venetoclax e ibrutinibe

Prévia do material em texto

Leukemia
https://doi.org/10.1038/s41375-019-0569-7
ARTICLE
Chronic lymphocytic leukemia
An in vitro assay for biomarker discovery and dose prediction
applied to ibrutinib plus venetoclax treatment of CLL
Sigrid S. Skånland 1,2,3,4 ● Andrea Cremaschi1,3,4,5 ● Henrik Bendiksen1,3 ● Johanne U. Hermansen2,4 ●
Deepak B. Thimiri Govinda Raj1,2,3,4 ● Ludvig A. Munthe3,6 ● Geir E. Tjønnfjord3,7 ● Kjetil Taskén1,2,3,4
Received: 21 February 2019 / Revised: 8 June 2019 / Accepted: 17 July 2019
© The Author(s), under exclusive licence to Springer Nature Limited 2019
Abstract
Recently, several small molecule drugs were approved for treatment of chronic lymphocytic leukemia (CLL), significantly
improving patient management. However, knowledge about how to combine these therapies for optimal effects and what
patients will best benefit from them is lacking. Here, we show that drug synergies can be identified by single cell signaling
analyses. We investigated the effects of idelalisib, ibrutinib, and venetoclax on 35 protein epitopes by phospho flow in CLL
cells. The activity of proteins in the B-cell receptor signalosome and the phosphatidylinositol 3-kinase pathway were altered
upon drug exposure. Combined treatment with ibrutinib and venetoclax give promising results in clinical studies and we
show that this combination exerted synergistic inhibitory effects on cell signaling and cell viability. Cell viability was
monitored by flow cytometry and with independent drug sensitivity screens. Our analyses indicate that the standard dosages
of ibrutinib and venetoclax can be lowered without loss of efficacy, potentially reducing drug costs, and toxicities. Observed
correlation between signaling and viability indicates that signaling molecules could serve as biomarkers to predict response
to therapy. We suggest that phospho flow should be considered as a novel approach for dose and synergy prediction in a
precision medicine setting for CLL.
Introduction
Chronic lymphocytic leukemia (CLL) is the most common
leukemia in the western world with an incidence rate
between 4 and 6 cases per 100,000 persons per year [1, 2].
The clinical outcome of CLL is very variable, but the
mutational status of the variable heavy chain gene region
(IgHV) of the B-cell receptor (BCR) has shown to be useful
in predicting disease progression [3]. The prognostic impact
of the IgHV mutational status suggests that signaling
through the BCR plays an important role in CLL patho-
genesis [4, 5]. This is further underscored by the recent
clinical successes of targeting and inhibiting different
components of the BCR pathway [3, 6].
Here, we performed single cell profiling of phospho-
protein levels in CLL cells treated with the targeted agents
idelalisib, ibrutinib, and venetoclax ex vivo. The aim was to
map the signaling responses of these drugs, as well as to
identify antagonistic, additive, or synergistic effects of drug
combinations. Combining novel drugs may result in more
patients with minimal residual disease-negative remission
and delay drug resistance [7]. Idelalisib inhibits the phos-
phatidylinositol 3-kinase δ isoform (PI3Kδ) (Fig. 1a).
* Sigrid S. Skånland
sigrid.skanland@medisin.uio.no
1 Department of Cancer Immunology, Institute for Cancer Research,
Oslo University Hospital, Oslo, Norway
2 K. G. Jebsen Centre for Cancer Immunotherapy, Institute of
Clinical Medicine, University of Oslo, Oslo, Norway
3 K. G. Jebsen Centre for B cell malignancies, Institute of Clinical
Medicine, University of Oslo, Oslo, Norway
4 Centre for Molecular Medicine Norway, Nordic EMBL
Partnership, University of Oslo, Oslo, Norway
5 Oslo Centre for Biostatistics and Epidemiology (OCBE),
University of Oslo, Oslo, Norway
6 Department of Immunology and Transfusion Medicine, Oslo
University Hospital, Oslo, Norway
7 Department of Haematology, Oslo University Hospital,
Oslo, Norway
Supplementary information The online version of this article (https://
doi.org/10.1038/s41375-019-0569-7) contains supplementary
material, which is available to authorized users.
12
34
56
78
90
()
;,:
12
34
56
78
90
();
,:
http://crossmark.crossref.org/dialog/?doi=10.1038/s41375-019-0569-7&domain=pdf
http://crossmark.crossref.org/dialog/?doi=10.1038/s41375-019-0569-7&domain=pdf
http://crossmark.crossref.org/dialog/?doi=10.1038/s41375-019-0569-7&domain=pdf
http://orcid.org/0000-0003-1630-356X
http://orcid.org/0000-0003-1630-356X
http://orcid.org/0000-0003-1630-356X
http://orcid.org/0000-0003-1630-356X
http://orcid.org/0000-0003-1630-356X
mailto:sigrid.skanland@medisin.uio.no
https://doi.org/10.1038/s41375-019-0569-7
https://doi.org/10.1038/s41375-019-0569-7
PI3Kδ is selectively expressed in lymphocytes and shows
increased activity in CLL cells [8]. The PI3Kδ pathway
regulates basic cellular functions such as proliferation,
metabolism, survival, and migration of lymphocytes. Ibru-
tinib inhibits the Bruton-tyrosine kinase (Btk) (Fig. 1a),
which tends to be overexpressed in CLL [9]. Btk is
upstream of phospholipase C gamma 2 (PLCγ2) and
nuclear factor kappa B (NF-κB) in the BCR-signaling
pathway and regulates proliferation and apoptosis of B cells
[10]. It binds irreversibly to the Cys-481 residue at the
active site of Btk, and suppresses the phosphorylation of
downstream proteins [11]. Venetoclax is a small molecule
highly selective for the proto-oncogene B-cell lymphoma 2
(Bcl-2) (Fig. 1a) which regulates the mitochondrial pathway
of apoptosis. Most CLL cases have elevated expression of
Bcl-2, resulting in evasion of apoptosis [12].
