| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
after vaccination with dendritic cells pulsed with MAGE peptides in patients with mage-A1/A3-positive tumors
,
* Unité de Thérapie Cellulaire et Moléculaire (U.T.C.M.),
Department of Medical Oncology,
Department of Dermatology,
Interdisciplinary Research Institute (IRIBHN), and
|| Department of Immuno-Hematology-Tranfusion, Erasme Hospital and Bordet Institute, Université Libre de Bruxelles, Brussels, Belgium
Correspondence: Dr. M. Toungouz, U.T.C.M., Erasme Hospital, 808 route de Lennik, B-1070 Brussels, Belgium. E-mail: toungouz{at}ulb.ac.be
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
secretion by cytokine flow
cytometry and ELISPOT. We also tested whether further KLH addition
could influence this response favorably. Monocyte-derived DC were
generated from leukapheresis products. They were pulsed with the
relevant MAGE peptide(s) alone in group A (n=10 pts) and
additionally with KLH in group B (n=16 pts). A specific but
transient increase in the number of peripheral blood T lymphocytes
secreting IFN- in response to the vaccine peptide(s) was observed in
6/8 patients of group A and in 6/16 patients of group B. We conclude
that anti-tumor vaccination using DC pulsed with MAGE peptides induces
a potent but transient anti-MAGE, IFN- secretion that is not
influenced by the additional delivery of a nonspecific, T-cell
help.
Key Words: T-cell response • keyhole limpet hemocyanin • ELISPOT • cytokine flow cytometry
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
.
Subsequently, reverse immunology allowed the identification of
tumor-derived peptides restricted by major histocompatibility complex
(MHC) class I molecules. The fact that peptide-MHC recognition by CD8+
CTL leads to tumor-cell lysis has prompted investigators to conceive
vaccination strategies based on the use of these peptides. Data from
murine models suggest that effective, anti-tumor responses including
the induction of an effector memory require the delivery of a T-cell
help [1
2
3
]. Peptide vaccination without the use of such
help has been associated with functional deletion of tumor-specific CTL
leading to a subsequent inability to reject tumors [4
].
This tolerization can be avoided in murine models by peptide
presentation on dendritic cells (DC), as opposed to other formulations
[5
]. Based on these data, we initiated a clinical trial
in a first group of patients with tumors expressing the MAGE-A1 and/or
-A3 antigens who were vaccinated with DC pulsed with HLA class
I-restricted, MAGE-derived peptides. In a second group of patients, we
evaluated the potential benefit of providing a further T-cell help by
pulsing DC with keyhole limpet hemocyanin (KLH). The primary goal of
this trial was the biological assessment of anti-tumor, immune
responses by cytokine flow cytometry (CFC) and enzyme-linked immunospot
(ELISPOT) of evaluating the frequency of T cells secreting IFN-
after in vitro restimulation with the peptide.
Herein, we show the kinetics of the response, the type of the
responding cells and on the influence of nonspecific T-cell help. |
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The first 10 patients evaluated in group A received DC pulsed with MAGE peptide(s) alone. In group B, 4 of the patients from group A and 12 additional patients were evaluated and received DC pulsed with MAGE peptide(s) and KLH (Table 1 ). For final evaluation, patients must have received at least three series of DC vaccination.
|
Clinical evaluation
The initial, clinical evaluation included a medical history; a
physical examination with evaluation of any potential tumor mass; a
complete blood count; blood chemistry and immunological analyses
including those for detecting auto-immune disorders; a bone scan; and
computed tomographies and/or magnetic resonance imagery of the chest,
abdomen, and any other appropriate localization. Evaluation was
performed after the fourth DC vaccination.
DC generation
Clinical-grade DC were generated in a closed system (Cell
Culture Container PL2417, Nexell Therapeutics, Irvine, CA) from
peripheral blood monocytes as previously described [6
].
