Search

Abstract Spatial transcriptomics (ST) assays represent a revolution in how the architecture of tissues is studied by allowing for the exploration of cells in their spatial context. A common element in the analysis is delineating tissue domains or “niches” followed by detecting differentially expressed genes to infer the biological identity of the tissue domains or cell types. However, many studies approach differential expression analysis by using statistical approaches often applied in the analysis of non-spatial scRNA data (e.g., two-sample t-tests, Wilcoxon’s rank sum test), hence neglecting the spatial dependency observed in ST data. In this study, we show that applying linear mixed models with spatial correlation structures using spatial random effects effectively accounts for the spatial autocorrelation and reduces inflation of type-I error rate observed in non-spatial based differential expression testing. We also show that spatial linear models with an exponential correlation structure provide a better fit to the ST data as compared to non-spatial models, particularly for spatially resolved technologies that quantify expression at finer scales (i.e., single-cell resolution).

Summary: Background: Insomnia is the most common sleep disorder in patients with epithelial ovarian cancer (EOC). We investigated the causal association between genetically predicted insomnia and EOC risk and survival through a two-sample Mendelian randomization (MR) study. Methods: Insomnia was proxied using genetic variants identified in a genome-wide association study (GWAS) meta-analysis of UK Biobank and 23andMe. Using genetic associations with EOC risk and overall survival from the Ovarian Cancer Association Consortium (OCAC) GWAS in 66,450 women (over 11,000 cases with clinical follow-up), we performed Iterative Mendelian Randomization and Pleiotropy (IMRP) analysis followed by a set of sensitivity analyses. Genetic associations with survival and response to treatment in ovarian cancer study of The Cancer Genome Atlas (TCGA) were estimated controlling for chemotherapy and clinical factors. Findings: Insomnia was associated with higher risk of endometrioid EOC (OR = 1.60, 95% CI 1.05–2.45) and lower risk of high-grade serous EOC (HGSOC) and clear cell EOC (OR = 0.79 and 0.48, 95% CI 0.63–1.00 and 0.27–0.86, respectively). In survival analysis, insomnia was associated with shorter survival of invasive EOC (OR = 1.45, 95% CI 1.13–1.87) and HGSOC (OR = 1.4, 95% CI 1.04–1.89), which was attenuated after adjustment for body mass index and reproductive age. Insomnia was associated with reduced survival in TCGA HGSOC cases who received standard chemotherapy (OR = 2.48, 95% CI 1.13–5.42), but was attenuated after adjustment for clinical factors. Interpretation: This study supports the impact of insomnia on EOC risk and survival, suggesting treatments targeting insomnia could be pivotal for prevention and improving patient survival. Funding: National Institutes of Health, National Cancer Institute. Full funding details are provided in acknowledgments.

Abstract Osteosarcoma is the most common bone sarcoma in children and young adults. While universally delivered, chemotherapy only benefits roughly half of patients with localized disease. Increasingly, intratumoral heterogeneity is recognized as a source of therapeutic resistance. In this study, we develop and evaluate an in vitro model of osteosarcoma heterogeneity based on phenotype and genotype. Cancer cell populations vary in their environment-specific growth rates and in their sensitivity to chemotherapy. We present the genotypic and phenotypic characterization of an osteosarcoma cell line panel with a focus on co-cultures of the most phenotypically divergent cell lines, 143B and SAOS2. Modest environmental (pH, glutamine) or chemical perturbations dramatically shift the success and composition of cell lines. We demonstrate that in nutrient rich culture conditions 143B outcompetes SAOS2. But, under nutrient deprivation or conventional chemotherapy, SAOS2 growth can be favored in spheroids. Importantly, when the simplest heterogeneity state is evaluated, a two-cell line coculture, perturbations that affect the faster growing cell line have only a modest effect on final spheroid size. Thus the only evaluated therapies to eliminate the spheroids were by switching therapies from a first strike to a second strike. This extensively characterized, widely available system, can be modeled and scaled to allow for improved strategies to anticipate resistance in osteosarcoma due to heterogeneity.

