Current evidence suggests that rituximab (Rituxan) works in vivo mainly
through complement-dependent cytotoxicity (CDC) and/or antibody-dependent cellular cytotoxicity (ADCC). Here we have investigated the
sensitivity of freshly isolated cells obtained from 33 B-cell chronic
lymphocytic leukemia (B-CLL), 5 prolymphocytic leukemia (PLL), and 6 mantle cell
leukemia (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 by
standard immunofluorescence or using calibrated beads, correlated linearly with
the lytic response (coefficient > 0.9, P < .0001). Furthermore, the
correlation remained highly significant when the six MCL patients were included
in the analysis (coefficient 0.91, P < .0001), suggesting that CD20 levels
primarily 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 (mean
fluorescence intensity [MFI] = 20-1,200 vs 20-250, respectively). Although
CD55 and CD59 levels did not permit us to predict complement susceptibility, the
functional block of these inhibitors demonstrated that they play an important
role in regulating CDC. Thus, lysis of poorly responding B-CLL samples was
increased five- to sixfold after blocking both CD55 and CD59, whereas that of
high responders was essentially complete in the presence of a single blocking
We have also investigated the lysis through ADCC of five fresh cases of CLL
and 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 was
lysed significantly by the natural killer (NK) cell line NKL in the presence of
rituximab, whereas the same cells were killed efficiently (by 30% to 70%) in the
presence of the anti-CD52 antibody alemtuzumab (Campath). Furthermore, three
leukemic B-cell lines could be lysed to a similar extent (48% to 62%) by either
rituximab or alemtuzumab and NKL cells.
CONCLUSION: Altogether, these data suggest that, although rituximab is able
to activate both CDC and ADCC on B-cell lymphoma lines, CDC is a more efficient
mechanism of action of this antibody than ADCC on freshly isolated B-CLL and
B-cell non-Hodgkin’s lymphoma cases. Furthermore, they show that CD20, CD55,
and CD59 are important factors in determining the response to rituximab and
complement in vitro, and indicate strategies to improve the clinical response to
this biologic therapy.