By applying phospho flow cytometry, 35 protein epi-
topes were screened before and after treatment with idela-
lisib, ibrutinib, and venetoclax alone or in combination. The
effects of BCR stimulation were also investigated. We show
that the novel agents primarily reduce phosphorylation of
proteins located in the BCR signalosome and in the PI3K
pathway.
Ibrutinib and venetoclax have been combined in clinical
trials in relapsed/refractory CLL [13] and in treatment-
naive high-risk CLL [14]. The combination is well toler-
ated and effective. Here, the ex vivo effects of ibrutinib
plus venetoclax were investigated in further detail. By
developing a novel tool to analyze drug synergy based on
signaling data we show that ibrutinib and venetoclax act in
a synergistic manner on several signaling proteins down-
stream of the BCR, as well as on apoptotic markers.
The drug synergy was confirmed in cell viability assays.
These findings are in line with clinical data [13, 14].
Our high-resolution synergy analyses indicate that the
doses of ibrutinib and venetoclax may be reduced without
loss of efficacy, which has been suggested by others [15].
Our study thus demonstrates that synergy detected at the
cell-signaling level may indicate synergistic effects on cell
viability, as well as on effective drug doses. This opens for
applying phospho flow analyses systematically in drug
sensitivity screens and to introduce signaling molecules as
biomarkers for drug sensitivity.
Materials and methods
Patient material and ethical considerations
Buffy coats from anonymized healthy blood donors were
procured from the Department of Immunology and
Transfusion Medicine, Oslo University Hospital and blood
samples from CLL patients from the Department of
Haematology, Oslo University Hospital following written
informed consent. The study was approved by the Regional
Committee for Medical and Health Research Ethics of
South-East Norway, and the research on human blood was
carried out in accordance with the Declaration of Helsinki
(2013).
Reagents and antibodies
Idelalisib, ibrutinib, and venetoclax were from Selleckchem
(Houston, TX, USA). The Alexa Fluor 647-conjugated
phospho-antibodies used in the study are listed in Supple-
mentary Table 1. Antihuman PerCP-Cy5.5 conjugated
CD19 was from eBioscience (San Diego, CA, USA).
Antihuman IgM (used at 10 µg/ml) was from Southern
Biotech (Birmingham, AL, USA). BD Phosflow Perm
Buffer III and Fix Buffer I were from BDBiosciences
(Franklin Lanes, NJ, USA). Fluorescent cell barcoding
fluorochromes Alexa Fluor 488-, Pacific Blue-, and Pacific
Orange Succinimidyl Ester were from Thermo Fisher Sci-
entific (Waltham, MA, USA).
Fig. 1 Titration of venetoclax, ibrutinib, and idelalisib. a Simplified
cartoon illustrating signaling networks downstream of CD19 and the
BCR. Action targets of venetoclax, ibrutinib, and idelalisib are indi-
cated. Signaling molecules shown in purple and green (dark shades)
were analyzed in this study by phospho flow. b–d CLL patient sam-
ples (n= 4) were treated with the indicated drug concentrations for
20 min before anti-IgM stimulation (5 or 30 min, indicated). The cells
were then fixed and processed for phospho flow analysis as described
in “Materials and methods” section. Statistical significance was cal-
culated relative to the control samples (DMSO) applying one-way
ANOVA for repeated measurements with multiple comparisons testing
(*p < 0.05, **p < 0.01, ***p < 0.001)
S. S. Skånland et al.
Isolation of B lymphocytes
Human peripheral blood CD19+ B cells were isolated from
buffy coats from healthy blood donors by negative selection
using a RosetteSep Human B-Cell Enrichment Cocktail
(1:50) (StemCell Technologies, Cambridge, UK) followed
by gradient centrifugation with Lymphoprep (Alere Tech-
nologies AS, Oslo, Norway) according to manufacturer’s
protocol. Peripheral blood mononuclear cells were isolated
from CLL patient blood samples following the same pro-
cedure, but without negative selection. Cell samples were
cryopreserved in liquid nitrogen. After thawing, the cells
were rested in RPMI 1640 GlutaMAX medium supple-
mented with 10% fetal calf serum for 1 h at 37 °C and 5%
CO2 [16]. Clinical characteristics of the CLL patients
included in this study are listed in Supplementary Table 2.
Phospho flow with fluorescent cell barcoding
Phospho flow assays were performed with fluorescent cell
barcoding as described [17, 18]. Controls from the Blood
Centre were drawn from a donor population where 45% were
>45 years of age. Antibody-stained cells were run on a BD
FACSCanto II (4-2-2) cytometer equipped with 405, 488, and
633 nm lasers. The data were analyzed in Cytobank (https://
cellmass.cytobank.org/cytobank/) as described [17].