Briefly, peripheral blood mononuclear cells (PBMC) were obtained by
leukapheresis, performed for each series of DC vaccination, on a COBE
SPECTRA separator (Cobe, Denver, CO). The monocytes-enriched fraction,
obtained after a modified-adherence step in the culture bag, was seeded
for 7 days in a serum-free medium (X-VIVO 20, Bio-Whittaker,
Walkersville, MD) supplemented with 800 U/ml granulocyte-macrophage
colony-stimulating factor (GM-CSF; Leukomax, Novartis, Basel,
Switzerland) and 1000 U/ml interleukin (IL)-4 (Schering-Plough,
Kenilworth, NJ). After collection, DC were pulsed for 2 h with the
relevant, clinical-grade peptide(s) [MAGE-A1.A1 (EADPTCHSY),
MAGE-A3.A2 (FLWGPRALV; UCB, Brussels, Belgium), MAGE-A3.A1 (EVDPIGHLY;
Clinalfa, Basel, Switzerland), MAGE-A3.B44 (MEVDPIGHLY; Peptisyntha,
Brussels, Belgium)] at 50 µg/ml and for patients of group B,
additionally pulsed with KLH (Biosyn, Fellbach, Germany) at 50 µg/ml.
The DC were then washed and resuspended in 1 ml phosphate-buffered
saline (PBS; Baxter, Fenwal Division, Deerfield, IL) supplemented with
5% human albumin (Belgian Red Cross, Brussels, Belgium) prior to
injection.
DC vaccination
Vaccinations with antigen-pulsed DC were performed every 3 weeks
(V1, V2, V3), and a fourth injection (V4) was performed 6 weeks later.
In case of clinical and/or biological response(s), additional DC
infusions were given every 2 months for 1 year, and then twice a year
for 2 years. Before each immunization, a leukapheresis was performed
for DC generation. Each vaccination consisted of 6–60 x
106 antigen-loaded DC, which were divided into three to
five aliquots. One aliquot was administered intravenously (i.v.), and
the remaining aliquots (two to four) were injected subcutaneously
(s.c.) into axillary and inguinal lymph nodes. Patients in group B
received only s.c. administration.
Monitoring anti-MAGE responses
Blood samples on heparin were collected on a weekly basis to
monitor the immune response against the MAGE peptide(s). An aliquot of
cells was used immediately for testing, and the remaining cells were
kept frozen for further analysis at 20 million cells/ml in Iscove’s
medium supplemented with 50% autologous serum.
For CFC analysis, unseparated, freshly drawn PBMC suspended in
Iscove’s medium (2x106 cells/ml) were incubated overnight
at 37°C in a humidified 7% CO2 atmosphere in
round-bottomed, sterile culture tubes (Nunc, Roskilde, Danemark) with
10 µg/ml brefeldin (Sigma Chemical Co., St. Louis, MO) and 2 µg/ml
of the relevant peptide(s). If more than one peptide were used for
vaccination, stimulation was performed individually for each peptide in
separate tubes to discriminate the response. As negative control, the
MAGE peptides, not used to pulse the DC that were injected to the
evaluated patient, or the MAGE-A1.Cw16 peptide (SAYCEPRKL; UCB) were
used. Cells were fixed and stained with a phycoerythrin (PE) or
peridinyl chorophil (PerCP)-conjugated, anti-CD3 monoclonal antibody
(mAb) and a combination of PE-conjugated, anti-CD16 and anti-CD56 mAb
(Becton Dickinson, Rutherford, NJ). Subsequently, the cells were
permeabilized using a commercially available kit (FIX and PERM; An der
Grub, Kaumberg, Austria), and intracytoplasmic staining was performed
using a fluorescein isothiocyanate (FITC)-conjugated, anti-IFN- mAb
(Becton Dickinson). Events acquisition was performed using a FACScan®
cytometer, and analysis of the IFN--positive events was performed
using the CellQuest® software (Becton Dickinson) after morphological
gating on lymphocytes and lymphoblasts, according to side scatter (SSC)
and forward scatter (FSC).
