Clinical and Immunological Outcomes in High-Risk Resected Melanoma Patients Receiving Peptide-Based Vaccination and Interferon Alpha, With or Without Dacarbazine Preconditioning: A Phase II Study

Abstract

Clinical studies based on novel rationales and mechanisms of action of chemotherapy agents and cytokines can contribute to the development of new concepts and strategies of antitumor combination therapies. In previous studies, we investigated the paradoxical immunostimulating effects of some chemotherapeutics and the immunoadjuvant activity of interferon alpha (IFN-α) in preclinical and clinical models, thus unraveling novel rationales and mechanisms of action of chemotherapy agents and cytokines for cancer immunotherapy. Here, we carried out a randomized, phase II clinical trial, in which we analyzed the relapse-free (RFS) and overall survival (OS) of 34 completely resected stage III-IV melanoma patients, treated with peptide-based vaccination (Melan-A/MART-1 and NY-ESO-1) in combination with IFN-α2b, with (arm 2) or without (arm 1) dacarbazine preconditioning. All patients were included in the intention-to-treat analysis. At a median follow-up of 4.5 years (interquartile range, 15.4-81.0 months), the rates of RFS were 52.9 and 35.3% in arms 1 and 2, respectively. The 4.5-year OS rates were 68.8% in arm 1 and 62.7% in arm 2. No significant differences were observed between the two arms for both RFS and OS. Interestingly, the RFS and OS curves remained stable starting from 18 and 42 months, respectively. Grade 3 adverse events occurred in 5.9% of patients, whereas grade 4 events were not observed. Both treatments induced a significant expansion of vaccine-specific CD8⁺ T cells, with no correlation with the clinical outcome. However, treatment-induced increase of polyfunctionality and of interleukin 2 production by Melan-A-specific CD8⁺ T cells and expansion/activation of natural killer cells correlated with RFS, being observed only in nonrelapsing patients. Despite the recent availability of different therapeutic options, low-cost, low-toxic therapies with long-lasting clinical effects are still needed in patients with high-risk resected stage III/IV melanoma. The combination of peptide vaccination with IFN-α2b showed a minimal toxicity profile and resulted in encouraging RFS and OS rates, justifying further evaluation in clinical trials, which may include the use of checkpoint inhibitors to further expand the antitumor immune response and the clinical outcome. Clinical Trial Registration: https://www.clinicaltrialsregister.eu/ctr-search/search, identifier: 2008-008211-26.

Keywords: chemotherapy; combination therapy; drug repurposing; immunotherapy; interferon-α; melanoma.

Figure 1

Treatment schedule and CONSORT flow diagram. (A) Roman numbers indicate the cycle of treatment. Tn indicates the day after beginning of treatment. Black arrow, dacarbazine (DTIC) intravenous infusion. White arrow, IFN-α2b subcutaneous injection. Gray arrow, vaccine (Melan-A/MART-1 and NY-ESO-1 peptides emulsified with Montanide ISA-51) intradermal injection. (B) Flow diagram showing the progress of patients throughout the trial.

Figure 2

Intention-to-treat analysis of relapse-free (RFS) (A,C) and overall survival (OS) (B,D) by Kaplan–Meier method. All enrolled patients were included in the analysis (n = 34). Months are calculated since time of randomization. Arm 1 patients (Pt) (n = 17) were treated with vaccination with Melan-A and NY-ESO-1 peptides (Vaccine) and interferon-α2b (IFN). Arm 2 patients (n = 17) received the same treatment of arm 1 patients with the addition of dacarbazine (DTIC) pretreatment. (A,B) Comparison between arms. (C,D) All patients (n = 34). p value by log-rank test.

Figure 3

Specific immune response. Frequencies of Melan-A– (A–F) and NY-ESO-1–specific (G–I) CD8⁺ T cells analyzed by tetramer staining (n = 29). (A) Kinetic analysis of the frequency of Melan-A–specific T cells between pretreatment and posttreatment samples (T21, T42, T63, T84, T105) in one representative patient (patient 29). (B,C) Variation of Melan-A–specific T cell percentage between pretreatment and posttreatment ex vivo samples in arm 1 (n = 15) vs. arm 2 patients (n = 14) (B) and in patients with no evidence of disease (NED) (n = 16) vs. relapsed patients (n = 13) (C). (D,E,G,H) Variation of Melan-A– (n = 28) (D,E) and NY-ESO-1–specific (n = 26) (G,H) T cell percentage after short term in vitro expansion, in arm 1 vs. arm 2 patients (D,G) and in patients with no evidence of disease (NED) vs. relapsed patients (E,H). (F,I). Representative staining (patient 09) of short term in vitro expansion before (pre) and after (post, T105) stimulation with Melan-A (F) and NY-ESO-1 (I) peptides. p values were calculated by Wilcoxon signed-rank test.

Figure 4

Polyfunctional analysis of Melan-A–specific CD8⁺ T cells. Peripheral blood mononuclear cells were analyzed by multiparameter flow cytometry after 5 h in vitro culture without peptide pulsing, followed by 1-h incubation with brefeldin A and monensin. CD8⁺ Melan-A tetramer–positive T cells were gated as shown in Supplementary Figure S1 and analyzed for their simultaneous expression of surface CD107a and intracellular IL-2, IFN-γ, and TNF-α. (A,B) Pie charts showing the proportion of cells expressing any combination of four (4 Fun), three (3 Fun), two (2 Fun), one (1 Fun), or zero (0 Fun) tested markers (CD107a, IL-2, IFN-γ, and TNF-α). Data are expressed as mean percentage of CD8⁺ Melan-A tetramer–positive T cells. (A) Patients with no evidence of disease (NED). (B) Patients with disease recurrence (Relapsed). p values by Wilcoxon signed-rank test. (C) Variation of Melan-A–specific T cell percentage expressing intracellular IL-2, TNF-α, IFN-γ, or surface CD107a (logarithmic scale) between pretreatment and posttreatment (4 months) in patients with no evidence of disease (NED) (n = 7) vs. relapsed patients (n = 6). p values by Wilcoxon signed-rank test. (D) Kaplan–Meier plot comparing the relapse-free survival of patients (Pts) characterized, or not, by a twofold expansion of IL-2–positive Melan-A–specific T cells in post vs. pre samples. p value by log-rank test.

Figure 5

Analysis of natural killer (NK) cell subsets and of their functionality. NK cells were identified and divided into four different subsets based on the expression of CD56 and CD16 within the CD3 negative lymphocytes (gating strategy depicted in Supplementary Figure S2). (A) Variation of the percentage of CD56^bright CD16^neg NK cells (logarithmic scale) between pretreatment and posttreatment (4 months) in patients with no evidence of disease (NED) (n = 6) vs. relapsed patients (n = 4). (B,C) NK cells were in vitro cultured with PMA/ionomycin (B) or with K562 target cells (C), and CD56^dim CD16^neg NK cells were analyzed for their percentage variation (B) and CD107a expression (C) pretreatment and posttreatment in NED (n = 6) and relapsed patients (n = 4). p values by Wilcoxon signed-rank test. (D) Functional analysis of CD56dim CD16neg NK cells in response to K562 target cells pretherapy and 4 months after therapy in one representative patient (patient 28). CD107a-positive cells increased after treatment.