Citations

Background: Rare, functionally deleterious genetic variants have been implicated as having substantially larger effect sizes than common variants, potentially accounting for much of the missing heritability in lung cancer (LC). Over 25 years, the Genetic Epidemiology of LC Consortium (GELCC) has curated an invaluable resource of specimens and data from individuals with a strong family history of LC. Methods: Whole-genome and exome sequencing data were analyzed in 129 high-risk familial LC (FLC) families, defined as having two first-degree relatives or three or more second-degree relatives affected by LC (177 affected FLC and 309 unaffected relatives). The discovery focused on rare (allele frequency < 1% in gnomAD global population) and deleterious variations (missense, stop-gained, frameshift) enriched in FLC. Promising candidates were further external validated in 2,408 sporadic lung cancer (SLC) cases and 885 population controls from the International Lung Cancer Consortium. Results: We identified 168 variants with moderate-to-large effects on FLC risk, 100 of which were validated for SLC risk (Table 1 Table 1. Top 20 candidates Gene Rare (MAF < 1%), deleterious variants No. of variant carrier counts Odds Ratio, LC vs Control FLC | Relative | SLC | Control FLC+SLC SLC Oncogene ERBB3 / HER3 p.A1131T 3 | 1 | 2 | 0 35 15 JAK p.V651M 2 | 3 | 7 | 0 62 51 NOTCH3 p.R1560P 3 | 1 | 4 | 0 48 29 ROS1 p.N785X 6 | 4 | 10 | 0 110 74 USP6 p.R522 fs del 3 | 6 | 6 | 0 62 45 Tumor suppressor gene ATM p.L2332P 3 | 2 | 8 | 0 76 59 LZTS2 p.K458R 4 | 4 | 18 | 0 152 133 NPIPB13 p.V270 fs del 13 | 14 | 10 | 0 159 74 PARK2 / PRKN p.P153R 3 | 3 | 8 | 0 76 59 RB1 p.A525G 3 | 0 | 6 | 0 62 45 SLX4 p.S1716T 3 | 2 | 9 | 0 83 66 SYNE1 p.M6566I 2 | 0 | 7 | 0 62 52 TGFBRAP1 p.T301R 5 | 0 | 11 | 0 110 82 WNK1 p.K583 fs del 4 | 0 | 23 | 0 187 171 Other cancer-associated gene CFTR p.R74W 4 | 2 | 4 | 0 55 30 FLG p.R3404 fs del 7 | 2 | 4 | 0 69 30 LTN1 p.R349H 6 | 1 | 10 | 0 110 74 MLNR p.Q334 fs del 3 | 2 | 19 | 0 152 134 SLC39A11 p.R38Q 4 | 2 | 13 | 0 117 96 TTYH2 p.R493C 5 | 4 | 14 | 0 131 104 Dose effect of the top candidates No. candidates 0 variant allele 67 | 194 | 1911 | 834 reference reference 1 variant allele 50 | 66 | 363 | 50 3.5 (2.6 - 4.7) 3.2 (2.3 - 4.3) 2 variant alleles 15 | 18 | 62 | 1 32 (4.5 - 230) 27 (3.8 - 195) 3+ variant alleles 45 | 31 | 72 | 0 98 (6 - 1506) 62 (3.9 -1012) lists 20 top hits). Among these, 15 variants mapped to our previously identified lung cancer linkage locus at 6q23-25 (ROS1, PARK2, SYNE1). Others were enriched in oncogenes (ERBB3, JAK1, MET), tumor suppressor genes (ATM, BRCA2, RB1), and genes involved in the extracellular matrix (COL6A3, FLG, MUC5B). Importantly, individuals carrying more than two combinations of these variants exhibited exceptionally strong dose effects, with odds ratios (OR) exceeding 27. Conclusion: Our findings underscore the significant role of rare, high-penetrance variants in FLC etiology and highlight promising targets for early detection and personalized treatment. Future in-depth mechanistic studies are planned to evaluate the pathogenic effects of these specific alleles. Citation Format: Yanhong Liu, Claudio Pikielny, Xiangjun Xiao, Bo Peng, Yafang Li, Jinyoung Byun, Chao Cheng, Dakai Zhu, Spiridon Tsavachidis, Colette Gaba, Elena Kupert, Ellen L. Goode, Erin L. Crawford, Kristen Purrington, Marshall Anderson, Michael Cole, Paul Brennan, Geoffrey Liu, James McKay, John K. Field, Rayjean J. Hung, David C. Christiani, Diptasri Mandal, James C. Willey, Ann Schwartz, Joan Bailey-Wilson, Susan M. Pinney, Christopher I. Amos. High-penetrance rare variants underlying familial lung cancer risk: Insights from the Genetic Epidemiology of Lung Cancer Consortium [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3815.