For ELISPOT analysis, PBMC (2x106 cells/ml) suspended in
Iscove’s medium supplemented with 10% human serum, IL-2 (Genzyme,
Cambridge, MA; 10 IU/ml), and mercaptoethanol (50 µM) were incubated
overnight with the same peptides as those used for the CFC. Cells were
then transferred for another 24 h on nitrocellulose microtiter
plates (Millipore, Bedford, MA) coated previously with a capture,
anti-IFN- mAb (MABTECH; Nacka, Sweden). Eight different dilutions of
the sample were tested always (200,000, 100,000, 50,000, 25,000,
12,500, 6250, 3125, and 1562 cells/well). Cytokine secretion was
revealed using a biotinylated, anti-IFN- mAb (MABTECH; Nacka),
extravidin peroxydase (Sigma), and aminoethylcarabazole (AEC; Sigma) as
substrate. Enumeration was performed under light microscopy.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Times to progression for the patients are given in Table 1 . In group A and B, 6 and 11 patients, respectively, had measurable tumor mass. Among them, five patients had stable disease for 90 (GEMA), 90 (DUJE), 90 (ANHU), 120 (VERL), and 182 (QULU) days, and two displayed tumor regression. As assessed by PET scan (data not shown), one patient had progressive regression (over more than 1 year) of axillary, metastatic melanoma (PAPE, from group A and B) and is still in remission more than 2 years after the first DC injection. Complete regression of some lung metastasis was also observed in another patient (DEJO) with stage IV melanoma with no tumor progression at >340 days. Among patients with no measurable tumor mass, two (PISU and BOLE) were free of relapse more than 1 year after DC vaccination (403 and >400 days, respectively; Table 1 ).
In vitro detection of anti-MAGE, IFN--producing cells
(Table 1
and Fig. 1
)
A clear increase in the frequency of cells secreting IFN- in
response to one of the immunizing, MAGE-derived peptides was observed
after vaccination by CFC and ELISPOT in six out of eight patients
(75%) receiving DC pulsed with peptide(s) alone (group A). The
biological response rate in patients immunized against the MAGE-A3.A2
peptide was 67% (4/6). Two out of two patients receiving DC pulsed
with the MAGE-A3.A1 peptide and one out of two patients receiving DC
pulsed with the MAGE-A1.A1 peptide responded to the vaccine peptide
(Table 1)
. In view of the good correlation between CFC and ELISPOT, the
second group of patients (group B) vaccinated with DC pulsed with both
peptide(s), and KLH was monitored only by CFC. Because of the
pre-vaccination presence of anti-M2 antibodies, one patient of group A
(ADMA) did not receive IV-DC injection. Nevertheless, this
patient disclosed a strong, anti-peptide response. These data, together
with the concomitant discovery that IV-administered DC do not
home to lymph nodes [7
], led us to stop IV-DC
injection in patients of group B. In this group, anti-peptide responses
were observed in 6 out of 16 patients (38%), and anti-KLH responses
were detected using the same technique in 10/15 patients (67%).
However, we must consider that the four patients enrolled initially in
group A and subsequently enrolled in group B did not respond to
peptide(s), although all of them were responsive in group A. Thus,
taking into account patients not previously treated with DC in group B,
the response rate was 6/12 (50%). The response rate to the peptide
used for DC pulsing in group B was 3/10 for MAGE-A3.A2 (3/8 for
exclusive, group-B patients), 2/4 for MAGE-A1.A1, 0/3 for MAGE-A3.B44,
and 1/3 for MAGE-A3.A1 (Table 1)
. In both groups, immune response of
some patients was not evaluable because of the unavailability of fresh
samples and the loss of sensitivity of the assessment of T-cell
reactivity in frozen samples as compared with fresh ones (Fig. 2
). On fresh samples, the response to the vaccine was detected by
CFC as soon as after the first vaccination in group A, whereas the
response was detected mostly after the second vaccination in group B.
Similar order of magnitude was observed in the induction of T cells,
when studied by CFC or by ELISPOT. However, these ELISPOT responses
were delayed by one or two weeks as compared with those detected by CFC
(Fig. 1) . The frequency of anti-MAGE-A3.A2 cells was <0.05% in all
patients before vaccination rose up to 3.97% as assessed by CFC and to
3.1% as assessed by ELISPOT. The most potent responses were observed
in group A. The median, best, overall immune response (BOIR) directed
specifically against the peptide(s) used to pulse the DC was 0.98%
(range: 0.13–3.97) of T cells in group A and 0.17% (range: 0.11–0.3)
in group B. In this group, the median BOIR against KLH was 0.29%
(range: 0.13–0.62). In all patients of both groups, the number of
peptide-specific T cells returned to pre-vaccination levels after the
fourth injection (V4), demonstrating the transient nature of the
response.
