Background Acute myeloid leukemia (AML) is a biologically heterogeneous malignancy and is usually associated with poor prognosis. Cytogenetic abnormalities remain a cornerstone of AML risk stratification, with complex karyotype traditionally defined as the presence of ≥3 unrelated chromosomal abnormalities in the absence of favorable-risk lesions. While it is well established that increasing karyotypic complexity correlates with adverse outcomes, the prognostic value of the absolute number of cytogenetic abnormalities has not been thoroughly studied. In this study, we aim to characterize the relationship between increasing karyotypic complexity and survival outcomes in AML.

Methods We conducted a retrospective cohort study of adult patients with newly diagnosed AML and complex karyotype treated at the Cleveland Clinic Foundation between January 2015 and September 2023. Cytogenetic data were reviewed to determine the maximum number of abnormalities within a single clone per patient. Patients were stratified into four groups based on karyotypic complexity: 3–4, 5–6, 7–8, and ≥9 abnormalities. Outcomes included response rate (composite complete remission (CCR) = CR + CR with incomplete hematologic recovery (CRi)), overall survival (OS), and event-free survival (EFS). OS and EFS were estimated using Kaplan-Meier analysis. Multivariable Cox proportional hazards (MV-CPH) models were used to assess the association between the number of cytogenetic abnormalities (as a continuous variable) and OS/EFS, adjusting for key covariates including age at diagnosis, TP53 status, presence of high-risk cytogenetic lesions (e.g., −5/ del(5q), −7/ del(7q), del(17p)), and comorbidities (coronary artery disease (CAD), heart failure, diabetes, hypertension).

Results We analyzed 200 patients with newly diagnosed AML and complex karyotype out of 971 patients treated during the study period, categorized by the number of cytogenetic abnormalities: 3–4 (n=39), 5–6 (n=34), 7–8 (n=37), and ≥9 (n=90). The median age ranged from 63 to 69 years (P=0.3). The −5/ del(5q) frequency increased with increasing complexity in the above groups (21%, 38%, 51%, 68%, P<0.01). Similarly, TP53 mutation frequency increased with increasing complexity (18%, 47%, 57%, 67%, P<0.01). The CCR rates in the above categories were 37%, 38%, 21% and 29% (P=0.4).

Median OS declined with increasing karyotypic complexity: 7.7 months (mo) (95%CI: 5.3-14) for 3–4 abnormalities, 7.8 mo (95%CI: 6-18) for 5–6, 5.0 mo (95%CI: 3-9.1) for 7–8, and 4.3 mo (95%CI: 1.7-6) for ≥9 (Log-rank <0.01). A similar pattern was observed for EFS, with medians of 5.3 mo (95%CI: 4.1-7.7) for 3–4, 5.3 mo (95%CI: 1.8-14) 5–6 abnormalities, 3.5 mo (95%CI: 1.8-7.2) for 7–8, and 2.5 mo (95%CI: 1.5-4.3) for ≥9 abnormalities (Log-rank= 0.03).

In MV-CPH analysis, treating the number of cytogenetic abnormalities as a continuous variable, each additional increase in one cytogenetic abnormality was associated with a 4% worsening in OS, but this was not statistically significant (HR per additional abnormality: 1.04; 95%CI: 1–1.09, P= 0.09). In this model, none of the covariates retained prognostic significance (TP53 status, other high-risk cytogenetic abnormalities…etc.) except for age (as a continuous variable: HR 1.03, 95%CI: 1.01-1.05, P<0.01) and presence of CAD (HR: 1.8, 95%CI: 1.2-2.8, P<0.01).

Similarly, each additional increase in one cytogenetic abnormality was associated with a 3% worsening in EFS, but this was not statistically significant (HR per abnormality: 1.03; 95%CI: 0.99–1.09, P= 0.11). Again, only increasing age and CAD at baseline predicted worse EFS (P<0.05).

Conclusion Our study found that increasing karyotypic complexity in AML was associated with poorer survival outcomes on unadjusted analysis. While MV-CPH analysis demonstrated a similar trend, the association did not reach statistical significance, likely due to limited sample size. These findings suggest that risk stratification based solely on the conventional definition of complex karyotype (≥3 abnormalities) may be insufficient. Future studies should evaluate the number of cytogenetic abnormalities as a continuous variable in all AML patients with abnormal karyotypes to determine whether this approach could improve existing risk classification systems.

This content is only available as a PDF.
2025
Sign in via your Institution