Combination of p53 Cancer Vaccine with Chemotherapy in Patients with Extensive Stage Small Cell Lung Cancer

Materials and Methods

Clinical protocol. Before enrolling patients, the protocol was reviewed and approved by the Food and Drug Administration (BB-IND 9792), the NIH Office of Biotechnology Activities' Recombinant DNA Advisory Committee (OBA no. 0205-538), the University of South Florida Institutional Review Board, and the University of South Florida Institutional Biosafety Committee. Patients aged 18 or older with a histologic diagnosis of ES SCLC were eligible to participate. Eastern Cooperative Oncology Group performance status of 0 to 2, and adequate organ function (WBC, >3,000/mm³; ANC, >1,500/mm³; platelets, >100,000/mm³; hematocrit, > 25%; bilirubin, <2.0 mg/dL; and creatinine, <2.0 mg/dL) were required. Patients with a preexisting autoimmune disorder, an immunodeficiency condition, a serious ongoing infection, or uncontrolled brain metastases were not eligible.

All patients were treated with conventional cytotoxic chemotherapy prior to receiving the investigational vaccine. Eighteen patients received carboplatin/VP-16, eight received cisplatin/VP-16, and three received cisplatin/CPT-11. Patients who had progressive disease after chemotherapy were eligible if they otherwise met all other inclusion criteria. At least 8 weeks after the last dose of chemotherapy, the patients underwent leukapheresis. Vaccines were produced, and administered by intradermal injection at four separate sites that drain to the bilateral axillary and inguinal lymph node basins. This was repeated on three separate occasions every 2 weeks. Two weeks after the third set of vaccines, the patients were re-staged. Those patients who did not exhibit progressive disease at this point underwent a second leukapheresis procedure, and received three additional sets of vaccines, this time every 4 weeks. Patients who developed progressive disease after the third or sixth vaccination were offered additional cytotoxic chemotherapy.

Vaccine production. Mononuclear cells for DC production were obtained after leukapheresis and kept stored in liquid nitrogen. After thawing, cells were placed in X-VIVO-15 medium (Biowhittaker, Walkersville, MD) in tissue culture flasks at a concentration of 1.3 to 1.7 × 10⁶ cells/cm² of available culturing surface. After 2 hours of culture, nonadherent cells were removed and the flasks were recharged with X-VIVO-15 medium supplemented with 5 ng/mL granulocyte macrophage colony-stimulating factor (Amgen, Thousand Oaks, CA) and 5 ng/mL interleukin-4 (R&D Systems, Minneapolis, MN). The flasks were incubated for 48 hours, at which time additional cytokine-supplemented medium was added to the flasks. The flasks were then incubated for an additional 72 hours. At the completion of incubation, the nonadherent and loosely adherent cells were collected and used for 2-hour infection with Ad-p53 at a viral particle to cell ratio of 15,000:1. This optimal dose of adenovirus that would produce the highest level of human p53 expression with the least amount of toxicity to the dendritic cells was determined in preliminary experiments. At the end of the 2-hour incubation, X-VIVO medium was added to a final cell concentration of 10⁶ cells/mL, and cells were incubated in flasks for an additional 46 hours, at which time the cells were harvested, washed, and analyzed. Vaccine release criteria includes, (a) negative Gram's staining, (b) negative Mycoplasma test by PCR analysis, (c) maximum endotoxin concentration of 5 EU/mL, and (d) a mature DC phenotype with evidence of intracellular p53 expression by flow cytometry analysis. Mature DC phenotype was defined as lineage (CD3, CD14, CD19, CD20, and CD56)-negative, HLA-DR-positive, and CD86-positive cells. To determine p53-positive DCs, cells were treated with “fix and perm” reagent (Caltag, Burling, CA). Staining with the p53-specific antibody was followed by staining with a phycoerythrin-labeled antimurine immunoglobulin antibody. After washing off excess antibody, surface staining for lineage markers with FITC-conjugated monoclonal antibodies, and for HLA-DR with phycoerythrin-conjugated monoclonal antibodies was done. The final product was analyzed by flow cytometry.