Abstract Background: Ovarian high-grade serous carcinomas (HGSC) comprise four distinct molecular subtypes based on mRNA expression patterns, with differential survival. Understanding risk factor associations is important to elucidate the etiology of HGSC. We investigated associations between different epidemiologic risk factors and HGSC molecular subtypes. Methods: We pooled data from 11 case–control studies with epidemiologic and tumor gene expression data from custom NanoString CodeSets developed through a collaboration within the Ovarian Tumor Tissue Analysis consortium. The PrOTYPE-validated NanoString-based 55-gene classifier was used to assign HGSC gene expression subtypes. We examined associations between epidemiologic factors and HGSC subtypes in 2,070 cases and 16,633 controls using multivariable-adjusted polytomous regression models. Results: Among the 2,070 HGSC cases, 556 (27%) were classified as C1.MES, 340 (16%) as C5.PRO, 538 (26%) as C2.IMM, and 636 (31%) as C4.DIF. The key factors, including oral contraceptive use, parity, breastfeeding, and family history of ovarian cancer, were similarly associated with all subtypes. Heterogeneity was observed for several factors. Former smoking [OR = 1.25; 95% confidence interval (CI) = 1.03, 1.51] and genital powder use (OR = 1.42; 95% CI = 1.08, 1.86) were uniquely associated with C2.IMM. History of endometriosis was associated with C5.PRO (OR = 1.46; 95% CI = 0.98, 2.16) and C4.DIF (OR = 1.27; 95% CI = 0.94, 1.71) only. Family history of breast cancer (OR = 1.44; 95% CI = 1.16, 1.78) and current smoking (OR = 1.40; 95% CI = 1.11, 1.76) were associated with C4.DIF only. Conclusions: This study observed heterogeneous associations of epidemiologic and modifiable factors with HGSC molecular subtypes. Impact: The different patterns of associations may provide key information about the etiology of the four subtypes.

Abstract Motivation Multiplex immunofluorescence (MxIF) enables the quantification of multiple protein markers at a single-cell level while preserving spatial information, offering a powerful tool for studying tissue microenvironments. However, the flexibility in MxIF panel design poses challenges in standardizing cell phenotyping. Results We present SpaFlow, an efficient, customizable pipeline for unsupervised clustering and classification of MxIF data, implemented using Nextflow. SpaFlow performs quality control, clustering, and postclustering analysis on segmented and quantified MxIF data, facilitating reproducible and scalable analyses across various computing platforms. The SpaFlow pipeline integrates three clustering and classification packages—Seurat, SCIMAP, and CELESTA—each providing unique methodologies for identifying cell types based on phenotypic markers. A novel “meta-clustering” approach condenses clusters across multiple regions of interest into common meta-clusters, streamlining the cell-type identification process in large datasets. SpaFlow’s robust quality control steps, including signal summation and cell density filtering, mitigate artifacts that may impact clustering accuracy. We demonstrate the utility of SpaFlow in a case study involving 297 ovarian tumor cores, where SpaFlow successfully identified biologically meaningful cell populations, including tumor-infiltrating lymphocytes, efficiently and rapidly. Additionally, SpaFlow’s reproducibility is validated using serial tonsil sections, confirming its capability to consistently identify distinctive cell populations across matched ROIs. Availability and implementation SpaFlow is freely available with detailed documentation and examples at https://github.com/dimi-lab/SpaFlow.

