Interactions between Ibrutinib and Anti-CD20 Antibodies: Competing Effects on the Outcome of Combination Therapy

Purpose: Clinical trials of ibrutinib combined with anti-CD20 monoclonal antibodies (mAb) for chronic lymphocytic leukemia (CLL) report encouraging results. Paradoxically, in preclinical studies, in vitro ibrutinib was reported to decrease CD20 expres-sionandinhibitcellulareffectormechanisms.Wethereforesetout to investigate effects of in vivo ibrutinib treatment that could explain this paradox. Experimental Design: Patients received single-agent ibrutinib (420 mg daily) on an investigator-initiated phase II trial. Serial blood samples were collected pretreatment and during treatment for ex vivo functional assays to examine the effects on CLL cell susceptibility to anti-CD20 mAbs. Results: We demonstrate that CD20 expression on ibrutinib was rapidly and persistently downregulated (median reduction 74%, day 28, P < 0.001) compared with baseline. Concomitantly, CD20 mRNA was decreased concurrent with reduced NF- k B signaling. An NF- k B binding site in the promoter of MS4A1 (encoding CD20) and downregulation of CD20 by NF- k B inhibitors support a direct transcriptional effect. Ex vivo , tumor cells from patients on ibrutinib were less susceptible to anti-CD20 mAb-mediated complement-dependent cytotoxicity than pretreatment cells (median reduction 75%, P < 0.001); however, opsonization by the complement protein C3d, which targets cells for phagocytosis, was relatively maintained. Expression of decay-accelerating factor (CD55) decreased on ibrutinib, providing a likely mechanism for the preserved C3d opsonization. In addition, ibrutinib signi ﬁ cantly inhibited trogocytosis, a major contributor to antigen loss and tumor escape during mAb therapy. Conclusions: Our data indicate that ibrutinib promotes both positive and negative interactions with anti-CD20 mAbs, suggest-ingthatsuccessfullyharnessingmaximalantitumoreffectsofsuchcombinationsrequiresfurtherinvestigation. Clin Res; for changes in CD20 expression and susceptibility to mAb-dependent killing. Ibrutinib substantially reduced CD20 expression on CLL cells in vivo and consequent-ly diminished complement-mediated cell killing in vitro . Con-versely, ibrutinib downregulated CD55, an inhibitor of complement activation, on CLL cells and reduced trogocytosis of anti-CD20 by macrophages, a major mechanism of antigen loss during mAb therapy. Ibrutinib therefore mediates both positive and negative effects on anti-CD20 mAb activity; harnessing these interactions could maximize the clinical ﬁ


Introduction
Ibrutinib is a small-molecule Bruton tyrosine kinase (BTK) inhibitor that has demonstrated exceptional safety and efficacy as a monotherapy for various B-cell malignancies (1), most notably chronic lymphocytic leukemia (CLL; ref. 2), mantle cell lymphoma (MCL; ref. 3), Waldenstr€ om's macroglobulinemia (4), and activated B-cell-like diffuse large B-cell lymphoma (ABC-DLBCL; ref. 5). While durable responses to single-agent ibrutinib are common in CLL (2) and MCL (3), most patients do not experience a complete response. In an effort to improve upon the already impressive activity of ibrutinib monotherapy, many clinical trials are currently exploring the value of combining ibrutinib with monoclonal antibodies (mAb) targeting CD20 (6)(7)(8)(9)(10). Anti-CD20 mAbs are foremost among promising therapeutic partners for ibrutinib because they are widely applicable, well-tolerated targeted agents that add clinical benefit to frontline chemotherapy (11). Furthermore, ibrutinib may enhance the effect of CD20targeting mAbs because it releases CLL cells into the blood from tissue sites (lymph nodes and bone marrow; ref. 12) and thus potentially facilitates the clearing of CLL cells through mAb effector mechanisms (13).
In addition to mediating CDC, mAbs also recruit NK cells and phagocytes to eliminate target cells (18)(19)(20). Interestingly, anti-CD20 mAbs have been shown to mediate potent antibodydependent cellular cytotoxicity (ADCC) even at low CD20 expression levels (16). Although ADCC may help to eliminate B-cell cancers with decreased CD20 expression, Kohrt and colleagues showed that ibrutinib decreases ADCC by directly inhibiting Fc receptor-dependent NK cell activation and cytotoxicity in vitro. Furthermore, ibrutinib diminished the in vivo antitumor efficacy of the HER2-specific mAb trastuzumab and the CD20-specific mAb rituximab when administered concurrently in a murine xenograft model of breast cancer and lymphoma, respectively (21). Similar to its effect on NK cells, ibrutinib has also been shown to decrease antibody-mediated phagocytosis of target cells by macrophages and neutrophils (22,23). Taken together, these findings strongly suggest that ibrutinib antagonizes anti-CD20 mAb activity, but are difficult to reconcile with growing clinical evidence that combination therapy can be more effective than either agent alone (6)(7)(8)(9)(10).
Here we investigated the effects of in vivo single-agent ibrutinib therapy on CLL cells of patients on an investigator-initiated phase II study (NCT01500733). Specifically, we sought to determine how ibrutinib-mediated CD20 downregulation compares with other mechanisms of antigen loss and what role the complement system might play in the context of ibrutinib and anti-CD20 mAb combination therapy. Our findings suggest complex and diverging interactions between ibrutinib and CD20-specific mAbs and may inform how these two agents may best be combined in ongoing and future clinical trials.

