Introduction
Biological drugs have revolutionized the management of patients with inflammatory bowel disease (IBD) by enabling the resolution of symptoms and mucosal healing. Research in tumour necrosis factor alpha (TNF-α) inhibitors, the first biological drugs approved for inflammatory bowel disease, have shown that some patients require higher drug doses to achieve remission than others. Individualized dosing based on serum drug concentrations may lead to better treatment outcomes. In addition to TNF-α inhibitors, two new agents have become available: vedolizumab, an inhibitor of the α4β7, and ustekinumab, an inhibitor of interleukin-12 and -23. The relationship between drug concentrations and treatment outcomes for these agents is not known.
Aims
We aimed:
- to study the association between serum concentrations of vedolizumab and ustekinumab with treatment outcomes in IBD;
- to study whether single serum concentration measurements of ustekinumab have the same predictive value for treatment outcomes as measures of cumulative drug exposure (area under the curve);
- to develop a pharmacokinetic-pharmacodynamic model for vedolizumab in Crohn’s disease.
Methods
We performed a prospective single-centre study of patients with IBD starting treatment with vedolizumab (Crohn’s disease [CD] and ulcerative colitis [UC]) or ustekinumab (CD). Serum vedolizumab concentrations were measured at trough prior to each infusion, serum concentrations of ustekinumab were measured immediately after the first intravenous infusion and at week 2, 4, and 8. Symptomatic remission was assessed using the PRO-2 questionnaire or the Harvey-Bradshaw index (in the ustekinumab study) and endoscopic remission with colonoscopy. For patients treated with ustekinumab, biochemical remission was defined as faecal calprotectin <100 μg/g. The pharmacokinetic-pharmacodynamic model was developed from the LOVE-CD study (LOw countries VEdolizumab in CD) with non-linear mixed-effects modelling.
Results
We included 51 patients (28 CD, 23 UC) in the vedolizumab study. Median vedolizumab trough concentrations at weeks 6 (25.7 vs 15.6 mg/L; P = 0.015) and 22 (15.1 vs 4.9 mg/L; P = 0.001) were higher in patients with combined symptomatic and endoscopic remission. A threshold of 22 mg/L at week 6 (area under the curve (AUC) 0.733; 95% confidence interval 0.567–0.899) and 8 mg/L at week 22 (AUC 0.819; 95% confidence interval 0.692–0.946) was associated with subsequent combined remission.
We included 41 patients (all with CD) in the ustekinumab study. Peak concentrations were associated with endoscopic remission (AUC, 0.717; 95% CI, 0.517–0.916); 6 of 13 patients (46%) with peak concentrations above 105 mg/L (upper tercile) achieved endoscopic remission, compared with only 1 of 14 patients (7%) with peak concentrations below 88 mg/L (lower tercile). There was no significant difference between the associations of peak concentrations, week 2 concentrations, AUC through week 2, or later exposure measures (at weeks 4 and 8) with endoscopic or biochemical remission, nor was there an advantage of using cumulative exposure metrics.
We developed a pharmacokinetic-pharmacodynamic model for vedolizumab in CD from the first 110 patients participating in the LOVE-CD study. Clearance was higher and vedolizumab exposure lower in patients with lower serum albumin concentrations, in the presence of antibodies to vedolizumab and with previous exposure to other biological therapy. A week 22 vedolizumab concentration of 20.0 mg/L was predicted to yield a 35% probability of achieving endoscopic remission at week 26. Model-based simulations suggest that endoscopic remission rates of 46.5% or 40.0% could be reached with every 4 weeks dosing in patients naïve or previously exposed to biological therapy, respectively.
Conclusions
We demonstrated an exposure-response relationship for both vedolizumab and ustekinumab. There was no significant difference in the strength of association between different ustekinumab exposure metrics and subsequent treatment outcomes – peak concentrations appear to be an attractive early prognostic marker. Our pharmacokinetic-pharmacodynamic model for vedolizumab in CD provides a foundation for future research aiming to maximise endoscopic remission rates with vedolizumab. It should be noted that all these findings require testing in future interventional studies.
|