Immunotherapy with dendritic cells and CpG oligonucleotides can be combined with chemotherapy without loss of efficacy in a mouse model of colon cancer
Article first published online: 30 DEC 2005
Copyright © 2005 Wiley-Liss, Inc.
International Journal of Cancer
Volume 118, Issue 11, pages 2790–2795, 1 June 2006
How to Cite
Bourquin, C., Schreiber, S., Beck, S., Hartmann, G. and Endres, S. (2006), Immunotherapy with dendritic cells and CpG oligonucleotides can be combined with chemotherapy without loss of efficacy in a mouse model of colon cancer. Int. J. Cancer, 118: 2790–2795. doi: 10.1002/ijc.21681
- Issue published online: 14 MAR 2006
- Article first published online: 30 DEC 2005
- Manuscript Accepted: 11 OCT 2005
- Manuscript Received: 29 JUL 2005
- Deutsche Forschungsgemeinschaft (DFG). Grant Number: En 169/7.1
- dendritic cells;
- CpG oligonucleotides;
Although immunotherapy has shown promising results in the treatment of cancer, clinical studies assessing immunologic approaches in patients with advanced cancer will seldom be conducted in the absence of conventional treatment strategies such as chemotherapy. Here we investigate the combination of chemotherapy with CpG oligonucleotide and dendritic cell-based immunotherapy in the C26 mouse model of colon carcinoma. The coinjection of antigen-pulsed, mature dendritic cells and CpG oligonucleotides together with a peritumoral injection of CpG oligonucleotides elicits a CD8 T-cell response resulting in tumor rejection and long-term protection in the C26 model. Tumor-bearing mice were treated weekly for 4 weeks by this immunotherapy protocol, by 5-fluorouracil plus leucovorin or irinotecan, or by the combination of immunotherapy and chemotherapy. We observed that immunotherapy was more effective in reducing tumor growth and increasing survival than 5-fluorouracil or irinotecan. Immunotherapy was well tolerated, whereas therapeutic doses of 5-fluorouracil or irinotecan were associated with dose-limiting toxicity. Furthermore, the efficacy of immunotherapy combined with either 5-fluorouracil or irinotecan was similar to that of immunotherapy alone. Addition of immunotherapy to either 5-fluorouracil or irinotecan treatment strongly decreased the toxicity of chemotherapy. Immunotherapy both with and without chemotherapy generated a memory immune response, leading to tumor rejection in mice rechallenged with C26 tumor cells up to several months after treatment. In summary, immunotherapy with a combination of dendritic cells and CpG oligonucleotides is superior to chemotherapy in the C26 tumor model. This immunotherapy protocol can be combined with current chemotherapy agents with no loss in therapeutic activity. © 2005 Wiley-Liss, Inc.
Immunotherapy for the treatment of cancer enhances antitumor immune responses and has shown considerable benefit in preclinical studies.1 Because immunologic approaches will however rarely be used alone in patients with advanced cancer, it is important to determine how immunotherapy may be affected by established treatment strategies such as chemotherapy. A major concern for combining cytotoxic chemotherapy with immunotherapy is the immunosuppression frequently associated with chemotherapy.2, 3 For instance, chemotherapy can be toxic to the dividing cells in the bone marrow and peripheral lymphoid tissues, and the antiproliferative effect of many chemotherapy agents may affect a developing antitumor T cell response.4
Synthetic oligonucleotides containing CpG motifs are recognized by cells of the innate immune system through the Toll-like receptor 9, leading to potent stimulation of both the innate and the adaptive immune responses.5, 6 In murine tumor models, effective antitumor immunity can be achieved by treatment with CpG oligonucleotides both as monotherapy and in combination with radiotherapy or chemotherapy.7, 8, 9, 10 In patients with metastatic melanoma, CpG oligonucleotides can be administered safely and exert antitumor activity.11 Furthermore, in a Phase-II study, CpG oligonucleotides improved tumor responses to standard chemotherapy in patients with non-small cell lung cancer.12 CpG oligonucleotides also increase the expression of costimulatory and MHC molecules on human B cell tumors.13
We have developed an effective protocol for dendritic cell-based therapy of subcutaneous tumors generated by injection of the syngeneic C26 colon adenocarcinoma cell line.14 Dendritic cells are loaded in vitro with irradiated tumor cells and are activated by CpG oligonucleotides prior to administration, resulting in enhancement of the antitumor response.15 The tumor-specific immune response, that depends on CD8 T cells, is further increased by coinjecting CpG oligonucleotides together with the dendritic cells and by a simultaneous peritumoral injection of CpG oligonucleotides.16
In the present study, we combined an established immunotherapy protocol with two chemotherapeutic drugs currently used in the treatment of advanced colon cancer, 5-fluorouracil and irinotecan. These agents have been shown to be moderately effective for therapy of C26 tumors.17, 18 We demonstrate that simultaneous administration of either 5-fluorouracil or irinotecan with dendritic cell-based immunotherapy does not affect the antitumor effect of immunotherapy.