Patients
The analyses presented here are based on samples from 48 patients with CLL enrolled in our ongoing, investigator-initiated phase II study of ibrutinib (NCT01500733). The study was approved by the local ethics committee; informed consent was obtained from all patients in accordance with the Declaration of Helsinki. Briefly, both treatment na€ ve and relapsed/refractory patients with either del(17p) or TP53 mutations or patients over the age of 65 years were eligible and treated with ibrutinib 420 mg orally once daily until disease progression or the occurrence of intolerable side effects. Mutation status of the immunoglobulin heavy chain variable (IGHV) gene was assessed as described and CLL patients with <2% mutations were classified as unmutated (12,24). Patient characteristics are summarized in Supplementary  Table S1.

Bone marrow immunohistochemistry
Bone marrow trephine biopsies were fixed in B-Plus fixative, embedded in paraffin, and processed for morphologic evaluation using standard procedures. Immunohistochemical studies with anti-CD20 and anti-CD79a antibodies (Ventana Medical System) were performed using immunoperoxidase staining procedures on an automated immunostainer (Ventana Medical System) according to the manufacturer's instructions. Results were assessed semiquantitatively by two trained hematopathologists on a scale from 0% to 100%. Images were obtained via digital microscopy using an Olympus BX-51 microscope (Olympus America) equipped with a DPlan 40/0Á65 numeric aperture objective and captured using an Olympus DP70 digital camera system.
Evaluation of CD20 antigen expression CD20 mean fluorescence intensity (MFI) values were obtained before and on ibrutinib using peripheral blood mononuclear cells (PBMC) prepared by density-gradient centrifugation (Lymphocyte Separation Media; ICN Biomedicals) and viably frozen in 90% FBS, 10% dimethyl sulfoxide (Sigma) in liquid nitrogen. Enumeration of anti-CD20-binding sites, antibody-binding capacity (ABC), was performed on fresh PBMCs, and fresh bone marrow cell suspensions in the clinical laboratory using the Quantibrite system (BD Biosciences) as previously described (25). Cells were analyzed on a FACSCanto II flow cytometer using FACS-DIVA 6.1.1, FCSExpress 4 (BD Biosciences) and FlowJo 10 software (TreeStar). In select experiments, the role of NF-kB on CD20 expression was measured using the NFkB inhibitor Bay 11-7028 (Santa Cruz Biotechnology).

Gene expression
Total RNA was extracted from CD19þ PBMCs or lymph node core biopsies using RNeasy kits (Qiagen) and cDNA was prepared using the High Capacity cDNA RT Kit (Applied Biosystems). Expression of MS4A1 (CD20) mRNA was quantified by RT-PCR using inventoried TaqMan primers on an ABI PRISM 7900HT Sequence Detection System (Applied Biosystems). For lymph node core biopsies, CD20 expression was normalized to CD19 mRNA expression. Expression of six previously validated NF-kB target genes were quantified by RT-PCR on TaqMan Custom Arrays (microfluidic cards; Applied Biosystems) and averaged into a NF-kB signature score as previously described (26).

