Zhigang Wang 1, Wannee Kantasiripitak 1, Bram Verstockt 2,3, João Sabino 2,3, Marc Ferrante 2,3, Paul Declerck 1, Geert D’Haens 4, David Laharie 5, Séverine Vermeire 2,3, Erwin Dreesen 1
1 Department of Pharmaceutical and Pharmacological Sciences, KU Leuven, Leuven, Belgium 2 Department of Gastroenterology and Hepatology, University Hospitals Leuven, Leuven, Belgium 3 Department of Chronic Diseases and Metabolism, KU Leuven, Leuven, Belgium 4 Department of Gastroenterology and Hepatology, Amsterdam UMC, Amsterdam, Netherlands 5 Department of Hepato-gastroenterology and Digestive Oncology, Haut-Lévêque Hospital, Bordeaux UMC, Bordeaux, France
Objectives: Monitoring of monoclonal antibody clearance has been hypothesized to be an appealing approach for predicting treatment outcomes in patients with inflammatory bowel diseases[1]. We aimed to investigate the benefits of monitoring infliximab and ustekinumab clearance in patients with Crohn’s disease (CD) based on data from clinical trials.
Methods: Data were obtained from patients with moderate-to-severe CD starting infliximab (n=108) or ustekinumab (n=80) therapy [2,3]. Except for dose optimization during infliximab maintenance therapy, all patients received a standard dosing regimen. Endoscopic remission (REM, CD Endoscopic Index of Severity <3) was assessed at week (w)12 and w54 of infliximab therapy. Endoscopic response (RESP, ≥50% decrease from baseline in Simple Endoscopic Score for CD) was assessed at w24 of ustekinumab therapy. A posteriori prediction of the antibody clearance at each sampling time point (Bayesian forecasting using the measured drug concentration and covariates at the same time point) was performed using previously built population pharmacokinetic models (NONMEM 7.5) [4,5]. Covariates fecal calprotectin (FC), albumin, CD Activity Index, and antibodies towards infliximab were used to estimate infliximab clearance. Similarly, albumin and body weight was used to estimate ustekinumab clearance. A linear mixed-effects model and Receiver Operating Characteristics (ROC) analysis were used to explore the predictive ability of drug clearance, drug concentration, FC, and C-reactive protein (CRP) for endoscopic outcomes.
Results: The measured infliximab trough concentration during standard 5 mg/kg induction therapy (at w0, 2, 6) was significantly higher (median [IQR]: 25.3[19.1-29.7] vs 19.7[12.2-25.9] mg/L, p <0.001) in patients who achieved REM at w12 of induction, but were not significantly different (p =0.780) during maintenance therapy (dose optimizations) between patients with REM (6.5[4.7-9.3] mg/L) and without REM (5.1[3.9-8.5] mg/L) at w54 of maintenance. Infliximab clearance during induction was significantly lower in patients who achieved REM at w12 of induction (0.28[0.25-0.31] vs 0.32[0.27-0.38] L/d, p <0.001). Infliximab clearance during maintenance was significantly lower in patients with REM at w54 of maintenance (0.25[0.22-0.30] vs 0.28[0.24-0.33] L/d, p <0.001).
The ustekinumab trough concentration during induction therapy (at w8) was not significantly different (p =0.240) between patients with RESP (8.5[6.5-9.0] mg/L) and without RESP (6.3[3.4-10.5] mg/L) at w24. The ustekinumab concentration during maintenance was significantly higher in patients with RESP at w24 (3.1[2.7-5.0] vs 2.1[0.8-4.0] mg/L). The ustekinumab clearance was significantly lower in patients with RESP, during both induction (0.25[0.19-0.31] vs 0.24[0.22-0.27], p =0.035) and maintenance therapy (0.20[0.19-0.25] vs 0.25[0.20-0.28], p =0.011)
During infliximab induction therapy, the infliximab concentration (area under ROC curve [AUROC] 0.65[0.58-0.72]), the infliximab clearance (AUROC 0.67[0.60-0.74]), and FC (AUC 0.70[0.64-0.77]) were associated with REM at w12 but the differences in terms of predictive ability among the three were insignificant (p >0.05). During infliximab maintenance therapy, infliximab clearance (AUROC 0.58[0.53-0.64]) and FC (AUROC 0.64[0.59-0.69]) were associated with REM at w54 while concentration (AUROC 0.51[0.45-0.56]) was not. More importantly, no difference in predictive ability between infliximab clearance and FC, the current golden standard biomarker for the endoscopic outcome, was observed for REM at w54 (p >0.05).
During both ustekinumab induction and maintenance therapy, the ustekinumab concentration (induction AUROC 0.61[0.52-0.71]; maintenance AUROC 0.72[0.61-0.83]), FC (induction AUROC 0.61[0.51-0.71]; maintenance AUROC 0.72[0.58-0.86]), and ustekinumab clearance (induction AUROC 0.68[0.59-0.77]; maintenance AUROC 0.75[0.65-0.86]) were predictive for RESP at w24, and the differences in terms of predictive ability among the three were insignificant (p >0.05).
Conclusion: When patients are dose optimized, drug concentrations lose their predictive ability for REM, while infliximab clearance still can predict and show no difference in predictive ability as the FC. Monitoring of infliximab clearance may be used to guide clinical decision-making when dose adjustments reduce the predictive value of TDM. Furthermore, model-based monitoring of drug clearance may replace inconvenient stool collection for FC measurement.
References:
[1] Kantasiripitak et al. Expert Rev Clin Pharmacol (2022) 1–12.
[2] D’Haens, G. et al. Gastroenterology (2018) 154, 1343-1351.e1.
[3] Verstockt, B. et al. J Crohns Colitis (2019) 13, 864–872.
[4] Dreesen, E. et al. Br J Clin Pharmacol (2021) 87, 106–118.
[5] Wang, Z. et al. Br J Clin Pharmacol (2021) 88 (1), 323-335.
Reference: PAGE 30 (2022) Abstr 9971 [www.page-meeting.org/?abstract=9971]
Poster: Drug/Disease Modelling - Other Topics