Mutational Signature Analysis Research Articles

Feasibility and clinical utility of cancer whole-genome sequencing in adult solid tumors.

e15055 Background: As sequencing costs decrease, whole-genome sequencing (WGS) is expected to be utilized in actual cancer patient care. However, there is insufficient evidence that WGS can be applied to the real-world cancer patient care, and how it can help in the diagnosis and treatment process. Methods: This prospective cohort study was designed to evaluate the feasibility and clinical utility of WGS in adult solid cancer care at a tertiary hospital. All solid cancer patients who were scheduled to undergo biopsy or surgery for diagnosis or treatment were eligible for the study. Either fresh or formalin-fixed paraffin-embedded tissue with match peripheral blood were obtained. Tumors were sequenced to depth of 40x, and WGS data were curated by Genome Insight Inc. The WGS reports were provided to clinician through a web-based, user-browsable reporting system in real time. Results: 97 patients were enrolled between Sep 2022 and Jan 2023. Besides 10 QC failed cases, 87 patients were provided WGS reports. Median turn-around-time was 14 day (range 12-23). The clinical utility of WGS was evaluated in 77 cancer patients, excluding 10 cases with estimated cancer cell fraction less than 15%. 31 patients (40.3%) had more than one actionable alteration targeted by on-label drugs. 49 patients (63.6%) also had more than one actionable alteration targeted by either on-label or off-label drugs. In selected cases (N = 6), WGS could suggest the diagnosis for cancer of unknown origin or the revision of pathologic diagnosis that was regarded as a double primary cancer into a metastasis of a single cancer. By providing information of total mutation counts, whole-genome-wide mutational signature analysis, copy number variation with high accuracy, germline alterations, detection of homologous recombination deficiency, and presence of loss-of-heterozygosity, WGS could further refine the actionable mutations in 39% cases, and estimate the etiology of cancer (genomic or environmental) or to determine hereditary or sporadic cancer in 60% cases. In 90% cases, WGS demonstrated clinical utility with supporting diagnosis, informing carcinogenesis, or annotating actionable alterations. Conclusions: WGS was applicable in real-world cancer care. WGS demonstrated clinical utility in the diagnosis and treatment of adult cancer patients. Clinical trial information: KCT0008118 . [Table: see text]

Integrated clinico-genomic characterization of pancreatic ductal adenocarcinoma in an Asian population.

e16291 Background: Pancreatic cancer is a devastating malignancy with an extremely poor prognosis. However, the clinico-genomic understanding of pancreatic cancer that could play an important role to improve prognosis is highly limited, especially in Asian population. Methods: Using blood sample and primary pancreas tissues by endoscopic ultrasound guided fine needle biopsy (EUS-FNB) from 237 patients diagnosed with pancreatic ductal adenocarcinoma (PDA) at Seoul National University Bundang Hospital, WES and WTS were performed. Integrated clinico-genomic analysis was conducted by combining genomic data with diverse clinical information. Results: WES and WTS data were obtained from 205 and 93 patients, respectively. In WES, recurrent somatic mutations were identified in KRAS (89.3%), TP53 (68.8%), SMAD4 (22.4%) and CDKN2A (20.5%), and the overall mutation profile was not significantly different from TCGA dataset. Germline mutations related with hereditary cancer syndrome were found in 19 (9.3%), and BRCA2 mutation was the most common as in the Western data. At least one actionable mutation was reported in 61 (29.8%), and KRAS p.G12C, the emerging actionable mutation, was found in five (2.4%). In mutational signature analysis with single-base substitution (SBS), SBS1, SBS3 and SBS13 were identified in this cohort, and SBS3, a potential biomarker for DDR pathway related therapy, was positive in 25 (12.2%). Since this cohort contained 112 stage I-III (47.3%) and 125 stage IV (52.7%), genetic features according to stage or metastatic patterns were analyzed. There was no difference of mutation frequency and burden according to stage. However, liver metastasis was associated with higher frequency of TP 53 mutation (OR = 3.40; P = 0.010) and mutation burden was significantly lower in the lung metastasis (35.4 vs 50.9; P = 0.023). In the gene-set enrichment analysis using WTS, distinct pathways were observed according to the potential mechanism related with each metastatic pattern. Conclusions: Integrative clinico-genomic analysis of Asian PDA cohort provided various clues to better understand the genomic landscape of PDA.

Metastatic breast cancer (MBC) with ultra-high tumor mutational burden (UHTMB): A comprehensive genomic profiling (CGP) study.

