Abstract 2990: Single-slide FFPE proteomic profiling enables therapeutic target quantification and molecular subtyping in non-Hodgkin’s lymphoma and bone marrow tumor biopsies

Abstract Molecular profiling of formalin-fixed paraffin-embedded (FFPE) tumors in clinical research is performed with low-plex immunohistochemistry or genome-wide transcriptomics. The application of unbiased, highly multiplexed proteomics to FFPE is limited by sample input requirements, assay throughput and complex bioinformatics. Here we report a proteomic method amenable to low-input FFPE profiling which complements transcriptomics. We applied Biognosys’ mass-spectrometry platform (TrueDiscovery, DIA-MS) to samples encountered in routine histopathology laboratories (e.g. B-cell malignancies [n = 35], and breast, pancreatic and colorectal tumors [n=8]). FFPE samples were deparafinzed, de-crosslinked followed by trypsin digestion, and resulted peptides were analyzed by directDIA. Impact of tissue staining by Hematoxylin and Eosin (H&E) and acid decalcification was also assessed. We compared a deep-proteome workflow on a Thermo Scientific Orbitrap Exploris 480 to a novel high throughput workflow on a Bruker timsTOF HT. The deep-proteome workflow with single-slide FFPE tissues (median area 35mm2) generated comprehensive proteome coverage: ∼12,000 proteins detected overall and ∼7,000 detected in the majority of samples. The high throughput Bruker workflow delivered similar data with 80% less input in 20% of the time. Proteomes from H&E stained, cover-slipped slides were nearly identical to adjacent unstained tissue. Interestingly, proteomics profiling was successful in acid decalcified leukemia bone marrow biopsies whereas RNA was not recoverable. To assess the detection of key tissue biomarkers for therapeutic decision-making, we compiled a list of FDA approved antibody based therapeutic targets and found 20 of 21 were detectable in our samples (e.g. PDL1, PD1, LAG3, CD19, CD20 and CD38). Similarly, >80% of 55 immuno-oncology targets undergoing clinical development were detected. A key effector function of antibody therapeutics is complement-dependent cytotoxicity. The presence of complement components was confirmed by proteomics, however, some were undetectable by transcriptomics. Diffuse Large B-cell Lymphoma samples in this study were molecularly subtyped for a key prognostic, Cell of Origin (COO), using a CLIA validated RNA assay (Lymph2Cx). Well studied proteins (e.g. MUM1 and CD10) showed expected distribution based on patient COO, demonstrating proteomics can recapitulate established subtyping methods. We demonstrate that proteomics profiling can be conducted on single FFPE slides even with pre-processed tissue. This high-throughput workflow enables real-time analysis of Phase I trials or large tissue banks. Furthermore, proteome profiling provides an opportunity to consolidate testing for multiple therapeutic targets instead of serial IHC assays, maximizing the use of precious tissue biopsies. Citation Format: Matthew Loya, Tine Casneuf, Helene Bon, Manling Ma-Edmonds, Angelo Harris, Daniel Martin, Anantharaman Muthuswamy, Dominique Kamber, Martin Mehnert, Amaury Lachaud, Yuehan Feng, Mark Fereshteh, Omar Jabado. Single-slide FFPE proteomic profiling enables therapeutic target quantification and molecular subtyping in non-Hodgkin’s lymphoma and bone marrow tumor biopsies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 2990.