CellTiter-Glo luminescent cell viability assay
The compounds were dissolved in DMSO and preprinted
into 384-well plates with an acoustic liquid handling device
(Echo 550, Labcyte Inc., CA, USA). Each compound was
tested at five different concentrations ranging from 1 to
10,000 nM. Combinations were designed using the fixed
molar concentration series identical to those used for single
agents. CLL cells were co-cultured with CD40L+, BAFF+,
and APRIL+ L cells (ratio 1:1:1) for 24 h prior to initiation
of the experiment to mimic the tumor microenvironment
and to prevent spontaneous apoptosis. The L cells were
removed by positive selection using PE anti-mouse CD47
antibody (Biolegend, CA, USA) and anti-PE microbeads
(Miltenyi Biotec, Bergisch Gladbach, Germany) according
to the manufacturer’s instructions. B cells from age-
matched healthy donors were treated the same way. Sin-
gle cell suspension (10,000 cells/well) was distributed to
each well using a peristaltic dispenser (MultiDrop Combi,
ThermoFisher Scientific, MA, USA). The plates were
incubated at 37 °C for 72 h. Cell viability was measured
using the CellTiter-Glo (Promega, WI, USA) luminescent
assay according to the manufacturer’s instructions. Lumi-
nescence was recorded with an EnVision 2102 Multilabel
reader (PerkinElmer, MA, USA). The response readout was
normalized to the negative (DMSO) and positive (100 µM
benzethonium chloride) controls. The raw dose-response
data were processed with the KNIME software (KNIME
AG, Zurich, Switzerland).
Synergy analyses and statistics
Statistical analyses were performed using GraphPad Prism 7
(San Diego, CA), applying the statistical tests indicated in
the respective figure legends. Normality of the data was
confirmed by a QQ plot. The phospho flow data obtained
when applying the ibrutinib/venetoclax combination matrix
were analyzed with R (https://www.r-project.org). The
responses were first transformed to the suitable range of
values (0,100) using a linear map, and then used to quantify
the synergy scores. In particular, the real-valued phospho
flow data ~y were transformed as follows:
y ¼ 100� ð~y� minð~yÞÞ=ðmaxð~yÞ �minð~yÞÞ, if they
exhibit a decreasing pattern, and
y ¼ 100� ðmax ~yð Þ � ~yÞ=ðmaxð~yÞ � minð~yÞÞ, if they
exhibit an increasing pattern. The type of pattern was
established by looking at the sign of the regression coeffi-
cient of the linear regression ~y ¼ β x1 � x2, for each pair of
concentrations (x1, x2) tested. Hence, the transformed
observations y belong to the desired interval (0,100).
To perform the analysis of the interactions, the reference
models Bliss [19] and highest single agent (HSA) [20] were
applied on the transformed data set, and synergy scores
were then calculated.
Results
Titration of idelalisib, ibrutinib, and venetoclax
The novel agents idelalisib, ibrutinib, and venetoclax target
distinct molecules downstream of the BCR or in the reg-
ulation of apoptosis in order to induce cell death (Fig. 1a).
Several studies have suggested that combinations of these
drugs may be beneficial in order to prevent drug resistance
and to achieve a deeper response [13, 14]. In order to, in the
future, identify biomarkers which may predict drug
responses in the individual patient, the first aim of the
present study was to investigate the impact of the novel
agents on signaling molecules by phospho flow cytometry.
This method allows high-throughput screening of signaling
molecules. Our hypothesis was that the activity of signaling
molecules may predict response to therapy. In order to
determine an appropriate drug concentration to be applied in
combination studies for subsequent identification of antag-
onistic, additive, or synergistic drug effects, the effect of
drugs on leukemic cells from CLL patients were analyzed
at five different concentrations (Fig. 1b–d). The Bcl-2
antagonist venetoclax induced significant phosphorylation
An in vitro assay for biomarker discovery and dose prediction applied to ibrutinib plus venetoclax. . .
https://cellmass.cytobank.org/cytobank/
https://cellmass.cytobank.org/cytobank/
https://www.r-project.org
of Histone H2AX (pS139), a marker of early apoptosis, at
concentrations starting at 1 µM (Fig. 1b). The BCR pathway
inhibitors ibrutinib and idelalisib inhibited receptor prox-
imal signaling and the PI3K pathway at concentrations from
1 and 0.001 µM, respectively (Fig. 1c, d). Based on these
analyses, the selected working concentrations were 1 µM
venetoclax, 1 µM ibrutinib, and 0.01 µM idelalisib.
Ibrutinib, but not idelalisib or venetoclax, inhibits
the phosphorylation of BCR proximal signaling
molecules
Upon BCR stimulation, the Src family kinase Lyn phos-
phorylates ITAMs located on CD79a and CD79b, which next
leads to recruitment and activation of SYK. This is followed
by formation of the BCR signalosome that includes BLNK,
PLCγ2, GRB2, VAV, and Btk (Fig. 1a). The effects of ide-
lalisib, ibrutinib, and venetoclax on the activity of BCR
proximal signaling molecules were investigated. Cells were
treated either with single agents or combinations of two drugs
(Fig. 2). Interestingly, only the Btk inhibitor ibrutinib, alone
or in combination with idelalisib or venetoclax, was able to
significantly reduce the phosphorylation of some of the BCR
proximal signaling molecules (Fig. 2). As expected, Btk
(pY551) was reduced in CLL cells (Fig. 2) as well as normal
B cells (Supplementary Fig. 1). Similarly, the downstream
proteins PLCγ2 (pY759) and BLNK (pY84) were inhibited in
normal B cells (Supplementary Fig. 1) and CLL cells (Fig. 2).
Src (pY418) was significantly reduced in both normal and
leukemic B cells upon ibrutinib treatment (Fig. 2, Supple-
mentary Fig. 1). SYK, which is downstream of the Srckinase
Lyn (Fig. 1a), showed reduced phosphorylation on epitope
Y352 after ibrutinib exposure, while Y525/Y526 phosphor-
ylation remained unaffected (Fig. 2, Supplementary Fig. 1).