|
|
antibodies. Analysis of IFN--positive cells by back-gating disclosed
that responding cells were T cells (Fig. 1)
, whereas background,
IFN- secretion was attributable to NK cells. As expected from the
HLA class I restriction of the peptide used, CD3+ IFN--producing T
cells were CD8+, although the level of expression of this antigen was
low (unpublished results), probably because of a physiological
down-regulation of the CD8 molecule consecutive to recognition of class
I MHC-peptide complexes [8
]. A similar mechanism could
occur with the CD4 molecule because KLH-responding cells also expressed
low levels of CD4 (unpublished results). |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
secretion in >50% of the patients,
(2) this anti-tumor response is transient, and (3) maintenance of
memory effectors cannot be achieved by providing a nonspecific, T-cell
help through class II presentation of KLH-derived peptides.
The decrease of the anti-MAGE, IFN- response in our patients after
the third vaccination and the impossibility to boost it with the fourth
injection are compatible with an exhaustion of effector cells through
activation-induced cell death (AICD). Such a mechanism could explain
why the four peptide-responding individuals treated previously in group
A failed to respond to the same peptide(s) after enrollment in group B.
These results are contradictory somehow with the murine data showing
the abrogation of the functional deletion of anti-tumor CTL after
presentation of class I-restricted peptides by DC in opposition to
other formulations [4
]. However, these results were
obtained using murine DC generated in the presence of fetal calf serum
(FCS), which, by itself, provides a significant, nonspecific help
[9
]. Because FCS should be avoided for preparation of
cellular product for clinical use, our DC were generated in serum-free
conditions with the consequence that they are probably not able to
provide such help. Although our study was not randomized, and the
numbers are small, the absence of improvement in the immunization with
KLH-pulsed DC does not support the role of a nonspecific T-cell help in
the induction of long-lasting, anti-MAGE responses in humans. In two
previous clinical studies, the contribution of nonspecific, T-cell help
was regarded as beneficial. Rosenberg et al.
[10
] described a clinical trial using an immunodominant
peptide from the gp100 melanoma-associated antigen in which patients
were vaccinated with the peptide in incomplete Freund’s adjuvant (IFA)
with or without IL-2. Tumor regressions were observed only when the
helper cytokine IL-2 was administered together with the peptide. In the
second study, Nestle et al. [11
] vaccinated
melanoma patients with DC pulsed with lysates or tumor peptide and KLH
as a source of nonspecific help. Again, tumor regressions were observed
in 5 out of 16 patients. In both studies, kinetic study of the
anti-tumor response was impossible because of the constraints of the
techniques used for the monitoring of immune responses. Rosenberg
et al. [10
] used an 11-day culture followed
by the assessment of IFN- release in the supernatant, whereas Nestle
et al. [11
] used the delayed-type
hypersensitivity (DTH) reaction. In this latter study, the DC were
generated in FCS-enriched medium. Thus, in these studies, no
information is available on the long-term maintenance of anti-tumor
responses that have been induced.
It might be argued that the absence of memory induction in our study is
a result of the use of immature DC. Most clinical trials described have
used immature DC [11
12
13
], but induction of in
vitro DC maturation is certainly appealing, because mature DC are
more stable and express higher levels of HLA and co-stimulatory
molecules [14
]. However, this may be viewed with caution
and the example of bacterial lipopolysaccharide (LPS), kept in mind.
LPS is a potent, DC-maturation agent that induces DC to produce high
amounts of IL-12, a cytokine pivotal for the Th1 differentiation of
CD4+ T cells. Low doses of LPS at the initiation of DC culture induce a
state of unresponsiveness characterized by the inhibition of their
capacity to produce IL-12 and tumor necrosis factor (TNF)-
on LPS
re-challenge [15
]. This suggests that it could be more
clinically relevant to preserve the possibility to achieve in
vivo DC maturation through vaccination with immature DC to allow
full, in vivo, T-cell activation into the draining lymph
nodes, notably through efficient IL-12 production. Recently, Thurner
et al. [16
] provided new, clinical insights
into the role of the maturation status of the immunizing cells. In
their trial, DC were pulsed with the MAGE-A3.A1 tumor peptide and
helper antigens (tetanus toxoid or tuberculin) and maturated using an
autologous, monocyte-conditioned medium. The vaccine was administered
at a 14-day interval three times in the skin (intradermal+s.c.)
and two times i.v. Essentially, these results were comparable
to those shown in our trial using immature DC. Although they
demonstrate regression of some metastatic sites in 6 out of 11
patients, this was in an overall setting of progressive disease.