Abstract Background Aspirin use has been associated with reduced ovarian cancer risk, yet the underlying biological mechanisms are not fully understood. To gain mechanistic insights, we assessed the association between prediagnosis low and regular‐dose aspirin use and gene expression profiles in ovarian tumors. Methods RNA sequencing was performed on high‐grade serous, poorly differentiated, and high‐grade endometrioid ovarian cancer tumors from the Nurses' Health Study (NHS), NHSII, and New England Case–Control Study (n = 92 cases for low, 153 cases for regular‐dose aspirin). Linear regression identified differentially expressed genes associated with aspirin use, adjusted for birth decade and cohort. False discovery rates (FDR) were used to account for multiple testing and gene set enrichment analysis was used to identify biological pathways. Results No individual genes were significantly differentially expressed in ovarian tumors in low or regular‐dose aspirin users accounting for multiple comparisons. However, current versus never use of low‐dose aspirin was associated with upregulation of immune pathways (e.g., allograft rejection, FDR = 5.8 × 10−10; interferon‐gamma response, FDR = 2.0 × 10−4) and downregulation of estrogen response pathways (e.g., estrogen response late, FDR = 4.9 × 10−8). Ovarian tumors from current regular aspirin users versus never users were also associated with upregulation in interferon pathways (FDR

Abstract Genomic and transcriptomic data have been generated across a wide range of prostate cancer (PCa) study cohorts. These data can be used to better characterize the molecular features associated with clinical outcomes and to test hypotheses across multiple, independent patient cohorts. In addition, derived features, such as estimates of cell composition, risk scores, and androgen receptor (AR) scores, can be used to develop novel hypotheses leveraging existing multi-omic datasets. The full potential of such data is yet to be realized as independent datasets exist in different repositories, have been processed using different pipelines, and derived and clinical features are often not provided or not standardized. Here, we present the curatedPCaData R package, a harmonized data resource representing >2900 primary tumor, >200 normal tissue, and >500 metastatic PCa samples across 19 datasets processed using standardized pipelines with updated gene annotations. We show that meta-analysis across harmonized studies has great potential for robust and clinically meaningful insights. curatedPCaData is an open and accessible community resource with code made available for reproducibility.

EWSR1 fusions are highly promiscuous and are associated with unique malignancies, clinical phenotypes, and molecular subtypes. However, rare fusion partners (RFP) of EWSR1 has not been well described. Here, we conducted a cross-sectional, retrospective study of 1,140 unique tumors harboring EWSR1 fusions. We identified 64 unique fusion partners. RFPs were identified more often in adults than children. Alterations in cell cycle control and DNA damage response genes as driving the differences between fusion partners. Potentially clinically actionable genomic variants were more prevalent in tumors harboring RFP than common fusions. While the data presented here is limited, tumors harboring RFP of EWSR1 may represent molecularly distinct entities and may benefit from further molecular testing to identify targeted therapeutic options.

Abstract Background Patients with relapsed and refractory solid and central nervous system (CNS) tumors have poor outcomes and need novel therapeutic options. Vincristine, irinotecan, and temozolomide (VIT) is a common chemotherapy regimen in relapsed pediatric tumors with an established toxicity profile. Metformin shows preclinical anti‐cancer activity through multiple pathways. Methods The objective of this Phase I trial was to establish the maximum tolerated dose (MTD) and recommended Phase II dose (RP2D) of metformin in combination with VIT in children with relapsed and refractory solid and CNS tumors. A 3 + 3 design was used to test the addition of metformin at five dose levels (666, 999, 1333, 1666, and 2000 mg/m2/day). Therapy toxicity, pharmacokinetics, and radiologic response to treatment were evaluated. Results Twenty‐six patients (median age 13 years, range 2–18 years) were enrolled with 22 evaluable for toxicity. The most common diagnoses were Ewing sarcoma (n = 8), rhabdomyosarcoma (n = 3) and atypical teratoid/rhabdoid tumor (n = 3). The MTD was exceeded at Dose Level 5 due to two dose‐limiting toxicities; both were Grade 3 diarrhea requiring prolonged hospitalization and intravenous fluids. The MTD was not determined due to study closure with less than six patients enrolled at Dose Level 4. Frequently observed toxicities were gastrointestinal (most notably diarrhea) and hematologic. Amongst 16 patients evaluable for best overall response, there was one complete response (Ewing sarcoma), three partial responses (Ewing sarcoma, glioblastoma multiforme, and alveolar rhabdomyosarcoma), and five patients with stable disease. Conclusions The MTD of VIT with metformin was not determined due to premature study closure. We recommend an RP2D of Dose Level 4, 1666 mg/m2/day. Radiographic responses were seen in multiple tumor types. Further evaluation for efficacy could be investigated in a Phase II trial.