Material and methods
Female Balb/c mice were purchased from Harlan-Winkelmann (Borchen, Germany) and maintained in the animal facility of the Medical Clinic of the Ludwig-Maximilian University. Mice were 5–12 weeks of age at the onset of experiments. Animal studies were approved by the local regulatory agency (Regierung von Oberbayern, Munich, Germany).
Reagents and cell line
The phosphorothioate-modified oligonucleotide 1826 (5′-TCCATGACGTTCCTGACGTT-3′) was kindly provided by Coley Pharmaceutical Group (Langenfeld, Germany). Endotoxin was undetectable by the limulus amoebocyte lysate assay. The Balb/c-derived C26 colon carcinoma cell line14 (Cell Lines Service, Heidelberg) was maintained in DMEM supplemented with 10% FCS, 2 mM L-glutamine, 100 μg/ml streptomycin and 1 IU/ml penicillin at 37°C and 5% CO2.
Preparation of bone marrow-derived dendritic cells
Bone marrow-derived dendritic cells were prepared as described.16 Bone marrow cells were harvested from murine femur and tibia. Erythrocytes were lysed with ammonium chloride buffer (BD Biosciences, Heidelberg, Germany), and lymphocytes and granulocytes were depleted by incubation with rat anti-mouse CD4, CD8a, CD45R, Ly6G and MHC ll mAbs (BD Biosciences) for 30 min at 4°C, followed by incubation with sheep anti-rat IgG conjugated to magnetic beads (Dynal Biotech, Oslo, Norway) for 30 min at 8°C. Unlabeled cells were isolated in a strong magnetic field and cultured at 5 × 105 cells/ml in RPMI 1640 medium supplemented with 10% FCS, 2 mM L-glutamine, 100 μg/ml streptomycin, 1 IU/ml penicillin, 20 ng/ml GM-CSF and 20 ng/ml IL-4 (Tebu Bio, Offenbach, Germany) (DC medium). On day 6, loosely adherent cells were harvested and cultured at 1 × 106 cells/ml with irradiated C26 cells (C26:dendritic cell ratio 1:5) in fresh DC medium. CpG 1826 (6 μg/ml) were added for dendritic cell activation on day 7. On day 8, loosely adherent cells were harvested and washed in Hank's buffered salt solution. Percentage and maturation status of dendritic cells were verified by flow cytometry. Dendritic cells generally represented 70% of the preparation.
For phenotypic analysis, dendritic cells were stained with anti-mouse CD86-FITC, CD80-PE, MHC-ll-PE, CD11b-PerCP, CD11c-APC mAbs and appropriate isotype controls (BD Biosciences). They were examined by flow cytometry (FACSCalibur, BD Biosciences). Data were analyzed using CellQuest software (BD Biosciences).
For tumor induction, C26 cells were washed in Hank's buffered salt solution and 2.5 × 105 cells were injected subcutaneously in the right flank in a volume of 250 μl. Tumor growth was monitored 3 times a week for 80 days after tumor challenge and expressed as the product of the perpendicular diameters of individual tumors. Animals were killed when tumor size exceeded 225 mm2 or when there were other signs of animal distress observed twice in 48 h.
