miR-122 and miR-21 are Stable Components of miRNA Signatures of Early Lung Cancer after Validation in Three Independent Cohorts

Several panels of circulating miRNAs have been reported as potential biomarkers of early lung cancer, yet the overlap of components between different panels is limited, and the universality of proposed biomarkers has been minimal across proposed panels. To assess the stability of the diagnostic potential of plasma miRNA signature of early lung cancer among different cohorts, a panel of 24 miRNAs tested in the frame of one lung cancer screening study (MOLTEST-2013, Poland) was validated with material collected in the frame of two other screening studies (MOLTEST-BIS, Poland; and SMAC, Italy) using the same standardized analytical platform (the miRCURY LNA miRNA PCR assay). On analysis of selected miRNAs, two associated with lung cancer development, miR-122 and miR-21, repetitively differentiated healthy participants from individuals with lung cancer. Additionally, miR-144 differentiated controls from cases specifically in subcohorts with adenocarcinoma. Other tested miRNAs did not overlap in the three cohorts. Classification models based on neither a single miRNA nor multicomponent miRNA panels (24-mer and 7-mer) showed classification performance sufficient for a standalone diagnostic biomarker (AUC, 75%, 71%, and 53% in MOLTEST-2013, SMAC, and MOLTEST-BIS, respectively, in the 7-mer model). The performance of classification in the MOLTEST-BIS cohort with the lowest contribution of adenocarcinomas was increased when only this cancer type was considered (AUC, 60% in 7-mer model).

Combining Low-Dose Computer-Tomography-Based Radiomics and Serum Metabolomics for Diagnosis of Malignant Nodules in Participants of Lung Cancer Screening Studies

Radiomics is an emerging approach to support the diagnosis of pulmonary nodules detected via low-dose computed tomography lung cancer screening. Serum metabolome is a promising source of auxiliary biomarkers that could help enhance the precision of lung cancer diagnosis in CT-based screening. Thus, we aimed to verify whether the combination of these two techniques, which provides local/morphological and systemic/molecular features of disease at the same time, increases the performance of lung cancer classification models. The collected cohort consists of 1086 patients with radiomic and 246 patients with serum metabolomic evaluations. Different machine learning techniques, i.e., random forest and logistic regression were applied for each omics. Next, model predictions were combined with various integration methods to create a final model. The best single omics models were characterized by an AUC of 83% in radiomics and 60% in serum metabolomics. The model integration only slightly increased the performance of the combined model (AUC equal to 85%), which was not statistically significant. We concluded that radiomics itself has a good ability to discriminate lung cancer from benign lesions. However, additional research is needed to test whether its combination with other molecular assessments would further improve the diagnosis of screening-detected lung nodules.

Short- and long-term effects of radiation exposure at low dose and low dose rate in normal human VH10 fibroblasts

Introduction

Experimental studies complement epidemiological data on the biological effects of low doses and dose rates of ionizing radiation and help in determining the dose and dose rate effectiveness factor.

Methods

Human VH10 skin fibroblasts exposed to 25, 50, and 100 mGy of 137Cs gamma radiation at 1.6, 8, 12 mGy/h, and at a high dose rate of 23.4 Gy/h, were analyzed for radiation-induced short- and long-term effects. Two sample cohorts, i.e., discovery (n = 30) and validation (n = 12), were subjected to RNA sequencing. The pool of the results from those six experiments with shared conditions (1.6 mGy/h; 24 h), together with an earlier time point (0 h), constituted a third cohort (n = 12).

Results

The 100 mGy-exposed cells at all abovementioned dose rates, harvested at 0/24 h and 21 days after exposure, showed no strong gene expression changes. DMXL2, involved in the regulation of the NOTCH signaling pathway, presented a consistent upregulation among both the discovery and validation cohorts, and was validated by qPCR. Gene set enrichment analysis revealed that the NOTCH pathway was upregulated in the pooled cohort (p = 0.76, normalized enrichment score (NES) = 0.86). Apart from upregulated apical junction and downregulated DNA repair, few pathways were consistently changed across exposed cohorts. Concurringly, cell viability assays, performed 1, 3, and 6 days post irradiation, and colony forming assay, seeded just after exposure, did not reveal any statistically significant early effects on cell growth or survival patterns. Tendencies of increased viability (day 6) and reduced colony size (day 21) were observed at 12 mGy/h and 23.4 Gy/min. Furthermore, no long-term changes were observed in cell growth curves generated up to 70 days after exposure.

Discussion

In conclusion, low doses of gamma radiation given at low dose rates had no strong cytotoxic effects on radioresistant VH10 cells.

Pathological changes or technical artefacts? The problem of the heterogenous databases in COVID-19 CXR image analysis

BACKGROUND

When the COVID-19 pandemic commenced in 2020, scientists assisted medical specialists with diagnostic algorithm development. One scientific research area related to COVID-19 diagnosis was medical imaging and its potential to support molecular tests. Unfortunately, several systems reported high accuracy in development but did not fare well in clinical application. The reason was poor generalization, a long-standing issue in AI development. Researchers found many causes of this issue and decided to refer to them as confounders, meaning a set of artefacts and methodological errors associated with the method. We aim to contribute to this steed by highlighting an undiscussed confounder related to image resolution.

METHODS

20 216 chest X-ray images (CXR) from worldwide centres were analyzed. The CXRs were bijectively projected into the 2D domain by performing Uniform Manifold Approximation and Projection (UMAP) embedding on the radiomic features (rUMAP) or CNN-based neural features (nUMAP) from the pre-last layer of the pre-trained classification neural network. Additional 44 339 thorax CXRs were used for validation. The comprehensive analysis of the multimodality of the density distribution in rUMAP/nUMAP domains and its relation to the original image properties was used to identify the main confounders.

