New Genetic Tool Could One Day Diagnose Rheumatic Disease Faster Than A Clinician

A novel diagnostic tool that combines a patient's own genetic data with genetic risk scores zeroed in on the most probable diagnosis among five different rheumatic diseases in a proof-of-principle study.

"Though much research has focused on using genetics to predict disease risk at a population level, we use genetics in the setting where a person has symptoms," Dr. Rachel Knevel of Leiden University told Reuters Health by email. "This increases the value of genetics and it fits the clinical situation much better."

"In contrast to traditional tests, instead of giving a positive/negative outcome, we give a probability," she said. "This fits the probabilistic thinking of a clinician and allows incorporation of our genetic probability into the clinical differential diagnoses."

Dr. Knevel and colleagues developed G-PROB (Genetic Probability tool), which uses genetic information combined with a genetic risk score from multiple diseases to calculate a given patient's conditional probabilities for each of those diseases, assuming one is present.

As reported in Science Translational Medicine, after validating the tool on simulated data, they used it to discriminate among rheumatoid arthritis, systemic lupus erythematosus, spondyloarthropathy, psoriatic arthritis, and gout in 1,699 patients with inflammatory arthritis and genetic and clinical data.

G-PROB was able to rule out at least one disease in all patients; in 45%, a likely diagnosis was identified with a 64% positive predictive value.

G-PROB also detected an incorrect clinical diagnosis in 35% of cases. Adding genetic risk score calculations ("G-probabilities") to the clinical data improved the tool's diagnostic accuracy to 51%, compared to 39% based on interpretations of the clinical data alone.

The authors note, "Pre-existing genetic data could be considered part of a patient's medical history given its potential to improve precision medicine in the modern outpatient clinic."

Dr. Knevel said, "Ideally, one would like to test G-PROB prospectively--i.e., give the score at the patient's first visit. I think this would be feasible if let's say 50% of the patients have genotyped data. My guess would be that this can happen within 10 years."

"Our method of disease differentiation in patients presenting with overlapping symptoms is applicable to many clinical situations beyond rheumatology," she added. These may include "different types of pulmonary hypertension or colitis or heart failure; diabetes (e.g., to differentiate adult-onset T1DM versus T2DM); and perhaps neurology and psychiatry."

Dr. Anca Askanase, Director, Columbia University lupus Center and Associate Professor of Medicine at Columbia University College of Physicians and Surgeons in New York City, commented in an email to Reuters Health, "This elegant study ushers in the future of rheumatology, where the diagnosis of rheumatic diseases will be a precise algorithm that factors in the carefully curated genetic risk profile, quantified signs and symptoms, and laboratory and radiographic information."

"For patients with symptoms of inflammatory arthritis, the G-PROB further guides the clinician on the correct path to order the appropriate battery of tests in order to quickly and efficiently make the diagnosis," she said. "These data are not ready for use in rheumatology clinics tomorrow and require further testing and refining; this is the first attempt to use the available DNA genotype data to help in the diagnosis of rheumatic disease."

--Marilynn Larkin

Source