Background ACPA and Rheumatoid Factor (RF) are included in the new 2010 ACR/EULAR criteria for RA diagnosis, although their presence is not compulsory. Indeed, 15 to 40% of RA patients do not produce ACPA during the course of the disease.
Objectives To study the evolution and differential diagnosis of ACPA-negative RA patients.
Methods In a bicentric retrospective study, 112 ACPA-negative RA patients diagnosed before the new ACR/EULAR criteria were followed regularly. Only 48 patients fulfilled the 2010 ACR/EULAR criteria for RA  and were further analysed. Their average age was 59 years old (range: 21-90), and the male-to-female sex ratio was 1 to 3.
Evolution of the initial diagnosis and the proportion of patients reclassified with alternative diagnoses were analysed.
Results Half of the patients (N=24) evolved to a different diagnosis than RA within 2.25 years [IQ25-75: 0,27-10]. New diagnoses were established from clinical signs, biological characteristics and patient history. Five had spondyloarthritis  including 3 psoriatic arthritis (fulfilling the Caspar criteria ), 5 had infectious diseases (1 Whipple disease, 1 Lyme borreliosis, 1 infective endocarditis, 1 helminthiasis and 1 parvovirus infection), 3 had polyarticular chondrocalcinosis, 1 had Sharp's syndrome, 1 had hyperthyroidism, 1 had Sjögren's syndrome, 1 had anti-synthetase syndrome, 1 had atrophic chronic polychondritis, 1 had Remitting seronegative symmetrical synovitis with pitting edema (RS3Pe), 1 had adult-onset Still's disease, 1 had granulomatosis with polyangiitis, 1 had Mediterranean fever, 1 had sarcoidosis and 1 had lymphoma. Six were RF positive (25%). Sixteen had received a treatment, combined or not, with: steroids (N=13), DMARDs (N=14), or biologics (N=10). Additionally 10 of the 16 treated patients (62%) were non- or bad responders.
The other half (N=24) had confirmed RA. Among them, 11 initially ACPA-negative became positive. ACPA positivity could be delayed up to five years after the first clinical signs. For the 13 remaining ACPA-negative patients, no new clinical event could lead towards an alternative diagnosis. Median follow-up was 9 years [IQ25-75: 7-14]; RF was positive for 17 (78%) patients; the majority had received a treatment, combined or not, with: steroids (N=14), DMARDs (N=19), or biologics (N=16). Forty percent were non- or bad responders.
Conclusions ACPA-negative polyarthritis patients, even those fulfilling ACR/EULAR 2010 criteria, should be followed up for several years to see whether their diagnosis evolves.
Among differential diagnoses, spondyloathropathies are the most common, especially psoriatic arthritis. Rare and atypical diseases must be considered when evolution seems non classical, all the more so in cases of non-response to DMARDs or biologics.
Rheumatologists must be extremely vigilant to extra-articular manifestations, particularly cutaneous involvement and patient history. RF positivity remains a solid argument in favor of RA diagnosis and the ACPA status should be regularly checked.
Aletaha D, et al. Arthritis Rheum 2010;62:2569–81
Rudwaleit M, et al. Ann Rheum Dis 2011;70:25–31.
Pedersen OB, Junker P. Ann Rheum Dis 2008;67:1495–6.
Disclosure of Interest None declared