Tracking Down and Interrogating Cancer’s Holdouts

The ability to forecast which cancers will likely come back after treatment could help improve patient outcomes and might also point to new vulnerabilities to exploit in future therapies. For clinical oncologists the obstinate cancer cells that withstand treatment and are most often undetectable by microscope, are known as ‘‘minimal residual disease.’’ For instance, individuals with acute myeloid leukemia (AML) typically have a massive burden of cancerous cells at the start of treatment, but after high-dose induction chemotherapy this is drastically reduced and many patients go into remission. Yet, for many of these the disease will eventually come back. How can relapse be predicted during remission? In a recent study, Jongen-Lavrencic et al. (2018) use next-generation sequencing of blood or bone marrow samples from individuals treated and in remission for acute myeloid leukemia (AML) to understand the mutation patterns of minimal residual disease and show which profiles have a higher association with disease recurrence. The authors conduct targeted sequencing to assess ‘‘molecular minimal residual disease’’ in 430 AML patients. When comparing the resulting molecular profiles with frequency of relapse and relapse-free survival, they show that individuals with persisting mutations in one or more of a trio of genes DNMT3A, TET2, and ASXL1 (so-called DTA mutations) do not have a high risk of recurrence, whereas those with persisting non-DTA mutations were at heightened risk over the 4-year period of follow up. What ties these three genes together is that they are frequently impacted with mutations that accrue during normal aging and so are not only associated with malignant disease. The authors also show that combining this type of molecular profiling with the standard-of-care flow cytometry methods to detect minimal residual disease improves the ability to predict outcomes over flow cytometry alone. This supports the idea that