Long-term clinical and immunological effects of p53-SLP® vaccine in patients with ovarian cancer
Article first published online: 9 MAY 2011
Copyright © 2011 UICC
International Journal of Cancer
Volume 130, Issue 1, pages 105–112, 1 January 2012
How to Cite
Leffers, N., Vermeij, R., Hoogeboom, B.-N., Schulze, U. R., Wolf, R., Hamming, I. E., van der Zee, A. G., Melief, K. J., van der Burg, S. H., Daemen, T. and Nijman, H. W. (2012), Long-term clinical and immunological effects of p53-SLP® vaccine in patients with ovarian cancer. Int. J. Cancer, 130: 105–112. doi: 10.1002/ijc.25980
- Issue published online: 27 OCT 2011
- Article first published online: 9 MAY 2011
- Accepted manuscript online: 16 FEB 2011 03:12PM EST
- Manuscript Accepted: 26 JAN 2011
- Manuscript Received: 15 OCT 2010
- Dutch Cancer Society (Amsterdam, The Netherlands). Grant Number: 2007-3919
- ISA Pharmaceuticals B.V. (Bilthoven, The Netherlands)
- ovarian cancer
Vaccine-induced p53-specific immune responses were previously reported to be associated with improved response to secondary chemotherapy in patients with small cell lung cancer. We investigated long-term clinical and immunological effects of the p53-synthetic long peptide (p53-SLP®) vaccine in patients with recurrent ovarian cancer. Twenty patients were immunized with the p53-SLP® vaccine between July 2006 and August 2007. Follow-up information on patients was obtained. Clinical responses to secondary chemotherapy after p53-SLP® immunizations were determined by computerized tomography and/or tumor marker levels (CA125). Disease-specific survival was compared to a matched historical control group. Immune responses were analyzed by flow cytometry, proliferation assay, interferon gamma (IFN-γ) ELISPOT and/or cytokine bead array. Lymphocytes cultured from skin biopsy were analyzed by flow cytometry and proliferation assay. Of 20 patients treated with the p53-SLP® vaccine, 17 were subsequently treated with chemotherapy. Eight of these patients volunteered another blood sample. No differences in clinical response rates to secondary chemotherapy or disease-specific survival were observed between immunized patients and historical controls (p = 0.925, resp. p = 0.601). p53-specific proliferative responses were observed in 5/8 patients and IFN-γ production in 2/7 patients. Lymphocytes cultured from a prior injection site showing inflammation during chemotherapy did not recognize p53-SLP®. Thus, treatment with the p53-SLP® vaccine does not affect responses to secondary chemotherapy or survival, although p53-specific T-cells do survive chemotherapy.
Ovarian cancer, which is generally treated with cytoreductive surgery and platinum-based chemotherapy, is the most frequent cause of death from gynecological malignancies. In an attempt to improve prognosis by inducing and/or enhancing tumor immune responses, we have recently performed a Phase II study with the p53-synthetic long peptide (p53-SLP®) vaccine.1 The vaccine proved to be safe, well tolerated and highly immunogenic, but no partial and/or complete clinical responses were observed.
Likewise, many previous p53-based immunotherapeutic strategies have disappointing clinical efficacy, although p53-specific immunity was induced.2–7 Interestingly, in patients with small cell lung cancer, a trend toward an increased response to secondary chemotherapy was observed after immunization with dendritic cells virally transduced with the wild-type p53 gene.6 Complete responses or partial responses (CR/PR) to second-line chemotherapy were seen in 75% of p53-responders as opposed to 30% of p53-nonresponders. Moreover, this clinical response rate of 75% seen after second-line chemotherapy in patients with immunological responses to the p53-transduced dendritic cells6 is much higher than observed in historical control groups treated with second-line chemotherapy for progression of disease (6–16%).8
Patients with immunological responses to p53-specific immunotherapy may thus be more likely to respond to secondary chemotherapy. Possible explanations for this synergy include (i) upregulation of p53 in tumor cells in response to chemotherapy, thus increasing chances of recognition and destruction by cytotoxic T-cells, and (ii) downregulation of immunosuppressive agents produced by tumor cells, thus enhancing destruction of tumor cells by cytotoxic T-lymphocytes.
