Page 43 - TD-4-3

Tumor Discovery HRD genomic alterations in Chinese NSCLC

alterations that occur at significantly higher frequencies signify patients with high genomic instability and therefore
in East Asian NSCLC patients compared to Western possible responders to therapeutics.
populations. The selected genes include some known
oncogenes and DDR-associated genes that are involved 2.6. Biomarker and clinical feature integration
in tumor development and therapy response. Sequencing Expression of PD-L1 was assessed using
libraries were prepared with 0.5 µg of good-quality DNA immunohistochemistry, and mutation status for key
per sample. DNA was sheared to 180 – 280 bp using a oncogenes was determined from sequencing data. This
Covaris M220 ultrasonicator (Massachusetts, USA). comprehensive biomarker assessment allows for an analysis
End-repair was then performed, followed by A-tailing of HRD’s clinical significance in the Chinese context.
and adapter ligation, and the samples were finally
subjected to polymerase chain reaction enrichment 2.7. Statistical analysis
using index primes. DNA samples were purified using In this study, we performed the statistical analysis and
AMPure XP beads (Beckman Coulter, USA) and all the visualization using the R package map tools. Landscape
libraries were then quantified using a high-sensitivity analyses, statistical tests, and other pertinent studies were
DNA assay kit from Agilent. DNA sequencing was done made easier by this package. For comparisons between
1
in the Illumina NovaSeq 6,000 where the tool used was 2 two categorical and continuous variables, Fisher’s exact
× 150 bp paired-end. With a significance level at p<0.05 test and the Mann–Whitney U tests were used. In addition,
and power of 0.90, a minimum average coverage depth the software used for differential analysis of the cancer
was set for targeted genes and SNPs loci at ×1,000 and genome atlas (TCGA) transcriptomic data was limma, and
×200, respectively, to ensure an adequate HRD scoring the software used for differential analysis of HRD-RNAseq
and mutation detection. data was DESeq2. The criteria for selecting differentially
expressed genes (DEGs) were |log2FC| > 1 and adjusted
2.4. Bioinformatics pipeline and variant calling p<0.05, and all other tests were also performed at the
The raw sequencing data were then filtered using FASTP statistically significance level <0.05.
version 0 (HaploX Biotechnology, China) with adapters
trimming and removing low-quality reads. The clean reads 3. Results
were then mapped to the target human genomic reference 3.1. Mutational landscape of HRR genes in Chinese
sequence, specifically the hg19 or the NCBI Build 37. NSCLC patients
Sambamba was used in the process of sorting the BAM
files and Samblaster in identifying the duplicate reads. To study the association between HRD score and NSCLC
Somatic and germline single nucleotide variants, as well as clinicopathological and genetic features, we employed
small insertion-deletions were called from the tumor and an NGS test covering 188 cancer-related genes and more
normal samples using VarScan2 (Washington University, than 37,000 SNPs distributed across the human genome.
USA). All variants were filtered with an in-house pipeline Among 158 NSCLC patients, 8.9% (14/158) harbored
and were then validated using Integrative Genomics somatic genomic alterations in HRR genes. Within these
Viewer (Broad Institute, USA) to ensure that these were 14 patients, we identified 17 mutations in HRR genes,
accurate. with the majority (88.2%) being missense mutations. In
addition, 24.7% (39/158) of patients exhibited germline
2.5. Calculation of HRD score and classification genomic alterations in the HRR genes. Among these
39 patients, we identified 49 mutations in the HRR genes,
For the evaluation of HRD, we utilized the scarHRD R with 18.4% being classified as pathogenic mutations
package (https://github.com/sztup/scarHRD) to calculate (Figure 1A). In our cohort of Chinese NSCLC patients,
HRD scores based on three genomic instability metrics: ATM emerged as the most frequently somatically mutated
loss of heterozygosity, telomeric allelic imbalance, and HRR gene, occurring in 4.4% of cases. The most common
large-scale state transitions. The use of 37,000 genome-wide germline mutated gene among the HRR genes was BRCA2,
SNP markers provided high-resolution detection of these present in 7.0% of cases, followed by ATM at 3.2%, BRIP1
events, tailored to the genetic background of Chinese at 3.2%, and BARD1 at 2.5% (Figure 1A). Our analysis
patients. The patients were further divided into two revealed that EGFR alterations were mutually exclusive
subgroups according to their scores of the HRD as follows: with many other genetic alterations, suggesting that EGFR
• HRD-High (HRD-H): Score ≥43 mutations represent the most potent driver events in
• HRD-Low (HRD-L): Score <43. NSCLC. Conversely, we observed that changes in multiple
This threshold has been taken from other previously 1 Accessible at http://bioconductor.org/packages/release/bioc/
validated works and is a biologically significant cut-off to vignettes/maftools/inst/doc/maftools.html
24

Volume 4 Issue 3 (2025) 35 doi: 10.36922/TD025180032