1036 Background: Pts with MBC whose tumors feature high TMB (≥ 10 mutations/Mb) are eligible for on label immune checkpoint inhibitor (ICI) treatment. This study evaluated the genomic landscape of MBC with “Ultra high” TMB, defined at > 20 mutations/Mb. Methods: 2049 MBC underwent hybrid capture-based CGP to evaluate all classes of genomic alterations (GA), TMB, microsatellite instability (MSI) and trinucleotide mutational signatures. HER2 IHC results were available in a subset of pts. PD-L1 expression on immunocytes was determined by IHC (Ventana SP142). Results: 45/2049 (2.2%) of MBC were UHTMB. 45 (100%) pts had metastatic disease. 38 (84%) had documented Stage IV disease and 7 documented axillary LN metastases at the time of sequencing. Local breast tumor was used for CGP in 19 (42.2%) MBC and metastatic site biopsy was used in 26 (57.8%). When compared with 2004 non-UHTMB pts with UHTMB were older (mean 64.6 yrs vs 58.2 yrs p < .0001), more often had lobular histology (40.0% vs 14.5% p < .0001) and ER+ disease (86.6% vs 70.0%). Of the 35 UHTMB cases with HER2 IHC data available, 11 (31.4%%) were HER2 IHC negative (0+), 21 (60.0%) were HER2-low status (9 1+ and 12 2+/ISH negative) and 3 (8.6%) were HER2 IHC positive (3+). 1/3 HER2 IHC2+ cases and 2/45 (4.4%) of all UHTMB cases were positive for HER2 copy number gain on CGP. UHTMB cases had more driver GA/tumor (mean 9.8 vs 5.7 p < .0001) and were less often TNBC (13.3% vs 27.0% p = .041) compared to non-UHTMB high cancers. Mutation signature analysis revealed APOBEC was predominant in UHTMB samples (82.5%); MMR signature was also observed in 10% of cases. MSI high status was significantly more frequent in UHTMB high cases (11.6% vs 0.4% p < .0001). GA more frequently identified in UHTMB cases included CDH1 (45.5% vs 14.3% p < .0001), PIK3CA (81.8% vs 37.9% p < .0001), CDKN2A (11.4% vs 3.2% p = .017), ARID1A (25.0% vs 5.0% p < .0001) and NF1 (20.5% vs 5.9% p = .0014). PD-L1 ( CD274) gene amplification (2.3% vs 1.3%) or protein expression by the Ventana SP142 assay (57.14% vs 51.10%) were not significantly different among groups. Conclusions: UHTMB MBC is a rare, yet clinically important subset of clinically advanced breast cancer driven by APOBEC mutagenesis, with high incidence of ER+ lobular histology and frequent alterations in CDH1 and PIK3CA. In addition to potential benefit from ICI based treatment, UHTMB MBC present with a high frequency of HER2-low status which may impact therapy decisions for this rare disease. [Table: see text]

Mutational signatures association with replication timing in normal cells reveals similarities and differences with matched cancer tissues

Mutational signatures’ association with replication timing (RT) has been studied in cancer samples, but the RT distribution of somatic mutations in non-cancerous cells was only minimally explored. Here, we performed comprehensive analyses of mutational signatures in 2.9 million somatic mutations across multiple non-cancerous tissues, stratified by early and late RT regions. We found that many mutational processes are active mainly or solely in early RT, such as SBS16 in hepatocytes and SBS88 in the colon, or in late RT, such as SBS4 in lung and hepatocytes, and SBS18 across many tissues. The two ubiquitous signatures, SBS1 and SBS5, showed late and early bias, respectively, across multiple tissues and in mutations representing germ cells. We also performed a direct comparison with cancer samples in 4 matched tissue-cancer types. Unexpectedly, while for most signatures the RT bias was consistent in normal tissue and in cancer, we found that SBS1’s late RT bias is lost in cancer.

14 Carcinogenicity Risk Assessment of Various Carbon Nanotubes by Intra-Tracheal Intra-Pulmonary Spray (TIPS) Dosing Followed by 2-Year Observation

Abstract Multi-walled carbon nanotubes (MWCNTs) are composed of multiple coaxially arranged graphene cylinders. MWCNTs range in diameter from 1 to over 100 nm depending on the number of constituting graphene cylinders. MWCNTs can be divided into two general subtypes, straight and tangled: increasing the number of constituent graphene cylinders increases the number of layers that compose the carbon nanotube wall, and as the number of wall layers increase, the wall thickness increases and the MWCNT becomes more rigid and straight. MWCNTs with a low number of wall layers are flexible and can assemble into tangled agglomerates. We initially showed MWCNT-N, a straight-type MWCNT with 30-40 walls, induced lung tumors and mesotheliomas after administration by intra-tracheal intra-pulmonary spraying (TIPS): rats were administered a total dose of 1mg/rat during the initial 2 weeks of the study, and animals were observed for 2 years without any further treatments. Using this short-term dosing and 2-year observation method we conducted carcinogenicity tests for other MWCNTs with different cylinder wall layers. We found that both straight type MWCNTs, MWCNT-7 (40 layers), MWCNT-A (150 layers), and tangled type MWCNTs, MWCNT-B (15 layers) and double-walled CNT (DWCNT) (2 layers) induced lung tumors and/or pleural mesotheliomas. In contrast to these results, we found carbon nanohorns, carbon nanostructures composed of single-walled carbon nanotubes (SWCNT), and carbon nanobrushes, a fibrous aggregate of carbon nanohorns, were not carcinogenic. Data on foreign body radical production, microarray and mutational signature analyses, and gene alterations will be presented.

