R.V. and N.L. contributed equally to this work
Potentiation of a p53-SLP vaccine by cyclophosphamide in ovarian cancer: A single-arm phase II study
Article first published online: 31 JAN 2012
Copyright © 2011 UICC
International Journal of Cancer
Volume 131, Issue 5, pages E670–E680, 1 September 2012
How to Cite
Vermeij, R., Leffers, N., Hoogeboom, B.-N., Hamming, I. L.E., Wolf, R., Reyners, A. K.L., Molmans, B. H.W., Hollema, H., Bart, J., Drijfhout, J. W., Oostendorp, J., van der Zee, A. G.J., Melief, C. J., van der Burg, S. H., Daemen, T. and Nijman, H. W. (2012), Potentiation of a p53-SLP vaccine by cyclophosphamide in ovarian cancer: A single-arm phase II study. Int. J. Cancer, 131: E670–E680. doi: 10.1002/ijc.27388
- Issue published online: 27 JUN 2012
- Article first published online: 31 JAN 2012
- Accepted manuscript online: 3 DEC 2011 02:16AM EST
- Manuscript Accepted: 21 OCT 2011
- Manuscript Received: 15 JUN 2011
- Dutch Cancer Society (Amsterdam, The Netherlands). Grant Number: 2007-3919
- ISA Pharmaceuticals B.V. (Bilthoven, The Netherlands)
- ovarian cancer;
- Top of page
- Material and Methods
- Supporting Information
The purpose of the current phase II single-arm clinical trial was to evaluate whether pretreatment with low-dose cyclophosphamide improves immunogenicity of a p53-synthetic long peptide (SLP) vaccine in patients with recurrent ovarian cancer. Patients with ovarian cancer with elevated serum levels of CA-125 after primary treatment were immunized four times with the p53-SLP vaccine. Each immunization was preceded by administration of 300 mg/m2 intravenous cyclophosphamide as a means to affect regulatory T cells (Tregs). Vaccine-induced p53-specific interferon-gamma (IFN-γ)-producing T cells evaluated by IFN-γ ELISPOT were observed in 90% (9/10) and 87.5% (7/8) of evaluable patients after two and four immunizations, respectively. Proliferative p53-specific T cells, observed in 80.0% (8/10) and 62.5% (5/8) of patients, produced both T-helper 1 and T-helper-2 cytokines. Cyclophosphamide induced neither a quantitative reduction of Tregs determined by CD4+FoxP3+ T cell levels nor a demonstrable qualitative difference in Treg function tested in vitro. Nonetheless, the number of vaccine-induced p53-specific IFN-γ-producing T cells was higher in our study compared to a study in which a similar patient group was treated with p53-SLP monotherapy (p ≤ 0.012). Furthermore, the strong reduction in the number of circulating p53-specific T cells observed previously after four immunizations was currently absent. Stable disease was observed in 20.0% (2/10) of patients, and the remainder of patients (80.0%) showed clinical, biochemical and/or radiographic evidence of progressive disease. The outcome of this phase II trial warrants new studies on the use of low-dose cyclophosphamide to potentiate the immunogenicity of the p53-SLP vaccine or other antitumor vaccines.
Epithelial ovarian cancer is the leading cause of death from gynecological malignancies in Western countries. The observation of improved prognosis in patients with intratumoral T lymphocytes1–3 has encouraged the development of immunotherapy for ovarian cancer. Immunotherapy aims to enhance antitumor immunity to eliminate malignant cells.