RESULTS

nUMAP revealed a hidden bias of neural networks towards the image resolution, which the regular up-sampling procedure cannot compensate for. The issue appears regardless of the network architecture and is not observed in a high-resolution dataset. The impact of the resolution heterogeneity can be partially diminished by applying advanced deep-learning-based super-resolution networks.

CONCLUSIONS

rUMAP and nUMAP are great tools for image homogeneity analysis and bias discovery, as demonstrated by applying them to COVID-19 image data. Nonetheless, nUMAP could be applied to any type of data for which a deep neural network could be constructed. Advanced image super-resolution solutions are needed to reduce the impact of the resolution diversity on the classification network decision.

Evaluation of zero counts to better understand the discrepancies between bulk and single-cell RNA-Seq platforms

Recent advances in sample preparation and sequencing technology have made it possible to profile the transcriptomes of individual cells using single-cell RNA sequencing (scRNA-Seq). Compared to bulk RNA-Seq data, single-cell data often contain a higher percentage of zero reads, mainly due to lower sequencing depth per cell, which affects mostly measurements of low-expression genes. However, discrepancies between platforms are observed regardless of expression level. Using four paired datasets with multiple samples each, we investigated technical and biological factors that can contribute to this expression shift. Using two separate machine learning models we found that, in addition to expression level, RNA integrity, gene or UTR3 length, and the number of transcripts potentially also influence the occurrence of zeros. These findings could enable the development of novel analytical methods for cross-platform expression shift correction. We also identified genes and biological pathways in our diverse datasets that consistently showed differences when assessed at the single cell versus bulk level to assist in interpreting analysis across transcriptomic platforms. At the gene level, 25 genes (0.12%) were found in all datasets as discordant, but at the pathway level, 7 pathways (2.02%) showed shared enrichment in discordant genes.

POLCOVID: a multicenter multiclass chest X-ray database (Poland, 2020-2021)

The outbreak of the SARS-CoV-2 pandemic has put healthcare systems worldwide to their limits, resulting in increased waiting time for diagnosis and required medical assistance. With chest radiographs (CXR) being one of the most common COVID-19 diagnosis methods, many artificial intelligence tools for image-based COVID-19 detection have been developed, often trained on a small number of images from COVID-19-positive patients. Thus, the need for high-quality and well-annotated CXR image databases increased. This paper introduces POLCOVID dataset, containing chest X-ray (CXR) images of patients with COVID-19 or other-type pneumonia, and healthy individuals gathered from 15 Polish hospitals. The original radiographs are accompanied by the preprocessed images limited to the lung area and the corresponding lung masks obtained with the segmentation model. Moreover, the manually created lung masks are provided for a part of POLCOVID dataset and the other four publicly available CXR image collections. POLCOVID dataset can help in pneumonia or COVID-19 diagnosis, while the set of matched images and lung masks may serve for the development of lung segmentation solutions.

Assessment of stained direct cytology smears of breast cancer for whole transcriptome and targeted messenger RNA sequencing

BACKGROUND

Rather than surgical resection, cytologic specimens are often used as first-line clinical diagnostic procedures due to higher safety, speed, and cost-effectiveness. Archival diagnostic cytology slides containing cancer can be equivalent to tissue biopsies for DNA mutation testing, but the accuracy of transcriptomic profiling by RNA sequencing (RNA-seq) is less understood.

METHODS

This study compares the results from whole transcriptome RNA-seq and a targeted RNA-seq assay of stained cytology smears (CS) versus matched tumor tissue samples preserved fresh-frozen (FF) and processed as formalin-fixed paraffin-embedded (FFPE) sections. Cellular cytology scrapes from all 11 breast cancers were fixed and stained using three common protocols: Carnoy's (CS_C) or 95% ethanol (CS_E) fixation and then Papanicolaou stain or air-dried then methanol fixation and DiffQuik stain (CS_DQ). Agreement between samples was assessed using Lin's concordance correlation coefficient.

RESULTS

Library yield for CS_DQ was too low, therefore it was not sequenced. The distributions of concordance correlation coefficient of gene expression levels in comparison to FF were comparable between CS_C and CS_E, but expression of genes enriched in stroma was lower in cytosmear samples than in FF or FFPE. Six signatures showed similar concordance to FF for all methods and two were slightly worse in CS_C and CS_E. Genomic signatures were highly concordant using targeted RNA-seq. The allele fraction of selected mutations calculated on cytosmear specimens was highly correlated with FF tissues using both RNA-seq methods.

CONCLUSION

RNA can be reliably extracted from cytology smears and is suitable for transcriptome profiling or mutation detection, except for signatures of tumor stroma.

Abstract PD9-09: Molecular differences between younger versus older estrogen receptor positive/human epidermal growth factor receptor-2 negative breast cancers

