Background The anti-TNF monoclonal antibody Infliximab (Ifx), has proven effective in treating rheumatoid arthritis (RA), although in 40% of cases may fail, mainly due to immunogenicity. A good clinical response is usually associated with high serum drug levels; however, it is not clear why some patients have a faster drug clearance since the beginning of the therapy. Formation of immunocomplexes between antibodies to Ifx (ATI) and Ifx can increase drug clearance, leading to treatment failure.
Objectives To analyze whether serum Infliximab trough levels (ITL) at the induction phase were related to Ifx disappearance and clinical outcomes at week (w) 54. The early development of immunogenicity as a related factor with low ITL was also investigated.
Methods In this observational retrospective study ITL were measured from 66 RA patients from the prospective biological cohort of La Paz Hospital. Serum samples were taken at w2, w6, w14 and w22. Serum-dependent receiver operating characteristics (ROC) curves were used to establish the ITL value that better predicts the absence of Ifx at w54. ATI were measured by bridging ELISA and by an acid-dissociation method without drug interference IDK (Immundiagnostik®, Germany). Patients were grouped as ITLpos if they had detectable Ifx at w54 and ITLneg otherwise.
Results ITLneg patients (n=25) had significantly lower levels at all time points than ITLpos (n=41). Based on ROC values ITL at w6 (4.44 μg/ml) had the best predictive value for disappearence of Ifx at w54 with a 70% sensitivity (95%CI 45.7–88.1), 95% specificity (95%CI 83.1–99.4) and positive likehood ratio of 14. Most patients in low disease activity or remission at w54 had at w6 ITL upper the predictive cut-off [20/44 (45%) upper cut-off vs 3/20 (15%) under cut-off p=0.02] and most EULAR responder at w54 had ITL upper the predictive cut-off at w6 [33/43 (77%) vs 10/21 (48%); p=0.08]. Treatment survival of patients with ITL upper 6w cut-off was longer: 5 years (1.6–5.0) vs 1.7 years (0.2–0.6); p=0.012.
In the multiple logistic regression analysis, after adjusting for confounders (age, sex, body mass index, baseline DAS28, PCR, TNF and IL6) with ITL at w2 and at w6, the absence of Ifx levels at w54 was significantly associated with ITL under the cut-off at w2 (OR: 15.85; 95%IC 2.95–85.03; p=0.01), at w6 (OR: 86.64; 95%IC 6.58–1139.99; p=0.001) and no MTX use (OR: 12.26; 95%IC 1.83–82.22; p=0.001 for w2; OR: 6.9; 95%IC 1.04–45.84; p=0.04 for w6)
Most patients with ITL under the cut-off at w6 were positive for ATI along the first year [15/20 (75%) under cut-off vs 5/44 (11%) upper cut-off, p<0.0001]. Most ATI were detected earlier by the IDK than with bridging.
Conclusions Low ITL at early stages (w2 and w6) are associated with the Ifx absence, the early drop-out of the treatment and the clinical outcome at w54, being the presence of ATI the main reason for the low early circulating drug levels. We also conclude that the cut-off value at w6 (4.44 μg/ml) provides the clinicians with a useful prognostic tool of treatment efficacy.
Acknowledgement This work is partially funded by a non restricted grant from Pzifer and from Leti laboratories.
Disclosure of Interest None declared