CDC, complement deposition, and complement regulatory protein expression
For analysis of CDC, cells were incubated with or without 10 mg/mL ofatumumab (commercial supply from GlaxoSmithKline) for 2 hours in the presence of 33% normal human serum (Innovative Research) as a source of complement. The proportion of

Translational Relevance
Ongoing clinical trials combine anti-CD20 monoclonal antibodies (mAb) and the Bruton tyrosine kinase (BTK) inhibitor ibrutinib with promising preliminary results. However, preclinical studies have identified a number of negative interactions between ibrutinib and anti-CD20 mAbs that may reduce the efficacy of combination therapy. Given discordant preclinical and clinical observations, we analyzed tumor cells of patients with chronic lymphocytic leukemia (CLL) treated with ibrutinib for changes in CD20 expression and susceptibility to mAb-dependent killing. Ibrutinib substantially reduced CD20 expression on CLL cells in vivo and consequently diminished complement-mediated cell killing in vitro. Conversely, ibrutinib downregulated CD55, an inhibitor of complement activation, on CLL cells and reduced trogocytosis of anti-CD20 by macrophages, a major mechanism of antigen loss during mAb therapy. Ibrutinib therefore mediates both positive and negative effects on anti-CD20 mAb activity; harnessing these interactions could maximize the clinical benefit of ibrutinib and anti-CD20 mAb combination therapy. lysed CD19 þ cells was assessed by TO-PRO3 (Life Technologies) staining. Complement deposition was measured using the FITC labeled 1H8 mAb (Cedarlane) that binds all forms of C3 (27). Cell surface expression levels were measured on CD19 þ cells using PE-labeled mAbs against CD55, CD46, and CD59 (BD Biosciences). Samples were analyzed using FACSCanto II and Fortessa LSR flow cytometers (BD Biosciences) and FlowJo 10 software (TreeStar). For all ex vivo experiments, CLL cells were cultured in AIM V media (Life Technologies).
FACS and in vitro analysis of CD20 subpopulations CLL PBMC samples were stained with CD19-FITC and CD20-PE mAbs (BD Biosciences), triple washed, and resuspended in PBS containing 1% FBS and penicillin-streptomycin (Life Technologies) before being sorted on a FACS Aria II cell sorter (BD Biosciences). CD19 þ cells that fell in the top or bottom <15% of the CD20 þ cell distribution were collected, acid washed as previously described (28) to remove staining mAbs, and then cultured in AIM-V media (Life Technologies). Cells were kept at 4 C at all times before being put into culture at 37 C. Cultured cells were treated with vehicle or 1 mmol/L ibrutinib for 48 hours as described previously (14) and finally assessed for CD20 expression changes and cell lysis resulting from ibrutinib treatment.

Trogocytosis assays
CLL cells from patients were stained with Vybrant CFDA SE Cell Tracer (Life Technologies), bound with Alexa-Fluor 594 (Life Technologies) labeled ofatumumab, coincubated with THP1 cells (ATCC), and then assessed for Cell Tracer and Alexa-Fluor 594 staining as previously described (29,30). THP1 cells were incubated in RPMI1640 media (Life Technologies) containing various concentrations of ibrutinib (Selleckchem) or vehicle for 1 hour before coincubation with CLL cells.

Statistical analyses
To determine whether data were normally distributed, the D 0 Agostino and Pearson (omnibus K 2 ) normality test was used. To compare measurements from patients either across time or treatment, two-tailed Student paired t tests were applied to normally distributed (Gaussian) datasets, otherwise matchedpairs Wilcoxon signed rank tests were used. Correlations were evaluated using Pearson test (Prism 5, GraphPad).

