Characterizing plasma circulating tumor RNA (ctRNA) landscape in advanced NSCLC: Preliminary insights into treatment resistance and tumor microenvironment.

167

Background: Plasma circulating tumor RNA (ctRNA) is currently used for gene fusion detection in non–small cell lung cancer (NSCLC). Beyond the detection of gene fusions, plasma ctRNA profiling has the potential to capture tumor-intrinsic transcriptional programs and features of the tumor microenvironment (TME). This may complement genomic profiling via ctDNA which is limited by tumor shedding. This exploratory study evaluated the feasibility of plasma ctRNA expression to identify molecular features associated with treatment response and acquired resistance in NSCLC. Methods: Plasma samples (n=115; 100 treatment-naïve and 15 EGFR -mutant at progression [PD]) were analyzed via amplicon-based targeted NGS. Expression levels were normalized to the geometric mean of control genes and log-transformed. Two primary analyses were conducted: (1) baseline differential expression comparing responders (R) versus non-responders (NR), and (2) comparative analysis of baseline EGFR -mutant samples versus those at PD to identify resistance-associated transcriptional shifts and pathway enrichment. Additionally, a curated panel of immune signatures—encompassing T-cell cytotoxicity, recruitment, exhaustion, B-cell/APC niches, and myeloid suppression—was utilized to characterize the baseline immune landscape from plasma. Results: At baseline, higher expression of CD70 and CD8B correlated with NR status in patients receiving chemotherapy plus immune checkpoint inhibitors (ICI). However, no significant differential expression patterns were identified in other treatment cohorts relative to response. The custom immune signature panel revealed significant inter-patient heterogeneity, successfully capturing distinct immune phenotypes—ranging from "inflamed" to "immune-excluded" (cold) barriers—directly from plasma. In the EGFR-mutant cohort at the point of TKI progression, significant upregulation was observed in genes associated with: TGF-β signaling and extracellular matrix organization (p < 10 ^-5 ), DNA damage response pathways (p <10 ^-4 ), immune regulatory and suppressive signatures (p <0.05). Conclusions: Preliminary evidence suggests that plasma ctRNA captures critical shifts in the TME that are not visible via ctDNA alone. The ability to profile immune exhaustion and myeloid suppression signatures using a specialized transcriptomic panel offers a potential framework for non-invasive, personalized immunotherapy selection. These data support the clinical utility of ctRNA as a tool for identifying therapeutic vulnerabilities. Further validation in larger, longitudinal cohorts is required to confirm these signatures as predictive biomarkers.