Postnatal origin of the chromosomal gains in older patients with high hyperdiploid acute lymphoblastic leukemia

High hyperdiploidy (HeH), characterized by nonrandom chromosomal gains resulting in 51-67 chromosomes, is the most common genetic subtype of childhood B-cell precursor (BCP) acute lymphoblastic leukemia (ALL). HeH ALL typically occurs in children aged 2-4 years and overall corresponds to 25% of cases in the pediatric (<18 years) population, but, for reasons unknown, becomes less frequent with increasing age; in adult ALL it is relatively rare. Several studies have shown clear evidence of HeH frequently arising already before birth in pediatric cases.^1-9 However, there are very little data on whether this also holds true for HeH arising in older children, adolescents and adults.

We previously utilized somatic single nucleotide variants (SNV) to study the “age” of the trisomies, i.e., whether their origin was recent or not in terms of clonal evolution, in 16 cases of pediatric HeH ALL.¹⁰ This was done by analyzing the mutant allele frequencies (MAF) of all somatic SNV in trisomic chromosomes to investigate whether they were present in one of two or two of three chromosomal homologues. Whereas the former could arise either before or after the chromosome became trisomic (B/ATRI SNV), the latter must have arisen before the trisomy was formed and become duplicated with the homologue (BTRI SNV) (Figure 1A). We found that BTRI SNV constituted only a small fraction of the SNV in most cases, suggesting that the chromosomal gains arose very early during clonal evolution and before the cell had acquired many passenger mutations, in line with a prenatal origin. However, the two adolescent patients in the study had much higher frequencies of BTRI SNV, indicating that older patients with HeH may differ from younger patients in this regard. Here, we have addressed this possibility further.

We ascertained whole genome sequencing (WGS) data from HeH cases from local biobanks (not included in Paulsson et al.¹⁰ and selected based on material being available except for ten cases selected based on age >16 years) (N=31), from the Therapeutically Applicable Research to Generate Effective Treatments (TARGET) Initiative (https://www.ncbi.nlm. nih.gov/projects/gap/cgi-bin/study.cgi?study_id=phs000464.v21.p8) (N=33), and from the St. Jude Cloud^11-14 (N=127). HeH was defined as 51-67 chromosomes without concurrent ETV6::RUNX1, TCF3::PBX1, BCR::ABL1, or KMT2A rearrangements. In addition, four cases with concurrent BCR::ABL1 and HeH were investigated (Online Supplementary Table S1). Analyses of BTRI data for 40 cases have previously been published.¹⁵ Informed consent was obtained according to the Declaration of Helsinki and the study was approved by the Swedish Ethical Review Authority. All somatic SNV in clonal trisomies were ascertained and categorized as BTRI (MAF=~0.67) or B/ATRI SNV (MAF=~0.33) (Figure 1A), excluding somatic SNV with MAF<0.15. Cases with large subclones for trisomic chromosomes, making the distinction between BTRI and B/ATRI unclear (N=18), with less than 65% leukemic cells based on MAF (N=15), or with <100 (N=26) or >1,000 (N=10) somatic SNV in trisomic chromosomes, were excluded. The final HeH cohort included 117 pediatric (1-17 years at diagnosis; median age 4 years) and five adult cases; 99 were aged 1-9 years and 23 ≥10 years at diagnosis (25 from local biobanks, 25 from TARGET, and 72 from St. Jude Cloud) (Online Supplementary Table S1). There were no differences in age (P=0.27; Mann-Whitney two-sided test) or sex distribution (P=0.81; Fisher’s exact test) in cases excluded (36% of initial cases) and those retained. Mann-Whitney two-sided test, Spearman correlation, and Fisher’s exact test were used as indicated below and P values <0.05 were considered significant. All statistical analyses were done using R x64 v4.4.1 (https:// www.r-project.org/).

The median numbers of somatic, BTRI and B/ATRI SNV were 277.5 (range, 101-842), eight (range, 0-79), and 265 (range, 98-825), respectively, in HeH cases. To adjust for possible differences related to read depths between the cohorts and to modal number, all subsequent analyses were done on the proportion of BTRI/all SNV in trisomic chromosomes, since this should not be affected by the total number of SNV.

The median BTRI/all SNV was 3.1% (range, 0-23%; Online Supplementary Table S1), indicating that the time period after the HeH arose was always longer than the time period before, in line with previous findings.¹⁰ No sex differences were seen (Mann-Whitney two-sided test; P=0.0531). However, whereas the BTRI/all SNV percentages appeared to be constant in patients aged 1-9 years, they increased linearly in those ≥10 years at diagnosis (Spearman correlation coefficient=0.64, P=0.00092; Figure 1B, C). The latter cases also had a significantly higher percentage of BTRI SNV (median 2.6% in patients 1-9 years vs. 9.8% in patients ≥10 years; Mann-Whitney two-sided test, P=5.26x10^-9; Figure 1D). That cases <10 years generally had the same proportion of BTRI SNV agrees well with a prenatal origin for the majority of these because their expected number of BTRI SNV would then be constant. In contrast, that cases ≥10 years show an increasing number of BTRI SNV by age suggests that the chromosomal gains generally arise postnatally at different time points in these cases, with the number of BTRI SNV being higher in older patients since the chromosomes that become duplicated will have acquired more passenger mutations with increasing age. Thus, our results show that the chromosomal gains in HeH ALL arising in older patients - outside the “age peak” in early childhood – may not arise in utero.