Analysis of IFN-γ-producing cells in ELISPOT assays. Peripheral blood mononuclear cells were collected from patients prior vaccination, 2 to 3 weeks after completion of the third vaccination (postvaccine time point), and then 2 months later. Samples were kept in aliquots in liquid nitrogen. Samples from each patient were thawed and analyzed simultaneously. ALVAC-p53, a recombinant canarypox virus containing full-length wild-type p53 was obtained from Aventis Pasteur (Toronto, Canada). Mononuclear cells were infected with ALVAC-p53 or ALVAC-control (empty vector) for 2 hours in serum-free medium at a multiplicity of infection of four plaque-forming units per cell. After infection, cells were seeded in quadruplicate in complete culture medium supplemented with interleukin-2 (1 × 10⁵ cells/well) in 96-well plates precoated with an anti-IFN-γ antibody and incubated for 36 hours. The number of IFN-γ-producing cells was evaluated using an automated ELISPOT reader (CTL) as described previously (9, 19). The mononuclear cells from HLA-A2-positive patients were incubated in parallel for 36 hours with 10 μg/mL of either a p53-derived HLA-A2-binding peptide (LLGRNSFEV) or a control PSA-derived peptide (FLTPKKLQCV). The number of IFN-γ-producing cells were evaluated in ELISPOT assays as described previously (19).

Tetramer staining. Tetramer HLA-A-0201/LLGRNSFEV was produced by the National Institute of Allergy and Infectious Diseases MHC tetramer core facility at Yerkes Regional Primate Research Center. Mononuclear cells were stained for 60 minutes at 4°C with an antigen-presenting cell–conjugated anti-CD8 antibody and the phycoerythrin-conjugated tetramer (1:100 dilution). As a control, we used a tetramer containing HLA-A2 peptide from the melanoma antigen gp100 (KTWGQYWQV). The proportion of tetramer-positive cells within the CD8⁺ T cell population was calculated.

Evaluation of humoral immune response. The levels of anti-p53 and antiadenovirus antibodies (IgG and IgM) were evaluated in ELISA using at least four serial dilutions. Internal controls provided by the manufacturers were used to establish a “cutoff” level. Samples were always assayed in duplicate. The absorbance was read on a spectrophotometer at a wavelength of 450 nm against a reference filter of 620 nm in order to compensate for differences in the material of the microtiter plate. The p53-Autoantibody Elisa^PLUS kit (Calbiochem, San Diego, CA) was used to measure circulating antibodies to p53 in human serum samples. Adenovirus IgG/IgM ELISA kits were purchased from IBL Immuno-Biological Laboratories (Hamburg, Germany).

Phenotype and function of T cells and DCs. After thawing, the mononuclear cells were cultured overnight in complete culture medium supplemented with 10% FCS. Cell phenotype was evaluated by multicolor flow cytometry using a FACSCalibur flow cytometer and monoclonal antibodies obtained from Becton Dickinson (Franklin Lakes, NJ).

To evaluate T cell proliferation, mononuclear cells (1 × 10⁵) were cultured in triplicate in U-bottomed 96-well plates in the presence of 0.1 μg/mL tetanus toxoid or 5 μg/mL phytohemagglutinin (Sigma, St. Louis, MO). These concentrations were selected after preliminary testing. ³H-thymidine (1 μCi) was added on day 3 and cells were harvested 18 hours later. Thymidine incorporation was evaluated using a liquid scintillation counter. The stimulation index was calculated as the ratio between cell proliferation in the presence versus absence of stimuli.

To evaluate DC function, responder T cells were isolated from control donors using T cell enrichment columns (R&D Systems). T cells (10⁵/well) were cultured with mononuclear cells obtained from cancer patients in U-bottomed 96-well plates. T cells with mononuclear cell ratios from 1:1 to 1:8 were used. All experiments were done in triplicate. ³H-thymidine (1 μCi) was added on day 3, and cells were harvested 18 hours later. Thymidine incorporation was evaluated using liquid scintillation counter. Each mononuclear cell sample was tested against T cells from at least two different donors and the maximum result was used.

