Cytokine 48 (2009) 231–238

Contents lists available at ScienceDirect

Cytokine
journal homepage: www.elsevier.com/locate/issn/10434666

In vivo antitumor efficacy of interleukin-21 in combination with chemotherapeutics

Kresten Skak a,*, Henrik Søndergaard a, Klaus Stensgaard Frederiksen b, Eva Ehrnrooth c
a
Immunopharmacology, Building F6.2.30, Novo Nordisk A/S, Novo Nordisk Park, 2760 Måløv, Denmark
b
Molecular Genetics, Novo Nordisk A/S, Novo Nordisk Park, 2760 Måløv, Denmark
c
Clinical Research Oncology, Novo Alle, Novo Nordisk A/S, 2880 Bagsv!rd, Denmark

a r t i c l e i n f o a b s t r a c t

Article history: Interleukin-21 (IL-21) is a class I cytokine with antitumor properties due to enhanced proliferation and
Received 16 January 2009 effector function of CD8+ T cells and natural killer (NK) cells. Here we have explored the magnitude
Received in revised form 27 May 2009 and time-course of cytostatics-induced lymphopenia in mice and investigated whether treatment with
Accepted 27 July 2009
cytostatics influences the antitumor effect of IL-21 in mouse tumor models. We show that pegylated lipo-
somal doxorubicin (PLD), irinotecan and oxaliplatin induced transient lymphopenia, whereas 5-fluoro-
uracil (5-FU) transiently increased lymphocyte counts. B cells were more sensitive than T cells towards
Keywords:
irinotecan and oxaliplatin. Additive antitumor effects were observed after combining IL-21 with PLD, oxa-
Immunotherapy
Cytokines
liplatin and to less extent 5-FU but not irinotecan, and larger effect was observed when IL-21 administra-
Chemotherapy tion was postponed relative to chemotherapy, suggesting that these agents may transiently impair
In vivo models immune function. However, the chemotherapies did not significantly alter the levels of circulating regu-
latory T cells and only marginally affected the ability of CD8+ T cells to respond to IL-21 measured as
increased granzyme B mRNA. Our results show that IL-21 therapy can be successfully combined with
agents from different chemotherapeutic drug classes, i.e. topoisomerase II inhibitors (PLD), anti-metab-
olites (5-FU) and platinum analogs (oxaliplatin) provided that IL-21 therapy is delayed relative to
chemotherapy.
! 2009 Elsevier Ltd. All rights reserved.

1. Introduction
Interleukin-21 (IL-21) is a novel four helix bundle class I cyto-
kine produced endogenously by activated CD4+ T cells and natural
killer T (NKT) cells [1,2]. IL-21 signals through a heterodimeric
receptor complex comprised of the unique IL-21 receptor protein
(IL-21R) and the common c chain. IL-21R is predominantly ex-
pressed by lymphocytes and dendritic cells, and in common with
other common c chain associated receptors, IL-21R signals primar-
ily through the transcriptional JAK/STAT activators pathway [3–5].
IL-21 has profound effects on most lymphocyte subsets, including
costimulation of B cell proliferation, differentiation and isotype
switching, enhancement of CD8+ T cell proliferation and effector
function, activation of NK cells and differentiation of CD4+ T cells
into Th17 cells (reviewed in [6]). In vivo studies using recombinant
murine IL-21 have demonstrated significant antitumor efficacy in a
variety of tumor models including melanoma and renal cell carci-
noma (RCC) [7,8], and studies in mice have identified cytotoxic T
cell and natural killer (NK) cell functions as likely mediators of
IL-21-induced antitumor activity [7,9–11]; a finding that is consis-
* Corresponding author. Tel.: +45 4443 1594; fax: +45 4443 4537.
E-mail address: skak@novonordisk.com (K. Skak).
tent with the ability of IL-21 to enhance effector functions of these
cell types in vitro and augment proliferation of cytotoxic T cells.
Clinical studies with recombinant human IL-21 administered as
monotherapy have been conducted in patients with stage IV mela-
noma and renal cell carcinoma. IL-21 demonstrated an acceptable
safety profile and was associated with antitumor activity when
administered as monotherapy in previously pretreated patients in
phase 1 and 2 trials [12–14]. IL-21 is currently being tested in a phase
II trial in combination with sorafenib in RCC (clinicaltrials.gov
identifier NCT00389285) and in phase I trials in combination with
rituximab in non-Hodgkin’s lymphoma (clinicaltrials.gov identifier
NCT00347971), with cetuximab in colorectal cancer (Eudract
No. 2006-004231-30) and with pegylated liposomal doxorubicin
(PLD) in epithelial ovarian cancer (clinicaltrials.gov identifier
NCT00523380) are also ongoing. Combining IL-21 with other drugs
is the next obvious step in the future development of IL-21, and since
the majority of cancer patients are undergoing chemotherapy at some
stage during their disease course it is highly relevant to investigate if
IL-21 therapy can be successfully combined with chemotherapy.
Traditionally combining chemotherapy with immunotherapy
has not been considered feasible due to the adverse effect of most
cytostatics on the bone marrow and on dividing lymphocytes,
which are required for raising an adaptive antitumor immune re-
sponse. However, more recently several studies have shown that

