Abstract A024: Tumor kinetic modeling of patients with gastric cancer reveals evolutionary factors associated with worse prognosis

Abstract Background: The failure of long-term treatment responses in cancer is associated with the evolution of resistance. While this resistance has been demonstrated in some cases to be due to the existence of resistant tumor subpopulations (subclones), it remains unclear whether the extent of pre-existing resistance or the rate of rebound of resistant subclones drives the time to treatment failure. Methods: We analyzed a clinical dataset consisting of 421 gastric tumor patients with repeated measurements of tumor volume over a four year period, as well as their final outcome (“living”, implying that patients were still alive at the time of study conclusion versus ”deceased”, for patients who died during the study timeframe). We used a two-population evolutionary tumor kinetics model consisting of a sensitive and resistant population, each with their own growth rate (Gs and Gr, respectively) and fraction of cells (Fs and Fr). We fitted each of the patient tumor kinetic profiles to either a two-population model or a one-population model (where the existence of a resistant or sensitive subpopulation was not visible). We then examined the distributions of Gs, Gr, Fs and Fr, as well as the initial and final tumor volumes in the living and deceased patient populations. Results: We were able to extract tumor kinetic parameters from ∼300 patients in this gastric tumor dataset. Several parameters differed strongly between living and deceased populations: Volume at baseline, Volume at final sample time, rate of growth of resistant cells (Gr). Growth rate of sensitive cells (Gs i.e., rate of tumor shrinkage) did not differ between the two populations. Further, the fraction of resistant cells (Fr) did not differ between living and deceased patients when considering bi-exponential curves (i.e., those in which some cells responded to treatment). Conclusions: The rate of rebound of the resistant tumor subclone (Gr) is a critical determinant of overall survival for patients in this dataset. Notably, for tumors showing response to treatment, neither the rate of tumor shrinkage (Gs), nor the fraction of resistant cells (Fr, related to the depth of RECIST response) was associated with increased survival. These results have implications for the design of clinical endpoints and therapeutic strategies for anticancer treatments. Citation Format: Luke Edwards, Christine Neumar, Madison Stoddard, Douglas E. White, Dean Bottino, Arijit Chakravarty. Tumor kinetic modeling of patients with gastric cancer reveals evolutionary factors associated with worse prognosis [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics; 2025 Oct 22-26; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2025;24(10 Suppl):Abstract nr A024.