A potential target for cancer immunotherapy is the tumor-suppressor protein p53. Mutation of the p53 gene is a frequent event in human oncogenesis,4, 5 which leads to persistent overexpression of p53 in 50–60% of ovarian cancer.6, 7 Therefore, the majority of patients with ovarian cancer might benefit from p53-directed immunotherapy. Despite the fact that p53 is a self-protein, studies both in mice and patients demonstrated that p53-specific immune responses can be induced.8–11 Therefore, we and others recently reported on clinical trials with vaccines targeting p53.12, 13
Although most immunotherapeutic strategies for ovarian cancer treatment investigated so far are capable of inducing antigen-specific immunity, unequivocal clinical benefit for these patients has not yet been demonstrated.14 Similarly, we showed that a p53-synthetic long peptide (SLP) vaccine induces p53-specific T cell responses in patients with ovarian and colorectal cancers, but observed no clinical benefit.12, 15, 16 The observed lack of clinical efficacy may be partly attributed to the presence of regulatory T cells (Tregs). In ovarian cancer, the presence of Tregs and especially the ratio between Tregs and effector T cells have been shown to be important for prognosis.2, 3, 17
Given the observed effects of cancer immunotherapy on immune inhibitory Tregs,18 strategies to eliminate or suppress Tregs are being explored in an attempt to improve clinical efficacy of cancer immunotherapy. One of these strategies is the treatment with low-dose cyclophosphamide, a well-known cytotoxic agent that has been widely applied in the treatment of ovarian cancer until the introduction of platinum-based chemotherapy.19 Dosages used in combination with immunotherapy are generally insufficient for cytotoxic reductions of tumor burden, but reduce Treg numbers and impair their function without deleting other immune cells.20–22 In a murine model, we showed a synergistic effect of cyclophosphamide and wild-type p53-specific cytotoxic cluster of differentiation (CD)8+ T lymphocytes in growth control of p53-overexpressing tumors.23 Its putative Treg-depleting and immune-potentiating characteristics make cyclophosphamide an interesting candidate drug for the elimination of Treg when combined with the p53-SLP vaccine, tested in our previous phase I/II study in patients with ovarian cancer.12
We report the results of a phase II single-arm study combining the p53-SLP vaccine with cyclophosphamide in an attempt to improve immunogenicity and deplete the number of Tregs determined by CD4+FoxP3+ T cell levels. Ten patients with ovarian cancer with (biochemical evidence) of recurrence of disease were immunized four times with the p53-SLP vaccine. Each immunization was preceded by the administration of low-dose cyclophosphamide. Next to immunological responses, clinical activity and safety were monitored.
Material and Methods
- Top of page
- Material and Methods
- Supporting Information
The study protocol for this uncontrolled phase II study was approved by the Central Committee on Research Involving Human Subjects (CCMO; NL21308.000.07) and conducted in compliance with the Declaration of Helsinki. All patients gave written informed consent. The trial was registered at the US National Institutes of Health clinicaltrials.gov (NCT00844506) and in The Netherlands National Trial Register (NTR1407). An independent agency (Trial Coordination Center, Groningen, The Netherlands) was contracted to monitor adherence to good clinical practice principles.
Patients with epithelial ovarian cancer with (biochemical) evidence of recurrent disease after prior cytoreductive surgery and chemotherapy, who were not eligible for renewed chemotherapy, were included. Other inclusion criteria were an adequate hepatic and renal function. Additional inclusion and exclusion criteria were similar to our previous study.12 Primary tumors were evaluated for p53 expression (BP53-12-1, 1:800, Biogenex) for study cohort characterization purpose only, as patients could participate in the study despite their p53 expression status. Tumors with >50% moderate or strong immunostaining were considered to have p53 overexpression.7, 24
Vaccine and treatment scheme
In this phase II single-arm clinical trial, patients with recurrent ovarian cancer were treated with the p53-SLP vaccine combined with low-dose cyclophosphamide. The p53-SLP vaccine consists of ten synthetic 25–30 amino acid long overlapping peptides (spanning amino acids 70–248 of the wt-p53 protein) dissolved in dimethyl sulfoxide (final concentration 20%) admixed with 20 mM phosphate buffer (pH 7.5) and emulsified with an equal volume of Montanide ISA-51.12 The p53-SLP vaccine was administered at a dose of 300 μg/peptide, subcutaneously four times with a 3-week interval. Two days before each immunization, patients were given an intravenous (i.v.) cyclophosphamide infusion (300mg/m2 in 30 min). Immunizations preceded by cyclophosphamide were administered between October 2008 and July 2009.
Our study was primarily designed to evaluate whether pretreatment with cyclophosphamide would (i) improve immunogenicity of the p53-SLP vaccine and (ii) affect Tregs, determined by CD4+FoxP3+ T cell levels. Secondary objectives were (i) evaluation of the clinical activity induced by the p53-SLP vaccine preceded by cyclophosphamide infusion and (ii) evaluation of the safety of the p53-SLP vaccine when preceded by cyclophosphamide infusion. Because of a lack in the decrease of FoxP3+ T cell levels measured, being one of the primary endpoints of the interim analysis of ten patients, the study stopped. Primary endpoint data of the current single-arm study were compared to data obtained in our previously conducted phase I/II clinical trial on the p53-SLP vaccine.12 Standard operating procedures (SOPs) used for study material analysis were the same in both trials.
