Targeted Next-generation Sequencing Data Research Articles

Microsatellite Instability and Mismatch Repair Deficiency Define a Distinct Subset of Lung Cancers Characterized by Smoking Exposure, High Tumor Mutational Burden, and Recurrent Somatic MLH1 Inactivation

IntroductionMicrosatellite instability (MSI) and mismatch repair (MMR) deficiency represent a distinct oncogenic process and predict response to immune checkpoint inhibitors (ICIs). The clinicopathologic features of MSI-high (MSI-H) and MMR deficiency (MMR-D) in lung cancers remain poorly characterized. MethodsMSI status from 5171 patients with NSCLC and 315 patients with SCLC was analyzed from targeted next-generation sequencing data using two validated bioinformatic pipelines. ResultsMSI-H and MMR-D were identified in 21 patients with NSCLC (0.41%) and six patients with SCLC (1.9%). Notably, all patients with NSCLC had a positive smoking history, including 11 adenocarcinomas. Compared with microsatellite stable cases, MSI-H was associated with exceptionally high tumor mutational burden (37.4 versus 8.5 muts/Mb, p < 0.0001), MMR mutational signatures (43% versus 0%, p < 0.0001), and somatic biallelic alterations in MLH1 (52% versus 0%, p < 0.0001). Loss of MLH1 and PMS2 expression by immunohistochemistry was found in MLH1 altered and wild-type cases. Similarly, the majority of patients with MSI-H SCLC had evidence of MLH1 inactivation, including two with MLH1 promoter hypermethylation. A single patient with NSCLC with a somatic MSH2 mutation had Lynch syndrome as confirmed by the presence of a germline MSH2 mutation. Among patients with advanced MSI-H lung cancers treated with ICIs, durable clinical benefit was observed in three of eight patients with NSCLC and two of two patients with SCLC. In NSCLC, STK11, KEAP1, and JAK1 were mutated in nonresponders but wild type in responders. ConclusionsWe present a comprehensive clinicogenomic landscape of MSI-H lung cancers and reveal that MSI-H defines a rare subset of lung cancers associated with smoking, high tumor mutational burden, and MLH1 inactivation. Although durable clinical benefit to ICI was observed in some patients, the broad range of responses suggests that clinical activity may be modulated by co-mutational landscapes.

Reproducibility of next-generation-sequencing-based analysis of a CRISPR/Cas9 genome edited oil seed rape

Next-generation-sequencing (NGS) becomes increasingly important for laboratories tasked with the detection of genetically modified organisms (GMOs) in food, feed and seeds. Its implementation into standardized workflows demands reliable intra- and inter-laboratory reproducibility. Here, we analyze the reproducibility of short- and long-read targeted NGS and long-read whole genome sequencing (WGS) data between three independent laboratories. Replicate samples were submitted for sequencing and comparatively analyzed. The targeted-NGS-samples consisted of oil seed rape (OSR) sampled from a commodity shipment spiked with a genome edited (GE) OSR and the WGS-samples consisted of leaf material from the GMOs’ parental line. All laboratories delivered highly reproducible high-quality targeted NGS data with little variation. The detection of GMO-related sequences works well regardless of the facility, while the mapping to the complex genome is superior using long read data. Long read WGS is currently not suitable for routine use in enforcement laboratories, due to a large inter-laboratory variation.

DREAMS: deep read-level error model for sequencing data applied to low-frequency variant calling and circulating tumor DNA detection

Circulating tumor DNA detection using next-generation sequencing (NGS) data of plasma DNA is promising for cancer identification and characterization. However, the tumor signal in the blood is often low and difficult to distinguish from errors. We present DREAMS (Deep Read-level Modelling of Sequencing-errors) for estimating error rates of individual read positions. Using DREAMS, we develop statistical methods for variant calling (DREAMS-vc) and cancer detection (DREAMS-cc). For evaluation, we generate deep targeted NGS data of matching tumor and plasma DNA from 85 colorectal cancer patients. The DREAMS approach performs better than state-of-the-art methods for variant calling and cancer detection.

