Predictive factors of abatacept therapy discontinuation in patients with rheumatoid arthritis.

The aim of this paper was to look for predictors of abatacept (ABA) therapy discontinuation in patients with rheumatoid arthritis (RA). Seventy-one RA patients treated with ABA were followed up. Demographical, clinical, and laboratory parameters of the patients, including peripheral blood T and B cell populations, different rheumatoid factor and anti-cyclic citrullinated peptide autoantibodies isotypes, and serum free light chains were evaluated. Comparing patients who discontinued ABA with those still in therapy we observed: a higher proportion of smokers (51.9 vs 25.6 %; p = 0.03); a non significant lower proportion of anti-cyclic citrullinated peptide positivity (76 vs 89.5 %; p = 0.13); a lower proportion of terminally differentiated effector memory cells (TDEM) among total CD8+ T lymphocytes at baseline (22.0 % (7.8-39.2) vs 38.7 % (20.7-55.9); p = 0.002). Logistic multivariate analysis showed that only the proportion of CD8+TDEM T cells was an independent predictive factor of therapy discontinuation (OR (95 % IC) = 6.2 (1.2 to 30.8); p = 0.026). Receiver-operating characteristic analysis showed a significant performance of this biomarker for prediction of therapy discontinuation (using a cut-off of 30.6 %: AUC: 0.760 ± 0.07; p = 0.002). Patients with a low proportion of CD8+TDEM at baseline had a higher probability of discontinuing the treatment during time (log-rank test: p < 0.01). T cell characterization for identification of TDEM CD8+ T cells might be a useful test to predict discontinuation of ABA therapy.