Germline variants in patients developing second malignant neoplasms after therapy for pediatric acute lymphoblastic leukemia—a case-control study

Highest


Candidate gene sequencing
We performed whole exome sequencing (WES) on a NovaSeq sequencing platform (Illumina GmbH, Berlin, Germany) using an IDT xGen Exome Research Panel V2 (Integrated DNA Technologies, Inc., Leuven, Belgium); annotation and final filtering on the 159 candidate genes was carried out with VarSeq 2.3 (Golden Helix, Inc., Bozeman, USA).In sequencing analyses of the exons (±70 base pairs), minimum depth was 100x.Copy number variations were not analyzed in this WES-based approach.The classification of pathogenicity according to the ACMG/AMP criteria and additional refinements of the variants was performed independently by two investigators and then extensively discussed in the context of the available literature and database information (gnomAD, ClinVar, ClinGen, LOVD3, INSiGHT, BRCA-XChange, etc.).Whenever available curated classifications based on the reviews of ClinGen expert panels were included.The classifications presented here are the result of this evaluation process.Overall, a total of 643 variants were determined and classified: 248(38.57%) in the finally included 75 cases and 395(61.43%) in the 148 control patients.According to Pearson's Chi-squared testing, the observed variant frequencies by pathogenicity tiers were not significantly different from the expected; further details can be obtained from Suppl. Figure 1.

Statistical analyses
Odds ratio (OR) was calculated by performing conditional Cox-Regression analysis with SPSS ("COXREG" procedure), an equivalent to conditional logistic regression analysis 1 .This procedure accounts for the matching aspect (case to control ratio 1:2) of our study design; we successfully validated our result with SAS ("SAS proc phreg" procedure).e Preventive cranial irradiation (CI) at 12 Gy was only applied in T-cell ALL and high-risk patients; CNS-positive patients received 18 Gy (<2 years, 12 Gy; <1 year, no CI).f Risk stratification based on MRD analysis for ERG: Standard risk, MRD-negative on treatment day33 and 78; high risk, leukemic cell load 5x10 -4 on treatment day 78; all other results correspond to intermediate risk.g Treatment group according to risk stratification including all relevant diagnostic parameters.h A (likely) pathogenic germline variant in one of the candidate genes was considered as a conspicuous finding; one of the SMN patients was determined with findings in two distinct genes, and one pathogenic KRAS (rs121913530) variant, determined in another SMN patient was identified to be of somatic origin and therefore not counted (compare Table 1 and Suppl.Abbreviations: Acute lymphoblastic leukemia (ALL), acute myeloid leukemia (AML), myelodysplastic syndrome (MDS), Non-B-cell non-Hodgkin lymphoma (NB-NHL), non-Hodgkin lymphoma (NHL), second malignant neoplasm (SMN), single nucleotide variants (SNV).a Approved gene symbols, according to the human gene nomenclature committee (HGNC); all variants determined in this project were assessed according to standard variant interpretation guidelines as indicated above for each gene; variants in the mismatch repair genes were classified according to the most recent version of the "InSiGHT Variant Interpretation Committee MMR gene variant classification criteria" published in 2018 (see https://www.insight-group.org/criteria/).b Included candidate genes were either considered to be relevant by us in previous investigations ( §) or recommended by Byrjalsen et al 2021 14 (ⱡ).c Approved gene identifier (Entrez ID/GeneID; National Center for Biotechnology Information (NCBI)'s database for gene-specific information).d Known inheritance modes: autosomal dominant (AD), autosomal recessive (AR) and Xlinked recessive (XLR); patients with a heterozygous deleterious mutation in a gene related to conditions/CPS with a known AR inheritance mode were considered to have a carrier status.3. Detailed clinical information on patients in our study population with (likely) pathogenic variants in the candidate genes.f Highest MAF observed in a subpopulation of gnomAD (v.2.1.1,exomes): non-Finnish European (NFE), African/African American (AFR), South Asian (SAS).g This rare CHEK2 variant, rs28909982 (c.349A>G p.(Arg117Gly), present in patients from both groups was previously identified to moderately increase the risk of hereditary breast cancer but not of ovarian cancer 15,16 .h This TP53 variant, rs121912651 (c.742C>T p.(Arg248Trp) is a frequent germline variant related to the Li-Fraumeni syndrome 17 .It was determined here with a VAF of 50% (125/248) and not considered to be caused by clonal hematopoiesis; DNA samples from additional tissues for subsequent testing were not available to us, due to the patients' death.i This KRAS variant, rs121913530 (c.34G>C p.(Gly12Arg, VAF = 45%), was considered to be of somatic origin and therefore excluded from subsequent frequency calculations.As germline variants at this residue are rare and the patient showed no Noonan (like) syndrome related conditions, additional DNA from hair follicles was analyzed to test for clonal hematopoiesis.Nevertheless, KRAS codon 12 is known to be frequently mutated in human cancers and associated with a poor prognosis in certain entities 18 .

