Adjuvant dendritic cell-based immunotherapy in melanoma: insights into immune cell dynamics and clinical evidence from a phase II trial

Abstract

Background: Dendritic cells (DCs) are the most efficient antigen-presenting cells and play a central role in the immune system, orchestrating immune response against tumors. We previously demonstrated that DC-based vaccination effectively induces anti-tumor immunity, yet at the same time showing a robust safety profile, making this treatment a potential candidate for effective adjuvant immunotherapy. To explore this possibility, we designed a randomized phase II trial (EudraCT no. 2014-005123-27) to provide a complementary autologous DC vaccination to patients (pts) with resected stage III/IV melanoma.

Methods: Overall, a total of 18 eligible pts were included in this study, 10 of whom received 6 monthly DC vaccination cycles combined with IL-2 administration (arm A), and 8 pts were enrolled in the follow-up observational cohort (arm B). A deep immune biomarkers profiling by multiplex immunoassay, human leukocyte antigens (HLA) typing, multiparametric flow cytometry and in situ tumor microenvironment analysis was performed for the entire pts cohort. The immunological response was assessed in vivo by DTH test and ex vivo against selected melanoma-associated antigens applying the IFN-γ ELISPOT assay.

Results: Pts receiving DC vaccination showed a better relapse-free survival compared to the observational cohort (median 6.6 months, 95% CI, 2.3-not reached (nr) (arm A) vs 5.2 months, 95% CI, 2.5-nr (arm B), not significant), with a favorable trends for female pts (median 15.5 months, 95% CI, 2.6-nr (female) vs 3.3, 95% CI, 2.3-nr (male)), pts with less than 60 years (median 22.5 months, 95% CI, 2.6-nr (age < 60) vs 4.7 months, 95% CI, 2.3-nr (age ≥ 60), and pts with wild-type BRAF status (median 22.5 months, 95% CI, 8.6-nr (BRAF wt) vs 3.8 months, 95% CI, 2.3-nr (BRAF mutated). The toxicity profile was favourable, with no severe adverse events and only mild, manageable reactions. Moreover, additional immune response data suggested increased immune modulation in vaccinated patients, which may reflect a shift in immune dynamics.

Conclusions: Our findings support the safety and tolerability of DC vaccination as an adjuvant treatment for melanoma, demonstrating significant immune modulation at both the tumor site and peripherally in relapsed and non-relapsed patients. These results highlight the potential of autologous, personalised DC-based therapies and pave the way for the development of innovative immunotherapy combinations in future treatment strategies. Trial registration ClinicalTrials.gov NCT02718391; EudraCT no. 2014-005123-27.

Keywords: Dendritic cell; Immune modulatory; Immunotherapy; Skin cancer; Tumor microenvironment.

Fig. 1

A ACDC clinical protocol study design. arm A pts were vaccinated with 6 doses of DC vaccine following total resection, while untreated arm B pts were followed up post-surgery over time. B Schematic table in which, for each study arm, the principal baseline pt characteristics and the results of the univariate analysis on RFS (time expressed in months) are summarized

Fig. 2

A Kaplan–Meier curve of the univariate analysis on RFS of HLA class I allelic heterozygosity/homozygosity distribution in arm A pts. B The table shows the DTH in vivo test best response to KLH and ATH in vaccinated pts (in columns from A to J). C Representative images of the TAA expression analysis by immunohistochemistry (IHC) on pre-vaccine biopsies collected from arm A pts (upper line). Dot plots with bars represent the staining intensity of each analyzed melanoma specific marker (Melan-A, Pmel, Tyrosinase and Ny-eso1 from left to right). Pts were divided in R and NR. D INFγ ELISPOT test steps and graphical representation of the median number of INFγ SFCs on 5 × 10⁵ PBMCs measured at baseline, VAX4 and at the EOT in arm A pts after Survivin and Ny-eso1 peptides stimulation, respectively

Fig. 3

A Boxplot of significantly modulated peripheral blood cell biomarkers, collected during treatment in arm A pts, and analyzed by non-parametric ANOVA test for repeated measures. B Boxplot showing the tendency of the ratios PLR, LMR and NLR during treatment. C Boxes with floating bars represent the lymphocytes percentage, the NLR and the LMR. Statistical analysis was performed with the nonparametric Wilcoxon signed rank test; the exact p value of the comparisons is shown on graphs

Fig. 4

A Boxes with floating bars representing the frequency of M-MDSCs, ratio of M-MDSCs on CD8 + lymphocytes and of regulatory T cells (CD4 + FOXP3 +) significantly modulated in arm A pt’s bloodstream compared to arm B pts. B Box plots with bars representing the abundance of total circulating monocytes (CD14 +), non classical monocytes (CD14-CD16 +) PDL1 + , eMDSCs and M-MDSC in pre and post-treatment blood samples of arm A pts with below panels representing the ratio between each population and the frequency of CD8 + cells. C Box plots representing the frequency of CD8 + EM in arm A pts before and after treatment. D Boxes with floating bars relative to CD4 + naive, CD4 + EM and CD4 + TE subsets frequencies in R and NR arm A pts before and after treatment. E Box plots with bars represent the concentration levels (pg/mL) of INFα, IL-8, IL-9 and IL-4 in vaccinated pts. Statistical analysis was performed with the nonparametric Wilcoxon signed rank test; the exact p value of the comparisons is shown on graphs

Fig. 5

A The number of intratumoral CD8 + T cells per mm² in pre-treatment biopsies of arm A pts (p = 0.2667) is plotted in graphs. Right graph shows the increment of CD8 expression after treatment in Pt#006 (R pt) biopsies. B The abundance of intratumoral CD8 + and CD163 + cells in pre- treatment biopsies is plotted in graphs. C Differences in PDL1 + tumor cells are illustrated in the graph as the percentage of PDL1 expressing tumor cells on the total tumor cell number (TPS 17.71 ± 9.007 vs. 33.67 ± 23.81, R vs NR, p = 0.5417). Right graph shows the increment of PDL1 expression after treatment in Pt#006 (R pt) biopsies. D The abundance of intratumoral FOXP3 + and CD68 + cells in pre-treatment biopsies is plotted in graph