Ibrutinib decreases CD20 expression on CLL cells in vivo
First, we confirmed the in vitro observations of CD20 downregulation on CLL cells by ibrutinib reported by Bojarczuk and colleagues, in clinical samples of patients treated with single-agent ibrutinib in our phase II clinical trial (NCT01500733). We measured CD20 expression on CLL cells in the peripheral blood at baseline and during treatment by flow cytometry. On day 2, after only one dose of ibrutinib, CD20 MFI on CLL cells was already decreased in half of the patients tested, albeit with wide interpatient variability (P ¼ 0.02; Fig. 1A). However, after one (day 28) and two cycles (day 56) of ibrutinib, CD20 MFI on CLL cells was decreased in all patients evaluated (n ¼ 22, P<0.0001; Fig. 1B). At day 28, the median decrease compared with baseline was 74% [interquartile range (IQR): 55-82] and at day 56 it was 78% (IQR: 56-83). Supplementary Figure S1 shows the decrease in CD20 expression at day 28 of a representative CLL patient. Reductions in CD20 by ibrutinib were independent of TP53 aberrations (24) and IGVH mutational status (Supplementary Fig. S2A-S2D). In 27 patients, anti-CD20 ABC was quantified using fresh PBMC samples obtained during treatment and after 6 months on ibrutinib using the BD Quantibrite system. In 25 of 27 patients (92.6%), CD20 expression was reduced on treatment compared with baseline. Notably, the two patients that did not show decreased CD20 on ibrutinib had unusually high baseline CD20 expression. The median pretreatment CD20 ABC was 7,572 sites per cell but varied widely from <500 to 67,000 sites per cell (Fig.  1C). The median decrease at 6 months was 64% (IQR, 43-69) compared with the matching pretreatment samples (P ¼ 0.001; Fig. 1C).
Bone marrow biopsies were obtained pretreatment, and at 2 months and 6 months on treatment. To assess the degree of tumor infiltration, biopsies were routinely stained with anti-CD20 and anti-CD79a antibodies by immunohistochemistry. Upon review, we noted that in addition to a reduction in tumor infiltration, evidenced by fewer cells staining for CD79a, the intensity of anti-CD20 staining decreased most notably at 2 months on treatment   Fig. 2B summarizes data for all patients evaluated). To confirm this decreased expression of CD20 on bone marrow-resident CLL cells, we obtained quantitative measurements of anti-CD20 ABC on CLL cells in fresh bone marrow aspirates. The anti-CD20 ABC on CLL cells in the bone marrow of thirteen patients was significantly decreased after two months on ibrutinib (n ¼ 13, P ¼ 0.003; Fig. 2C, left) and remained suppressed after 6 months of treatment (n ¼ 13, P ¼ 0.001; Fig. 2C, right). The median decrease in CD20 expression at 2 months was 69% (IQR, 60-85) and at 6 months it was 60% (IQR, 38-65). Thus, a decrease in the expression of CD20 is equally present in peripheral blood and bone marrow-resident CLL cells throughout at least the first 6 months of therapy.

Inhibition of CD20 mRNA expression by ibrutinib correlates with reduced NF-kB activity
To assess whether decreased CD20 cell surface expression was correlated with a reduction in CD20 mRNA (encoded by the MS4A1 gene), we performed quantitative RT-PCR on RNA from purified CLL cells. Already within 24 hours of initiating ibrutinib (day 2), a trend toward decreased CD20 mRNA expression was seen in most patients, although it did not reach statistical significance (Fig. 3A). After one cycle (day 28), CD20 mRNA expression in the CLL cells was decreased in all patients assessed, with a median reduction of 67% (n ¼ 12, P<0.001; Fig. 3A). Reduction in CD20 mRNA and CD20 cell surface expression correlated well (n ¼ 8, R ¼ 0.72, P ¼ 0.04; Fig. 3B). In vivo ibrutinib treatment reduces CD20 expression on bone marrowresident CLL cells. A, bone marrow immunohistochemistry staining for CD20 and CD79a in specimens taken from the same patient before ibrutinib and at 2 months (day 56) and 6 months on treatment. CD79a staining is shown for comparison of CLL infiltration in matched areas of the bone marrow. B, semiquantitative assessment of CD79a and CD20 staining intensity in bone marrow specimens during treatment (n ¼ 11). C, CD20 antibody binding capacity (ABC) values at 2 months (left) and 6 months (right) in bone marrow of patients treated with ibrutinib (n ¼ 13). ABC measurements were made on fresh samples using Quantibrite beads. P values were calculated using the matched-pairs Wilcoxon signed-rank test. The whiskers on all box plots depict the range (minimum and maximum values).
We previously demonstrated that ibrutinib decreases the expression of NF-kB target genes in CLL cells (26). To investigate the possibility that NF-kB acts as a mechanistic link between BTK inhibition and CD20 downregulation, we scanned the MS4A1 gene (NCBI Entrez Gene ID: 931) for the canonical NF-kB binding consensus sequence (31,32) and found one (GGGACTTTCC, 6594-6603) downstream of the transcriptional start site (6567) and another (GGGGTCCCCC, 6957-6966) between exons 1 and 2. We next compared the effect of ibrutinib on the averaged expression of six NF-kB target genes, referred to as NF-kB signature score (26), with its effect on CD20 expression by quantitative RT-PCR. There was a strong positive correlation (R ¼ 0.65, P<0.01) between the treatment-induced downregulation of CD20 mRNA and the NF-kB signature score (n ¼ 22; Fig. 3C).
To further examine the role of NF-kB in regulating CD20 expression, we tested the in vitro effect of the NF-kB inhibitor Bay11-7028 on primary CLL cells collected from patients prior to the start of ibrutinib. CD20 surface staining was decreased in a dose-dependent manner by in vitro treatment with the NF-kB inhibitor (n ¼ 8, P<0.001; Supplementary Fig. S3). Together, these data suggest that inhibition of NF-kB signaling by ibrutinib decreases the expression of CD20 through a transcriptional mechanism.

