Next-generation sequencing based post-transplant monitoring of acute myeloid leukemia

TaeHyung Kim, Joon Ho Moon, Jae-Sook Ahn, Yeo-Kyeoung Kim, Seung-Shin Lee, Seo-Yeon Ahn, Sung-Hoon Jung, Deok-Hwan Yang, Je-Jung Lee, Seung Hyun Choi, Ja-yeon Lee, Marc S. Tyndel, Myung-Geun Shin, Yoo Jin Lee, Sang Kyun Sohn, Seong-Kyu Park, Zhaolei Zhang, Hyeoung-Joon Kim and Dennis Dong Hwan Kim

Key points

  • Higher allelic burden at day 21 of post-HCT is associated with higher risk of relapse and mortality.

  • Longitudinal tracking of AML patients receiving HCT is feasible and provides clinically relevant information.


Next generation sequencing (NGS) has been applied to define clinically relevant somatic mutations and classify subtypes in acute myeloid leukemia (AML). Persistent allelic burden after chemotherapy is associated with higher relapse incidence, but presence of allelic burden in AML patients after receiving allogeneic hematopoietic cell transplantation (HCT) has not been examined longitudinally. As such, we aimed to assess the feasibility of NGS in monitoring AML patients receiving HCT. Using targeted gene panel, we performed NGS in 104 AML patients receiving HCT using samples collected at diagnosis, pre-HCT, and post-HCT at day 21. NGS detected 256 mutations in 90/104 patients at diagnosis, which showed step-wise clearances after chemotherapy and HCT. In a subset of patients, mutations were still detectable pre-HCT and post-HCT. Most post-HCT mutations originate from mutations initially detected at diagnosis. Post-HCTD21 allelic burdens in relapsed patients were higher than in non-relapsed patients. Post-HCTD21 mutations in relapsed patients all expanded at relapse. Assessment of variant allele frequency (VAF) revealed that overall VAF post-HCT day 21 (VAF0.2%-Post-HCTD21) is associated with an increased risk of relapse (56.2% vs. 16.0% at 3-years, p < 0.001) and worse overall survival (36.5% vs. 67.0% at 3 years, p = 0.006). Multivariate analyses confirmed VAF0.2%-Post-HCTD21 is an adverse prognostic factor for overall survival (hazard ratio 3.07, p = 0.003) and relapse incidence (hazard ratio 4.75, p < 0.001), independent of the revised European LeukemiaNet risk groups. Overall, current study demonstrates that NGS-based post-transplant monitoring in AML patients is feasible and can distinguish high-risk patients for relapse.

  • Submitted April 27, 2018.
  • Accepted August 6, 2018.