A Population Pharmacokinetic and Exposure–Response Model of Golimumab for Targeting Endoscopic Remission in Patients With Ulcerative Colitis

Background Unlike other anti–tumor necrosis factor alpha antibodies, golimumab does not deliver on its promise of effectiveness for treating patients with ulcerative colitis. We investigated the value of therapeutic drug monitoring for optimizing golimumab therapy. Methods We analyzed the golimumab pharmacokinetics data of 56 patients with moderate to severe ulcerative colitis. Induction and maintenance golimumab concentrations (296 venipuncture, 414 serum) were used to develop a population pharmacokinetic model. Exposure–response relationships were analyzed using the data of 40/56 patients with available endoscopy data. Receiver operating characteristic curve analysis was performed, and an exposure–response Markov model was developed, linking golimumab exposure to probabilities of transitioning between Mayo endoscopic subscore (MES) states from baseline to week (w)14. Results Golimumab pharmacokinetics was best described by a 2-compartment model with linear absorption and elimination. Antibodies to golimumab and previous biological therapy reduced golimumab exposure. Still, interindividual pharmacokinetic variability (IIVPK) remained largely unexplained. Endoscopic remission (ER; MESw14 ≤ 1) was achieved in 14/40 (35%) patients. Golimumab serum trough concentration thresholds of 7.4 mg/L (w6) and 3.2 mg/L (w14) predicted ER at w14 (positive predictive values [pv+] 83% and 91%, pv- 82% and 67%, respectively). The 3.2-mg/L target predicted 38% and 44% chances of achieving ER in patients with MESbaseline of 3 and 2, respectively. Conclusions Personalized, model-based induction dosing aiming at here-established target concentrations may account for IIVPK and thus provide patients with more equal chances of achieving ER. As <50% of patients attained the exposure targets, higher golimumab induction dosing requires investigation to secure its future in clinical practice.