Lymphocytes and sera
Blood for immunological assays was obtained before immunization, after two and four immunizations. Two days after each delivery of cyclophosphamide, peripheral blood mononuclear cells (PBMCs) and serum were collected. Serum was cryopreserved and PBMCs were frozen until use in liquid nitrogen.
Three weeks after the second and fourth immunizations, a 6-mm skin biopsy was obtained from the most recent immunization site to determine the presence of p53-specific T cells. Biopsy tissue was manually cut into small pieces and infiltrating lymphocytes were expanded by homeostatic proliferation according to SOPs as reported previously.12, 25
Antigens used in immunological assays
Vaccine peptides were divided into four pools: p1-2 (aa 70–115), p3-4 (aa 102–155), p5-7 (aa 142–203) and p8-10 (aa 190–248). Memory recall mix (MRM), a mixture of tetanus toxoid (0.75 limus flocculentius/ml; Netherlands Vaccin Institute, Bilthoven, The Netherlands), tuberculine PPD (0.4 μg/ml; Netherlands Vaccin Institute, Bilthoven, The Netherlands) and C. albicans (0.015% Greenlabs, Lenoir), was used as a positive control.
Interferon-gamma ELISPOT assay
The interferon-gamma (IFN-γ) ELISPOT assay optimized to measure p53-specific T cell responses was performed according to SOPs as extensively described previously.12, 25–27 Cells were seeded in quadruplicates at 105 cells per well. Responses were considered p53-specific if [(mean number of spots in experimental wells) − (mean number of spots in medium + 2 × standard deviation, SD)] ≥ 10 spots/105 PBMCs. A response was considered to be vaccine-induced when (i) the p53-specific response exceeded the pre-existing immune response at least threefold26 or (ii) when a p53-specific response could be detected in a patient without a p53-specific response before immunization.
To evaluate proliferative capacity of freshly isolated PBMCs in response to stimulation with vaccine and nonvaccine peptides, a 6-day lymphocyte stimulation test (LST) was performed according to SOPs as previously described.12, 26 Skin biopsy-derived lymphocytes were analyzed for the presence of p53-specific T cells by stimulation with autologous monocytes pulsed overnight with the indicated vaccine peptides in a standard 3-day proliferation assay.12 Supernatants were cryopreserved. A proliferative response was considered p53-specific when corrected counts per minute (corrected cpm) ≥ 0 (corrected cpm = [(mean 3H-thymidine incorporation) − (mean 3H-thymidine incorporation in medium control + 3 × SD 3H-thymidine incorporation in medium control)]) and stimulation index (SI) ≥ 3 [SI = (mean of p53-induced proliferation)/(mean of medium control)].28 A vaccine-induced response was defined as (i) a p53-specific response with at least a twofold increase in corrected counts per minute compared to preimmunization levels or (ii) a p53-specific response in a patient without p53-specific response before immunization.
Cytokine bead array
To evaluate production of interleukin (IL)-2, IL-4, IL-5, IL-10, IFN-γ and tumor necrosis factor-alpha (TNF-α), we analyzed supernatants of proliferation assays by cytokine bead array (LINCOplex kit, Linco Research, St. Charles, MO) as described earlier.12, 25 P53-specific cytokine production was defined as concentration of cytokine minus concentration of medium control ≥ 100 pg/ml (IFN-γ) or ≥ 10 pg/ml (other cytokines) and concentration ≥ 2 times medium control. A vaccine-induced response was defined as a p53-specific postimmunization concentration ≥ 2 preimmunization concentration.
Both PBMCs and skin biopsy-derived lymphocytes were evaluated for CD3, CD4, CD8, CD19, CD25, CD56 (IQ Products, Groningen, The Netherlands) and FoxP3 (eBioscience, San Diego, CA) by flow cytometry [fluorescence-activated cell sorting (FACS)Calibur from BD Biosciences, Erembodegem, Belgium] according to the manufacturer's instructions.