1043-4666/$ - see front matter ! 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.cyto.2009.07.039
232 K. Skak et al. / Cytokine 48 (2009) 231–238
cytostatics can as well augment tumor immunity due to their abil-
ity to induce immunogenic cell death, which may subsequently
lead to pro-inflammatory responses at the tumor site, uptake of tu-
mor antigen by resident antigen-presenting cells and subsequent
augmentation of antitumor T cell responses [15]. In vitro results
indicate that adjuvant IL-2 based immunotherapy combines effec-
tively with a number of cytostatics e.g. 5-fluorouracil (5-FU)1 and
doxorubicin [16,17]. Similarly, 5-FU permitted efficient CTL-medi-
ated killing of target cells in vitro [16]. In this study we have inves-
tigated how administration of chemotherapy affects the number of
circulating lymphocyte subsets in a time-course study and the abil-
ity of CD8+ T cells to respond to IL-21 stimulation in vivo measured
as increased granzyme B mRNA expression. Furthermore, we have
addressed whether the antitumor effect of IL-21 can be maintained
when chemotherapy is administered prior to or simultaneous with
IL-21 therapy. To this end we chose four different chemotherapeu-
tics, i.e. the slow release topoisomerase II inibitor PLD, the third gen-
eration platinum analog oxaliplatin, the topoisomerase I inhibitor
irinotecan, and the anti-metabolite 5-FU which is used in combina-
tion with leucovorin and either oxaliplatin or irinotecan. These che-
motherapeutics are used either as monotherapy or in different
combinations for the treatment of various cancers e.g. gastrointesti-
nal, breast, bladder and ovarian cancers. In two different mouse tu-
mor models we have addressed whether IL-21 can be successfully
combined with any of these chemotherapeutic compounds.
2. Materials and methods
2.1. Animals
C57BL/6 and Balb/c female mice, 8–10 weeks-old, were pur-
chased from Taconics. All experiments were conducted in accor-
dance with local guidelines on use of experimental animals and
approved by the Danish National Ethics Committee on Experimen-
tal Animals.
2.2. Cell culture
C57BL/6 derived B16 (F0) melanoma cells (American Type Cul-
ture Collection (ATCC), CRL-6322) and BALB/c derived RenCa renal
cell carcinoma cells (kindly provided by Dr. Robert H. Wiltrout, NCI
at Frederick, MD, USA) were cultured in RPMI 1640 with GlutaMAX
supplemented with 10% heat-inactivated FCS, sodium pyruvate
(RenCa only), nonessential amino acids (RenCa only), and 5% pen-
icillin–streptomycin (all from GIBCO Cell Culture, Invitrogen,
Denmark).
2.3. Reagents
Recombinant murine IL-21 produced by Novo Nordisk A/S was
used throughout all experiments. The test substance was formu-
lated as 10.5 mg/ml IL-21 in 0.16% L-histidine + 4.66% D-mannitol,
pH 5.1. Before injection IL-21 was diluted in PBS. PBS was used
as vehicle control for IL-21.
PLD (Schering-Plough), irinotecan (Mayne Pharma), 5-FU (May-
ne Pharma) and oxaliplatin (Mayne Pharma) were all purchased
through Nomeco (Copenhagen, Denmark).
2.4. Chemotherapy-induced lymphocytopenia in mice
PLD, and oxaliplatin were administered to C57BL/6 or Balb/c
mice as indicated in the figure legends by intravenous injection
1
Abbreviations used: 5-FU, 5-fluorouracil; GZMB, granzyme B; PLD, pegylated
liposomal doxorubicin; RCC, renal cell carcinoma.
once on day 0. Irinotecan and vehicle were administered intraper-
itoneally once on day 0. 5-Fluorouracil (5-FU) was administered
intraperitoneally twice on day 0 and 1. All cytostatics were admin-
istered using a dose close to MTD for mice as the highest dose, or if
the MTD was not available the dose was based on published papers
using these compounds in mice. Blood samples of up to 100 ll
were drawn from the periorbital plexus under isoflurane anaesthe-
sia into heparinized tubes at the following time points: pre-dosing,
24 h, 72 h, 7 days, 14 days, 21 days and 28 days post-dosing. Two
alternating cohorts of 4 mice each were used for each group in or-
der to reduce blood sampling from individual mice to maximum
two times within one week. From each blood sample 50 ll blood
was transferred to True-count tubes (BD Biosciences) and stained
for 15–20 min at RT with the antibodies TCR-b PE-CY5.5 (clone
H97–597, eBioscience) CD4 PE (clone GK1.5, BD Biosciences),
CD8 APC (clone 56–6.7, BD Biosciences), CD19 APC-Cy7 (clone
1D3, BD Biosciences) and NK1.1 FITC (clone PK136, BD Biosci-
ences). To each tube 1 ml FACS lyzing solution (BD Biosciences)
was added, and the samples were incubated for 15 min at RT and
transferred to ice. For enumeration of granulocyte and monocyte
20 ll whole blood was stained with CD11b APC-Cy7 (clone M1/
70, BD Biosciences), CD16/32 FITC (clone 2.4G2, BD Biosciences),
CD4 PE and NK1.1 in true-count tubes as above. Granulocytes
and monocytes were defined as CD11b+CD16/32+NK1.1!CD4!
combined with FSC/SSC gating. For enumeration of Treg cells
80 ll erythrocytes were lyzed in whole blood samples, and the
samples were stained with TCR-b PE-Cy5.5 (clone H57–597, eBio-
science), CD4 Pacific Orange (clone RM4–5, Caltag) and CD25 FITC
(clone PC61, eBioscience) followed by fixation/permeabilization
and staining with Foxp3 PE (clone FJP-16, eBioscience) according
to the manufacturer’s recommendations. Data acquisition was per-
formed on a LSR II flow cytometer (BD Biosciences) and data were
analyzed with BD FACSDiva software (BD Biosciences) for the abso-
lute number of lymphocytes. Data are normalized to a vehicle-trea-
ted group in order to compensate for day-to-day variations.
2.5. Effect of chemotherapy on IL-21 induced granzyme B
Cohorts of 8 mice were injected on day 0 with PLD (200 lg/
mouse), oxaliplatin (5 mg/kg), irinotecan (150 mg/kg) or 5-FU
(50 mg/kg on two consecutive days) as described above. At time
points 24 h, 72 h, 7 days and 14 days half of the mice in each cohort
were injected s.c. with 50 lg IL-21, the other half were injected
with equi-volume PBS. An additional cohort of mice (N = 8) was in-
jected only with IL-21 or PBS on day 0 to serve as baseline for the
mice injected with chemotherapy. Four hours after IL-21 or PBS
injection the mice were sacrificed and spleens were removed for
single cell preparation followed by lysis of erythrocytes in ACK buf-
fer. CD8+ T cells were isolated from 2 " 107 splenocytes by positive
selection using magnetic beads according to the manufacturer
instruction (Miltenyi, Bergisch Gladbach, Germany). The purity
was generally 80–90%.
Sorted CD8+ T cell samples were lyzed in TRIzol reagent (Invitro-
gen, Carlsbad, CA) and chloroform added. Samples were vortexed for
15 s and phases separated by centrifugation for 15 min at 10,000g at
4 "C. Total RNA from the aqueous phase was further purified using
the RNeasy 96 Qiagen vacuum system following the ‘‘Clean-up” pro-
cedure. RNA integrity was confirmed on a 2100 Bioanalyzer (Agilent
Technologies, Palo Alto, CA) using the RNA 6000 Pico LabChip kit
(Agilent Technologies). cDNA was prepared from total RNA using
random hexamer primers and TaqMan Reverse Transcription re-
agents (Applied Biosystems, Foster City, CA, USA) according to the
manufacturer instructions. Quantitative PCR was performed in
duplicate on each of the cDNA samples (10" dilutions of cDNA) using
TaqMan Universal PCR Master Mix (Applied Biosystems) and the ABI
PRISM# 7900HT Sequence Detection System (Applied Biosystems).
 K. Skak et al. / Cytokine 48 (2009) 231–238 233