Multiple Myeloma (MM) is an incurable plasma cell malignancy often treated by autologous stem cell transplant (ASCT). Clinical response to ASCT has been associated with DNA repair efficiency. Here we interrogated the role of the base excision DNA repair (BER) pathway in MM response to ASCT. Across 450 clinical samples and six disease stages, expression levels of genes in the BER pathway were found to be highly upregulated during the development of MM. In a separate cohort of 559 patients with MM treated with ASCT, expression of BER pathway members MPG and PARP3 was positively associated with overall survival (OS) while expression of PARP1, POLD1, and POLD2 was negatively associated with OS. In a validation cohort of 356 patients with MM treated with ASCT, PARP1 and POLD2 findings were replicated. In patients with MM who never received ASCT (n=319), PARP1 and POLD2 were not associated with OS, suggesting that the prognostic effect of these genes may be treatment-dependent. In preclinical models of MM, synergy was observed in anti-tumor activity when poly (ADPribose) polymerase (PARP) inhibitors (olaparib, talazoparib) were used in combination with melphalan. The negative prognosis associated with PARP1 and POLD2 expression along with the apparent melphalan-sensitizing effect of PARP inhibition may suggest this pathway as a potential biomarker in patients with MM in the setting of ASCT. Further understanding of the role of the BER pathway in MM is vital to improve therapeutic strategies related to ASCT.

Abstract Background Greater ovulatory years is associated with increased ovarian cancer risk. Although ovulation leads to an acute pro-inflammatory local environment, how long-term exposure to ovulation impacts ovarian carcinogenesis is not fully understood. Thus, we examined the association between gene expression profiles of ovarian tumors and lifetime ovulatory years to enhance understanding of associated biological pathways. Methods RNA sequencing data was generated on 234 invasive ovarian cancer tumors that were high-grade serous, poorly differentiated, or high-grade endometrioid from the Nurses’ Health Study (NHS), NHSII, and the New England Case Control Study. We used linear regression to identify differentially expressed genes by estimated ovulatory years, adjusted for birth decade and cohort, overall and stratified by menopausal status at diagnosis. We used false discovery rates (FDR) to account for multiple testing. Gene set enrichment analysis (GSEA) with Cancer Hallmarks, KEGG, and Reactome databases was used to identify biological pathways associated with ovulatory years. Results No individual genes were significantly differentially expressed by ovulatory years (FDR > 0.19). However, GSEA identified several pathways that were significantly associated with ovulatory years, including downregulation of pathways related to inflammation and proliferation (FDR