Statistical analysis. We used the SAS (SAS, Inc. v. 9.2., Cary, NC) and GraphPad Prizm software for statistical analyses. p53-specific immune responses to immunization were analyzed by descriptive statistics. The Wilcoxon Mann-Whitney test was used to determine a relationship between two groups with continuous variables. Univariate associations between frequency of responders and groups were analyzed by Fisher's exact test considering small sample sizes. No attempt was made to adjust to multiple tests in this exploratory study. Any results which seem to be statistically significant require further confirmation. Overall survival was estimated using the duration from the first vaccine administration to the end of follow-up by the Kaplan-Meier method. Two-sided tests were used for all calculations. The statistical significance for all analyses was determined with P < 0.05.

Results

Patient characteristics and treatment summary. Between January 2003 and June 2005, 29 patients were treated with the vaccine. All patients had ES SCLC at the time of vaccination (17 patients with newly diagnosed ES disease and 12 with relapsed disease; Table 1). The median age was 63 years (range 39-76). Twenty patients were vaccinated after only one prior chemotherapy regimen (six patients after two regimens, and three patients after three regimens). All patients had received prior platinum therapy (Table 1).

DCs were generated from peripheral blood mononuclear cell precursors, and then infected with an adenoviral construct containing wild-type p53 (ADVEXIN; Ad-p53). A typical example of the cell phenotype after Ad-p53 infection is presented in Fig. 1. Patients were scheduled to receive three doses of vaccine intradermally at 2-week intervals. If patients showed stable disease or better, they were given three or more doses of the vaccine, once per month. The phase I component of the trial had an initial goal to escalate the vaccine dose from 5 × 10⁶ to 5 × 10⁷ p53⁺ DC. However, generation of >5 × 10⁶ p53⁺ DC per dose was difficult to achieve. Therefore, due to feasibility, the doses of p53⁺ DC were not escalated beyond 5 × 10⁶ cells. On average, 7.7 × 10⁷ total DC and 8.6 × 10⁶ p53⁺ DC were generated for each dose (Table 2). The number of injected p53⁺ DC was limited to 5 × 10⁶ even if more cells were generated.

Antigen-specific cellular immune response to the vaccine. p53-specific immune responses were evaluated in IFN-γ ELISPOT assays using a canarypox virus (ALVAC) containing wild-type p53 or control virus for stimulation. Use of ALVAC containing the full-length p53 gene allowed for evaluation of p53-specific responses regardless of the patients' HLA haplotype. Response to vaccination was considered significant if a p53-specific response after immunization (2-3 weeks after third vaccination) was more than 2 SD higher than the p53-specific response before immunization and at least 2 SD higher than the response to control ALVAC. Representative ELISPOT data are shown in Fig. 2A. Immune responses were further evaluated in a subset of 12 HLA-A2-positive patients, using p53-derived or control peptides presented by HLA-A2. Illustrative results are shown in Fig. 2B. In these HLA-A2-positive patients it was also possible to evaluate the antigen-specific CD8⁺ T cells using tetramer staining (data not shown).

Modest but significant p53-specific T cell responses to vaccination were found in 13 out of 25 patients (52%) using ALVAC-p53, and in 7 out of 12 patients (58.3%) using the p53-derived peptide (Fig. 2C and D). Three patients who had a significant response to vaccination measured using the p53-derived peptide were not tested with ALVAC-p53 due to technical reasons. Overall, 16 out of the 28 tested patients (57.1%; termed p53 responders) had statistically significant p53-specific responses to immunization. When both assays were used to test peripheral blood mononuclear cells from the same patient, the response rate to ALVAC-p53 compared with p53-derived peptide was not significantly different (P > 0.1), however, a lower frequency was seen using tetramer staining. Only 3 out of 11 tested patients (27.2%) showed a significant increase in tetramer staining. The prevaccination level of p53-specific immunity was similar in patients who responded immunologically to the vaccine to those who did not respond (data not shown).

We analyzed a possible link between the number of DCs injected with each vaccine and p53-specific immune response developed as the result of immunization. On average, 14 patients received the maximum dose (5 × 10⁶ p53⁺ DCs per vaccine), 2 patients received more than 4 × 10⁶ cells, 5 patients received more than 3 but less than 4 × 10⁶ cells, 6 patients received more then 2 but less than 3 × 10⁶ cells, and 2 patients received less than 2 × 10⁶ p53⁺ DCs. No correlation between the number of injected DCs and p53-specific immune response was found.