Abstract Multiplexed Immunofluorescence (MxIF) enables detailed immune cell phenotyping, providing critical insights into cell behavior within the tumor immune microenvironment (TIME). However, signal integrity can be compromised due to the complex cyclic staining processes inherent to MxIF. Hematoxylin and Eosin (H&E) staining, on the other hand, offers complementary information through its depiction of cell morphology and texture patterns and is often visually cross-referenced with MxIF in clinical settings. In this study, we proposed a novel framework to align H&E and MxIF images for precise cross-modal cell feature validation. Using cell detection outputs from each modality as anchors, we formulated the multimodal image registration problem as point set alignment. Coherent Point Drift (CPD) is employed for initial alignment, followed by Graph Matching (GM) for refinement. Evaluations on ovarian cancer tissue microarrays (TMAs) demonstrate that our method achieves high alignment accuracy, enabling reliable validation of cell-level features across modalities for both restained and serial sections. Our results indicate that restained H&E enhances confidence in findings derived from MxIF. Additionally, we demonstrated the feasibility of generating high-quality virtual H&E images from MxIF data when restained H&E is unavailable, offering a viable alternative for integrated multimodal analysis.

BACKGROUND Despite an overall poor prognosis, about 15% of patients with advanced-stage tubo-ovarian high-grade serous carcinoma (HGSC) survive 10 or more years after standard treatment. METHODS We evaluated the tumor microenvironment of this exceptional, understudied group using a large international cohort enriched for long-term survivors (LTS; 10+ years; n = 374) compared with mid-term (MTS; 5–7.99 years; n = 433) and short-term survivors (STS; 2–4.99 years; n = 416). Primary tumor samples were immunostained and scored for intraepithelial and intrastromal densities of 10 immune-cell subsets (including T cells, B cells, plasma cells, myeloid cells, PD-1 + cells, and PD-L1 + cells) and epithelial content. RESULTS Positive associations with LTS compared with STS were seen for 9 of 10 immune-cell subsets. In particular, the combination of intraepithelial CD8 + T cells and intrastromal B cells showed near 5-fold increased odds of LTS compared with STS. All of these associations were stronger in tumors with high epithelial content and/or the C4/Differentiated molecular subtype, despite immune-cell densities generally being higher in tumors with low epithelial content and/or the C2/Immunoreactive molecular subtype. CONCLUSION The tumor microenvironment of HGSC LTS is distinguished by the intersection of T and B cell coinfiltration, high epithelial content, and C4/differentiated molecular subtype, features which may inspire new approaches to immunotherapy. FUNDING Ovarian Cancer Research Program (OCRP) of the Congressionally Directed Medical Research Program (CDMRP), U.S. Department of Defense (DOD); American Cancer Society; BC Cancer Foundation; Canada’s Networks of Centres of Excellence; Canadian Cancer Society; Canadian Institutes of Health Research; Cancer Councils of New South Wales, Victoria, Queensland, South Australia, and Tasmania, Cancer Foundation of Western Australia; Cancer Institute NSW; Cancer Research UK; Deutsche Forschungsgesellschaft; ELAN Funds of the University of Erlangen-Nuremberg; Fred C. and Katherine B. Andersen Foundation; Genome BC; German Cancer Research Center; German Federal Ministry of Education and Research, Programme of Clinical Biomedical Research; Instituto de Salud Carlos III; Mayo Foundation; Minnesota Ovarian Cancer Alliance; Ministerio de Economía y Competitividad; Medical Research Council (MRC); National Center for Advancing Translational Sciences; National Health and Medical Research Council of Australia (NHMRC); Ovarian Cancer Australia; Peter MacCallum Foundation; Sydney West Translational Cancer Research Centre; Terry Fox Research Institute; The Eve Appeal (The Oak Foundation); UK National Institute for Health Research Biomedical Research Centres at the University of Cambridge; University of Pittsburgh School of Medicine; U.S. National Cancer Institute of the National Institutes of Health; VGH & UBC Hospital Foundation; Victorian Cancer Agency.

