Search

TPS486 Background: A heterologous prime-boost vaccination strategy using GVAX pancreas and CRS-207 is showing promise in patients with pancreatic adenocarcinoma (PDA) (Le, JCO 2015). Furthermore, blockade of the immune checkpoint programmed death-1 (PD-1) is active in some cancers. Combinatorial strategies aimed at priming tumor antigen-specific T cells while simultaneously blocking negative checkpoints may be necessary to improve outcomes in PDA. GVAX is composed of allogeneic pancreatic cancer cells modified to express GM-CSF and induces a broad response against multiple tumor antigens. GVAX is given with low-dose cyclophosphamide (CY) to inhibit regulatory T cells. CRS-207 is live-attenuated Listeria monocytogenes engineered to express the tumor-associated antigen mesothelin. CRS-207 boosts responses against mesothelin and is unique in its capacity to stimulate both innate and adaptive immunity by activating T cells and NK cells. Nivolumab is an antibody against PD-1. Methods: This is a phase 2 study comparing CY/GVAX and CRS-207 with or without nivolumab in subjects with PDA who failed only one chemotherapy regimen for metastatic disease. Subjects are randomized in a 1:1 ratio to receive either 2 doses of CY/nivolumab/GVAX and 4 doses of nivolumab/CRS-207 (Arm A) or 2 doses of CY/GVAX and 4 doses of CRS-207 (Arm B). The primary objective is to compare OS between Arms A and B. Secondary/exploratory objectives include: assessment of safety and clinical responses (tumor assessments and CA19-9 levels) and correlation of Lm- and mesothelin-specific T cell and other immunological responses with OS, progression-free survival and best overall response. Clinical trial information: NCT02243371.

A major obstacle to the development of immunotherapy for pancreatic ductal adenocarcinoma (PDA) has been identifying immune relevant antigens that are not expressed by normal cells. DNA sequencing technologies that allow for the genome-wide identification of tumor mutations have shown that human PDAs harbor an average of 63 mutations (called the “mutome”). Since these mutations are not present prior to the onset of tumorigenesis, each mutation generates a novel and essentially foreign antigen (termed “epitope”). Unlike most classical tumor epitopes, mutome-derived epitopes are uniquely specific to each patient’s tumor. Furthermore, T cells recognizing them may avoid immunological tolerance mechanisms that inhibit or deplete T cells that recognize classical tumor epitopes that are also expressed by normal cells. Vaccines designed to activate mutome-specific T cells have the potential to induce robust and specific antitumor responses and provide a new approach for personalized tumor immunotherapy. We are using the Panc02 murine PDA model to develop and explore the potential of this approach. Sequence analysis identified 596 point mutations in Panc02 cells, significantly more than the number found in human PDAs. The NetMHC prediction server was used to screen for epitopes predicted to bind H2-Db and H2-Kb. In total, 15 strong candidates (NetMHC scores 100 spots per 1x105 T cells). Antitumor efficacy was evaluated by immunizing mice 3 days following subcutaneous Panc02 challenge. Only immunization against the 5 robust peptides was capable of preventing tumor outgrowth, and was significantly more effective than treatment with GM-CSF-secreting whole-cell vaccination (GVAX), another approach currently being investigated in the clinic by our group. We are currently testing the efficacy of this approach against larger established tumors and metastatic liver lesions. We are also working to optimize this approach by testing additional adjuvants and vaccine vectors, including recombinant Listeria monocytogenes that express the mutant gene sequences. Exact epitopes are being mapped in order to evaluate the utility of MHC-binding predictions. Ultimately, we intend to translate our optimized approach into the clinic.