Cumulative exposure to infliximab during induction therapy predict remission in patients with Crohn’s disease and ulcerative colitis
Juma Ayubu Mohamedi (1), Debby Thomas (1), Jean-François Rahier (2), Edouard Louis (3), Filip Baert (4), Pieter Dewint (5), Guy Lambrecht (6), Séverine Vermeire (7), Peter Bossuyt (8), Denis Franchimont (9), Claire Liefferinckx (9), Erwin Dreesen (1)
(1) Department of Pharmaceutical and Pharmacological Sciences, KU Leuven, Leuven, Belgium, (2) Department of Gastroenterology, CHU UCL Namur, Yvoir, Belgium, (3) Department of Gastroenterology, Centre Hospitalier Universitaire Sart-Tilman, ULG, Liège, Belgium, (4) Department of Gastroenterology, AZ Delta, Roeselare-Menen-Torhout, Belgium, (5) Department of Gastroenterology, AZ Maria Middelares, Ghent, Belgium, (6) Department of Gastroenterology, AZ Damiaan, Oostende, Belgium, (7) Department of Gastroenterology and Hepatology, University Hospitals Leuven, Leuven, Belgium, (8)Department of Gastroenterology, Imelda General Hospital, Bonheiden, Belgium, (9) Department of Gastroenterology, Hospital Erasme, ULB, Brussels, Belgium
Objectives:Infliximab, an anti-tumor necrosis factor monoclonal antibody, has revolutionized the pharmacological management of inflammatory bowel diseases (IBDs), but non-response to induction therapy remains a challenge. While cumulating data reinforce the evidence base for trough concentration (TC)-guided therapeutic drug monitoring (TDM) during infliximab maintenance therapy, consistent evidence for a role during induction therapy is still lacking [1]. It was previously suggested that the overall exposure, expressed as the area under the infliximab concentration–time curve (AUC), is the true driver of efficacy of induction treatment, yet this has not been confirmed [2].
The objective of this study was to investigate the infliximab pharmacokinetics (PK) and exposure–response relationship during induction therapy of patients with Crohn’s disease (CD) and ulcerative colitis (UC). More specifically, we aimed to compare the predictive performance of various exposure metrics (TCs and AUCs) for treatment efficacy.
Methods:We performed a prospective, multicenter Phase 4 study to investigate the PK and exposure–response relationship of infliximab induction therapy in IBD patients (EudraCT 2015-004618-10). A rich PK sampling schedule contributed 13 infliximab serum concentrations per patient between the start of therapy and week (w)30 (3 peak, 5 intermediate, 5 trough; determined using an enzyme-linked immunosorbent assay). Antibodies to infliximab (ATI) were determined using a drug-tolerant affinity-capture elution assay in trough samples at w6, w14, and w30.
A nonlinear mixed-effects population PK (popPK) model was developed to describe the infliximab concentration–time course from w0 to w30 of treatment (Monolix Suite 2023R1, Simulations Plus, California, USA). The final popPK model including covariate effects was built using the Stochastic Approximation for Model Building Algorithm. Tested covariates were body weight, age, sex assigned at birth, disease duration, serum albumin, and ATI.
The primary endpoint was remission at w30, defined as combined clinical and biological remission (Harvey–Bradshaw Index [HBI] ≤4 and C-reactive protein ≤5 mg/L for CD, and partial Mayo score ≤2 and fecal calprotectin <250 µg/g for UC). Receiver operating characteristics (ROC) analyses were used to evaluate the predictive performance of infliximab TCs at w2, w6, and w14 for remission at w30 in comparison with cumulative AUC from w0 to w2, w6, and w14.
Results: A total of 74 patients contributed to 900 infliximab concentrations. Eighteen (24.3%) out of 74 patients achieved remission (13 CD, 5 UC). ATIs were detected in ten samples of nine patients.
A two-compartment popPK model with linear elimination best described the infliximab concentration–course. Serum albumin was a covariate on clearance (CL), and body weight was a covariate on CL and the central volume of distribution (Vc). Final popPK model parameters estimates were 0.31 L/d (relative standard error, 3.8%) for CL, 2.93 L (3.0%) for Vc, 1.98 L (8.5%) for the peripheral volume of distribution (Vp), and 0.39 L/d (14.3%) for the intercompartmental clearance (Q).
The TCs at w2 (28.5 ±13.4 mg/L vs 23.1 ±11.4 mg/L, P=0.156), w6 (21.6 ±12.74 mg/L vs 14.71 ±9.7 mg/L, P=0.059) and w14 (6.4 ±4.9 mg/L vs 4.5 ±4.5 mg/L, P=0.187) did not show significant differences between patients with and without remission, resp.
The AUC from w0 to w2 (776 ±140 mg×d/L vs 668 ±163 mg×d/L, P=0.018), w6 (2,264 ±485 mg×d/L vs 1,916 ±582 mg×d/L, P=0.031), and w14 (4,026 ±943 mg×d/L vs 3,303 ±1165 mg×d/L, P=0.022) was significantly higher in patients achieving remission than in those who did not, respectively.
ROC analysis showed numerically superior predictive performance of AUCs as compared to TCs at all time points, w2 (AUROC 0.73, 95% confidence interval [CI] 0.57–0.88 vs 0.62, 95%CI 0.45–0.78), w6 (0.68, 95%CI 0.52–0.84 vs 0.66, 95%CI 0.49–0.83), and w14 (0.70, 95%CI 0.55–0.86 vs 0.62, 95%CI 0.46–0.79).
Conclusions:Cumulative exposure to infliximab during induction predicts remission at w30. This key observation may impact the management of IFX therapy which is currently based on TCs. As opposed to TC-guided TDM, AUC-guided infliximab (precision) dosing would demand an informative sampling scheme and creates an opportunity for model-informed dosing tools.
References:
[1] Syversen SW et al. JAMA (2021) 325:1744–1754.
[2] Dreesen E et al. Br J Clin Pharmacol (2019) 85:782–795.