Prevalence of pathogenic variants in cancer-predisposing genes in second cancer after childhood solid cancers.

Second malignant neoplasms (SMNs) are one of the most severe late complications after pediatric cancer treatment. However, the effect of genetic variation on SMNs remains unclear. In this study, we revealed germline genetic factors that contribute to the development of SMNs after treatment of pediatric solid tumors. We performed whole-exome sequencing in 14 pediatric patients with SMNs, including three brain tumors. Our analysis revealed that five of 14 (35.7%) patients had pathogenic germline variants in cancer-predisposing genes (CPGs), which was significantly higher than in the control cohort (p < 0.01). The identified genes with variants were TP53 (n=2), DICER1 (n=1), PMS2 (n=1), and PTCH1 (n=1). In terms of the type of subsequent cancer, leukemia and multiple episodes of SMN had an exceptionally high rate of CPG pathogenic variants. None of the patients with germline variants had a family history of SMN development. Mutational signature analysis showed that platinum drugs contributed to the development of SMN in three cases, which suggests the role of platinum agents in SMN development. We highlight that overlapping effects of genetic background and primary cancer treatment contribute to the development of second cancers after treatment of pediatric solid tumors. A comprehensive analysis of germline and tumor samples may be useful to predict the risk of secondary cancers.

Abstract 6073: Distinct subtype of multiple myeloma revealed by whole genome and transcriptome sequencing

Abstract Background: Advanced sequencing technologies have contributed to identifying genomic and transcriptomic features of various types of tumors. Despite several sequencing data from multiple myeloma (MM) being generated, understanding various properties of MM remained insufficient. Methods: The patients confirmed with MM by bone marrow examination were analyzed in the study. Myeloma cells were enriched from the bone marrow aspirates of the patients using CD138. Whole genome sequencing (WGS) and whole transcriptome sequencing (WTS) were performed for isolated myeloma cells. Mutational signatures were analyzed using somatic mutations detected by WGS. Results: A total of 37 MM patients (median age 67) were enrolled in this study. WGS revealed somatic mutations of an average of 8872 single nucleotide variants (SNVs) and 934 insertions and deletions (Indels) from the patients, respectively. By the analysis of structural variants, 7 chromothripsis and 4 chromoplexy patients were identified. Mutational signature analysis showed that single base substitution (SBS) 9 signature largely varied between patients. SBS9 signature high patients were mostly diagnosed with non-IgG/non-IgA heavy chain and lambda light chain type myeloma. Notably, IgD myeloma patients were solely detected in the SBS9 signature high group compared to the low group (p = 0.001). Gene expression patterns obtained from WTS were well-divided into SBS9 signature high and low groups, and analysis of differentially expressed genes (DEG) between SBS9 signature high and low groups implied IgD myeloma features including high expression of IGHD. As a validation test, we tried mutational signature analysis on whole exome sequencing (WES) data of 784 MM patients from the MMRF database. Of the 38 patients with more than 500 exonic mutations, only the SBS9 signature high group had IgD myeloma candidates, whereas the SBS9 signature low group had no IgD myeloma candidate (p = 0.014). Conclusions: DNA and RNA sequencing of myeloma patients could classify specific subtypes of disease categories. Further study for SBS 9 signature high and IgD myeloma is needed to discover the underlying mechanism of distinct features and to find out their clinical implication. Citation Format: Sheehyun Kim, Hyundong Yoon, Youngil Koh, Sung-Soo Yoon. Distinct subtype of multiple myeloma revealed by whole genome and transcriptome sequencing. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 6073.

Abstract 3505: Cancer disparities in race exposed through geospatial analysis of mutational signatures and environmental exposure

Abstract Background: Cancer has been disproportionally affecting minorities in the US due to socioeconomic, environmental, and genetic disparities. Cancer disparities are usually measured in incidence, mortality, survival, etc. Genomic-based cancer disparity analyses have been less common. In the past decade, mutational signatures were proposed as characteristic footprints of endogenous or exogenous carcinogens, which remarkably propelled the progress of genomic cancer research. The disparities caused by uneven exposure to environmental pollutants may be recorded in mutational signatures. Method: The Cancer Genome Altas is one of the richest genomic cancer datasets, empowering hundreds of secondary and tertiary cancer studies. One aspect of TCGA data that is seldomly utilized is geospatial information. Based on the contribution hospital location, we estimated the patients’ approximate location. Furthermore, utilizing data on 450 pollutants released by the Environmental Protection Agency from 2007-2017, we were able to establish associations between mutational signatures and certain pollutants. To remove the noise introduced by approximating the location and average pollution level, stringent statistical procedures were applied. Results: First, we showed that mutation frequencies varied substantially between different races. For example, TP53 was mutated in 46% of Black breast cancer patients, compared to 31% in Whites. Such differences in single genes translated to the level of mutational signatures. Liver cancer has aflatoxin as an affirmed carcinogen, and we compared the quantities of aflatoxin signature across races. It was found that aflatoxin contributed significantly higher (FDR = 0.002) to mutations in Asian than in Caucasians. In-depth analyses revealed that aflatoxin had a strong effect on hepatitis C virus infection in addition to the previously reported effect on hepatitis B virus infection. Similarly, as a known critical player in head and neck tumors, APOBEC signature was found at a significantly higher level (FDR = 0.01) in patients with positive infection of human papillomavirus. Furthermore, analyses from the geospatial perspective revealed several potential associations between pollutants and mutational signatures. For example, the pollutant carbofuran is positively associated with mutational signature SBS10a with the etiology of Polymerase epsilon exonuclease domain mutations (FDR = 0.02). Summary: We conducted a thorough mutational signature-based cancer disparity analysis. Our results reinforced previously known facts and expectations of oncogene mutations. The novel results demonstrated representative links between various forms of cancer disparity and genomic mutation footprints. Most importantly, such disparities can be interrogated at the mutational signature level and correlated to environmental pollutants. Citation Format: Yan Guo, Judy Bai, Katherine Ma, Shangyang Xia, Shuguang Leng, Hui Yu, Xi Gong. Cancer disparities in race exposed through geospatial analysis of mutational signatures and environmental exposure [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 3505.

