Contributions
Abstract: P732
Type: Poster
Abstract Category: Therapy - disease modifying - Tools for detecting therapeutic response
Background: We have previously reported the existence of two subsets of MS subjects that were defined using transcriptomic profiles of peripheral blood mononuclear cells (PBMC) from untreated subjects, subjects treated with Glatiramer Acetate (GA) and interferon β. The smaller subset was reported to be more likely to have inflammatory disease activity
Objective: To confirm (1) our report of structure in the MS patient population and (2) the association of the smaller subset of subjects with greater inflammatory disease activity.
Methods: Relapsing remitting MS subjects (RRMS, n=163) from the Comprehensive Longitudinal Investigation of MS at the Brigham and Women"s Hospital (CLIMB) that are independent from the set of subjects used in the prior study (Ottoboni et al. Sci Trans Med 2012) and had been treated with GA for > 6 months were selected for this study. The outcome measure was the length of event-free time following profiling, (event= a clinical relapse, new T2 hyperintense and/or gadolinium-enhancing lesion, or a >6 months sustained increase of 1 EDSS point). . For each of the 163 PBMC samples, a cDNA library was constructed and sequenced to an average depth of 40 million paired end reads . As in the prior study, non-negative matrix factorization (NMF) was used to cluster subjects, and a Cox proportional hazards regression model was used to compare the time to an inflammatory event between the clusters of MS subjects.
Results: In the new dataset of 163 GA-treated subjects, NMF analysis returns an optimal solution of k=2, as in the prior study: the data are best explained by the presence of two subsets of RRMS subjects, we also find that the smaller subgroup of subjects is more likely to have an inflammatory event after the date of sampling (Cox proportional hazard ratio = 0.33, p = 0.017), consistent with the original study.
Conclusion: Among RRMS subjects treated with GA, we confirm the existence of two subsets that can be distinguished using PBMC RNA profiles. Given the difference in inflammatory disease activity, this population structure could be leveraged to enhance the design of translational studies in MS and should be further investigated to assess whether it can inform clinical decision-making, when considered along with current parameters.
Disclosure:
De Jager: This work was funded by a sponsored research agreement from Biogen.
Drs. Carulli, Harris and Goyal are employees of Biogen.
Patrick: nothing to disclose
Bargiela:nothing to disclose
Ottoboni: nothing to disclose
Weiner: no pertinent disclosuresJG, JPC are full time employees of Biogen and hold stock/stock options in Biogen. TH currently at SV Lifesciences and was a Biogen employee at the time this study was executed and held stock/stock options in Biogen
Abstract: P732
Type: Poster
Abstract Category: Therapy - disease modifying - Tools for detecting therapeutic response
Background: We have previously reported the existence of two subsets of MS subjects that were defined using transcriptomic profiles of peripheral blood mononuclear cells (PBMC) from untreated subjects, subjects treated with Glatiramer Acetate (GA) and interferon β. The smaller subset was reported to be more likely to have inflammatory disease activity
Objective: To confirm (1) our report of structure in the MS patient population and (2) the association of the smaller subset of subjects with greater inflammatory disease activity.
Methods: Relapsing remitting MS subjects (RRMS, n=163) from the Comprehensive Longitudinal Investigation of MS at the Brigham and Women"s Hospital (CLIMB) that are independent from the set of subjects used in the prior study (Ottoboni et al. Sci Trans Med 2012) and had been treated with GA for > 6 months were selected for this study. The outcome measure was the length of event-free time following profiling, (event= a clinical relapse, new T2 hyperintense and/or gadolinium-enhancing lesion, or a >6 months sustained increase of 1 EDSS point). . For each of the 163 PBMC samples, a cDNA library was constructed and sequenced to an average depth of 40 million paired end reads . As in the prior study, non-negative matrix factorization (NMF) was used to cluster subjects, and a Cox proportional hazards regression model was used to compare the time to an inflammatory event between the clusters of MS subjects.
Results: In the new dataset of 163 GA-treated subjects, NMF analysis returns an optimal solution of k=2, as in the prior study: the data are best explained by the presence of two subsets of RRMS subjects, we also find that the smaller subgroup of subjects is more likely to have an inflammatory event after the date of sampling (Cox proportional hazard ratio = 0.33, p = 0.017), consistent with the original study.
Conclusion: Among RRMS subjects treated with GA, we confirm the existence of two subsets that can be distinguished using PBMC RNA profiles. Given the difference in inflammatory disease activity, this population structure could be leveraged to enhance the design of translational studies in MS and should be further investigated to assess whether it can inform clinical decision-making, when considered along with current parameters.
Disclosure:
De Jager: This work was funded by a sponsored research agreement from Biogen.
Drs. Carulli, Harris and Goyal are employees of Biogen.
Patrick: nothing to disclose
Bargiela:nothing to disclose
Ottoboni: nothing to disclose
Weiner: no pertinent disclosuresJG, JPC are full time employees of Biogen and hold stock/stock options in Biogen. TH currently at SV Lifesciences and was a Biogen employee at the time this study was executed and held stock/stock options in Biogen