90 Utilizing spatial transcriptomics to define potential biomarkers of immune checkpoint blockade in appendiceal cancer – a proof of concept pilot study

<h3>Background</h3> Appendiceal cancer’s mucinous nature and scant cellularity significantly restrict traditional molecular analyses. Single cell sequencing overcomes technical challenges of bulk sequencing but lacks spatial information. To address this, we employed digital spatial profiling to select regions of interest exclusive of mucin pools to better isolate the tumor compartment. Specimens from patients in a recently completed trial of combination immune checkpoint blockade (ICB) and VEGF inhibition (Atezolizumab-Bevacizumab) in metastatic appendiceal cancer¹ were utilized. This was the first trial of ICB in appendix cancer and demonstrated efficacy in this rare malignancy. This offers a mechanism to deconvolute the microenvironment of appendiceal tumors while uncovering biomarkers of response to ICB. <h3>Methods</h3> Spatial whole transcriptome sequencing was performed on FFPE tissue sections from two moderately-differentiated and two well-differentiated metastatic appendiceal cancers from patients on trial (NCT03074513)¹ using nanoString’s GeoMX Digital Spatial Profiling Platform. Sequential slides were stained with either H&amp;E or fluorescent markers for DNA (SYTO 13), epithelium (PanCK), bacteria (16s) and immune cells (CD45). In collaboration with an appendiceal expert gastrointestinal pathologist, 95 representative regions including tumor, normal epithelium, stroma and immune aggregates were selected. Unsupervised dimension reduction was performed using Uniform Manifold Approximation and Projection (UMAP). <h3>Results</h3> From 95 regions collected, 76 passed quality control focused on filtering regions with low sequencing reads (&lt;1000), nuclei count (&lt;10) and gene detection rate (&lt;400 genes). Overall, we detected ~14,000 genes above the limit of quantification. Survey of the tumor microenvironment identified small tumor islands (10-200 cells) within expansive pools of mucin (figure 1). We analyzed 6 tumor islands with only one specimen demonstrating bacterial co-localization with tumor. UMAP analysis differentiated transcriptionally distinct normal appendix and adenocarcinoma populations (figure 2). Furthermore, immune aggregates potentially representing two molecularly distinct subtypes of Tertiary Lymphoid Structures (TLSs) were noted in the peri-tumoral vicinity, with TLS1 containing more plasma cells (p=0.015) and a trend towards higher M2 macrophages (p=0.060). TLS2 aggregates demonstrated higher clonality (Gini coefficient p&lt;0.001) and lower diversity (Shannon H p&lt;0.001) (figure 3). <h3>Conclusions</h3> We successfully identified transcriptionally distinct populations of cells in adenocarcinoma vs. normal tissue and uncovered two potential TLS subtypes in the peri-tumoral landscape. This work represents a significant advancement in studying these rare tumors. Additional specimens from trial participants are undergoing analysis to expand this pilot study. By profiling the landscape of the tumor and peri-tumoral tissue immune microenvironment and correlating with clinical outcomes, we hope to identify biomarkers of therapeutic response. <h3>Trial Registration</h3> NCT03074513 Atezolizumab and Bevacizumab in Treating Patients With Rare Solid Tumors. <h3>Reference</h3> Hornstein NJ, Zeineddine MA, Gunes BB, Pellatt AJ, Knafl M, Zhu H, Willett AF, Yousef A, Liu S, Sun R, Futreal A, Woodman SE, Taggart MW, Overman MJ, Halperin DM, Raghav KP, Shen JP. Efficacy and safety of atezolizumab and bevacizumab in appendiceal adenocarcinoma. <i>Cancer Res Commun</i> 2024 May 29;<b>4</b>(5):1363-1368. doi: 10.1158/2767-9764.CRC-24-0019. PMID: 38709066; PMCID: PMC11135244. <h3>Ethics Approval</h3> This study was conducted under a protocol approved by the University of Texas M.D. Anderson Cancer Center‘s Institutional Review Board. Specific IRB identifier has been withheld due to Institutional Policy on cybersecurity.