Comparison of karyotype scoring guidelines for evaluating karyotype complexity in chronic lymphocytic leukemia

Matthew R. Avenarius,Ying Huang,Adam S. Kittai,Seema A. Bhat,Kerry A. Rogers,Michael R. Grever,Jennifer A. Woyach,Cecelia R. Miller

DOI: https://doi.org/10.1038/s41375-024-02177-y

2024-02-19

Leukemia

Abstract:Karyotype complexity is established by enumerating the number of chromosome abnormalities present in a karyotype. In chronic lymphocytic leukemia (CLL), karyotype complexity, typically defined as ≥3 chromosome abnormalities, is an important prognostic indicator associated with poor outcomes [1]. Additionally, stratifying patients by highly complex karyotypes (≥5 chromosome abnormalities) has shown significantly poorer outcomes compared to patients with 3–4 abnormalities [2]. Further, we have previously shown that complexity is significant as a continuous variable enumerated up to 20 abnormalities [3]. In principle, counting chromosome abnormalities is straightforward; however, in practice it can be nuanced, and the number of abnormalities counted for the same karyotype can differ across laboratories [4]. A lack of standardized rules has resulted in variability with how complexity is evaluated across CLL studies [5]. We and others have previously adapted the international working group on myelodysplastic syndrome cytogenetics (IWGMC) framework [4] for counting chromosome abnormalities for use in CLL [1, 3]. Additionally, a separate set of guidelines were published in the 2020 International System for Human Cytogenomic Nomenclature (ISCN) [6]. These two guidelines differ in several ways, such as how to count unbalanced abnormalities and whether to sum abnormalities across multiple subclones. In a review by Nguyen-Khac et al. the discussion was raised that the differences between these two guidelines can result in discordant complexity assignments for patients [5], which may confound the ability to compare results across clinical trials. We sought to investigate what impact the use of different guidelines for counting chromosome abnormalities may have in CLL correlative studies. Here we evaluated karyotype complexity using both the IWGMC and ISCN guidelines in a previously reported, well-characterized, dataset with long term follow-up of patients with CLL treated with ibrutinib [3]. 456 patient karyotypes underwent independent review by two cytogeneticists with enumeration of chromosomes abnormalities up to 20 using each set of guidelines. Discordant scores were discussed to reach consensus. Patients were categorized as not complex (0–2 abnormalities), complex (3–4), or highly complex (≥5). The weighted kappa statistic was used to evaluate agreement between the categories designated by the two guidelines. Progression free survival (PFS) and overall survival (OS) from ibrutinib start were evaluated by the Kaplan-Meier method. Univariable Cox regression model was used to associate the number of abnormalities as a continuous variable or a categorical variable with PFS and OS, and the discrimination power of the models were evaluated by Uno's concordance statistic.

oncology,hematology