Background: The RxPONDER and TAILORx trials demonstrated benefit from adjuvant chemotherapy in patients < 50 years with node-positive breast cancer and Recurrence Score (RS) 0-25, and with node-negative disease and RS 16-25, respectively. Neither trial showed benefit in older women with RS < 26. It is unclear what explains the interaction between age and adjuvant chemotherapy benefit. Methods: We analyzed transcriptomic and genomic data from n=4,507 ER+/HER2- breast cancers to compare differences in estrogen receptor (ER), proliferation, and immune-related gene expressions, and somatic mutation patterns and mutation burden between younger (< 50 years of age) and older (>55 years) patients. We restricted our analysis to patients in the lower 80% range of in silico RS distribution to mimic the RxPONDER and TAILORx populations. Results: Five data sets were analyzed independently to assess consistency of results (TCGA n=530; microarray cohort A n=865; Cohort B n=609, METABRIC n=867, SCAN-B n=1636). Older patients had significantly higher somatic mutation burden and more frequent copy number gain in ESR1, LATS1, ARID1B, SGK1, and MYB genes (odds ratio [OR] > 8.5, FDR< 0.05), but lower frequency of GATA3 mutations (12% versus 26%, P< 0.0001). Younger patients had higher rate of ESR1 copy number loss (OR: 0.45, FDR: 0.03). There was no difference in proliferation-related gene expression. ESR1 mRNA expression was significantly lower in younger women in all cohorts (P < 0.001). A regression model of ESR1 mRNA expression using age and ER IHC positivity indicated that lower ER expression in younger patients is primarily driven by lower ESR1 mRNA per cancer cell and not by fewer ER positive cells. We also assessed four gene signatures associated with endocrine therapy sensitivity including a 4-gene ERS, a 7-gene ERS-Lum, a 106-gene ERS-Pos signature, and a 59-gene ERS-Neg signature associated with endocrine resistance. In the TCGA and METABRIC cohorts, the ERS, ERS-Lum, and ERS-Pos signatures were all lower (FDR< 0.03) while the ERS-Neg signature was higher (FDR< 0.001) in younger patients. Similarly, in both microarray cohorts, and in the SCAN-B-cohort, the ERS-Pos signature was lower and the ERS-Neg signature was higher in younger patients (FDR< 0.002). Next, we assessed 4 different immune cell signatures that have been associated with response to chemotherapy. In the TCGA, B-cell, T-cell, Mast-cell, and TIS signatures were significantly higher (FDR<.05). In the microarray Cohort-A, B cells and mast cells were significantly higher, and the T cell and TIS signatures showed a trend for higher expression. In Cohort-B, T cells, B cells, TIS, and dendritic cells signatures were significantly higher in younger patients. Significantly higher expression of immune gene signatures in younger patients were also seen in the METABRIC and SCAN-B data sets. The ER-related and immune-related gene signatures showed negative correlation and joint analysis defined three subpopulations in younger women: (i) immune-high/ER-low, (ii) immune-intermediate/ER-intermediate and (iii) immune-low/ER-intermediate, whereas in older women the dominant pattern was immune-low/ER-high. Conclusion: ESR1 mRNA and ER-associated gene expression is lower in ER positive cancers of younger compared to older patients, while immune infiltration is higher. The cytotoxic and endocrine effects of adjuvant chemotherapy could both contribute to the survival benefit seen in younger patients, but the relative contributions of these effects may vary by ER and immune phenotype. We hypothesize that in immune-high/ER-low cancers, the cytotoxic effect of chemotherapy may drive the benefit, whereas in immune-low/ER-intermediate cancers chemotherapy induced ovarian suppression may play a more important role.

Citation Format: Tao Qing, Thomas Karn, Mariya Rozenblit, Julia Foldi, Michal Marczyk, Naing Lin Shan, Kim Blenman, uwe Holtrich, Kevin Kalinsky, Funda Meric-Bernstam, Lajos Pusztai. Molecular differences between younger versus older estrogen receptor positive/human epidermal growth factor receptor-2 negative breast cancers [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 PD9-09.

Abstract P1-04-12: Systemic immune response to a Phase I/II trial of Durvalumab concomitant with neoadjuvant chemotherapy in early stage TNBC

Background Peripheral blood cells and secreted products are important regulators of the systemic immune response. Circulating cytokines have been shown to predict severity of immune-related toxicity in melanoma patients that received combination anti-PD-1-based immunotherapy. In this study, we evaluated 38 serum cytokines and the peripheral blood t cell receptor (TCR) immune repertoire of 66 patients with TNBC for associations with pCR, treatment phase (PRE; POST), and immune related treatment emergent adverse events (TEAEs). Methods Serum and peripheral blood buffy coats were collected at pre-treatment (week 0) and at post-treatment (~ week 24). MILLIPLEX MAP Human Cytokine/Chemokine Magnetic Bead Panel was run in duplicate and read on the Luminex platform. Genomic DNA was isolated using QIAGEN Kits per manufacturer’s instructions. TCR immune repertoire profiling was performed using the Immunoseq at Adaptive Biotechnologies. Statistical analysis was performed in R. P-values < 0.05 for serum cytokines and P-values < 0.05 for TCR were considered significant. Results Pielou’s diversity index showed no difference between patient groups for TCR (P>0.319). Baseline samples had increased sCD40L, EGF, and IL-10 in patients with RD compared to pCR (P< 0.05). Baseline samples had decreased FGF2 and IFN gamma in patients with immune related TEAEs compared to those with no immune related TEAEs (P< 0.05). Comparison of Pre- vs Post-treatment revealed that EGF, MIP1 alpha, IL-1 alpha, IL-8, and MDC were increased in patients with pCR compared to those with RD. Comparison of Pre- vs. Post- also revealed increased levels of cytokines (EGF, IL-7, IFN gamma, GM-CSF) in samples in patients with immune related TEAEs (P< 0.05) compared to those patients without immune-related TEAEs. It also revealed that there was an increase in a subset of cytokines (IL-7, IL-4, MCP3, and IL-1 alpha) in patients that presented with serious immune related TEAEs (P< 0.05) compared to those patients without serious immune-related TEAEs. Conclusions There are subsets of circulating inflammatory cytokines that may be associated with treatment emergent adverse events in patients with TNBC treated with Durvalumab concomitant with standard of care chemotherapy.

Citation Format: Kim Blenman, Michal Marczyk, Julia Foldi, Vignesh Gunasekharan, Andrea L.M. Silber, Lajos Pusztai. Systemic immune response to a Phase I/II trial of Durvalumab concomitant with neoadjuvant chemotherapy in early stage TNBC [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 P1-04-12.