Ibrutinib decreases ofatumumab-mediated CDC
To characterize the functional consequences of reduced CD20 expression on CLL cells during treatment with ibrutinib, we measured ofatumumab induced CDC ex vivo. Compared with the matched pretreatment samples, CDC was already significantly reduced in CLL cells obtained on day 2 of ibrutinib and decreased further with continued therapy (n ¼ 12, P 0.02; Fig. 4A). After 2 and 6 months of therapy, the median reduction in ofatumumabinduced CDC was 75% (IQR 38-86) and 84% (IQR 50-91), respectively. Across all patients and time points, there was a strong correlation between the decrease in CD20 expression and the ability of ofatumumab to induce CDC (R ¼ 0.88; P<0.001; Fig. 4B).
To assess the possibility that the observed loss of CD20 was due to selective elimination of CLL cells with the highest CD20 expression, we sorted PBMCs obtained from CLL patients into CD20-bright and -dim fractions and measured their CD20 expression and viability after 48 hours of in vitro treatment with ibrutinib or vehicle. Subpopulations from the same CLL patient responded similarly to treatment in terms of the change in percent viability and CD20 expression (n ¼ 4, data not shown). These results argue against a major role for preferential killing of CD20-bright CLL cells in the observed downregulation of CD20 during ibrutinib therapy.
To investigate whether ibrutinib-mediated downregulation of CD20 and the resultant loss of CDC are permanent, we cultured CLL cells from patients on ibrutinib that demonstrated a treatment-induced decrease in CD20 expression for up to 96 hours in vitro and reassessed CD20 expression and CDC at 24-hour intervals. The recovery in CD20 expression was rapid, already apparent after 24 hours of culture, and reached 60% to 70% of pretreatment levels by 96 hours (Fig. 4C). In parallel, CDC also recovered reaching 70% to 80% of pretreatment values by 96 hours (Fig.  4C). These results indicate that the effects of ibrutinib on CDC and CD20 expression are rapidly reversible. Inhibition of CD20 mRNA expression by ibrutinib correlates with reduced NF-kB activity. A, expression of CD20 mRNA was quantified by RT-PCR in CD19 þ PBMCs on day 2 (n ¼ 8) and after 1 month (day 28, n ¼ 12) on ibrutinib. The percent change on treatment compared with baseline is shown. B, correlation between CD20 mRNA (MS4A1 gene) expression and CD20 cell surface expression on CLL cells on day 28, n ¼ 8. C, correlation between NF-kB signature score and CD20 gene expression. LN core biopsies (day 2, n ¼ 11) are shown as triangles, whereas day 28 peripheral blood samples are represented as circles (n ¼ 12