Treg suppression assay
To evaluate Treg function, we used the “Treg Suppression Inspector” (Miltenyi Biotec, Utrecht, The Netherlands) designed for the functional characterization of human CD4+CD25+ T cells by in vitro suppression assays.29, 30 CD4+CD25− responder T cells were cocultured with CD4+CD25+ T cells, isolated with the “CD4+CD25+ Regulatory T Cell Isolation Kit” (Miltenyi) according to the manufacturer's instructions, in three different responder cell:suppressor cell ratios (R:S ratio) [1:0.1 (105:104), 1:0.2 and 1:0.3, respectively]. For T cell stimulation of both the CD4+CD25− responder T cells and the CD4+CD25+ T cells, the Treg Suppression Inspector (CD2, CD3 and CD28 antibodies) was added to the culture. As control, CD4+CD25− responder T cells alone were cultured without any stimulus. Proliferation of T cells was determined by measuring 3H-thymidine incorporation. Suppression is evaluated by the percentage of maximal proliferation, which was calculated as [mean suppressed responder cells (multiple R:S ratios)]/(mean responder cells) × 100%. The isolated CD4+CD25+ T cells were analyzed for FoxP3 positivity, if sufficient material was available for analysis.
Clinical responses were monitored by serum CA-125 measurements at screening, before each delivery of cyclophosphamide and 3 weeks after the last immunization, combined with computerized tomography (CT) at screening and after study completion. Clinical activity was evaluated according to Gynecologic Cancer InterGroup criteria31 by combining serum CA-125 levels with CT performed 6–9 weeks after the last immunization and evaluated according to response evaluation criteria in solid tumors.32 Clinically responding patients are defined as patients with either stable disease, partial or complete clinical response after treatment.
For the evaluation of safety, severity of adverse events was graded according to the Common Terminology Criteria (CTC) for Adverse Events v3.0.33 Relationship to the p53-SLP vaccine and/or cyclophosphamide was evaluated for all adverse events. A full blood count with differential and serum biochemistry was obtained before each gift of cyclophosphamide. The clinical monitoring was similar to our previous study.12
The number of patients needed to achieve sufficient power to exclude the absence of a clinical response (one-proportion test with alpha level 5%, one-sided) was calculated as 19. Furthermore, with 19 patients included in the study, the upper limit of the 95% exact confidence interval of a true response rate in the absence of clinical responses in a trial is less than 15%, which is the clinical response rate of most drugs registered for the treatment of recurrent ovarian cancer.34 Unfortunately, because of the outcome of the interim analysis, the study stopped at inclusion of ten patients. Therefore, we were unable to determine our secondary objective clinical efficacy. Differences between preimmunization and postimmunization were tested for normality by the Shapiro–Wilk test. Normally distributed data (p ≥ 0.05) were evaluated using a T-test for paired comparisons, and when normality was rejected (p < 0.05) a Wilcoxon's signed-ranks test was used. The remainder of tests was similar to the tests used in our previous study.12 Statistical significance was defined as p < 0.05. Statistical Analysis System version 9.1 (SAS Institute, Cary, NC) was used for all analyses.
- Top of page
- Material and Methods
- Supporting Information
Informed consent was obtained from 12 patients. One patient (P106) failed screening because of rapid progressive disease (Fig. 1 in the Supporting Information Appendix). Consent was withdrawn for personal reasons by one patient (P104) after receiving a single immunization preceded by cyclophosphamide. According to the protocol, this patient was used for the safety evaluation only. Two patients (P101 and P108) withdrew prematurely from the study because of progressive disease after three and four immunizations combined with cyclophosphamide, respectively. These patients were evaluable for both safety and efficacy analyses consistent with our protocol. All other patients (n = 8) successfully completed the study. Overall, ten patients were qualified for clinical and immunological analyses. P53 overexpression in the primary tumor was demonstrated by immunohistochemical staining in 45.5% (5/11) of patients (Table 1).
