Rituximab Kills Freshly Isolated B-NHL and B-CLL More Efficiently by Complement Than by ADCC In Vitro: Role of CD20 Expression Levels and of the CD55 and CD59 Complement Inhibitors

Current evidence suggests that rituximab (Rituxan) works in vivo mainlythrough complement-dependent cytotoxicity (CDC) and/or antibody-dependent cellular cytotoxicity (ADCC). Here we have investigated thesensitivity of freshly isolated cells obtained from 33 B-cell chroniclymphocytic leukemia (B-CLL), 5 prolymphocytic leukemia (PLL), and 6 mantle cellleukemia (MCL) patients to be lysed by rituximab and complement in vitro.

The results showed that in B-CLL and PLL, the levels of CD20, measured bystandard immunofluorescence or using calibrated beads, correlated linearly withthe lytic response (coefficient > 0.9, P < .0001). Furthermore, thecorrelation remained highly significant when the six MCL patients were includedin the analysis (coefficient 0.91, P < .0001), suggesting that CD20 levelsprimarily determine lysis regardless of diagnostic group.

The role of the complement inhibitors CD55 and CD59 was also investigated.All B-CLL and PLL cells expressed these molecules, but at variable levels (meanfluorescence intensity [MFI] = 20-1,200 vs 20-250, respectively). AlthoughCD55 and CD59 levels did not permit us to predict complement susceptibility, thefunctional block of these inhibitors demonstrated that they play an importantrole in regulating CDC. Thus, lysis of poorly responding B-CLL samples wasincreased five- to sixfold after blocking both CD55 and CD59, whereas that ofhigh responders was essentially complete in the presence of a single blockingantibody.

We have also investigated the lysis through ADCC of five fresh cases of CLLand two of MCL. Of these samples, four expressed relatively high levels of CD20(MFI > 200) and were lysed significantly with rituximab and complement (40%to 65%). On the contrary, we have found that none of the samples examined waslysed significantly by the natural killer (NK) cell line NKL in the presence ofrituximab, whereas the same cells were killed efficiently (by 30% to 70%) in thepresence of the anti-CD52 antibody alemtuzumab (Campath). Furthermore, threeleukemic B-cell lines could be lysed to a similar extent (48% to 62%) by eitherrituximab or alemtuzumab and NKL cells.

CONCLUSION: Altogether, these data suggest that, although rituximab is ableto activate both CDC and ADCC on B-cell lymphoma lines, CDC is a more efficientmechanism of action of this antibody than ADCC on freshly isolated B-CLL andB-cell non-Hodgkin’s lymphoma cases. Furthermore, they show that CD20, CD55,and CD59 are important factors in determining the response to rituximab andcomplement in vitro, and indicate strategies to improve the clinical response tothis biologic therapy.