The inhibitory effects of ibrutinib were similar or slightly
more potent in normal cells compared with CLL cells
(Fig. 2b). Surprisingly, ibrutinib only inhibited phosphoryla-
tion of BCR proximal signaling in unstimulated cells. The
same effects could not be detected in CLL or normal B cells
stimulated with anti-IgM for 5 min (data not shown). Of note,
however, it was observed that the receptor proximal mole-
cules exerted higher phosphorylation levels in normal B cells
than in CLL cells after anti-IgM stimulation (Supplementary
Fig. 2). This is in agreement with our earlier report [17] and
suggests that CLL cells are hyporesponsive to BCR activa-
tion. Of the three novel drugs, only ibrutinib inhibits receptor
proximal signaling both in CLL and in normal B cells.
Novel drugs inhibit the PI3K-signaling pathway
downstream of the BCR
Once the BCR signalosome is formed, downstream-
signaling pathways are activated, including the PI3K-Akt-
mTOR pathway (Fig. 1a). Idelalisib is a PI3Kδ inhibitor,
and its effects on PI3K signaling are well established [21].
Here, the phosphorylation level of Akt (pT308 and pS473),
which is downstream of PI3K, and the further downstream
molecules mTOR (pS2448), S6-ribosomal protein (pS235/
S236), and NF-κB p65 (pS529 and pS536) were investi-
gated after drug treatment and BCR stimulation. In agree-
ment with earlier reports [21, 22], both ibrutinib and
idelalisib were found to reduce the level of pAkt (Fig. 3a).
The combination of these inhibitors appeared to have an
additive inhibitory effect on pAkt, as indicated by increased
difference from the control condition (Fig. 3a). When ana-
lyzing the absolute value of the averaged signals from
normal B cells (n= 10) and CLL cells (n= 10), the Akt
Fig. 2 Ibrutinib, but not idelalisib or venetoclax, inhibits the phos-
phorylation of BCR proximal molecules in CLL cells. a CLL patient
samples (n= 10) were treated with idelalisib (0.01 µM), ibrutinib
(1 µM), venetoclax (1 µM), or combinations as indicated, for 20 min.
The cells were then fixed and processed for phospho flow analysis (see
“Materials and methods” section). Statistical significance was calcu-
lated relative to the control samples (DMSO) applying one-way
ANOVA with multiple comparisons testing (**p < 0.01, ***p <
0.001). b Averaged data from ten healthy donors and ten CLL patients
(see a) presented in heatmaps as arcsinh ratio
S. S. Skånland et al.
(pS473) level was indeed further reduced upon combination
treatment (ID, IB= arcsinh ratio 0.09; Fig. 3b) as compared
with single drug treatment (ID= arcsinh ratio 0.31 and IB
= arcsinh ratio 0.22; Fig. 3b). Interestingly, venetoclax also
significantly reduced the Akt (pS473) level (Fig. 3a) and
showed an additive inhibitory effect in combination with
both idelalisib and ibrutinib (Fig. 3b). All three drugs also
inhibited the downstream molecule S6-ribosomal protein
(pS235/S236), with venetoclax displaying the most potent
effect (Fig. 3a). mTOR and NF-κB p65 phosphorylation
was not reduced upon drug treatment of CLL cells (Fig. 3a).
In normal B cells, however, the PI3K pathway was more
generally inhibited by the three drugs (Supplementary
Fig. 3). Even so, the drug effects were much more potent in
CLL cells, as illustrated in Fig. 3b. No effects wer-
e observed after drug treatment in unstimulated cells
(Supplementary Fig. 4 and data not shown). In summary,
the novel agents inhibit signaling molecules downstream of
the PI3K pathway both in CLL cells and normal B cells, but
most efficiently in CLL cells.
Correlation analysis of cell signaling and cell
viability responses to ibrutinib plus venetoclax
treatment
Several clinical trials are investigating the combined treat-
ment of ibrutinib and venetoclax, and initial reports have
demonstrated that this combination is well tolerated and
effective [13, 14]. In order to investigate if signaling aber-
rations observed after ibrutinib plus venetoclax exposure
can predict cell viability, we performed phospho flow
experiments with 35 protein readouts, including multiple
markers for apoptosis. Figure 4 shows the correlation matrix
of the 35 analyzed proteins with the significant correlations
indicated in the upper right triangle. Interestingly, as the
only signaling molecule S6-ribosomal protein (pS235/S236)
displayed a statistically significant negative correlation with
the apoptotic markers cleaved PARP, Histone H2AX
(pS139), and p53 (pS37) (see first row in Fig. 4). A negative
correlation was also observed with cleaved caspase-3
(Fig. 4), but this correlation was not found statistically
significant (see first column in Fig. 4). Importantly,
expression of the apoptotic markers cleaved caspase-3,
cleaved PARP, Histone H2AX (pS139), and p53 (pS37) all
displayed pair-wise positive correlations (Fig. 4). These
findings confirm that activity of signaling molecules cor-
relates with induction of cell death, and they suggest that
S6-ribosomal protein (pS235/S236) may serve as a sensi-
tivity marker for response to ibrutinib plus venetoclax.
