Li, Hu

4185 Background: Elraglusib is a first-in-class inhibitor of GSK-3ß, a well-credentialed target in cancer implicated in intrinsic oncologic processes and tumor immune response. Preliminary results of the 1801 Part 3B trial (NCT03678883) showed statistically significant benefits for elraglusib+GnP versus GnP for 1-year survival and mOS in mPDAC (companion abstract: Mahalingam et al.). We investigated plasma levels of cytokines/chemokines/soluble cell receptors/growth factors (CCSG) as potential biomarkers of favorable outcomes on elraglusib. Methods: Forty CCSGs were evaluated in pre-dose plasma from patients with previously untreated mPDAC enrolled in 1801 Part 3B treated with GnP (n = 78) or elraglusib+GnP (n = 155) using a Luminex immunoassay. Using Kaplan-Meier statistics and cutpoint determination, CCSGs were assessed for OS predictive ability in the elraglusib+GnP arm. Multivariate models were constructed to predict binarized survival at 12 months using machine learning (ML). Fivefold cross-validation was used in both analyses, and all methods were applied to the GnP arm to identify elraglusib+GnP specific predictors. Results: Data shown as of November 15, 2024. Multiple CCSGs significantly stratified elraglusib+GnP patients. The most extreme prognosticators were IFN-ß (average HR = 2.34), and PD-L1 (average HR = 0.52) (HR shown as high CCSG vs low CCSG). Screening of different ML approaches ranked logistic regression at the top, and hyperparameter grid search identified stochastic gradient solver with ridge regularization as the optimal method. The model had an accuracy of 88% (SD: 3.9%) and a balanced accuracy of 80% (SD: 8.5%). IFN-beta had the most substantial effect size (odds ratio (OR) = 72.28), followed by IL-18 (OR = 23.38) and PD-L1 (OR = 15.32). All other CCSGs had an OR between 0.03 and 6.71. When this model was applied to patients in the GnP arm, accuracy was 68.4%, and balanced accuracy was 43.1%. Conclusions: Many CCSG biomarkers were identified as promising predictors of survival benefit in mPDAC patients treated with elraglusib+GnP. Both univariate statistical and multivariate ML approaches show predictive significance with high interpretability. The initial ML model is specific to elraglusib treatment, not GnP alone. These results indicate that patients’ initial immune state plays a role in response to elraglusib+GnP. However, single CCSGs for enrichment currently exclude too many patients ( > 75%), and thus, panels of biomarkers are being investigated with ML to overcome this limitation. The 1801 Part 3B clinical trial recruitment has been completed, and updated biomarker models reflective of topline OS data, available by April 2025, will be presented.

Abstract Cytokines mediate strong immunomodulatory effects in cancer. Elraglusib (9-ING-41), an inhibitor of GSK-3, is a potent antitumor and immune-modulatory agent in phase 2 clinical trials for the treatment of various cancers. Here, we examined the immune environment of elraglusib-treated cancer patients by measuring peripheral blood protein levels to explore whether the expression level of cytokines/chemokines/soluble cell receptors/growth factors (CCSG) is correlated with clinical outcome. Using a customized Luminex platform, the expression of forty CCSG hypothesized to be linked to GSK-3 activity was assessed in pre-dose plasma samples obtained from 45 cancer patients with relapsed/refractory metastatic disease (NCT03678883; the 1801 trial). All patients in the 1801 trial were treated with elraglusib (dose range: 1-15 mg/kg) either as a single agent (Part 1; n=21) or in combination with chemotherapy (Part 2; n=24). Pre-dose peripheral blood protein levels were examined as binary predictors of clinical outcome using optimal cutoff points determined by maximally selected rank statistics. In the aggregate Part 1 and 2 populations (n=45), we identified 13 CCSG that significantly stratified the cohort by overall survival (OS). High pre-dose levels of CD95 and TNFRSF10C were associated with better OS. CD95 was a better prognostic of OS than TNFRSF10C (HR: 2.27, 95% CI: 1.10-4.67, p-value: 0.023). Low pre-dose levels of 11 CCSG were also significantly associated with favorable OS, with IL-8 being the most predictive (HR: 0.137, 95% CI: 0.0545-0.346, p-value: <0.0001). Part 1 and 2 predose samples were also analyzed separately, and many of the biomarkers identified in the aggregate analysis retained significance in the separated analysis. In Part 1, CXCL5, IFN-alpha, and IL-18 emerged as unique markers, and in Part 2, CCL22 was the single unique marker.The results of our exploratory study identified several putative biomarkers of elraglusib clinical benefit and demonstrated potential immunomodulatory mechanisms of elraglusib that will be used to inform the further clinical development of elraglusib for the treatment of metastatic cancer. Citation Format: Taylor Weiskittel, Kelsey Huntington, Leiqing Zhang, Austin Koukol, Benedito A. Carneiro, Hu Li, Andrey Ugolkov, Wafik El-Deiry, Andrew Mazar. Identification of potential immune biomarkers for GSK-3 inhibitor elraglusib (9-ING-41) in patients with relapsed/refractory metastatic cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 6426.

Abstract 5-Fluorouracil (5-FU) is a cornerstone therapy for the treatment of several aggressive tumors. However, nearly 1/3 of patients receiving 5-FU experience severe (grade 3+) adverse 5-FU toxicity. Genetic variations in dihydropyrimidine dehydrogenase (DPD; DPYD gene) are major contributors of 5-FU toxicity. With continued progress in next-generation sequencing technologies, the number of reported genetic variants is substantially increasing; to date, there are more than 450 reported missense variants in the DPYD gene. Only a handful of those variants, including *2A, I560S, D949V, and rs75017182, have been established as predictors of toxicity. Because most of the remaining variants are unstudied, mainly due to their rare occurrence and unequal racial distribution, it is challenging to conduct statistically powered clinical association studies. Therefore, in vitro and computational methods to identify potential toxicity-associated variants have major appeal. The primary goal of the current study was to develop a highly accurate gene-specific in silico classifier to predict deleterious DPYD variants. For this purpose, using a recombinant cellular model, we characterized 69 missense DPYD variants reported in online databases. We then utilized DPD activity data from the 69 variants in combination with data from 87 previously characterized variants (see Offer, SM et al. Cancer Research 2013 & 2014) to develop the DPYD variant classifier model (DPYD-Varifier). DPD protein-specific biochemical and structural features and amino acid conservation scores were integrated into the model's random forest classifier-based algorithm. DPYD-Varifier showed improved accuracy (85% accurate) compared to other in silico tools. Additionally, the model provided mechanistic insights into alterations of DPD function conferred by DPYD variants. In general, deleterious variants were located in closer proximity to important binding/coordinating sites of the DPD protein. Additionally, DPYD-Varifier allowed us to estimate the cumulative allele frequency of all reported deleterious variants in various geographical/racial groups, which can substantially add to our knowledge of DPD deficiency and 5-FU toxicity observed in these populations. In conclusion, the integration of in vitro and accurate computational data from our study permitted classification of relevant DPYD variants, which could be useful to more effectively individualize 5-FU therapy in the future. Citation Format: Shikshya Shrestha, Cheng Zhang, Calvin R. Jerde, Hu Li, Steven M. Offer, Robert B. Diasio. DPYD-Varifier, a computational model to identify 5-FU toxicity-associated DPYD variants [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 3894.