a
Abbreviations: Acute lymphoblastic leukemia (ALL), second malignant neoplasm (SMN), single nucleotide variants (SNV), central nervous system (CNS), identifier (ID), minor allele frequency (MAF), variant allele fraction (VAF).a Treatment group according to risk stratification including all relevant diagnostic parameters; groups were high (HR), intermediate (IR) and standard risk (SR).b White blood cell count at diagnosis of ALL.c Defined by cytogenetics (> 50 chromosomes) or by flow cytometric analyses of the ratio of DNA content of leukemic G0/G1 cells to normal diploid lymphocytes (≥ 1.16).d Preventive cranial irradiation (CI) at 12 Gy was only applied in T-cell ALL and high-risk patients; CNS-positive patients received 18 Gy (<2 years, 12 Gy; <1 year, no CI).e Time from diagnosis of ALL to diagnosis of SMN [years].f Highest MAF observed in a subpopulation of gnomAD (v.2.1.1,exomes): non-Finnish European (NFE), African/African American (AFR), South Asian (SAS).g This rare CHEK2 variant, rs28909982 (c.349A>G p.(Arg117Gly), present in patients from both groups was previously identified to moderately increase the risk of hereditary breast cancer but not of ovarian cancer15,16 .h This TP53 variant, rs121912651 (c.742C>T p.(Arg248Trp) is a frequent germline variant related to the Li-Fraumeni syndrome17 .It was determined here with a VAF of 50% (125/248) and not considered to be caused by clonal hematopoiesis; DNA samples from additional tissues for subsequent testing were not available to us, due to the patients' death.i This KRAS variant, rs121913530 (c.34G>C p.(Gly12Arg, VAF = 45%), was considered to be of somatic origin and therefore excluded from subsequent frequency calculations.As germline variants at this residue are rare and the patient showed no Noonan (like) syndrome related conditions, additional DNA from hair follicles was analyzed to test for clonal hematopoiesis.Nevertheless, KRAS codon 12 is known to be frequently mutated in human cancers and associated with a poor prognosis in certain entities18 .

Table 1 .
Characteristics of all 223 ALL patients included in this candidate gene approach, according to their SMN status.Defined by cytogenetics (> 50 chromosomes) or by flow cytometric analyses of the ratio of DNA content of leukemic G0/G1 cells to normal diploid lymphocytes (≥ 1.16).
a P-values resulting from X 2 or Fisher's exact test: Patients of our study population with ALL and subsequent SMN versus ALL patients without SMN.b The immunophenotype of one patient was not available to us.c

Table 3 ). Supplementary Table 2. Selected candidate genes. Gene a Panel b Gene ID c Associated phenotype Inheritance d
§, ⱡ Familial diffuse gastric cancer with or without cleft lip and/or palate AD CDKN2A 2 §, ⱡ Familial atypical multiple mole melanoma-pancreatic carcinoma syndrome (FAMMPC) AD