Complement plays an important role in the immunotherapeutic action of the anti-CD20 mAb rituximab, and therefore we investigated whether complement might be the limiting factor in rituximab therapy. Our in vitro studies indicate that at high cell densities, binding of rituximab to human CD20+ cells leads to loss of complement activity and consumption of component C2. Infusion of rituximab in chronic lymphocytic leukemia patients also depletes complement; sera of treated patients have reduced capacity to C3b opsonize and kill CD20+ cells unless supplemented with normal serum or component C2. Initiation of rituximab infusion in chronic lymphocytic leukemia patients leads to rapid clearance of CD20+ cells. However, substantial numbers of B cells, with significantly reduced levels of CD20, return to the bloodstream immediately after rituximab infusion. In addition, a mAb specific for the Fc region of rituximab does not bind to these recirculating cells, suggesting that the rituximab-opsonized cells were temporarily sequestered by the mononuclear phagocytic system, and then released back into the circulation after the rituximab-CD20 complexes were removed by phagocytic cells. Western blots provide additional evidence for this escape mechanism that appears to occur as a consequence of CD20 loss. Treatment paradigms to prevent this escape, such as use of engineered or alternative anti-CD20 mAbs, may allow for more effective immunotherapy of chronic lymphocytic leukemia.

Clinical investigations have revealed that infusion of immunotherapeutic mAbs directed to normal or tumor cells can lead to loss of targeted epitopes, a phenomenon called antigenic modulation. Recently, we reported that rituximab treatment of chronic lymphocytic leukemia patients induced substantial loss of CD20 on B cells found in the circulation after rituximab infusion, when rituximab plasma concentrations were high. Such antigenic modulation can severely compromise therapeutic efficacy, and we postulated that B cells had been stripped (shaved) of the rituximab/CD20 complex by monocytes or macrophages in a reaction mediated by FcγR. We developed an in vitro model to replicate this in vivo shaving process, based on reacting rituximab-opsonized CD20+ cells with acceptor THP-1 monocytes. After 45 min at 37°C, rituximab and CD20 are removed from opsonized cells, and both are demonstrable on acceptor THP-1 cells. The reaction occurs equally well in the presence and absence of normal human serum, and monocytes isolated from peripheral blood also promote shaving of CD20 from rituximab-opsonized cells. Tests with inhibitors and use of F(ab′)2 of rituximab indicate transfer of rituximab/CD20 complexes to THP-1 cells is mediated by FcγR. Antigenic modulation described in previous reports may have been mediated by such shaving, and our findings may have profound implications for the use of mAbs in the immunotherapy of cancer.

The CD20 mAb ofatumumab (OFA) is more effective than rituximab (RTX) in promoting complement-dependent cytotoxicity (CDC) of B cells via the classical pathway (CP) of complement. CP activation is initiated by C1q binding to cell-bound IgG. Therefore, we examined the role of C1q in the dynamics of complement activation and CDC of B cell lines and primary cells from patients with chronic lymphocytic leukemia, reacted with OFA or RTX. C1q binding, complement activation, and colocalization of C1q with cell-bound mAbs were determined by flow cytometry and high-resolution digital imaging. C1q binds avidly to OFA-opsonized Raji and Daudi cells (KD = 12–16 nM) and colocalizes substantially with cell-bound OFA. Cells opsonized with OFA undergo high levels of complement activation and CDC in C1q-depleted serum supplemented with low concentrations of C1q. Under comparable conditions, RTX-opsonized cells bind less C1q; in addition, even when higher concentrations of C1q are used to achieve comparable C1q binding to RTX-opsonized cells, less complement activation and CDC are observed. Greater CDC induced by OFA may occur because C1q is bound in close proximity and with high avidity to OFA, resulting in effective CP activation. Moreover, OFA binds to the small, extracellular CD20 loop, placing the mAb considerably closer to the cell membrane than does RTX. This may facilitate effective capture and concentration of activated complement components closer to the cell membrane, potentially shielding them from inactivation by fluid phase agents and promoting efficient generation of the membrane attack complex.

The CD20 mAb ofatumumab (OFA) induces complement-mediated lysis of B cells. In an investigator-initiated phase II trial of OFA plus chemotherapy for chronic lymphocytic leukemia (CLL), OFA treatment promoted partial CLL B cell depletion that coincided with reduced complement titers. Remaining CLL B cells circulated with bound OFA and covalently bound complement breakdown product C3d, indicative of ongoing complement activation. Presumably, neither complement- nor effector cell-based mechanisms were sufficiently robust to clear these remaining B cells. Instead, almost all of the bound OFA and CD20 was removed from the cells, in accordance with previous clinical studies that demonstrated comparable loss of CD20 from B cells after treatment of CLL patients with rituximab. In vitro experiments with OFA and rituximab addressing these observations suggest that host effector mechanisms that support mAb-mediated lysis and tumor cell clearance are finite, and they can be saturated or exhausted at high B cell burdens, particularly at high mAb concentrations. Interestingly, only a fraction of available complement was required to kill cells with CD20 mAbs, and killing could be tuned by titrating the mAb concentration. Consequently, maximal B cell killing of an initial and secondary B cell challenge was achieved with intermediate mAb concentrations, whereas high concentrations promoted lower overall killing. Therefore, mAb therapies that rely substantially on effector mechanisms subject to exhaustion, including complement, may benefit from lower, more frequent dosing schemes optimized to sustain and maximize killing by cytotoxic immune effector systems.

