Myeloproliferative neoplasms are clonal hematopoietic neoplasms that typically present in the later decades of life. This year, seminal studies shed new light onto the tempo of their evolution, applying advanced genomics to unravel the clonal trajectories and disease latency, and revealing that the founding mutations were likely acquired much earlier in life than was previously suspected.
Prior to the availability of techniques such as those applied in recent studies, our understanding of the temporal relationship between the clinical manifestation of myeloid neoplasms with the initiating genetic lesion was based on research in individuals who developed secondary cancers after a specific genotoxic insult. Such observations, like those in survivors of the Hiroshima and Nagasaki atomic bombings, suggested that the secondary cancers typically manifest after two to eight years, after which the risk of malignancy decreases.1 The recent data in which the lineage histories of individual hematopoietic stem cells were reconstructed using somatic mutation patterns challenge this model, indicating that the latency between the acquisition of a driver mutation and the progression to overt myeloid malignancy can, in many cases, be numerous decades.2-4
In the current studies, the timing of the original genetic lesion was estimated using the acquisition of somatic mutations across the genome as a “genetic clock,” as these are presumed to occur at a constant rate over human ontogeny. The first of the series of papers was published at the end of 2021 by a group from Harvard2 who performed whole-genome sequencing (WGS) of single cell–derived colonies grown from hematopoietic stem and progenitor cells (HSPCs) isolated from two patients with essential thrombocythemia and reconstructed the phylogenies of the cells. This estimated that the initiating JAK2 V617F mutation was acquired 20 to 40 decades prior to clinical presentation — substantially earlier than previously thought.
In a comprehensive article published in Nature by the Wellcome Sanger Institute group in February 2022, Dr. Nicholas Williams and colleagues3 performed WGS of single cell–derived colonies in a larger cohort of 12 patients with MPNs between 20 and 81 years of age. In total, they sequenced 1,013 colonies, identifying 560,978 single nucleotide variants (SNVs) and 19,155 small insertions and deletions (“indels”). Tracing the detected SNVs across colonies for individual patients, they were able to decipher the phylogenetic trees and arrange the colonies in “clades.” Remarkably, for all cases studied, the acquisition of the driver mutation occurred early in the patients’ lives — in childhood, or in some cases, even in utero. In the five patients where JAK2 V617F was the first driver event within the MPN clone, the mean latency between JAK2 V617F acquisition and development of disease was more than three decades (range, 19-54 years). In all cases, within the premalignant clades, the earliest driver mutations were acquired within the first 200 mutations after the start of life. Equally important, the colonies harboring no driver mutations shared only a few variants with each other, indicating that the remaining wild-type HSPCs remain highly polyclonal with a mutation rate similar to that of heathy individuals.
In a more focused study published later in the year, Dr. Nikolaos Sousos and colleagues4 described the case of fraternal twin brothers who both developed primary myelofibrosis that was driven by a 52-base-pair deletion in calreticulin (CALR), confirming that MPNs may be initiated before birth, but not manifest until adulthood. They used WGS and tracked the onset, showing that the CALR mutation was somatic in origin and not germline (i.e., not present in T cells) and demonstrating that the premalignant clone likely emerged in one twin and engrafted the other via intraplacental, twin-to-twin transfer. Nine additional SNVs affecting noncoding regions of the genome were also shared by both twins. Genetic analysis of single-cell HSPC colonies showed that the shared somatic mutations detected by WGS were present in colony-derived material from both twins, while unique somatic mutations for each brother were only detected in the respective twin. Notably, single-HSPC index sorting and CALR genotyping revealed the presence of CALR-mutant HSCs, supporting HSCs as the propagating population for the MPN clone. They supported the finding in these twins by also detecting JAK2 V617F in a neonatal blood spot of a patient who presented with an MPN in later adult life.
Collectively, these studies shift the paradigm in cancer biology, confirming that cancers presenting in late adulthood are often initiated before birth. Importantly, the prolonged latency means that there is a long pre-disease phase, and consequently, the window for preventative intervention therapies may be larger than previously thought. Future studies of the evolutionary dynamics during this phase will improve early detection of MPNs. Moreover, early detection, in combination with robust polygenic risk scores5 and potentially disease-modifying treatment options,6 may result in early intervention strategies aiming to modify clonal evolution and the corresponding clinical course of the disease, improving overall outcomes for patients. Given the growing appreciation that mutations frequently occur and are under selection in histologically normal tissues,7-11 the model suggested by these findings may be broadly applicable to the evolution of solid tumors.12,13
Dr. Sousos is the first author of one of the studies mentioned in this article. Dr. Psaila indicated no relevant conflicts of interest.