We hypothesized that patients treated with the p53-SLP® vaccine would have a higher response rate to “secondary” chemotherapy than generally described for palliative chemotherapy for ovarian cancer. Furthermore, we investigated whether p53-specific immunity previously induced by the p53-SLP® vaccine was influenced by secondary chemotherapy.
Patients and Methods
In a Phase II study, patients with epithelial ovarian cancer were subcutaneoulsy immunized four times with the p53-SLP® vaccine.1 The vaccine consisted of 10 synthetic long overlapping peptides, spanning amino acids (aa) 70–248 of the wt-p53 protein. Clinical response to immunizations was determined 6–9 weeks after the last immunization. Subsequent follow-up information for patients who participated in this Phase II study was prospectively collected. Patients treated with chemotherapy after immunization were invited to give a blood sample to measure the level of p53-specific immune responses. Written informed consent was obtained specifically for the collection of this additional blood sample.
Evaluation of long-term immunogenicity
Lymphocytes and sera
Blood for immunological assays was obtained at least 4 weeks after secondary chemotherapy. Serum was isolated from clotted blood and cryopreserved. Peripheral blood mononuclear cells (PBMCs) were isolated from heparinized blood by Ficoll-Paque density centrifugation and were freshly used and/or frozen until use in liquid nitrogen.
Antigens (for 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). Thirteen overlapping 30-mers spanning the first and last part of wt-p53 protein not included in the vaccine were divided into two pools: aa 1–78 and aa 241–393. Memory recall mix (MRM), a mixture of tetanus toxoid (0.75 limus flocculentius per milliliter; Netherlands Vaccine Institute, Bilthoven, The Netherlands), tuberculine PPD (0.4 μg/mL; Netherlands Vaccine Institute) and C. albicans (0.015% Greenlabs, Lenoir), was used as a positive control.
IFN-γ ELISPOT assay
Cryopreserved PBMC obtained before immunization, after immunization and after subsequent chemotherapy were available for seven patients. P53-specific responses in cryopreserved PBMC obtained before immunization, after four immunizations and after chemotherapy of a single patient were simultaneously determined by interferon gamma (IFN-γ) ELISPOT as previously described.5, 9 PBMC were stimulated with vaccine and non-vaccine p53-peptide pools (10 μg/peptide/mL) or MRM (1:50). A response was considered p53-specific if [(mean number of spots in experimental wells) − (mean number of spots in medium + 2× SD medium)] ≥ 10 spots per 105 PBMC, where SD is the standard deviation. A vaccine-induced response was defined as a p53-specific response that exceeded the pre-existing immune response at least threefold.10
PBMC were evaluated for CD3, CD4, CD8, CD19, CD56 (IQ Products, Groningen, The Netherlands) and FoxP3 expression (eBioscience, San Diego, CA) by flow cytometry (FACSCalibur from BD Biosciences, Erembodegem, Belgium) according to the manufacturer's instructions. Skin biopsy-derived lymphocytes were analyzed for CD4, CD8 and Foxp3 expression.
Freshly isolated PBMC were stimulated with vaccine and non-vaccine p53-peptide pools (10 μg/peptide/mL) or MRM (1:156) as previously described.10 Proliferation was considered p53 specific if the stimulation index ≥ 3 and [(mean cpm in experimental well) – (mean cpm in medium + 3× SD medium)] ≥ 0, where cpm is counts per minute. A vaccine-induced response was defined as a p53-specific response with [(mean cpm in medium + 3× SD medium after treatment)/(mean cpm in medium + 3× SD medium before immunization)] ≥ 2.