Molecular differences between younger versus older ER-positive and HER2-negative breast cancers

The RxPONDER and TAILORx trials demonstrated benefit from adjuvant chemotherapy in patients age ≤ 50 with node-positive breast cancer and Recurrence Score (RS) 0-26, and in node-negative disease with RS 16-25, respectively, but no benefit in older women with the same clinical features. We analyzed transcriptomic and genomic data of ER+/HER2- breast cancers with in silico RS < 26 from TCGA (n = 530), two microarray cohorts (A: n = 865; B: n = 609), the METABRIC (n = 867), and the SCAN-B (n = 1636) datasets. There was no difference in proliferation-related gene expression between age groups. Older patients had higher mutation burden and more frequent ESR1 copy number gain, but lower frequency of GATA3 mutations. Younger patients had higher rate of ESR1 copy number loss. In all datasets, younger patients had significantly lower mRNA expression of ESR1 and ER-associated genes, and higher expression of immune-related genes. The ER- and immune-related gene signatures showed negative correlation and defined three subpopulations in younger women: immune-high/ER-low, immune-intermediate/ER-intermediate, and immune-low/ER-intermediate. We hypothesize that in immune-high cancers, the cytotoxic effect of chemotherapy may drive the benefit, whereas in immune-low/ER-intermediate cancers chemotherapy induced ovarian suppression may play important role.

Tumor immune microenvironment of self-identified African American and non-African American triple negative breast cancer

Differences in the tumor immune microenvironment may result in differences in prognosis and response to treatment in cancer patients. We hypothesized that differences in the tumor immune microenvironment may exist between African American (AA) and NonAA patients, due to ancestry-related or socioeconomic factors, that may partially explain differences in clinical outcomes. We analyzed clinically matched triple-negative breast cancer (TNBC) tissues from self-identified AA and NonAA patients and found that stromal TILs, PD-L1 IHC-positivity, mRNA expression of immune-related pathways, and immunotherapy response predictive signatures were significantly higher in AA samples (p < 0.05; Fisher's Exact Test, Mann-Whitney Test, Permutation Test). Cancer biology and metabolism pathways, TAM-M2, and Immune Exclusion were significantly higher in NonAA samples (p < 0.05; Permutation Test, Mann-Whitney Test). There were no differences in somatic tumor mutation burden. Overall, there is greater immune infiltration and inflammation in AA TNBC and these differences may impact response to immune checkpoint inhibitors and other therapeutic agents that modulate the immune microenvironment.

Cancer Relevance of Human Genes

BACKGROUND

We hypothesize that genes that directly or indirectly interact with core cancer genes (CCGs) in a comprehensive gene-gene interaction network may have functional importance in cancer.

METHODS

We categorized 12 767 human genes into CCGs (n = 468), 1 (n = 5467), 2 (n = 5573), 3 (n = 915), and more than 3 steps (n = 416) removed from the nearest CCG in the Search Tool for the Retrieval of Interacting Genes/Proteins network. We estimated cancer-relevant functional importance in these neighborhood categories using 1) gene dependency score, which reflects the effect of a gene on cell viability after knockdown; 2) somatic mutation frequency in The Cancer Genome Atlas; 3) effect size that estimates to what extent a mutation in a gene enhances cell survival; and 4) negative selection pressure of germline protein-truncating variants in healthy populations.

RESULTS

Cancer biology-related functional importance of genes decreases as their distance from the CCGs increases. Genes closer to cancer genes show greater connectedness in the network, have greater importance in maintaining cancer cell viability, are under greater negative germline selection pressure, and have higher somatic mutation frequency in cancer. Based on these 4 metrics, we provide cancer relevance annotation to known human genes.

CONCLUSIONS

A large number of human genes are connected to CCGs and could influence cancer biology to various extent when dysregulated; any given mutation may be functionally important in one but not in another individual depending on genomic context.

Predictive Markers of Response to Neoadjuvant Durvalumab with Nab-Paclitaxel and Dose-Dense Doxorubicin/Cyclophosphamide in Basal-Like Triple-Negative Breast Cancer

PURPOSE

We examined gene expression, germline variant, and somatic mutation features associated with pathologic response to neoadjuvant durvalumab plus chemotherapy in basal-like triple-negative breast cancer (bTNBC).

EXPERIMENTAL DESIGN

Germline and somatic whole-exome DNA and RNA sequencing, programmed death ligand 1 (PD-L1) IHC, and stromal tumor-infiltrating lymphocyte scoring were performed on 57 patients. We validated our results using 162 patients from the GeparNuevo randomized trial.

RESULTS

Gene set enrichment analysis showed that pathways involved in immunity (adaptive, humoral, innate), JAK-STAT signaling, cancer drivers, cell cycle, apoptosis, and DNA repair were enriched in cases with pathologic complete response (pCR), whereas epithelial-mesenchymal transition, extracellular matrix, and TGFβ pathways were enriched in cases with residual disease (RD). Immune-rich bTNBC with RD was enriched in CCL-3, -4, -5, -8, -23, CXCL-1, -3, -6, -10, and IL1, -23, -27, -34, and had higher expression of macrophage markers compared with immune-rich cancers with pCR that were enriched in IFNγ, IL2, -12, -21, chemokines CXCL-9, -13, CXCR5, and activated T- and B-cell markers (GZMB, CD79A). In the validation cohort, an immune-rich five-gene signature showed higher expression in pCR cases in the durvalumab arm (P = 0.040) but not in the placebo arm (P = 0.923) or in immune-poor cancers. Independent of immune markers, tumor mutation burden was higher, and PI3K, DNA damage repair, MAPK, and WNT/β-catenin signaling pathways were enriched in germline and somatic mutations in cases with pCR.

CONCLUSIONS

The TGFβ pathway is associated with immune-poor phenotype and RD in bTNBC. Among immune-rich bTNBC RD, macrophage/neutrophil chemoattractants dominate the cytokine milieu, and IFNγ and activated B cells and T cells dominate immune-rich cancers with pCR.