Abstract Purpose: The purpose of this study was to evaluate RB1 expression and survival across ovarian carcinoma histotypes and how co-occurrence of BRCA1 or BRCA2 (BRCA) alterations and RB1 loss influences survival in tubo-ovarian high-grade serous carcinoma (HGSC). Experimental Design: RB1 protein expression was classified by immunohistochemistry in ovarian carcinomas of 7,436 patients from the Ovarian Tumor Tissue Analysis consortium. We examined RB1 expression and germline BRCA status in a subset of 1,134 HGSC, and related genotype to overall survival (OS), tumor-infiltrating CD8+ lymphocytes, and transcriptomic subtypes. Using CRISPR-Cas9, we deleted RB1 in HGSC cells with and without BRCA1 alterations to model co-loss with treatment response. We performed whole-genome and transcriptome data analyses on 126 patients with primary HGSC to characterize tumors with concurrent BRCA deficiency and RB1 loss. Results: RB1 loss was associated with longer OS in HGSC but with poorer prognosis in endometrioid ovarian carcinoma. Patients with HGSC harboring both RB1 loss and pathogenic germline BRCA variants had superior OS compared with patients with either alteration alone, and their median OS was three times longer than those without pathogenic BRCA variants and retained RB1 expression (9.3 vs. 3.1 years). Enhanced sensitivity to cisplatin and paclitaxel was seen in BRCA1-altered cells with RB1 knockout. Combined RB1 loss and BRCA deficiency correlated with transcriptional markers of enhanced IFN response, cell-cycle deregulation, and reduced epithelial–mesenchymal transition. CD8+ lymphocytes were most prevalent in BRCA-deficient HGSC with co-loss of RB1. Conclusions: Co-occurrence of RB1 loss and BRCA deficiency was associated with exceptionally long survival in patients with HGSC, potentially due to better treatment response and immune stimulation.

Abstract We used deep learning (DL) to predict multiple genetic biomarkers of colorectal cancer (CRC) from routine histopathology images, aiming to identify morphological changes associated with the mutational profile of tumors. A new Transformer-based prediction model was developed which can simultaneously predict multiple biomarkers from a single histopathology image. We compared the performance of the multi-target approach to conventional single-target models. We analyzed five patient cohorts from the Genetics and Epidemiology of CRC Consortium (GECCO) with colorectal cancer (N=1,385 patients). Our model was trained on an internal data set (N=739) to predict the presence of >200 genetic alterations, focusing on genes with non-silent mutations and mutational signatures, which were assessed using targeted sequencing. We used a 7-fold cross validation and validated the model on an independent, external data set (N=646). Morphological features of microsatellite instability (MSI) are detectable by DL models in histopathology images, making MSI a potential confounding factor. Thus, we also assessed whether our model can predict genetic alterations regardless of MSI status, or whether the model identifies the MSI phenotype. For the majority of biomarkers, the multi-target transformer reached a higher prediction performance as measured by the mean Area Under the Receiver Operating Characteristic (AUROC) Curve and lower standard deviation than single-target DL models. For the external validation set, the mean AUROC of the multi-target versus the single-target model was 0.78 (+/- 0.01) versus 0.72 (+/- 0.06) for BRAF mutations, 0.88 (+/- 0.01) versus 0.86 (+/- 0.03) for hypermutation status, 0.94 (+/- 0.01) versus 0.91 (+/- 0.02) for MSI, 0.86 (+/- 0.01) versus 0.80 (+/- 0.05) for RNF43 mutations and 0.72 (+/- 0.02) versus 0.69 (+/- 0.05) for TP53 mutations, respectively. However, our analysis of the individual prediction scores suggests that the model does not identify the phenotype of the alterations themselves. Instead, the model seems to determine the status of specific genetic alterations based on their correlation with MSI. Consequently, biomarkers which are associated with microsatellite stability have prediction scores negatively correlated with predicted MSI scores. Conversely, biomarkers which are associated with MSI have prediction scores positively correlated with predicted MSI scores. Our results demonstrate potential benefits of multi-target DL models in improving the predictive power and efficacy of histology-based biomarker extraction compared to single-target DL models. However, our results crucially highlight the importance of considering potential confounders, such as MSI status in CRC, and emphasize the need for extensive analysis beyond AUROC in studies focusing on DL-based biomarker detection. Citation Format: Marco Gustav, Marko van Treeck, Zunamys I. Carrero, Chiara M. Loeffler, Nic G. Reitsam, Bruno Märkl, Asier Rabasco Meneghetti, Lisa A. Boardman, Amy J. French, Ellen L. Goode, Andrea Gsur, Stefanie Brezina, Marc J. Gunter, Neil Murphy, Paul Limburg, Stephen Thibodeau, Sebastian Foersch, Robert Steinfelder, Tabitha Harrison, Ulrike Peters, Amanda Phipps, Jakob N. Kather. Assessing microsatellite instability dominance in colorectal cancer phenotype: A multi-study initiative using multi-target transformers for genomic biomarker prediction [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 4924.

