SOHO 2018 | Uses and caveats of NGS at diagnosis for acute myeloid leukemia

At the 6^th Annual Meeting of the Society of Hematologic Oncology, Christopher J. Gibson from the Dana-Farber Cancer Institute, Boston, MA, gave a talk about the uses and caveats surrounding next-generation sequencing (NGS) at the time of acute myeloid leukemia (AML) diagnosis.¹

The speaker first discussed the advantages of using NGS at the time of diagnosis of AML. In a study published in 2015 in Blood by Lindsley et al., the genetic basis of distinct pathways of AML development was investigated by performing targeted mutational analysis of patients with rigorously defined secondary (s-AML) or therapy-related (t-AML) and unselected AML patients. It was found that the presence of a mutation in SRSF2, SF3B1, U2AF1, ZRSR2, ASXL1, EZH2, BCOR, or STAG2 was > 95% specific for the diagnosis of s-AML. The results of this study suggest that NGS at diagnosis allows for a molecular dissection of the AML ontogeny.²

In a study published in the New England Journal of Medicine (NEJM) in 2016, Papaemmanuil et al. aimed to understand how genetic diversity in AML defines the pathophysiology and informs clinical practice. Using targeted sequencing, it was observed that the driver landscape in AML is associated with several distinct molecular subgroups that reflect the discrete evolution of AML. These subgroups were associated with different outcomes thus implying that it can be used for informing disease classification and prognostic stratification.³ This study shows that NGS at diagnosis allows for prognostication in patients with AML.

Published data from a study by Welch et al., 2016 in the NEJM, which investigated the molecular determinants of clinical responses to decitabine therapy in patients with AML or myelodysplastic syndromes demonstrated that response rates were higher among patients with unfavorable-risk cytogenetics and patients with TP53 mutations.⁴ This study suggests that NGS at diagnosis can aid in the prediction of treatment responsiveness. NGS can also be used for the identification of targetable variants. In the phase III randomized RATIFY study, sequencing allowed for stratification according to the subtype of FLT3 mutation.⁵

The speaker then discussed the caveats of NGS at AML diagnosis. Christopher Gibson highlighted that technology tradeoffs will always exist in the sequencing market. He noted that there were some considerations to take into account including the cost per assay, time to usable data, and the sequencing breadth versus depth. Furthermore, distinguishing pathogenic variants and real mutations requires knowledge which indicates that there is a need for a higher quality annotation in NGS. Although many genes have been found to have a prognostic impact for patients with AML in univariate analysis, very few are predictive of treatment responses, which indicates that NGS at present is limited by knowledge.