Association between cellular and humoral immune responses to the vaccine. A detectable preimmunization level of anti-p53 antibody was observed in 10 out of 22 (45%) tested patients, and only 3 patients showed a significant increase in the level of anti-p53 antibodies after immunization. Interestingly, all three of these patients had a detectable prevaccine level of the antibody. Four out of 10 patients (40%) with a detectable preimmunization level of anti-p53 antibody developed a positive p53-specific T cell response to vaccination, which was not statistically different from the response rate in patients with no preexisting level of anti-p53 antibody (data not shown).

Antiadenovirus antibodies could potentially play a critical role in limiting the effectiveness of adenovirus-based cancer vaccines. We measured antiadenovirus IgG and IgM antibodies by ELISA using serial dilution of the patients' sera. Most patients had a detectable preimmunization level of antiadenovirus IgG antibody. After three rounds of immunization with adenovirus-p53 DC, the titer of antiadenovirus antibody increased in 16 out of 28 tested patients (57.1%). A moderate increase (>2 and <8-fold) was observed in 14 patients and a substantial (>8-fold) increase in 2 patients. Both patients with the substantial increase in antiadenovirus antibodies did not develop p53-specific cellular responses to immunization. p53-specific cellular immune responses to vaccination were observed in 12 out of 14 patients (85.7%) who had a moderately increased titer of antiadenovirus antibodies, and in only 5 out of 13 patients (38.5%) who had no detectable increase in antibody titer (P = 0.018; Fig. 2E). The antiadenovirus antibodies detected in those patients had neutralizing activity with a median neutralizing titer of 1:1,600.

Association of immune response to vaccination with prevaccination levels of T cell function, DC activity, and the presence of immature myeloid cells. The generation of a p53-specific T cell response depends not only on the quality of antigen stimulation but also on the functional activity of T cells and antigen-presenting cells in the host, specifically DCs. In order to determine whether the preexisting condition of the host immune system influenced the outcome of vaccination, mononuclear cells isolated from patients prior to vaccination were stimulated with either tetanus toxoid or phytohemagglutinin, and cell proliferation was measured by uptake of ³H-thymidine. Nine out of 20 (45%) tested patients had a tetanus toxoid response below the lower limit of normal samples (stimulation index, <15), and 9 out of 19 (47.3%) tested patients had a phytohemagglutinin response below the control range (stimulation index, <25). However, patients with a decreased T cell response to tetanus toxoid or phytohemagglutinin developed p53-specific immunity to vaccination at a rate similar to the patients with normal T cell responses (P > 0.1; data not shown).

Recent studies have suggested that CD4⁺CD25⁺ regulatory T cells (T_reg) might play an important role in the down-regulation of antitumor immune responses (reviewed in ref. 20). As an initial evaluation of the T_reg population, CD25^high cells were enumerated within the total population of CD3⁺CD4⁺ T cells. No differences in the proportion of these cells were found between the groups of healthy subjects and patients with SCLC prior to vaccination or 2 to 3 weeks after completion of vaccination (Fig. 3A). Furthermore, no statistically significant link was found between the presence of these cells in the patients' blood before or after vaccination, and p53-specific T cell responses to vaccination (Fig. 3B).

The possible association between the number and functional activity of DCs prior to vaccination, and antigen-specific responses to vaccination was also investigated. Although no statistically significant decrease in the median proportion of DCs (Lin⁻HLA-DR⁺) nor their mature CD83⁺ subset was found in patients with SCLC as compared with normal subjects, a substantial number of individual patients had decreased levels of DCs (below minimum values of the control group; Fig. 3C). Patients with a reduced level of the total DC population did not exhibit a significant decrease in the ability to develop a p53-specific cellular immune response to vaccination compared with those who had a normal level of these cells (Fig. 3D). In contrast, there was a trend toward a lower rate of p53-specific responses in patients with reduced levels of CD83⁺ mature DCs. Five out of six patients with control levels of CD83⁺ DCs prior to vaccination developed p53-specific immune response to vaccination, whereas only 5 out of 12 patients with reduced level of these cells responded to the vaccine (P = 0.15; Fig. 3D). Expression of HLA-DR on SCLC patients' DCs was significantly lower than in healthy subjects (Fig. 3E). Furthermore, a substantial decrease in the function of DC was seen in allogeneic mixed leukocyte reactions (Fig. 3F). Analysis of p53-specific immunologic response to vaccination in this case was not relevant because very few patients showed control levels of these variables.