Comprehensive Analysis of Metabolic Isozyme Targets in Cancer

Metabolic reprogramming is a hallmark of malignant transformation, and loss of isozyme diversity (LID) contributes to this process. Isozymes are distinct proteins that catalyze the same enzymatic reaction but can have different kinetic characteristics, subcellular localization, and tissue specificity. Cancer-dominant isozymes that catalyze rate-limiting reactions in critical metabolic processes represent potential therapeutic targets. Here, we examined the isozyme expression patterns of 1,319 enzymatic reactions in 14 cancer types and their matching normal tissues using The Cancer Genome Atlas mRNA expression data to identify isozymes that become cancer-dominant. Of the reactions analyzed, 357 demonstrated LID in at least one cancer type. Assessment of the expression patterns in over 600 cell lines in the Cancer Cell Line Encyclopedia showed that these reactions reflect cellular changes instead of differences in tissue composition; 50% of the LID-affected isozymes showed cancer-dominant expression in the corresponding cell lines. The functional importance of the cancer-dominant isozymes was assessed in genome-wide CRISPR and RNAi loss-of-function screens: 17% were critical for cell proliferation, indicating their potential as therapeutic targets. Lists of prioritized novel metabolic targets were developed for 14 cancer types; the most broadly shared and functionally validated target was acetyl-CoA carboxylase 1 (ACC1). Small molecule inhibition of ACC reduced breast cancer viability in vitro and suppressed tumor growth in cell line- and patient-derived xenografts in vivo. Evaluation of the effects of drug treatment revealed significant metabolic and transcriptional perturbations. Overall, this systematic analysis of isozyme expression patterns elucidates an important aspect of cancer metabolic plasticity and reveals putative metabolic vulnerabilities.

SIGNIFICANCE

This study exploits the loss of metabolic isozyme diversity common in cancer and reveals a rich pool of potential therapeutic targets that will allow the repurposing of existing inhibitors for anticancer therapy. See related commentary by Kehinde and Parker, p. 1695.

Abstract P5-17-01: Targeting Acetyl-CoA carboxylase in pre-clinical breast cancer models

Background: Cancer cells and normal cells of the same lineage differ in their metabolism. We previously described large scale shifts in isoenzyme distribution between matching cancer and normal tissues and identified Acetyl-coA carboxylase (ACC1/ACACA) as a cancer dominant enzyme that is overexpressed in multiple cancer types. ACC1 catalyzes the initial rate-limiting step in de novo fatty acid synthesis, the conversion of acetyl-CoA to malonyl-CoA. Gene knock-out experiments demonstrated that this enzyme is essential for cancer growth. In this study, we evaluated the in vitro and in vivo efficacy of a small molecule ACC inhibitor, PF05175157 as a potential anticancer drug. This drug has been tested in clinical trials for diabetes, but development was discontinued due to grade 2 thrombocytopenia.Methods: We performed in vitro cytotoxicity assays in 15 breast cancer cell lines and in normal mammary epithelial HMEC cells, examined effect on apoptosis and cell cycle progression, and tested for synergy with alpelisib, docetaxel, doxorubicin, everolimus, iniparib, neratinib and TEPP46 (PKM2 and PKLR activator). We next assessed in vivo single agent activity in a triple negative patient derived (PDX) model (J000102184) in NSGTM mice and in MDAMB468 xenografts implanted into Rag2/IL2RG double knockout mice. We performed RNA sequencing and metabolomic profiling of cells treated with PF05175157 to study metabolic and transcriptomic effects of the drug. Results: PF05175157 induced time and dose dependent growth inhibition in all but 1 of the 15 cancer cell lines. The estimated EC50 after 72h exposure ranged from 0.95 to 76 μg/mL in T47D and BT549 cells, respectively (Cmax of 20 μg/mL can be achieved in human serum). There was no significant inhibitory effect on HMEC cells. In cancer cell lines, the % of apoptotic cells increased from 4% to 8% in BT474 and from 7.7% to 17.8% in MDMBA468 cells upon treatment with the compound, and there was a trend towards G2/M cell cycle arrest in both cell lines after 72 hours of exposure (10μg/mL). In drug combination experiments, PF05175157 added to iniparib, or to the PKM2 activator, TEPP46, decreased cell viability compared to single agent therapies in several cell lines. PF05175157 significantly delayed tumor growth compared to vehicle, when administered orally (20mg/kg gavage BID) in a TNBC PDX model (median tumor volume after 33 days: 334.1 mm3 in PF05175157-treated vs. 490.5 mm3 in methylcellulose-treated mice; p=3.39e-7) and intraperitoneally (20 mg/kg in DMSO) in an MDAMB468 xenograft model (median tumor volume after 40 days: 244.5 mm3 in PF05175157-treated vs. 303.3 mm3 in DMSO-treated mice; p&lt;0.05). Transcriptomic and metabolic profiling of MDAMB468 and BT474 cells treated with 10 ug/ml PF05175157 for 6 and 24 hours revealed activation of immune signaling, epigenetic regulation and DNA damage repair pathways along with down-regulation of a broad range of metabolic pathways. Conclusions: The small molecule ACC inhibitor, PF05175157, has significant single agent in vitro and in vivo growth inhibitory effect on a range of breast cancer cell lines at concentrations that can be achieved in human serum. It showed synergy with iniparib and another metabolic inhibitor (TEPP46). Targeting de novo fatty acid synthesis by inhibiting ACC is a promising therapeutic strategy.

Citation Format: Julia Foldi, Michal Marczyk, Vignesh Gunasekharan, Tao Qing, Raghav Sehgal, Naing Lin Shan, Viswanathan Muthusamy, Sheila Umlau, Yulia V. Surovtseva, Richard Kibbey, Lajos Pusztai. Targeting Acetyl-CoA carboxylase in pre-clinical breast cancer models [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P5-17-01.