CNV Deep Learning based Methodology for Recognition

Background: Copy number variants (CNVs) account for a significant amount of genetic variation. Since many CNVs include genes that result in differential levels of gene expression, substantial normal phenotypic variation can be explained. Current efforts are directed toward a more comprehensive characterization of CNVs that will provide the basis for determining how genomic diversity impacts biological function, evolution and common diseases in human as well as plants. Methods: The analytical variability in next generation sequencing (NGS) and artifacts in coverage data along with lack of robust bioinformatics tools for CNV detection have limited the utility of targeted NGS data to identify CNVs. Literature has the evidence of development of deep learning-based pipeline that incorporates a machine learning component to identify CNVs from targeted NGS data. Result: It is believed that combining this with clinical “gold standard” (e.g. FISH) information, the CNV detection could be more accurate. This would lead to a new research direction, supplementing the existing NGS methods.

Homologous Recombination Abnormalities Associated With BRCA1/2 Mutations as Predicted by Machine Learning of Targeted Next-Generation Sequencing Data.

Homologous recombination deficiency (HRD) is the hallmark of breast cancer gene 1/2 (BRCA1/2)-mutated tumors and the unique biomarker for predicting response to double-strand break (DSB)-inducing drugs. The demonstration of HRD in tumors with mutations in genes other than BRCA1/2 is considered the best biomarker of potential response to these DSB-inducer drugs. We explored the potential of developing a practical approach to predict in any tumor the presence of HRD that is similar to that seen in tumors with BRCA1/2 mutations using next-generation sequencing (NGS) along with machine learning (ML). We use copy number alteration (CNA) generated from routine-targeted NGS data along with a modified naïve Bayesian model for the prediction of the presence of HRD. The CNA from NGS of 434 targeted genes was analyzed using CNVkit software to calculate the log2 of CNA changes. The log2 values of various sequencing reads (bins) were used in ML to train the system on predicting tumors with BRCA1/2 mutations and tumors with abnormalities similar to those detected in BRCA1/2 mutations. Using 31 breast or ovarian cancers with BRCA1/2 mutations and 84 tumors without mutations in any of 12 homologous recombination repair (HRR) genes, the ML demonstrated high sensitivity (90%, 95% confidence interval [CI] = 73%-97.5%) and specificity (98%, 95% CI = 90%-100%). Testing of 114 tumors with mutations in HRR genes other than BRCA1/2 showed 39% positivity for HRD similar to that seen in BRCA1/2. Testing 213 additional wild-type (WT) cancers showed HRD positivity similar to BRCA1/2 in 32% of cases. Correlation with proportional loss of heterozygosity (LOH) as determined using whole exome sequencing of 51 samples showed 90% (95% CI = 72%-97%) concordance. The approach was also validated in an independent set of 1312 consecutive tumor samples. These data demonstrate that CNA when combined with ML can reliably predict the presence of BRCA1/2 level HRD with high specificity. Using BRCA1/2 mutant cases as gold standard, this ML can be used to predict HRD in cancers with mutations in other HRR genes as well as in WT tumors.

Copy Number Loss at Chromosome 14q11.2 Correlates With the Proportion of T Cells in Biopsies and Helps Identify T-Cell Neoplasms.

Evidence of T-cell clonality is often critical in supporting the diagnosis of a T-cell lymphoma. To retrospectively explore the significance of copy number losses at the 14q11.2 T-cell receptor α locus in relation to the presence of a T-cell neoplasm and proportion of T cells by targeted next-generation sequencing. Targeted next-generation sequencing data from 139 tissue biopsies, including T-cell lymphomas, B-cell lymphomas, classic Hodgkin lymphomas, nonhematopoietic malignancies, and normal samples, were reviewed for copy number losses involving the T-cell receptor α gene segments at chr14q11.2. We found that biallelic or homozygous deletion of 14q11.2 was found in most (28 of 33, 84.8%) T-cell lymphomas. The magnitude of 14q11.2 loss showed a statistically significant correlation with the proportion of T cells in lymphoma tissue samples. Copy number losses could also be detected in other lymphomas with high numbers of T cells (8 of 32, 25% of B-cell lymphomas, 4 of 4 classical Hodgkin lymphomas), though biallelic/homozygous deletion of 14q11.2 was not significantly observed outside of T-cell lymphomas. Most nonhematopoietic neoplasms and normal tissues (59 of 64, 92.2%) showed no significant copy number losses involving the T-cell receptor α locus at chr14q11.2. Analysis of copy number losses at the T-cell receptor α locus chr14q11.2 with targeted next-generation sequencing can potentially be used to estimate the proportion of T cells and detect T-cell neoplasms.

