Ibrutinib inhibits BCR and NF-κB signaling and reduces tumor proliferation in tissue-resident cells of patients with CLL

Abstract

Chronic lymphocytic leukemia (CLL) cells depend on microenvironmental factors for proliferation and survival. In particular, tissue-resident CLL cells show prominent activation of both B-cell receptor (BCR) and NF-κB pathways. We evaluated the in vivo effects of ibrutinib, a Bruton tyrosine kinase (BTK) inhibitor on tumor cell activation and proliferation in the blood, lymph node, and bone marrow of patients with CLL. Applying validated pathway-specific gene signatures, we detected a rapid and sustained downregulation of BCR and NF-κB signaling in CLL cells from both the peripheral blood and tissue compartments during ibrutinib treatment. Ibrutinib reduced phosphorylation of PLCγ2 and ERK and decreased nuclear protein expression of NF-κB p50. Ibrutinib significantly decreased tumor proliferation and expression of surface activation markers CD69 and CD86, independent of prognostic factors such as IGHV mutational status, chromosome 17p deletion, or prior treatment history. Interestingly, stronger inhibition of BCR signaling in lymph node resident CLL cells after one dose of ibrutinib was associated with a higher rate of nodal response at the end of cycle 2. Together, these data validate on-target effects of BTK inhibition in the tissue compartments and demonstrate that ibrutinib effectively inhibits pathways that promote tumor cell activation and proliferation in vivo. This study is registered at www.clinicaltrials.gov as #NCT01500733.

Figures

Figure 1

Ibrutinib induces rapid and sustained inhibition of BCR and NF-κB signaling in vivo in circulating CLL cells. (A-C) Change in BCR and NF-κB signature scores (identified in reference 1 and described in Materials and methods) in purified CLL cells on treatment. (A) Mean (± SEM) percent reduction in signature scores (n = 8 patients, in whom all time points were available). Comparisons by paired Student t test: *P < .05, **P < .01, and ***P < .001. (B) Percent reduction on day 28 compared with pretreatment. Each dot represents a different patient (n = 25). The line represents the median. (C) Correlation of reduction in BCR and NF-κB signature scores. Each dot represents one patient. R and P values of Pearson correlation are displayed (n = 25). (D-E) PBMCs were fixed, permeabilized, and stained with the indicated antibody. Results shown are for the CLL population (CD5+/CD19+). (D) A representative histogram of pPLCγ2 staining. Isotype control is represented by the gray shaded area, the dashed line represents pretreatment, and the solid black line day 28. (E) The mean (± SEM) percent of pPLCγ2 and pERK expressing CLL cells pretreatment (Pre) and on day 28 of ibrutinib treatment is shown (n = 30). (F) Quantification of nuclear p50 was done on nuclear lysates from CLL cells pretreatment (Pre) and on day 28 of ibrutinib treatment (n = 10).

Figure 2

Ibrutinib has a strong on-target effect in tissue-resident CLL cells. (A) BTK occupancy in lymph node (LN) biopsies on day 2 compared with pretreatment as determined by probe assay. Each dot represents a different patient (n = 6). (B-C) Percent reduction in BCR and NF-κB signature scores in either (B) LN biopsies (n = 24) or (C) CLL cells purified from bone marrow (BM) aspirates (n = 8). The line represents the median. (D) Correlation of reduction in BCR and NF-κB signature scores in LN biopsies (n = 24). (E-F) Mononuclear cells from LN biopsies (n = 8) or BM aspirates (n = 10) were fixed, permeabilized, and stained with the indicated antibody. Results shown are for the CLL population (CD5+/CD19+). The mean (± SEM) percent of (E) pPLCγ2 and (F) pERK expressing CLL cells pretreatment (Pre) and on ibrutinib is shown.

Figure 3

Ibrutinib inhibits cellular activation. (A-B) The percentage of CLL cells expressing the cell surface marker (A) CD69 and (B) CD86 on CLL cells from peripheral blood (PB, n = 20) was compared between pretreatment (Pre) and various treatment time points. Comparisons by paired Student t test: *P < .05, **P < .01, and ***P < .001. (C-D) The percentage of CLL cells expressing the cell surface marker (C) CD69 and (D) CD86 on CLL cells from PB (n = 30) was compared between pretreatment (Pre) and day 28. Comparisons are by paired Student t test. (E-F) Mononuclear cells from lymph node biopsies (LN, n = 8) or bone marrow (BM) aspirates (n = 10) were stained with the indicated antibody. Results shown are for the CLL population (CD5+/CD19+). The mean (± SEM) percent of (E) CD69 and (F) CD86 expressing CLL cells pretreatment (Pre) and on ibrutinib is shown. Comparisons are by paired Student t test.

Figure 4

Ibrutinib inhibits tumor proliferation in lymph node and bone marrow. (A) A representative histogram of Ki67 staining on CLL cells in the lymph node (LN). Isotype control is represented by the gray shaded area, the dashed line represents pretreatment, and the solid black line day 2. (B) Mononuclear cells from LN (n = 8) or bone marrow (BM) aspirates (n = 10) were fixed, permeabilized, and stained with the indicated antibody. Results shown are for the CLL population (CD5+/CD19+). The mean (± SEM) percent of Ki67 expressing CLL cells pretreatment and on ibrutinib is shown. Comparisons are by paired Student t test. (C) Images of LN tissue obtained pretreatment and on day 2 of ibrutinib stained with hematoxylin and eosin (H&E) and anti Ki67 were captured at ×40 magnification on an Olympus Bx41 microscope. (D) The percent of CLL cells expressing Ki67 at various clinical time points in the PB is shown (n = 20). Significance compared with pretreatment was determined by a paired Student t test and is indicated by asterisks: ***P < .001.

Figure 5

Inhibition of BCR signaling in the LN correlates with nodal response. (A-B) BCR gene signature scores are shown for patients with mutated (n = 9) and unmutated IGHV (n = 14). Comparisons within patients are by paired Student t test, whereas comparisons between mutational subgroups are by an unpaired Student t test. (A) Comparison of baseline BCR scores in CLL cells sampled from the peripheral blood (PB) and the matching lymph node (LN). (B) Comparison of pretreatment and day 2 BCR scores in LN core biopsies. (C) Correlation between percent reduction in lymphadenopathy and percent reduction in BCR signaling on day 2 in the lymph node. Gray squares represent patients who did not achieve a 50% reduction in lymphadenopathy after 2 months on treatment. Dashed lines represent the 50th percentile. (D) Percent reduction in lymphadenopathy in patients divided by the degree of reduction in the BCR signature score. Comparison is by Student t test.