OS07.8.A INNOVATIVE DYNAMIC STRATIFICATION IN GLIOBLASTOMA: ANALYZING TUMOR CELL CHARACTERISTICS AND MOBILITY STRATEGIES THROUGH SINGLE CELL BEHAVIOR LIVE IMAGING (SCBLI)

Abstract BACKGROUND Glioblastoma, a highly aggressive brain tumor, presents a daunting prognosis with an average survival of under 15 months due to its invasive nature. Understanding the specific behaviors of individual tumor cells in terms of volume, shape, and movement dynamics is crucial for improved patient stratification and understanding tumor mechanisms. METHODS we developed Single-Cell Behavior Live Imaging (ScBLI) analysis, a novel approach combining live imaging with advanced mathematical processing. Our method enables precise analysis of morphological and behavioral traits in patient-derived tumor cells cultured in vitro. Leveraging sophisticated cell tracking methods, we analyzed 21 glioblastoma and 1 healthy astrocyte (HA) cell cultures, to capture detailed movement information. Our analyses extend beyond mere tracking; we extrapolated an overarching mobility strategy employed by the cells within each culture. By measuring the delta scaling parameter, we discern between subdiffusive motion, random Brownian motion, and more organized efficient strategies such as a typical Levy Walk. We correlate variations in delta scaling with cell applied stimuli, such as invasion and antineoplastic agents, enhancing our understanding of tumor cell behavior. RESULTS We analyzed 100 cell trajectories per culture, totaling 2200 trajectories. Parameters as distance traveled, volume, track displacement, directional change rate, circularity, mean velocity, and confinement ratio were used to identify 3 distinct cell populations with different distributions within each GBM and HA culture facilitating the stratification of samples into 3 clusters. Moreover, the delta scaling was divided into three groups (0.28 to 0.80), and its integration with clinical data revealed an increased percentage of patients with shorter overall survival as scaling increased. Higher delta scaling indicated a more organized system adapted to challenging environments. Transcriptome analysis unveiled gene expression differences, highlighting potential therapeutic targets. CONCLUSION This approach provides a deeper understanding of how tumor cells navigate and occupy the available space, shedding light on their dynamic behavior in vitro. Our results indicate that cancer cells must preserve the control of their locomotion to explore space in an organized and efficient way for survival, similar to the organized and functional behavior of healthy cells. By integrating the unique characteristics of each patient’s tumor cell behavior with clinical data, this approach offers a new method for stratifying patients with different survival probabilities. Overall, by combining imaging, image processing algorithms, and dynamic modeling, we provide a valuable framework for understanding the complex dynamics of glioblastoma and paving the way for more personalized and effective treatment approaches.