Pathology

FFPE-compatible Hi-C reveals oncogenic rearrangements and enhancer hijacking in lymphoid malignancies

An FFPE-compatible Hi-C assay was applied to 44 lymphoid neoplasm biopsies, including large B-cell lymphomas and plasma cell neoplasms. Hi-C showed high concordance with FISH for expected oncogene rearrangements and additionally revealed enhancer hijacking involving BCL2, CCND1, and MYC. The assay detected unexpected non-coding rearrangements, including PD-1 ligand loci and other potentially targetable regions missed by standard workups. Hi-C also distinguished functionally divergent BCL6 rearrangements and clarified the topologic context of variant MYC rearrangements. These data support Hi-C as a genome-wide structural assay that can complement conventional cytogenetics in lymphoid malignancy diagnostics. 1

Shallow whole-genome sequencing (LeukoPrint) improves CNA detection and risk stratification in myeloma

LeukoPrint, a shallow whole-genome sequencing assay, profiled copy number aberrations in 423 multiple myeloma patients across three hospitals. Compared with karyotyping, LeukoPrint detected abnormalities much more frequently and identified CNAs in most karyotype-negative cases. Concordance with FISH for key prognostic CNAs, including amp(1q), del(1p), del(13q), and del(17p), was high at 94%. Integrating LeukoPrint plus FISH into the Mayo risk model reclassified 11.5% of standard-risk patients as high-risk, suggesting candidates for intensified therapy. The authors propose replacing karyotyping with LeukoPrint plus FISH for routine myeloma cytogenetic profiling. 2

Interpretable AI distinguishes prefibrotic PMF from ET on digitized marrow biopsies

An interpretable AI framework was trained on digitized H&E bone marrow biopsies from myeloproliferative neoplasm patients with thrombocytosis. The model differentiated prefibrotic primary myelofibrosis from essential thrombocythemia with an AUROC of 0.89 and 92.3% accuracy. Synthetic, AI-generated images highlighted disease-specific morphologies and showed marked diagnostic discordance with hematopathologists for some ET-like images. Quantitative analysis revealed a higher proportion of marrow fat content in ET compared with prefibrotic primary myelofibrosis. These results suggest AI leverages adiposity and other subtle morphologic cues beyond current criteria that could refine MPN diagnosis. 3

Updated classification of IBD-associated colorectal dysplasia guides surveillance and management

This review summarizes the most recent pathologic classification of inflammatory bowel disease-associated neoplastic precursor lesions. It explains the historical rationale for revising dysplasia terminology, morphology, and category boundaries in the IBD setting. Newly defined morphologic, endoscopic, and molecular subtypes are described, with early evidence linking some patterns to distinct outcomes. The review highlights persisting knowledge gaps but provides a practical framework for applying the new system to surveillance and therapeutic decisions. For pathologists, it clarifies preferred terminology and key features that should be communicated in reports to guide risk stratification. 4