Primers and FAM-labeled-probes for murine granzyme B mRNA and
18S rRNA were ordered as Assays-on-Demand (Applied Biosystems).
Probe sequences for these assays were: granzyme B (GZMB)
(CCAGGACAAAGGCAGGGGAGATCAT (assay Mm00442834_m1)),
and 18S rRNA (TGGAGGGCAAGTCTGGTGCCAGCAG; (assay Hs999
99901_s1)). Data were analyzed using ABI Prism SDS 2.2 software
(Applied Biosystems), and expression levels for GZMB were normal-
ized to 18S rRNA levels.
2.6. Tumor experiments
RenCa renal cell carcinoma cells (2 " 105 per mouse) or B16
melanoma cells (2 " 105 per mouse) were injected subcutaneously
in the right flank of syngenic mice, i.e. Balb/c mice for RenCa cells
and C57BL/6 mice for B16 cells. Tumor growth was monitored by
measuring the longest diameter and the perpendicular diameter
with at digital calliper. Tumor volume was calculated according
to the following formula:
Volume = (longest diameter) " (perpendicular diameter)2/2.
When tumors were visible the mice were allocated to 5 treat-
ment groups: vehicle, IL-21 ‘‘early”, cytostatic, cytostatic + IL-21
‘‘early” and cytostatic + IL-21 ‘‘late”. Mice were allocated according
to tumor size to obtain as similar as possible average tumor size in
all groups. Unless otherwise indicated chemotherapy was initiated
when tumors were small (10–50 mm3) in order obtain maximal ef-
fect of IL-21, which has most effect on small tumors [7]. Murine IL-
21 was administered subcutaneously on 5 consecutive days start-
ing on the day of chemotherapy (‘‘early”) or 7 days after chemo-
therapy (‘‘late”). The exact scheduling of therapy is indicated in
the figure legend for each experiment.
The tumor size was measured three times weekly. Tumor size
and Kaplan–Meyer survival analysis of time to reach tumor si-
ze = 750 mm3 were used as endpoints. This size was chosen as
cut-off value as many tumors had a tendency to ulcerate when
they became larger. Mean growth curves were calculated until
more than 20% of the mice in each group had exceeded maximum
allowed tumor size (1000 mm3). Since mice with tumors exceeding
1000 mm3 had to be terminated according to our guidelines the
last observation was carried forward for mice terminated due to
tumor size exceeding 1000 mm3 before the analysis of mean tumor
size was carried out.
2.7. Statistics
A two-sided, unpaired, equal variance Student’s t test was used to
compare the mean tumor size between groups and to compare cell
counts between mice treated with vehicle and test compound. The
log-rank test was used to compare Kaplan–Meyer survival curves.
3. Results
3.1. Chemotherapy-induced lymphocytopenia in mice
The effect of cytostatics injection on the number of circulating
blood lymphocytes was measured by flow cytometry on blood
samples drawn from mice treated with a panel of cytostatics. The
absolute number of CD4+ T cells (defined as TCR-b+CD4+), CD8+ T
cells (defined as TCR-b+CD8+), B cells (defined as TCR-b!CD19+)
and NK cells (defined as TCR-b!NK1.1+) were measured at defined
time points. In the section below the results are presented sepa-
rately for each drug. IL-21 administration up to 50 lg/mouse did
not have any measurable effect on the above lymphocyte subsets
at 24 h up to 7 days after administration (unpublished data). We
have therefore administered the chemotherapies only as mono-
therapy assuming that IL-21 would not substantially change
counts when given in combination with chemotherapy.
3.1.1. Pegylated liposomal doxorubicin (PLD)
PLD is a pegylated liposomal formulation of doxorubicin used
among others for the treatment of ovarian cancer. In C57Bl/6 mice
a single dose of 200 lg/mouse (approximately 8 mg/kg) resulted in
a decrease to 60–70% of controls in the circulating CD4+ and CD8+ T
cells on day 3 (CD8: p < 0.01 and CD4: p = 0.064 vs vehicle) with
full recovery on day 7. B cells and NK cells decreased to 40–50%
of controls (p < 0.001 vs vehicle on day 3, 7 and 14 [B cells] and
day 1 and 3 [NK cells]) but returned to baseline after 2–3 weeks
(Fig. 1b). At 30 lg/mouse (approximately 1.2 mg/kg), only a minor,
non-significant decrease in lymphocyte counts was observed

a PLD, 30 µg/mouse
b PLD, 200 µg/mouse
c Irinotecan, 150 mg/kg
d 5-FU, 50 mg/kg
140% 140% 140% 250%
120% 120% 120%
200%
5-FU
% of vehicle

100% 100% 100%

C57Bl/6

80% 80% 80% 150%

% of vehicle

60% 60% 60% 100%

40% 40% 40%
50%
20% 20% 20%
n=7 n=7 n=7 n=4
0% 0% 0% 0%
0 7 14 21 28 0 7 14 21 28 0 7 14 21 28 0 7 14 21 28
180% 180% 180% Day
Day Day Day
160% 160% 160%
Legend:
140% 140% 140%
CD19
% of vehicle

120% 120% 120%

Balb/c

100% 100% 100% CD4

80% 80% 80% CD8
60% 60% 60%
NK1.1
40% 40% 40%
20% 20% 20%
n=3 n=3 n=3
0% 0% 0%
0 7 14 21 28 0 7 14 21 28 0 7 14 21 28
Day Day Day

Fig. 1. Transient lymphopenia induced by PLD, irinotecan and 5-FU. Vehicle (isotonic NaCl) or 30 lg/mouse PLD (a), 200 lg/mouse PLD (b), 150 mg/kg irinotecan (c) or
50 mg/kg 5-FU (d) was injected intravenously (PLD) or intraperitoneally (irinotecan and 5-FU) in C57Bl/6 mice (upper panel) and Balb/c (lower panel, only PLD and
irinotecan) mice on day 0. Injection of 5-FU was repeated on day 1. The absolute numbers of CD4+, CD8+, NK1.1+ and CD19+ lymphocytes were measured in the blood samples
drawn pre-dose and day 1, 3, 7, 14, 21 and 28 by flow cytometry. The data have been normalized to vehicle-treated mice. Error bars are SEM.
234 K. Skak et al. / Cytokine 48 (2009) 231–238

