Medical Student Icahn School of Medicine at Mount Sinai, United States
Introduction: Although considered benign, up to 20% of pituitary adenomas (PAs) may invade the cavernous sinus, causing significant morbidity and complicating surgical resection. Predicting likelihood of invasion may allow for early, aggressive intervention in at-risk patients. Tumor features such as location and lineage are hypothesized to impact invasion, but a distinct molecular profile and mechanism underlying invasion has yet to be characterized. Here, we examine the largest transcriptomic cohort of PAs across lineages to date, identifying a distinct transcriptomic signature associated with cavernous invasion.
Methods: For n=568 cross-lineage PAs (285 nonfunctioning, 47 lactotroph, 168 somatotroph, 43 corticotroph, 7 gonadotroph, 4 thyrotroph,14 plurihormonal), RNA-sequencing or microarray data was obtained from publicly available sources. Data was harmonized across datasets and residualized for covariates using SVA. Differential expression analysis was performed using linear mixed models with limma-voom across 10,048 genes. Significance was determined using Benjamini-Hochberg multiple testing correction FDR < 0.05. Functional enrichment of biological processes was assessed using gene ontology with FGSEA. To assess gene co-regulation, network analysis was performed using WGCNA.
Results: Cavernous invasion was significantly associated with differential expression of 116 genes (81 upregulated, 35 downregulated) across lineages. In contrast, lineage-specific markers of invasion were not found. Invasion-implicated genes included known mediators of cell migration, such as PLXND1 (p=3.30e-06) and mitotic regulators PSRC1 (p=2.14e-06) and ACTB (p=3.15e-06). Upregulated genes enriched for migratory and angiogenic processes (p=3.98e-07), while hormone secretion was downregulated (p=1.06e-04). Co-expressed gene networks involved angiogenesis (p=3.80e-02), proto-oncogene activity (p=6.70e-03), and reduced immune surveillance (p=6.20e-07), implicating key hub gene and interferon-modulator IRF6.
Conclusion : A distinct transcriptomic signature implicating angiogenesis, cell migration, and immune escape is associated with cavernous invasion, suggesting that molecular profiling may distinguish invasive PA subtypes across lineages.