AFTER HOW MANY TIME DOES IT BECOME USELESS TO TRY TO CLASSIFY AN INFLAMMATORY RHEUMATISM?

Background: There is several models to predict the development ofrheumatoid arthritis in patients with early undifferentiated arthritis (UA). No study have concerned the future of the patients with UA in the medium and long term. Objectives: 1) to describe the population of UA at inclusion. 2) to identify predictive factors of the preservation of the status of UA. 3) to determine a time from which it is useless to try to classify a UA. Methods: Community-based adults (n = 310) who had never taken disease-modifying antirheumatic drugs (DMARDs) or steroids with swelling of ≥2 joints persisting for >4 weeks and lasting <6 months were recruitedbetween 1998 and 2002. They have been followed during 10 years. The diagnoses were annually analyzed by experts’ committee. UA were defined according to 2 approaches : “academic”, that is the impossibility to classify the rheumatism according to the official criteria; “pragmatics” in which were not any more considered as unclassified the patients with a diagnostic hesitation without therapeutic impact. UA were compared with the classified arthritis (CA) at inclusion; a survival analysis of the status of UA have been made using a Cox regression. Results: according to the “academic” method, there were 180 UA / 312 patients at inclusion. There was 73,9 % of women versus 60,6 % to the CA ( p=0,014 ); the number of swollen joints/ 44 ( NSJ) was lesser (7,2 vs 12,1; p < 0,001); lower limbs less often affected (50,0 % vs 72,7 %; p < 0,001); the DAS of 2,5 vs 2,9 (p < 0,001); ESR and CRP less high (p < 0,05); rheumatoid factor and anti-CCP-2 antibodies less frequent (p < 0,001), finally, the erosion Sharp score less high (0,3 vs 0,9; p=0,001). According to the “pragmatic” filter, 162/312 remained unclassified. There was no significant feminine ascendancy, the NSJ was lesser (6,8 vs 9,3; p < 0,001); the big joints (56,8 % vs 62,2 %; p=0,047) and lower limbs (50,6 % vs 59,6 %; p < 0,001) less often affected; the DAS lesser (2,4 vs 2,7; p < 0,001); rheumatoid factor and anti-CCP-2 antibodies less often positive (p < 0,01); finally, the erosion Sharp score less high (0,3 vs 0,6; p=0,002). The survival analysis of the status of UA according to the “academic” approach found, as predictive factor of classification : a rate of anti-CCP-2 > 3 times the normal (Hazard Ratio ( HR) 3,01, p < 0,001), the painful involvement of the only lower limbs - protective factor (HR=0,24; p=0,017). According to the “pragmatic” approach: a morning stiffness > 60 ′ (HR=1,64, p=0,002), a NSJ > median (HR=1,62; p=0,003), a swelling of upper limbs joints (HR=2,67; p=0,023), an involvement of the lower and upper limbs (HR=1,44; p=0,047). The protective factors were : swelling of the small joints (HR=0,17; p < 0,001) and involvement of the small and big joints (HR=0,57; p=0,002). The inventory of the diagnoses done according to time shows that after 6 years of evolution, only 3 patients lost the status of UA. Conclusions: UA distinguishes from CA by a lower level of activity. The predictive factors of the preservation of the status of UA are in touch with a less important severity of the rheumatism. It seems useless to try to classify a rheumatism after 6 years of evolution. Disclosure of Interest: None DeclaredCitation: , volume 72, supplement s3, year 2013, page Session: Poster session Friday ( )