Importance of SNP Dependency Correction and Association Integration for Gene Set Analysis in Genome-Wide Association Studies

A typical genome-wide association study (GWAS) analyzes millions of single-nucleotide polymorphisms (SNPs), several of which are in a region of the same gene. To conduct gene set analysis (GSA), information from SNPs needs to be unified at the gene level. A widely used practice is to use only the most relevant SNP per gene; however, there are other methods of integration that could be applied here. Also, the problem of nonrandom association of alleles at two or more loci is often neglected. Here, we tested the impact of incorporation of different integrations and linkage disequilibrium (LD) correction on the performance of several GSA methods. Matched normal and breast cancer samples from The Cancer Genome Atlas database were used to evaluate the performance of six GSA algorithms: Coincident Extreme Ranks in Numerical Observations (CERNO), Gene Set Enrichment Analysis (GSEA), GSEA-SNP, improved GSEA for GWAS (i-GSEA4GWAS), Meta-Analysis Gene-set Enrichment of variaNT Associations (MAGENTA), and Over-Representation Analysis (ORA). Association of SNPs to phenotype was calculated using modified McNemar's test. Results for SNPs mapped to the same gene were integrated using Fisher and Stouffer methods and compared with the minimum p-value method. Four common measures were used to quantify the performance of all combinations of methods. Results of GSA analysis on GWAS were compared to the one performed on gene expression data. Comparing all evaluation metrics across different GSA algorithms, integrations, and LD correction, we highlighted CERNO, and MAGENTA with Stouffer as the most efficient. Applying LD correction increased prioritization and specificity of enrichment outcomes for all tested algorithms. When Fisher or Stouffer were used with LD, sensitivity and reproducibility were also better. Using any integration method was beneficial in comparison with a minimum p-value method in specific combinations. The correlation between GSA results from genomic and transcriptomic level was the highest when Stouffer integration was combined with LD correction. We thoroughly evaluated different approaches to GSA in GWAS in terms of performance to guide others to select the most effective combinations. We showed that LD correction and Stouffer integration could increase the performance of enrichment analysis and encourage the usage of these techniques.

Analysis of the Applicability of microRNAs in Peripheral Blood Leukocytes as Biomarkers of Sensitivity and Exposure to Fractionated Radiotherapy towards Breast Cancer

Biomarkers for predicting individual response to radiation and for dose verification are needed to improve radiotherapy. A biomarker should optimally show signal fidelity, meaning that its level is stable and proportional to the absorbed dose. miRNA levels in human blood serum were suggested as promising biomarkers. The aim of the present investigation was to test the miRNA biomarker in leukocytes of breast cancer patients undergoing external beam radiotherapy. Leukocytes were isolated from blood samples collected prior to exposure (control); on the day when a total dose of 2 Gy, 10 Gy, or 20 Gy was reached; and one month after therapy ended (46-50 Gy in total). RNA sequencing was performed and univariate analysis was used to analyse the effect of the radiation dose on the expression of single miRNAs. To check if combinations of miRNAs can predict absorbed dose, a multinomial logistic regression model was built using a training set from eight patients (representing 40 samples) and a validation set with samples from the remaining eight patients (15 samples). Finally, Broadside, an explorative interaction mining tool, was used to extract sets of interacting miRNAs. The most prominently increased miRNA was miR-744-5p, followed by miR-4461, miR-34a-5p, miR-6513-5p, miR-1246, and miR-454-3p. Decreased miRNAs were miR-3065-3p, miR-103a-2-5p, miR-30b-3p, and miR-5690. Generally, most miRNAs showed a relatively strong inter-individual variability and different temporal patterns over the course of radiotherapy. In conclusion, miR-744-5p shows promise as a stable miRNA marker, but most tested miRNAs displayed individual signal variability which, at least in this setting, may exclude them as sensitive biomarkers of radiation response.

Abstract 423: Pooled analysis of matching score and patient outcome in I-PREDICT and WINTHER studies

Introduction: Comprehensive genomic profiling (CGP) of tumors is available at many specialized cancer centers, potentially allowing oncologists to individualize patient therapy based on tumor molecular features. However, clinicians still lack a systematic approach to deal with the wealth of information provided by CGP. The proportion of plausible drug-matched alterations over the total number of alterations detected by CGP (Matching Score, MS) has been postulated to predict the likelihood of patient benefit. We performed a systematic analysis of MS and its association with the outcome of patients treated according to tumor DNA alterations within two recent precision oncology trials.

Materials and Methods: We downloaded the supplementary data from WINTHER (NCT01856296) and I-PREDICT (NCT02534675) studies. We included patients who received treatment for one or more single-gene alterations, excluding eight patients that received immunotherapy due to high tumor mutational burden, PD-L1 positivity or microsatellite instability. We used MS cut-offs of 0.25 (as in WINTHER) and 0.5 (as in I-PREDICT) and also MS as a continuous variable to study its association with clinical variables such as age, sex, tumor type, and number of previous treatment lines (Kruskall-Wallis and Wilcoxon-rank tests), as well as disease control rate (DCR), progression-free survival (PFS) and overall survival (OS) (Cox proportional hazards models).

Results: The pooled analysis included 144 patients. Median age was 60 years (range 21-86) and 64% had received ≥ 2 lines of therapy. Most patients had colorectal (37%), lung (14%), or breast (11%) tumors, and 58% were treated with 2 or more drugs. MS was not associated with any clinical variable nor with DCR. MS &gt; 0.25 (n=98; 68%) was associated with longer PFS (median 3.5 vs. 1.9 months, HR [95%CI]: 0.60 [0.41-0.87]) but not with OS. In contrast, MS &gt; 0.5 (n=41; 28%) was associated with longer OS (median 14.1 vs. 6.5 months, HR [95%CI]: 0.50 [0.29-0.54]) but not with PFS. As a continuous variable, increasing MS values were associated with better OS (HR [95% CI]: 0.43 [0.20-0.95]) and with a non-significant trend towards better PFS (HR [95% CI]: 0.54 [0.21-1.05]). In a multivariable model adjusted for Age and study, PFS/OS probabilities improved consistently with increasing MS values from 0 to 0.5. With MS values &gt; 0.5, OS probabilities kept increasing while PFS probabilities plateaued.

