Characterization of murine CEACAM1 in vivo reveals low expression on CD8(+) T cells and no tumor growth modulating activity by anti-CEACAM1 mAb CC1

McLeod, Robbie L.; Angagaw, Minilik H.; Baral, Toya Nath; Liu, Liming; Moniz, Raymond Joseph; Laskey, Jason; Hsieh, SuChun; Lee, Mike; Han, Jin-Hwan; Issafras, Hassan; Javaid, Sarah; Loboda, Andrey; Sadekova, Svetlana; O'Connor, Joann A.; Tse, Archie; Punnonen, Juha

The PK properties of CC1 were studied in Balb/c mice ( Table 1 ). The PK parameters were non-linear between 10 mg/kg and 30 mg/kg with a significantly over-dose proportional increase for both CMax and AUC0-Inf and a significant difference in CL/F. At 10 mg/kg, CC1 exhibited fast clearance (CL/F 111 ± 34 mL/h/kg) with concentration dropping from ~40 ug/mL at ~6 h (CMax) to below 1 ug/ mL 48 hr post dosing. This is an uncommon PK feature for an IgG1 mAb, suggesting that there is a significant sink effect mediated by target-associated drug disposition which could in part be due to an abundant expression of CEACAM1 on myeloid cells, including neutrophils [24] . At and above 30 mg/kg, the CC1 clearance was significantly reduced (CL/F 13 ± 5 mL/h/kg at 30 mg/kg and 11 ± 3 for 45 mg/kg), resembling a profile of non-targeted IgG1. Interestingly, above the 30 mg/kg dose, PK was linear because the CMax and AUC0-Inf were dose proportional between 30 mg/kg and 45 mg/kg. Consequently, the 30 mg/kg dose was chosen for further experiments in tumor models. Moreover, the simulated multi-dose (every 2 days) PK profile for the 30 mg/ kg dose shows that the trough level of CC1 is approximately 313 nM, which is expected to sufficiently block CC1 binding in vivo. Indeed, full receptor occupancy of CEACAM-1 at a cellular level after in vivo CC1 treatments (30 mg/kg) was demonstrated by flow cytometry staining on CT26 tumor infiltrates using a fluorescence-labeled CC1. Both in the CC1 monotherapy group and the muDX400+CC1 combination therapy group, fluorescence-labeled CC1 was not able to stain the target receptor, CEACAM-1 (staining of CD8 T cells shown, ( Figure 3B ). This complete blocking was consistent in all mice in the CC1 monotherapy and muDX400+CC1 combination therapy groups ( Figure 3B ). These findings CEACAM1 interaction with mTIM-3 was evaluated using mTIM3-CHO cells (C) and anti-TIM-3 mAb was used as a positive control (▲). Effects of CC1 antibody antibody (■) or isotype control (○) on the homophilic interaction of mCEACAM1 were evaluated using (D) mCEACAM1-His in protein ELISA, or (E) mCEACAM1-CHO cells in cELISA. Effect of Fab fragment of CC1 (■) and isotype control (○) on mCEACAM1-hFc binding to mCEACAM1-CHO cells was studied in cELISA (F). www.oncotarget.com confirm that CC1 treatment has fully engaged CEACAM-1 on tumor infiltrates at a dose level of 10 mg/kg. We studied the impact of muDX400 at a dose level of 5 mg/kg Q5d as this has been used in many syngeneic mouse tumor models with good efficacy and exposure [25] .