Downregulation of CD55 by ibrutinib rescues ofatumumabmediated complement deposition
While ofatumumab-mediated CDC was clearly compromised concurrent with CD20 downregulation, the biologic effects of the complement system are not limited to direct cytotoxicity. In particular, the complement component C3 is deposited, and remains covalently attached, to cells targeted by a complementactivating antibody such as ofatumumab; priming them for phagocytosis. We therefore measured the various forms of deposited C3 (denoted simply as C3d) on CLL cells after ofatumumab treatment ex vivo. The C3d opsonization of CLL cells correlated well to the degree of CDC achieved, both pretreatment (n ¼ 16, R ¼ 0.81, P<0.001; Fig. 5A, left) and after 2 months on ibrutinib (n ¼ 16, R ¼ 0.73, P<0.001; Fig. 5A, right). In parallel to CDC, C3d opsonization was also diminished on ibrutinib; however, the reduction in CDC in ibrutinib-treated samples was significantly more pronounced (median reduction 67%; IQR 53-79) than the decrease in C3d (median reduction 30%; IQR 11-54;n ¼ 16, P<0.001; Fig. 5B). Thus, complement deposition was relatively preserved despite decreased CD20 expression, prompting us to investigate the expression of regulators of complement activation on CLL cells.
We compared expression of complement regulatory proteins (CRP) on CLL cells at 2 months compared with baseline. Notably, expression of CD55 (decay accelerating factor) was significantly diminished in patients on ibrutinib (median decrease 37%, IQR 32%-46%, P<0.001; Fig. 5C). CD55 promotes the decay of proteases that amplify C3d opsonization (15,33). The reduction in CD55 expression on ibrutinib was inversely correlated to the decrease in C3d on CLL cells treated with ofatumumab ex vivo (n ¼ 16, R ¼ À0.60, P ¼ 0.01; Fig. 5D). Downregulation of CD55 by ibrutinib could explain the minimal reduction in C3d opsonization compared with CDC (Fig. 5B). In contrast, CD46 (membrane cofactor protein) and CD59 (protectin) levels were not consistently affected (median change À5.4% and 4.2%, respectively; Supplementary Fig. S4A and S4B). These results implicate CD55 as an important inhibitor of CLL cell surface opsonization by C3d. Furthermore, diminished expression of CD55 may play a significant role in augmenting the complement activation mediated by ofatumumab in the face of downregulated CD20 expression.
Ibrutinib reduces ofatumumab-induced loss of CD20 by inhibiting trogocytosis CLL cells have relatively lower CD20 expression than other Bcell malignancies. In addition, it is well established that anti-CD20 antibody therapy promotes the rapid loss of the antigen through trogocytosis (34). We used the methods of Beum and colleagues to study whether ibrutinib affects trogocytosis mediated by ofatumumab (28)(29)(30). Figure 6A shows a representative experiment in which CFSElabeled CLL cells were coincubated with THP-1 cells (monocyte cell line). At the beginning of coincubation, Alexa Fluor 594 labeled ofatumumab-stained CLL cells but did not bind to the monocyte cell line (Fig. 6A, left). After 2 hours of coculturing THP-1 cells with ofatumumab-bound CLL cells, CD20-labeled antibody was transferred to THP-1 cells, while a CD20-negative CLL population emerged (Fig. 6A, middle). In contrast, THP-1 cells treated with ibrutinib for 1 hour showed a dose-dependent decrease in trogocytosis (Fig. 6A, right; Fig. 6B; n ¼ 9, P <  Ibrutinib decreases ofatumumabmediated complement-dependent cytotoxicity. A, flow cytometric analysis of CDC mediated by ofatumumab in 33% normal human serum (NHS) against CLL cells obtained from 12 patients before treatment (pre), the day after treatment began (day 2), and at 2 and 6 months. B, correlation of ibrutinib effect on CD20 expression (x-axis) and CDC (y-axis) in the patients shown in A, n ¼ 12. C, CLL samples from the 10 patients with >20% decrease in CD20 expression and CDC were reassessed for CD20 expression and susceptibility to CDC after 1 to 4 days of culture in AIM V media. Symbols represent mean values. P values for all comparisons were calculated using the two-tailed Student paired t test. Correlations were determined using the Pearson test. The whiskers on all box plots depict the range (minimum and maximum values).