ObjectiveOptimal debulking with no macroscopic residual disease strongly predicts ovarian cancer survival. The ability to predict likelihood of optimal debulking, which may be partially dependent on tumor biology, could inform clinical decision-making regarding use of neoadjuvant chemotherapy. Thus, we developed a prediction model including epidemiological factors and tumor markers of residual disease after primary debulking surgery.MethodsUnivariate analyses examined associations of 11 pre-diagnosis epidemiologic factors (n=593) and 24 tumor markers (n=204) with debulking status among incident, high-stage, epithelial ovarian cancer cases from the Nurses’ Health Studies and New England Case Control study. We used Bayesian model averaging (BMA) to develop prediction models of optimal debulking with 5x5-fold cross-validation and calculated the area under the curve (AUC).ResultsCurrent aspirin use was associated with lower odds of optimal debulking compared to never use (OR=0.52, 95%CI=0.31-0.86) and two tissue markers, ADRB2 (OR=2.21, 95%CI=1.23-4.41) and FAP (OR=1.91, 95%CI=1.24-3.05) were associated with increased odds of optimal debulking. The BMA selected aspirin, parity, and menopausal status as the epidemiologic/clinical predictors with the posterior effect probability ≥20%. While the prediction model with epidemiologic/clinical predictors had low performance (average AUC=0.49), the model adding tissue biomarkers showed improved, but weak, performance (average AUC=0.62).ConclusionsAddition of ovarian tumor tissue markers to our multivariable prediction models based on epidemiologic/clinical data slightly improved the model performance, suggesting debulking status may be in part driven by tumor characteristics. Larger studies are warranted to identify those at high risk of poor surgical outcomes informing personalized treatment.

Cell type classification is an important problem in cancer research, especially with the advent of single cell technologies. Correctly identifying cells within the tumor microenvironment can provide oncologists with a snapshot of how a patient’s immune system reacts to the tumor. Wide and deep learning (WDL) is an approach to construct a cell-classification prediction model that can learn patterns within high-dimensional data (deep) and ensure that biologically relevant features (wide) remain in the final model. In this paper, we demonstrate that regularization can prevent overfitting and adding a wide component to a neural network can result in a model with better predictive performance. In particular, we observed that a combination of dropout and ℓ2 regularization can lead to a validation loss function that does not depend on the number of training iterations and does not experience a significant decrease in prediction accuracy compared to models with ℓ1, dropout, or no regularization. Additionally, we show WDL can have superior classification accuracy when the training and testing of a model are completed data on that arise from the same cancer type but different platforms. More specifically, WDL compared to traditional deep learning models can substantially increase the overall cell type prediction accuracy (36.5 to 86.9%) and T cell subtypes (CD4: 2.4 to 59.1%, and CD8: 19.5 to 96.1%) when the models were trained using melanoma data obtained from the 10X platform and tested on basal cell carcinoma data obtained using SMART-seq. WDL obtains higher accuracy when compared to state-of-the-art cell classification algorithms CHETAH (70.36%) and SingleR (70.59%).

Abstract Background Unsupervised clustering represents one of the most widely applied methods in analysis of high-throughput ‘omics data. A variety of unsupervised model-based or parametric clustering methods and non-parametric clustering methods have been proposed for RNA-seq count data, most of which perform well for large samples, e.g. N ≥ 500. A common issue when analyzing limited samples of RNA-seq count data is that the data follows an over-dispersed distribution, and thus a Negative Binomial likelihood model is often used. Thus, we have developed a Negative Binomial model-based (NBMB) clustering approach for application to RNA-seq studies. Results We have developed a Negative Binomial Model-Based (NBMB) method to cluster samples using a stochastic version of the expectation-maximization algorithm. A simulation study involving various scenarios was completed to compare the performance of NBMB to Gaussian model-based or Gaussian mixture modeling (GMM). NBMB was also applied for the clustering of two RNA-seq studies; type 2 diabetes study (N = 96) and TCGA study of ovarian cancer (N = 295). Simulation results showed that NBMB outperforms GMM applied with different transformations in majority of scenarios with limited sample size. Additionally, we found that NBMB outperformed GMM for small clusters distance regardless of sample size. Increasing total number of genes with fixed proportion of differentially expressed genes does not change the outperformance of NBMB, but improves the overall performance of GMM. Analysis of type 2 diabetes and ovarian cancer tumor data with NBMB found good agreement with the reported disease subtypes and the gene expression patterns. This method is available in an R package on CRAN named NB.MClust. Conclusion Use of Negative Binomial model based clustering is advisable when clustering over dispersed RNA-seq count data.

