Multiple Myeloma Sequencing Reveals Subclonality and Timing of Genetic Alterations

Abstract 2897

Multiple myeloma is considered to be a homogenous disease within a given patient, and current classification systems and treatment algorithms are based on this assumption. We have asked if there is genetic heterogeneity of multiple myeloma within a patient and if this heterogeneity can be quantified.

To address this question, we have used massively parallel whole exome and whole genome sequencing of tumors and matched normal controls of 64 patients with multiple myeloma. We present an analytic strategy to distinguish potential driver mutations based on their clonality. We demonstrate that in some patients there are many somatic mutations that are only present in a subclonal fraction of the malignant plasma cells, and the subclonal fraction comprises up to 50%. These mutations are therefore less likely to confer a selective clonal advantage and are less attractive therapeutic targets because they only affect a small fraction of the myeloma cells. As an example, we found KRAS to be one of the most prevalent mutated genes in multiple myeloma, and KRAS mutations are significantly more likely clonal than subclonal, while other mutations in other genes are predominantly subclonal.

We also used this approach to investigate how copy number variations are related to somatic mutations, i.e. to define the temporal sequence of these events. This question is particularly relevant for hyperdiploidy in multiple myeloma, since this is associated with trisomies of odd numbered chromosomes. However, these trisomies do not occur with the same frequency in all odd numbered chromosomes and some hyperdiploid samples are also associated with trisomies of various even numbered chromosomes. It is unclear if these trisomies occur as a single catastrophic event, or rather in a sequential fashion. By assuming a constant rate of somatic mutations and utilizing this rate as a “timer” for chromosomal duplications we demonstrate that trisomies of odd-numbered chromosomes appear to occur early in a distinct order, whereas trisomies of even-numbered chromosomes and chromosome 1q occur late.

Our analyses allow us to determine which somatic mutations occurred before chromosomal duplication and may therefore give insight in the time course of pathogenic genetic alterations in multiple myeloma. Our work may also play an important role in prioritizing somatic mutations for therapeutic targeting in multiple myeloma.