5-Fluorouracil and irinotecan were injected intraperitoneally in 0.9% NaCl solution. 5-Fluorouracil was always combined with 100 μg/g body weight leucovorin i.p. injected in 2 doses of 50 μg/g body weight 1 hr before 5-fluorouracil application and again at the time of 5-fluorouracil application. For dendritic cell immunization, mice received 2 × 105 bone marrow-derived dendritic cells s.c. together with 100 μg CpG 1826 in 200 μl HBSS in the non tumor-bearing (contralateral) flank and another 100 μg CpG 1826 in 100 μl PBS peritumorally. Treatment with chemotherapy and dendritic cell immunization was initiated at a tumor size of 30–60 mm2. The interval between the first and second applications was 5 days, and the remaining 2 applications were performed at 7- to 8-day intervals.
Quantification of cytokines
Cytokine concentrations were determined by ELISA for IL-6 (BioSource, Solingen, Germany) and IL-12p70 (BD Biosciences) in serum of tumor-bearing Balb/c mice 2 hr after the first therapy, according to the manufacturers' protocol.
Mean tumor size curves were plotted until 2 mice died or were killed. The last measured value for a dead mouse was included in calculation of the mean for remaining time points. Comparisons in tumor size among groups were made using the Mann-Whitney test for various time points. Comparisons among groups regarding survival time were made using the log-rank test. For cytokine ELISAs, significance was assessed by the Student's t test. Fisher's exact test was used for comparing tumor incidence in rechallenged mice. Statistical analyses were performed using SPSS software.
Immunotherapy with dendritic cells and CpG oligonucleotidesis more effective than 5-fluorouracil or irinotecan in thetreatment of C26 tumors
Mice bearing C26 tumors on one flank were treated 4 times at 5- to 7-day intervals with either immunotherapy or increasing doses of 5-fluorouracil. The immunotherapy protocol was based on a combination of dendritic cells and CpG oligonucleotides established previously.16 According to this treatment protocol, tumor antigen-pulsed bone marrow-derived dendritic cells were matured overnight with CpG oligonucleotides and coinjected together with CpG oligonucleotides in the flank opposite the tumor. At the same time, mice received an injection of CpG oligonucleotides in the peritumoral area. Alternatively, mice were treated with the thymidine synthase inhibitor 5-fluorouracil, administered together with leucovorin to enhance inhibition of DNA synthesis.19 50 mg/kg of 5-fluorouracil had no effect on tumor progression, whereas 75 mg/kg 5-fluorouracil transiently arrested tumor growth for one week (Fig. 1a). Tumor growth then resumed at the same rate as in the untreated control group. The transient reduction in tumor progression did not result in a significant increase in survival (Fig. 1b). 5-Fluorouracil (100 mg/kg) led to tumor regression in 4 of 5 mice, but was associated with severe toxicity. In this group, 4 of 5 mice died 7 days after the second application at the peak of the hematologic toxicity of 5-fluorouracil.20 In contrast, immunotherapy with dendritic cells and CpG oligonucleotides led to complete tumor regression in 4 of 5 mice and to complete cure in 2 of 5 mice. The increase in survival in this treatment group was highly significant compared to untreated and to all 5-fluorouracil-treated groups (Fig. 1b). While some of the mice receiving immunotherapy displayed transient signs of discomfort such as piloerection and reduced activity, these signs were observed for a maximum of 48 hr after application. No signs of severe toxicity were observed in mice treated by immunotherapy.
In the same experiment, we compared dendritic cell and CpG oligonucleotide-based immunotherapy with the topoisomerase inhibitor irinotecan (Fig. 2a). The effect on tumor growth of 250 mg/kg irinotecan, administered at 5- to 7-day intervals, was similar to that of 75 mg/kg 5-fluorouracil: an initial arrest in tumor growth lasting 7 days was followed by a rapid increase in tumor size. This transient reduction in tumor growth did not increase survival (Fig. 2b). In mice treated with the higher dose of 350 mg/kg irinotecan, lethal toxicity was observed in 2 of 5 mice 5 days after the first administration. In the remaining 3 mice of this group, tumor progression was reduced compared to untreated mice. However, no tumor regression was observed.
