Poster #56: Janvi Garg

Computational Synthesis of Inflammatory and Nutritional Biomarker Signatures for Precision Risk Stratification in NSCLC Immunotherapy: A Translational Bioinformatics Approach

Twesha Singh, ATLAS Researcher, ATLAS (Advocates in Technology Leadership and Science) Research Group; Sriram Siva, ATLAS Researcher, ATLAS (Advocates in Technology Leadership and Science) Research Group; Shivam Loyalka, ATLAS Researcher, ATLAS (Advocates in Technology Leadership and Science) Research Group; Tanvi Chakka, ATLAS Researcher, ATLAS (Advocates in Technology Leadership and Science) Research Group; Janvi Garg, BS (Baylor University), MS (Columbia University), ATLAS (Advocates in Technology Leadership and Science) Research Group.

Systemic inflammation and nutritional status are increasingly recognized as quantifiable biological signals shaping prognosis and treatment response in non-small cell lung cancer (NSCLC) patients receiving immune checkpoint inhibitors (ICIs). This project synthesizes circulating cytokine profiles, composite inflammatory-nutritional indices (CRP/albumin-based scores, neutrophil-to-lymphocyte ratio [NLR], prognostic nutritional index [PNI]), and clinical outcome data to build computationally derived risk-stratification frameworks applicable to precision oncology workflows. Frequency data, effect estimates (odds ratios, hazard ratios), and significance values were synthesized across NSCLC cohorts, spanning baseline cytokine profiling, longitudinal changes in NLR/PNI and composite scores during ICI therapy, and downstream outcomes including early mortality, ICI-associated acute kidney injury (AKI), and metabolic sequelae of chronic inflammation. Effect sizes were benchmarked using standardized interpretive thresholds to enable cross-study comparability, consistent with translational bioinformatics approaches to heterogeneous clinical datasets. Advanced NSCLC patients showed significantly elevated IL-6, IL-8, IL-10, and TNF-RI relative to controls; elevated IL-6 correlated with an exhausted T-cell phenotype (increased PD-1, impaired STAT1 signaling) and a threefold reduction in survival. Composite CRP/albumin indices (Glasgow Prognostic Score, CRP-to-albumin ratio, PNI) consistently stratified survival and recurrence risk (HR 2 to 4, p < 0.01). A dynamic scoring model integrating on-treatment NLR/PNI trajectories (IMMUNO-SCORE) achieved an AUC of approximately 0.79 for predicting bone metastasis response, with tenfold differences in response rates across score strata. Early mortality (≤60 days) occurred in approximately 13% of patients, predicted by liver metastasis and poor performance status (adjusted OR 8 to 10); AKI incidence approached 25%. This synthesis demonstrates that integrating multi-marker inflammatory and nutritional data into dynamically updated composite scores yields robust, generalizable predictive models for immunotherapy response, toxicity, and survival. These findings support computational risk-stratification tools that combine longitudinal biomarker trajectories with clinical variables, an approach directly extensible to multi-omic data integration for future precision immuno-oncology applications.