P53-specific T cell responses analyzed by IFN-γ ELISPOT
Before immunization, responses against vaccine peptides were present in patients P108 [26 specific spots/105 PBMCs (p1-2)], P111 [62, 90 and 44 specific spots/105 PBMCs (p1-2, p3-4 and p5-7), respectively] and P112 [20 specific spots/105 PBMCs (p3-4)]. None of these pre-existing p53-specific responses were boosted by the vaccine (Table 1a in the Supporting Information Appendix). Vaccine-induced IFN-γ-producing p53-specific T cells, however, could be detected in 90% (9/10) of patients after two immunizations (Table 1a in the Supporting Information Appendix) and in 87.5% (7/8) of the evaluable patients who received all four immunizations. The strength of the vaccine-induced p53-specific response was robustly enhanced to multiple epitopes within the vaccine as demonstrated by the enhanced reactivity against p1-2 (p = 0.039), p3-4 (p = 0.002), p5-7 (p = 0.004) and p8-10 (p = 0.001) after two immunizations. The p53-specific response to p1-2 was the weakest reaction (Fig. 1a). After four immunizations, the response against peptide pools p5-7 and p8-10 (p = 0.039 and p = 0.010, respectively) was significantly higher than the preimmunization levels. The number of circulating IFN-γ-producing p53-specific T cells remained relatively stable throughout immunizations as no difference in strength of the response against individual peptide pools was observed after the second and fourth immunizations (p ≥ 0.236).
Comparing preimmunization to postimmunization levels, the responsiveness to the mix of recall antigens (MRM) decreased (p = 0.016; Table 1a in the Supporting Information Appendix). This indicates that changes in p53 reactivity resulted from p53-SLP immunization and did not reflect a generalized increase in responsiveness of the adaptive immune system.
Proliferation assay analysis of P53-specific T cell responses in PBMCs
Pre-existing p53-SLP vaccine-peptide specific proliferative responses were detected in patients P103 (SI 4, 18, 12 and 4 against each subsequent peptide pool), P105 [SI 9 (p5-7)], P108 [SI 5 (p5-7)] and P111 [SI 5 and 3 (p5-7 and p8-10), respectively]. The pre-existing responses in patients P103 (p1-2, p5-7 and p8-10) and P108 (p5-7) were boosted upon two immunizations (Table 1b in the Supporting Information Appendix). The remainder of pre-existing responses was not boosted. After two immunizations, vaccine-induced p53-specific responses against vaccine peptides were observed in 80.0% (8/10) of patients and after four immunizations in 62.5% (5/8) of patients (Table 1b in the Supporting Information Appendix).
Similar to the results of the IFN-γ ELISPOT, the strength of the vaccine-induced p53-specific response was robustly enhanced after two immunizations in comparison to preimmunization levels (p1-2: p = 0.045; p3-4: p = 0.004; p5-7: p = 0.006 and p8-10: p = 0.004). The strength of the p53-specific proliferative response against peptide pools p3-4, p5-7 and p8-10 was higher than responses against peptide pool p1-2 (p ≤ 0.017). The highest responses were detected against p5-7 (Fig. 1b). Compared to preimmunization, the response after four immunizations only was increased against peptide pool p8-10 (p = 0.022).
The proliferative capacity of p53-specific T cells after four immunizations was a bit weaker than after two vaccinations albeit that this difference was not statistically different when the proliferation against individual peptide pools was compared between the second and fourth immunizations (p ≥ 0.068). The proliferative responsiveness to MRM remained stable over time (p = 0.994; Table 1b in the Supporting Information Appendix).
T-helper 1 and T-helper 2 cytokine characterization by cytokine bead array
P53-specific proliferation of PBMCs coincided with the production of both T-helper (Th)1 and Th2 cytokines (Fig. 2). Th1 cytokines (median [range] IL-2 1.9 [0.0–3.9]; IFN-γ 40.6 [0.0–190.5] pg/ml; TNF-α 250.9 [0.0–747.2] pg/ml) were detected in p53 peptide-stimulated PBMCs isolated after four immunizations. The same holds true for Th2 cytokines (IL-4 1.7 [0.0–2.6] pg/ml; IL-5 272.8 [0.0–495.3] pg/ml; IL-10 46.1 [23.6–124.3] pg/ml). The cytokine levels increased after the second immunization for IFN-γ, IL-10, IL-4 and IL-5 (p = 0.016, p = 0.009, p = 0.043 and p = 0.016). After four immunizations, the cytokine levels for IFN-γ and IL-10 (p = 0.016 and p = 0.014, respectively) decreased when compared to those measured after two immunizations, but remained higher than preimmunization levels.