(Fig. 1a). The effect of PLD was compared between C57Bl/6 and
Balb/c mice. No difference between the strains was found at
200 lg/mouse. Due to the lack of the NK1.1 molecule in Balb/c
mice NK cells were not measured in this mouse strain.
3.1.2. Irinotecan
After a single injection of 150 mg/kg irinotecan into C57Bl/6
mice, lymphocyte counts decreased to 60% of controls on day 1
after dosing (p < 0.01 vs vehicle for all subsets) with full recovery
around day 3–7. The effect of irinotecan was more profound in
Balb/c mice (Fig. 1c) leading to lower counts (28–48% of controls,
p < 0.05 vs vehicle on day 3 for all subsets and day 1 for B cells
only) and slower recovery.
3.1.3. 5-FU
At 50 mg/kg given on two consecutive days a small (27–46%) drop
in all lymphocyte subsets was observed on day 3 (day 1 and 3:
p < 0.05 vs vehicle for B and NK cells), with rebound effect reaching
200% of controls on day 14 for B and T cells (p < 0.05 vs vehicle)
and 150% of controls for NK cells (non-significant) (Fig. 1d).
3.1.4. Oxaliplatin
At a dose of 5 mg/kg a decrease to 50–70% of controls was ob-
served for all subsets on day 7 (day 3 and 7: p < 0.05 vs vehicle
for all subsets), full recovery (B cells and NK cells) or partial recov-
ery (T cells) was observed on day 21 (Fig. 2a). No significant effect
was observed at 1 mg/kg except an increase in NK cells day 7
(p < 0.001 vs vehicle). At 10 mg/kg a much more profound lympho-
penia was observed. T cell numbers were decreased to 25–35% of
controls on day 1 (day 1: p < 0.01 and day 3: p < 0.001 vs vehicle
for all subsets) with partial recovery on day 7 (p < 0.05 vs vehicle
for all subsets). B cells and NK cells decreased to 6% of controls
at day 7, at which time point the mice had to be terminated due
to weight loss and toxicity induced by oxaliplatin.
3.2. Antitumor effect of IL-21 in mice treated with chemotherapy
Chemotherapy may interfere with immunotherapy due to the ad-
verse effect of many chemotherapies on the hematopoietic system
a 200% Oxaliplatin
180%
160%
140%
% of vehicle
and immune function. We wanted to test whether IL-21 treatment
would preserve its antitumor effect in mice treated with chemother-
apy. To this end we used the subcutaneous RenCa renal cell carci-
noma and B16 melanoma tumor models, as previous studies have
shown that IL-21 has antitumor effect in both these syngeneic
mouse tumor models [7]. We tested the effect of IL-21 therapy in
combination with PLD and irinotecan in both RenCa and B16 models,
and additionally in combination with oxaliplatin and 5-FU in the Re-
nCa model. Our results above show that for all chemotherapies ex-
cept oxaliplatin circulating T lymphocyte numbers were
normalized 7 days after chemotherapy treatment. Since timing of
immunotherapy relative to the chemotherapy may be crucial for
success, the possible additive effect of IL-21 and chemotherapy
was tested in two different schedules, where treatment with IL-21
and chemotherapy were initiated either on the same day (‘early’)
or IL-21 therapy was initiated 7 days after chemotherapy (‘late’). Be-
low the results are presented separately for each compound.
3.2.1. PLD
Since this compound has a potent antitumor effect against Re-
nCa tumors, a dose-titration was initially performed to establish
a dose where PLD would lead to partial inhibition of tumor growth.
A single injection of PLD reduced growth of RenCa tumors dose-
dependently, and treatment at small tumor size was more efficient
than treatment at larger tumor size (Fig. 3a). The highest dose
(200 lg/mouse) led to partial regression and slow recovery,
whereas the lower doses tested (20 and 50 lg/mouse) led to partial
reduction in growth rate without any tumor regression. Based on
these results 30 lg/mouse of PLD was chosen in the following
study to give partial tumor inhibition.
The combination of IL-21 and PLD (30 lg/mouse) was tested in
the RenCa tumor model. Compared to PLD as single therapy, a signif-
icant additive antitumor effect was obtained when IL-21 therapy was
initiated one week after PLD injection (Fig. 3b). A small, non-signifi-
cant reduction in tumor growth was observed when IL-21 therapy
was initiated on the same day as PLD injection. A similar study was
subsequently performed in the B16 model. Since these tumors are
much more resistant to PLD therapy than RenCa tumors (data not
shown) the mice were treated with 200 lg/mouse PLD. This study
showed a similar additive effect of IL-21 in the B16 model as in the
RenCa model, with the ‘late’ IL-21 treatment being superior to ‘early’
120% IL-21 treatment (Fig. 3c). Thus, we conclude that IL-21 and PLD has an
10 mg/kg additive antitumor effect in both B16 and RenCa models.
100%
80%

120% 3.2.2. Irinotecan

% of vehicle

100% 60% A pilot study showed that only administration of irinotecan at

80%
40% the MTD (150 mg/kg) resulted in a minor reduction in tumor
60%
40% growth (Supplementary Fig. 1a). This dose was selected for the
20%
20%
n=4 n =4 combination studies.
1 mg/kg
0% 0% Two combination studies were performed in the RenCa model
0 7 14 21 28 0 7 14 21 28
Day
and one study in the B16 model. In all experiments the mice re-
180% ceived a single injection of 150 mg/kg irinotecan. Whereas IL-21
160%
Legend: alone reduced the growth of both RenCa and B16 tumors, irinotec-
140%
CD19 an given in combination with IL-21 did not inhibit tumor growth
% of vehicle

120%
100% CD4 compared to either compound alone; only irinotecan + IL-21 ‘late’
80% CD8 had a small, non-significant effect on tumor growth (Fig. 4). Thus,
60%
NK1.1 there was no additive antitumor effect of irinotecan and IL-21 in
40% the B16 and RenCa models.
20%
5 mg/kg n=4
0%
0 7 14 21 28 3.2.3. Oxaliplatin
Day A titration study was performed to establish the optimal dose of
oxaliplatin in Balb/c mice (data not shown). Based on this study
Fig. 2. Oxaliplatin induces transient, dose-dependent lymphopenia. Vehicle (iso- 5 mg/kg oxaliplatin was established as MTD in mice and chosen
tonic NaCl) or oxaliplatin was injected intravenously in C57Bl/6 mice on day 0. The
absolute numbers of CD4+, CD8+, NK1.1+ and CD19+ lymphocytes were measured in
for combination studies with IL-21. The combination of oxaliplatin
the blood samples drawn pre-dose and day 1, 3, 7, 14, 21 and 28 by flow cytometry. and IL-21 was performed in the RenCa tumor model. In this exper-
The data have been normalized to vehicle-treated mice. Error bars are SEM. iment both combination regimens reduced tumor growth com-
 K. Skak et al. / Cytokine 48 (2009) 231–238 235

1000
a Control

Tumor volume (mm3)

750 20 µg PLD d9
50 µg PLD d9
100 µg PLD d9
500 100 µg PLD d16
PLD 200 µg PLD d16
PLD
250

0
0 5 10 15 20 25 30 35 40 45
Days after tumor inoculation

1000

% mice with tumors < 750 mm3

100
b Vehicle
PLD + PBS
Tumor volume (mm3)