Abstract Introduction: Identifying dysregulated gene expression in ovarian cancers has been limited by a deficit of available normal tissues. Here, we generated the largest set of high-grade serous ovarian cancer (HGSOC) tumors with normal precursor tissues for transcriptome analyses to identify lncRNA expression patterns, provide insight into their cellular contexts and functions, and assess their relevance to HGSOC pathogenesis and genetic susceptibility. Methods: We performed RNA sequencing on 220 primary HGSOCs and 116 benign epithelia (micro-dissected fallopian tube, ovarian surface, and inclusion cyst epithelia), and combined samples with 428 HGSOCs from TCGA, 60 HGSOCs from a prior study, and 180 bulk ovary tissues from GTEx. Reads were processed with a uniform bioinformatic and quality control pipeline, combined and batch corrected. Normalized TPM expression values were compared using the Wilcoxon rank sum test with >2-fold change and FDR <1% considered statistically significant. We computed weighted gene co-expression networks for each tissue type and tested for enrichment of differentially expressed genes (DEGs) and gene ontologies within co-expression modules. Results: Transcriptomes included 27,700 expressed genes (8,202 lncRNAs) across 706 HGSOCs, 180 bulk ovary, and 88 ovarian epithelia tissues. Most genes were expressed in all tissues, though 5% each showed HGSOC- or normal tissue-specific expression which comprised ≥50% lncRNA. Comparing HGSOCs to normal tissues identified 11,804 DEGs with 4,522 lncRNAs (DElncRNA) of which ~50% were tissue-specific. DEGs included MUC16 and multiple GWAS/TWAS implicated susceptibility genes including RAD51, BRIP1, BNC2, TIPARP-AS1, PRC1, KANSL1, ANKLE1, CHMP4C, ESRP2, and CCNE1. The most highly expressed DElncRNA in HGSOC were upregulated RMRP (P=1.4 × 10−39), SNHG1 (P=3.0 × 10−27), and HAGLR (P=2.0 × 10−24) at the HOXD risk locus. Lower proportions of lncRNA were seen in the HGSOC co-expression modules compared to those in normal tissues and less (3% vs. 15%) HGSOC modules were considered strongly lncRNA regulated (>50% lncRNA in module). Tissue-specific DElncRNA tended toward upregulation in HGSOC compared to ovarian epithelia with enriched modules associated with cell cycle regulation (P=1.5 × 10−38); and toward downregulation in HGSOC compared to bulk ovary with enriched modules associated with chemokine signaling/response (P=1.7 × 10−113). Conclusions: The lower proportion of lncRNA involved in HGSOC co-expression networks indicates a marked loss of regulation by lncRNA. Preliminary interrogation of HGSOC-dysregulated lncRNA expression revealed tissue-specific differences that highlight unique biological pathways in precursor epithelia and the ovarian microenvironment that contribute to HGSOC pathogenesis. Our results provide additional support for previously nominated risk genes. eQTL analysis has recently been completed and integration with GWAS are underway to elucidate lncRNA and lncRNA-regulated susceptibility genes and will be presented at the conference. Citation Format: Brett M. Reid, Ann Chen, Zhihua Chen, Florian A. Karreth, Peter Kanetsky, Jennifer B. Permuth, Ozlen Saglam, Jamie K. Teer, Xiaoqing Yu, Simon Gayther, Ellen Goode, Paul Pharoah, Thomas A. Sellers, Kate Lawrenson, Jonathan Tyrer. Patterns of lncRNA-regulated gene expression in high-grade serous ovarian carcinomas [abstract]. In: Proceedings of the AACR Special Conference on Ovarian Cancer; 2023 Oct 5-7; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(5 Suppl_2):Abstract nr A004.

