Personalized Viral-Based Prime/Boost Immunotherapy Targeting Patient-Specific Or Shared Neoantigens: Immunogenicity, Safety, And Efficacy Results From Two Ongoing Phase I Studies.

3137 Background: Neoantigens are key targets of a tumor-specific immune response and CD8 T cells targeting neoantigens drive clinical benefit in patients (pts) treated with checkpoint inhibitors. Methods: Two Phase I studies are being conducted to assess the safety, immunogenicity, and early clinical activity of a viral-based neoantigen-targeting prime/boost immunotherapy aimed at maximizing the CD8 T cell response. Both studies use a chimpanzee adenovirus prime followed by increasing doses of repeat boosts with a self-amplifying mRNA in combination with IV nivolumab +/- SC ipilimumab. In the first study, GO-004, patient-specific neoantigens are predicted using Gritstone's EDGE model and incorporated into both prime/boost vectors. In GO-005, shared neoantigens derived from common driver mutations (including several from KRAS) are encoded in off-the-shelf prime/boost vectors. Results: To date, 12 pts have been treated: 6 pts with GEA, NSCLC, or MSS-CRC (GO-004) and 6 pts with NSCLC, MSS-CRC, or PDA (GO-005) with all pts receiving IV nivolumab and 5 pts also receiving SC ipilimumab. Nine pts continue to receive study treatment. No DLTs have been observed. Treatment-related AEs are reversible and include Grade 1/2 fever (7/12), injection site reactions (4/12), fatigue (3/12), diarrhea (2/12), hypotension (2/12), pruritus (2/12), skin reactions (2/12), anorexia (1/12), dyspnea (1/12), hyponatremia (1/12), infusion-related reactions (1/12), myalgia (1/12), and asymptomatic Grade 3 CK elevation (1/12). At the time of analysis, 8 of 12 pts with ≥ 1 radiographic assessment have a best response of stable disease (SD) (3) and progressive disease (PD) (4), and one pt with no evaluable disease at baseline continues on study > 8 months. In GO-005, 1 pt with SD has a 20% reduction in tumor dimensions that correlates with a decrease in ctDNA. In 4 pts in GO-004 analyzed to date, all pts showed substantial neoantigen-specific CD8 T cell responses to multiple neoantigens after priming which increase further in 2 of 3 pts analyzed after subsequent boosts. In GO-005, 1 of 3 pts showed a robust KRAS G12C-specific CD8 T cell response. Induced T cells express IFNg and granzyme B, consistent with an effector response. Conclusions: Taken together, these early data support the tolerability of a viral-based prime/boost immunotherapy, demonstrate marked immunogenicity, and are consistent with potential clinical activity. Additional pts and data at higher dose levels will be presented. Clinical trial information: NCT03639714, NCT03953235 .