Genetic Signatures Predict Response to Treatment for RA

Kate Johnson

November 17, 2016

WASHINGTON, DC — Three genetic signatures could one day help rheumatologists decide which rheumatoid arthritis therapies to choose for which patients, according to a subanalysis of data from the Optimal Management of Rheumatoid Arthritis Patients Requiring Biologic Therapy (ORBIT) study.

"We're opening up the possibility of stratified medicine," said ORBIT investigator Duncan Porter, MD, consultant rheumatologist at the Queen Elizabeth University Hospital in Glasgow, United Kingdom.

Generally, about 60% of rheumatoid arthritis patients will respond to a specific drug, "yet far more than 60% of our patients have got good disease control on a biologic," he said during a press conference here at the American College of Rheumatology 2016 Annual Meeting.

"If they don't respond to the first drug, they go on to the second or third until they get a good response," he explained. "Right now, that's mostly a flip of a coin," but this subanalysis shows that baseline genetic analysis "could very accurately predict 3-month response to TNF or rituximab or both."

These results "present many reasons for optimism," Dr Porter told Medscape Medical News.

In the original study (Lancet. 2016;388:239-247), the ORBIT investigators compared TNF inhibitors with rituximab in biologic-naïve rheumatoid arthritis patients, and found that the two therapies are equivalent "as long as patients moved from one to the other if they didn't have a good response," he explained.

"Around 20% needed to change from the first drug to get a response," he reported. "Wouldn't it have been ideal if we could have identified the patients who were not going to respond to the first drug and given them the right drug from the beginning?"

For their subanalysis of the ORBIT data, Dr Porter and his colleagues looked at baseline genetic markers in 241 patients. The team used 70% of the cohort to identify response to therapy and the other 30% to validate the findings.

The three genetic signatures they identified were highly predictive, with a sensitivity of 93%, specificity of 91%, positive predictive value of 96%, and negative predictive value of 86%.

When the signatures were applied to the original ORBIT sample, "we see that about 10% of patients aren't expected to respond to either TNF inhibition or rituximab," said Dr Porter. "Maybe these are patients who would do best with IL-6 blockade or T-cell costimulation blockade."

The team also found that about 50% of patients respond to either drug and about 40% respond to one drug or the other.

Many proposed biomarkers of treatment-response prediction have failed replication in subsequent studies.

"This is a very interesting study," said Antonio Gonzalez, MD, PhD, from the rheumatology unit at Complejo Hospitalario Universitario de Santiago in Santiago de Compostela in Spain. "The findings seem very promising."

"It could lead to useful biomarkers for the prediction of response to treatment in patients with rheumatoid arthritis, but this will require additional work," said Dr Gonzalez.

Still, he cautioned, the "precedents are not very encouraging. Many proposed biomarkers of treatment-response prediction have failed replication in subsequent studies."

Dr Porter reports receiving research grants from Roche Pharmaceuticals, sponsorship from AbbVie, and serving on the speakers' bureau for Bristol-Myers Squibb and Janssen. Dr Gonzalez has disclosed no relevant financial relationships.

American College of Rheumatology (ACR) 2016 Annual Meeting: Abstract 3089. Presented November 15, 2016.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.