Abstract Background: Women with an inherited pathogenic variant in BRCA1 or BRCA2 have a greatly increased risk of developing ovarian cancer, but the importance of behavioral factors is less clear. We used a case-only design to compare the magnitude of associations with established reproductive, hormonal, and lifestyle risk factors between BRCA mutation carriers and noncarriers. Methods: We pooled data from five studies from the Ovarian Cancer Association Consortium including 637 BRCA carriers and 4,289 noncarriers. Covariate-adjusted generalized linear mixed models were used to estimate interaction risk ratios (IRR) and 95% confidence intervals (CI), with BRCA (carrier vs. noncarrier) as the response variable. Results: IRRs were above 1.0 for known protective factors including ever being pregnant (IRR = 1.29, 95% CI; 1.00–1.67) and ever using the oral contraceptive pill (1.30, 95% CI; 1.07–1.60), suggesting the protective effects of these factors may be reduced in carriers compared with noncarriers. Conversely, the IRRs for risk factors including endometriosis and menopausal hormone therapy were below 1.0, suggesting weaker positive associations among BRCA carriers. In contrast, associations with lifestyle factors including smoking, physical inactivity, body mass index, and aspirin use did not appear to differ by BRCA status. Conclusions: Our results suggest that associations with hormonal and reproductive factors are generally weaker for those with a pathogenic BRCA variant than those without, while associations with modifiable lifestyle factors are similar for carriers and noncarriers. Impact: Advice to maintain a healthy weight, be physically active, and refrain from smoking will therefore benefit BRCA carriers as well as noncarriers.

Abstract Background: The genotoxin colibactin causes a tumor single-base substitution (SBS) mutational signature, SBS88. It is unknown whether epidemiologic factors’ association with colorectal cancer risk and survival differs by SBS88. Methods: Within the Genetic Epidemiology of Colorectal Cancer Consortium and Colon Cancer Family Registry, we measured SBS88 in 4,308 microsatellite stable/microsatellite instability low tumors. Associations of epidemiologic factors with colorectal cancer risk by SBS88 were assessed using multinomial regression (N = 4,308 cases, 14,192 controls; cohort-only cases N = 1,911), and with colorectal cancer–specific survival using Cox proportional hazards regression (N = 3,465 cases). Results: 392 (9%) tumors were SBS88 positive. Among all cases, the highest quartile of fruit intake was associated with lower risk of SBS88-positive colorectal cancer than SBS88-negative colorectal cancer [odds ratio (OR) = 0.53, 95% confidence interval (CI) 0.37-0.76; OR = 0.75, 95% CI 0.66–0.85, respectively, Pheterogeneity = 0.047]. Among cohort studies, associations of body mass index (BMI), alcohol, and fruit intake with colorectal cancer risk differed by SBS88. BMI ≥30 kg/m2 was associated with worse colorectal cancer–specific survival among those SBS88-positive [hazard ratio (HR) = 3.40, 95% CI 1.47–7.84], but not among those SBS88-negative (HR = 0.97, 95% CI 0.78–1.21, Pheterogeneity = 0.066). Conclusions: Most epidemiologic factors did not differ by SBS88 for colorectal cancer risk or survival. Higher BMI may be associated with worse colorectal cancer–specific survival among those SBS88-positive; however, validation is needed in samples with whole-genome or whole-exome sequencing available. Impact: This study highlights the importance of identification of tumor phenotypes related to colorectal cancer and understanding potential heterogeneity for risk and survival.