Understanding the Origins of Myeloproliferative Neoplasia One Mutation at a Time

In this study, published in Nature this year, Dr. Nicholas Williams and colleagues used advanced sequencing methods to understand the clonal evolution of myeloproliferative neoplasms (MPNs). The authors performed whole-genome sequencing of more than 1,000 individual single-cell–derived hematopoietic colonies from 12 patients with different MPN subtypes, sampled at 17 time points. This facilitated the construction of phylogenetic trees whereby driver mutations could be traced to their origin. We are thought to be born with approximately 50 somatic mutations within our blood cells, and we continue to acquire mutations in a linear fashion as we age.^{1, 2}  As time marches on, some of these acquired errors lead to problematic disease states. Remarkably, the modeling in this study suggests that causative driver mutations of MPNs can emerge very early in life, with the earliest appearance of JAK2 V617F in utero at 33 weeks gestation (a prediction that has a 95% CI of 11 weeks post-conception to 1.1 years). The earliest predicted acquisition of DNMT3A was also in utero, at between 12 and 14 weeks post-conception.

The phylogenetic tree approach, facilitated by deep sequencing of the whole genome of each single-cell–derived colony, allowed the authors to identify both initial and subsequent driver mutations as well as the hundreds of passenger mutations that occurred during the decades between initial mutational insult and overt disease. Modeling of the “continuous birth-to-death process” demonstrated the fastest growth in clades with multiple mutations (with an estimated doubling time of 8 months). The least fit clones, according to the model, were those where a DNMT3A mutation was acquired in utero. The unique samples, combined with the deep sequencing approach and the sophisticated modeling, has given novel insights into the kinetics of MPN development.

The JAK2 V617F mutation was studied in a similar fashion by another group using single-cell transcriptomics and whole-genome sequencing of hematopoietic stem and progenitor cells from individuals with MPNs.³  Dr. Debra Van Egeren and colleagues also reported childhood acquisition of JAK2 V617F at age nine ± two years in a 34-year-old patient, and at age 19 ± three years in a 63-year-old patient. These studies highlight that, in the absence of selection pressure, a decades-long latency period can exist between the acquisition of a known driver mutation and the development of clinically significant disease.