Prognostic impact of mutations in the DNA methylation pathway on clinical outcomes in acute myeloid leukemia: A retrospective cohort study
Abstract
Abstract Background DNA methylation in CpG islands plays a crucial role in epigenetic regulation and is frequently altered in acute myeloid leukemia (AML). Somatic mutations in genes involved in this pathway are common in AML and may impact treatment response and patient survival. In this study, we evaluated the prognostic significance of these mutations in newly diagnosed (ND) AML patients treated at our institution. Methods We conducted a retrospective cohort study including adult ND-AML patients from 2015–2023. Baseline demographics, treatments, and molecular profiles via next-generation sequencing were collected. Associations between mutations in DNMT3A, TET2, IDH1, IDH2, WT1 and outcomes were analyzed. Endpoints included composite clinical response (CCR = CR + CRi), overall survival (OS), and event-free survival (EFS). Multivariable Cox regression (MV-CPH) was used for OS and EFS, while multivariate logistic regression (MV-LR) evaluated response. All models were adjusted for age, gender, high-risk cytogenetics, TP53 status, and comorbidities including cardiovascular disease, heart failure, diabetes, and hypertension. Results A total of 971 ND-AML patients were included. DNMT3A was the most frequently mutated gene (29%, N=186/642), followed by TET2 (24%, N=153/634), IDH2 (13%, N=79/629), IDH1 (9.6%, N=60/624), and WT1 (6.7%, N=42/629). Compared to wildtype (wt), patients with DNMT3A-mutated (mt) AML were more often female (51% vs 41%), had higher peripheral blasts (34% vs 24%), bone marrow blasts (58% vs 47%), WBC (10 vs 6 K/µL), and platelets (62 vs 49 K/µL) (all p<0.05). DNMT3A-mt cases had lower high-risk cytogenetics (23% vs 33%) and TP53 co-mutations (12% vs 21%) (p<0.05). On MV-LR, DNMT3A was not significantly associated with CCR (OR 1.32, 95% CI 0.88–2.00) or cytogenetic response (OR 1.11, 95% CI 0.70–1.77). On MV-CPH, DNMT3A was associated with improved EFS (HR 0.79, 95% CI 0.64–0.98) but not OS (HR 0.85, 95% CI 0.68–1.05). Median OS and EFS were 12 and 8.7 months; 12-month OS and EFS were 52% and 38%. TET2-mt AML patients were older (69 vs 64 yrs), had higher WBC (12 vs 6 K/µL), absolute neutrophil count (ANC) (1.6 vs 0.9 K/µL), and peripheral blasts (31% vs 24%) (all p<0.05) than TET2-wt AML. TET2-mt was not significantly associated with CCR (OR 1.32, 95% CI 0.85–2.07), cytogenetic response (OR 0.95, 95% CI 0.58–1.57), OS (HR 1.14, 95% CI 0.91–1.43), or EFS (HR 1.09, 95% CI 0.88–1.36). Median OS and EFS were 9.6 and 6.8 months, with 12-month OS and EFS of 48% and 35%. IDH1-mt patients were older (69 vs 65 yrs) and had lower WBC (3 vs 7 K/µL) and ANC (0.3 vs 1.1 K/µL) (all p<0.05). IDH1-mt AML was significantly associated with higher CCR on MV-LR (OR 1.87, 95% CI 1.00–3.57), but not with cytogenetic response (OR 1.53, 95% CI 0.76–3.18), OS (HR 0.92, 95% CI 0.66–1.29), or EFS (HR 1.05, 95% CI 0.76–1.46). Median OS and EFS were 15 and 7.4 months; 12-month OS and EFS were 56% and 34%. IDH2-mt patients had higher platelets (77 vs 50 K/µL), favorable cytogenetics (82% vs 67%), and wt TP53 (94% vs 80%) (all p<0.05). IDH2 was not associated with CCR (OR 1.26, 95% CI 0.72–2.25), cytogenetic response (OR 1.19, 95% CI 0.63–2.30), OS (HR 0.98, 95% CI 0.73–1.31), or EFS (HR 0.93, 95% CI 0.69–1.24). Median OS and EFS were 