Treatment of chronic lymphocytic leukemia (CLL) patients with standard dose infusion of rituximab (RTX), 375 mg/m2, induces clearance of malignant cells from peripheral blood after infusion of 30 mg of RTX. After completion of the full RTX infusion, substantial recrudescence of CLL cells occurs, and these cells have lost >90% of CD20. To gain insight into mechanism(s) of CD20 loss, we investigated the hypothesis that thrice-weekly low-dose RTX (20 or 60 mg/m2) treatment for CLL over 4 wk would preserve CD20 and enhance leukemic cell clearance. During initial infusions in all 12 patients, the first 30 mg of RTX promoted clearance of >75% leukemic cells. Four of six patients receiving 20 mg/m2 RTX retained ≥50% CD20, and additional RTX infusions promoted further cell clearance. However, four of six patients receiving 60 mg/m2 had CD20 levels <20% baseline 2 days after initial infusions, and additional RTX infusions were less effective, presumably due to epitope loss. Our results suggest that when a threshold RTX dose is exceeded, recrudesced RTX-opsonized cells are not cleared, due to saturation of the mononuclear phagocytic system, but instead are shaved of RTX-CD20 complexes by acceptor cells. Thrice-weekly low-dose RTX may promote enhanced clearance of circulating CLL cells by preserving CD20.

We previously reported that 1 h after infusion of CD20 mAb rituximab in patients with chronic lymphocytic leukemia (CLL), >80% of CD20 was removed from circulating B cells, and we replicated this finding, based on in vitro models. This reaction occurs via an endocytic process called shaving/trogocytosis, mediated by FcγR on acceptor cells including monocytes/macrophages, which remove and internalize rituximab–CD20 immune complexes from B cells. Beers et al. reported that CD20 mAb-induced antigenic modulation occurs as a result of internalization of B cell-bound mAb–CD20 complexes by the B cells themselves, with internalization of ∼40% observed after 2 h at 37°C. These findings raise fundamental questions regarding the relative importance of shaving versus internalization in promoting CD20 loss and have substantial implications for the design of mAb-based cancer therapies. Therefore, we performed direct comparisons, based on flow cytometry, to determine the relative rates and extent of shaving versus internalization. B cells, from cell lines, from patients with CLL, and from normal donors, were opsonized with CD20 mAbs rituximab or ofatumumab and incubated for varying times and then reacted with acceptor THP-1 monocytes to promote shaving. We find that shaving induces considerably greater loss of CD20 and bound mAb from opsonized B cells in much shorter time periods (75–90% in <45 min) than is observed for internalization. Both shaving/trogocytosis and internalization could contribute to CD20 loss when CLL patients receive rituximab therapy, but shaving should occur more rapidly and is most likely to be the key mechanism of CD20 loss.

Binding of the CD20 mAb rituximab (RTX) to B lymphocytes in normal human serum (NHS) activates complement (C) and promotes C3b deposition on or in close proximity to cell-bound RTX. Based on spinning disk confocal microscopy analyses, we report the first real-time visualization of C3b deposition and C-mediated killing of RTX-opsonized B cells. C activation by RTX-opsonized Daudi B cells induces rapid membrane blebbing and generation of long, thin structures protruding from cell surfaces, which we call streamers. Ofatumumab, a unique mAb that targets a distinct binding site (the small loop epitope) of the CD20 Ag, induces more rapid killing and streaming on Daudi cells than RTX. In contrast to RTX, ofatumumab promotes streamer formation and killing of ARH77 cells and primary B cells from patients with chronic lymphocytic leukemia. Generation of streamers requires C activation; no streaming occurs in media, NHS-EDTA, or in sera depleted of C5 or C9. Streamers can be visualized in bright field by phase imaging, and fluorescence-staining patterns indicate they contain membrane lipids and polymerized actin. Streaming also occurs if cells are reacted in medium with bee venom melittin, which penetrates cells and forms membrane pores in a manner similar to the membrane-attack complex of C. Structures similar to streamers are demonstrable when Ab-opsonized sheep erythrocytes (non-nucleated cells) are reacted with NHS. Taken together, our findings indicate that the membrane-attack complex is a key mediator of streaming. Streamer formation may, thus, represent a membrane structural change that can occur shortly before complement-induced cell death.

More than 20 years ago clinical investigations in the immunotherapy of cancer revealed that infusion of certain immunotherapeutic mAbs directed to tumor cells induced loss of targeted epitopes. This phenomenon, called antigenic modulation, can compromise mAb-based therapies. Recently we reported that rituximab (RTX) treatment of chronic lymphocytic leukemia patients induced substantial loss of targeted CD20 on B cells found in the circulation after RTX infusion; this “shaving” of RTX-CD20 complexes from B cells is also promoted in vitro by THP-1 monocytes and by PBMC in a reaction mediated by Fcγ receptors. The mechanism responsible for shaving appears to be trogocytosis, a process in which receptors on effector cells remove and internalize cognate ligands and cell membrane fragments from target cells. We now report that three therapeutic mAbs approved by the U.S. Food and Drug Administration for the treatment of cancer, RTX, cetuximab, and trastuzumab, as well as mAb T101, which has been shown to induce antigenic modulation in the clinic, promote trogocytosis in vitro upon binding to their respective target cells. Trogocytosis of the mAb-opsonized cells is mediated by THP-1 monocytes and by primary monocytes isolated from PBMC. In view of these results, it is likely that these mAbs and possibly other anticancer mAbs now used in the clinic may promote trogocytic removal of the therapeutic mAbs and their cognate Ags from tumor cells in vivo. Our findings may have important implications with respect to the use of mAbs in cancer immunotherapy.