Because immature myeloid cells have been implicated in the immunosuppressive state of tumor-bearing hosts (21, 22) the patients with SCLC were examined for the presence of these cells. Patients with SCLC had elevated levels of Lin⁻HLA-DR⁻CD33⁺ immature myeloid cells prior to vaccination (P = 0.01). After vaccination, their presence increased even further (P = 0.002; Fig. 3G). All six patients with a normal level of immature myeloid cells prior to vaccination developed p53-specific immune responses to vaccination (100%), compared with only 4 out of 12 patients who had an elevated level of immature myeloid cells (25.0%; P = 0.012; Fig. 3H). Thus, it seems that patients with an increased presence of immature myeloid cells were less likely to develop responses to vaccination compared with the patients with a normal level of immature myeloid cells.

Clinical response to vaccination and its association with an antigen-specific immune response. Adverse events associated with the administration of the vaccine were infrequent and mostly mild. Only two patients experienced grade 2 adverse events (one for fatigue and one for arthralgia) with vaccine administration, and vaccinations were never withheld due to the presence of any adverse events. The occurrence of adverse events was independent of the number of vaccines previously received.

Out of 29 patients treated with the vaccine, 1 patient achieved a partial response, 7 had stable disease, and 21 patients developed progressive disease. The patient with a partial response received four cycles of cisplatin and etoposide concurrent with thoracic radiation therapy immediately after the initial diagnosis. She subsequently progressed 2 months after her last dose of cisplatin, with the appearance of several positron emission tomography–positive enlarged retroperitoneal lymph nodes. She received three vaccines at that time, and was re-staged 2 weeks later. Overall Response Evaluation Criteria in Solid Tumors measurements revealed a 60% decrease in the size of all of her measurable lesions (Fig. 4A). This patient, and five out of seven patients with stable disease after vaccination, eventually progressed 2 to 6 months later. Twenty-three of the 27 patients with progressive disease after vaccine administration were treated with additional chemotherapy (four patients declined). At this time, clinical follow-up was completed for 21 patients. Of these, 13 were platinum-resistant (progressed within 90 days of receiving a platinum-containing regimen). Fourteen patients received paclitaxel, three received carboplatin/CPT-11, two received CDDP/CPT-11, one received carboplatin/VP-16, and one received epirubicin. Previously published objective response rate to second-line chemotherapy in patients with platinum-resistant ES SCLC is 2% to 5%, and for studies where >50% patients had resistant disease (as in our patient population; 6% to 16%; ref. 23). However, we observed objective clinical responses (partial response + complete response) in 61.9% of the 21 patients treated with second-line chemotherapy (Table 3). Only four patients (19%) progressed after second-line chemotherapy. Thirteen of the platinum-resistant patients treated with the vaccine received various chemotherapeutic regimens when they progressed after receiving the vaccine, the response rate was 61.5% (Table 3). The median survival of these platinum-resistant patients (n = 13) from the time of the first vaccine administration was 9.3 months with a lower 95% confidence interval of 7.1 months. The median overall survival of all 29 patients was 11.8 months from the time of the first vaccine administration, with a lower 95% confidence interval of 7.9 months (Fig. 4B). Eleven of the 29 patients (38.1%) were alive 1 year after the administration of the first vaccine.

Twenty-eight treated patients are thus far evaluable for immune responses. One patient was removed from the immune response analysis due to the loss of a blood specimen. We evaluated the connection between the development of p53-specific T cell responses to immunization and clinical response to second-line chemotherapy. As was described above, immune response was measured 2 to 3 weeks after the end of vaccination just before the start of second-line chemotherapy. Nine out of 12 patients (75%) who had a positive immunologic response to immunization developed a complete response or partial response to second-line chemotherapy compared with 3 out of 10 patients (30.0%) who had no detectable immunologic response (P = 0.08; Fig. 4C). Patients with positive immunologic response to vaccination showed a slight trend toward improved overall survival (median 12.1 months) compared with patients who did not respond immunologically to vaccination (median survival, 9.6 months).