80
750
Vehicle + mIL-21 d10-14
60 PLD + mIL-21 d10-14
500 PLD + mIL-21 d17-21
PLD 40
250
20
**
0 0
0 10 20 30 40 50 0 20 40 60 80
Days after tumor inoculation Days after tumor inoculation

1500
% mice with tumors < 750 mm3

100

1250
c Vehicle
Tumor volume (mm3)

PLD + PBS
75
1000 Vehicle + mIL-21 d4-8
PLD + mIL-21 d4-8
750 50 PLD + mIL-21 d11-15
500 PLD
25
250 * #
* *
0 * 0
0 5 10 15 20 25 0 10 20 30 40 50
Days after tumor inoculation Days after tumor inoculation

Fig. 3. Additive antitumor effect of PLD and IL-21 in the RenCa model. (a) RenCa cells were injected s.c. on day 0. PLD was injected intravenously at the indicated doses on day
9 or 16. N = 5. (b) RenCa cells were injected s.c. on day 0. PLD (30 lg/mouse) or vehicle was injected intravenously on day 10. Murine IL-21 (50 lg/injection) or PBS was
administered s.c. daily for 5 days on the days shown. N = 11. (c) B16 cells were injected s.c. on day 0. PLD (200 lg/mouse) or vehicle was injected intravenously on day 4.
Murine IL-21 (50 lg/injection) or PBS was administered s.c. daily for 5 days on the days shown. N = 10. Tumors were measured thrice weekly and the volume was calculated.
The charts to the left show mean tumor size, whereas the charts to the right show Kaplan–Meyer survival analysis of mice with tumors <750 mm3. *p < 0.05 vs PLD only by
Student’s t test. #p < 0.05 vs PLD by Log-rank test. Error bars are SEM.

pared to oxaliplatin alone, with the largest effect observed when gate whether treatment with chemotherapeutics or IL-21 affect Treg
IL-21 therapy was initiated one week after oxaliplatin injection numbers we injected these drugs in mice and subsequently moni-
(Fig. 5). Thus, we conclude that oxaliplatin and IL-21 has additive tored the circulating numbers of Treg cells over a 4 week period. A
antitumor effect in the RenCa model. minor, non-significant decrease in CD4+CD25+Foxp3+ Treg cells was
observed around day 7 after injection of the chemotherapies, consis-
3.2.4. 5-FU tent with a general decrease in CD4+ T cells, (Fig. 7a). IL-21 did not af-
In the clinic 5-FU is normally administered as a bolus followed fect the number of circulating Treg cells (Fig. 7a) nor did it affect the
by 48 h infusion. In order to mimic this regimen we administered numbers of CD4+ and CD8+ T cells in the blood (data not shown).
5-FU as a once daily injection on two consecutive days. A dose- In the same experiment we also monitored circulating granulo-
titration of 5-FU in the RenCa model showed a small antitumor ef- cytes and monocytes. Here, oxaliplatin treatment resulted in 3-fold
fect at the highest dose tested, 50 mg/kg per injection (Supplemen- increase in neutrophils on day 7 whereas 5-FU led to profound neu-
tary Fig. 1b). This dose was used in the following combination tropenia on day 7 leading to subsequent recovery day 14 slightly
study in the RenCa model. For technical reasons the group treated above baseline. PLD led to 70–90% increase in granulocyte numbers
with IL-21 alone was not included in this experiment. In the 5- from day 7 (non-significant) whereas IL-21 did not affect granulocyte
FU + IL-21 ‘late’ group a reduction in tumor size up to 75% com- numbers (Fig. 7b). Both oxaliplatin and 5-FU led to an initial reduc-
pared to 5-FU alone was observed (Fig. 6). In the Kaplan–Meyer tion in monocytes which subsequently recovered beyond baseline.
survival analysis of time to reach tumor volume = 750 mm3, 5- PLD led to an approximately 3-fold increase in monocyte numbers
FU + IL-21 ‘late’ therapy significantly increased ‘survival’ time com- from day 7 whereas IL-21 did not affect monocyte numbers (Fig. 7c).
pared to 5-FU alone, whereas the difference in tumor volume was A comparison between spleen and blood on day 7 showed that
non-significant. It should be noted that in this experiment varia- the relative effect of the drugs compared to vehicle was similar in
tion was high and the group size was small (N = 7–10 mice). spleen and blood (data not shown).

3.3. Effect of chemotherapeutics and IL-21 on regulatory T cells 3.4. IL-21 mediated upregulation of granzyme B mRNA in CD8+ T-cells

The additive effect of oxaliplatin, PLD and 5-FU combined with IL- To examine whether the chemotherapeutic compounds sup-
21 could be due to a decrease in regulatory T cells (Tregs). To investi- pressed IL-21 mediated effects in a relevant immune-cell subset,
236 K. Skak et al. / Cytokine 48 (2009) 231–238

700
RenCa tumors and clearly upregulated after 4 h in all of the IL-21 dosed groups
600 Vehicle (on average #8-fold upregulation). When comparing the GZMB
Tumor volume (mm3)

Irinotecan + PBS mRNA levels in the IL-21 dosed groups, there were generally no
500
Vehicle + mIL-21 d11-15
400 Irinotecan + mIL-21 d11-15
significant differences between the control group (receiving no
300 Irinotecan + mIL-21 d15-21 chemotherapy), and the chemotherapeutics dosed groups. The only
Irinotecan two groups which were found to be marginally suppressed was the
200
irinotecan dosed group at day 3 and the 5-FU dosed group at day 1
100
(p < 0.05 with no correction for multiple testing). In the PLD dosed
0
0 5 10 15 20 25 30 group a trend towards increasing responses to IL-21 was observed
Days after tumor inoculation at day 7 and 14 after compound injection.

