User login
STOCKHOLM –
Further, “combined omics improves diagnostic accuracy” over oligoclonal band (OCB) status alone in differentiating patients with multiple sclerosis (MS) from a group of normal control patients, said Fay Probert, PhD, speaking at a poster session at the annual congress of the European Committee for Treatment and Research in Multiple Sclerosis.
Not everyone with clinically isolated syndrome (CIS) converts to clinically definite MS, and there is large variability in the time to progression, explained Dr. Probert, a postdoctoral fellow in the department of pharmacology at Oxford (England) University, and colleagues. Though the revised McDonald criteria now allow earlier diagnosis of MS, individuals who will be early converters cannot be identified by these criteria, they noted, citing earlier work showing that just over half (52%) of CIS patients who are OCB positive will have clinically definite MS at the 3-year mark.
To see whether combining analysis of multiple proteins and metabolites improved diagnostic accuracy, Dr. Probert and colleagues examined cerebrospinal fluid (CSF) samples from 41 patients with clinically definite MS, 71 patients with CIS, and 64 control participants without MS. In their analysis, the investigators used nuclear MR metabolomics and a commercially available proteomics assay that identifies and quantifies more than 5,000 proteins.
The multivariate analysis strategy achieved 10-fold external cross-validation of the samples, repeating training and testing of the analysis model while shuffling data. This, explained Dr. Probert and colleagues, “ensures that any discrimination observed cannot have occurred by chance.” Further analysis “identifies the optimal combination of proteomics and metabolomics features which results in the highest diagnostic accuracy.”
Both the nuclear MR metabolomics and the proteomic analyses were able to discriminate between those with clinically definite MS and the control participants, with accuracy of 71% and 75%, respectively.
The levels of seven metabolites present in CSF were predictive of clinically definite MS, compared with non-MS status, independent of OCB status. In fact, noted Dr. Probert and colleagues, “the CSF myoinositol concentration alone diagnosed [clinically definite] MS in this cohort with a specificity of 74% but did not outperform OCB status overall.”
Using the combined omics approach, though, “significantly improved the discrimination” between the non-MS control CSF samples and those of patients with clinically definite MS, wrote Dr. Probert and colleagues. Using a combination of up to five CSF proteins and metabolites yielded accuracy of 85 plus or minus 2%, sensitivity of 85 plus or minus 3%, and specificity of 85 plus or minus 3%. For comparison, using just OCB status provides accuracy of 74%, sensitivity of 88% and specificity of 63%.
Then, Dr. Probert and colleagues turned to the CSF samples from patients with CIS to look for predictors of “fast” (4 years or less) or “slow” (greater than 4 years) conversion to clinically definite MS. “While important for diagnosis, OCB status was not predictive of early conversion,” the investigators noted. However, baseline CSF proteomics analysis alone did differentiate the fast from the slow converters among the CIS subgroup, with an accuracy of 77%.
For patients with CIS who were OCB positive, their baseline metabolite and proteomic profiles were “indistinguishable” from those with clinically definite MS, wrote Dr. Probert and colleagues. The omics analysis was also able to distinguish between OCB-positive CIS patients and the non-MS control patients.
“These results indicate that combined metabolomics and proteomics analysis could not only be used as an adjunct in diagnosis of [clinically definite] MS but could be used as a prognostic test to identify CIS patients at high risk of a second clinical attack within 4 years of onset,” wrote Dr. Probert and coauthors. They noted that the method reported in the poster is the first to offer this prognostic accuracy, but that more work is needed before routine clinical use.
Dr. Probert reported that she had no financial conflicts of interest. One coauthor reported being a consultant to Novartis. Two coauthors reported financial relationships with multiple pharmaceutical companies, including Merck, which partially funded the study. Numares Health, the U.K. Medical Research Council, and the Multiple Sclerosis Society also provided funding support.
SOURCE: Probert F et al. ECTRIMS 2019, Abstract P586.