Abstract 2055: Prevalent focal amplification of oncogene in lung adenocarcinomas without canonical driver alterations

Abstract Lung adenocarcinoma (LADC) is the most common type of lung cancer and a leading cause of death worldwide. Typical oncogenes that drive LADCs when mutated (referred to as canonical drivers) include EGFR (30-60%), KRAS (10-30%), and fusion oncogenes (~10%) involving ALK, RET, and ROS1. However, approximately 30% of lung adenocarcinomas lack canonical drivers, implying unseen kinds of genomic alterations. Here, we explore alterations beyond the scope of canonical drivers in LADCs (referred to as non-canonical drivers) using large-scale whole-genome sequencing (WGS) of LADCs. We gathered total 183 WGS of LADCs, 138 of which were from our previous study (Lee et al. Cell. 2019) and 45 were newly sequenced from surgically resected LADCs. For raw data processing, we used a standard WGS analysis pipeline from read mapping to variant calling. Variant filtering and downstream analyses were performed using in-house scripts. Genomic variants analyzed in this study include single-base substitution (SBS), indel, structural variation (SV), and copy number variation (CNV). As a result, we found that 53 cases (29%) had no canonical drivers (ND group). Intriguingly, LADCs without canonical drivers had a contrasting genomic landscape compared to LADCs with canonical drivers (CD group). First, the point-mutation burden was ~3-fold higher in the ND group than in the CD group (52,500 vs. 17,900 for SNVs, p&amp;lt;0.001). Mutational signature analysis revealed that these additional mutations are attributable mainly to tobacco smoking (by mutational signature SBS4) and APOBEC-mediated mutagenesis (SBS12 and SBS13) and that they are more enriched in the ND group (4,800 vs 2,600, p=0.03). Second, the ND group harbored more frequent focal amplifications. Recurrently amplified genes include TERT, IL7R, NKX2-1, CCND1, CCND3, CCNE1, and EGFR. The pattern of SVs suggest that the amplifications occurred through the formation of extra-chromosomal DNA involving multiple chromosomes or following chromothripsis. Overall, this study broadens our understanding of non-canonical drivers of LADCs. Similar analyses of thousands of cancers with clinical information of the patient will likely yield the functional impact of these driver mutations and may offer more opportunities for LADC treatment in the future. Citation Format: Seongyeol Park, Joonoh Lim, Kijong Yi, Boram Yi, Ryul Kim, Jaemo Koo, Kwon Joong Na, Samina Park, In Kyu Park, Chang Hyun Kang, Jeong Seok Lee, Young Seok Ju, Young Tae Kim. Prevalent focal amplification of oncogene in lung adenocarcinomas without canonical driver alterations [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 2055.

Abstract 3506: High-grade serous ovarian cancer somatic mutational signatures in Black and White individuals

