Treatment parameters modulating regression of human melanoma xenografts by an antibody-drug conjugate (CR011-vcMMAE) targeting GPNMB

Purpose To investigate the pharmacological properties of the CR011-vcMMAE fully human antibody-drug conjugate (ADC), such as dose titrations, quantitation of the time (days) to complete regression, pharmacokinetics, and schedule dependency. Our prior study characterized a fully human antibody to GPNMB covalently linked to monomethylauristatin E, CR011-vcMMAE, and further demonstrated cell surface staining of melanoma lines susceptible to the immunoconjugate's cytotoxicity (Clin Cancer Res 2005; 12(4): 1373-1382). Methods The human SK-MEL-2 and SK-MEL-5 melanoma xenografts were used in athymic mice to assess anti-tumor efficacy. After s.c. implantation, tumors became established (60-100 mg), and treatment commenced by i.v. injection of the immunoconjugate or vinblastine or paclitaxel. Short-term anti-tumor effects (inhibition of tumor growth) and long-term effects (complete regression) were observed. Results CR011-vcMMAE induced regression of established human SK-MEL-2 and SK-MEL-5 xenografts at doses from 1.25 to 80 mg/kg treatment when administered intravenously every 4 days (4 treatments); strikingly, regressions were not associated with re-growth during the observation period (200 days). The disappearance rate of implants was dose dependent (minimum time, 18.5 days). Detectable serum CR011-vcMMAE >= 1 mu g/mL (similar to 0.01 mu M) was observed for > 30 days post-dose; CR011-vcMMAE showed an elimination half-life of 10.3 days. A low volume of distribution suggested that CR011-vcMMAE was confined to blood and interstitial fluid. CR011-vcMMAE could be delivered by either a single bolus dose or by intermittent dosing (i.e., every 1, 2, 4, 8, or 16 days) with no discernible differences in the proportion of tumor-free survivors, indicating a lack of schedule dependency. The antibody-drug conjugate produced complete regressions, but the equivalent doses of free monomethylauristatin E or unconjugated antibody did not show anti-tumor effects. In addition, decreases in plasma tumor-derived human interleukin-8 coincided with tumor nodule disappearance. Conclusions Short-term anti-tumor effects and long-term effects (complete regression) were observed with CR011-vcMMAE, but not with the reference agents. These results suggest that CR011-vcMMAE may provide therapeutic benefit in malignant melanoma.