The administration of second line chemotherapy started in most patients 3 to 4 weeks after the end of vaccination. To follow-up the status of the specific immune response in these patients, we evaluated p53-specific immune responses 2 months after the last vaccination. In most patients, there was a significant decrease in the p53-specific immune responses (Fig. 4D and E). This decrease was not associated with significant chemotherapy-induced lymphopenia (Fig. 4F).

Discussion

The goal of this study was to develop a vaccine that would improve the clinical outcome of ES SCLC. We and others have established in preclinical experiments that p53 could be a good target for cancer immunotherapy (7–9, 14, 24–29). In several clinical trials using p53 peptides or a viral vector containing the wild-type p53 gene, generation of a specific immune response in some patients has been reported. However, no objective clinical responses have been observed (11, 30, 31). We hypothesized that the use of the most potent antigen-presenting cells, DCs, transduced with wild-type p53 might induce a more potent p53-specific immune response that would translate into a clinical response. Adenovirus is not only an excellent tool for gene delivery into DCs (reviewed in refs. 17, 18) but also induces activation of these cells manifested by up-regulation of MHC class II and costimulatory molecules on DC, production of interleukin-12 and other proinflammatory cytokines as well as enhancing functional potency (14–16, 32, 33). Thus, adenovirus offers a unique ability to combine Ag delivery and DC activation and might provide additional benefits for DC-based cancer immunotherapy.

Vaccination resulted in the development of p53-specific T cell responses in just over half of the treated patients (57.1%). These data indicate that despite our efforts, from an immunologic standpoint, the vaccine was effective in only half of the patients. There are several factors that could contribute to a suboptimal immune response: defective T cell function, T cell suppression induced by regulatory T cells, the inability of host DCs to maintain antigen-specific immune responses, the presence of immune suppressive populations of DCs or macrophages, suppression of antigen-specific immune responses by immature myeloid cells, the induction of antiadenoviral immune responses, etc. In one study, it was not possible to evaluate all possible mechanisms. Our analysis indicated that inhibition of overall T cell reactivity and accumulation of T_reg were not associated with failure to develop p53-specific immune responses. Our data do not necessarily indicate a lack of involvement of T_reg in SCLC. T_reg may migrate from peripheral blood to the site of the tumors. In addition, because all patients were treated with platinum-based chemotherapy 8 weeks before the analysis, it is possible that chemotherapy could have eliminated some of these cells as was previously reported for cyclophosphamide (34). Investigation is currently under way to address this issue.

Our data showed a trend in the association between a decreased presence of mature CD83⁺ DCs in the peripheral blood of patients prior to vaccination and a lack of immune response to the vaccine. These data may point out an important role of host DCs in the development of antigen-specific immune responses. Nearly all patients enrolled in the trial had significantly reduced levels of DC function, suggesting that improvement of DC function prior to vaccination might be necessary to achieve improved efficacy of the vaccine.

Patients with ES SCLC had increased levels of immature myeloid cells, the cells implicated in tumor-associated antigen-specific immune suppression (21, 22, 35). Importantly, 100% of patients with normal prevaccine levels of immature myeloid cells developed p53-specific responses to vaccination, compared with only 25% of patients with elevated prevaccine levels of immature myeloid cells (P = 0.012). Those differences show that an increase in the number of immature myeloid cells might negatively affect antigen-specific responses to vaccine. After vaccination, the presence of immature myeloid cells increased even further, with only two patients having a normal level of these cells (both developed positive responses to vaccination). The increase in immature myeloid cells may have been caused by the fact that most of the patients had progressive disease by the time of evaluation, thus, removal of immature myeloid cells could potentially be beneficial in enhancing the effect of cancer vaccines.

The induction of antiadenovirus antibody response is considered one of the major limiting factors for the use of this vector in gene therapy. However, our results suggested that a moderate increase in the titer of antiadenovirus antibody was associated with p53-specific responses to vaccination. Because both adenovirus- and p53-derived antigens are presented on the same DC, evaluation of antiadenovirus immunity may serve as a correlate for functional activity of DC. These data are consistent with the results obtained in animal models that showed a limited antiadenovirus response after immunization of mice with DCs transduced with different adenoviral constructs which did not affect antigen-specific CTL activity (14, 36).