1400
B16 tumors
1200 Vehicle 4. Discussion
Tumor volume (mm3)

1000
800
600
Irinotecan
400
200
0
0 5 10
Fig. 4. No additive antitumor effect of irinotecan and IL-21. (a) RenCa cells were
injected s.c. on day 0. Irinotecan (150 mg/kg) or vehicle was injected intraperito-
neally on day 11. Murine IL-21 (50 lg/injection) or PBS was administered s.c. daily
for 5 days as indicated. N = 12. The data shown are representative of two
independent experiments. (b) B16 cells were injected s.c. on day 0. Irinotecan
(150 mg/kg) or vehicle was injected intravenously on day 3. Murine IL-21 (50 lg/
injection) or PBS was administered s.c. daily as indicated. N = 10–12. The charts
show the mean tumor size. Error bars are SEM.
we analyzed the mRNA expression of the cytotoxic effector mole-
cule granzyme B (GZMB) in MACS-sorted CD8+ T cells from spleens.
Mice were dosed with 5-FU, oxaliplatin, irinotecan, or PLD on day
0, and IL-21 or vehicle was injected at day 1, 3, 7 or 14. The effect of
IL-21 dosing was analyzed 4 h after injection of IL-21 (Fig. 8). The
GZMB mRNA levels were very low in the PBS injected groups,
Irinotecan + PBS
Vehicle + mIL-21 d3-6
Irinotecan + mIL-21 d3-6+11-14
Chemotherapy has a long history in the treatment of cancer, but
Irinotecan + mIL-21 d13-17 so far it is only in a few indications that these agents can cure can-
cer. Immunotherapy is another approach that has proven success-
ful in some clinical trials, but is still only used in certain narrow
indications. Combining immunotherapy with chemotherapy has
15 20 usually been considered unfeasible due to the known suppressive
effects of many chemotherapies on the bone marrow, although, a
growing body of evidence is emerging that these modalities may
complement rather than antagonize each other [15,18].
In this study we demonstrate an additive antitumor effect
in vivo of IL-21 in combination with chemotherapeutic agents with
different mechanisms of action, i.e. topoisomerase II inhibiting
drugs (PLD), anti-metabolites (5-FU) and platinum analogs (oxa-
liplatin) provided that IL-21 therapy is delayed relative to chemo-
therapy. To investigate the possible antitumor effect of combining
IL-21 therapy with chemotherapy we used the subcutaneous Re-
nCa renal cell carcinoma and B16 melanoma tumor models. These
cell lines were chosen since they have both proven to be respon-
sive to IL-21 therapy in vivo [7]. Generally, IL-21 dosing initiated
7 days after chemotherapy showed improved antitumor efficacy
compared to IL-21 therapy initiated simultaneously with chemo-
therapy. The antitumor activity of IL-21 has generally been as-

100
% mice with tumors < 750 mm 3

800 Vehicle
Oxaliplatin + PBS
Tumor volume (mm 3)

75 Vehicle + mIL-21 d8-12

600
Oxaliplatin + mIL-21 d8-12
50 Oxaliplatin + mIL-21 d15-19
400
Oxaliplatin
200 25
* *
0 ** * ** ** 0
0 10 20 30 40 0 10 20 30 40 50 60 70
Days after tumor inoculation Days after tumor inoculation

Fig. 5. Additive antitumor effect of oxaliplatin and IL-21. RenCa cells were injected s.c. on day 0. Oxaliplatin (5 mg/kg) or vehicle was injected intravenously on day 8. Murine
IL-21 (50 lg/injection) or PBS was administered s.c. daily for 5 days on the days shown. The left chart shows mean tumor size. The right chart shows Kaplan–Meyer survival
analysis of mice with tumors <750 mm3. N = 9. *p < 0.05 and **p < 0.01 vs oxaliplatin alone by Student’s t test. Error bars are SEM.

100
% of mice withtumors < 750 mm3

800
Vehicle
80 5-FU + PBS
Tumor volume (mm3)

600 5-FU+ mIL-21 d 11-15

60 5-FU+ mIL-21 d 18-22
400
40 #
5-FU
200 20

0 0
0 10 20 30 40 0 10 20 30 40 50 60 70 80
Days after tumor inoculation Days after tumor inoculation

Fig. 6. Additive antitumor effect of 5-FU and IL-21. RenCa cells were injected s.c. on day 0. 5-FU (50 mg/kg) or vehicle was injected intraperitoneally on day 11 and day 12.
Murine IL-21 (50 lg/injection) or PBS was administered s.c. daily for 5 days as indicated. N = 7–10. The left chart shows mean tumor size. The right chart shows Kaplan–Meyer
survival analysis of mice with tumors <750 mm3. #p < 0.05 vs 5-FU alone by Log-rank test. Error bars are SEM.
 K. Skak et al. / Cytokine 48 (2009) 231–238 237

a 160% Treg
exception of oxaliplatin given at 10 mg/kg, which was discontin-
ued due to associated toxicity.
140% Legend: Our results suggest that the antitumor effect of chemotherapy
120% PLD, 30 µg/mouse and IL-21 is not related to their effect on monocytes and granulo-
% of vehicle

100% 5-FU, 50 mg/kg x 2 cytes. Thus on day 7, the optimal time point for combination ther-
80% Oxaliplatin, 5 mg/kg apy, 5-FU had opposite effect on monocytes and granulocytes
60% IL-21, 50 µg/mouse numbers compared to PLD and oxaliplatin, and IL-21 did not have
40% any effect at all. T cells generally appeared to recover faster from
20%
n=4
the treatment than B and NK cells, which may be important consid-
0% ering that the primary antitumor effect of IL-21 is through activa-
0 7 14 21 28
tion of CD8+ T cells. Treatment with 5-FU even caused B and T cell
b Granulocytes
c Monocytes
350% 450%
numbers to increase following an initial drop, suggesting that 5-FU
400% could have a beneficial effect on lymphocytes. Considering the
300%
*** 350% modest decrease in lymphocyte numbers observed after chemo-
% of vehicle

250%
300% * therapy it is not likely that the lymphopenia per se will confound
200% 250% * the effect of IL-21 therapy, especially if IL-21 is given after the na-
150% 200% ** * *** dir. Also, we did not observe a significant change in the number of
150%
100% circulating Tregs following chemotherapy or IL-21 injection, so the
100%
50% 50%
antitumor effect of the compounds is not likely to be caused by a
0% *** 0% ** * contraction of Tregs. Rather, it is possible that in the recovery
0 7 14 21 28 0 7 14 21 28 phase following the modest lymphopenia, homeostatic prolifera-
Day Day tion may produce memory-like T cells. This process may be rein-
forced by IL-21 therapy, leading to enhanced T cell activity, as
Fig. 7. Effect of chemotherapy on granulocytes and monocytes but not Tregs.
Vehicle (isotonic NaCl), PLD, oxaliplatin, 5-FU or IL-21 was injected in C57Bl/6 on demonstrated in a study where IL-21 was shown to enhance
day 0. The absolute numbers of CD4+CD25+Foxp3+ Tregs (a), granulocytes (b) and homeostatic proliferation leading to rapid T cell turnover and auto-
monocytes (c) were measured in the blood samples drawn pre-dose and on day 1, 3, immunity [20]. Thus, the ability of IL-21 to promote homeostatic
7, 14, 21 and 28 by flow cytometry. The data have been normalized to vehicle- proliferation may contribute to the antitumor activity of IL-21.
treated mice. *p < 0.05, **p < 0.01 and ***p < 0.001 vs vehicle by Student’s t test.
Error bars are SEM.
Establishing similar kinetics of chemotherapy-induced lymphope-
nia in humans may be important to clinically translate our results.
In contrast to these results, combining IL-21 with the topoiso-
Granzyme B mRNA in CD8+ cells merase I inhibitor irinotecan produced no additive effect or even
5
GZMB mRNA (18S normalized)

