No broad differences in gut bacterial composition, however, are associated with risk of relapse, according to the investigators. The findings were presented at the Joint European Committee for Treatment and Research in Multiple Sclerosis–Americas Committee for Treatment and Research in Multiple Sclerosis (ECTRIMS–ACTRIMS) 2020, this year known as MSVirtual2020.
Dr. Mary Horton
Previous research has found an association between Blautia stercoris and disease activity in other immune-mediated diseases such as systemic lupus. Although the current study is the largest in patients with MS that includes data about the microbiome and relapses, its findings require replication, said Mary Horton, a doctoral candidate in epidemiology at the University of California, Berkeley.
Gut microbes digest food, produce vitamins (for example, B12 and K), create a barrier against pathogens, and regulate the immune system, among other tasks. Most current knowledge about the gut microbiome in MS comes from studies of patients with adult-onset MS. In 2016, Tremlett et al. found an increase in Desulfovibrionaceae and a decrease in Lachnospiraceae and Ruminococcaceae in patients with pediatric-onset MS. They also found that a decrease in Fusobacteria was associated with risk of relapse in this population.
Advanced analytical methods
Using a larger sample size and newer analytical methods than in the study by Tremlett and colleagues, Ms. Horton’s group sought to determine whether features of the gut microbiome are associated with relapse. From 2014 to 2018, the investigators recruited 53 patients with pediatric-onset MS from the University of California, San Francisco, and six centers in the U.S. Network of Pediatric MS Centers. At baseline, they collected stool samples, blood samples, information about past relapses, medication records, demographics, and environmental factors. At each relapse, the investigators collected information about the patient’s current and past medication use and about relapses that the patient had had since the previous visit.
Ms. Horton and colleagues analyzed the stool samples using 16S rRNA sequencing of the V4 region. They identified amplicon sequence variants (ASVs), which are used to define species of bacteria, with the Divisive Amplicon Denoising Algorithm-2 (DADA2). Taxonomies were assigned using the naive Bayesian classifier method, and the read count was normalized using multiple rarefaction.
The investigators identified ASV clusters using weighted genetic correlation network analysis (WGCNA). To evaluate whether individual ASVs were associated with relapse, they used a Prentice, Williams, and Peterson (PWP) recurrent event model, an extension of the Cox proportional hazards model.
The role of methanogenesis
Ms. Horton and colleagues included 53 patients (72% girls) in their study. The population’s mean age was 14.3 years at disease onset and 15.5 years at stool sample collection. About 70% of patients were White, and about 36% were Hispanic. Mean disease duration was 1.3 years, and median Expanded Disability Status Scale score was 1.0.
Approximately 45% of participants had one relapse, and 30% had more than one relapse during the subsequent mean follow-up of 2.5 years. About 91% of patients used a disease-modifying therapy during follow-up.
Gut bacterial abundance was broadly similar between patients who relapsed during the study period and those who did not. Of 270 ASVs included in the analyses, 20 were nominally associated with risk of relapse. Blautia stercoris had the most significant association with relapse risk (hazard ratio, 2.50). Blautia massiliensis also was among the 20 ASVs associated with risk of relapse.
WGCNA identified six ASV clusters. Higher values of one cluster’s eigengene were significantly associated with higher relapse risk (HR, 1.23). The following four ASVs nominally associated with higher relapse risk were in this cluster: Blautia massiliensis, Dorea longicatena, Coprococcus comes, and an unknown species in genus Subdoligranulum.
When Ms. Horton and colleagues examined the pathways from these bacterial species, they found 10 that were significantly associated with the risk of relapse. Four of these 10 pathways are involved in methane production, which suggests the involvement of methanogenesis pathways in relapse.
Although the investigators used advanced techniques for genetic and statistical analysis, the study’s sample size is small, Ms. Horton acknowledged. In addition, the conclusions that can be drawn from observational data are limited.
These suggest several avenues for future research. “There is a big question about how the different treatments that people are on when they are experiencing relapses might impact the microbiome,” said Ms. Horton. “Is the microbiome impacting your treatment response, or is it the reverse?” Investigators also could examine why the methane production pathway is overrepresented among people with MS who have relapses. “Which specific archaea might be leading to that increase in methane is a ripe future study question. Just what that means for health is really unknown.”
The National MS Society and the National Institute of Neurological Disorders and Stroke provided funding for the study. Ms. Horton had no disclosures.
SOURCE: Horton M et al. MSVirtual2020, Abstract LB01.05.