Characterization and comparison of innate and adaptive immune responses at vaccine sites in melanoma vaccine clinical trials

Abstract

The strength and durability of systemic anti-tumor immune responses induced by cancer vaccines depends on adjuvants to support an immunogenic vaccine site microenvironment (VSME). Adjuvants include water-in-oil emulsions with incomplete Freund's adjuvant (IFA) and combinations of toll-like receptor (TLR) agonists, including a preparation containing TLR4 and TLR9 agonists with QS-21 (AS15). IFA-containing vaccines can promote immune cell accumulation at the VSME, whereas effects of AS15 are largely unexplored. Therefore, we assessed innate and adaptive immune cell accumulation and gene expression at the VSME after vaccination with AS15 and compared to effects with IFA. We hypothesized that AS15 would promote less accumulation of innate and adaptive immune cells at the VSME than IFA vaccines. In two clinical trials, patients with resected high-risk melanoma received either a multipeptide vaccine with IFA or a recombinant MAGE-A3 protein vaccine with AS15. Vaccine site biopsies were obtained after one or multiple vaccines. T cells accumulated early after vaccines with AS15, but this was not durable or of the same magnitude as vaccination in IFA. Vaccines with AS15 increased durable expression of DC- and T cell-related genes, as well as PD-L1 and IDO1, suggesting complex activation and regulation of innate and adaptive immune function with AS15. These changes were generally greater with vaccines containing IFA, but IFA induced reduction in myeloid suppressor cells markers. Evidence of tertiary lymphoid structure (TLS) formation was observed with both adjuvants. Our findings highlight adjuvant-dependent changes in immune features at the VSME that may impact systemic immune responses.

Keywords: Cancer vaccine adjuvant; Clinical trial; Immune response; Melanoma; Tertiary lymphoid structure; Vaccine site.

Fig. 1

Clinical study designs for MEL48 and MEL55

Fig. 2

Example IHC stains for CD4 on MEL55 VSME biopsies for patients 15,341 (a) and 16,578 (b) one week after vaccination, with high power images in panels (c) and (d), respectively. Deep and superficial perivascular dermal lymphoid aggregates are evident in both cases

Fig. 3

Number of immune cells per mm2 of vaccine site biopsies in both the superficial and mid deep layers of the skin. Displayed are number of CD83 + cells (a), CD1A + cells (b), and square root of Eosinophils (c), CD8⁺ cells (d), CD4⁺ cells (e), CD20⁺ cells (f), GATA3⁺ cells (g), Tbet⁺ cells (h) or FoxP3⁺ cells (i) week 1 and week 7 after the first vaccine in MEL48 (with IFA) and MEL55 (with AS15). All p values have been corrected for false-discovery rate as stated in the methods and statistical significance was determined at p < 0.005

Fig. 4

a Ratio of GATA3⁺ cells to Tbet⁺ cells in the VSME in MEL48 and MEL55 both week 1 and week 7 after the first vaccine. b Ratio of CD8⁺ cells to FoxP3⁺ cells in the VSME in MEL48 and MEL55 both week 1 and week 7 after the first vaccine. For panels a and b, means and standard deviations are shown in addition to values for each sample. c Relative proportions of FoxP3⁺, Tbet⁺, and GATA3⁺ cells in the VSME dermis are shown for both trials and both time points. All p values have been corrected for false-discovery rate as stated in the methods and statistical significance was determined at p < 0.005

Fig. 5

Examples of PNAd staining in vaccine sites of MEL55 in superficial dermis (a) and deep dermis/subcutaneous (b). Normal lymph node was used as control (c). Small hematoxylin-staining nuclei clustered around PNAd + vessels in a and b are consistent with lymphocytes and other immune cells

Fig. 6

Individual gene expression of eighteen genes that have been previously associated with TLS formation: a BAFF, b APRIL, c LIGHT, d Lymphotoxin alpha, e Lymphotoxin beta, f CD20, g CCL2, h CCL3, i CCL4, j CCL5, k CCL8, l CXCL9, m CXCL10 n CXCL11, o CXCL13 p CCL18, q CCL19, r CCL21. Expression data was obtained from vaccine site biopsies of patients treated with IFA (MEL48) or AS15 (MEL55), as well as normal tissue obtained pre-vaccination for control purposes (n = 3). For patients treated with IFA, gene expression is shown at week (w) 1 (n = 5), and week 3 (n = 4), following initial vaccination at a site separate from the biopsied tissue. For patients treated with AS15, gene expression is shown at week 1 (n = 10) and week 7 (n = 9), following initial vaccination at a distant site. For factors of significance: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; derived from differential gene expression

Fig. 7

Individual gene expression of MDSC-related genes (a–d), inhibitory molecule genes (e–g) and genes involved in FAS-mediated apoptosis (h–m). Expression data was obtained from vaccine site biopsies of patients treated with IFA (MEL48) or AS15 (MEL55), as well as normal tissue obtained pre-vaccination for control purposes (n = 3). For patients treated with IFA, gene expression is shown at week (w) 1 (n = 5), and week 3 (n = 4), following initial vaccination at a site separate from the biopsied tissue. For patients treated with AS15, gene expression is shown at week 1 (n = 10) and week 7 (n = 9), following initial vaccination at a distant site. For factors of significance: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; derived from differential gene expression