Background/Purpose:  Predictors of rheumatoid arthritis (RA) response to anti-TNF agents have been described. Except for certolizumab, the value of interim analysis before 24 weeks of treatment is not clear to predict response. The purpose of this observational study was to evaluate predictors of response of anti-TNF agents in RA patients in a real-life situation tertiary center. 

Methods: An inception cohort of all RA patients (ACR 1987) who received the first anti-TNF dose in the rheumatology outpatient infusion center between July, 2007 and July, 2012 was analyzed. Demographic, clinical and laboratorial data, including disease activity score (DAS28) were recorded at baseline and each 6-8 weeks and at each therapeutic change or interruption. Drug response at 24 weeks was evaluated. Remission was defined as DAS28 was ≤ 2.6 at 24 weeks. Drug survival was also analyzed. Baseline data, early (6-8 weeks) and late (14-16 weeks) significant (≥0.6) and very significant (≥1.2) changes in DAS28 were evaluated in order to predict remission at 24 weeks. Any cycle of anti-TNF was included.

Results:   One hundred sixty-six RA patients were included, corresponding to 215 cycles of anti-TNF treatments. Patients were predominantly female (n= 91%;n=151); and had a positive rheumatoid factor (RF) (80%; n=133). Mean age and mean disease duration: 49.2 ± 12.2 and 12.5 ± 8.8 years, respectively. The cycles analyzed corresponded to the first one in 54% (n=116), the second in 40% (n=85) and the third in 6% (n=14). Infliximab corresponded to the main used drug (41%; n=89 cycles), followed by adalimumab (32%; n=69) and etanercept (27%; n=57). The analysis per cycle showed that, after 24 weeks, only 35 (16%) patients achieved remission. Patients who achieved remission and those who did not were similar according to age, gender, RF positivity, disease duration, number of cycles of anti-TNF, drug used, or association with methotrexate, leflumomide or any traditional disease-modifying antirheumatic drug (p>0.05). Patients who achieved response had lower disease severity at baseline according to DAS28 (4.7 ± 1.1 vs. 5.7 ± 1.2; p<0.001), number of painful joints (10 ± 9 vs. 19 ± 14; p<0.001), pain visual analogue scale (VAS) (5 ± 2.6 vs. 6.4 ± 2.2; p=0.01), patient global VAS (4.9 ± 2.4 vs. 6.5 ± 3; p<0.001), physician VAS (4.9 ± 2.1 vs. 6.4 ± 2.7; p<0.001), ESR (20 ± 14 vs. 32 ± 24; p=0.006), and HAQ (1.2 ± 0.7 vs. 1.5 ± 0.7; p=0.012). Patients who achieved remission at 24 weeks also had deeper falls in DAS28 at 6-8 weeks (1.77 ± 1.27 vs. 1.06 ± 1.38; p=0.009). Moreover, significant (≥ 0.6) early (weeks 6-8) falls in DAS28 [ 26 (84%) vs. 102 (64%); p=0.032) and very significant (≥1.2) late falls (weeks 14-16) were also related to remission [ 22 (71%) vs. 76 (48%); p=0.018). Significant rises (≥0.6) in DAS28 in any interim analysis were less frequent among patients who achieved remission than in patients who did not [5 (17%) vs. 47(36%), p=0.047]. Mean drug survival was 57.1 ± 50.1 weeks, without differences among the anti-TNF drugs (p>0.05). At 52 weeks, remission rate was stable (14%), but has fallen continuously every 6 months: 7% to 5% (18 to 24 months), 3% to 1% (30 to 48 months), up to 0 in 54 months. 

Conclusion:  The low and not sustained remission rate, regardless of the anti-TNF treatment, with short drug survival, shows the actual difficulties in treating RA in a real-life setting. Interim positive and negative significant variations in disease activity before 24 weeks of treatment can help to avoid prolonged and unnecessary exposition to such drugs. 

« Back to 2016 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/early-das28-drop-is-a-predictor-for-clinical-response-to-anti-tnf-agents-in-patients-with-rheumatoid-arthritis-an-observational-study-of-a-real-life-inception-cohort/