PBS an antagonizing effect of IL-21 even with postponed initiation of

4 IL21 IL-21, suggesting a negative impact of irinotecan on the target cells.
As with the other chemotherapies, T cell numbers were almost
3 back to normal 7 days after irinotecan injection; nevertheless iri-
2
notecan inhibited the antitumor effect of IL-21 therapy initiated
7 days after irinotecan injection. It is possible that the initial de-
1 cline in lymphocytes is reducing the responsiveness to IL-21, but
the fact that the antitumor effect of IL-21 was lost in mice treated
0
with irinotecan suggests that not only the number of circulating
14 (Oxaliplatin)

14 (Irinotecan)
1 (Oxaliplatin)
3 (Oxaliplatin)
7 (Oxaliplatin)

1 (Irinotecan)
3 (Irinotecan)
7 (Irinotecan)
14 (5-FU)

14 (PLD)
No Compound

1 (5-FU)
3 (5-FU)
7 (5-FU)

1 (PLD)
3 (PLD)
7 (PLD)

lymphocytes is important for the antitumor effect of IL-21. There-

fore, we also addressed whether the immune cells may be func-
tionally impaired after exposure to certain chemotherapeutics,
leading to a longer lasting unresponsiveness to IL-21 stimulation.
Days after injection of chemo (compound)
Previously, we have demonstrated that GZMB mRNA was upregu-
Fig. 8. No chemotherapeutic impairment of IL-21 mediated granzyme B (GZMB) lated in CD8+ T cells and CD56+ NK cells from metastatic melanoma
mRNA upregulation in CD8+ cells. GZMB mRNA was quantified by qRT-PCR in CD8+ patients administered with recombinant human IL-21 [12,21], and
cells from mice (N = 4) 4 h after s.c. administration with either PBS or murine IL-21 hence used this biologically relevant marker in the present study to
(50 lg/injection). The effect of IL-21 dosing on GZMB mRNA levels was analyzed at
examine whether the chemotherapeutic compounds suppressed
day 1, 3, 7, and 14 after injection of either 5-FU, oxaliplatin, irinotecan, or PLD. A
group receiving no chemotherapeutic compounds was included as a reference. All IL-21-induced activation of CD8+ T cells. The strong IL-21 mediated
GZMB levels were normalized to 18S, and mean values with SEM error bars are upregulations of GZMB mRNA in CD8+ T cells observed in all treat-
shown. ment groups suggests that the ability of the CD8+ T-lymphocytes to
respond to IL-21 was intact after administration of chemotherapy.
Thus, our observed differences in antitumor activity of the various
cribed to costimulation and expansion of antigen-specific T cells as combinations of IL-21 and chemotherapy must be due to other
well as costimulation of NK cell activity [8,11,19]. A possible expla- parameters. Obviously, there might still be differences in the spe-
nation for our findings is the fact that chemotherapy induces short cific immunomodulatory effects of the chemotherapeutics used
term immune suppression which may be harmful to the effect of here, which have to be determined.
IL-21, whereas at a later stage the immune system may recover, Differences in the immunogenicity of dying tumor cells induced
and dying tumor cells may facilitate presentation of tumor anti- by the various chemotherapeutics could be another important fac-
gens and activation of phagocytes and DCs leading to augmented tor [22]. Here, doxorubicin, the active component of PLD, has been
antitumor T cell responses, which can be enhanced by IL-21 ther- thoroughly studied and showed superior induction of immuno-
apy. This explanation is supported by our differential lymphocyte genic cell death when compared to a panel of distinct chemother-
counts, which showed varying degrees of lymphopenia after injec- apies [23]. Furthermore, tumor cell death induced by doxorubicin
tion of the chemotherapeutics but generally returned to normal or as well as oxaliplatin has also been shown to boost tumor immune
slightly above normal levels approx. 7 days post injection, with the responses via stimulation of TLR-4 [24]. In combination with IL-2,
238 K. Skak et al. / Cytokine 48 (2009) 231–238
doxorubicin also resulted in immune mediated cures of metastatic
murine breast cancer [25]. Thus, the dissimilar effect of different
chemotherapies on the immunogenicity of dying tumor cells may
provide an explanation to the different ability to provide additive
antitumor effects with IL-21, and furthermore this ability to induce
immunogenic cell death may to some extent counteract the ad-
verse effect of chemotherapy on immune cells and bone marrow.
The chemotherapeutics combined with IL-21 in this study are
used as monotherapy or in different combinations in the treatment
of various human solid tumors such as ovarian cancer and gastro-
intestinal malignancies, whereas IL-21 primarily is expected to be
used in indications with known immunotherapeutic success, i.e.
melanoma and RCC. Immune surveillance mechanisms have also
been shown to be active in selective patients with ovarian, colorec-
tal and esophageal cancer, and profoundly affected the natural
course and clinical outcome of these diseases [26–30]. Thus, these
results suggest that other indications may also be amenable to IL-
21-mediated immunotherapy and/or chemo-immunotherapy. A
phase 1 trial combining recombinant human IL-21 with liposomal
doxorubicin in ovarian cancer patients is currently ongoing (clini-
caltrials.gov identifier NCT00523380).
In summary, we show that IL-21 can be successfully combined
with certain chemotherapeutic agents without loosing its immune
stimulating and antitumor activity, supporting clinical trials of this
concept. Our data also highlight the importance of selecting the
proper time point for commencing IL-21 therapy after chemother-
apy, a notion which should be considered in future clinical trials.
Conflict of interest
K.S., H.S. and K.S.F. are employees of Novo Nordisk A/S who
sponsored this study and who has been involved in the develop-
ment of IL-21 as oncology drug.
Acknowledgements
We thank Ann-Christin Løff, Semra Esiyok, Ken Heding, Lene