Anti-peptide CTL were detected in 8 out of 11 patients after the three
administrations in the skin but declined thereafter. Analysis of
peptide-specific, IFN--producing cells by ELISPOT disclosed a
response in 2 out of 11 patients. The low number of ELISPOT-positive as
compared with CTL-positive samples in this study could be related to
the absence of a preincubation step, which, in our hands, enhances
greatly the sensitivity of the technique.
MAGE antigens are poor immunogens notably, and the frequency of CTL precursors (CTLp) susceptible to recognize them is very low. The frequency of anti-MAGE T cells detected in our study is higher than that expected from previous works using limiting dilution assays (LDA; 4–17x10-7) [17 , 18 ]. However, techniques based on long-term culture such as LDA do not assess the same type of responses as cytokine-based assays [19 ]. Indeed, Altman et al. [20 ], using tetrameric complexes of HLA-A2 and epitope peptides from HIV gag and pol, stained 2% of CD8+ T cells of HIV-infected patients, a situation in which LDA estimated the precursor frequency at 1/40.000–1/20.000 previously [21 ]. We interpret the high frequency of anti-MAGE-A3.A2 T cells detected in the present study as the assessment by CFC and ELISPOT of effectors already primed in vivo that are able to be activated in vitro by restimulation with the relevant antigen but subsequently undergo activation-induced cell death. These cells would not proliferate on a long-term basis and therefore would not be detected by LDA. This could explain the apparent paradox constituted by the observation of tumor regressions after vaccination with peptide alone without evidence of circulating, anti-tumor CTL when studied by LDA [22 ].
This study provides a warning signal on the use of class I-restricted peptide in humans even with DC as natural adjuvant and KLH as a source of nonspecific, T-cell help. Apparently, in vitro DC maturation also fails to trigger the induction of memory cells [16 ]. Our hypothesis is that providing a tumor-specific, T-cell help is crucial for the induction of memory and for effective, tumor-cell lysis in humans. The recent identification of human, tumor antigens recognized by CD4+ T cells in the context of MHC class II molecules provides new tools for investigating the role of specific, T-cell help [23 24 25 ]. This will be the question addressed in our next trial.
|
ACKNOWLEDGEMENTS |
|---|
Received January 9, 2001; revised January 25, 2001; accepted January 26, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and prostaglandin E2 Blood 91,3112-3117This article has been cited by other articles:
|
|
 |
|
  G. J. Ullenhag, J.-E. Frodin, M. Jeddi-Tehrani, K. Strigard, E. Eriksson, A. Samanci, A. Choudhury, B. Nilsson, E. D. Rossmann, S. Mosolits, et al. Durable Carcinoembryonic Antigen (CEA)-Specific Humoral and Cellular Immune Responses in Colorectal Carcinoma Patients Vaccinated with Recombinant CEA and Granulocyte/Macrophage Colony-Stimulating Factor Clin. Cancer Res., May 15, 2004; 10(10): 3273 - 3281. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Nakahara, T. Tsunoda, T. Baba, S. Asabe, and H. Tahara Dendritic Cells Stimulated with a Bacterial Product, OK-432, Efficiently Induce Cytotoxic T Lymphocytes Specific to Tumor Rejection Peptide Cancer Res., July 15, 2003; 63(14): 4112 - 4118. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. H. Butterfield, A. Ribas, V. B. Dissette, S. N. Amarnani, H. T. Vu, D. Oseguera, H.-J. Wang, R. M. Elashoff, W. H. McBride, B. Mukherji, et al. Determinant Spreading Associated with Clinical Response in Dendritic Cell-based Immunotherapy for Malignant Melanoma Clin. Cancer Res., March 1, 2003; 9(3): 998 - 1008. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
