Abstract

VACCIMEL plus BCG and GM-CSF is an immunotherapeutic treatment (IT) that has been assayed in stages IIB, IIC, and III cutaneous melanoma patients. In the adjuvant randomized Phase II study CASVAC-0401, VACCIMEL-treated patients had longer distant-metastases-free survival (DMFS) than those treated with IFNα2b. Five years after locking the data, an actualization was performed. The benefit in DMFS was maintained in the VACCIMEL group versus the IFNα2b treated group (p=0.035), with a median DMFS of 96 months for VACCIMEL and 13 months for IFNα2b. DMFS was also analyzed as a single cohort in patients (n=30) who had been treated with VACCIMEL. The median DMFS was 169 months, and at 48 months follow-up it was 71.4 %, which was not statistically different from DMFS of previously published results obtained in adjuvancy with ipilimumab, pembrolizumab, nivolumab or dabrafenib/trametinib. VACCIMEL induced a polyclonal cellular immune response against melanocytic differentiation antigens (MDA), cancer-testis antigens (CTA), and neoantigens derived from both the patient’s tumor and VACCIMEL, as demonstrated in PRE- and POST-VAC PBMC cultured ex vivo with HLA-restricted peptides and quantified by IFNγ ELISPOT.

The feasibility of combining VACCIMEL with anti-immune checkpoints inhibitors (ICKi) was analyzed. Remarkably, 5 pts relapsing after VACCIMEL achieved complete responses (CR) with anti-PD-1 IT without added toxicity. Assuming that anti-PD-1 synergized VACCIMEL responses, we performed analogous ex vivo assays, stimulating available PBMC from 8 pts treated with VACCIMEL, with HLA-restricted peptides ± nivolumab (10 µg/ml) and evaluated their responses by IFNγ ELISPOT. Overall, IFNγ+ responses increased for 13/30 peptides tested. Two pts (#2, #5) were studied in depth. Pt#2 who relapsed 29 months after VACCIMEL, received pembrolizumab (200 mg Q3W), and achieved CR lasting >24 months. PBMC collected after 10 months of therapy (anti-PD1IT-PBMC) retained reactivity to Tyrosinase and gp100, which was induced only after vaccination, since they had been detected in POST but not in PRE-VAC PBMC. Compared to PRE and POST-VAC PBMC, anti-PD1 IT-PBMC had higher Ki67+ CD4 and CD8 T cells, higher HLA-DR+, CD69+, CD137+ activation markers, the highest effector memory and lowest TEMRA fractions. Upon peptide stimulation, POST-VAC and anti-PD1 IT-PBMC similarly increased effector memory and reduced central memory/naïve T cells, which prevailed in PRE-VAC PBMC. Anti-PD1 IT PBMC after ex vivo TAA stimulation enhanced their activation phenotype and showed the lowest proportion of PD1+ T cells.

In Pt#5 (HLA-A0201+), PRE- and POST-VAC PBMC were cultured ± nivolumab and tested in a calcein-lysis assay targeting an HLA-A0201+ CM line expressing relevant TAA. Nivolumab increased POST-VAC PBMC lysis from 11.5% to 24.2% (E:T 30:1), whereas  PRE-VAC PBMC remained weakly cytotoxic.

These results demonstrate that VACCIMEL-induced T cell responses persisting long after treatment and remain functional under anti-PD-1 IT, suggesting that VACCIMEL expands clones whose enhanced lytic activity may contribute to favorable outcomes with checkpoint blockade.