Abstract Causes of racial disparities in ovarian cancer survival are likely multifactorial, including socio-environmental, structural, and biological factors. Mutational signatures reflect endogenous and exogenous exposures, which may differ by race and ethnicity. This study aims to characterize mutational signatures and their associated etiologies among Black high-grade serous ovarian cancer (HGSC) cases, and to compare the occurrence and distribution of signatures with White TCGA cases. Mutational profiling with whole exome sequencing in matched blood and tumor formalin-fixed paraffin-embedded specimens was performed in 271 HGSC cases identifying as Black or African American and participating in the African American Cancer Epidemiology Study (AACES) or the North Carolina Ovarian Cancer Study (NCOCS). After filtering, we identified 28,601 high-confidence coding mutations in 256 samples that passed quality control assessments. The median number of mutations per sample was 67, similar to that in TCGA at 66. Mutation frequencies of the major genes identified in TCGA were nearly identical in Black AACES/NCOCS cases; mutations in TP53 were present in 96% and 89%, respectively, and mutation frequencies in BRCA1/2, CSMD3, NF1, CDK12, FAT3, GABRA6, and RB1 differed by only 0-2%. We performed mutational signature analysis using SigProfiler and refitting to the Catalogue of Somatic Mutations in Cancer (COSMIC) previously published single base-pair substitution (SBS) signatures reflecting mutational processes resulting from known endogenous and exogenous exposures. SBS signatures are defined by 96 mutation features including the single nucleotide variant class (e.g., C&amp;gt;A, C&amp;gt;G, C&amp;gt;T, T&amp;gt;A, T&amp;gt;C, and T&amp;gt;G) and the identity of the 5’/3’ flanking bases. We observed much higher frequencies of the mismatch repair deficiency, homologous recombination deficiency, ultraviolet light, and treatment exposure signatures in Black AACES/NCOCS cases (26%, 22%, 9%, and 8%, respectively) than in White TCGA cases (3%, 10%, 3%, and 2%, respectively). The frequencies of the remaining signatures with known etiologies (clock-like, reactive oxygen species, base excision repair, mutagen exposure, DNA polymerase epsilon, and tobacco signatures) differed by less than 5% between Black and White cases. Many mismatch repair deficiency signatures identified in Black AACES/NCOCS cases are dominated by C&amp;gt;T or T&amp;gt;C mutations and may represent an artifact from formalin fixation. While we observed that the somatic mutation frequencies in major genes associated with HGSC were similar in Black and White cases, the frequencies of some known mutational signatures were considerably higher in Black HGSC. In particular, despite similar frequencies of BRCA1/2 mutations, the homologous recombination deficiency signature was considerably more common in Black than White HGSC. Citation Format: Katherine A. Lawson-Michod, David Nix, Lindsay Collin, Natalie Davidson, Ariel Hippen, Chad Huff, Aaron Atkinson, Lucas A. Salas, Lauren Peres, Casey Greene, Joellen Schildkraut, Jeffrey Marks, Jennifer A. Doherty. High-grade serous ovarian cancer somatic mutational signatures in Black and White individuals [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 3506.

Molecular Biomarkers of Disease Outcomes and Mechanisms of Acquired Resistance to First-Line Osimertinib in Advanced EGFR-Mutant Lung Cancers.

Preferred first-line treatment for patients with metastatic EGFR-mutant lung cancer is osimertinib, yet it is not known whether patient outcomes may be improved by identifying and intervening on molecular markers associated with therapeutic resistance. All patients with metastatic EGFR-mutant lung cancer treated with first-line osimertinib at the Memorial Sloan Kettering Cancer Center (n= 327) were identified. Available pretreatment and postprogression tumor samples underwent targeted gene panel sequencing and mutational signature analysis using SigMA algorithm. Progression-free survival (PFS) and overall survival were estimated using the Kaplan-Meier method. Using multivariate analysis, baseline atypical EGFR (median PFS= 5.8 mo, p < 0.001) and concurrent TP53/RB1 alterations (median PFS= 10.5 mo, p= 0.015) were associated with shorter PFS on first-line osimertinib. Of 95 patients with postprogression biopsies, acquired resistance mechanisms were identified in 52% (off-target, n= 24; histologic transformation, n= 14; on-target, n= 12), with MET amplification (n= 9), small cell lung transformation (n= 7), and acquired EGFR amplification (n= 7), the most frequently identified mechanisms. Although there was no difference in postprogression survival on the basis of identified resistance (p= 0.07), patients with subsequent second-line therapy tailored to postprogression biopsy results had improved postprogression survival (hazard ratio= 0.09, p= 0.006). The paired postprogression tumors had higher tumor mutational burden (p= 0.008) and further dominant APOBEC mutational signatures (p= 0.07) compared with the pretreatment samples. Patients with EGFR-mutant lung cancer treated with first-line osimertinib have improved survival with treatment adaptation on the basis of identified mechanisms of resistance at time of progression using tissue-based genomic analysis. Further survival gains may be achieved using risk-based treatment adaptation of pretreatment genomic alterations.

Undifferentiated and Dedifferentiated Metastatic Melanomas Masquerading as Soft Tissue Sarcomas: Mutational Signature Analysis and Immunotherapy Response

The distinction between undifferentiated melanoma (UM) or dedifferentiated melanoma (DM) from undifferentiated or unclassifiable sarcoma can be difficult and requires the careful correlation of clinical, pathologic, and genomic findings. In this study, we examined the utility of mutational signatures to identify patients with UM/DM with particular attention as to whether this distinction matters for treatment because the survival of patients with metastatic melanoma has dramatically improved with immunologic therapy, whereas durable responses are less frequent in sarcomas. We identified 19 cases of UM/DM that were initially reported as unclassified or undifferentiated malignant neoplasm or sarcoma and submitted for targeted next-generation sequencing analysis. These cases were confirmed as UM/DM by harboring melanoma driver mutations, UV signature, and high tumor mutation burden. One case of DM showed melanoma in situ. Meanwhile, 18 cases represented metastatic UM/DM. Eleven patients had a prior history of melanoma. Thirteen of 19 (68%) of the tumors were immunohistochemically completely negative for 4 melanocytic markers (S100, SOX10, HMB45, and MELAN-A). All cases harbored a dominant UV signature. Frequent driver mutations involved BRAF (26%), NRAS (32%), and NF1 (42%). In contrast, the control cohort of undifferentiated pleomorphic sarcomas (UPS) of deep soft tissue exhibited a dominant aging signature in 46.6% (7/15) without evidence of UV signature. The median tumor mutation burden for DM/UM vs UPS was 31.5 vs 7.0 mutations/Mb (P < .001). A favorable response to immune checkpoint inhibitor therapy was observed in 66.6% (12/18) of patients with UM/DM. Eight patients exhibited a complete response and were alive with no evidence of disease at the last follow-up (median 45.5 months). Our findings support the usefulness of the UV signature in discriminating DM/UM vs UPS. Furthermore, we present evidence suggesting that patients with DM/UM and UV signatures can benefit from immune checkpoint inhibitor therapy.

