WT1-mRNA dendritic cell vaccination of patients with glioblastoma multiforme, malignant pleural mesothelioma, metastatic breast cancer, and other solid tumors: type 1 T-lymphocyte responses are associated with clinical outcome

Abstract

Cell therapies, including tumor antigen-loaded dendritic cells used as therapeutic cancer vaccines, offer treatment options for patients with malignancies. We evaluated the feasibility, safety, immunogenicity, and clinical activity of adjuvant vaccination with Wilms’ tumor protein (WT1) mRNA-electroporated autologous dendritic cells (WT1-mRNA/DC) in a single-arm phase I/II clinical study of patients with advanced solid tumors receiving standard therapy. Disease status and immune reactivity were evaluated after 8 weeks and 6 months. WT1-mRNA/DC vaccination was feasible in all patients, except one. Vaccination was well tolerated without evidence of systemic toxicity. The disease control rate and overall response rate among a total of 39 evaluable patients were 74.4% and 12.8%, respectively. Median overall survival (OS) was 43.7 months among 13 patients with glioblastoma multiforme, 41.9 months among 12 patients with metastatic breast cancer, and 48.8 months among 10 patients with malignant pleural mesothelioma, comparing favourably with historical controls reported in the literature. OS was longer in patients with stable disease at 8 weeks and disease control at 6 months versus patients without disease control at either time point. Disease control and higher OS were associated with antigen-specific type 1 CD4⁺ and/or CD8⁺ T-lymphocyte responses, mainly induced by WT1-mRNA/DC vaccination. Antigen-nonspecific type 2 CD8⁺ T-cell responses were common before WT1-mRNA/DC vaccination but did not show any association with clinical outcome. Collectively, these data indicate that WT1-mRNA/DC vaccination is feasible, safe, and immunogenic and shows clinical activity in patients with advanced solid tumors, suggesting that it has the potential to help improve their survival.

To the Editor,

We report on 40 patients with poor-prognosis advanced tumors who received adjuvant treatment in a single-arm phase I/II study (NCT01291420) with WT1-mRNA/DC, i.e. autologous dendritic cells exposed to keyhole limpet hemocyanin (KLH) and electroporated with mRNA encoding the Wilms’ tumor protein (WT1) tumor-associated antigen [1, 2] (Additional File 1: Methods). WT1-mRNA/DC production and vaccination were feasible in 40/40 and 39/40 patients, respectively (Additional File 2: Tables S1–S6). Side effects were limited to local reactions at the injection site. The disease control and overall response rates were 64.1% and 5.1% at 8 weeks (time point T1) and 48.7% and 12.8% at 6 months (time point T2) after the initiation of WT1-mRNA/DC vaccination, respectively (Table 1). Median overall survival (OS) was 43.7 months in patients with glioblastoma multiforme, 41.9 months in patients with metastatic breast cancer, and 48.8 months in patients with malignant pleural mesothelioma, comparing favourably with previously reported figures of 16.0, 37.2, and 23.4 months, respectively [3,4,5] (Table 1; Additional File 2: Figure S1). Stable disease was observed at T1 and/or T2 in 24/39 patients (without any concomitant treatment in 12/39 patients), potentially explaining the observed benefits in terms of OS [2, 6, 7]. Achieving stable disease at T1 with disease control at T2 was associated with a longer median OS compared with failing to achieve disease control at T1 or T2 (N = 34, P = 0.0356). All patients achieving partial or complete responses also received concomitant treatment during WT1-mRNA/DC vaccination.

Table 1 Clinical responses and overall survival by patient group

All but 1/38 patients tested (SOL36) had a delayed type hypersensitivity reaction. Most patients also exhibited type 1 T-lymphocyte responses at T1 and/or T2, comprising interferon (IFN)-γ⁺ and/or tumor necrosis factor (TNF)-α⁺ CD4⁺ T-lymphocyte reactivity and IFN-γ⁺ CD8⁺ T-lymphocyte response against WT1, and TNF-α⁺ CD4⁺ T-lymphocyte reactivity against KLH (Fig. 1a). High frequencies (> 0.1%) of antigen-nonspecific interleukin (IL)-5⁺ CD8⁺ T-lymphocytes were detected in many patients at baseline (Fig. 1a), decreasing by at least 50% in 5/27 patients, increasing by at least 50% in 3/27 patients, and remaining within these bounds in 19/27 patients after WT1-mRNA/DC vaccination. Accordingly, most patients displayed antigen-nonspecific type 2 cytotoxic T-lymphocyte (CTL) responses at baseline, which were not consistently affected by WT1-mRNA/DC vaccination, whereas antigen-specific type 1 helper T-lymphocyte and CTL responses emerged de novo in a majority of cases after WT1-mRNA/DC vaccination.

High frequencies of WT1-specific tetramer⁺ CD8⁺ T-lymphocytes, especially those specific for the WT1_187–195 and WT1_235–243 epitopes [8], were detected in HLA-A*02:01⁺ patients at baseline (Fig. 1a; Additional File 2: Figure S2). About 1/3rd of patients showed an increase post-vaccination in tetramer⁺ CD8⁺ T-cells specific for each of the epitopes tested, except WT1_{126 − 134}, for which half of the patients showed increased reactivity. IgG responses specific for WT1 [7] were also detected in most patients, primarily directed against the WT1_235–248 epitope, many of them already present at baseline (Additional File 2: Figure S3; Fig. 1a).