Cytokine bead array
Production of Th1 cytokines interleukin (IL)-2, IFN-γ and tumor necrosis factor alpha (TNF-α) and Th2 cytokines IL-4, IL-5 and IL-10 was evaluated for three patients. Supernatants of proliferation assays were used in a cytokine bead array (LINCOplex Kit; Linco Research, St. Charles, MO) according to a previously validated standard operating procedure.9 p53-specific cytokine production was defined as concentration of cytokine ≥ 2 × concentration of medium control and if concentration ≥ 100 pg/mL (IFN-γ), or ≥ 10 pg/mL (other cytokines).
Evaluation of long-term clinical activity
Historical control group
To evaluate the effect of p53-SLP® treatment on subsequent chemotherapy and survival, a historical control group was formed with three control patients for each patient treated with the p53-SLP® vaccine. Historical controls were obtained from an anonymous password-protected database containing clinicopathological and follow-up data of all patients with epithelial ovarian cancer treated with primary debulking surgery according to standard treatment protocols by gynecological oncologists of the University Medical Center Groningen (Groningen, The Netherlands) between May 1985 and May 2006. As no patient's identity can be eluded from this computerized database, further approval from our Institutional Review Board was not required for the use of these historical control patients according to the Dutch law. p53-SLP®-treated patients were matched with historical controls based on FIGO stage, histological tumor type, amount of residual disease after primary debulking surgery and histological grade (in the order of importance).
Evaluation of clinical responses to chemotherapy
For historical control patients, tumor response to second-line chemotherapy was evaluated based on CA-125 levels and reports of imaging when available. For p53-SLP®-treated patients, tumor response to postimmunotherapy chemotherapy was evaluated by serum CA-125 levels (GCIG criteria11) and comparison of computerized tomography before and after chemotherapy assessed according to RECIST criteria12 by an experience radiologist (RW)).
Differences in PBMC composition of cryopreserved samples and cytokine levels were evaluated with the Wilcoxon signed rank test. Whether matching resulted in similar distributions of clinicopathological characteristics was evaluated by likelihood ratio statistics (categorical variables) or independent samples t-test (normally distributed continuous variables). Differences between historical controls and p53-SLP® vaccine-treated patients in response rates to chemotherapy for first recurrence were evaluated with χ2 test. Survival differences were plotted using Kaplan–Meier curves and tested by log-rank test. Disease-specific survival was defined as time of diagnosis to date of death due to ovarian cancer, treatment-related fatalities or last follow-up.
All analyses were performed using SPSS version 16.0.2 software package for windows (SPSS, Chicago, IL). p values < 0.05 were considered significant (tested two-sided).
Twenty patients were treated with the p53-SLP® vaccine.1 Three patients did not receive subsequent chemotherapy (P01, P12 and P14). Responses to p53-SLP treatment and postimmunotherapy chemotherapy are shown in Table 1. Eight patients consented to the donation of an additional blood sample after chemotherapy. No differences in clinicopathological characteristics of immunized patients and matched historical controls were observed (Table 2).
Persistent p53-specific proliferative T-cell responses after chemotherapy
p53-specific proliferative responses were observed in 5/8 patients (63%) after postimmunotherapy chemotherapy as measured by proliferation assay (Fig. 1). In three patients (P17, P20 and P23), responses present after the fourth immunization were no longer detectable after subsequent chemotherapy, whereas the opposite held for P09. Persisting responses were frequently more pronounced after chemotherapy than after the fourth immunization. Interestingly, after chemotherapy proliferative responses to aa 241–393, part of the p53 protein not covered by the p53-SLP® vaccine, increased in number as well as strength.
