Background: Primary non-response (PNR) to infliximab (IFX) has been reported in up to one third of patients with inflammatory bowel disease (IBD) and represents a major therapeutic challenge. This study aimed to asses if IFX trough levels during induction therapy differed between patients with and without PNR to IFX and to calculate therapeutic thresholds.
Methods: Retrospective cohort study including all 166 IFX-naïve IBD patients (Crohn's disease (CD) n=93; ulcerative colitis (UC) n=72) who had received IFX induction therapy (5 mg/kg at weeks 0, 2 and 6) at a tertiary center since 2009. Corresponding to each infusion time point during induction therapy (weeks 2 and 6) and just after induction therapy (week 14) IFX trough levels were measured in a total of 322 blood samples using a validated automated immunofluorometric assay. Clinical response was defined as a reduction of >3 points in clinical scores from baseline (Harvey-Bradshaw Index or Partial Mayo Score) and was assessed at week 14. PNR was defined as no clinical improvement during IFX induction therapy resulting in discontinuation of IFX treatment.
Results: 18 patients (11%) (UC n=7; CD n=11) were classified as having PNR, whereas 148 (89%) responded to IFX induction therapy and were classified as responders. Patients with PNR had significantly lower IFX trough levels compared to responders at week 2 (p<0.05): Mean (SD) 17.2 (8.6) vs. 22.7 (8.1) μg/mL. The difference between the two group was even more pronounced at week 6 (p<0.05): Median (IQR) 8.4 (4.3–16.4) and mean (SD) 17.9 (10.2) μg/mL, respectively. IFX cut-off values of 22.9 μg/mL (area under the ROC curve (AUC) = 0.667, sensitivity=51%, specificity=80%) at week 2 (p<0.05) and 11.8 μg/mL (AUC = 0.707, sensitivity = 72%, specificity = 77%) at week 6 (p<0.05) were associated with an overall clinical response to the induction therapy. In 8 patients with PNR, who had blood samples available also at week 14, IFX trough levels were compared to corresponding values from responders (n=58). Also at this time point IFX trough levels were significantly lower in patients with PNR compared to responders to the preceding IFX induction therapy (median (IQR) 0.5 (0.1–3.4) vs. 5.6 (3.0–10.0) μg/mL, p<0.01). An IFX cut-off value of 1.4 μg/mL (AUC = 0.851, sensitivity = 91%, specificity = 75%) was associated with an overall clinical response (p<0.01).
Conclusions: IBD patients with PNR to IFX treatment have less circulating IFX than responders throughout the induction phase. This indicates a need for early dose intensification which is likely due to a high inflammatory load. Therapeutic drug monitoring may have value also during the IFX induction phase.