Association of Serum Levels of Ustekinumab, Vedolizumab, and Faecal Calprotectin in Paediatric Patients with Inflammatory Bowel Diseases: A Prospective Observational Study

Abstract

Background and objectives: Ustekinumab (USTE) and vedolizumab (VEDO) are increasingly used in paediatric patients with inflammatory bowel diseases (pIBD). However, data on the usefulness of therapeutic drug monitoring (TDM) in children are scarce. The primary objective of this study was to evaluate the association between disease activity, measured by faecal calprotectin (F-CPT), and serum trough levels (TLs) of USTE and VEDO. Secondary outcomes were to explore factors potentially associated with the outcome and exposure, to determine the optimal USTE or VEDO dose that predicts remission (defined as F-CPT < 250 µg/g), to validate our hypothesis using a proof-of-concept cohort (POCC) and to assess the occurrence of serum antibodies to USTE and VEDO.

Methods: This was a prospective single-centre observational study performed at the University Hospital Motol, Prague, Czech Republic. Of the 87 patients (51 Crohn's disease (CD), 30 ulcerative colitis (UC), and 6 IBD unclassified (IBD-U)), drug serum TLs and antibodies were measured in 282 observations (49 treatment courses) of USTE and 359 observations (38 courses) of VEDO. Serum and stool samples were collected before each study drug application during both the induction and maintenance phases of the treatment throughout the entire study period (January 2020 to June 2024). Clinical and laboratory data were obtained from the nationwide prospective registry CREdIT. Patients with perianal disease and those with previous major bowel surgery were not excluded from the study. As a POCC, we analysed a group of pIBD treated at our centre with anti-TNF agents-adalimumab or infliximab.

Results: In a linear multiple regression mixed model, an association was observed between logF-CPT levels and USTE treatment duration (β -0.0010, 95% confidence interval (CI) -0.0015 to -0.0006, p < 0.001) but not with USTE TLs (p = 0.12). VEDO TLs and logF-CPT levels were negatively associated both in the linear (β -0.0173, 95% CI -0.0292 to -0.0053, p = 0.005) and categorical models (p = 0.026), even after adjusting for time. A VEDO TL of 15.1 µg/mL showed the best, though still poor, combination of sensitivity (0.82) and specificity (0.32) to predict F-CPT < 250 µg/g (area under the curve (AUC) 0.56, 95% CI 0.49-0.63). Intensification, induction phase, undetectable TLs, and type of IBD (CD, UC, IBD-U) were not associated with logF-CPT. Slightly elevated anti-drug antibodies were detected in 5 USTE and 16 VEDO observations, with no clinical implications.

Conclusions: TDM of USTE does not appear to be useful in pIBD. TDM of VEDO may assist in therapeutic strategy decisions, although establishing clinically useful cut-offs remains challenging.