0.01
). This appears to be a direct inhibitory effect of ibrutinib as trogocytosis mediated by THP-1 cells is restored when BTK blockade by ibrutinib is bypassed by pretreatment with the PKC activator phorbol myristate acetate (Fig. 6B, far right).

Discussion
Combinations of kinase inhibitors with monoclonal antibodies (mAb) may improve upon the depth of responses obtained with single-agent therapy and are currently being explored in clinical studies (6)(7)(8)(9)(10). However, preclinical reports show multiple negative interactions between these two classes of drugs (14,(21)(22)(23). Here, we demonstrate that the BTK inhibitor ibrutinib exerts several effects on CLL cells and mAb effector mechanisms, some of which could reduce, while others could enhance, the efficacy of the combination. We found that CD20 expression was downregulated on tumor cells in peripheral blood and bone marrow of CLL patients treated with ibrutinib. We demonstrate that CD20 downregulation by ibrutinib is transcriptional and tightly correlated to inhibition of NF-kB signaling. In parallel to the decrease in CD20 expression, we observed a striking reduction in CDC mediated by the CD20-specific mAb ofatumumab.
Despite diminished CDC, complement deposition on CLL cells was relatively well maintained, which may be due to the concomitant downregulation of the complement regulatory protein CD55 on CLL cells by ibrutinib. In addition, ibrutinib inhibited antibody-mediated trogocytosis in vitro, a major mechanism of CD20 loss during mAb infusion (28)(29)(30)(35)(36)(37)(38), suggesting that ibrutinib may have positive as well as negative effects on CD20 expression.
We provide the first demonstration of CD20 downregulation on CLL cells by ibrutinib in vivo, confirming previous in vitro studies by others (14). Previously, Bojarczuk and colleagues have shown that in vitro treatment with ibrutinib and other BCR pathway inhibitors decreases CD20 mRNA and protein expression. Our data implicate NF-kB, which is downstream of BTK in the BCR pathway (39,40), as a link between BCR signaling and CD20 expression. Previous work by our group demonstrated that BCR and NF-kB signaling are both diminished upon ibrutinib treatment in a closely correlated manner (26). Here, we identified an NF-kB binding site at the CD20 gene transcription start site and show a tight correlation between ibrutinib-mediated changes in NF-kB and CD20 mRNA expression. In addition, we found a direct NF-kB inhibitor decreased CD20 surface expression on CLL A, correlation between cell lysis and the degree of C3d opsonization of CLL cells exposed to ofatumumab and normal human serum ex vivo. CLL cells were obtained from patients (n ¼ 16) pretreatment and at 2 months on ibrutinib. B, the percent change in ofatumumab-mediated CDC and C3d opsonization on ibrutinib compared with baseline (n ¼ 16 cells in vitro. Taken together, our data suggest inhibition of BCRmediated NF-kB signaling as a mechanism for the in vivo reduction in CD20 surface expression during ibrutinib treatment. Next, we examined the functional consequences of CD20 downregulation on the efficacy of anti-CD20 mAb therapy. CDC mediated by ofatumumab against CLL cells from ibrutinib-treated patients was severely impaired and the reduction of CDC ex vivo was strongly correlated to the in vivo downregulation of CD20 expression. These findings are consistent with the well-established connection between CD20 expression levels and the efficacy of anti-CD20 mAb-mediated CDC (15)(16)(17). Notably, in vivo ibrutinib-treated CLL cells recovered CD20 expression and CDC susceptibility during in vitro culture, indicating that ibrutinib-mediated CD20 downregulation is reversible.
Sequential administration of these agents could conceivably mitigate the effects of ibrutinib-mediated antigen loss. Indeed, administering ibrutinib before or after the anti-CD20 mAb rituximab yielded better antitumor activity in a mouse lymphoma model than concurrent administration (21). Furthermore, a clinical trial in CLL showed that starting ibrutinib first, followed by concurrent administration of ibrutinib and ofatumumab resulted in an objective response rate of 100% and was superior to sequential treatment with ofatumumab before ibrutinib or simultaneous administration of ibrutinib and ofatumumab (8). These data suggest that the benefit of combination therapy may depend on the sequence of administration.