Background: A major challenge in implementing personalized medicine in pediatrics is identifying appropriate drug dosages for children. The majority of drug dosing studies have been based on adult populations, often with modification of the dosing for children based on size and weight. However, the growth and development experienced by children between birth and adulthood represents a dynamically changing biological system, with implications for effective drug dosing, efficacy as well as potential drug toxicity. The purpose of this study was to apply a metabolomics approach to gain preliminary insights into the ontogeny of liver function from newborn to adolescent.Methods: Metabolites were measured in 98 post-mortem pediatric liver samples in two experiments 3 batches of samples, allowing for both technical and biological validation. After extensive quality control, imputation and normalization, non-parametric tests were used to determine which metabolite levels differ between the four age groups (AG) ranging in age from newborn to adolescent (AG1—children

Endometrial cancer is the most common invasive gynaecological cancer in developed countries. Here a meta-analysis identifies an additional nine novel endometrial cancer risk loci and eQTL analysis reveals risk variants associate with reduced expression of negative regulators of oncogenic signal transduction proteins.

Four gene expression subtypes of high-grade serous ovarian cancer (HGSC) have been previously described. In these early studies, a fraction of samples that did not fit well into the four subtype classifications were excluded. Therefore, we sought to systematically determine the concordance of transcriptomic HGSC subtypes across populations without removing any samples. We created a bioinformatics pipeline to independently cluster the five largest mRNA expression datasets using k-means and nonnegative matrix factorization (NMF). We summarized differential expression patterns to compare clusters across studies. While previous studies reported four subtypes, our cross-population comparison does not support four. Because these results contrast with previous reports, we attempted to reproduce analyses performed in those studies. Our results suggest that early results favoring four subtypes may have been driven by the inclusion of serous borderline tumors. In summary, our analysis suggests that either two or three, but not four, gene expression subtypes are most consistent across datasets.

A region on chromosome 19p13 is associated with the risk of developing ovarian and breast cancer. Here, the authors genotyped SNPs in this region in thousands of breast and ovarian cancer patients and identified SNPs associated with three genes, which were analysed with functional studies.

Systemic lupus erythematosus (SLE) is a chronic systemic autoimmune disease resulting from abnormal interactions between T and B cells. The acquisition of SLE is linked to genetic susceptibility, and diverse environmental agents can trigger disease onset in genetically susceptible individuals. However, the strongest risk factor for developing SLE is being female (9:1 female to male ratio). The female sex steroid, estradiol, working through its receptors, contributes to the gender bias in SLE although the mechanisms remain enigmatic. In a small clinical trial, monthly administration of the estrogen receptor (ERα) antagonist, ICI182,780 (fulvestrant), significantly reduced disease indicators in SLE patients. In order to identify changes that could account for improved disease status, the present study utilized fulvestrant (Faslodex) to block ERα action in cultured SLE T cells that were purified from blood samples collected from SLE patients (n = 18, median age 42 years) and healthy control females (n = 25, median age 46 years). The effects of ERα antagonism on estradiol-dependent gene expression and canonical signaling pathways were analyzed. Pathways that were significantly altered by addition of Faslodex included T helper (Th) cell differentiation, steroid receptor signaling [glucocorticoid receptor (GR), ESR1 (ERα)], ubiquitination, and sumoylation pathways. ERα protein expression was significantly lower (p 0.05) between SLE patients and controls. A previously undetected interaction between GR and ERα signaling pathways suggests posttranslational modification of steroid receptors in SLE T cells may alter ERα/GR actions and contribute to the strong gender bias of this autoimmune disorder.

