” could support clinical decision making and facilitate application of MRI and cerebrospinal fluid (CSF) biomarkers in daily practice.
The research was published online October 16 in JAMA Neurology.
Beyond Group Statistics
This work helps clinicians and patients in several ways, first author, Ingrid van Maurik, MSc, from Amsterdam Neuroscience, VU University Medical Center, Amsterdam, the Netherlands, told Medscape Medical News.
“It provides clinicians and patients with more specific information, instead of group statistics. It aids clinicians to understand better what biomarker results mean and thereby could improve communication between clinicians and patients. And, as soon as future disease-modifying treatment becomes available, this tool will have high relevance,” van Maurik explained.
The investigators used data from 525 patients with MCI (40% female; mean age, 63 years) from the longitudinal Amsterdam Dementia Cohort to construct prognostic models that estimate individualized probabilities of progression to AD dementia.
The models are based on patient characteristics (age, sex, and Mini-Mental State Examination [MMSE] score), MRI biomarkers (hippocampal volume, normalized whole-brain volume), CSF biomarkers (amyloid-β1-42, tau) and combined biomarkers. All participants had a baseline visit to a memory clinic from September 1997 through August 2014.
Over an average of 2.4 years, 38.3% of patients progressed to AD dementia, 9.9% developed another type of dementia, and 51.8% remained stable.
In a model based on sex, age, and MMSE score, the 3-year risk for progression ranged from 26% in younger men with an MMSE score of 29 to 76% in older women with an MMSE score of 24. The 1-year progression risk ranged from 6% in younger men with an MMSE score of 29 to 24% in older women with an MMSE score of 24.
The 3-year and 1-year progression risks were 86% and 27%, respectively, when MRI results were abnormal, 82% and 26% when CSF test results were abnormal, and 89% and 26% when both MRI and CSF results were abnormal.
Conversely, 3- and 1-year progression risks were 18% and 3% when MRI test results were normal, 6% and 1% when CSF results were normal, and 4% and 0.5% when both MRI and CSF results were normal.
“The prognostic value of models determining any type of dementia were in the same order of magnitude although somewhat lower,” the researchers say. “The models show particularly high negative predictive values and external validation showed that our models were highly robust.”
Ready for the Clinic?
van Maurik believes the models are ready for clinical use.
“On request, we provide a spreadsheet calculator for academic use that only requires the user to fill in a few patient characteristics and biomarker values,” she told Medscape Medical News.
“Although the models are validated in an independent cohort, we do think that improvements could be made and that we should still be cautious while using it. The generalizability of the models is restricted to the use of identical methods for CSF analysis and volumetric MRI measurements. For that reason, we are currently working on a ‘supra model’ to enhance generalizability,” van Maurik added.
Commenting on the research for Medscape Medical News, Marc Gordon, MD, chair of neurology at Zucker Hillside Hospital in Glen Oaks, New York, said, the biomarkers in the study are not novel.
“They are very well validated biomarkers. What’s interesting is not the specific biomarkers but the concept of trying to take it to the next step and move biomarkers from population-based to individual prognostic-based.”
Dr Gordon said there are challenges in using these models. As part of the Amsterdam Dementia Cohort, participants received a standardized diagnostic workup: MRI and lumbar puncture. “I think in this country, for really unwarranted reasons, lumbar punctures are unpopular. I think people are unrealistically afraid of lumbar puncture even though clearly that’s the best validated biomarker,” he commented.
“What will be challenging,” said Dr Gordon, “is to try and standardize imaging across multiple machines, having people willing to do lumbar punctures and how those lumbar punctures will be processed differentially and making sure there is a standardized process, which is a big challenge in biomarkers in general.”
The study was supported by the National Institutes of Health and the Department of Defense to the Alzheimer’s Disease Neuroimaging Initiative. Ms van Maurik has disclosed no relevant financial relationships. A complete list of author disclosures appears with the original paper.
JAMA Neurol. Published online October 16, 2017. Abstract
For more Medscape Neurology news, join us on Facebook and Twitter
Tidak ada komentar:
Posting Komentar