Objective: Antibodies to CA 125 have been used to predict relapse of ovarian cancer, but have performed poorly as therapeutic agents. One rationale for this is antibody binding to circulating shed antigen. Our aim in this study was to develop antibodies to human CA 125 that have enhanced selectivity for the cell-associated form of the antigen.
Methods: Monoclonal antibodies were raised to a recombinant fragment of CA 125 that included sequence proximal to the putative membrane attachment site. Antibodies were characterized in terms of their binding site, affinity and selectivity for cell-associated CA 125.
Results: In assays using patient-derived CA 125, a subset of high-affinity (KD <5 nM) monoclonal antibodies demonstrated a 10- to greater than 200-fold increase in selectivity for cell-associated CA 125 when compared with controls. Based on mapping of the various monoclonal antibodies obtained, it was determined that shedding of CA 125 most likely occurs in the most C-terminal repeat domain.
Conclusion: Results from competition analysis using patient-derived shed antigen predict that the antibodies described in this study may have significantly enhanced tumor-targeting properties when compared with existing antibodies to CA 125 in a tumor environment having high concentrations (>10,000 CA 125 units) of shed CA 125.