Non-Clinical Pharmacology of Anti-CD22 Antibody Drug Conjugates.

CD22 is a B-cell specific antigen expressed on both normal and malignant cells. It is internalized rapidly upon binding to its antibody, making it an ideal target for immunotoxin therapy. Anti-CD22 is a humanized monoclonal antibody that recognizes human and cynomolgus monkey CD22 but not the rodent CD22. When conjugated with mitotic inhibitors such as DM1 (a maytansinoid) or monomethylauristatin F (MMAF) as anti-CD22-MCC-DM1 or anti-CD22-MC-MMAF, the conjugates have been shown to regress tumors completely after a single IV bolus in a variety of xenograft mouse models. Pharmacokinetics (PK), pharmacodynamics (PD), efficacy and safety studies were conducted to explore the non-clinical pharmacology of the two anti-CD22 conjugates (ADC). Dose response of anti-tumor activity, characterized in the Bjab-Luc xenograft model, yielded an ED50 of ∼1 mg/kg for MC-MMAF. MCC-DM1 was about 2-fold less potent compared to MC-MMAF in this model. Anti-CD22 antibody demonstrated very limited antibody dependent cell mediated cytotoxicity and no complement dependent cytotoxicity in vitro. Antigen-independent pharmacokinetics were assessed in mouse (single IV bolus, 0.5 or 5 mg/kg) and rat (single IV bolus, 10 mg/kg), while the effect of target mediated disposition was studied in cynomolgus monkey (two IV bolus with three weeks apart, vehicle, 10, 20 or 30 mg/kg). Conjugation with DM1 or MMAF appeared to have minimal impact on the PK of anti-CD22 when the PK of naked anti-CD22 was compared with the total antibody PK profile in all three species. Pharmacokinetic parameters of both total antibody and ADCs were generally dose proportional at the dose range studied. Free DM1 in plasma increased with dose, while free MMAF was not detected. Safety and PD were also characterized in the cynomolgus monkey. Both ADCs were well tolerated up to 30 mg/kg through out the entire study period (42 days). For PD, B-cell depletion was analyzed via FACS. Substantial decreases of CD20+ B-cells were observed in peripheral blood at all dose levels. In tissues, reduction of CD20+ B-cells was observed in spleen, bone marrow and in the lymph nodes, although to a lesser extent. No anti-drug antibodies were detected in the cynomolgus monkey at the doses tested. Based on the preclinical data, human efficacious dose and therapeutic window were estimated to be ∼2–5 mg/kg and 5–10, respectively. In summary, both ADCs exhibit an encouraging nonclinical efficacy, pharmacokinetic, pharmacodynamic, and safety profile that supports clinical development for the potential treatment of NHL.