Abstract
Background:
Loss of chromosome Y (LOY) is one of the most common acquired age-related somatic genomic alterations in men. However, it is also a frequent cytogenetic abnormality in myelodysplastic neoplasms (MDS) and is also observed in clonal cytopenia of undetermined significance (CCUS). Thus, in clinical practice it is challenging to assess the role and the prognostic and diagnostic relevance of a sole LOY in patients with (suspected) myeloid neoplasm (MN).
Aim:
Analysis of genetic and clinical characteristics of patients with (suspected) MN with sole LOY to assess the relevance of LOY in routine hematologic diagnostics.
Methods:
The cohort comprised peripheral blood or bone marrow samples of 2,390 male individuals sent to our laboratory between 2006 and 2024 with (suspected) MN (median age: 78 years [44-96]). All showed sole LOY by standard cytogenetic diagnostics in chromosome banding analysis (CBA, ≥20 metaphases investigated) and in FISH (≥100 cells investigated). In all cases cytomorphology results were available. Molecular data of ≥24 genes associated with MN was analyzed by NGS in 1,015/2,390 cases. Cases were categorized based on their final diagnosis: group 1 (gr1): no hematologic neoplasm according to cytomorphology (n=1,081), group 2 (gr2): MN possible (n=506), group 3 (gr3): MN (n=803).
Results:
The percentage of LOY (clone size) was larger in CBA compared to FISH (median: CBA: 70%, FISH: 63%; p<0.001). However, clone size in both CBA and FISH was highly correlated (Pearson's r=0.80). High proliferation activity of the LOY clone during culturing (defined as CBA clone size ≥20% larger than FISH) was observed in 20% of cases. In the following, only CBA was considered for LOY clone size definition, though patterns and statistical significance remained the same when FISH was used. LOY clone size showed a significant but weak correlation with age (Pearson's r=0.16). Comparison of the clone size between the different groups revealed a significant increase from gr1 to gr2 to gr3 (median: 56%, 65%, 85%, p<0.001), showing a strong association of LOY clone size with disease state. Median age was similar between gr1/2/3 (78 [49-96], 79 [49-94], 78 [44-95] years). Large LOY clones (defined as ≥70% in CBA, which roughly bisects the dataset) showed significantly higher numbers and higher variant allele frequencies (VAF) of mutations compared to smaller clones. 79% of cases with large LOY clones carried at least one mutation, most commonly in TET2 (42%), SF3B1 (17%) and ZRSR2 (12%). Only 51% of smaller clones carried mutations, most commonly in TET2 (15%), DNMT3A (14%) and SF3B1 (8%). The median VAF of mutations was also higher in cases with large LOY clones: 31% vs. 8%. Cases with high proliferation activity (LOY CBA > FISH) did not show a higher incidence of mutations compared to other cases. The proportion of cases carrying mutations increased with disease state, from gr1 (48%) to gr2 (60%) to gr3 (90%) as well as the number of genes mutated per case (median 0 [0-6], 1 [0-5], 2 [0-8]).
To further quantify the association of LOY clone size with MN we created a multivariate logistic regression model for cases with sole LOY. We modeled the probability of being associated with MN (i.e., being in gr2 or gr3) based on binary variables for age (≥75 years vs. <75 years), mutational status (mutated vs. wildtype) and CBA LOY clone size (≥70% vs. <70%). All three variables contributed significantly to the model (age: p=0.02, mutations and LOY: both p<0.001). Mutational status had the largest effect in the model with a hazard ratio (HR) of 3.5 for mutated cases relative to wildtype, followed by LOY clone size (HR=2.4 for large clones relative to small clones) and age (HR=1.4 for younger patients relative to older patients). Thus, not only the presence of mutations but also a large LOY clone is a predictor for MN independent of age. This clearly demonstrates that LOY is not only an age-associated phenomenon in these cases.
Conclusions: Our results demonstrate a significant correlation between LOY clone size and the presence of MN. Thus, a molecular analysis should be performed in cases with a large LOY clone even if no definitive diagnosis of MN can be derived from cytomorphology, since mutations are an even stronger indicator of MN. Since individuals with large LOY clones, especially with co-occurring mutations, have a high probability of MN, close monitoring of these cases for a timely therapeutic intervention is indicated.