Although ofatumumab-mediated CDC was strongly reduced, we noted that ex vivo opsonization of CLL cells by the complement protein C3 upon exposure to the mAb was relatively well-maintained. C3 is the central component of all complement activation pathways and a necessary element of convertases that deposit additional C3 molecules and initiate the formation of the membrane attack complex that is responsible for CDC (41). This prompted us to investigate whether ibrutinib modulated the expression of complement regulatory proteins, of which CD55 (decay accelerating factor) was found to be potently inhibited on ibrutinib. CD55 catalyzes the inactivation of convertases that amplify complement deposition (33).
The decrease in CD55 expression correlated with cell surface deposited C3 (C3d). Thus, the nonlytic functions of complement are partially preserved, despite reduced CD20 expression, when CD55 is concurrently downregulated. This suggests complementopsonized cells could be eliminated by complement receptormediated phagocytosis, a mechanism of cell clearance distinct from Fc-receptor dependent elimination of antibody-opsonized cells (18,(42)(43)(44). The biologic implications of complement opsonization warrant further study in combination therapy clinical trials.
Fc-receptor-mediated phagocytosis is an important mechanism of antibody activity (45,46). However, Fc-receptor bearing cells can also contribute to antigen loss through trogocytosis, a process in which the effector cells takes up the antibody-antigen complex from a target cell but fails to engulf it entirely (34). It has been shown that trogocytosis causes near complete antigen loss within hours of mAb infusion in CLL patients (35,36). Different strategies for preventing trogocytosis have been explored, including low or fractionated dosing of anti-CD20 mAbs (47)(48)(49). However, no pharmacologic inhibitors of trogocytosis have been developed for clinical use. Here, we report that ibrutinib potently inhibits trogocytosis of the ofatumumab-CD20 complex from CLL cells in vitro. Thus, ibrutinib could potentially avert antigen loss through trogocytosis during treatment with anti-CD20 antibodies. Assessing CD20 expression and C3d deposition on tumor cells before and after infusion of therapeutic mAbs in clinical trials that investigate combinations of antibody therapy with ibrutinib could clarify the relevance of these diverging effects.
A recent clinical trial demonstrated that combination therapy with ibrutinib and ofatumumab can be effective against relapsed/ refractory CLL. Specifically, lead-in treatment with single-agent ibrutinib followed by concurrent administration of ofatumumab and ibrutinib achieved a 100% overall response rate and an 88.7% 12-month progression-free survival rate (8). Notably, this schedule appeared to be superior to immediately beginning concurrent administration of ibrutinib and ofatumumab. The results of this clinical trial suggest that the combination can yield clinical benefit and that the outcome of combination therapy depends heavily on the schedule of administration. Further data will be necessary to conclusively determine the optimal administration sequence that can minimize the negative impact of ibrutinib on anti-CD20 mAb activity and make the most of the positive interactions between these therapeutic agents.
Clinical trials of ibrutinib in combination with anti-CD20 antibodies in CLL have reported encouraging results (6,8,9). Ibrutinib-mediated egress of tissue-resident CLL cells into the peripheral blood (12) and inhibition of CLL homing to tissue sites (50) could increase exposure of CLL cells to antibodies and serum complement and thus enhance the effect of mAb therapy. In addition, downregulation of CD20 and CD55, inhibition of effector cell function and mitigation of Fc-receptor dependent CD20 loss through trogocytosis likely impact the efficacy of mAb therapy. In extension to previous studies demonstrating a multitude of negative effects of ibrutinib on anti-CD20 mAb activity (14,(21)(22)(23), the data presented here suggest that beneficial interactions between ibrutinib and anti-CD20 mAbs could mitigate some of the negative effects. The kinetics and extent of positive and negative interactions and their net effect should be examined further in clinical trials that assess the utility of combining ibrutinib with anti-CD20 mAbs. Our data indicate that maximizing the antitumor activity of combination therapy requires successful harnessing of the opposing effects of ibrutinib on tumor cells and effector functions of anti-CD20 mAbs.

Disclosure of Potential Conflicts of Interest
C.U. Niemann is a consultant/advisory board member for Janssen and Roche. G.E. Marti is a consultant/advisory board member for Amerimmune Laboratory. A. Wiestner reports receiving a commercial research grant from and is a consultant/advisory board member for Pharmacyclics. No potential conflicts of interest were disclosed by the other authors.