Discovering cryptic splice mutations in cancers via a deep neural network framework.

Somatic mutations can disrupt splicing regulatory elements and have dramatic effects on cancer genes, yet the functional consequences of mutations located in extended splice regions is difficult to predict. Here, we use a deep neural network (SpliceAI) to characterize the landscape of splice-altering mutations in cancer. In our in-house series of 401 liver cancers, SpliceAI uncovers 1244 cryptic splice mutations, located outside essential splice sites, that validate at a high rate (66%) in matched RNA-seq data. We then extend the analysis to a large pan-cancer cohort of 17714 tumors, revealing>100 000 cryptic splice mutations. Taking into account these mutations increases the power of driver gene discovery, revealing 126 new candidate driver genes. It also reveals new driver mutations in known cancer genes, doubling the frequency of splice alterations in tumor suppressor genes. Mutational signature analysis suggests mutational processes that could give rise preferentially to splice mutations in each cancer type, with an enrichment of signatures related to clock-like processes and DNA repair deficiency. Altogether, this work sheds light on the causes and impact of cryptic splice mutations in cancer, and highlights the power of deep learning approaches to better annotate the functional consequences of mutations in oncology.

TaME-seq2: tagmentation-assisted multiplex PCR enrichment sequencing for viral genomic profiling

BackgroundPreviously developed TaME-seq method for deep sequencing of HPV, allowed simultaneous identification of the human papillomavirus (HPV) DNA consensus sequence, low-frequency variable sites, and chromosomal integration events. The method has been successfully validated and applied to the study of five carcinogenic high-risk (HR) HPV types (HPV16, 18, 31, 33, and 45). Here, we present TaME-seq2 with an updated laboratory workflow and bioinformatics pipeline. The HR-HPV type repertoire was expanded with HPV51, 52, and 59. As a proof-of-concept, TaME-seq2 was applied on SARS-CoV-2 positive samples showing the method’s flexibility to a broader range of viruses, both DNA and RNA.ResultsCompared to TaME-seq version 1, the bioinformatics pipeline of TaME-seq2 is approximately 40× faster. In total, 23 HPV-positive samples and seven SARS-CoV-2 clinical samples passed the threshold of 300× mean depth and were submitted to further analysis. The mean number of variable sites per 1 kb was ~ 1.5× higher in SARS-CoV-2 than in HPV-positive samples. Reproducibility and repeatability of the method were tested on a subset of samples. A viral integration breakpoint followed by a partial genomic deletion was found in within-run replicates of HPV59-positive sample. Identified viral consensus sequence in two separate runs was > 99.9% identical between replicates, differing by a couple of nucleotides identified in only one of the replicates. Conversely, the number of identical minor nucleotide variants (MNVs) differed greatly between replicates, probably caused by PCR-introduced bias. The total number of detected MNVs, calculated gene variability and mutational signature analysis, were unaffected by the sequencing run.ConclusionTaME-seq2 proved well suited for consensus sequence identification, and the detection of low-frequency viral genome variation and viral-chromosomal integrations. The repertoire of TaME-seq2 now encompasses seven HR-HPV types. Our goal is to further include all HR-HPV types in the TaME-seq2 repertoire. Moreover, with a minor modification of previously developed primers, the same method was successfully applied for the analysis of SARS-CoV-2 positive samples, implying the ease of adapting TaME-seq2 to other viruses.

A Biterm Topic Model for Sparse Mutation Data

Mutational signature analysis promises to reveal the processes that shape cancer genomes for applications in diagnosis and therapy. However, most current methods are geared toward rich mutation data that has been extracted from whole-genome or whole-exome sequencing. Methods that process sparse mutation data typically found in practice are only in the earliest stages of development. In particular, we previously developed the Mix model that clusters samples to handle data sparsity. However, the Mix model had two hyper-parameters, including the number of signatures and the number of clusters, that were very costly to learn. Therefore, we devised a new method that was several orders-of-magnitude more efficient for handling sparse data, was based on mutation co-occurrences, and imitated word co-occurrence analyses of Twitter texts. We showed that the model produced significantly improved hyper-parameter estimates that led to higher likelihoods of discovering overlooked data and had better correspondence with known signatures.