Drug synergy analysis of an ibrutinib plus
venetoclax combination matrix
Among the rationales for combining ibrutinib and venetoclax
are results from preclinical models showing that the drugs
have synergistic effects [14, 23]. Here, we aimed at investi-
gating whether synergy can be detected at the cell-signaling
level. A drug matrix was designed in which ibrutinib and
venetoclax was included in tenfold dilution steps ranging
from 1 to 10,000 nM and combined at each concentration,
generating a total of 36 data points (Fig. 5a). CLL cells from
ten patients with IgHV unmutated (UM-CLL) status were
exposed to the drug matrix followed by anti-IgM stimulation
for 5 min. Traditionally, drug synergy is calculated from cell
viability measurements for which the data points range
between 0 and 100% viability. In order to be able to perform
Fig. 3 Idelalisib, ibrutinib, and venetoclax inhibit the PI3K pathway
downstream of the BCR in CLL cells. a B cells from CLL patients
(n= 10) were treated with drugs as indicated for 20 min before 5 or
30 min (S6-rib prot) of anti-IgM stimulation. The cells were then fixed
and processed for phospho flow analysis (see “Materials and methods”
section). Statistical significance was calculated relative to the control
samples (DMSO) applying one-way ANOVA with multiple compar-
isons testing (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001). b
Averaged data from ten healthy donors and ten CLL patients (see a)
presented in heatmaps as arcsinh ratio
An in vitro assay for biomarker discovery and dose prediction applied to ibrutinib plus venetoclax. . .
similar analyses on signaling measurements, the data were
first transformed to range between 0 and 100 (not shown).
The averaged synergy scores of the combination matrix are
presented in Fig. 5b for each of the 35 proteins studied above.
Among the signaling proteins that clustered together as
positive responders were several from the BCR signalosome
or the PI3K pathway, including S6-ribosomal protein (pS235/
S236) (Fig. 5b, red box). Interestingly, a second branch of the
same cluster contained all four apoptotic markers; Histone
H2AX (pS139), cleaved caspase-3, p53 (pS37), and cleaved
PARP (Fig. 5b, blue box). This finding demonstrates that
synergy can be calculated from phospho flow data, and shows
that ibrutinib plus venetoclax acts synergistically on both
signaling molecules and apoptotic markers. When the synergy
matrices for S6-ribosomal protein (pS235/S236) and cleaved
caspase-3 were analyzed in 2D and 3D by the HSA (Fig. 5c)
and Bliss reference models (Fig. 5d), very similar synergy
patterns were observed, supporting an interaction between
these molecules. Importantly, the high-resolution synergy
analysis provides detailed information on effective drug
doses. Ibrutinib is commonlyadministered at 420mg/d,
which results in a plasma concentration of about 100 ng/mL
[24]. This corresponds to an in vitro concentration of about
200 nM. Venetoclax is administered at up to 400mg/d, which
gives a peak plasma concentration of about 1.75 µg/mL [25]
and an in vitro concentration of about 2000 nM. As shown for
patient CLL138 in Fig. 5c, ibrutinib gave an optimal syner-
gistic effect at doses > 20 nM combined with venetoclax >
100 nM, suggesting that both ibrutinib and venetoclax doses
may be reduced considerably and still induce synergy. Similar
results were found in patients CLL105, CLL153, and
CLL171 (Supplementary Fig. 5). However, we observed large
variation in responses between patients. For example, patient
CLL135 required almost 100 nM of both ibrutinib and
venetoclax to achieve synergy. Results demonstrated a
marked drug synergy in 8/10 patients (with less effect in
Fig. 4 Phosphorylation of S6-
ribosomal protein negatively
correlates with markers for
apoptosis. B cells from CLL
patients (n= 10) were exposed
to ibrutinib (1 µM) plus
venetoclax (1 µM) for 20 min
followed by anti-IgM
stimulation for 5 min. The cells
were then fixed and processed
for phospho flow analysis with
the indicated antibodies (see
“Materials and methods”
section). The matrix shows
paired protein correlations
(lower left triangle). Increasing
correlation is indicated by darker
color (red for negative, blue for
positive). Statistically significant
correlations were assessed by
Pearson’s r-test, p < 0.05, and
are plotted in the upper, right
triangle of the correlation matrix
S. S. Skånland et al.
CLL149 and CLL174) suggesting advantage of drug com-
binations (Supplementary Fig. 5).
In summary, combined treatment with ibrutinib and
venetoclax induces synergistic effects on both signaling
molecules and apoptotic markers. Synergy analyses based
on phospho flow data may guide dose prediction and pre-
cision medicine.
Ibrutinib plus venetoclax selectively and
synergistically kill CLL cells
In order to assess whether drug combinations that display
synergy in perturbing cell signaling translates into synergy
with respect to reducing cell viability, 5 normal and 20 CLL
samples were subjected to treatment with combinations of
Fig. 5 Combination of ibrutinib
and venetoclax exerts
synergistic inhibitory effects on
CLL cell signaling and cell
viability. a Illustration of the
applied ibrutinib plus venetoclax
combination matrix. b B cells
from CLL patients (n= 10) were
treated with the drug matrix
shown in a for 20 min before
5 min of anti-IgM stimulation.
The cells were then fixed and
processed for phospho flow
analysis (see “Materials and
methods” section). Hierarchical
agglomerative clustering
(Ward’s method) of averaged
synergy scores on the indicated
protein epitopes calculated by
the Highest Single Agent (HSA)
model. c 2D and 3D
visualization of HSA synergy
scores on S6-ribosomal protein
(pS235/S236) and cleaved
caspase-3 for patient CLL138.
d As in c, but synergy scores
were calculated with the Bliss
model. e Illustration of the
applied diagonal of the ibrutinib
plus venetoclax combination
matrix. f Primary B cells from
CLL patients (n= 20) or normal
controls (n= 5) were co-
cultured with CD40+, BAFF+,
and APRIL+ (ratio 1:1:1) L
cells for 24 h prior to initiation
of the experiment to mimic the
tumor microenvironment. The L
cells were then removed and the
CLL cells were exposed to the
drugs for 72 h. Cell viability was
measured with the CellTiter-Glo
assay. Bliss synergy scores were
calculated and are shown for
each concentration in the drug
combination. g Visualization of
the Bliss synergy scores
achieved after treatment with
100 nM ibrutinib plus
venetoclax in f. Bars indicate
mean synergy score.