Genomic Landscapes and Hallmarks of Mutant RAS in Human Cancers.

The complex genomic landscape of RAS-mutant tumors is reflective of selection processes in a cancer lineage-specific and context-dependent manner, highlighting differential therapeutic vulnerabilities that can be clinically translated.

Benchmarking of 5 algorithms for high-resolution genotyping of human leukocyte antigen class I genes from blood and tissue samples.

BackgroundSpecific alterations in human leukocyte antigen class I (HLA-I) loci are associated with clinical outcomes for immune checkpoint inhibitors, which increase the clinical relevance of accurate high-resolution HLA genotyping in immuno-oncology applications. Numerous algorithms have been developed for high- to full-resolution HLA genotyping by next-generation sequencing (NGS) data; however, Sanger sequencing-based typing (SBT) remains the gold standard. With the increasing use of NGS for clinical oncology, it is important to identify the computational tool with comparable performance as the gold standard. This study aimed to benchmark 5 algorithms against SBT for the high-resolution typing of classical HLA-I genes for targeted NGS data from blood and tissue samples.MethodsPaired white blood cell (WBC), plasma, and tissue deoxyribonucleic acid (DNA) samples derived from 22 cancer patients with known HLA genotypes were sequenced using a panel of all the following exons of classical HLA-I genes: HLA-A, HLA-B, and HLA-C. NGS-based genotypes were generated by the 5 different algorithms, including HLA-HD, HLAscan, OptiType, Polysolver, and xHLA. Accuracy was defined as the concordance between the SBT and NGS-based algorithms. Accuracy was computed as the fraction of all the alleles with concordant genotype using the SBT and any of the algorithm over the total number of alleles.ResultsIn relation to the WBC, plasma, and tissue samples, all 5 algorithms were highly accurate at low-resolution HLA-I genotyping, but had more varied accuracy at high-resolution HLA-I genotyping, particularly in HLA-A. The in-silico analyses revealed that high-resolution genotyping by all 5 algorithms achieved approximately 90% accuracy at sequencing depths of 6,000× – 100× for the WBC samples, at 6,000× – 700× for the plasma samples, and at 1,000× – 100× for the tissue samples. Among the 5 algorithms, HLA-HD was consistently accurate at high-resolution HLA-I genotyping, and had an accuracy of 93.9% for the WBC samples, 87.9% for the plasma samples, and 94.2% for tissue samples even at a 50× sequencing depth.ConclusionsWe found that HLA-HD was an accurate algorithm for the high-resolution genotyping of classical HLA-I genes sequenced by our targeted panel, particularly at a sequencing depth ≥300× for blood and tissue samples.

Concurrent TP53 Mutations Facilitate Resistance Evolution in EGFR-Mutant Lung Adenocarcinoma

IntroductionPatients with EGFR-mutant NSCLC experience variable duration of benefit on EGFR tyrosine kinase inhibitors. The effect of concurrent genomic alterations on outcome has been incompletely described. MethodsIn this retrospective study, targeted next-generation sequencing data were collected from patients with EGFR-mutant lung cancer treated at the Dana-Farber Cancer Institute. Clinical data were collected and correlated with somatic mutation data. Associations between TP53 mutation status, genomic features, and mutational processes were analyzed. ResultsA total of 269 patients were identified for inclusion in the cohort. Among 185 response-assessable patients with pretreatment specimens, TP53 alterations were the most common event associated with decreased first-line progression-free survival and decreased overall survival, along with DNMT3A, KEAP1, and ASXL1 alterations. Reduced progression-free survival on later-line osimertinib in 33 patients was associated with MET, APC, and ERBB4 alterations. Further investigation of the effect of TP53 alterations revealed an association with worse outcomes even in patients with good initial radiographic response, and faster acquisition of T790M and other resistance mechanisms. TP53-mutated tumors had higher mutational burdens and increased mutagenesis with exposure to therapy and tobacco. Cell cycle alterations were not independently predictive, but portended worse OS in conjunction with TP53 alterations. ConclusionsTP53 alterations associate with faster resistance evolution independent of mechanism in EGFR-mutant NSCLC and may cooperate with other genomic events to mediate acquisition of resistance mutations to EGFR tyrosine kinase inhibitors.