Normann Nielsen, and Heidi Winther for technical assistance with
the experiments.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.cyto.2009.07.039.
References
[1] Parrish-Novak J, Dillon SR, Nelson A, Hammond A, Sprecher C, Gross JA, et al.
Interleukin 21 and its receptor are involved in NK cell expansion and
regulation of lymphocyte function. Nature 2000;408:57–63.
[2] Coquet JM, Kyparissoudis K, Pellicci DG, Besra G, Berzins SP, Smyth MJ, et al. IL-
21 is produced by NKT cells and modulates NKT cell activation and cytokine
production. J Immunol 2007;178:2827–34.
[3] Asao H, Okuyama C, Kumaki S, Ishii N, Tsuchiya S, Foster D, et al. Cutting edge:
the common gamma-chain is an indispensable subunit of the IL-21 receptor
complex. J Immunol 2001;167:1–5.
[4] Strengell M, Sareneva T, Foster D, Julkunen I, Matikainen S. IL-21 up-regulates
the expression of genes associated with innate immunity and Th1 response. J
Immunol 2002;169:3600–5.
[5] Brenne A, Ro TB, Waage A, Sundan A, Borset M, Hjorth-Hansen H. Interleukin-
21 is a growth and survival factor for human myeloma cells. Blood
2002;99:3756–62.
[6] Spolski R, Leonard WJ. Interleukin-21: basic biology and implications for
cancer and autoimmunity. Annu Rev Immunol 2008;26:57–79.
[7] Søndergaard H, Frederiksen KS, Thygesen P, Galsgaard ED, Skak K, Kristjansen
PE, et al. Interleukin 21 therapy increases the density of tumor infiltrating CD8+
T cells and inhibits the growth of syngeneic tumors. Cancer Immunol
Immunother 2007;56:1417–28.
[8] Zeng R, Spolski R, Finkelstein SE, Oh S, Kovanen PE, Hinrichs CS, et al. Synergy
of IL-21 and IL-15 in regulating CD8+ T cell expansion and function. J Exp Med
2005;201:139–48.
[9] Brady J, Hayakawa Y, Smyth MJ, Nutt SL. IL-21 induces the functional
maturation of murine NK cells. J Immunol 2004;172:2048–58.
[10] Wang G, Tschoi M, Spolski R, Lou Y, Ozaki K, Feng C, et al. In vivo antitumor
activity of interleukin 21 mediated by natural killer cells. Cancer Res
2003;63:9016–22.
[11] Moroz A, Eppolito C, Li Q, Tao J, Clegg CH, Shrikant PA. IL-21 enhances and
sustains CD8+ T cell responses to achieve durable tumor immunity:
comparative evaluation of IL-2, IL-15, and IL-21. J Immunol 2004;173:
900–9.
[12] Davis ID, Skrumsager BK, Cebon J, Nicholaou T, Barlow JW, Moller NPH,
et al. An open-label, two-arm, phase I trial of recombinant human
interleukin-21 in patients with metastatic melanoma. Clin Cancer Res
2007;13:3630–6.
[13] Thompson JA, Curti BD, Redman BG, Bhatia S, Weber JS, Agarwala SS, et al.
Phase I study of recombinant interleukin-21 in patients with metastatic
melanoma and renal cell carcinoma. J Clin Oncol 2008;26:2034–9.
[14] Davis ID, Brady B, Kefford RF, Millward M, Cebon J, Skrumsager BK, et al.
Clinical and biological efficacy of recombinant human interleukin-21 in
patients with stage IV malignant melanoma without prior treatment: a
phase IIa trial. Clin Cancer Res 2009;15:2123–9.
[15] Lake RA, Robinson BW. Immunotherapy and chemotherapy – a practical
partnership. Nat Rev Cancer 2005;5:397–405.
[16] Fujii S, Shimizu K, Smith C, Bonifaz L, Steinman RM. Activation of natural
killer T cells by alpha-galactosylceramide rapidly induces the full
maturation of dendritic cells in vivo and thereby acts as an adjuvant for
combined CD4 and CD8 T cell immunity to a coadministered protein. J Exp
Med 2003;198:267–79.
[17] Giaccone G, Punt CJ, Ando Y, Ruijter R, Nishi N, Peters M, et al. A phase I study
of the natural killer T-cell ligand alpha-galactosylceramide (KRN7000) in
patients with solid tumors. Clin Cancer Res 2002;8:3702–9.
[18] Zitvogel L, Apetoh L, Ghiringhelli F, Kroemer G. Immunological aspects of
cancer chemotherapy. Nat Rev Immunol 2008;8:59–73.
[19] Kasaian MT, Whitters MJ, Carter LL, Lowe LD, Jussif JM, Deng BJ, et al. IL-21
limits NK cell responses and promotes antigen-specific T cell activation: a
mediator of the transition from innate to adaptive immunity. Immunity
2002;16:559–69.
[20] King C, Ilic A, Koelsch K, Sarvetnick N. Homeostatic expansion of T cells during
immune insufficiency generates autoimmunity. Cell 2004;117:265–77.
[21] Frederiksen KS, Lundsgaard D, Freeman JA, Hughes SD, Holm TL, Skrumsager
BK, et al. IL-21 induces in vivo immune activation of NK cells and CD8(+) T cells
in patients with metastatic melanoma and renal cell carcinoma. Cancer
Immunol Immunother 2008;57:1439–49.
[22] Lake RA, van der Most RG. A better way for a cancer cell to die. N Engl J Med
2006;354:2503–4.
[23] Obeid M, Tesniere A, Ghiringhelli F, Fimia GM, Apetoh L, Perfettini JL, et al.
Calreticulin exposure dictates the immunogenicity of cancer cell death. Nat
Med 2007;13:54–61.
[24] Apetoh L, Ghiringhelli F, Tesniere A, Obeid M, Ortiz C, Criollo A, et al. Toll-like
receptor 4-dependent contribution of the immune system to anticancer
chemotherapy and radiotherapy. Nat Med 2007;13:1050–9.
[25] Ewens A, Luo L, Berleth E, Alderfer J, Wollman R, Hafeez BB, et al. Doxorubicin
plus interleukin-2 chemoimmunotherapy against breast cancer in mice.
Cancer Res 2006;66:5419–26.
[26] Coukos G, Conejo-Garcia JR, Roden RBS, Wu TC. Immunotherapy for
gynaecological malignancies. Expert Opin Biol Ther 2005;5:1193–210.
[27] Zhang L, Conejo-Garcia JR, Katsaros D, Gimotty PA, Massobrio M, Regnani G,
et al. Intratumoral T cells, recurrence, and survival in epithelial ovarian cancer.
N Engl J Med 2003;348:203–13.
[28] Naito Y, Saito K, Shiiba K, Ohuchi A, Saigenji K, Nagura H, et al. CD8+ T cells
infiltrated within cancer cell nests as a prognostic factor in human colorectal
cancer. Cancer Res 1998;58:3491–4.
[29] Galon J, Costes A, Sanchez-Cabo F, Kirilovsky A, Mlecnik B, Lagorce-Pages C,
et al. Type, density, and location of immune cells within human colorectal
tumors predict clinical outcome. Science 2006;313:1960–4.
[30] Schumacher K, Haensch W, Roefzaad C, Schlag PM. Prognostic significance of
activated CD8+ T cell infiltrations within esophageal carcinomas. Cancer Res
2001;61:3932–6.