Significance was calculated
using Welch’s t test (p < 0.05)
An in vitro assay for biomarker discovery and dose prediction applied to ibrutinib plus venetoclax. . .
drugs at concentrations along the diagonal of the drug
matrix for 72 h (Fig. 5e). Importantly, the B cells were
primed with environmental factors for 24 h prior to drug
exposure in order to mimic the tumor environment and to
prevent induction of spontaneous apoptosis. As shown in
Fig. 5f, g, drug synergy was observed in the majority of the
patient samples while the normal B cells were spared.
Collectively, these findings demonstrate that synergy
detected at the level of cell signaling can translate to
synergy in CLL cell killing.
Discussion
BCR pathway inhibitors have demonstrated significant
clinical activity in CLL. However, in order to provide
optimally tailored treatment, more knowledge is needed on
response versus dosing regimens and with respect to pre-
dictive biomarkers and companion diagnostics. The aim of
the present study was to characterize the effects of the tar-
geted agents idelalisib, ibrutinib, and venetoclax on cell
signaling, and to investigate if drug-induced alterations in
phosphorylation of cell signaling molecules can predict
drug sensitivity and synergy.
Initially, we focused on two signaling pathways
downstream of the BCR; BCR proximal molecules
including the BCR signalosome and the PI3K pathway.
Only the Btk inhibitor ibrutinib reduced the activity of
BCR proximal molecules suggesting little crosstalk from
idelalisib and venetoclax, which, unlike ibrutinib, target
molecules that are not part of the BCR signalosome.
Ibrutinib has been shown to have inhibitory effects on
BCR signaling also in vivo [26]. When studying the PI3K
pathway, inhibitory effects of all three drugs were
observed. Idelalisib targets PI3Kδ, and inhibition of
downstream molecules was therefore expected and have
been reported earlier [17, 21]. It has also been shown that
ibrutinib reduces Akt phosphorylation in acute myeloid
leukemia [27] and CLL [22]. To the best of our knowl-
edge, similar effects have not previously been reported for
venetoclax. However, it has been shown that acquired
resistance to venetoclax results in Akt activation and
enhanced sensitivity towards PI3K inhibition, providing a
rationale for combination therapy [28–30]. In CLL, clin-
ical trials are combining treatment with venetoclax and the
PI3K inhibitors duvelisib (ClinicalTrials.gov Identifier
NCT03534323), umbralisib/TGR-1202 (NCT03379051,
NCT03801525), or idelalisib (NCT03639324).
Here, we studied in detail the responses to combined
treatment with ibrutinib plus venetoclax. Several clinical
trials are investigating this combination, and encouraging
results have recently been reported [13, 14]. The rationale
for combining these therapies includes nonoverlapping
mechanisms of action, nonoverlapping toxicity profiles,
complementary activity in treating disease compartments,
and preclinical models showing synergy [14]. However, de
novo resistance to ibrutinib plus venetoclax has been
reported [31]. In order to investigate the effects of ibrutinib
plus venetoclax treatment on cell signaling, we performed
phospho flow experiments with 35 protein readouts,
including apoptotic markers. We identified a significant
negative correlation between the apoptotic markers and the
signaling molecule S6-ribosomal protein (pS235/S236),
which is downstream of PI3K. This means that when the
cells are becoming apoptotic, i.e. the expression level of the
apoptotic markers increases, the activity of S6-ribosomal
protein decreases. This molecule may therefore serve as an
indicator of drug response.
Our high-resolution synergy analysis revealed that the
combination of ibrutinib and venetoclax is active over a
relatively large range of doses. The concentration ranges
were patient dependent, but cells from 8/10 patients
responded markedly to drug combinations. These findings
suggest that drug doses may be lowered 10- to 100-fold
without reducing activity. Importantly, this is in agreement
with a recent study showing that ibrutinib dose may be
reduced without loss of biological effect [15]. Our results
suggest an extension of this finding to drug synergy with
venetoclax. Clinical validation of ourprediction model is
needed and would introduce a completely novel approach to
dose prediction. If drug doses can be lowered as indicated
by the present study, drug costs could be dramatically
reduced.
Unmutated IgVH status has been associated with
increased sensitivity to ibrutinib ex vivo [32], while IgVH
mutational status does not significantly affect PFS of
patients treated with ibrutinib [33, 34]. In our signaling
analysis on a mixed population of UM-CLL and M-CLL
samples (Figs. 1–3), we observed that drug-induced sig-
naling alterations were more readily observed in UM-CLL
samples. This is likely due to higher signaling amplitudes
induced by anti-IgM stimulation in these cells [17].
Therefore, for the high-resolution signaling analysis of
ibrutinib plus venetoclax treatment only UM-CLL patient
samples were included (Figs. 4 and 5). In the relatively
small cohort of patient samples included in the drug sen-
sitivity analysis (n= 20), no correlation was observed
between patient characteristics and response to therapy. Of
the eight highest responders (response above mean), five
patients had UM-CLL and three had M-CLL. In an ex vivo
drug sensitivity screen on 26 CLL samples, sensitivity
towards venetoclax was also shown to be independent
of IgVH mutational status [35], supporting the present
findings.