Support for the hypothesis that pediatric HeH ALL may have a prenatal origin has previously been obtained by detection of clonotypic IGH rearrangements in Guthrie cards, observations of monozygotic twins with concurrent HeH ALL, and detection of trisomies in saved cord blood cells from children who later developed leukemia.^1-9 Whereas the latter two of these methods have provided unequivocal evidence of the HeH itself being present before birth in some cases, studies of clonotypic IGH rearrangements without analysis of chromosome 14 copy number can only provide evidence for the presence of a (pre)leukemic clone, not HeH itself. In total, we found 39 reported HeH cases where prenatal origin was investigated and the patient age was given (Table 1);^1-9 both positive (N=29) and negative (N=10) cases had a median age of 2 years at diagnosis. Two of three HeH cases older than 10 years at diagnosis had evidence of a preleukemic clone - albeit not conclusive for HeH since chromosomal copy numbers were not addressed - at birth.⁷ This agrees with our findings because the proportion of BTRI/all SNV varied also in patients ≥10 years, with some having low levels of BTRI SNV and thereby possibly a prenatal origin of the chromosomal gains.

A second question arising from our findings is whether chromosomal gains arising postnatally in older patients are the primary event or if they occur as a secondary aberration to another leukemia-initiating lesion. We have previously shown that in BCP ALL with concurrent BCR::ABL1-fusion and HeH, the chromosomal gains are most likely secondary to the fusion gene.¹⁶ Thus, such cases can serve as a model for the characteristics of HeH as a secondary aberration. We investigated four ALL with concurrent BCR::ABL1 and HeH, finding very high BTRI/all SNV in all (median 22%; range 19-44%), including a 4-year-old case (Online Supplementary Table S1), in line with HeH being relatively new in terms of leukemic evolution. In fact, the proportion of BTRI SNV was higher than in all but one of the HeH cases without BCR::ABL1 investigated. Thus, the analysis corroborated the view that HeH as a secondary event would indeed have high BTRI/all SNV.

Figure 1.The fraction of BTRI mutations, arising before the chromosomal gains, is higher in older patients with high hyperdiploid acute lymphoblastic leukemia. (A) Example of somatic single nucleotide variants (SNV) in trisomic chromosomes from case L1, showing 2 distinct peaks corresponding to B/ATRI and BTRI SNV, respectively. (B) The fraction of BTRI/all SNV in relation to age is constant in cases 1-9 years old at diagnosis but increases with age in cases 10 years or older at diagnosis. The center of the boxplot is the median and lower/upper hinges correspond to the first/third quartiles; whiskers are 1.5-times the interquartile range and the individual data points are shown as black points. The median value is shown as red points for each group. (C) Spearman correlation shows that the fraction of BTRI/all SNV increases linearly with age in patients older than 10 years at diagnosis. (D) Boxplot of the fraction of BTRI/all SNV in patients 1-9 and ≥10 years old at diagnosis, respectively, showing a higher fraction in the older age group. The center of the boxplot is the median and lower/upper hinges correspond to the first/third quartiles; whiskers are 1.5-times the interquartile range and the individual data points are shown as black points.

Table 1.High hyperdiploid acute lymphoblastic leukemia reported in the literature investigated for prenatal origin and with age information available.

We attempted to find potential other primary events in cases with high BTRI/all SNV levels, but analysis of somatic WGS data for these cases did not reveal any unusual chromosomal events, structural rearrangements, targeted deletions, or coding mutations (Online Supplementary Table S2). Furthermore, the pattern of chromosomal gains, chromosomal modal numbers, and frequencies of the most common additional somatic events were similar (Online Supplementary Table S2). Thus, the only difference we can identify between HeH ALL with high BTRI/all SNV levels and the remaining HeH cases is that the former are, on average, older at diagnosis. Importantly though, some cases with high BTRI/all SNV were found in patients <10 at diagnosis. Thus, if there is an underlying primary event in these cases, it seems to be relatively age-independent although population-based cohorts would be needed to determine this.

In conclusion, we show that there is a marked difference in the percentage of BTRI/all SNV between young children and older children, adolescents, and adults in HeH ALL. We interpret this as most HeH ALL occurring in patients <10 years at diagnosis, constituting the majority of cases, have a prenatal origin of the chromosomal gains, in line with previous findings.^1-9 In contrast, HeH ALL occurring in older children, adolescents, and adults generally have a later, most likely postnatal, origin of the chromosomal gains. Notably, some cases in young children may also fall into this group, suggesting a hitherto unknown heterogeneity within HeH ALL. Considering that we have recently shown that the chromosomal gains occur simultaneously and very early in the leukemogenesis of HeH ALL,¹⁵ we deem it likely that this means that also the ALL arises postnatally in these cases, although we cannot definitely exclude another, as yet hidden, primary event.

Footnotes

• Received August 21, 2023
• Accepted February 8, 2024

Correspondence

K. PAULSON - kajsa.paulsson@med.lu.se

Disclosures

No conflicts of interest to disclose.

Contributions

MY, RG, ND, MZ, JZ, BJ and KP performed research. MY and KP performed data analysis. KP supervised the study and wrote the manuscript with input from all authors.

Funding

Acknowledgments

The results published here are in part based upon data generated by the Therapeutically Applicable Research to Generate Effective Treatments (TARGET) Initiative (phs000464). This study makes use of data generated by the St. Jude Children’s Research Hospital Genomes for Kids Study¹³ This study makes use of data generated by St. Jude Children’s Research Hospital.¹⁴