The observation that monozygotic twins share identical but non constitutive and clone-specific fusion gene sequences (e.g., ETV6 RUNX1) in pediatric ALL provided the first unambiguous evidence (in 2003) that genetic lesions, generated by chromosomal translocation, arise in utero. These studies initiated by Mel Greaves and his colleagues contributed a wealth of knowledge about founder mutations, subclonal development, clonal origin, and the evolution of dis ease. The initiating lesion and premalignant clone is shared by the twins as a consequence of intraplacental vascular anastomoses and blood cell chimerism. The twin data were endorsed by backtracking of prenatal-initiating genetic lesions in the archived blood spots, or Guthrie cards, of patients with ALL. These data were interpreted to suggest that ETV6-RUNX1 was likely a critical initiating lesion for ETV6-RUNX1-positive ALL. However, such fusions were detectable in cord blood from newborn infants at rates approximately 100-fold higher than the incidence of ALL, suggesting an obligatory requirement for additional mutations for leukemia development to occur. Over a period of 10 years these results were confirmed and the in utero origin of MLL, BCR-ABL1, and RUNX1-RUNXT1 fusion rearrangements further documented, providing direct evidence for a prenatal origin of many childhood leukemias.
Results from genome sequencing of ETV6-RUNX1 fusion region suggests it arises as a consequence of nonhomologous end-joining in the pro-B-cell stage with possible self-renewal capacity to down stream B-cell precursors. Additional evidence was provided by screening for trisomies in stored cord blood and the data indicated that ∼6% of enriched CD34+/CD19+ B progenitor cells carry trisomies frequently seen in hyperdiploid childhood ALL. Using novel SNP as well as other technologies, it has become apparent that ALL has multiple genome copy number variations (CNVs), mostly deletions, and these CNVs are distinctive between a pair of twins, indicating a secondary, postnatal origin.
When genotyping sequencing combined with other molecular technologies of five pairs of monozygotic twins with concordant ETV6-RUNX1-positive ALL, Greaves and his colleagues demon strated that all recurrent CNVs (32 in total) were different within twin pairs providing strong evidence that they are probably secondary events and postnatal in origin in both twins as well as in non-twins with ALL. Another two twin pairs who shared a monochorionic placenta and had Ph-positive ALL, were also studied by Greaves and his colleagues. These studies provided confirmation of the previous observation that BCR-ABL1 is not sufficient to cause Ph-positive leukemia. Twin A presented with ALL at age 3.8 years and twin B presented at age 4.1 years. Both had an identical BCR-ABL1 fusion transcript and both received an allogeneic stem cell transplantation from the same matched sibling donor but 7 months later twin B died, whereas twin A remained in good health 8 years following the trans plantation. SNP analysis revealed that twin A had a subclone with trisomies for chromosomes 4, 6, 9, 14, 17, and X, tetrasomy 21, and a gain of 22q11.1–q11.23 region, whereas twin B had deletions of EBF1 and IKZ1. These results provide direct evidence that rearrangements additional to BCR-ABL1 are postnatal in origin, they represent subclonal evolution, and BCR-ABL1 by itself is not sufficient for development of Ph-positive ALL. Recently two groups have reported, using mathematical models, that some patients’ MPNs acquire the JAK2 mutation in utero.
These mutation-driven natural history studies provide evidence for sequential multistep pathogenesis of hematologic malignancies with sequential accumulation of genetic changes, which may be linear through clonal succession but that clonal evolution in most leukemias is complex and may represent a branching structure, as predicted by Darwin.
Clonal Hematopoiesis of Indeterminate Potential (CHIP) and Age-Related Clonal Hematopoiesis
High-throughput sequencing has provided novel observations which have added to our current understanding of the initial mutations that occur in premalignant stem cells and their clonal development that eventually leads to myeloid malignancies.
Clonal mosaicism for large chromosomal anomalies (duplications, deletions, and copy number neutral (CN-LOH)), using SNP microarray data from 50,000 subjects in the GENEVA study, indicated that CH is infrequent (< 0.5%) from birth until 50 years of age after which its frequency rapidly increases to 2% to 3% of the elderly. It has been estimated that individuals with clonal hematopoiesis (CH) have 10-fold higher risk of developing a subsequent hematologic malignancy. These age-related mutations were confirmed by analysis of mutation acquisition in hematopoietic stem cells, over time, through whole-exome sequencing of single hematopoietic stem cell-derived colonies, which showed that the total number of mutations in healthy individual stem cells increases with age. Therefore, the term CH describes the acquisition of somatic mutations in hematopoietic stem and progenitor cells resulting in clonal expansion in individuals with no overt hematologic disease. Moreover, sequencing of multiple elderly women provided evidence of CH based on X inactivation pattern and recurrent somatic mutation in TET2 gene. Subsequent analysis of 182 additional elderly women with CH determined that more than 5% of these individuals had mutations in TET2.