Conclusions: The significant and consistent relationship between MS and OS (in contrast with PFS) may be explained by patients with higher MS values having more targeted therapies available for successive treatment lines, although we cannot exclude the existence of other confounding prognostic variables. We note that, as better drugs are developed and new knowledge is incorporated into the clinic, the value of CGP will probably increase over time, improving the number of patients with actionable alterations and their outcome.

Citation Format: David Casadevall, Michal Marczyk, Xavier Monzonis, Lajos Pusztai, Joan Albanell. Pooled analysis of matching score and patient outcome in I-PREDICT and WINTHER studies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 423.

Treatment scheduling effects on the evolution of drug resistance in heterogeneous cancer cell populations

The effect of scheduling of targeted therapy combinations on drug resistance is underexplored in triple-negative breast cancer (TNBC). TNBC constitutes heterogeneous cancer cell populations the composition of which can change dynamically during treatment resulting in the selection of resistant clones with a fitness advantage. We evaluated crizotinib (ALK/MET inhibitor) and navitoclax (ABT-263; Bcl-2/Bcl-xL inhibitor) combinations in a large design consisting of 696 two-cycle sequential and concomitant treatment regimens with varying treatment dose, duration, and drug holiday length over a 26-day period in MDA-MB-231 TNBC cells and found that patterns of resistance depend on the schedule and sequence in which the drugs are given. Further, we tracked the clonal dynamics and mechanisms of resistance using DNA-integrated barcodes and single-cell RNA sequencing. Our study suggests that longer formats of treatment schedules in vitro screening assays are required to understand the effects of resistance and guide more realistically in vivo and clinical studies.

Characterization of the tumor immune microenvironment of triple-negative breast cancer (TNBC) patients who self-identify as African American (AA) or non-African American (NonAA)

564

Background: What tumor biological differences, if any, contribute to the higher incidence and worse prognosis of triple negative breast cancer (TNBC) in African American (AA) compared to NonAA patients are unknown. We hypothesized that differences in the tumor immune microenvironment may contribute to the outcome disparities. The purpose of this study was to characterize and compare the immune microenvironment of TNBC between patients self-identified as NonAA or AA. Methods: Formalin fixed paraffin embedded surgically resected cancer and paired normal tissues collected before any systemic therapy and the corresponding clinical data were collected for NonAA (n = 56) and AA (n = 54) stage I-III TNBC treated at Yale Cancer Center between 2000-2017. The two cohorts were matched for clinical stage, age of diagnosis, and year of diagnosis. We performed somatic and germline whole exome sequencing (WES), bulk RNA sequencing, and immunohistochemistry to assess PD-L1 expression (SP142). Stromal tumor infiltrating lymphocytes (sTILs) were assessed on H&E slides. Mutation load, mutation frequencies, and gene expression differences were compared at gene and pathway level. Immune cell composition was estimated through gene expression deconvolution analyses (TIDE). Results: Tumor mutational burden was similar between the two cohorts. At gene level, few genes had significantly different somatic mutation frequencies, or differential mRNA expression between AA and NonAA samples. Pathway level alterations showed inflammation, immunity (adaptive; innate), antigen presentation, and allograft rejection pathways were more affected by somatic mutations in AA samples. The affected genes differed from cancer to cancer and were not recurrent and therefore were missed at gene level analysis. Gene set enrichment and co-expression analysis also showed higher immune related pathway expression in AA samples. Unsupervised co-expression cluster analysis confirmed coordinated overexpression of genes involved in immunity, inflammation, and cytokine/chemokine signaling in AA patients. Two immunotherapy response predictive signatures, immune inflamed and the IFNG as well as sTILs score and PD-L1 positivity were also higher in AA samples. These findings raise the possibility that immune checkpoint inhibitors might be particularly effective in AA patients. In NonAA samples, the EMT transition, angiogenesis, adipogenesis, myogenesis, fatty acid metabolism, TGFβ signaling, UV-response, and hypoxia pathways were overexpressed. TIDE analysis suggested higher levels of TAM M2, overall TIDE score, and the Immune Exclusion score in NonAA samples. Conclusions: TNBC in AA patients more frequently harbor somatic mutations in genes involved with immune functions and overexpress immune and inflammatory genes compared to NonAA patients.

Treatment Efficacy Score (TES), a continuous residual cancer burden-based metric to compare neoadjuvant chemotherapy efficacy between trial arms in the I-SPY 2 trial

587

Background: Residual cancer burden (RCB) is a continuous score that captures the amount of residual cancer after neoadjuvant chemotherapy and predicts disease recurrence and survival across all breast cancer subtypes. RCB score 0 corresponds to pathological complete response (pCR; ypT0, ypN0). We hypothesize that comparison of the distributions of RCB scores between randomized treatment arms of a trial could predict treatment effect on recurrence free survival better than comparison of pCR rates only. Methods: The cancer Treatment Efficacy Score (TES) compares efficacies of two treatments using non-continuous RCB results. We examined (i) area between cumulative distribution (ABC) functions; (ii) density ratio of RCB scores; and (iii) density difference of RCB scores from two treatments, to select the most efficient metric to compute TES. A random permutation procedure was used to estimate the p-value from each test. These methods were applied to data from the durvalumab/olaparib arm and corresponding controls of the I-SPY2 trial, separately by molecular subtype. In subsampling and simulation experiments we assessed robustness of results including power and false positive rate control under variable sample sizes to select the most robust TES metric. The other 11 experimental arms of I-SPY2 were used to assess the performance of the final metric. We calculated correlation between TES and (i) pCR rate difference, and 3- and 5-year (ii) event-free (EFS) and (iii) distant recurrence free survivals (DRFS). Results: RCB scores are multimodal and do not follow normal distribution.In simulated data ABC provided more stable results than the other methods, had good power, performed well with small sample sizes, resulted in low false positive rate, required the least computational time, and therfore was selected as the TES metric for validation in 11 arms of I-SPY2. We found a high correlation between difference in pCR rate and TES value across all molecular subtypes in each of the 11 trial arms (r = 0.92, p = 1.7e-8). There was also significant linear relationship between TES and survival estimates in EFS (r = 0.58, p = 9.3e-3 for 3-years survival; r = 0.62, p = 4.8e-3 for 5-years survival) and DRFS (r = 0.56, p = 1.2e-2 for 3-years survival; r = 0.54, p = 1.8e-2 for 5-years survival). Statistically significant TES score correlated significantly with higher benefit in 3-years survival (p = 9.7e-4 for EFS; p = 5.7e-3 for DRFS) and 5-years survival (p = 9.7e-4 for EFS; p = 3.0e-3 for DRFS). In most instances, this correlation with survival was higher than seen with pCR difference. Conclusions: TES is a novel more optimal metric to identify the more effective cytotoxic neoadjuvant regimen from the entire distribution of pathologic response that significantly correlates with event and recurrence free survival and may serve as a better surrogate than pCR rate difference.

