SAN DIEGO – Researchers say they’ve developed a novel blood-based assay that can differentiate patients with seropositive or seronegative rheumatoid arthritis from healthy people and those with other inflammatory diseases or osteoarthritis.
While cautioning that the results need to be confirmed, University of Oxford (England) rheumatologist Peter Taylor, PhD, MA, told an audience at the annual meeting of the American College of Rheumatology that the test has an overall mean sensitivity of 90.8% (standard deviation, 0.94%; 95% confidence interval, 83.2%-95.4%) and mean specificity of 96.1% (SD, 0.64%; 95% CI, 92.7%-97.9%). The mean area under the curve (AUC) is 0.991 (SD, 0.001; 95% CI, 97.2%-99.6%).
Dr. Peter Taylor
“That is to say that it can correctly identify 96% of people without rheumatoid arthritis, and it can correctly identify over 90% of those who do have rheumatoid arthritis,” Dr. Taylor said. In the big picture, he said, “we’ve developed a blood-based assay that detects both organ-specific and systemic biological processes in patients with rheumatoid arthritis.”
Rheumatologist Kevin W. Byram, MD, of Vanderbilt University, Nashville, Tenn., who did not take part in the study but is familiar with its findings, noted in an interview that “there is a constant search for biomarkers to help aide in more accurate and faster diagnosis of all rheumatic conditions, including RA.”
He added that “a common clinical scenario for the rheumatologist is a patient presenting with painful and/or swollen joints and other features that might suggest a few different diagnoses. A diagnostic assay like this might help distinguish patients with RA from other inflammatory conditions or non-inflammatory conditions that might mimic RA.”
In his presentation, Dr. Taylor noted that “we haven’t yet integrated precision medicine into routine clinical practice in rheumatology.” While blood-based assays are prevalent in other types of clinical diagnostics, rheumatology relies on synovial biopsies that are “rarely used in routine clinical practice,” he said.
The new test is a “non-invasive DNA capture assay that can identify specific gene expression from synovium-specific signatures in blood plasma of patients with rheumatoid arthritis,” Dr. Taylor said. Specifically, it focuses on the “unique patterns and sizes of cell-free DNA,” he said. “Analysis of [long] fragments has the potential to give us a great deal of information about disease progression, potentially about customizing treatments and even evaluating the effectiveness of therapies.”
For the new study, researchers examined 229 samples from 191 patients, of whom 63.3% were White and 67.9% were female, with a median age of 56. A total of 89 patients with RA provided 89 samples and 102 without RA provided 140 samples, including 29 healthy controls (66 samples) and others with conditions such as psoriatic arthritis, ulcerative colitis, and osteoarthritis.
The machine learning model “identified 3,425 epigenetic features with statistically significant discrimination between the patients with and without rheumatoid arthritis,” Dr. Taylor said. These features were mapped to 929 genes which had some overlap with known blood pathway genes.
“Over and above that, there’s a whole set of these epigenetic features which represent novel pathways and potentially rich hunting ground for therapeutic targets and other translational investigation,” he said.
For seronegative cases, mean AUC was 0.971 (SD, 0.001; 95% CI, 93.8%-99.2%), sensitivity was 83.7% (SD, 2.03; 95% CI, 63.3%-91.8%) and specificity was 95.4% (SD, 0.69; 95% CI, 90.8%-97.5%).
Specificity for RA versus healthy controls was 100 (SD, 0; 95% CI, 94.4-100.0).
Dr. Byram described the study as small but intriguing. He cautioned that “there is always some likelihood that the actual components of the test are just recognizing some combination of things we are already testing in the clinic,” he said. Details about the patients in the study can offer insight into “whether the assay is actually just recognizing something about patients with RA that is truly different, or rather is it recognizing how a common factor among patients with RA is transcribed by the cell.”
Moving forward, “it is important to get a grasp of how these biomarkers might perform in various settings,” he said.
Dr. Taylor did not discuss the potential cost of the assay in his presentation. “Tests like these have to strike a real balance in being useful and cost-effective and, since they are still made by commercial companies with commercial interests, also make a margin for their owner,” Dr. Byram said. “Turnaround time is also an important factor to think about.”
Aqtual funded the study. Dr. Taylor reports consulting for AbbVie, Aqtual, Biogen, Fresenius Kabi, Galapagos, Gilead, GlaxoSmithKline, Janssen, Lilly, Nordic Pharma, Pfizer, Sanofi, and UCB and receiving grant support from Galapagos. The other study authors all have relationships with Aqtual, and some report various other disclosures. Dr. Byram has no disclosures.