Investigation of 82 paired bone marrow and blood samples from carefully selected healthy adult volunteers using a panel of 41 genes known to be mutated in myeloid malignancies, with 1% threshold of detection revealed that bone marrow clones were found in almost 40% of healthy volunteers over 50 years of age. The most frequent mutations included DNMT3A and TET2. Variant allele frequencies were highly concordant between blood and bone marrow samples.
However, individuals with CH may live for many years and decades without developing hematologic malignancies though they are at increased risk as compared with those without mutations. These studies may explain the higher frequency of myeloid malignancies in the elderly population.
Loss of Y chromosome
Loss of Y chromosome (LOY) was first described in 1963 in cultured blood leukocytes as a sole cytogenetic abnormality in bone marrow cells which represents one of the most common acquired, post zygotic somatic genomic alterations in elderly men. Approximately 6% of male bone marrow karyotypes from normal individuals and 16% of bone marrow karyotypes with hematologic malignancies show a LOY. The incidence of LOY in bone marrow cells increases proportion ally with aging with up to 20% of healthy men over 80 years of age showing LOY using standard G-banding techniques. Mosaic LOY in blood has been reported to be associated with higher risk of all-cause mortality and overall lower life expectancy. LOY has been described most frequently in patients with myelodysplastic disorders and has been reported in MPN, AML, and other myeloid diseases as well as in various solid tumors. Since MDS and MPN are primarily diseases of elderly individuals it is often difficult to separate age-associated LOY from disease-associated LOY. Although the incidence of LOY in MDS is up to 30%, the sole LOY is lower, ranging from 4% to 10%; the WHO does not recognize LOY by itself to be useful in cytogenetically identifying a case of MDS unlike other abnormalities such as del(5q), del(7q), +8, and del(20q).
Recent observations that a higher LOY burden is associated with a higher likelihood for progression to a myeloid malignancy represents the basis for a proposal to include LOY as another form of CH with striking similarities with CHIP. Among 73 studied male patients with sole LOY and a median age of 75 (range 29 to 90), those who had equal or greater than 75% of marrow metaphase cells with LOY had a significantly greater chance of developing a myeloid neoplasm and a higher lifetime incidence of diagnosis of MDS (P < .0001). Male individuals with greater than 75% of metaphase cells with LOY also had greater likelihood of having somatic mutations (P = .0009) and greater numbers of these mutations (P = .0002).6 The presence of metaphase cells with equal or less than 25% LOY was associated with normal bone marrow morphology and a lower likelihood of progression to MDS. Multivariate analysis showed that these observations are consistent with samples showing 75% or more metaphase cells with LOY are myeloid disease-associated and not related due to aging. Therefore, it has been proposed that LOY represent a marker of CH and as such should be included in the definitions of CHIP in clinical and research settings.
Loss of X chromosome
The relationship between isolated loss of X chromosome (LOX) and aging was first documented in 1995 using interphase FISH method applying to peripheral blood cells of 90 healthy females (age: 1 week to 93 years) with a frequency of 3.2% below 51 years to 5.1% at age 51 to 93 years. This observation was confirmed a few years later in a cohort of 655 females and the frequency of LOX is ranging from 0.07% in patients older than 16 years to 7.3% in individuals older than 65 years. LOX was occasionally reported in BM cells at the frequency of 0.1% of females with the median age of 71.
Acquired trisomy 15 as an isolated abnormality is rarely detected. Approximately 0.1% to 0.2% of elderly patients show +15, frequently in association with LOY. In bone marrow cells, when I-FISH is used, +15 may represent a small clone of up to 24% of nondividing cells but none of these patients developed a myeloid neoplasm after a median follow-up of 28 months (range 0 to 149 months). These observations suggest that +15, like LOY, when present as a small clone is not disease-related but represents an abnormality associated with CH and is most likely age-related.
Chronic Myeloproliferative Neoplasms The World Health Organization (WHO) characterizes MPNs as clonal stem cell disorders. CML has a unique place among hematologic malignancies and is described separately from the other Ph-negative MPNs.
