Clonal dynamics in a single AML case tracked for 9 years reveals the complexity of leukemia progression.

Most types of cancers are made up of heterogeneous mixtures of genetically distinct subclones. In particular, acute myeloid leukemia (AML) has been shown to undergo substantial clonal evolution over the course of the disease. AML tends to harbor fewer mutations than solid tumors, making it challenging to infer clonal structure. Here, we present a 9-year, whole-exome sequencing study of a single case at 12 time points, from the initial diagnosis until a fourth relapse, including 6 remission samples in between. To the best of our knowledge, it covers the longest time span of any data set of its kind. We used these time series data to track the hierarchy and order of variant acquisition, and subsequently analyzed the evolution of somatic variants to infer clonal structure. From this, we postulate the development and extinction of subclones, as well as their anticorrelated expansion via varying drug responses. In particular, we show that new subclones started appearing after the first complete remission. The presence and absence of different subclones during remission and relapses implies differing drug responses among subclones. Our study shows that time series analysis contrasting remission and relapse periods provides a much more comprehensive view of clonal structure and evolution.