Spatial transcriptomics for profiling the microenvironment of cervical precancer progression from routine H&E slides: Applications for pharmaceutical development.

Abstract

e17536 Background: High-grade cervical intraepithelial neoplasia (HG-CIN) affects 200,000 U.S. women annually, with rising prevalence among those aged 30 & older, many of whom missed the opportunity for HPV vaccination. Treatment remains predominantly surgical, with a notable absence of pharmacologic interventions. Clinical trial endpoints rely on histological regression, but standard microscopy cannot accurately distinguish between regressing and progressing lesions. Compelling evidence underscores the tumor microenvironment’s (TME) critical role in driving immune suppression and malignant transformation. Leveraging a novel spatial imaging platform, this study provides high-resolution molecular insights into HG-CIN directly from routine H&E-stained slides, enabling precise characterization of immune marker distributions and their implications for disease progression. This technology has the potential to transform tissue analysis and significantly impact drug development. Methods: Sixteen formalin-fixed paraffin-embedded (FFPE) tissues were obtained from the health system’s Biospecimen Repository, with twelve used to optimize the sequencing protocol directly from H&E slides. Sample was categorized as CIN3, benign, or micro-invasive cervical mucosa based on the history, Pap smear/HPV test results, and an expert review by a Gynecologic Pathologist. Library generation and sequencing were conducted using Visium Gene Expression (both SD & HD platforms) kits for FFPE (10x Genomics). Specific genetic markers for immunosuppressive cell activities were highlighted to demonstrate its differential presence in CIN3 vs normal cervical tissue. Results: Sixteen FFPE samples were analyzed, with 11 used to optimize sequencing protocols. Three CIN3, three normal, and one micro-invasive cervical cancer sample revealed distinct tumor, stromal, and immune microenvironments. CIN3 tissues had elevated Tregs markers (FOXP3, CD4), MDSCs (CD11b, CD33, HLA-DR), and the Treg trafficking molecules CCR7 and TGF-β within stromal regions. Additionally, CIN3 had high gene expression for CDKN2A, which encodes the immunosuppressive ligand PDL1. Conclusions: The CIN and cancer microenvironment demonstrated strong immunosuppressive markers across immune, stromal, and epithelial compartments, aligning with key known features of tumorigenesis and progression. This spatial imaging platform & optimized workflow captures biologically relevant data directly from standard routinely stained H&E slide, achieving near single-cell resolution, enabling histopathologic comparison to gene expression data. Our study revealed unique immune suppression profiles in CIN3 linked to HPV progression, highlighting the platform’s potential to transform tissue analysis and advance drug development applications.