FACS analysis of major lymphocyte subpopulations in PBMCs
CD3+, CD4+ and CD8+ T cells increased throughout immunizations (CD3+: p ≤ 0.017, CD4+: p ≤ 0.022 and CD8+: p = 0.020) in comparison to preimmunization levels, and CD19+ and CD56+ lymphocytes decreased after immunizations (CD19+: p ≤ 0.008 and CD56+: p ≤ 0.021) compared to preimmunization levels (Fig. 2 in the Supporting Information Appendix). The increase of CD3+ and CD4+ cells throughout immunizations was not observed in all analyzed patients but was predominantly observed in patients P101, P111 and P112 who had an increase of more than 25% (data not shown). No change in the percentage of CD4+FoxP3+ T cells were observed comparing PBMCs at preimmunization and after each subsequent delivery of cyclophosphamide (p ≥ 0.118; Fig. 3a).
Qualitative analysis of Tregs
To evaluate Treg function we performed a qualitative analysis using the PBMCs isolated before and 3 weeks after the last dose of cyclophosphamide from patients P111 and P112. The purity of CD4+CD25− T cells and CD4+CD25+ T cells analyzed was 97% in both cases, as assessed by flow cytometry. In the samples analyzed, CD4+CD25+ T cells indeed were FoxP3 positive. We observed a decreased percentage of maximal proliferation of 70, 50 and 40%, respectively, with increasing R:S ratios (Fig. 3b). We observed no difference in percentage of maximal proliferation before treatment or after four immunizations preceded by four cycles of cyclophosphamide.
P53-specific T cell migration in the skin analyzed by proliferation assays
To analyze the migratory capacity of p53-specific T cells to sites where p53 antigen is presented, we cultured lymphocytes from skin biopsies taken from the second and fourth injection sites (n = 9 and n = 7, respectively). After four immunizations, a difference in proliferating T cells is observed between different peptide pools, i.e., responses against p5-7 and p8-10 are higher than p1-2 and p3-4 (p = 0.002; Fig. 1c). One patient (P111) displayed very high responses only against peptide pool p3-4 after two immunizations (Fig. 1c). Proliferation against p1-2 and p8-10 was more pronounced in biopsies taken from the fourth vaccination site when compared to what was observed in the biopsies from the second vaccine site (p = 0.028). Vaccine-site infiltrating p53-specific T cells were detected in 44.4% (4/9) of biopsies taken from the second vaccination site and in 85.7% (6/7) of the fourth vaccination site. Phenotyping of the infiltrating lymphocytes from these biopsies by flow cytometry revealed a preponderance of CD3+ T cells [mean ± standard error of the mean: 64.4% ± 9.8 and 73.8% ± not applicable (n = 1), second vs. fourth vaccination site, respectively], of which 52.5% ± 8.5 and 53.5% ± 7.0 were CD4+, 23.0% ± 6.2 and 18.0% ± 2.7 were CD8+, respectively. CD4+FoxP3+ cells were not detected among the infiltrating lymphocytes.
Two patients (20%) had stable disease as evaluated by CA-125 and CT scan (P108, P109). In both patients, vaccine-induced p53-specific responses were present. The other patients (8/10; 80%) had clinical, biochemical and/or radiographic evidences of progressive disease. Seven patients had decreasing or stable CA-125 levels during the study (P101, P103, P107, P108, P109, P110 and P111; Fig. 4a). The strength of the p53-specific responses measured by ELISPOT and LST after two and four immunizations from these seven patients was not significantly different from the patients showing rising CA-125 levels (p > 0.293; Fig. 4b).
No vaccine-related CTC grade 3 or 4 adverse events were observed (Table 2 in the Supporting Information Appendix). Patients P101 and P108 developed ileus and omental metastases, respectively, because of disease progression for which they were hospitalized. No major vaccine- or cyclophosphamide-induced alterations were observed in serum biochemistry and complete blood count. Mild to moderate inflammatory symptoms at the injection site occurred in nearly all patients; redness and swelling lasted 15–16 days, and pain and itch 4–5 days on average. (Re)activation of local inflammatory symptoms at prior injection sites was observed in the majority of patients (90%) after subsequent immunizations. We did not find evidence of vaccine-induced autoimmunity, assessed by clinical symptoms and serum antinuclear antibody levels (data not shown).