Immunotherapy with dendritic cells and CpG oligonucleotidescan be combined with chemotherapy without loss oftherapeutic activity
To assess therapeutically relevant interactions between chemotherapy and dendritic cell and CpG-based immunotherapy, we examined the efficacy of simultaneous administration of 5-fluorouracil with immunotherapy. Mice received 4 administrations of immunotherapy with dendritic cells and CpG oligonucleotides, or 100 mg/kg 5-fluorouracil with leucovorin, or both treatments simultaneously. As expected, tumors regressed in 4 of 5 mice treated by immunotherapy (Fig. 3a and 3b) and none of them relapsed. In mice receiving both immunotherapy and 5-fluorouracil, therapy was equally effective: complete tumor regression was observed in 4 of 5 mice and a complete cure was observed in 3 mice in this group. Survival was highly increased in mice treated simultaneously by immunotherapy and 5-fluorouracil as well as in mice receiving immunotherapy alone (Fig. 3b). This indicates that the simultaneous application of high-dose 5-fluorouracil does not reduce the overall therapeutic effect seen with dendritic cell and CpG-based immunotherapy. Interestingly, although administration of 100 mg/kg 5-fluorouracil was associated with severe toxicity, resulting in death of 4 of 5 mice 5 to 8 days after the second application, the mice treated simultaneously by immunotherapy and the same dose of 5-fluorouracil displayed only transient signs of discomfort and no toxic deaths.
We then assessed the efficacy of simultaneous administration of irinotecan with dendritic cell and CpG oligonucleotide-based immunotherapy (Fig. 3c). To better differentiate between treatment groups, treatment was initiated when tumors reached a size of 60 mm2 (instead of 30 mm2 as in previous experiments). The addition of irinotecan at 200 or 400 mg/kg to the immunotherapy protocol did not affect the efficacy of immunotherapy. Survival was increased both in the group receiving immunotherapy alone and in the groups receiving immunotherapy combined with irinotecan compared to untreated animals. Between the 3 groups treated with immunotherapy, no significant differences were observed.
In vivo application of CpG oligonucleotides leads to production of high levels of Th1-type cytokines that are important for tumor rejection.3 As a measure of the suppressive activity of simultaneous chemotherapy on the immunotherapy protocol, we assessed serum levels of IL-12p70 and IL-6 2 hr after treatment. Dendritic cell and CpG oligonucleotide-based immunotherapy induced high serum levels of both cytokines that were not affected by simultaneous administration of 5-fluorouracil (Fig. 4). In contrast, addition of irinotecan to the immunotherapy protocol resulted in lower serum cytokine levels that were significant for IL-6.
Chemotherapy does not affect the long-term protectiveeffect of immunotherapy with dendritic cells andCpG oligonucleotides
Mice cured by dendritic cell and CpG oligonucleotide-based immunotherapy show long-term protection against rechallenge with the same tumor (Heckelsmiller et al., 2002). To evaluate the effect of simultaneous administration of 5-fluorouracil on this memory response, we rechallenged a total of 15 cured mice from 3 different experiments (91–293 days after last application of treatment) with C26 tumor cells (Table I). Tumor formation was observed in 3 of 8 mice that had received immunotherapy alone, whereas none of the 7 mice cured by simultaneous administration of 5-fluorouracil and immunotherapy developed a tumor. The decrease in tumor incidence between cured and naive mice was only significant for mice cured by 5-fluorouracil plus immunotherapy. This indicates that the immunotherapy-induced memory response against tumor antigens is not impeded by simultaneous administration of chemotherapy; on the contrary, the combination of both treatment modalities induces an effective antitumor memory response.
