Aneuploidy in neoplasia: Single‐cell data on 83,862 tumors

Fredrik Mertens,Jakob Hofvander,Nils Mandahl,Felix Mitelman

DOI: https://doi.org/10.1002/ijc.35163

2024-09-04

International Journal of Cancer

Abstract:What's new? Chromosomal aneuploidy, reflecting alterations in chromosome number, is a well‐known feature of cancer. However, varying definitions of aneuploidy and inferences from sequencing‐ and array‐based approaches using bulk DNA have limited its characterization. Here, the authors analyzed chromosome number using single‐cell cytogenetic data from 83,862 tumors, representing 112 different tumor types. Aberrations in chromosome number split into two major clusters, one featuring near‐diploid neoplasms and the other neoplasms with extensive aneuploidy and whole genome doubling. Single‐cell analyses further reveal extensive heterogeneity in chromosomal aneuploidy in both benign and malignant tumors, with certain tumor types potentially utilizing shared pathogenetic mechanisms. Chromosomal aneuploidy, that is, numerical chromosome aberrations, is one of the molecular hallmarks of cancer. However, when neoplasms are studied with sequencing‐ and array‐based approaches, chromosome numbers and ploidy states are typically inferred from bulk DNA data. Furthermore, published molecular estimates of neoplasia‐associated aneuploidy often also include genomic imbalances resulting from various types of structural rearrangement, which likely result from other mechanisms than numerical chromosome aberrations. We thus analyzed chromosome numbers using single‐cell cytogenetic data from 83,862 tumors, and show that both benign and malignant tumors are highly heterogeneous with regard to deviations from the normal, diploid state. Focusing on the chromosome numbers in 112 specific tumor types, defined by both exact morphologic diagnosis and organ location and from which data from ≥50 cases were available, we found two major clusters: one predominated by near‐diploid neoplasms and one by neoplasms with extensive aneuploidy and one or more whole genome doublings. The former cluster included most benign solid tumors, myeloid neoplasms, and malignant gene fusion‐associated solid tumors, whereas the latter was predominated by malignant solid tumors and lymphomas. For 16 malignant tumor types, the distribution of chromosome numbers could be compared to TCGA ploidy level data. Cytogenetic and molecular data correlated well, but the former indicates a higher level of clonal heterogeneity. The results presented here suggest shared pathogenetic mechanisms in certain tumor types and provide a reference for molecular analyses.

oncology