Page 44 - Haematologica - Vol. 105 n. 6 - June 2020

G. Pavlasova and M. Mraz
et al. observed in the Ramos cell line that TGFβ signaling led to SMAD2/3 binding directly to the MS4A1 transcrip- tion start site, resulting in CD20 repression.99 However, we and others have shown that in CLL, the chemokine CXCL12 (also known as SDF1) produced by stromal cells in immune niches induces CD20 expression, and that the intraclonal CLL cell subpopulation that recently exited the lymph nodes is characterized by high levels of CD20.34 This has an important consequence for the mech- anism of rituximab’s action since, in vivo, rituximab infu- sion leads to rapid and preferential elimination of this aggressive, proliferative CLL cell subpopulation. The remaining large proportion of CLL cells can survive the rituximab therapy because of relatively weak cell-surface levels of CD20, but these cells have a gene-expression profile of non-activated CLL cells, which are relatively less able to activate the BCR pathway, and do not prolif- erate. It remains unclear which molecular pathways pro- vide CLL cells in the lymph node microenvironment with resistance to rituximab, despite having high levels of CD20.97 The resistance to rituximab in the microenviron- ment seems to be limited to rituximab-mediated apopto- sis and CDC.
Combinatorial therapy of novel drugs and anti-CD20 monoclonal antibodies
For over a decade, scientists and clinicians have become accustomed to the empirical experience that adding ritux- imab to other therapies leads to increased therapeutic efficacy in B-cell malignancies. This also prompted stud- ies for strategies to induce higher CD20 levels on the B- cell surface to potentially sensitize malignant cells to anti- CD20 monoclonal antibodies (summarized in Table 2). Several of these “CD20 inducers” are being explored in preclinical or phase I/II clinical trials, including aurora kinase inhibitors, FOXO1 inhibitors, and chromatin mod- ulators. For example, it has been shown that in lym- phomas, the aurora kinase inhibitor alisertib can be safely and successfully used in combination with vincristine and rituximab (phase I/II trial).72 Nevertheless, none of the “CD20 inducers” has been prioritized for phase III trials yet (see Table 2). Combining the BTK inhibitor ibrutinib with rituximab was expected to increase the BTK inhibitor’s clinical efficacy. This hypothesis was support- ed by the observation that CLL/lymphoma cells become more sensitive to apoptosis and anti-CD20 monoclonal antibodies when mobilized from immune niches (a typi- cal effect of BCR inhibitors).96,97 However, results from a phase II study12 demonstrated no benefit from adding rit- uximab to ibrutinib. Recent studies have shown that CD20 levels are repressed during ibrutinib therapy and that ibrutinib affects cells responsible for effector mecha- nisms such as T/NK cells and macrophages. The reduc- tion of CD20 levels by ibrutinib has a clear impact on rit- uximab-mediated CDC and apoptosis. However, CD20 is not completely lost, which still allows for anti-CD20 monoclonal antibodies to bind to cells. Skarzynski et al.100 also suggested that ibrutinib reduces complement inhibitor CD55 levels, which might partially counterbal- ance the effects of lower CD20, but it seems that in the sum of all effects, rituximab or ofatumumab efficacy decreases during ibrutinib therapy.
CD20 levels appear to play an essential role in CDC induced by rituximab, but they seem to be less relevant for ADCC.101 Unfortunately, ibrutinib also affects the
functions of T cells and NK cells by inhibiting their BTK or a related ITK. BTK is critical for regulating the func- tions of NK cells as BTK-less NK cells have impaired cyto- toxic activity.102 ITK signals downstream of the T-cell receptor and is required to activate NK cells through FcγRIII.103 Inhibiting BTK or ITK impairs the cytotoxic functions of NK cells (degranulation, cytokine secretion) and ADCC mediated by type I and II anti-CD20 mono- clonal antibodies.104,105 Thus ibrutinib may impair T- and NK-cell functions through either BTK or ITK, or both.104 It has also been suggested that phagocytosis by macrophages is affected by ibrutinib, but it is unclear if this is due to BTK inhibition in these cells or an off-target effect.104-106 Based on the results from the iLLUMINATE study,49 the Food and Drug Administration has already approved the combination of ibrutinib plus obinutuzum- ab for treatment-naïve patients with CLL (Table 2). However, the control arm of the study with chlorambucil plus obinutuzumab did not allow a conclusion on whether obinutuzumab provided a real benefit. Alternatively, sequential ibrutinib administration after the anti-CD20 monoclonal antibody could conceivably allow for better antibody effects. However, a clinical trial in CLL showed that sequential treatment with ofatumumab before ibrutinib was inferior to starting ibrutinib first, fol- lowed by the administration of ibrutinib and ofatumum- ab.107 Since obinutuzumab acts in part through mecha- nisms different from those of type I antibodies (ritux- imab, ofatumumab), its combination with ibrutinib may lead to different effects and increased clinical efficacy. Alternatively, other BTK inhibitors, such as acalabrutinib, are more selective with less off-target activity and likely do not interfere with antibody-dependent cellular phago- cytosis or ADCC.108 It needs to be determined whether acalabrutinib is more suitable than ibrutinib for therapeu- tic combination with anti-CD20 monoclonal antibodies. Recently, a phase III study showed that acalabrutinib combined with obinutuzumab is highly efficient in pro- longing progression-free survival when compared to obinutuzumab and chlorambucil in patients with previ- ously untreated CLL.109
An interesting case is also the PI3Kδ inhibitor idelalisib, which is currently approved for use in combination with rituximab to treat relapsed CLL based on a comparison to rituximab alone (Table 2).110 A similar phase III study showed that the progression-free survival of participants treated with a combination of idelalisib and ofatumumab was significantly longer than that of the group treated with ofatumumab only (Table 2).111 However, it is unclear whether adding rituximab or ofatumumab to idelalisib actually provides any clinical benefit, and based on the understanding of CD20 regulation, it is very likely that PI3K inhibitors also repress CD20 expression. In fact, data indicate that any inhibitor repressing Akt or NFκB or CXCR4 activity in B cells, such as a SYK inhibitor, BTK inhibitor, PI3Kδ inhibitor, or CXCR4 antagonist, also reduces CD20 expression, leading to decreased binding of anti-CD20 monoclonal antibodies.44,70,100,112,113 Notably, SRC inhibitors such as dasatinib also repress CD20 tran- scription and impair NK cell functions.112 Additionally, PI3Kδ plays a critical role in maturation, development, and effector functions of NK cells, which would indicate that it might impair ADCC. However, some studies indi- cate that this might be less prominent than ibrutinib’s effects114 or that idelalisib does not reduce ADCC at all.115
1502
haematologica | 2020; 105(6)