|Experiment No.||Days after end of therapy||Proportion of mice with tumor in therapy groups|
|Naive||DC plus CpG||DC plus CpG plus5-FU|
|Control||–||4 of 4|
|1||91||1 of 2||0 of 2|
|2||153||2 of 4||0 of 3|
|3||293||0 of 2||0 of 2|
|Total||4 of 4||3 of 8||0 of 7|
Encouraging results from a clinical study in patients with metastatic non-small cell lung cancer show that immunotherapy with CpG oligonucleotides improves the response rate to standard chemotherapy.12 While patients with certain tumor types may benefit substantially from immunotherapy, future studies focusing on immunologic treatments in patients with advanced cancer will rarely be possible in the absence of established treatments such as chemotherapy or radiotherapy. A major concern for the successful combination of chemotherapy and immunotherapy is the significant immunosuppression associated with chemotherapy.2, 3 The antiproliferative effect of many chemotherapeutic agents, in particular antimetabolites such as 5-fluorouracil, may impair a developing T-cell response. For instance, generation of CTLs in vitro is markedly reduced by 5-fluorouracil when cells are exposed to the drug during the sensitization phase.21 Furthermore, dendritic cell function and blood counts are decreased in patients receiving various types of chemotherapy.22
In particular, the long-term cyclic administration of chemotherapy agents usually applied in clinical practice may affect the immune system more strongly than the same drug administered over a short period. In contrast to other studies focusing on the combination of chemotherapy and immunotherapy with a short treatment schedule of a few days,9, 23, 24 we administered high-dose chemotherapy weekly over the course of 4 weeks. Despite this extended schedule, the addition of 5-fluorouracil or irinotecan to dendritic cell and CpG oligonucleotide-based immunotherapy did not result in a loss of therapeutic activity. Although the relative contribution of immunotherapy and chemotherapy in the combination treatment protocol is difficult to assess, it is clear that the generation of a long-term protective immune response lasting up to several months after end of treatment was not impaired by chemotherapy. On the contrary, protection tended to be more efficient in mice treated with combination therapy than in mice treated by immunotherapy alone. Thus, simultaneous administration of 5-fluorouracil with immunotherapy inhibited neither the priming of a de novo immune response nor the formation of immunological memory.
As an indicator of the suppressive activity of 5-fluorouracil and irinotecan, we assessed cytokine production at an early time point after therapy when cytokine levels induced by our immunotherapy protocol are high. Interestingly, 5-fluorouracil with leucovorin had no effect on immunotherapy-induced IL-12p70 and IL-6 production, whereas serum levels of both cytokines were lower after administration of irinotecan. This decrease did not however correlate with a reduction in the therapeutic efficiency of immunotherapy. The partial immune activation may have been sufficient to mediate an effective anti-tumor response, or alternatively, other mechanisms may play a role at later time points in mediating the antitumor response.
Combination of chemotherapy with immunization by immature dendritic cells has been studied recently by several groups.23, 24, 25 Immature dendritic cells, however, activate tumor-specific CTL inefficiently. Additional signals, such as those generated by Toll-like receptor ligands, are necessary to enable dendritic cells to generate a productive antitumor immune response.26 Indeed, a direct comparison of immature and mature dendritic cells in patients with metastatic melanoma showed that mature dendritic cells were superior inducers of antigen-specific T cell responses.27 Indeed, the strong antigen-specific T cell response induced by antigen-pulsed dendritic cells matured with LPS is not affected by the myelosuppressive drugs doxorubicin and melphalan.28 In the present study, dendritic cells were matured by CpG oligonucleotides before immunization to generate a strong immune response. Furthermore, CpG oligonucleotides were both coinjected with the dendritic cells and injected in the peritumoral area to further increase therapeutic activity.16 Local administration of CpG oligonucleotides activates dendritic cells in vivo and conditions draining lymph nodes, leading to increased CTL and Th1-type immune responses.29, 30