Abstract P2-23-09: Ultra-high Tumor Mutation Burden in Metastatic/Clinically Advanced Breast Cancer (MBC)

Abstract Background: High (HTMB) tumor mutation burden (TMB) defined as ≥10 mutations/megabase (Mb) identifies breast cancer patients who could benefit from pembrolizumab. The higher the TMB the greater the likelihood of benefit. The goal of this analysis was to determine the frequency and genomic landscape of MBC with ultra-high TMB (UHTMB) defined as a TMB &amp;gt; 20 mutations/Mb. Design 2,049 MBC patients (pts) underwent hybrid capture based comprehensive genomic profiling for genomic alterations (GA) in at least 324 genes including determination of TMB to guide therapy decisions using the FoundationOne®CDx assay. ER, PR and HER2 expression were abstracted from submitted pathology reports. Results: 165 of 2049 MBC (8.1%) had HTMB &amp;gt; 10 mutations/Mb, among these 45 (2.2% of all cases) had UHTMB. When compared with the 2,004 non-UHTMB MBC pts with TMB &amp;lt; 20 mutations/Mb, the 45 UHTMB pts were older (mean 64.6 yrs vs 58.2 yrs; p&amp;lt;.0001), more often had lobular histology (40.00% vs 14.5%; p&amp;lt;.0001) and ER+ disease (86.6% vs 70.0%), had higher average driver GA/tumor (9.84 vs 5.7; p&amp;lt;.0001), and less often had TNBC (13.3% vs 27.0%; p=.041) compared to non-UHTMB high cancers. There were no significant differences in ancestry. Mutation signature analysis revealed that APOBEC was predominant in UHTMB samples (82.5%) with a minor portion with an MMR signature (10%), however, MSI-H status was significantly higher in UHTMB high cases (11.6% vs 0.40%; p&amp;lt;.0001). GA more frequently identified in UHTMB cases included CDH1 (45.50% vs 14.32%; p&amp;lt;.0001), PIK3CA (81.80% vs 37.86%; p&amp;lt;.0001), CDKN2A (11.40% vs 3.19%; p=.017), ARID1A (25.00% vs 5.01%; p&amp;lt;.0001) and NF1 (20.50% vs 5.94%; p=.0014). PD-L1 (CD274) gene amplification (2.3% vs 1.3%) or protein expression by the Ventana SP142 assay (57.14% vs 51.10%) were not significantly different. Conclusions: UHTMB MBC is a rare but clinically important subset in breast cancer that could have high response rates to single agent pembrolizumab. This phenotype is driven by APOBEC mutagenesis, more often seen in ER+ lobular cancers, and have higher frequencies of MSI-high status and mutations in CDH1 and PIK3CA. Citation Format: Kristina Fanucci, maryam lustberg, Neal Fischbach, Maureen Pelletier, Abirami Sivapiragasam, Prashanth Ashok Kumar, Mansi Kallem, Natalie A. Danziger, Ethan Sokol, Smruthy Sivakumar, Dean Pavlick, Jeffrey S. Ross, Lajos Pusztai. Ultra-high Tumor Mutation Burden in Metastatic/Clinically Advanced Breast Cancer (MBC) [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P2-23-09.

Spatial genomic diversity associated with APOBEC mutagenesis in squamous cell carcinoma arising from ovarian teratoma.

Although the gross and microscopic features of squamous cell carcinoma arising from ovarian mature cystic teratoma (MCT-SCC) vary from case to case, the spatial spreading of genomic alterations within the tumor remains unclear. To clarify the spatial genomic diversity in MCT-SCCs, we performed whole-exome sequencing by collecting 16 samples from histologically different parts of two MCT-SCCs. Both cases showed histological diversity within the tumors (case 1: nonkeratinizing and keratinizing SCC and case 2: nonkeratinizing SCC and anaplastic carcinoma) and had different somatic mutation profiles by histological findings. Mutation signature analysis revealed a significantly enriched apolipoprotein B mRNA editing enzyme catalytic subunit (APOBEC) signature at all sites. Intriguingly, the spread of genomic alterations within the tumor and the clonal evolution patterns from nonmalignant epithelium to cancer sites differed between cases. TP53 mutation and copy number alterations were widespread at all sites, including the nonmalignant epithelium, in case 1. Keratinizing and nonkeratinizing SCCs were differentiated by the occurrence of unique somatic mutations from a common ancestral clone. In contrast, the nonmalignant epithelium showed almost no somatic mutations in case 2. TP53 mutation and the copy number alteration similarities were observed only in nonkeratinizing SCC samples. Nonkeratinizing SCC and anaplastic carcinoma shared almost no somatic mutations, suggesting that each locally and independently arose in the MCT. We demonstrated that two MCT-SCCs with different histologic findings were highly heterogeneous tumors with clearly different clones associated with APOBEC-mediated mutagenesis, suggesting the importance of evaluating intratumor histological and genetic heterogeneity among multiple sites of MCT-SCC.

