Trial Design, Patients, and Endpoints
In a single-center, investigator-initiated phase 1 clinical trial, patients diagnosed with surgically resectable pancreatic ductal adenocarcinoma (PDAC) without distant metastases and with five or more neoantigens as predicted by a computational pipeline underwent a treatment protocol. The protocol involved sequential surgery followed by adjuvant atezolizumab (anti-PD-L1), autogene cevumeran (an individualized vaccine encoding up to 20 neoantigens), and mFOLFIRINOX. All patients had an Eastern Cooperative Oncology Group performance status of 0-1. Patients with pathologically confirmed PDAC and a surgical margin status of R0/R1 were included, while those with metastatic, borderline, or locally unresectable PDAC were excluded. Neoadjuvant therapy recipients were also not eligible. The primary endpoint was safety, with secondary endpoints being 18-month recurrence-free survival and 18-month overall survival. Detailed eligibility criteria, procedures, treatments, and ethical study conduct were outlined in the protocol.
Treatment and Patient Cohort
Nineteen patients received atezolizumab in the safety-evaluable cohort, with 16 of those patients proceeding to receive autogene cevumeran in the biomarker-evaluable cohort. Of the 16 patients who received autogene cevumeran, 15 also received mFOLFIRINOX. Patients who elicited robust T cell responses to vaccine neoantigens, as determined by an ex vivo IFNγ ELISpot assay, were classified as autogene cevumeran responders. Recurrence was defined according to response evaluation criteria in solid tumors (RECIST, version 1.1), with recurrence-free survival calculated from the date of surgery or the last priming dose of autogene cevumeran to the date of recurrence or death. Overall survival was calculated from the date of surgery. The cutoff for data analysis was December 1, 2023, extending the median follow-up to 38 months.
Ethical Considerations and Study Approval
The trial was conducted in compliance with the Declaration of Helsinki and good clinical practice guidelines. Institutional Review Board approval was obtained from Memorial Sloan Kettering Cancer Center (MSK), the United States Federal Drug Administration (FDA), and registration on clinicaltrials.gov (NCT04161755). All participants provided written informed consent.
Mutation Identification and Neoantigen Selection for Personalized Vaccines
The process of selecting vaccine neoantigens involved identifying expressed non-synonymous mutations and HLA types using whole-exome sequencing of patient-specific tumor-normal pairs and tumor RNA sequencing. Bioinformatic estimation and ranking of neoantigens based on immunogenicity were performed to select the final neoantigen candidates.
Manufacture of Autogene Cevumeran
Individualized mRNA neoantigen vaccines were manufactured for each patient under good manufacturing practice conditions. The vaccines were constructed as two uridine-based mRNA strands encoding up to 10 MHCI and MHCII neoantigens in lipoplex nanoparticles for intravenous delivery.
Patient Samples, Cell Lines, and Cell Culture
Patient PBMCs were purified and cultured in a medium supplemented with various components. Single clone reactivity was mapped by culturing purified human T cells in medium supplemented with specific factors. Different cell lines were maintained in specific culture media according to the requirements of the experimental procedures.
HLA Cloning and Transduction
Patient-matched HLA alleles were cloned and transduced into specific cells to study antigen presentation. The process involved transfecting cells with vectors containing HLA alleles and sorting them to obtain cells expressing the desired alleles for further experimentation.
TCR Reconstruction, Cloning, Transduction, and Peptide Stimulation
TCRs were reconstructed, cloned, and transduced into human T cells for subsequent peptide stimulation experiments. The selection of neoantigen epitopes for TCR specificity testing was based on predicted binding affinities to patient HLA-I molecules. TCR-transduced T cells were then stimulated with peptides to assess reactivity and binding affinities.
Immune Response
TCR Vβ Sequencing
Genomic DNA was extracted from various cell sources, and TCR Vβ sequencing was performed to analyze clonal diversity and responses to vaccination. Differential gene expression analysis, single-cell immune profiling, and TCR sequencing were essential components of the immune response evaluation.
Single-Cell RNA and TCR Sequencing
The prepared libraries for single-cell immune profiling were subjected to sequencing and data processing at the Epigenomics Core at Weill Cornell Medicine. This approach provided valuable insights into single-cell gene expression and TCR diversity.
Sample Preparation
Bulk T cells from patient-derived PBMCs were purified and sequenced for detailed gene expression analysis. The preparation of scRNA-seq libraries allowed for single-cell immune profiling and further characterization of T cell responses.
Sequencing and Data Processing
Libraries were prepared for gene expression and TCR sequencing, generating data for analysis. Integrated sequencing data were processed and analyzed to identify gene expression patterns, TCR diversity, and immune response profiles.
CloneTrack
T cell clones were defined based on nucleotide CDR3 sequences and analyzed for expansion, contraction, and memory phases. The identification of vaccine-induced clones, characterization of their lifespans, and survival curves were critical components of the analysis.
PhenoTrack
Phenotyping of cells from specific T cell clones was performed through the GeneVector pipeline. The analysis involved fitting a Gaussian mixture model to assess cell phenotypes and visualize their distribution over various timepoints and phases.
Pgen
The probability of generating each amino acid CDR3 TRB sequence by VDJ recombination, known as Pgen, was computed using OLGA software. This analysis provided insights into the likelihood of specific TRB sequences being generated through genetic recombination.
Gene-Expression Analysis
Differential gene expression analysis was conducted on cells belonging to particular T cell clones, focusing on upregulated genes across different immune response phases. Gene set enrichment analysis was performed to confirm the expression of genes associated with T cell functions.
Whole-Exome Sequencing, Mutation Identification, and Phylogeny
Whole-exome sequencing was utilized to identify somatic mutations and reconstruct phylogenetic trees to infer cancer cell fraction changes between primary and recurrent tumors. This analysis provided valuable insights into tumor evolution and response to treatment.
In Vitro Peptide Stimulation
Peptide pools containing neoantigen sequences were used to stimulate PBMCs in vitro, assessing T cell degranulation, TNFα, and IFNγ production. This approach helped evaluate T cell responses to specific neoantigens.
Epitope Spreading
To assess epitope spreading in responders, PBMCs were stimulated with neoantigens not included in the vaccine. This analysis aimed to identify T cell responses to a broader range of antigens and assess the immune response specificity.
Flow Cytometry and Cell Sorting
Flow cytometry and cell sorting were used to analyze various cell populations and identify antigen-specific T cell clones through in vitro neoantigen rechallenge. Specific markers and staining protocols were employed to characterize T cell responses and activation.
Statistical Analyses
Statistical analyses were performed to evaluate survival curves, compare groups, analyze gene expression data, and assess immune responses. Various tests, including log-rank tests, Mann-Whitney tests, and Wilcoxon tests, were used to determine statistical significance in different analyses.
In conclusion, the study design, patient characteristics, treatment protocols, immune response evaluations, and data analysis procedures described in the article provide valuable insights into personalized cancer vaccine therapy and the assessment of T cell responses in patients with PDAC. The comprehensive approach to assessing immune responses, identifying specific T cell clones, and evaluating gene expression patterns contributes to a better understanding of the dynamics of immunotherapy in cancer treatment.