Interobserver reliability and digital morphometric characterization of constitutional chromosomal heteromorphisms in...

Constitutional chromosomal heteromorphisms are common cytogenetic variants involving constitutive heterochromatin and acrocentric short-arm structures whose visual identification and classification remain challenging in routine cytogenetic practice. This study evaluated interobserver reliability in the visual classification of constitutional chromosomal heteromorphisms and characterized differences between homologous chromosomes using digital morphometry. A retrospective cross-sectional study included peripheral blood cytogenetic studies performed between 2021 and 2024. Sixty-six cases with previously reported heteromorphisms were included, yielding 73 heteromorphism observations. Four cytogeneticists, blinded to the original classification and clinical information, independently evaluated selected Q-banded metaphases using predefined visual criteria. Interobserver agreement was assessed using Krippendorff's alpha and Gwet's AC1. Digital morphometric analysis was performed in Fiji by measuring the chromosomal region associated with each heteromorphism, defined as the target region (TR), and calculating its absolute size and its ratio to total chromosome size (TR/T). Homologous chromosomes were compared using the Wilcoxon signed-rank test. Overall observed agreement was 70.1%, and complete agreement among the four cytogeneticists was observed in 45.5% of metaphases. Krippendorff's alpha was 0.340 (95% CI 0.276-0.398), and Gwet's AC1 was 0.453 (95% CI 0.380-0.532). Positive specific agreement was higher than negative specific agreement (77.1% vs. 56.9%). Significant morphometric asymmetry was identified in 11.0% of observations using absolute TR measurements and in 15.1% using TR/T. Findings for the acrocentric subgroup were based on only six observations and should be interpreted cautiously. Overall, visual classification showed limited reproducibility, particularly for pericentromeric heterochromatic variants. Digital morphometry provided complementary quantitative descriptors but did not resolve the limitations of classification based on chromosome morphology and should not be used as a confirmatory criterion.