Instead of leading to immunosuppression, chemotherapy may in some cases enhance the antitumor effect of immunotherapy. For instance, cyclophosphamide increases the antitumor immune response following immunization by suppressing T regulatory cells and increasing the Th1-type immune response.31, 32 Alternatively, agents such as 5-fluorouracil and irinotecan can sensitize colon carcinoma cells to the lytic effects of antigen-specific CTLs. This drug-mediated sensitization involves components of the caspase pathway.33 In the present study, we observed no increase in the antitumor effect by combining immunotherapy and chemotherapy. It is possible that the strong efficacy of immunotherapy alone may have obscured the contribution of the chemotherapy to the antitumor effect of the combined treatment. Indeed, 2 studies combining a less effective immunotherapy protocol consisting solely of CpG oligonucleotides with another topoisomerase inhibitor, topotecan, report better results for the combination therapy versus immunotherapy alone.9, 34
In our study, the dose of 5-fluorouracil or irinotecan necessary to induce tumor regression was associated with lethal toxicity, whereas the highly effective immunotherapy with dendritic cells and CpG oligonucleotides was associated with only transient side effects. Unexpectedly, addition of immunotherapy to either 5-fluorouracil or irinotecan treatment protected from the toxicity of chemotherapy. As CpG oligonucleotides can stimulate hematopoiesis,35 we hypothesize that our immunotherapy protocol may counteract the hematologic toxicity of 5-fluorouracil and irinotecan. Indeed, CpG oligonucleotides can reverse the decrease in NK cell activity induced by 5-fluorouracil in a murine hepatoma model.36 In the literature, sequence of administration appears to be crucial: while sequential administration of topotecan followed by a CpG oligonucleotide slowed tumor growth in athymic nude mice, a protocol alternating treatment by CpG oligonucleotides and topotecan every 3–4 days resulted in toxic deaths.34 The mechanism for this unexpected toxicity is unclear.
Although the role of both T cells and NK cells in mediating antitumor effects of CpG-based immunotherapy is established,7, 8 the contribution of CpG-stimulated B cells to an effective antitumor response remains to be determined. In mice bearing C26 tumors as well as in patients with colon cancer, circulating antibodies recognizing intracellular tumor-associated antigens are described.37 The role of these antigens in antitumor immunity is unclear. CpG oligonucleotides activate human B cells to proliferate and to produce cytokines and immunoglobulins both directly via the Toll-like receptor 9 expressed on B cells and indirectly via interferon-α produced by activated plasmacytoid dendritic cells.5, 38, 39 In particular, memory B cells as well as memory T cells may contribute to the long-term protective effect of CpG-based immunotherapy.
In conclusion, we have shown that a potent immunotherapy protocol involving dendritic cells and CpG oligonucleotides may be combined with chemotherapy agents without loss of efficacy in the C26 tumor model. Furthermore, combination of immunotherapy with either 5-fluorouracil or irinotecan substantially decreased toxicity of chemotherapy. These results are important for the development of dendritic cell-based immunotherapy toward clinical application. In the future, novel Toll-like receptor ligands may play a role in the maturation and activation of dendritic cell vaccines. One such ligand are short immunostimulatory RNA sequences that bind to Toll-like receptor 7 and induce interferon-α production both in vitro and in vivo.40
This work was supported by Deutsche Forschungsgemeinschaft (DFG) grant En 169/7.1 to S.E. We thank Susanne Wenk for technical assistance.
- 11TLR9-targeted CpG immunostimulatory treatment of metastatic melanoma: a phase II trial with CpG 7909 (ProMune). J Clin Oncol 2004; 22: 7513., , , , , , .
- 12CpG 7909, a TLR9 agonist, added to first Line taxane/platinum for advanced non-small cell lung cancer, a randomized, controlled phase II study. J Clin Oncol, in press., , , , , , , .
- 21Treatment of colon and breast carcinoma cells with 5-fluorouracil enhances expression of carcinoembryonic antigen and susceptibility to HLA-A(*)02.01 restricted, CEA-peptide-specific cytotoxic T cells in vitro. Int J Cancer 2003; 104: 437–45., , , , , , , , , , , , et al.