STOCKHOLM –
Further, “combined omics improves diagnostic accuracy” over oligoclonal band (OCB) status alone in differentiating patients with multiple sclerosis (MS) from a group of normal control patients, said Fay Probert, PhD, speaking at a poster session at the annual congress of the European Committee for Treatment and Research in Multiple Sclerosis.
Not everyone with clinically isolated syndrome (CIS) converts to clinically definite MS, and there is large variability in the time to progression, explained Dr. Probert, a postdoctoral fellow in the department of pharmacology at Oxford (England) University, and colleagues. Though the revised McDonald criteria now allow earlier diagnosis of MS, individuals who will be early converters cannot be identified by these criteria, they noted, citing earlier work showing that just over half (52%) of CIS patients who are OCB positive will have clinically definite MS at the 3-year mark.
To see whether combining analysis of multiple proteins and metabolites improved diagnostic accuracy, Dr. Probert and colleagues examined cerebrospinal fluid (CSF) samples from 41 patients with clinically definite MS, 71 patients with CIS, and 64 control participants without MS. In their analysis, the investigators used nuclear MR metabolomics and a commercially available proteomics assay that identifies and quantifies more than 5,000 proteins.
The multivariate analysis strategy achieved 10-fold external cross-validation of the samples, repeating training and testing of the analysis model while shuffling data. This, explained Dr. Probert and colleagues, “ensures that any discrimination observed cannot have occurred by chance.” Further analysis “identifies the optimal combination of proteomics and metabolomics features which results in the highest diagnostic accuracy.”
Both the nuclear MR metabolomics and the proteomic analyses were able to discriminate between those with clinically definite MS and the control participants, with accuracy of 71% and 75%, respectively.
The levels of seven metabolites present in CSF were predictive of clinically definite MS, compared with non-MS status, independent of OCB status. In fact, noted Dr. Probert and colleagues, “the CSF myoinositol concentration alone diagnosed [clinically definite] MS in this cohort with a specificity of 74% but did not outperform OCB status overall.”
Using the combined omics approach, though, “significantly improved the discrimination” between the non-MS control CSF samples and those of patients with clinically definite MS, wrote Dr. Probert and colleagues. Using a combination of up to five CSF proteins and metabolites yielded accuracy of 85 plus or minus 2%, sensitivity of 85 plus or minus 3%, and specificity of 85 plus or minus 3%. For comparison, using just OCB status provides accuracy of 74%, sensitivity of 88% and specificity of 63%.
Then, Dr. Probert and colleagues turned to the CSF samples from patients with CIS to look for predictors of “fast” (4 years or less) or “slow” (greater than 4 years) conversion to clinically definite MS. “While important for diagnosis, OCB status was not predictive of early conversion,” the investigators noted. However, baseline CSF proteomics analysis alone did differentiate the fast from the slow converters among the CIS subgroup, with an accuracy of 77%.
For patients with CIS who were OCB positive, their baseline metabolite and proteomic profiles were “indistinguishable” from those with clinically definite MS, wrote Dr. Probert and colleagues. The omics analysis was also able to distinguish between OCB-positive CIS patients and the non-MS control patients.
“These results indicate that combined metabolomics and proteomics analysis could not only be used as an adjunct in diagnosis of [clinically definite] MS but could be used as a prognostic test to identify CIS patients at high risk of a second clinical attack within 4 years of onset,” wrote Dr. Probert and coauthors. They noted that the method reported in the poster is the first to offer this prognostic accuracy, but that more work is needed before routine clinical use.
Dr. Probert reported that she had no financial conflicts of interest. One coauthor reported being a consultant to Novartis. Two coauthors reported financial relationships with multiple pharmaceutical companies, including Merck, which partially funded the study. Numares Health, the U.K. Medical Research Council, and the Multiple Sclerosis Society also provided funding support.
SOURCE: Probert F et al. ECTRIMS 2019, Abstract P586.