An IFN-γ ELISPOT was performed for seven patients to evaluate whether these p53-specific cells were IFN-γ-producing Th1-type cells (Fig. 2). Comparable to the previously published results,1 no differences existed in composition of cells obtained before immunization, after immunization and after chemotherapy as analyzed simultaneously by flow cytometry (data not shown). After chemotherapy, two patients (29%) showed p53-specific IFN-γ production in response to stimulation of PBMC with vaccine peptides p3-p4 and p8-p10 (P06 and P20), whereas no such responses were observed in these patients after four immunizations. In patients with responses against vaccine peptides after immunotherapy (P11, P17 and P22), no IFN-γ-producing T-cells could be detected after subsequent chemotherapy. Therefore, IFN-γ production by cells that proliferate in response to stimulation with p53-specific peptides does not seem to be augmented by chemotherapy after immunotherapy with the p53-SLP® vaccine.
Th1/Th2 cytokine production is not altered by postimmunization chemotherapy
Th1 and Th2 cytokine production was evaluated in three patients. p53-specific cytokine production was observed at all time points. Although overall vaccine-induced cytokine concentrations do not seem to be different between postimmunotherapy and postchemotherapy samples in this small sample group of patients, p53-specific production of Th2 cytokines (i.e., IL5 and IL-10) as well as Th1 cytokine TNF-α seemed to be more common after postimmunotherapy chemotherapy (Fig. 3).
Reactivation of injection sites is not caused by p53-specific T-lymphocytes
Reactivation of prior injection sites during chemotherapy was reported by several patients (Fig. 4). For two patients (P19 and P22), an additional blood sample was obtained for IFN-γ ELISPOT at the time of reactivation during chemotherapy for a second recurrence, in addition to the sample obtained after the first course of postimmunotherapy chemotherapy. Similar to the results after the first course of postimmunotherapy, no p53-specific responses were observed for P19 at the time of reactivation of the vaccine injection sites. Although P22 had p53-specific responses against p3-p4, p5-p7 and p8-p10, only responses against p8-p10 could be considered vaccine induced (43, 57 and 189 specific spots per 105 cells, respectively).
P22 also consented to a skin biopsy taken from a reactivated immunization site. This biopsy was used for the culture of T-lymphocytes as previously described,1 yielding 12.3 × 106 lymphocytes after 4 weeks of culturing (CD4+ 15%, CD8+ 5%). Cells were subsequently used for a 6-day proliferation assay. The lymphocytes cultured from the biopsy taken from the reactivated injection site were not p53-specific, indicating that reactivation of prior injection sites is not a p53-specific event (data not shown).
Long-term clinical activity
Information on clinical response to second-line chemotherapy was available for 30 historical controls. Response rates to second-line chemotherapy were similar for p53-SLP® vaccine-treated patients and historical controls (CR/PR 60.0% vs. 61.5%, p = 0.925). Likewise, median disease-specific survival did not differ between p53-SLP®-treated patients with ovarian cancer and historical controls (median 44.0 vs. 47.4 months, p = 0.601). Analyses of only those patients who received all four p53-SLP SLP® immunizations or exclusion of patients not treated with chemotherapy after immunotherapy did not result in differences in survival either.
The p53-SLP® vaccine was recently shown to induce p53-specific T-cell responses in patients with ovarian cancer.1 We investigated long-term immunological and clinical effects of the p53-SLP® vaccine as it has been suggested that responses to chemotherapy might improve in patients with vaccine-induced immune responses.6 Despite the presence of p53-specific immune responses in patients treated with the p53-SLP® vaccine, neither clinical responses to chemotherapy for recurrent disease nor survival differed from the response rates and survival of historical controls. After chemotherapy for recurrent disease subsequent to p53-SLP® immunotherapy, p53-specific immune responses could be detected in 6/8 patients willing and able to give an additional blood sample. Moreover, after chemotherapy, epitope spreading was observed as proliferative responses to parts of the p53 protein not covered by the p53-SLP® vaccine increased in number as well as strength. Although reactivation of inflammatory reactions at prior injection sites during chemotherapy for a second recurrence was accompanied by circulating p53-specific T-cells recognizing p8-p10 in one patient, no locally active p53-specific T-cells could be detected.