Aggregating information across multiple variants in a gene or region can improve power for rare variant association testing. Power is maximized when the aggregation region contains many causal variants and few neutral variants. In this paper, we present a method for the localization of the association signal in a region using a sliding-window based approach to rare variant association testing in a region. We first introduce a novel method for analysis of rare variants, the Difference in Minor Allele Frequency test (DMAF), which allows combined analysis of common and rare variants, and makes no assumptions about the direction of effects. In whole-region analyses of simulated data with risk and protective variants, DMAF and other methods which pool data across individuals were found to outperform methods which pool data across variants. We then implement a sliding-window version of DMAF, using a step-down permutation approach to control type I error with the testing of multiple windows. In simulations, the sliding-window DMAF improved power to detect a causal sub-region, compared to applying DMAF to the whole region. Sliding-window DMAF was also effective in localizing the causal sub-region. We also applied the DMAF sliding window approach to test for an association between response to the drug gemcitabine and variants in the gene FKBP5 sequenced in 91 lymphoblastoid cell lines derived from white non-Hispanic individuals. The application of the sliding window test procedure detected an association in a sub-region spanning an exon and two introns, when rare and common variants were analyzed together.

New technologies, such as multiplex immunofluorescence microscopy (mIF), are being developed and used for the assessment and visualization of the tumor immune microenvironment (TIME). These assays produce not only an estimate of the abundance of immune cells in the TIME, but also their spatial locations. However, there are currently few approaches to analyze the spatial context of the TIME. Therefore, we have developed a framework for the spatial analysis of the TIME using Ripley's K, coupled with a permutation-based framework to estimate and measure the departure from complete spatial randomness (CSR) as a measure of the interactions between immune cells. This approach was then applied to epithelial ovarian cancer (EOC) using mIF collected on intra-tumoral regions of interest (ROIs) and tissue microarrays (TMAs) from 160 high-grade serous ovarian carcinoma patients in the African American Cancer Epidemiology Study (AACES) (94 subjects on TMAs resulting in 263 tissue cores; 93 subjects with 260 ROIs; 27 subjects with both TMA and ROI data). Cox proportional hazard models were constructed to determine the association of abundance and spatial clustering of tumor-infiltrating lymphocytes (CD3+), cytotoxic T-cells (CD8+CD3+), and regulatory T-cells (CD3+FoxP3+) with overall survival. Analysis was done on TMA and ROIs, treating the TMA data as validation of the findings from the ROIs. We found that EOC patients with high abundance and low spatial clustering of tumor-infiltrating lymphocytes and T-cell subsets in their tumors had the best overall survival. Additionally, patients with EOC tumors displaying high co-occurrence of cytotoxic T-cells and regulatory T-cells had the best overall survival. Grouping women with ovarian cancer based on both cell abundance and spatial contexture showed better discrimination for survival than grouping ovarian cancer cases only by cell abundance. These findings underscore the prognostic importance of evaluating not only immune cell abundance but also the spatial contexture of the immune cells in the TIME. In conclusion, the application of this spatial analysis framework to the study of the TIME could lead to the identification of immune content and spatial architecture that could aid in the determination of patients that are likely to respond to immunotherapies.

ObjectivesCentral nervous system metastases (CNSm) secondary to endometrial cancer (EC) are rare. As a result, prognostic factors for this patient population are not well described.MethodsEC patients with CNSm were identified retrospectively from two academic centers. EC patients without CNSm (non-CNSm) were used as controls. Chi-square and Fisher's exact tests were used for analysis of categorial variables. Wilcoxon tests were used for quantitative measures. Overall survival (OS) was compared with Log-rank test. Cox proportional hazard models were used to estimate hazard ratios for OS.Results22 EC patients with CNSm and 354 non-CNSm patients were included. Compared to non-CNSm EC, the CNSm cohort was younger (58.5 vs 62.0 years, p = 0.018) with lower BMI (27.7 vs. 33.7 kg/m2, p = 0.005), and had more advanced stages (p = ≤ 0.001), grade 3 tumors (81.8% CNSm vs 25.1% non CNSm, p≤0.001) and serous histology (22.7% vs 8.5%, p = 0.010). Median survival after CNSm diagnosis was 9 months (95% CI 4, NA). CNSm was a strong poor prognostic factor (HR death 4.96, p = 0.022). Improved OS was seen with CNS as the only disease site (83m CNSm only vs 30m additional sites, p = 0.007) and less than five CNSm (49m ConclusionsCNSm is a poor prognostic factor in EC, however, low volume disease with aggressive treatment may result in more favorable survival outcomes.