STOCKHOLM –
Further, “combined omics improves diagnostic accuracy” over oligoclonal band (OCB) status alone in differentiating patients with multiple sclerosis (MS) from a group of normal control patients, said Fay Probert, PhD, speaking at a poster session at the annual congress of the European Committee for Treatment and Research in Multiple Sclerosis.
Not everyone with clinically isolated syndrome (CIS) converts to clinically definite MS, and there is large variability in the time to progression, explained Dr. Probert, a postdoctoral fellow in the department of pharmacology at Oxford (England) University, and colleagues. Though the revised McDonald criteria now allow earlier diagnosis of MS, individuals who will be early converters cannot be identified by these criteria, they noted, citing earlier work showing that just over half (52%) of CIS patients who are OCB positive will have clinically definite MS at the 3-year mark.
To see whether combining analysis of multiple proteins and metabolites improved diagnostic accuracy, Dr. Probert and colleagues examined cerebrospinal fluid (CSF) samples from 41 patients with clinically definite MS, 71 patients with CIS, and 64 control participants without MS. In their analysis, the investigators used nuclear MR metabolomics and a commercially available proteomics assay that identifies and quantifies more than 5,000 proteins.
The multivariate analysis strategy achieved 10-fold external cross-validation of the samples, repeating training and testing of the analysis model while shuffling data. This, explained Dr. Probert and colleagues, “ensures that any discrimination observed cannot have occurred by chance.” Further analysis “identifies the optimal combination of proteomics and metabolomics features which results in the highest diagnostic accuracy.”
Both the nuclear MR metabolomics and the proteomic analyses were able to discriminate between those with clinically definite MS and the control participants, with accuracy of 71% and 75%, respectively.
The levels of seven metabolites present in CSF were predictive of clinically definite MS, compared with non-MS status, independent of OCB status. In fact, noted Dr. Probert and colleagues, “the CSF myoinositol concentration alone diagnosed [clinically definite] MS in this cohort with a specificity of 74% but did not outperform OCB status overall.”
Using the combined omics approach, though, “significantly improved the discrimination” between the non-MS control CSF samples and those of patients with clinically definite MS, wrote Dr. Probert and colleagues. Using a combination of up to five CSF proteins and metabolites yielded accuracy of 85 plus or minus 2%, sensitivity of 85 plus or minus 3%, and specificity of 85 plus or minus 3%. For comparison, using just OCB status provides accuracy of 74%, sensitivity of 88% and specificity of 63%.
Then, Dr. Probert and colleagues turned to the CSF samples from patients with CIS to look for predictors of “fast” (4 years or less) or “slow” (greater than 4 years) conversion to clinically definite MS. “While important for diagnosis, OCB status was not predictive of early conversion,” the investigators noted. However, baseline CSF proteomics analysis alone did differentiate the fast from the slow converters among the CIS subgroup, with an accuracy of 77%.
For patients with CIS who were OCB positive, their baseline metabolite and proteomic profiles were “indistinguishable” from those with clinically definite MS, wrote Dr. Probert and colleagues. The omics analysis was also able to distinguish between OCB-positive CIS patients and the non-MS control patients.
“These results indicate that combined metabolomics and proteomics analysis could not only be used as an adjunct in diagnosis of [clinically definite] MS but could be used as a prognostic test to identify CIS patients at high risk of a second clinical attack within 4 years of onset,” wrote Dr. Probert and coauthors. They noted that the method reported in the poster is the first to offer this prognostic accuracy, but that more work is needed before routine clinical use.
Dr. Probert reported that she had no financial conflicts of interest. One coauthor reported being a consultant to Novartis. Two coauthors reported financial relationships with multiple pharmaceutical companies, including Merck, which partially funded the study. Numares Health, the U.K. Medical Research Council, and the Multiple Sclerosis Society also provided funding support.
SOURCE: Probert F et al. ECTRIMS 2019, Abstract P586.
REPORTING FROM ECTRIMS 2019