Impact of Cytogenetic Response on Long-Term Survival in Acute Myeloid Leukemia: A Retrospective Cohort Study

Background: Cytogenetic evaluation after first-line induction therapy is an important prognostic tool in acute myeloid leukemia (AML). Further research is necessary to determine its impact on long-term survival. This study aims to elucidate the role of cytogenetic response in predicting treatment outcomes and survival in AML patients.

Methods: We conducted a single-center retrospective cohort study on AML patients (≥18 years) treated at Cleveland Clinic to investigate the impact of cytogenetic response on overall survival (OS), event-free survival (EFS), and clinical response rates. Baseline variables included demographics, comorbidities, AML subtypes, cytogenetics, next-generation sequencing, performance status, treatment responses, and follow-up data. Response to first-line treatment was assessed through composite complete response (CCR), including complete response (CR) and complete response with incomplete count recovery (CRi), and measurable residual disease (MRD) by flow cytometry. Complete cytogenetic response was defined as complete remission from pathological changes to a completely normal karyotype. Normal cytogenetic response was defined as no change in karyotype from a normal exam at the time of diagnosis. Partial cytogenetic response was defined as any partial remission from pathological changes to a normal karyotype (e.g., 46,XX,del(7)(q22)[20] → 46,XX,del(7)(q22)[5]/46,XX[15]). Persistent cytogenetic response was defined as no change in karyotype. Gain of cytogenetic abnormality was defined as developing any new pathologic abnormality in the karyotype not present at diagnosis. EFS was defined from treatment start to refractoriness, progression, or death, while OS was defined from treatment start to death. The Kaplan-Meier method estimated survival probabilities, with differences tested using the log-rank test. A landmark analysis at 60 days post-treatment accounted for the timing of cytogenetic response assessment.

Results: From May 2017 to September 2023, 730 AML patients were treated at the Cleveland Clinic Foundation. Among them, 379 patients had a cytogenetic exam available for analysis at the response landmark: 280 (74%) achieved normal cytogenetics/complete cytogenetic remission (NL-Cy group), 14 (3.7%) achieved partial cytogenetic remission (Partial-Cy group), and 85 (22%) had persistent/gain of cytogenetic abnormalities (Abnl-Cy group). The median age of the 3 comparison groups was 61 (IQR: 52-68), 68 (IQR: 56-74), and 67 (IQR: 58-75), respectively. Average bone marrow blasts at diagnosis were 54%, 45%, and 38% in the NL-Cy, Partial-Cy, and Abnl-Cy groups, respectively. The most common mutations were NPM1 in NL-Cy group (35%), TET2 in Partial-Cy group (42%), and DNMT3A in the Abnl-Cy group (33%).

Treatment regimens varied: In the NL-Cy group, 67% received intensive therapy, 12% Venetoclax + Hypomethylating (Ven+HMA) agents, 3.9% HMA, and 17% other treatments. In the Partial-Cy group, 64% received intensive therapy and 36% Ven+HMA. In the Abnl-Cy group, 25% received intensive therapy, 39% Ven+HMA, 11% HMA, and 26% other treatments. AML response rates were significantly different across groups: 86% (n=241) for NL-Cy, 79% (n=11) for Partial-Cy, and 34% (n=29) for Abnl-Cy group (p<0.001). Multivariable logistic regression showed lower odds ratio (OR) of response for Abnl-Cy group (OR: 0.09, 95%CI: 0.05-0.16) compared to NL-Cy group (reference) but not for Partial-Cy group (OR: 0.67, 95%CI: 0.19-3.11).

The median follow-up for OS was 32.76 months (range: 2.99 - 191.18). The 12-month OS probabilities were 68% (95%CI: 63%-74%) for NL-Cy group, 72% (95%CI 49%-100%) for Partial-Cy group, and 49% (95%CI: 39%-61%) for Abnl-Cy group (log-rank p<0.001). The median follow-up for EFS was 33.1 months (range: 1.65 - 88.2). The 12-month EFS probabilities were 55% (95%CI: 50%-62%) for NL-Cy group, 32% (95%CI: 13%-80%) for Partial-Cy group, and 38% (95%CI: 28%-51%) for Abnl-Cy group (log-rank p<0.001).

Conclusions: Cytogenetic response at first evaluation is a significant predictor of treatment response and long-term survival in AML patients. Persistent or gained cytogenetic abnormalities after induction are associated with markedly lower response rates and poorer survival outcomes. Future studies should explore the integration of cytogenetic response to refine risk stratification and optimize therapeutic strategies in AML.