Longitudinal plasma proteomic analysis: A monitoring strategy for NSCLC patients treated with immunotherapy.

8579 Background: Real-time monitoring is critical for tailoring treatments to individual patient responses in clinical oncology. Plasma proteomics offers a comprehensive systemic view of disease progression, tumor activity, immune responses, and various biological processes, making it a powerful tool for clinical decision-making. This study explores the feasibility of three specific plasma proteomic signatures for longitudinal monitoring of treatment responses in patients with non-small cell lung cancer (NSCLC) undergoing therapy with immune checkpoint inhibitors (ICIs). Methods: Plasma samples were collected from patients with advanced NSCLC receiving PD-1/PD-L1 inhibitor-based regimens. Cohort-1 (n=225) includes samples collected before treatment (T0) and 4-6 weeks after treatment initiation (T1). Cohort-2 (n=56) included samples collected pre-treatment and every three months, up to 36 months. Aptamer-based proteomic profiling quantified ~7,000 plasma protein analytes per sample. Three proteomic signatures were derived from T0–T1 changes in Cohort-1 and tracked in Cohort-2, then compared with radiologic imaging-based response evaluation. Results: Three distinct plasma proteomic signatures were identified. The first, featuring soluble PD-1 and PD-L1, indicates drug presence in circulation. The second reflects T-cell activation (e.g., CD8A, LAG3, IL2R), linked to drug uptake, without confirming a favourable tumor response. The third includes intracellular proteins indicative of lung tissue damage, allowing dynamic disease monitoring. Lung tissue damage signature correlated with radiologic imaging-based response evaluation (PR: n = 79, –4.19 [–12.47, 3.58]; SD: n = 125, 1.03 [–1.87, 5.01]; PD: n = 30, 3.37 [0, 7.27]; KW P-value = 0.01). Longitudinal analysis of these signatures facilitated early detection of non-responders in an average of 6.6 months [4 - 9.2 months, n=13] prior to radiologic evaluation. Among progressors, nine cases identified responders who later developed acquired resistance, distinguishing them from patients who did not respond to therapy at all. These findings highlight the potential of proteomic profiling to provide comprehensive systemic insights. A comparative analysis with ctDNA will also be presented to further validate these results. Conclusions: Our study demonstrates the feasibility of using plasma proteomic signatures to monitor responses to ICIs in NSCLC. We highlight the potential and emphasize the need to further develop these plasma-based monitoring tools through more extensive prospective studies. Such advancements are essential for establishing proteomic signatures as dependable decision-support tools in NSCLC treatment protocols.