Results of p53-SLP preceded by cyclophosphamide compared to p53-SLP alone
Analyses of the clinicopathological stage, histology, grade, residual disease, prior courses of chemotherapy, evidence of disease at inclusion and p53 overexpression of primary tumor revealed no differences between patients treated in the current and the previous study, suggesting that the two different patient groups are similar and to some extent comparable (Table 3 in the Supporting Information Appendix).
Notably, analysis of the results obtained by IFN-γ ELISPOT revealed that the reactivity against each individual peptide pool differed between the current and the previous study (p ≤ 0.012; Fig. 5a). Stronger reactivity was found against all peptide pools after two and four immunizations, except for p1-2 (p = 0.060) after two immunizations. In the previous study, a strong reduction was observed in the number of circulating IFN-γ-producing p53-specific T cells after four immunizations. This reduction, however, was not observed in our study, rather responses remained stable. Comparison of the cumulative response to peptide pools between the current and previous trial showed that the reactivity, as reflected in median number of cumulative spots, differed between the two trials at two immunizations [peptide pools p1-10 median (interquartile range, IQR) current study: 329 (242–388); previous study: 70 (42–205)] as well as at four immunizations [peptide pools p1-10 median (IQR) current study: 285 (102–698); previous study: 40 (5–82); Fig. 5b].
No differences were observed between the two trials with respect to the proliferative capacity of p53-specific T cells as measured by LST (data not shown) or in the amounts of p53-specific proliferation-associated cytokine production (Fig. 5c).
- Top of page
- Material and Methods
- Supporting Information
Combination of low-dose cyclophosphamide with a p53-SLP vaccine induced no quantitative or qualitative reduction of Tregs in ten patients with advanced-stage epithelial ovarian cancer. However, we did observe a strong Th1/Th2 immune response to p53 in this phase II single-arm study. This shows that the use of low-dose cyclophosphamide 2 days before each p53-SLP immunization does not impair de novo induction of vaccine-induced p53-specific Th cell immunity. Furthermore, the number of vaccine-induced p53-specific IFN-γ-producing T cells was higher in our study, compared to a study in which a similar patient group was treated with p53-SLP monotherapy. Moreover, the strong reduction in the number of circulating p53-specific T cells observed previously after four immunizations was absent in our study. Our results warrant new studies on the use of low-dose cyclophosphamide to potentiate the immunogenicity of antitumor vaccines.
Cyclophosphamide in dosages half or less the equivalent of those used for chemotherapy in combination with immunotherapy have been described to decrease Treg numbers and impair their function.20, 35 Cyclophosphamide appears to be most effective when administered several days before immunization.36, 37 Contrary to our expectations, we observed that CD4+FoxP3+ T cells were not depleted by addition of 300 mg/m2 i.v. cyclophosphamide 2 days before the p53-SLP vaccine, nor did cyclophosphamide affect the in vitro suppressive capacity of CD4+CD25+ suppressor cells. These observations are supported by Audia et al., who previously reported that cyclophosphamide combined with immunotherapy failed to reduce the frequency of Tregs or significantly modulate their function.38
Addition of cyclophosphamide to the p53-SLP vaccine resulted in induction of high numbers of p53-specific IFN-γ+-producing T cells against multiple p53 epitopes in the majority of patients with ovarian cancer. Induction of tumor-specific immune responses in patients with cancer treated with cyclophosphamide combined with specific vaccinations were also reported by others,39, 40 suggesting that targeting function and frequency of Tregs by cyclophosphamide in patients with cancer unmasks and potentially enhances tumor-specific T cell responses. When comparing the results obtained in this trial to the results of our previous trial, we observed that the number of IFN-γ-producing T cells was higher and remained more stable in the group of patients receiving the p53-SLP vaccine combined with cyclophosphamide, suggesting that cyclophosphamide treatment augmented Th1 reactivity. However, no difference was observed in the p53-specific proliferation and associated cytokine production of these proliferating cells, implying that rather than the expansion of p53-specific T cells their in vivo lifespan or function was altered. Recently, Ding et al. showed that pretreatment with cyclophosphamide may augment tumor-specific CD4+ T cell immunity by rescuing these cells from apoptosis through the prevention of programmed death-1 upregulation and IL-7R downregulation on CD4+ effector T cells.41 A similar phenomenon may play a role in our study.