Detection of a pathogenic Alu element insertion in PALB2 gene from targeted NGS diagnostic data.

Despite routine analysis of a large panel of genes, pathogenic variants are only detected in approximately 20% of families with hereditary breast and/or ovarian cancer. Mobile element insertions (MEI) are known to cause genetic diseases in humans, but remain challenging to detect. Retrospective analysis of targeted next-generation sequencing (NGS) data from 359 patients was performed using a dedicated MEI detection pipeline. We detected one MEI in exon 9 of the PALB2 gene in a woman with a family history of breast cancer. The pathogenic variant, c.2872_2888delins114AluL2, disrupts the PALB2 coding sequence and leads to the production of a truncated protein, p.(Gln958Valfs*38). This is the first report of a pathogenic MEI in PALB2. This study illustrates that MEI analysis may help to improve molecular diagnostic yield and can be performed from targeted NGS data used for routine diagnosis.

Somatic Mutations of TP53 Identified by Targeted Next-Generation Sequencing Are Poor Prognostic Factors for Primary Operable Breast Cancer: A Single-Center Study.

Few studies have reported on the clinical utility of targeted next-generation sequencing (NGS) for breast cancer in Korea. We retrospectively reviewed the targeted NGS data of 219 patients with breast cancer who underwent surgical resection between August 2018 and April 2021. Here, we described the mutational profiles of breast cancer and examined their prognostic implications. The most frequently mutated gene was PIK3CA (n = 97/219, 44.3%), followed by TP53 (n = 79/219, 36.1%), AKT1 (n = 23/219, 10.5%), and GATA3 (n = 20/219, 9.1%). TP53 mutations were associated with aggressive histologic features. We followed up for 31 (range, 1-39) months and observed 11 (5.0%) recurrences: nine were TP53 mutant and two were TP53 wild-type. Multivariable analysis revealed that TP53 mutation was an independent prognostic factor for recurrence (p = 0.012). Although no drug is currently available for TP53 mutations, it is valuable to know the mutational status of TP53 for the precise management of breast cancer.

Genomic and Transcriptomic Correlates of Thyroid Carcinoma Evolution after BRAF Inhibitor Therapy.

Targeted inhibition of BRAF V600E achieves tumor control in a subset of advanced thyroid tumors. Nearly all tumors develop resistance, and some have been observed to subsequently undergo dedifferentiation. The molecular alterations associated with thyroid cancer dedifferentiation in the setting of BRAF inhibition are unknown. We analyzed targeted next-generation sequencing data from 639 advanced, recurrent and/or metastatic thyroid carcinomas, including 15 tumors that were treated with BRAF inhibitor drugs and had tissue sampled during or posttreatment, 8 of which had matched pretherapy samples. Pre- and posttherapy tissues from one additional patient were profiled with whole-exome sequencing and RNA expression profiling. Mutations in genes comprising the SWI/SNF chromatin remodeling complex and the PI3K-AKT-mTOR, MAPK, and JAK-STAT pathways all increased in prevalence across more dedifferentiated thyroid cancer histologies. Of 7 thyroid cancers that dedifferentiated after BRAF inhibition, 6 had mutations in these pathways. These mutations were mostly absent from matched pretreatment samples and were rarely detected in tumors that did not dedifferentiate. Additional analyses in one of the vemurafenib-treated tumors before and after anaplastic transformation revealed the emergence of an oncogenic PIK3CA mutation, activation of ERK signaling, dedifferentiation, and development of an immunosuppressive tumor microenvironment. These findings validate earlier preclinical data implicating these genetic pathways in resistance to BRAF inhibitors, and suggest that genetic alterations mediating acquired drug resistance may also promote thyroid tumor dedifferentiation. IMPLICATIONS: The possibility that thyroid cancer dedifferentiation may be attributed to selective pressure applied by BRAF inhibitor-targeted therapy should be investigated further.

