Perioperative inflammation, anesthesia modulation and cellular signaling in lung cancer: an integrated transcriptomic and single-cell analysis.

Background: Patients undergoing lung cancer surgery experience significant systemic inflammatory responses and anesthetic interventions during the perioperative period, which can affect immune status and the tumor microenvironment. Different anesthetic techniques variably regulate inflammation, cytokine signaling, and tissue protection. It remains unclear, however, whether perioperative inflammatory signals leave identifiable molecular patterns at the tissue level. Furthermore, analyses integrating multi-layered data are currently lacking. We performed an integrated transcriptomic and single-cell analysis to identify shared molecular patterns linking perioperative inflammation, anesthetic modulation, and the lung cancer tumor microenvironment, and to localize these pathways to specific cellular communication axes. Methods: Integrated analysis was performed across three levels using publicly available transcriptomic data: peripheral blood, myocardial tissue, and single-cell data from lung cancer. Peripheral blood data delineated acute inflammatory features, identified differentially expressed genes, and performed pathway enrichment to construct a risk score reflecting inflammatory burden. Paired myocardial tissue samples were compared to assess pathway alterations before and after anesthesia with propofol versus sevoflurane, evaluating the directional influence of the anesthetic regimen on inflammation-related regulation. Cell annotation and composition statistics were then performed on single-cell data from lung cancer and adjacent tissues. This analysis localized the cellular origins of major pathways and reconstructed cell communication networks from ligand‑receptor interactions to identify signaling hub cells and key communication axes. Results: Analysis of peripheral blood revealed enrichment in autophagy, protein processing, and inflammatory pathways. The risk score increased progressively across the Healthy, D1 sepsis, and D2 sepsis groups, reflecting the intensity of the acute inflammatory response. A comparison of anesthesia methods showed that samples from both groups exhibited enrichment in cytokine signaling and inflammatory pathways, though the specific genes involved, the degree of enrichment, and the regulatory patterns differed. In the single-cell data, T cells, natural killer (NK) cells, and macrophages constituted the major cell populations, with inflammatory and immune-related pathways active across multiple cell types. CellChat analysis revealed that macrophages/monocytes and epithelial/endothelial cells participated in dense communication networks characterized by antigen-presentation-related interactions [human leukocyte antigens (HLA) class I and II signaling with CD8A and CD4] and extracellular matrix-associated interactions (FN1- and collagen-based signaling through integrins, CD44, and syndecans), distinguishing tumor tissue from adjacent normal lung tissue. Signals related to MHC, FN1, and COLLAGEN were enhanced in tumor tissue, where multiple ligand-receptor pairs demonstrated high communication probabilities. Conclusions: A continuous transcriptional link connects acute inflammation, anesthetic techniques, and the tumor microenvironment. Within lung cancer tissues, monocytes/macrophages, T cells, and epithelial cells primarily mediate the pathways regulated by inflammation and anesthesia, establishing a communication network via antigen presentation, cell adhesion, and matrix‑related signaling. These results offer a molecular foundation for understanding how perioperative management relates to the lung cancer immune microenvironment.