Routine hand and foot radiographs in more than 700 patients being evaluated for rheumatoid arthritis (RA) made almost no difference in diagnosing the condition nor in predicting its course, Dutch researchers said.
Fewer than 5% of x-rays showed RA-type erosions, and diagnostic classifications were changed in just two of the 724 patients because of these findings, according to Evy Ulijn, MD, of Sint Maartenskliniek in Nijmegen, Netherlands, and colleagues.
“The recommendation of conducting routine [hand and foot x-rays] in patients with newly presenting arthritis suspected of RA might be reconsidered, due to low prevalence of early erosive disease and lack of diagnostic and prognostic value,” Ulijn and colleagues concluded.
Guidance published in 2010 by the American College of Rheumatology and the European League Against Rheumatism (ACR/EULAR) called for x-rays to be performed in most patients under evaluation for possible RA, even though its authors acknowledged that erosions would be uncommon in early RA.
Diagnosis under those criteria was to be based primarily on exam findings of joint swelling and pain, with blood testing for rheumatoid factor (RF), anti-citrullinated protein antibodies (ACPA), and acute-phase reactants (C-reactive protein and erythrocyte sedimentation rate) helping to distinguish between RA and other forms of arthritis. Presence of erosions would, of course, confirm an RA diagnosis; in fact, the 2010 criteria indicated that radiographic erosions could justify an RA diagnosis without considering other factors.
But Ulijn and colleagues took the opposite view: with erosions being relatively rare in early RA, how much value could there be in routine x-rays?
They looked at records of patients seen at their Nijmegen clinic for suspected RA from 2016 to 2019, identifying 724 who had joint inflammation at presentation with testing performed for RF and ACPA, and in whom hand and foot x-rays were taken. Patients were excluded if the order for radiographs mentioned conditions other than RA.
Of the 724 cases, 299 eventually were given RA diagnoses. About 80% of these came from application of the 2010 ACR/EULAR criteria, the rest primarily from clinicians’ judgment. In two cases, patients were negative for RF and ACPA and also for acute-phase reactants, but had erosions on x-ray; these were the only patients for whom the radiographs were instrumental in reaching an RA diagnosis.
Similarly, erosions indicating an unfavorable prognosis were detected in only three patients.
In other words, radiographs were normal in about 99% of patients who ultimately received an RA diagnosis, and thus could have been dispensed with. Erosions were rare, too, among the larger group of patients who were negative for blood-based markers, most of whom ended up with non-RA diagnoses.
“Therefore, there seems no need to recommend routine [x-rays] in patients with newly presenting arthritis suspected of RA,” Ulijn and colleagues wrote. They emphasized, though, that x-rays may be ordered to evaluate specific concerns involving a particular hand or foot and these may be of substantial value.
Notably, the study methods considered x-rays showing erosion at a single joint to be positive, whereas the 2010 ACR/EULAR guidance required that erosions be presented in three joints to justify a finding of erosive disease. Using those criteria, Ulijn and colleagues commented, “would yield an even lower prevalence of RA-associated erosions” than their data indicated.
The researchers also slapped down the notion that radiographs might alert clinicians to otherwise unsuspected diagnoses. “[O]nly a handful of exceedingly uncommon alternative diagnoses in early oligo-/polyarthritis are associated with these specific radiographical abnormalities which does not make it rational to routinely perform” hand and foot x-rays, they wrote.
The study had no external funding.
Authors declared they had no relevant financial interests.
Arthritis Care & Research
Source Reference: Ulijn E, et al “Limited diagnostic and prognostic value of routine radiographs in newly presenting arthritis suspected of rheumatoid arthritis: a retrospective study” Arthritis Care Res 2023; DOI: 10.1002/acr.25271.