Abstract 244: A computation method for noise reduction based on ultra-deep targeted sequencing data

Abstract Detection of circulating tumor DNA is a promising approach for the diagnosis and monitor of solid tumors. However, sensitive detection of ctDNA remains challenging when minor allele frequency (MAF) is ultra-low. Numerous noise reduction data algorithms have been developed to reduce next-generation sequencing (NGS) errors, but variant callers that can take advantage of ultra-deep sequencing data are lacking. Here we develop a bin-based genotyping (BinGo) method for canceling sequencing noise using ultra-deep (&amp;gt; 100,000 X depth) targeted sequencing data. A mutation reference sample with 44 mutations was diluted to different MAFs at 1% and 0.5%. One blood sample for proficiency test was included. We constructed targeted sequencing libraries by anchored multiplex enrichment for a panel of targets that cover the 44 mutations, and sequenced to a very high depth (Mean X ranged: 71K to 350K).After raw NGS reads were mapped by BWA MEM, BinGo tags unique molecular index (UMI) to each alignment (BAM) to preserve alignment signal of all bases in all raw reads. BinGo then bins all reads with the same UMI into one bin, generating millions of bins. Millions of parallel Samtools mpileup was enabled by the GNU parallel tool. Typical mpileup analyses use the human genome (requires ~4GB RAM) as a reference, making millions of such analyses essentially impossible in a regular computing server. To overcome this hurdle, BinGo uses tailor-made reference sequences comprising the targets only (&amp;lt; 100KB), drastically reduces computation memory consumption and enables millions of mpileup analyses. BinGo recompiles Samtools to increase its default max depth from 8,000 to 8,000,000 to accommodate the ultra-deep data. Variants derived from the same bin are effectively phased, thereby enabling calling of di-nuclei mutations such as BRAF V600K (GT&amp;gt;TT) and V600R (GT&amp;gt;AG). This binning also allows for sensitive calling of insertions and deletions even down to a single read. To profile noise patterns, BinGo uses Samtools tview to acquire every base in every alignment position of the target sequences in all raw reads. Finally, BinGo uses multiple statistical testing to call variants, including calling nucleotide changes over the background noise model that is defined by nucleotide change and sequence context; filtering out strand imbalance likely due to DNA template damage, and a low number of supporting ligation starting sites. With 10 ng DNA as input and at an expected MAF of 0.5%, BinGo demonstrates a 100% (44/44) analytical sensitivity and 100% specificity. When using 1 ng DNA as input and at an expected MAF of 0.5%, only one mutation was missed, leading to a sensitivity of 97.73% (43/44) and 100% specificity. Furthermore, BinGo detected in the blood DNA sample a 15-bp EGFR exon 19 deletion at 0.05%.Bingo integrates UMI-bin derived variants and raw sequencing background noise modeling, enabling analysis of ultra-deep targeted NGS data for rare variant detection. Citation Format: Weiwei Bian, Firaol Kebede, Zongli Zheng. A computation method for noise reduction based on ultra-deep targeted sequencing data [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 244.

Population-Scale Identification of Clonal Hematopoiesis Derived Mutations in Solid Tumor Mutation Profiling