A trial of p53 immunization in patients with small cell lung cancer showed a trend toward improved response rates to secondary chemotherapy for immunological responders.6 As there was no distinct group of non-responders in our previous trial,1 a similar comparison between immunological responders and non-responders could not be made. We therefore compared clinical response rates to chemotherapy for recurrent disease with a historical control group. No differences in response rates to chemotherapy were observed, which may be attributed to several causes. First, although an attempt was made to match p53-SLP®-treated patients with historical controls based on some well-known prognostic factors, such retrospective comparisons remain prone to selection bias. To obtain a truly reliable insight in the effect of p53-SLP® treatment on chemotherapy and survival, a randomized controlled trial should be performed with patients allocated to p53-SLP treatment or a control arm receiving no or placebo treatment. Second, one could argue that the likelihood of improved clinical responses after p53-SLP® treatment is limited as immunization with the p53-SLP® vaccine was shown to induce predominantly Th2 CD4+ T-cells, which are less likely to contribute to effective antitumor responses than Th1 CD4+ T-cells.1 Moreover, after chemotherapy, the number of patients with IFN-γ-producing p53-specific T-cells was lower than the number of patients with p53-specific proliferating T-cells. Although the latter results do not reach statistical significance in this small study, it is noteworthy that proliferative responses were frequently more pronounced after chemotherapy. Within the limitations of our study, the data suggest that especially IFN-γ-producing Th1 CD4+ T-cells are prone to destruction by chemotherapy, whereas Th2 CD4+ T-cells seem to be less affected. Such an effect of chemotherapy on Th1 and Th2 T-cells was previously described for patients with breast cancer.13 Corroborating cytokine profiles were observed in the our study, with increased frequency of Th2 cytokine production after chemotherapy.
A separate, although related observation was the rise in circulating p53-specific T-cells after chemotherapy. Next to boosting of proliferative responses by chemotherapy, this rise may be attributable to the loss of activation markers necessary for extravasation and recruitment of lymphocytes to affected tissues. This would result in increased numbers of circulating p53-specific lymphocytes, which corresponds to observations in patients with cervical cancer in whom responses increased in time without additional immunizations or cytotoxic treatment.10
Interestingly, several patients reported reactivation of immunization sites during chemotherapy for recurrent disease. p53-specific immune responses in blood samples obtained were observed in one of two patients. In a skin biopsy of the injection site from this patient, no p53-specific T-cells could be detected. This suggests that the inflammatory reactions at the immunization sites during chemotherapy may be immune responses enhanced by Montanide ISA51, the adjuvant used in the p53-SLP® vaccine. Next to its depot function, this water-in-oil emulsion allows slow release of antigens and has been reported to promote inflammation (innate immune responses) and recruitment of antigen-presenting cells as well as lymphocytes (adaptive immune responses).14
In summary, we show that vaccine-induced p53-specific T-cells can still be detected after chemotherapy. Our results indicate that immunotherapy of recurrent ovarian cancer with the p53-SLP® vaccine does not affect responses to subsequent chemotherapy or prognosis.
The authors thank all patients who were willing to donate additional blood samples and/or skin biopsies. C. Melief is a part-time employee of ISA Pharmaceuticals B.V., license holder of the p53-SLP® vaccine.
- 2Wild-type p53 peptide vaccine can generate a specific immune response in low burden ovarian adenocarcinoma. Proc Am Soc Clin Oncol 22: 2003 (abstr 678). Abstract 660., , , , , , , , , .
- 4Vaccination with p53 peptide-pulsed dendritic cells is associated with disease stabilization in patients with p53 expressing advanced breast cancer; monitoring of serum YKL-40 and IL-6 as response biomarkers. Cancer Immunol Immunother 2007; 56: 1485–99., , , , , , , , , , .
- 12New 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., , , , , , , , , , .
- 14Function of carriers and adjuvants in induction of immune responses. In: Nervig RM, Gough PM, Kaeberle ML, Whetstone CA, eds. Advances in carriers and adjuvants for veterinary biologics. Ames, IA: Iowa State University Press, 1986. 11–23..