13 and 9.7 months; 12-month OS and EFS were 55% and 40%. WT1-mt patients were younger (55 vs 66 yrs), had higher peripheral blasts (42% vs 26%) and WBC (19 vs 7 K/µL), and more often had wt TP53 (97% vs 81%) and favorable cytogenetics (86% vs 68%) (all p<0.05). WT1-mt AML trended toward lower CCR (OR 0.46, 95% CI 0.20–1.06) and had significantly worse EFS (HR 1.54, 95% CI 1.01–2.35), but was not associated with cytogenetic response (OR 0.62, 95% CI 0.25–1.64) or OS (HR 1.52, 95% CI 0.96–2.41). Median OS and EFS were 17 and 11 months; 12-month OS and EFS were 61% and 40%. Conclusions In this large retrospective cohort of ND-AML patients, mutations in DNA methylation pathway genes demonstrated variable associations with treatment response and survival. DNMT3A mutations were significantly associated with improved EFS, whileWT1 mutations were significantly linked to inferior EFS.IDH1 mutations were predictive of significantly higher clinical response without impacting survival. TET2 and IDH2 mutations were not significantly associated with response or survival. These findings suggest that specific epigenetic mutations, particularly DNMT3A, WT1, and IDH1, may have both prognostic and predictive value in clinical settings.
Article Details
Authors (26)
John Hanna
Emily Zabor
2Cleveland Clinic, Quantitative Health Sciences, Cleveland, United States
Moath Albliwi
1Department of Internal Medicine, Cleveland Clinic Foundation, Cleveland, United States
Jessica El-Asmar
2Cleveland Clinic Taussig Cancer Institute, Department of Hematology and Medical Oncology, Cleveland, United States
Daniel Nurse
1Cleveland Clinic, Department of Internal Medicine, Cleveland, United States
Ameed Bawwab
1Cleveland Clinic, Department of Internal Medicine, Cleveland, United States
Hasan Abuamsha
1Cleveland Clinic, Department of Internal Medicine, Cleveland, United States
Yomna Abu-Farsakh
2Cleveland Clinic Taussig Cancer Institute, Department of Hematology and Medical Oncology, Cleveland, United States
Heya Batah
1Cleveland Clinic, Department of Internal Medicine, Cleveland, United States
Asad Rauf
1Cleveland Clinic, Department of Internal Medicine, Cleveland, United States
Ahmed Nabil
Muaz Alsabbagh Alchirazi
1Cleveland Clinic, Internal Medicine, Cleveland, United States
Lana Khatib
Mark Chen
Joy Nakitandwe
2Cleveland Clinic Taussig Cancer Institute, Department of Hematology and Medical Oncology, Cleveland, United States
Elizabeth Azzato
5Cleveland Clinic Diagnostics Institute, Department of Pathology and Laboratory Medicine, Cleveland, United States
David Bosler
5Cleveland Clinic Diagnostics Institute, Department of Pathology and Laboratory Medicine, Cleveland, United States
Akriti Jain
1Cleveland Clinic, Internal Medicine, Cleveland, United States
John Molina
1Cleveland Clinic Foundation, Department of Hematology and Medical Oncology, Cleveland, United States
Sophia Balderman
1Cleveland Clinic, Internal Medicine, Cleveland, United States
Abhay Singh
1Cleveland Clinic, Internal Medicine, Cleveland, United States
Aaron Gerds
3Cleveland Clinic Taussig Cancer Institute, Cleveland, United States
Sudipto Mukherjee
1Cleveland Clinic, Internal Medicine, Cleveland, United States
Anjali Advani
6Cleveland Clinic Taussig Cancer Institute, Cleveland, OH
Hetty Carraway
1Cleveland Clinic, Internal Medicine, Cleveland, United States
Moaath Mustafa Ali
1Cleveland Clinic, Internal Medicine, Cleveland, United States