Abstract 5808: Differential clinical outcomes of immune vs non-immune based therapeutics in gastrointestinal malignancies: A retrospective, single center experience of early phase studies

Abstract Introduction: Evidence suggest that the immunosuppressive environment of liver parenchyma, leads to a specific resistance to immunotherapy (IO) in patients with liver metastasis (LM). This study seeks to evaluate the outcomes of patients (pts) with advanced gastrointestinal (GI) cancer who have undergone treatment with novel therapeutics (IO vs non-IO agents) within early-phase clinical trials. Further, we aim to identify oncogenic molecular signaling pathways and phenotypes that may indicate a poor response to IO in patients with LM. Methods: This retrospective study focused on patients treated at Northwestern University from August 2014 to August 2023, who were enrolled in early-phase clinical trials and had CLIA certified next-generation sequencing available. Standard radiologic evaluations were per study, and efficacy based on progression-free survival (PFS) were defined as responders (PFS ≥ 3 months) vs non-responders (PFS < 3 months). Survival curves were calculated using the Kaplan-Meier method. For responder vs non-responder analyses, logistic regression assessed the impact of clinical variables on response, with significance defined as p ≤ 0.05. Results: 265 pts participated in this study with 11 pts participating in two studies during the course of their visits to the unit. The median age was 65 years (range 55-72), with 52% being male. Three tumor types were represented: colorectal (49%), pancreatic (43%), and gastroesophageal (8%) cancers. 218 (82%) pts had documented LM. 102(39%) patients were treated with IO based vs 157 (61%) in non-IO based therapeutics. When evaluating efficacy (n=276), the median PFS was 2.28 months, and 43% of patients had a PFS rate of ≥ 3 months. We found that pts with treated non-IO-based therapy were associated with higher PFS, median 3.48m (1.68-6.12m) vs 1.8m (1.4-3.48) for IO-based therapy (p<0.0001). In contrast to other studies, LM did not have prognostic significance for patients treated with IO (p=0.49). Among IO-based LM patients, KRAS (OR: 6.35, p=0.028) and TP53 (OR:4.23, p=0.044) pathway mutations were significantly associated with higher odds of non-response. No significant differences were found in the pathways involving APC, CDKN2A, ARID1A, PIK3CA, PTEN, SMAD4, NOTCH1, RB1, FBXW7, and FGF4. Conclusion: In contrast to findings from earlier studies, our data did not establish a significant association between LM and survival outcomes in patients receiving IO. This discrepancy may result from the variety of IO regimens used in early-phase trials and the limited number of patients. Nonetheless, our findings contribute valuable insights into this complex area, offering important perspectives for future research and potential therapeutic strategies, particularly as IO treatments evolve and more extensive patient data becomes available. Citation Format: Tarik Demir, Nicole Altomare, Katrina Dobinda, Ruohui Chen, Carolyn Moloney, Pedro Hermida de Viveiros, Aparna Kalyan, Young Kwang Chae, Devalingam Mahalingam. Differential clinical outcomes of immune vs non-immune based therapeutics in gastrointestinal malignancies: A retrospective, single center experience of early phase studies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 5808.