▪Background: Clonal hematopoiesis (CH) is the commonly observed age-related acquisition of somatic mutations in hematopoietic lineage leading to clonal expansion of mutated cells. We recently showed 25% of advanced solid tumor patients harbor CH related mutations in their blood. As tumor tissue is typically infiltrated by blood cells, we hypothesized that sequencing tumors without the utilization of matched blood in a paired analysis could erroneously attribute CH mutations to be of tumor-specific origin. In review of patient clinical records, we encountered two patients (one endometrial adenocarcinoma and one mucosal melanoma) with DNMT3A R882H mutations reported by an external tumor-only NGS assay. In each case, paired tumor/blood sequencing subsequently confirmed the hematopoietic origin. Here, we show the prevalence of CH derived mutations in the solid tumors by analyzing a cohort of patients where tumor/normal sequencing is performed using MSK-IMPACT.Methods: We analyzed deep coverage, targeted next-generation sequencing data from January 2014 - May 2017 of paired blood and tumor samples from 15,432 patients with advanced solid malignancies using the MSK-IMPACT platform. MSK-IMPACT detects point mutations and indels in coding exons of currently 468 genes with older versions targeting 341 and 410 genes. We called mutations in blood DNA and identified those associated with CH using a custom filtering schema: First, germline mutations were eliminated based on variant allele fraction (VAF) > 0.35, except for mutations reported in COSMIC on at least 10 occasions. Second, we required mutations to be supported by at least 8 reads and present at twice the variant allele fraction (VAF) or greater in blood cells compared to the tumor. Third, we retained mutations that were detected with a VAF greater than 0.02 for a selected panel of leukemia genes and 0.05 for other genes. For cases where at least one mutation exceeded these thresholds, we reduced the VAF threshold in blood to 0.01 to detect subsequent events for leukemia genes and 0.03 for non-leukemia genes.Results: We identified 4,044 (26.2%) patients with at least one CH derived mutation in their blood corresponding to 6,588 presumptive somatic, non-silent mutations in 350 genes. Out of these, we identified 963 CH mutations from 808 patients (5.2%) that could have been classified as tumor-derived mutations in the absence of paired blood sequencing given a tumor VAF > 0.02. The number of mutations per patient ranged from 1-7 with a median VAF of 0.04 and 0.17 in paired tumor and blood samples, respectively. 34.5% of these variants are present in the COSMIC database with occurrence counts greater than 5 and include well-characterized driver alterations such as DNMT3A R882H, NRAS G12D, and IDH2 R140Q, which could readily be mistaken for somatic mutations in the corresponding solid tumor.Conclusion: With the potential for admixed leukocytes in biopsy specimens, these data demonstrate that failure to employ a paired blood sample in solid tumor sequencing can lead to the misclassification of CH mutations as somatic in the tumor, with clear diagnostic, prognostic, and therapeutic implications. DisclosuresCoombs:H3 Biomedicine: Honoraria; Pharmacyclics: Honoraria. Arcila:Raindance Tecnologies: Honoraria; Archer: Honoraria; Invivoscribe: Honoraria. Levine:Roche: Research Funding; Qiagen: Equity Ownership; Celgene: Research Funding; Celgene: Research Funding; Roche: Research Funding; Qiagen: Equity Ownership.

Clinical implications of copy number alteration detection using panel-based next-generation sequencing data in myelodysplastic syndrome

Recent advancements in next-generation sequencing (NGS) technologies allow the simultaneous identification of targeted copy number alterations (CNAs) as well as somatic mutations using the same panel-based NGS data. We investigated whether CNAs detected by the targeted NGS data provided additional clinical implications, over somatic mutations, in myelodysplastic syndrome (MDS). Targeted deep sequencing of 28 well-known MDS-related genes was performed for 266 patients with MDS. Overall, 215 (80.8 %) patients were found to have at least one somatic mutation; 67 (25.2 %) had at least one CNA; 227 (85.3 %) had either a somatic mutation or CNA; and 12 had CNA without somatic mutations. Considering the clinical variables and somatic mutations alone, multivariate analysis demonstrated that sex, revised International Prognostic Scoring System (IPSS-R), and NRAS and TP53 mutations were independent prognostic factors for overall survival. For AML-free survival, these factors were sex, IPSS-R, and mutations in NRAS, DNMT3A, and complex karyotype/TP53 mutations. When we consider clinical variables along with somatic mutations and CNAs, genetic alterations in TET2, LAMB4, U2AF1, and CBL showed additional significant impact on the survivals. In conclusion, our study suggests that the concurrent detection of somatic mutations and targeted CNAs may provide clinically useful information for the prognosis of MDS patients.