In summary, we have characterized the effects of tar-
geted agents on cell signaling in CLL cells, alone and in
S. S. Skånland et al.
http://ClinicalTrials.gov
combination. We found that drug-induced alterations in cell
signaling might predict drug sensitivity. The present study
confirms and expands on the synergistic mechanism
of ibrutinib plus venetoclax treatment of CLL, and our
findings encourage further studies to identify and validate
signaling molecules as biomarkers for drug sensitivity. We
suggest that phospho flow can serve as a powerful and
appropriate method for biomarker discovery and prediction
of drug doses.
Acknowledgements The authors are grateful to all the study partici-
pants without whom this study would not have been possible. We
thank Martine Schrøder, Marianne Enger, and Gladys M. Tjørhom for
expert technical assistance. We are grateful to Alexandra Gade,
Johannes Landskron, and Eirin Solberg at the High-Throughput
Chemical Biology Screening Platform at Centre for Molecular Medi-
cine Norway, University of Oslo, for assistance with drug sensitivity
screens. This work was supported by the Norwegian Cancer Society,
the Regional Health Authority for South-Eastern Norway, the
Research Council of Norway and Stiftelsen Kristian Gerhard Jebsen
(grant number SKGJ-MED-019). SSS was funded by a Career
Development Research Fellowship from the Norwegian Cancer
Society.
Author contributions SSS designed the research. SSS, HB, and JUH
performed the experiments and analyzed the data together with AC,
LAM, and KT. DBTGR provided methodology. GET and LAM
contributed with patient samples and interpreted data. SSS wrote the
paper. All authors read and commented on draft versions of the paper
and approved the final version.
Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of
interest.
Publisher’s note: Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
References
1. Scarfo L, Ferreri AJ, Ghia P. Chronic lymphocytic leukaemia. Crit
Rev Oncol Hematol. 2016;104:169–82.
2. Lenartova A, Johannesen TB, Tjønnfjord GE. National trends in
incidence and survival of chronic lymphocytic leukemia in Nor-
way for 1953-2012: a systematic analysis of population-based
data. Cancer Med. 2016;5:3588–95.
3. Kipps TJ, Stevenson FK, Wu CJ, Croce CM, Packham G, Wierda
WG, et al. Chronic lymphocytic leukaemia. Nat Rev Dis Prim.
2017;3:16096.
4. Zhong Y, Byrd JC, Dubovsky JA. The B-cell receptor pathway: a
critical component of healthy and malignant immune biology.
Semin Hematol. 2014;51:206–18.
5. Ten Hacken E, Gounari M, Ghia P, Burger JA. The importance of
B cell receptor isotypes and stereotypes in chronic lymphocytic
leukemia. Leukemia. 2019;33:287–98.
6. Arnason JE, Brown JR. Targeting B cell signaling in chronic
lymphocytic leukemia. Curr Oncol Rep. 2017;19:61.
7. Khan M, Siddiqi T. Targeted therapies in CLL: monotherapy
versus combination approaches. Curr Hematol Malig Rep. 2018;
13:525–33.
8. Brown JR. The PI3K pathway: clinical inhibition in chronic
lymphocytic leukemia. Semin Oncol. 2016;43:260–4.
9. Falchi L, Baron JM, Orlikowski CA, Ferrajoli A. BCR signaling
inhibitors: an overview of toxicities associated with ibrutinib and
idelalisib in patients with chronic lymphocytic leukemia. Mediterr
J Hematol Infect Dis. 2016;8:e2016011.
10. Vela CM, McBride A, Jaglowski SM, Andritsos LA. Ibrutinib for
treatment of chronic lymphocytic leukemia. Am J Health Syst
Pharm. 2016;73:367–75.
11. Vitale C, Burger JA. Chronic lymphocytic leukemia therapy: new
targeted therapies on the way. Expert Opin Pharm. 2016;17:1077–89.
12. Itchaki G, Brown JR. The potential of venetoclax (ABT-199) in
chronic lymphocytic leukemia. Ther Adv Hematol. 2016;7:
270–87.
13. Hillmen P, Rawstron A, Brock K, Munoz Vicente S, Yates F,
Bishop R, et al. Ibrutinib plus venetoclax in relapsed/refractory
CLL: results of the bloodwise TAP Clarity Study. Blood. 2018;
132:182.
14. Jain N, Keating MJ, Thompson PA, Ferrajoli A, Burger JA,
Borthakur G, et al. Ibrutinib and venetoclax for first-line treatment
of CLL. N Engl J Med. 2019;380:2095–103.
15. Chen LS, Bose P, Cruz ND, Jiang Y, Wu Q, Thompson PA, et al.
A pilot study of lower doses of ibrutinib in patients with chronic
lymphocytic leukemia. Blood. 2018;132:2249–59.
16. Hermansen JU, Tjønnfjord GE, Munthe LA, Taskén K, Skånland
SS. Cryopreservation of primary B cells minimally influences
their signaling responses. Sci Rep. 2018;8:17651.
17. Myhrvold IK, Cremaschi A, Hermansen JU, Tjønnfjord GE,
Munthe LA, Taskén K, et al. Single cell profiling of phospho-
protein levels in chronic lymphocytic leukemia. Oncotarget.
2018;9:9273–84.
18. Skånland SS. Phospho flow cytometry with fluorescent cell bar-
coding for single cell signaling analysis and biomarker discovery.
J Vis Exp. 2018;140:e58386.
19. Bliss C. The toxicity of poisons applied jointly. Ann Appl Biol.
1939;26:585–615.