Despite the induction of p53 antigen-specific immunity in more than half of the patients, an objective clinical response was observed in only one patient after vaccine treatment. Although this objective response indicates that the vaccine could have direct antitumor effects, the frequency of such responses was unsatisfactory. Importantly, it was similar to the response described in previous clinical trials (3). However, unexpected results were observed after treatment of patients with second-line chemotherapy. Most of the vaccinated patients had objective clinical responses (complete response or partial response) to subsequent chemotherapy. Given the very low historic response rate in this group of patients; these data suggest an improvement in the clinical outcome of the disease caused by vaccination. Patients with ES SCLC have a poor prognosis (reviewed in ref. 16), with treated patients having a median survival of 7 to 11 months. Although first line chemotherapy produces tumor responses 40% to 70% of the time, the response rate to second-line chemotherapy is much lower. Patients who develop progressive disease within 90 days after receiving a platinum compound (platinum resistant) have a particularly poor prognosis, with response rates reported with single agent chemotherapy of 3.7% to 6.4% with corresponding median survivals of 3 to 6.9 months (37–39). One trial where patients were treated with a combination of paclitaxel, cisplatin, and ifosfamide showed a response rate of 70%. Although the median survival of the platinum-resistant patients was not reported separately, the median survival of all patients treated (20 with platinum resistant and 13 with platinum sensitive disease) was 7 months (40).

It is important to point out that the dramatic difference in historic rate of response to second-line chemotherapy of platinum-resistant patients reported previously (<8%) and the one observed in our study (>60%) could be partly caused by the bias introduced during selection of patients. Further studies are required to rigorously evaluate this treatment regimen in the setting of a controlled clinical trial. Importantly, clinical response to vaccination was associated with immunologic response. Only 30% of patients who did not develop a p53-specific response to vaccination responded clinically to second-line chemotherapy, whereas 75% of p53 cellular immune responders had objective clinical responses to chemotherapy after vaccination (P = 0.08). The median overall survival of all patients from the time of the first vaccine administration was 11.8 months with a lower limit of the 95% confidence interval of 7.8 months. Furthermore, 38% of patients were alive 1 year after the first vaccine injection. The median survival of the patients treated with second-line chemotherapy in most phase II clinical trials has been ∼6 months with <20% of patients living beyond 1 year (39). These data support the hypothesis that vaccination may be beneficial for patients with SCLC receiving chemotherapy. These data are consistent with recent unexpected observations made in cancer patients immunized with the carcinogen activator cytochrome P450 1B1 (CYP1B1). Most of the 17 vaccinated patients progressed, and all but 1 of 11 patients who did not develop immunity to CYP1B1 progressed and did not respond to salvage therapy. Five patients who developed immunity to CYP1B1 required salvage therapy for progressive metastatic disease and showed marked response to their next treatment regimen, most of which lasted >1 year (41). In another recent phase I study, similar results were obtained in patients with glioma (42).

Chemotherapy may blunt the antigen-specific T cell response because p53 immunity was significantly reduced after starting chemotherapy. However, apparently this time frame was sufficient to produce a positive clinical result. It seems that induction of anti-p53 cellular immunity synergizes with subsequent chemotherapy to provide potent systemic antitumor activity. There are a number of potential mechanisms that could explain the observed effect. Chemotherapy might down-regulate the effect of tumor-produced immunosuppressive factors that prevent CTLs from killing tumor cells, it can up-regulate p53 in tumor cells, which could make them more susceptible to recognition by CTLs, it may activate CTLs by up-regulating the level of expression of perforin or granzymes, or the proapoptotic effect of granzymes and chemotherapy might synergize at a molecular level. We are currently investigating these possibilities.

Preclinical data accumulated in recent years suggest that the combination of cancer immunotherapy and chemotherapy might provide significant benefit (43). Here, we provide direct clinical evidence in support of this emerging paradigm in cancer immunotherapy. Cancer vaccines may be most effective not as a single modality but rather in close combination with other methods of treatment, specifically, chemotherapy. This concept may have broad implications for the further development of new vaccine strategies.