Abstract 4335: Clinical value of EGFR gene amplifications detected using amplicon based targeted next generation sequencing data in lung adenocarcinoma patients

Abstract Objective The relevance of EGFR gene amplification in patients with EGFR-mutated advanced non-small cell lung cancer (NSCLC) has not been fully elucidated. Moreover, a limited number of studies have been published on methods to estimate gene amplifications based on routinely performed targeted next generation sequencing (NGS). In this study we aimed to determine whether PCR-based targeted NGS data on tumor tissue can be used to estimate presence of gene amplifications and explored the prognostic value of EGFR gene amplification in EGFR mutated NSCLC patients. Materials and methods A total of 3,194 good quality targeted NGS data files were retrieved from 2014 to 2017. Among those, 1,729 NSCLC samples originated from 1,586 NSCLC patients of whom 134 had an EGFR mutation (8.2%). Clinical data were available for 66 of the patients. Raw sequencing data were re-analyzed using a custom designed pipeline. The presence of an amplification was based on the read depth of a given amplicon relative to a set of reference amplicons from the sample or relative to a set of normal control samples. Reference amplicons were selected based on low degree of variation in read depth amongst all tested samples. Amplifications were regarded significant when the ratio was ≥3 and the z score was ≥3.5. Technical validation was done by FISH and MLPA. Cox regression analysis on overall survival was done with each of the amplification parameters using SPSS. Results No amplifications were detected for the ALK, KIT, NRAS, PDGFRA, GNAQ and MAP2K1 gene loci, whereas amplifications for BRAF, ESR1, GNA11, HRAS, KRAS, MET and PIK3CA were observed at a low frequency. Depending on the amplification analysis strategy (within sample or relative to normal controls), 19% and 13% of the EGFR mutated group had an EGFR amplification, respectively. In EGFR wild type patients, amplifications were detected in 5% and 4% of the patients using the two methods, respectively. The sensitivity and specificity of the NGS based estimation of amplifications for EGFR was 94% (14/15) and 97% (34/35) respectively for both data analysis approaches. Patients with concurrent EGFR mutations and amplifications (estimated within sample) treated with EGFR-TKI had a significantly worse overall survival compared with those without concurrent EGFR amplifications (ratio, p=0.047 and z score, p=0.015). Cox regression analysis indicated a borderline significant interaction between ratio and z score (p=0.052). Conclusion Routinely obtained amplicon-based NGS data can be used to identify gene amplifications. The presence of EGFR amplifications in EGFR mutant patients is predictive of a worse overall survival. Keywords: Lung adenocarcinoma, EGFR, survival, tyrosine kinase inhibitor Citation Format: Pei Meng, Jiacong Wei, Miente Martijn Terpstra, Anke van Rijk, Menno Tamminga, Frank Scherpen, Arja ter Elst, Mohamed Z. Alimohamed, Lennart F. Johansson, T. Jeroen N. Hiltermann, Harry J. Groen, Klaas Kok, Anthonie J. van der Wekken, Anke van den Berg. Clinical value of EGFR gene amplifications detected using amplicon based targeted next generation sequencing data in lung adenocarcinoma patients [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 4335.

Abstract 846: Microsatellite instability determined by clinical targeted next generation sequencing

Abstract Microsatellite instability (MSI)/mismatch repair (MMR) status is a positive predictor for immune checkpoint inhibitors. Conventional detection method for MSI status relies on polymerase chain reaction (PCR) of 5 microsatellite loci. With the advancement of next-generation sequencing (NGS) applied to clinical samples, we developed a MSI prediction algorithm to determine MSI status from large-panel clinical targeted NGS data accounting for more than 500 microsatellite loci. A set of colorectal cancer clinical samples with known MSI status by PCR were used to train a MSI prediction model, which predicts a sample as microsatellite instability high (MSI-H) or microsatellite stable (MSS). The model was constructed based on the change of number of repeats of different lengths from a pooled MSS baseline across microsatellites in a comprehensive cancer genomic panel targeting 440 genes. Using a quantitative measure based on more than 500 microsatellites loci, the model achieves high robustness and efficiency for MSI status detection from NGS data without matched normal. An independent dataset of clinical samples across various cancer types were used to validate the model showing an accuracy of 96.8%. In summary, MSI status can be determined by the MSI prediction model from clinical targeted NGS panel without normal paired sample. Citation Format: Ya-Chi Yeh, Chien-Hung Chen, Kuan-Ying Chen, Ying-Ja Chen, Shu-Jen Chen. Microsatellite instability determined by clinical targeted next generation sequencing [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 846.

