Abstract Background: Inefficient prediction of functional tumor neoantigens and subsequent host anti-tumor immune responses limit optimization of immunotherapeutic approaches. We recently demonstrated that programmed cell death 1 inhibitors have greater efficacy in those with charged human leukocyte antigen (HLA)-B binding pockets whose tumors harbor mutation(s) leading to “motif” neoepitopes (those that generate new charged HLA anchors). Whether peptides derived from motifs are capable of inducing host tumor-specific immune responses was previously unknown. Methods: Matched tumor and peripheral blood mononuclear cell (PBMC) biospecimens from participants (pts) in NCT03546361 underwent whole exome sequencing (WES) and HLA/supertype/neoepitope prediction using established pipelines. Pts were grouped as (1) 1+ B44/B27 allele with 1+ motif, (2) B44/B27 without motifs, and (3) no B44/B27 alleles. For (3), neoepitope prediction was run with classic B44/B27 supertype alleles (B*40:01, B*27:05) as a negative control. Nonamer peptides were created from pt motif and non-motifs and their wildtype sequence. For those with more than 2 non-motifs, the highest variant allele fraction (VAF), lowest IC50 (binding), and/or highest fold change (fold) were selected. T-cell activation and expansion assays were run using pt PBMC collections on days 0, 21, 42, and 63 by co-incubating peptide-pulsed dendritic cells (2x105) with autologous CD8+ T cells (2x106) in RPMI containing 10% pooled human serum, 10 mM HEPES, 2mM L-Glutamine, and 50 µM β-mercaptoethanol supplemented with IL-15 (10 ng/mL) and IL-2 (50 IU/mL) for both 18 hours and 10 days. Negative controls included vehicle only (0.1% DMSO); positive controls included PMA (25 ng/mL), ionomycin (1 µg/mL) and BFA. Activated T cells were assessed by flow cytometry based on intracellular staining of INF-ɣ and TNF-ɑ. Results: 18 pts underwent analysis: 14 had sufficient WES, 11 exhibited a B44/B27 allele. Of these, 8 pts had predicted neoepitopes with VAF ≥10%: 3 pts were (1), 2 were (2), 3 were (3). From this cohort, 23 peptides represented 6 motif, 6 VAF, 4 binding, 4 fold, and 3 multiple categories. Positive controls yielded activated T cell peaks at day 42 (4.78%). One participant in the B44/B27 motif category (1) had one peptide that elicited T cell activation at day 42 (2.51%), which was motif/VAF, and confirmed in a subsequent assay (0.63%). All other peptide experiments did not demonstrate T cell activation, and there were no pts who demonstrated meaningful clinical benefit from the trial. Conclusion: Current neoepitope prediction continues to provide limited insight into functionality. In a limited sample size of those without clinical benefit to NCT03546361, possible functionality of one motif neoepitope was demonstrated. Additional studies are ongoing to validate these findings and further define the functional relevance of motif neoepitopes. Citation Format: Amy Lauren Cummings, Andy Han, Michael Oh, Camelia Dumitras, Seung J. Park, Sai S. Kollapaneni, Daniel Li, Arjan Gower, Maria Antonia Velez Velez, Aaron Lisberg, Monica Cappelletti, Steven M. Dubinett, Bin Liu, Edward B. Garon. Motif neoepitopes demonstrate immunogenic signal in CCL21-gene modified dendritic cell vaccination trial (NCT03546361) abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 1605.
Cummings et al. (2026) studied this question.