Spatial transcriptomic prioritization of anesthesia-associated candidate genes in triple-negative breast cancer: an exploratory multi-omics reappraisal

Abstract Background Perioperative anesthetic exposure has been linked to tumor biology in experimental and clinical studies, but public-data analyses can overstate causal and clinical implications when genetic instruments, disease endpoints and validation datasets are incomplete. Triple-negative breast cancer (TNBC) remains an aggressive breast cancer subtype with substantial immune and stromal heterogeneity. We performed a conservative multi-omics reappraisal of anesthesia-associated candidate genes in TNBC, emphasizing evidence that could be supported by the available public datasets. Methods Thirty-nine anesthesia-associated genes were curated from public gene-disease and toxicogenomic resources. Genetic evidence was audited using available breast cancer GWAS and eQTL resources, with Mendelian randomization (MR), SMR and colocalization outputs classified by data completeness and interpretability. TNBC spatial transcriptomics from GSE210616 was analyzed at the patient level across three Visium samples and 5,103 spots. Moran's I was summarized per patient and combined using meta-analytic evidence. TLS-associated transcriptional hotspots, spatial niches and ligand-receptor co-expression were evaluated with explicit recognition that spots are technical observations nested within patients. Results The genetic evidence did not support a definitive causal claim for the main TNBC candidates. In the analyzed MR resources, a complete pipeline-level result was available for SCN9A, whereas corresponding complete eQTL-supported MR results were not available for SDC1, ZNF521, LPO, TRPA1 and GSDME. Spatial transcriptomics provided the strongest directly usable evidence. SDC1 showed consistent positive spatial autocorrelation in all three TNBC patients, with a mean Moran's I of 0.3042 and a patient-level Stouffer meta-analysis P value of 7.92e-06. TLS-associated transcriptional hotspots comprised 192 of 5,103 spots (3.8%) and were patient-dependent, with 125, 67 and 0 hotspot spots in the three samples. SDC1 expression was higher in TLS-associated hotspots (P = 5.75e-11), and SDC1-CD44 spatial co-expression was positive (Spearman r = 0.3307, P = 1.69e-130). In contrast, SDC1 showed weak negative correlation with the M2 macrophage score (Spearman r=-0.0336, P = 0.0165), indicating that this axis is not supported as a strong positive macrophage mechanism. Conclusions The available evidence supports an exploratory spatial and genetic-prioritization study rather than a definitive causal study of anesthesia-driven TNBC metastasis. SDC1 is the most supportable candidate because it shows consistent patient-level spatial clustering and TLS-associated spatial enrichment in TNBC tissue. Further TNBC-specific genetic outcomes, independent external cohorts and experimental validation are needed before clinical or mechanistic claims can be made.