Evaluation of CNV detection tools for NGS panel data in genetic diagnostics

Although germline copy-number variants (CNVs) are the genetic cause of multiple hereditary diseases, detecting them from targeted next-generation sequencing data (NGS) remains a challenge. Existing tools perform well for large CNVs but struggle with single and multi-exon alterations. The aim of this work is to evaluate CNV calling tools working on gene panel NGS data and their suitability as a screening step before orthogonal confirmation in genetic diagnostics strategies. Five tools (DECoN, CoNVaDING, panelcn.MOPS, ExomeDepth, and CODEX2) were tested against four genetic diagnostics datasets (two in-house and two external) for a total of 495 samples with 231 single and multi-exon validated CNVs. The evaluation was performed using the default and sensitivity-optimized parameters. Results showed that most tools were highly sensitive and specific, but the performance was dataset dependant. When evaluating them in our diagnostics scenario, DECoN and panelcn.MOPS detected all CNVs with the exception of one mosaic CNV missed by DECoN. However, DECoN outperformed panelcn.MOPS specificity achieving values greater than 0.90 when using the optimized parameters. In our in-house datasets, DECoN and panelcn.MOPS showed the highest performance for CNV screening before orthogonal confirmation. Benchmarking and optimization code is freely available at https://github.com/TranslationalBioinformaticsIGTP/CNVbenchmarkeR.

HemoMIPs-Automated analysis and result reporting pipeline for targeted sequencing data.

Targeted sequencing of genomic regions is a cost- and time-efficient approach for screening patient cohorts. We present a fast and efficient workflow to analyze highly imbalanced, targeted next-generation sequencing data generated using molecular inversion probe (MIP) capture. Our Snakemake pipeline performs sample demultiplexing, overlap paired-end merging, alignment, MIP-arm trimming, variant calling, coverage analysis and report generation. Further, we support the analysis of probes specifically designed to capture certain structural variants and can assign sex using Y-chromosome-unique probes. In a user-friendly HTML report, we summarize all these results including covered, incomplete or missing regions, called variants and their predicted effects. We developed and tested our pipeline using the hemophilia A & B MIP design from the “My Life, Our Future” initiative. HemoMIPs is available as an open-source tool on GitHub at: https://github.com/kircherlab/hemoMIPs

Secondary myelodysplastic syndrome identified via next-generation sequencing in non-small cell lung cancer patients.

e19531 Background: Chemotherapy and radiotherapy used in cancer treatment are associated with increased risk of secondary myelodysplastic syndrome (MDS). MDS is characterized with specific chromosomal abnormalities and genomic alterations in JAK2, KRAS, CBL, DNMT3A, TET2, IDH1/2, EVI1, RUNX1, GATA2, EZH2, ASXL1, SF3B1, U2AF1, SRSF2 and ZRSR2. Since next generation sequencing (NGS) techniques have been widely used for the detection of actionable genomic variation in solid tumor patients, analysis of NGS data for chromosomal abnormalities and gene mutations may be supplementary for diagnosis of secondary MDS. Methods: We retrospectively analyzed targeted NGS data of paired tumor and blood samples from patients with non-small cell lung cancer (NSCLC) for chromosomal abnormalities and gene mutations. Sequence panel covered whole exons of 381 genes associated with cancers. Sequence data were processed using a customized analysis pipeline designed to accurately detect multiple classes of genomic alterations in routine clinical specimens. All testing was done in a CLIA-certified, CAP-accredited laboratory. Clinical information of cases with chromosomal abnormalities was reviewed, including treatment for previous cancer, blood and bone marrow tests. Results: We found 4 cases carrying MDS associated chromosome abnormalities among 3523 NSCLC patients with the prevalence of0.1%. They were old (from 66 to 71 years old), and were more likely to be male (3 males and 1 females). All of 4 cases included common abnormalities associated with MDS like 5q-, 7q-, -7, +21. 3 cases had complex (≥3) abnormalities and 1 case had double abnormalities. Point mutations were identified in blood samples of 3 cases. Mutations in TP53 occurred in 3 cases, mutations in ASXL1, DNMT3A, FLT3 and CBL occurred in one case respectively. Among the 4 cases, 2 cases were confirmed as MDS by blood and bone marrow tests. Conclusions: We identified 4 MDS associated chromosomal abnormalities from 3523 NSCLC patients through analyzing NGS data. NGS techniques could serve as supplementary for diagnosis of secondary MDS. [Table: see text]