Different scenarios with varying implications for the predictability of transformation. (A) In this scenario, transformation results from the acquisition of genetic alterations that occur after diagnosis of preceding FL. Transformation cannot be predicted by a test performed at the time of diagnosis. This would be a probabilistic scenario as no feature from the initial FL determines subsequent transformation. (B) In some cases, the transformed clone can be detected when patients present with what is thought to be FL. Detection is typically challenging as it requires experimental strategies such as deep sequencing of DNA obtained from the FL tumor biopsy or of circulating DNA in plasma. This would be either a probabilistic or deterministic scenario as it is unclear whether the presence of a minor aggressive subclone results in increased risk of transformation. (C) FL tumor cells harbor characteristics that increase the likelihood of transformation. These characteristics could be distinct alterations that increase the mutation rate, or even a permissive microenvironment. A biomarker could be particularly useful in this setting. This latter scenario is deterministic. We would like to emphasize that scenarios A-C are simplified conceptualizations of the large number of possible ways that transformation can potentially evolve from underlying low-grade lymphoma and are certainly not mutually exclusive.