Impact of Smoking and TP53 Mutations in Acute Myeloid Leukemia and Myelodysplastic Syndromes

Objective: With the routine use of next generation sequencing (NGS) for diagnosis and prognosis, a growing number of pathogenic mutations have been discovered in patients with acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS). TP53 is the most frequently mutated gene in human cancer. Mutations in TP53 are found in approximately 8% of de novo AML, but occur more frequently, up to 30% in secondary-AML (s-AML) and therapy-related AML (t-AML), as well as 5-20% of patients with MDS. They are associated with older age, chemotherapy resistance, and worse overall survival. Previous studies have shown potent carcinogens in tobacco smoke can induce transversion mutations along the p53 gene in patients with lung cancers, laryngeal cancers, and head and neck cancers. The association between cigarette smoking and TP53 mutations and its prognostic implications in AML and MDS have not been characterized before.

Methods: We performed a retrospective review of AML and MDS patients with or without TP53 mutations at Thomas Jefferson University Hospital between April 2016 and December 2018. Data on patient age, gender, smoking status (current or past history of smoking), disease type, induction regimen, cytogenetics, and genetic mutations via NGS from bone marrow or peripheral blood were collected. Overall response rate (ORR) defined as complete remission and partial response and overall survival (OS) were analyzed. Kaplan-Meier method was used to estimate OS and compared using the log-rank test. Chi-square test was used where appropriate in comparison. P<0.05 was considered as statistically significant.

Results: From April 2016 to December 2018, 50 and 74 patients were identified with AML or MDS with and without a TP53 mutation, respectively. In patients with a TP53 mutation, median age at diagnosis was 70 years (range 51 to 91). 27 patients were diagnosed with AML (including 4 with s-AML, 3 with t-AML), 22 with MDS (including 2 with t-MDS), and 1 with acute leukemia with ambiguous lineage. A past or current history of smoking was found in 70% of patients with a TP53 mutation and 51% of patients without a TP53 mutation (p=0.006). Complex karyotype was observed in 54% of smokers with a TP53 mutation, and 73% of non-smokers with a TP53 mutation (p=0.21). In smokers with TP53 mutation, the most common co-occurring mutations were DNMT3A (6 patients), ASXL1 (5 patients), and TET2 (4 patients). In non-smokers with TP53 mutation, the most common co-occurring mutation was TET2 (3 patients). 26% of patients received intensive chemotherapy (IC), 38% of patients received a hypomethylating agent (HMA), 6% of patients received lenalidomide, and 22% of patients received supportive care only. 8% of patients were lost to follow-up. ORR was 40% in smokers with a TP53 mutation, compared to 20% in non-smokers (p=0.26). Response rate was 27% in patients who received IC and 75% in patients who received HMA (p=0.01). All patients with a TP53 mutation had died at the time of analysis, with a median OS 7.1 months (range 0.0 - 55.9) in smokers, and 6.3 months (range 0.2 - 30.3) in non-smokers (p=0.40).

Conclusion: A higher prevalence of smoking was found in patients with AML or MDS with a TP53 mutation. Response rate was higher in patients who received HMA compared to IC. Overall survival was poor in all patients with a TP53 mutation, but the difference was not statistically significant in smokers versus non-smokers. Limitations of the study include its retrospective nature, small number of patients, and single institution experience. More studies are needed to elucidate the effect of cigarette smoking on TP53 mutations in patients with AML and MDS, especially in the setting of evolving treatment paradigms using newer agents.