The gut microbiome of RA patients can be analysed to predict improvements in disease activity, according to a small study conducted by Mayo Clinic researchers.
The study, published in Genome Medicine, used artificial intelligence to show proof-of-concept evidence that microbiota from stool samples accurately predicted minimum clinically important improvement (MCII) in RA patients.
Forecasting if a patient will have a good prognosis would, according to the researchers, address “a steep challenge” in the clinical practice of RA and enable more precise treatment for patients.
In this retrospective, observational cohort study, researchers compared the baseline gut microbiome composition between two groups of patients who had been diagnosed with RA between 1988 and 2014: those who had achieved MCII in disease activity and those who did not.
Two stool samples were collected from each of the 32 participants at two separate time points 6-12 months apart. At baseline all patients were taking bDMARDs, csDMARDs or prednisone. The patients’ Clinical Disease Activity Index (CDAI) was measured at both time points to determine whether they’d achieved MCII.
The samples underwent a comprehensive precision genomic analysis called shotgun metagenomic sequencing. Irrelevant sequences and potential human contamination were filtered out and taxonomic profiling was performed. Then an artificial intelligence (AI) deep-learning neural network model was trained on baseline microbiome, clinical and demographic data to determine how a prediction of disease outcomes could be made based on the microbiome composition.
The team investigated the connection between the gut microbiome and the smallest meaningful changes in clinical disease activity and found several traits of the microbiome linked to future prognosis.
And in determining disease trajectory, microbiome data were revealed to be more important predictive factors, as determined by the neural network, than clinical and demographic characteristics.
The deep-learning neural network model was found to predict which patients would show clinical improvement with 90% accuracy. Researchers said they hoped that their work would provide a cornerstone for future leveraging of AI capabilities to create data-based tools to detect, diagnose and treat RA.
Apart from the small sample size and lack of longer-term follow-up, researchers listed several limitations including the possible geographical/cultural bias created by patients being mostly from the mid-west of USA and the particular dietary preferences of that region.
Gut microbiome has been linked in recent years to RA disease detection, classification and treatment efficacy. What this novel study has potentially provided is a non-invasive screening tool for predicting clinical improvement.
“This is the first study to date that uses gut microbiome data to predict clinical improvement in rheumatoid arthritis disease activity independent of the initial measurement of their condition or prior treatment,” said co-senior author Dr Jaeyun Sung, a computational biologist within Mayo Clinic’s Center for Individualized Medicine, in a press release.
“With further development, such prognostic biomarkers could identify patients who will achieve early clinical improvement with a given therapy, thereby sparing them the expense and risk of other therapies that are less likely to be effective,” added co-senior author and Mayo Clinic rheumatologist, Dr John Davis.
“Conversely, such tools can detect patients whose disease symptoms are less likely to improve, and perhaps allow clinicians to target and monitor them more closely. Much is left to be done, but we’re on the right path toward advancing our understanding of this disease in order to individualise medicine for patients with rheumatoid arthritis,” said Dr Davis.
The researchers also suggested that the gut microbiome could be implicated in future treatment scenarios.
“Ultimately, our study reveals that modifying the gut microbiome to enhance clinical outcome may hold promise as a future treatment for rheumatoid arthritis,” said Dr Sung.
“This could revolutionise how we deliver care to our patients.”