Predominant Complement Mediated Lysis of B-CLL Cells by Therapeutic MAbs Rituximab and Campath-1H in Whole Blood Assays.

Abstract 3445

Poster Board III-333

The mechanism of action of unconjugated MAbs such as Rituximab and Campath-1H in vivo is still a matter of debate. Most in vitro assays with antibodies rely upon purified effector cells or proteins taken outside their natural context, and on target cell lines rather than patients cells. In order to analyse the activity of therapeutic MAbs on circulating leukemic cells in more physiological conditions and in a system the least manipulated as possible, we have set up a whole blood assays using Rituximab and Campath-1H. Peripheral blood samples were drawn from B-CLL patients or normal donors in sodium citrate and antibodies were directly added at different concentrations. We first demonstrated that neither apoptosis, induced by cross-linked anti-CD20 antibody, nor complement mediated cytotoxicity (CDC) induced by Campath-1H or Rituximab were significantly inhibited by citrate used at the standard concentration (0.1 M). We then performed a number of experiments using whole blood samples in citrate, into which increasing concentrations of Rituximab or Campath-1H were added. Lysis was analysed by FACS analysis after different incubation times at 37°C. We observed that Campath-1H very rapidly and efficiently lysed normal B cells or B-CLL targets in vitro in whole blood: maximal lysis was reached within 4 hours and was observed already with 1 and 10 μg/ml antibody (61 %), even though it was still more effective at 25 or 50 μg/ml (up to 90 % lysis). 25 μg/ml is known to be reached in the circulation after 30mg infusions of the antibody 3 times a week. Lysis by Campath-1H was fully complement dependent since it was inhibited by 90% in presence of excess blocking anti-C5 antibody Eculizumab (200 μg/ml). Eculizumab alone in contrast had no effect on cell viability. We then analysed the efficacy of increasing concentrations of Rituximab in the same assay conditions. We observed in general a much reduced lysis with Rituximab compared to Campath-1H, even using antibody up to 200 μg/ml, a concentration that is reached in the circulation after standard 375 mg/m² administration of the antibody once a week. Lysis showed also slower kinetics, with limited lysis at 4 hours (mean 6.4%) and maximal lysis with Rituximab reached only after 24 hours incubation (mean 18.8%). Also in this case, target cell death was inhibited by at least 90% in presence of Eculizumab, suggesting a major role of complement. Lysis by Rituximab correlated directly with CD20 expression levels (R=0.8) in 13 B-CLL samples analysed, as expected for a mechanism complement dependent. Indeed a mean 29.3% and 73.2% killing could be observed in the two CD20 bright B-CLL, at 4 and 24 hours respectively, whereas a mean of 3.1% and 10.9% lysis was observed in the 11 low-intermediate CD20 samples analysed at the same time points. These data in whole blood confirm our previously published results on the role of CD20 expression levels in CDC of isolated B-CLL cells (Golay et al., Blood 98, 3383-3389, 2001).

In contrast to CDC and apoptosis, ADCC was strongly inhibited by citrate as well as several anti-coagulants tested and therefore could not be analysed in this type of assay. Nonetheless in B-CLL samples, NK cells were below detection limit (<0.1%) in most cases analysed, suggesting that ADCC in the circulation is not a major mechanism of lysis in this disease subtype.

Finally we determined the effect of citrate on phagocytosis mediated by Rituximab and in vitro differentiated human macrophages. Phagocytosis could be observed in presence of 0.1M citrate (31%, compared to 44% in absence of citrate). Phagocytosis of B-CLL in whole blood was therefore analysed by layering samples directly onto the macrophages. We observed that phagocytosis of B-CLL targets in whole blood was very low (less than 1% over background) compared to a mean of 47% for purified B-CLL targets phagocytosed in normal culture medium. Phagocytosis in whole blood was low presumably due to the presence of high concentration of human IgG in whole blood since as low as 50 μg/ml human IgG is known to inhibit phagocytosis by 90%.

We conclude that the major activity of Campath-1H and Rituximab in the circulation is through complement. Apoptosis, ADCC and phagocytosis appear to play a marginal role in this context but may become more important in tissues. The method presented could be used to rapidly screen novel antibodies for their efficacy through either as apoptosis or CDC directly on unmanipulated patients material.