Comprehensive genomic profiling of upper tract urothelial carcinoma and urothelial carcinoma of the bladder identifies distinct molecular characterizations with potential implications for targeted therapy & immunotherapy.

Despite the genomic landscape of urothelial carcinomas (UC) patients, especially those with UC of bladder (UCB), has been comprehensively delineated and associated with pathogenetic mechanisms and treatment preferences, the genomic characterization of upper tract UC (UTUC) has yet to be fully elucidated. A total of 131 Chinese UTUC (74 renal pelvis & 57 ureter) and 118 UCB patients were enrolled in the present study, and targeted next-generation sequencing (NGS) of 618 cancer-associated genes were conducted to exhibit the profile of somatic and germline alterations. The COSMIC database, including 30 mutational signatures, were utilized to evaluate the mutational spectrums. Moreover, TCGA-UCB, MSKCC-UCB, and MSKCC-UTUC datasets were retrieved for preforming genomic alterations (GAs) comparison analysis between Western and Chinese UC patients. In our cohort, 93.98% and 56.63% of UC patients were identified with oncogenic and actionable somatic alterations, respectively. Meanwhile, 11.24% of Chinese UC patients (of 14.50% and 7.63% of UTUC and UCB cases, respectively) were identified to harbor a total of 32 pathogenic/likely-pathogenic germline variants in 22 genes, with DNA damage repair (DDR)-associated BRCA1 (1.20%) and CHEK2 (1.20%) being the most prevalent. Chinese UTUC and UCB patients possessed distinct somatic genomic characteristics, especially with significantly different prevalence in KMT2D/C/A, GNAQ, ERCC2, RB1, and PPM1D. In addition, we also found notable differences in the prevalence of ELF3, TP53, PMS2, and FAT4 between renal pelvis and ureter carcinomas. Moreover, 22.90% and 33.90% of UTUC and UCB patients, respectively, had at least one deleterious/likely deleterious alteration in DDR related genes/pathways. Subsequently, mutational signature analysis revealed that UC patients with mutational signature 22, irrespective of UTUC or UCB, consistently had the markedly higher level of tumor mutational burden (TMB), which was proved to be positively correlated with the objective complete/partial response rate in the IMvigor210 cohort. By comparison, Chinese and Western UTUC patients also differed regrading GAs in oncogenic-related genes/pathways, especially in TP53, RTK/RAS, and PI3K pathways; besides, more alterations in WNT pathway but less TP53, RTK/RAS, HIPPO, and PI3K pathways were identified in Chinese UCB. The in-depth analysis of genomic mutational landscapes revealed distinct pathogenetic mechanisms between Chinese UTUC and UCB, and specific genomic characterizations could identify high risk population of UTUC/UCB and provided information regarding the selection of alternative therapeutic regimens.

Mutational Landscape of Normal Human Skin: Clues to Understanding Early-Stage Carcinogenesis in Keratinocyte Neoplasia

Normal skin contains numerous clones carrying cancer driver mutations. However, the mutational landscape of normal skin and its clonal relationship with skin cancer requires further elucidation. The aim of our study was to investigate the mutational landscape of normal human skin. We performed whole-exome sequencing using physiologically normal skin tissues and the matched peripheral blood (n= 39) and adjacent-matched skin cancers from a subset of patients (n= 10). Exposed skin harbored a median of 530 mutations (10.4/mb, range= 51-2,947), whereas nonexposed skin majorly exhibited significantly fewer mutations (median= 13, 0.25/mb, range= 1-166). Patient age was significantly correlated with the mutational burden. Mutations in six driver genes (NOTCH1, FAT1, TP53, PPM1D, KMT2D, and ASXL1) were identified. De novo mutational signature analysis identified a single signature with components of UV- and aging-related signatures. Normal skin harbored only three instances of copy-neutral loss of heterozygosity in 9q (n= 2) and 6q (n= 1). The mutational burden of normal skin was not correlated with that of matched skin cancers, and no protein-coding mutations were shared. In conclusion, we revealed the mutational landscape of normal skin, highlighting the role of driver genes in the malignant progression of normal skin.

Comprehensive analysis of mutational signatures reveals distinct patterns and molecular processes across 27 pediatric cancers

Analysis of mutational signatures can reveal underlying molecular mechanisms of the processes that have imprinted the somatic mutations found in cancer genomes. Here, we analyze single base substitutions and small insertions and deletions in pediatric cancers encompassing 785 whole-genome sequenced tumors from 27 molecularly defined cancer subtypes. We identified only a small number of mutational signatures active in pediatric cancers, compared with previously analyzed adult cancers. Further, we report a significant difference in the proportion of pediatric tumors showing homologous recombination repair defect signatures compared with previous analyses. In pediatric leukemias, we identified an indel signature, not previously reported, characterized by long insertions in nonrepeat regions, affecting mainly intronic and intergenic regions, but also exons of known cancer genes. We provide a systematic overview of COSMIC v.3 mutational signatures active across pediatric cancers, which is highly relevant for understanding tumor biology and enabling future research in defining biomarkers of treatment response.