BERLIN – Patients with at least three erosive joints in their hands and feet have rheumatoid arthritis regardless of whatever other signs and symptoms they may or may not have, according to a task force of the American College of Rheumatology and the European League Against Rheumatism (Ann. Rheum. Dis. 2012;71[Suppl3]:25).

This unanimous decision from the 18-member international task force closes the circle for redefining the classification of rheumatoid arthritis patients, a process that began with the 2010 release of a new signs and symptoms–driven rheumatoid arthritis (RA) classification that had no role for radiographic evidence of joint erosions (Arthritis Rheum. 2010;62:2569-81).

Although "radiographs are not required in the ACR/EULAR classification criteria, the presence of typical erosions allows classification of RA even without fulfillment of the [2010] scoring system," task force leader Dr. Désirée van der Heijde said at the meeting. "Radiographs should be taken in an unclassified patient in whom long-standing, inactive disease is suspected, and is likely misclassified," said Dr. van der Heijde, professor rheumatology at Leiden (the Netherlands) University Medical Center. Another common scenario is when a radiograph already exists for a patient who might have early arthritis because it was taken before rheumatologic referral.

The key premise of the new erosive-joint classification was that because radiographs alone could classify a patient as having RA, the erosive joint criteria had to be very specific and produce a minimal number of false positives. Lack of sensitivity was not an issue, as patients could also be classified with RA by the 2010 criteria.

To get a sense of how many erosive joints are needed to produce a reliable identification of RA, the task force studied data collected from two early-arthritis groups, the Early Arthritis Cohort assembled in Leiden with 902 patients, and the Etude et Suivi des Polyarthrites Indifferenciées Récentes (ESPOIR) cohort assembled in Montpellier, France with 811 patients. The task force focused on the subgroup of people in either cohort who did not meet the 2010 classification criteria for RA, and examined the link between various numbers of erosive joints in these people and the incidence of three outcomes the task force considered pathognomonic for RA: on methotrexate treatment after 1 year, on treatment with any disease-modifying antirheumatic drug after 1 year, or disease persistence for 5 years. They found that all three outcomes occurred at similar rates.

In the ESPOIR cohort, with the end point of 5-year disease persistence, people with at least three erosive joints developed RA with a specificity of 91% and a sensitivity of 24%. (An erosive joint was any hand or foot joint with at least one radiographic erosion visible as a cortical break.) In contrast, a threshold of at least two erosive joints carried a specificity of 82% – putting it below the 90% the task force sought – and a sensitivity of 30%. In the Leiden cohort, three or more erosive joints had a sensitivity of 95% and a sensitivity of 15%, while two or more joints was 91% specific and 20% sensitive.

"Some rheumatologists might be surprised [at the need for three erosive joints] because they think that just one erosion is very specific for RA. We showed that one erosion is not specific; you really need more to be very specific for RA," she said in an interview. "After we saw the data it was very clear to us that three would be the right cutoff. It was a unanimous decision."

It also turned out that a criterion of three or more erosive joints kept a lid on the number of positive classifications. In the Leiden cohort, 31 (10%) of the 308 enrollees who did not meet the 2010 clinical criteria met the radiographic threshold; in the ESPOIR cohort, 18 (12%) of 147 had at least three erosive joints.

"There will not be many patients who get classified," based on their erosive joints, she noted. Despite that, "it’s very important to have this definition, because many people were asking for it. It’s important to have a clear definition. And it’s important for patients who do not meet the classification criteria but have several erosions."

"Erosions are the hallmark of RA; if patients have erosions they don’t need anything else," commented Dr. Alan Silman, medical director of Arthritis Research UK in Chesterfield, England. The new findings show "there are very few people with RA who just have erosions. The results support the classification criteria," by showing that the 2010 criteria capture roughly 90% of all RA patients, he said in an interview.

Dr. van der Heijde and Dr. Silman said that they had no disclosures.

Three Erosive Joints Made New RA Definition