A major international collaboration has culminated in the creation of new draft systemic lupus erythematosus classification criteria that more accurately reflects our current understanding of the disease.
Presenting an outline of the criteria to the press Dr Sindhu Johnson, co-chair of the 12 member steering committee, said the new and improved draft criteria was unique because it was data, expert and patient driven.
It could be used by the global rheumatology community and would allow better comparison between studies, she said.
The draft criteria, that involved 189 global lupus experts and data from more than 3500 patients was developed in four phases. The first phase involved collection of possible criteria from review of the published literature, analysis of data, an expert delphi process, and cross-sectional survey of people with SLE.
From this information the steering committee identified 145 candidate criteria which were then streamlined to 40 criteria which were weighted differently to reflect that not all criteria are equal.
“For example, in early cohort studies we realised that fever is a manifestation that may differentiate people with lupus from people who don’t have lupus very early on. But we also know that kidney involvement can be a very serious manifestation.
Through multi-criteria decision analysis we were able to retain fever and kidney involvement but they were weighted differently,” explained Dr Johnson who is also Director of the Toronto Scleroderma Program at the University of Toronto in Canada. The committee then asked 36 international lupus centres to contribute 100 cases and 100 controls to a data set. Each case was independently adjudicated by three lupus experts from three different centres.
They then selected 500 cases and 500 control cases in order to test the draft criteria.
What does the classification system look like?
Dr Johnson explained there was an entry criteria of an antinuclear antibody test (ANA) titre of 1: 80 measured using a hep-2 immunofluorescence assay.
Next, there are clinical domains that include skin manifestations, arthritis, neurological symptoms, serositis, kidney and haematological manifestations. Immunological domains included antiphospholipid antibodies, complement proteins and lupus specific antibodies.
The domains have sub-criteria and each criteria is weighted from the lowest to highest. Within any one domain the criteria with the highest score must be counted. If the subject has a score of 10 or more they are classified as having SLE.
Dr Johnson said that when they tested the performance of the draft criteria in the test cohort of 1000 patients they found it had a sensitivity of 98% and specificity of 97% .
When they compared the same data set with previous lupus criteria they found that the draft classification system had improved sensitivity over the old American College of Rheumatology criteria and maintained excellent specificity.
“We have defined a system of criteria which produces a measure of the relative probability that a particular case with a combination of clinical symptoms or features has SLE… the differential weighting more accurately reflects our current understanding of the disease,” Dr Johnson said.
The proposed classification criteria are in draft form and the ACR meeting was a chance for rheumatologists to give their feedback.
“It is truly an exciting time for SLE research. We are finally developing therapies that may help our patients, and it is important to study them properly,” Dr Johnson said.