Maria Dorn-Rasmussen[1], Sine Buhl[1], Jørn Brynskov[1], Tobias W Klausen[2], Nils Bolstad[3], David J Warren[3], Mark A Ainsworth[1], Casper Steenholdt[1]

Affiliates

1 Department of Gastroenterology, Copenhagen University Hospital Herlev, Denmark 2 Department of Haematology, Clinical Research Unit, Copenhagen University Hospital Herlev, Denmark 3 Department of Medical Biochemistry, Oslo University Hospital, Oslo, Norway

Abstract

Background: Therapeutic drug monitoring (TDM)-based strategies to optimize infliximab (IFX) therapy are hampered by lack of well-defined, clinically validated IFX concentration thresholds that distinguishes therapeutic and sub-therapeutic drug levels. Furthermore, it remains unknown whether such thresholds are constant over the course of therapy.

Methods: Prospective cohort study including all bio-naïve inflammatory bowel disease (IBD) patients (n=163) who had received IFX induction followed by standard maintenance therapy (5 mg/kg q8) for up to one year, and with ≥1 biobanked trough maintenance sample available (n=690). Clinical remission assessed at each IFX q8 maintenance infusion was defined as Harvey-Bradshaw Index ≤3 and/or Perianal Disease Activity Index 2, or Partial Mayo Scoring Index ≤1. IFX and anti-IFX antibodies (Abs) were measured by immunofluorometric assays. Optimal thresholds were defined by ROC analysis and Youden Index.

Results: IFX levels were significantly higher in IBD patients in clinical remission compared to those not in remission at all time points during the first year of IFX q8 maintenance therapy (w14: IFX median 6.8 µg/mL vs. 3.4, p<0.0001; w22: 4.6 vs. 2.2, p<0.01; w30: 4.6 vs. 1.9, p<0.001; w38: 4.4 vs. 1.7, p=0.06; w46: 4.3 vs. 2.1, p<0.01; w54: 4.6 vs. 1.4, p<0.01). The association between IFX concentration and remission was not influenced by the q8 infusion week as assessed by a generalized estimating equation, and it was thus appropriate to pool data from all weeks to calculate a common ROC curve for all weeks to establish a therapeutic threshold for the entire first year of maintenance therapy (IFX median 4.9 µg/mL vs. 2.4, p<0.01). Thus, a common therapeutic IFX threshold of 4.5 µg/mL was applicable during the first year of standard IFX q8 maintenance therapy (sensitivity 54% [49-59], specificity 73% [67-78], AUCROC 0.65 [0.60-0.69], p<0.0001). Exploratory analyses of patients stratified for disease type revealed comparable results. Anti-IFX Abs and markers of inflammatory load (CRP, HBI, albumin, IFX monotherapy) generally associated with low IFX levels.

Conclusion: A therapeutic IFX threshold of 4.5 µg/mL applies during the entire first year of standard IFX maintenance therapy.