20. Berenbaum MC. What is synergy? Pharm Rev. 1989;41:93–141.
21. Brown JR, Byrd JC, Coutre SE, Benson DM, Flinn IW, Wagner-
Johnston ND, et al. Idelalisib, an inhibitor of phosphatidylinositol
3-kinase p110delta, for relapsed/refractory chronic lymphocytic
leukemia. Blood. 2014;123:3390–7.
22. Cheng S, Ma J, Guo A, Lu P, Leonard JP, Coleman M, et al. BTK
inhibition targets in vivo CLL proliferation through its effects on
B-cell receptor signaling activity. Leukemia. 2014;28:649–57.
23. Cervantes-Gomez F, Lamothe B, Woyach JA, Wierda WG,
Keating MJ, Balakrishnan K, et al. Pharmacological and protein
profiling suggests venetoclax (ABT-199) as optimal partner with
ibrutinib in chronic lymphocytic leukemia. Clin Cancer Res.
2015;21:3705–15.
24. Advani RH, Buggy JJ, Sharman JP, Smith SM, Boyd TE, Grant
B, et al. Bruton tyrosine kinase inhibitor ibrutinib (PCI-32765) has
significant activity in patients with relapsed/refractory B-cell
malignancies. J Clin Oncol. 2013;31:88–94.
25. Salem AH, Agarwal SK, Dunbar M, Enschede SL, Humerick-
house RA, Wong SL. Pharmacokinetics of venetoclax, a novel
BCL-2 inhibitor, in patients with relapsed or refractory chronic
lymphocytic leukemia or non-hodgkin lymphoma. J Clin Pharm.
2017;57:484–92.
26. Herman SE, Mustafa RZ, Gyamfi JA, Pittaluga S, Chang S, Chang
B, et al. Ibrutinib inhibits BCR and NF-kappaB signaling and
reduces tumor proliferation in tissue-resident cells of patients with
CLL. Blood. 2014;123:3286–95.
27. Zaitseva L, Murray MY, Shafat MS, Lawes MJ, MacEwan DJ,
Bowles KM, et al. Ibrutinib inhibits SDF1/CXCR4 mediated
migration in AML. Oncotarget. 2014;5:9930–8.
An in vitro assay for biomarker discovery and dose prediction appliedto ibrutinib plus venetoclax. . .
28. Choudhary GS, Al-Harbi S, Mazumder S, Hill BT, Smith MR,
Bodo J, et al. MCL-1 and BCL-xL-dependent resistance to the
BCL-2 inhibitor ABT-199 can be overcome by preventing PI3K/
AKT/mTOR activation in lymphoid malignancies. Cell Death Dis.
2015;6:e1593.
29. Pham LV, Huang S, Zhang H, Zhang J, Bell T, Zhou S, et al.
Strategic therapeutic targeting to overcome venetoclax resistance in
aggressive B-cell lymphomas. Clin Cancer Res. 2018;24:3967–80.
30. Bodo J, Zhao X, Durkin L, Souers AJ, Philips DC, Smith MR,
et al. Acquired resistance to venetoclax (ABT-199) in t(14;18)
positive lymphoma cells. Oncotarget. 2016;7:70000–10.
31. Jayappa KD, Portell CA, Gordon VL, Capaldo BJ, Bekiranov S,
Axelrod MJ, et al. Microenvironmental agonists generate de novo
phenotypic resistance to combined ibrutinib plus venetoclax in
CLL and MCL. Blood Adv. 2017;1:933–46.
32. Amin NA, Balasubramanian S, Saiya-Cork K, Shedden K, Hu N,
Malek SN. Cell-intrinsic determinants of ibrutinib-induced
apoptosis in chronic lymphocytic leukemia. Clin Cancer Res.
2017;23:1049–59.
33. Barr PM, Robak T, Owen C, Tedeschi A, Bairey O, Bartlett NL,
et al. Sustained efficacy and detailed clinical follow-up of first-line
ibrutinib treatment in older patients with chronic lymphocytic
leukemia: extended phase 3 results from RESONATE-2. Hae-
matologica. 2018;103:1502–10.
34. Fraser G, Cramer P, Demirkan F, Silva RS, Grosicki S, Pristupa
A, et al. Updated results from the phase 3 HELIOS study of
ibrutinib, bendamustine, and rituximab in relapsed chronic lym-
phocytic leukemia/small lymphocytic lymphoma. Leukemia.
2019;33:969–80
35. Thimiri Govinda Raj DB, Cremaschi A, Skånland SS, Gade A,
Schjesvold FH, Tjønnfjord GE, et al. In-vitro drug sensitivity
screening in chronic lymphocytic leukemia (CLL) primary patient
samples identifies drug candidates for precision cancer therapy.
Blood. 2018;132:4677.
S. S. Skånland et al.
	An in�vitro assay for biomarker discovery and dose prediction applied to ibrutinib plus venetoclax treatment of CLL
	Abstract
	Introduction
	Materials and methods
	Patient material and ethical considerations
	Reagents and antibodies
	Isolation of B lymphocytes
	Phospho flow with fluorescent cell barcoding
	CellTiter-Glo luminescent cell viability assay
	Synergy analyses and statistics
	Results
	Titration of idelalisib, ibrutinib, and venetoclax
	Ibrutinib, but not idelalisib or venetoclax, inhibits the phosphorylation of BCR proximal signaling molecules
	Novel drugs inhibit the PI3K-signaling pathway downstream of the BCR
	Correlation analysis of cell signaling and cell viability responses to ibrutinib plus venetoclax treatment
	Drug synergy analysis of an ibrutinib plus venetoclax combination matrix
	Ibrutinib plus venetoclax selectively and synergistically kill CLL cells
	Discussion
	Compliance with ethical standards
	ACKNOWLEDGMENTS
	References