Introduction: The underlying cellular mechanisms of successful immunotherapy are still insufficiently characterized. Especially, priming of tumor-specific T cells against relevant antigens and in situ tumor-recognition is crucial for therapy success in the development of active immunotherapeutic approaches, but remain a serious challenge. Thermal ablation of cancer, e.g. percutaneous radiofrequency ablation (RFA) is a well-established treatment option, particularly for patients with contraindications for surgical resection or in non-resectable primary or secondary liver cancers. RFA induces local tumor destruction by coagulation necrosis, leading to the release of danger signals (e.g. heat shock proteins) and cellular content that can induce and stimulate adaptive and innate immune responses, including tumor-antigen specific T cells. By breaking local immune tolerance, RFA may support the development of antigen-specific anti-tumor responses. Here, we analyzed patient-individual immune responses against autologous HLA-presented tumor antigens after RFA treatment.
Materials and Methods: Patients (n=6) with liver metastases of colorectal carcinoma (CRC) in different liver segments were included. First, non-resectable manifestations were ablated using RFA. Subsequently, patients underwent metachronous liver surgery for resectable residual tumors. Using surgical specimens, immunopeptidome analyses of tumor and corresponding non-malignant liver tissues were performed by uHPLC-coupled tandem mass spectrometry (MS) after immunoprecipitation of HLA-ligands. Ligandome data were combined with results from whole transcriptome sequencing (WTS) in order to select for patient-specific HLA-ligands potentially relevant for the patient's individual tumor. Furthermore, identified peptides were compared with an extensive in house database of HLA ligands, derived from different non-transformed tissues including non-malignant colon (NMC) specimens. Peripheral blood mononuclear cells (PBMC) obtained at pre-defined time points before and after RFA and surgery were stimulated in vitro with these T cell epitope candidates. Functional T cell responses were characterized by intracellular cytokine staining (ICS).
Results: Combining WTS data and analyses of autologous tumor-exclusive HLA-ligandomes, we selected 32 individual T cell epitope candidates for the respective patients. Antigen-specific T cells could be detected against approx. one third of the candidate epitopes. Both boosting of pre-existing as well as newly induced T cell responses against these epitopes could be detected after RFA. With regard to pre-existing immune responses, CD8+ and CD4+ T cell reactivities could be demonstrated against both HLA class I and class II epitopes. In 4/6 patients (66%) analyzed, de novo induced antigen-specific T cells could be detected after RFA. These reactivities comprised CD8+ as well as CD4+ T cells. Of particular interest, in one patient (diagnosed with mCRC) antigen-specific CD4+ T cells against a mutated HLA-class II peptide were detectable, which were induced by RFA. Of note, this epitope could only be detected on the transcriptome level but not directly as an HLA-ligand by MS.
Conclusions: Employing this individualized approach, tumor-antigen specific T cells were more frequently detected as compared to our previous observations employing broadly expressed tumor-associated antigens. These findings suggest that RFA leads to induction of antigen-specific T cells more frequently than previously appreciated, and interestingly also encompasses responses against neo-antigens. The detected immune responses induced by RFA themselves may not be solely sufficient for effectively rejecting established tumors. However, these findings justify the clinical investigation of combinatorial therapies, using immune stimulatory agents or immune checkpoint inhibitors together with thermal ablation.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.