In a clinical case report on NY-ESO-1-specific immunotherapy, it was shown that tumors regressed completely after adoptive transfer of NY-ESO-1-specific CD4+ T cells, even though NY-ESO-1 was not uniformly expressed by the tumor cells.42 This complete regression of tumor was explained by the induction of T cell responses to two other tumor-associated antigens displayed by the tumor, i.e., MART-1 and MAGE-3, after infusion of NY-ESO-1-specific CD4+ T cells.42 This additional response is likely to have been triggered by the NY-ESO-1 CD4+ Th cell, which induced activation of antigen-presenting dendritic cells that had ingested autogenic material from cancer cells. We observed that the p53-SLP vaccine strongly activates p53-specific CD4+ T cells but not CD8+ T cells. Analogous to the study described above, a strong p53-specific CD4+ T cell response may trigger the responses of CD8+ T cells to other ovarian cancer expressed antigens, e.g., the highly immunogenic Wilms' tumor protein 1 antigen.43, 44
Finally, our results fit in with the safety and immunogenicity experience gathered thus far with vaccines consisting of SLPs dissolved in Montanide ISA-51 adjuvant, showing only low-grade toxicity and strong immunogenicity. Despite the observed lack of influence on CD4+FoxP3+ T cells in our study, cyclophosphamide treatment, at this dose schedule, is a promising immune-potentiating strategy for anti-p53 vaccines. The information gained from this study will serve as a baseline for further clinical investigation to better define the full potential of this strategy to ultimately achieve antitumor immunity with clinical impact. Future studies will be needed to establish possibilities to induce robust antitumor immunity by induction of antigen-specific cytotoxic T lymphocyte alongside p53-SLP-induced p53-specific Th cells.
- Top of page
- Material and Methods
- Supporting Information
The authors thank all patients who participated in this study and the Trial Coordination Center of the UMCG for study monitoring and statistical analyses. C. Melief is a part-time employee of ISA Pharmaceuticals B.V., and J. Oostendorp receives funding for research projects from ISA Pharmaceuticals. The LUMC holds a patient on long peptide vaccines on which S.H. van der Burg is named as inventor.
- Top of page
- Material and Methods
- Supporting Information
- 19The International Collaboration Ovarian Neoplasm (ICON) group. Paclitaxel plus carboplatin versus standard chemotherapy with either single-agent carboplatin or cyclophosphamide, doxorubicin, and cisplatin in women with ovarian cancer: the ICON3 randomised trial. Lancet 2002; 360: 505–15.
- 31Gynecologic Cancer InterGroup (2005) GCIG. CA 125 definitions. Available at: http://ctcp.cancer.gov/resources/cig/respact.html.
- 32New guidelines to evaluate the response to treatment in solid tumors. European Organization for Research and Treatment of Cancer, National Cancer Institute of the United States, National Cancer Institute of Canada. J Natl Cancer Inst 2000; 92: 205–16., , , , , , , , , , .
- 33Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Available at: http://evs.nci.nih.gov/ftp1/CTCAE.
- 38Increase of CD4+ CD25+ regulatory T cells in the peripheral blood of patients with metastatic carcinoma: a Phase I clinical trial using cyclophosphamide and immunotherapy to eliminate CD4+ CD25+ T lymphocytes. Clin Exp Immunol 2007; 150: 523–30., , , , , , , , , , , , et al.
- 40Allogeneic granulocyte macrophage colony-stimulating factor-secreting tumor immunotherapy alone or in sequence with cyclophosphamide for metastatic pancreatic cancer: a pilot study of safety, feasibility, and immune activation. Clin Cancer Res 2008; 14: 1455–63., , , , , , , , , , , .
- Top of page
- Material and Methods
- Supporting Information
Additional Supporting Information may be found in the online version of this article.
|IJC_27388_sm_SuppAppendix.doc||293K||Supporting Information Appendix.|
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.