A short course of prednisolone may help rheumatologists differentiate between patients with rheumatoid arthritis and osteoarthritis, a proof of concept study shows.

However, the investigators caution that a positive response to the 3-day steroid course does not confirm a diagnosis of rheumatoid arthritis (RA).

For many years, rheumatologists have been using short courses of prednisolone in unclear clinical situations to differentiate between inflammatory and non-inflammatory arthritis, according to the investigators led by Uta Kiltz, MD , from Rheumazentrum Ruhrgebiet, Herne, Germany.

“The basic idea is that RA is due to inflammation, whereas OA [osteoarthritis] is either not based on the same pathophysiology, or at least not to the same degree,” they wrote.

The pilot part of the TryCort study involved 15 patients with confirmed osteoarthritis and 15 with rheumatoid arthritis who were given 1 g of paracetamol (acetaminophen) a day for 5 days, and on days 3-5, they were given a 20-mg dose of prednisolone ( Arthritis Res Ther. 2017;19:73. doi: 10.1186/s13075-017-1279-z ).

Results showed that the patients with RA had greater improvements in their pain scores (0-10 on a numerical rating scale), compared with OA patients. The mean percentage improvement in pain scores at day 5 was 52.3% in the RA group and 22.0% in the OA group.

The research team considered that a 40% improvement in pain scores was the best choice between sensitivity and specificity regarding a diagnosis of RA.

At this 40% improvement cut-off, the “pred-test” was positive in 11 patients with RA and in four patients with OA (P = .012), with a sensitivity and specificity for a diagnosis of RA of 73.3% for both measures.

In order to validate the test, the researchers enrolled 95 patients with pain in their fingers and hands but without a clear diagnosis. These patients completed the 5-day intervention, and then at week 12 a rheumatologist diagnosed 47 as having RA and 48 were thought to not have RA.

The patients with diagnosed RA had a higher reduction in pain scores during the treatment with prednisolone, compared with patients without RA.

The median percentage of improvement at day 5 was higher in patients with RA than in those without RA (50% [interquartile range, 30%-60%] vs. 20% [IQR, 10%-30%]; P = .001). Overall, 40 of the 95 patients had an improvement of more than 40% in pain levels on day 5, fulfilling the criteria of a positive pred-test.

However, the authors noted that 31 patients with RA had a positive pred-test (77.5%), compared with nine (22.5%) patients without RA (P greater than .001).

The sensitivity of the pred-test for a diagnosis of RA was 0.6 (95% confidence interval, 0.5-0.8) and the specificity was 0.8 (95% CI, 0.7-0.9). The positive and negative predictive values were 0.77 and 0.70, respectively.

The authors concluded that the pred-test “performed well” but not “perfectly well.”

“We are aware that the pred-test without confirmation of other surrogate markers is not helpful in clinical decision-making processes,” they said. “We, therefore, recommend use of the test in light of other confirming factors, such as history, physical examination, imaging, and laboratory results.”

The test could be used to triage patients from primary care to rheumatologist care, they suggested.

The study was financially supported by Rheumazentrum Ruhrgebiet. The authors declared no conflicts of interest.