Anti-CD20 Therapy Reliance on Antibody-Dependent Cellular Phagocytosis Affects Combination Drug Choice

The efficacy of many therapeutic unconjugated monoclonal antibodies (mAbs), including those targeting CD20 in CLL, requires immune cell-mediated cytotoxicity. mAbs have often been optimized for natural killer (NK) cell antibody-dependent cellular cytotoxicity (ADCC) activity. However, in vivo mouse studies have shown that antibody-dependent cellular phagocytosis (ADCP) by macrophages is the major mechanism of clearance of circulating B cells by anti-CD20 mAbs. To directly compare ADCC versus ADCP, we previously used a panel of anti-CD20 mAbs (rituximab, ofatumumab, obinutuzimab, ocaratuzumab) to test cytotoxicity of paired human NK cells and monocyte-derived macrophages (hMDM) against CLL cells in vitro. All mAbs demonstrated ADCP activity at least 10-fold greater than ADCC as measured by CLL cell depletion per effector cell. Moreover, ADCC and ADCP activity levels did not correlate, meaning that ADCC cannot be used as a surrogate measure of ADCP for these mAb. This could explain why mAb optimization for ADCC activity has often failed to translate into more efficacious treatment. Thus, ADCP may be an effective translational measurement of anti-CD20 mAb performance.

Because of the clinical interest in combining anti-CD20 mAb with targeted small molecule inhibitors, we began studying the effects of the Bruton tyrosine kinase (BTK) inhibitors on anti-CD20 mAb-mediated ADCP. Our initial studies showed that ibrutinib (IBR), but not acalabrutinib (Acala), significantly decreased anti-CD20 ADCP as measured by a flow cytometry-based assay that measures single timepoint cell collections. These types of assays cannot easily determine the kinetics and individual effector cell activity of ADCP. Thus, to more fully study the BTK inhibitor effects on ADCP, we developed a live cell time-lapse imaging method for measuring ADCP, utilizing recent advances in microscopy, cellular dye labeling, digital imaging, imaging software and computing. Whole-cell labeling of macrophages enabled visualization of internalized CLL cells as regions of dye exclusion or "voids". Because of the vast number of images acquired during live cell time-lapse imaging, we utilized computer software-aided image recognition and enumeration to measure the number of macrophages and voids inside each macrophage. As a measure of phagocytic engulfment, we developed a void index, which provides a relative measure of phagocytic engulfment per macrophage. Measuring ADCP in this manner replicates clinical observation of mAb therapeutic activity. Clinically, intravenous anti-CD20 mAb therapy typically induces a rapid decrease in circulating CLL cells (within hours), followed by a long period (days) of stable to increased levels of circulating cells. Similarly, our live cell time-lapse video assay shows initial rapid ADCP over the first 2 hours followed by a prolonged period of "hypophagia" with little ADCP for the remainder of the assay (imaged every 2 minutes for 8 hours). This "hypophagia" phenomenon may explain the resistance to therapeutic mAb observed clinically. With these new tools for quantitation of ADCP, we compared the effects of serial dilutions of IBR or Acala on ADCP. Overall, as measured by Area Under the Curve (AUC) analysis, IBR decreased phagocytic capacity of anti-CD20 mediated CLL cell ADCP at concentrations of 0.41 μM and above. By contrast, Acala did not begin to decrease AUC measurements until 3.7 μM, and subsequent AUC values were higher in Acala versus IBR-treated ADCP assays up to the highest tested drug concentration (100 μM). Similarly, the initial ADCP kinetics (void index / min over the first hour) reflected a decrease with IBR treatment at 0.41 μM that continued until a nadir was reached at 33 μM. In contrast, Acala did not induce a decrease in this kinetic measurement until 3.7 μM and a nadir was not reached (up to 100 μM). Thus, IBR significantly decreases ADCP by hMDM at concentrations much lower than a more specific BTK inhibitor, Acala. This result suggests that BTK inhibition has little to no effect on ADCP and furthermore suggests that IBR off-target effects decrease ADCP. IBR off-target candidates include other tyrosine kinases in the TEC (tyrosine kinase expressed in hepatocellular carcinoma) family. These data suggest that a a highly selective BTK inhibitor with little effect on ADCP could be a more suitable drug to combine with therapeutic mAb(s).