Disease control was associated with type 1 cytokine reactivity (Fig. 1b). Direct associations were detected between disease control and IFN-γ⁺ and/or TNF-α⁺ CD4⁺ and/or CD8⁺ T-lymphocyte responses specific for WT1 or KLH. IFN-γ⁺ CD4⁺ T-lymphocyte responses specific for WT1 associated directly with OS (Fig. 1b). Accordingly, antigen-specific type 1 CD4⁺ and/or CD8⁺ T-lymphocyte responses were associated with clinical outcome, as reported previously [9].

Additional associations were detected among immunological parameters, including IFN-γ⁺ and TNF-α⁺ CD4⁺ versus CD8⁺ WT1-specific T-lymphocyte responses, suggesting a degree of synchronicity in the overall immune response, mainly induced by WT1-mRNA/DC vaccination (Fig. 1b). IFN-γ⁺ CD4⁺ and CD8⁺ WT1-specific T-lymphocyte responses also associated with those of IFN-γ⁺ KLH-specific CD8⁺ T-lymphocytes. These associations are compatible with the importance of antigen-targeted helper functions for the induction of optimal CTL activity [10] and with the phenomenon of heterologous CD8⁺ T-lymphocyte helper functionality, which can stimulate tumor-specific CTL responses induced by cancer immunotherapy [11].

In summary, our early-phase clinical study has demonstrated the safety, feasibility, type 1 immunogenicity and clinical activity of adjuvant WT1-mRNA/DC vaccination in patients with various solid tumors, and suggests a potential beneficial impact on OS. Antigen-specific type 1 T-lymphocyte response was associated with clinical outcome (disease control and OS). Since this study was performed on a pre-selected group of patients, further clinical studies are warranted to confirm the clinical efficacy of WT1-mRNA/DC.

Fig. 1

Proportions of patients with WT1-specific, KLH-specific, and antigen-nonspecific immune responses (a) and statistical associations among immunological and clinical parameters (b). (a) Responses were determined using intracellular cytokine staining of CD4⁺ and CD8⁺ T-lymphocytes, by WT1-specific HLA-A*02:01 tetramer staining of CD8⁺ T-lymphocytes and by measuring the production of IgG antibodies against WT1. Antigen-nonspecific responses were determined after culture with control medium. Baseline response is defined as values at the PRE time point > 0.1% of cytokine⁺ CD4⁺ or CD8⁺ T-lymphocytes (KLH-responsive or WT1-responsive cases were counted if PRE cytokine⁺ values were at least 50% higher than PRE antigen-nonspecific values); or percentage of WT1 HLA-A*02:01 tetramer⁺ CD8⁺ T-lymphocytes > 0.1% at the PRE time point; or values at PRE > 0.15 for WT1 IgG concentrations in the ELISA. Post-vaccine response was defined as increase of at least 50% in response to WT1 (response to non-KLH-exposed WT1-mRNA/DCs and/or to the WT1 peptide pool) or to KLH or to control medium at the POST and/or FIN time points compared with the PRE time point (as defined in Additional File 1: Methods); or increase of at least 50% in the percentage of WT1 HLA-A*02:01 tetramer⁺ CD8⁺ T-lymphocytes at the POST and/or FIN time points compared with the PRE time point (as defined in Additional File 2: Figure S2); or increase in WT1 IgG concentrations from below (in PRE) to above (in POST and/or FIN) the cutoff absorbance value of 0.15 in the ELISA or increase in WT1 IgG concentrations of at least 50% at the POST and/or FIN time points (as defined in Additional File 1: Methods) from a PRE value of > 0.15 in the ELISA. Numbers in bold indicate where ≥ 50% of patients examined showed a response. (b) Statistical analysis of the data summarized in Table 1 and Fig. 1a was performed to assess associations between clinical response and immune response parameters or between immune response parameters. Each mosaic plot represents a statistically significant 2 × 2 contingency analysis (Additional File 1: Methods) between the parameters indicated. P values are indicated under the mosaic plot

^a = 13 patients with glioblastoma multiforme (GBM), 12 patients with metastatic breast cancer (MBC), 10 patients with malignant pleural mesothelioma (MPM), and 4 patients with other tumors

^b = 12 patients with GBM, 11 patients with MBC, 9 patients with MPM, and 4 patients with other tumors

^c = 13 patients with GBM, 5 patients with HR⁺/HER2⁻ MBC, and 10 patients with MPM

* GBM OS from diagnosis > 16.0 months vs. ≤16.0 months & HR⁺/HER2⁻ MBC OS from diagnosis of metastatic disease > 42.1 months vs. ≤42.1 months & MPM OS from diagnosis > 23.4 months vs. ≤23.4 months

Abbreviations: WT1 = W1-reactive. KLH = KLH-reactive. CD4⁺ = reactivity in CD4⁺ T-lymphocytes. CD8⁺ = reactivity in CD8⁺ T-lymphocytes. IFN-γ⁺ = interferon-γ production. TNF-α⁺ = tumor necrosis factor-α production. IL-5⁺ = interleukin-5 production. ≥2 epit = number of cases positive for 2 or more WT1 epitopes. # = number of patients. ± = and/or. + = reactivity. − = no reactivity. n = number of patients analyzed. OS = overall survival