The Influence of Underlying Disease on Rituximab Pharmacokinetics May be Explained by Target-Mediated Drug Disposition.

BACKGROUND AND OBJECTIVES

Rituximab is an anti-CD20 monoclonal antibody approved in several diseases, including chronic lymphocytic leukemia (CLL), diffuse large B-cell lymphoma (DLBCL), follicular lymphoma (FL), rheumatoid arthritis (RA), and anti-neutrophil cytoplasmic antibody-associated vasculitis (AAV). The influence of underlying disease on rituximab pharmacokinetics has never been investigated for several cancer and non-cancer diseases simultaneously. This study aimed at assessing this influence using an integrated semi-mechanistic model accounting for target-mediated elimination of rituximab.

METHODS

Rituximab concentration-time data from five studies previously published in patients with CLL, DLBCL, FL, RA, and AAV were described using a two-compartment model with irreversible binding of rituximab to its target antigen. Both underlying disease and target antigen measurements were assessed as covariates.

RESULTS

Central volume of distribution was [95% confidence interval] 1.7-fold [1.6-1.9] higher in DLBCL than in RA, FL, and CLL, and it was 1.8-fold [1.6-2.1] higher in RA, FL, and CLL than in AAV. First-order elimination rate constants were 1.8-fold [1.7-2.0] and 1.3-fold [1.2-1.5] higher in RA, DLBCL, and FL than in CLL and AAV, respectively. Baseline latent antigen level (L) was 54-fold [30-94], 20-fold [11-36], and 29-fold [14-64] higher in CLL, DLBCL, and FL, respectively, than in RA and AAV. In lymphoma, L increased with baseline total metabolic tumor volume (p = 6.10). In CLL, the second-order target-mediated elimination rate constant (k) increased with baseline CD20 count on circulating B cells (CD20, p = 0.0081).

CONCLUSIONS

Our results show for the first time that rituximab pharmacokinetics is strongly influenced by underlying disease and disease activity. Notably, neoplasms are associated with higher antigen amounts that result in decreased exposure to rituximab compared to inflammatory diseases. Our model might be used to estimate unbound target amounts in upcoming studies.