Diverse immune response of DNA damage repair-deficient tumors

Tumors with DNA damage repair (DDR) deficiency accumulate genomic alterations that may serve as neoantigens and increase sensitivity to immune checkpoint inhibitor. However, over half of DDR-deficient tumors are refractory to immunotherapy, and it remains unclear which mutations may promote immunogenicity in which cancer types. We integrate deleterious somatic and germline mutations and methylation data of DDR genes in 10,080 cancers representing 32 cancer types and evaluate the associations of these alterations with tumor neoantigens and immune infiltrates. Our analyses identify DDR pathway mutations that are associated with higher neoantigen loads, adaptive immune markers, and survival outcomes of immune checkpoint inhibitor-treated animal models and patients. Different immune phenotypes are associated with distinct types of DDR deficiency, depending on the cancer type context. The comprehensive catalog of immune response-associated DDR deficiency may explain variations in immunotherapy outcomes across DDR-deficient cancers and facilitate the development of genomic biomarkers for immunotherapy.

Tumor immunogenicity is associated with DNA damage repair deficiencies (DDR-ds)

The immunogenicity of DDR alterations varies by pathways and cancer types

DDR-d tumors with high immune infiltrates correlate with immunotherapy response

Technical Validity of a Customized Assay of Sensitivity to Endocrine Therapy Using Sections from Fixed Breast Cancer Tissue

BACKGROUND

We translated a multigene expression index to predict sensitivity to endocrine therapy for Stage II-III breast cancer (SET2,3) to hybridization-based expression assays of formalin-fixed paraffin-embedded (FFPE) tissue sections. Here we report the technical validity with FFPE samples, including preanalytical and analytical performance.

METHODS

We calibrated SET2,3 from microarrays (Affymetrix U133A) of frozen samples to hybridization-based assays of FFPE tissue, using bead-based QuantiGene Plex (QGP) and slide-based NanoString (NS). The following preanalytical and analytical conditions were tested in controlled studies: replicates within and between frozen and fixed samples, age of paraffin blocks, homogenization of fixed sections versus extracted RNA, core biopsy versus surgically resected tumor, technical replicates, precision over 20 weeks, limiting dilution, linear range, and analytical sensitivity. Lin's concordance correlation coefficient (CCC) was used to measure concordance between measurements.

RESULTS

SET2,3 index was calibrated to use with QGP (CCC 0.94) and NS (CCC 0.93) technical platforms, and was validated in two cohorts of older fixed samples using QGP (CCC 0.72, 0.85) and NS (CCC 0.78, 0.78). QGP assay was concordant using direct homogenization of fixed sections versus purified RNA (CCC 0.97) and between core and surgical sample types (CCC 0.90), with 100% accuracy in technical replicates, 1-9% coefficient of variation over 20 weekly tests, linear range 3.0-11.5 (log2 counts), and analytical sensitivity ≥2.0 (log2 counts).

CONCLUSIONS

Measurement of the novel SET2,3 assay was technically valid from fixed tumor sections of biopsy or resection samples using simple, inexpensive, hybridization methods, without the need for RNA purification.

Neoadjuvant durvalumab plus weekly nab-paclitaxel and dose-dense doxorubicin/cyclophosphamide in triple-negative breast cancer

The goal of this Phase I/II trial is to assess the safety and efficacy of administering durvalumab concurrent with weekly nab-paclitaxel and dose-dense doxorubicin/cyclophosphamide (ddAC) neoadjuvant therapy for stages I-III triple-negative breast cancer. The primary endpoint is pathologic complete response (pCR:ypT0/is, ypN0). The response was correlated with PDL1 expression and stromal tumor-infiltrating lymphocytes (sTILs). Two dose levels of durvalumab (3 and 10 mg/kg) were assessed. PD-L1 was assessed using the SP263 antibody; ≥1% immune and tumor cell staining was considered positive; sTILs were calculated as the area occupied by mononuclear inflammatory cells over the total intratumoral stromal area. 59 patients were evaluable for toxicity and 55 for efficacy in the Phase II study (10 mg/kg dose). No dose-limiting toxicities were observed in Phase I. In Phase II, pCR rate was 44% (95% CI: 30-57%); 18 patients (31%) experienced grade 3/4 treatment-related adverse events (AE), most frequently neutropenia (n = 4) and anemia (n = 4). Immune-related grade 3/4 AEs included Guillain-Barre syndrome (n = 1), colitis (n = 2), and hyperglycemia (n = 2). Of the 50 evaluable patients for PD-L1, 31 (62%) were PD-L1 positive. pCR rates were 55% (95% CI: 0.38-0.71) and 32% (95% CI: 0.12-0.56) in the PD-L1 positive and negative groups (p = 0.15), respectively. sTIL counts were available on 52 patients and were significantly higher in the pCR group (p = 0.0167). Concomitant administration of durvalumab with sequential weekly nab-paclitaxel and ddAC neoadjuvant chemotherapy resulted in a pCR rate of 44%; pCR rates were higher in sTIL-high cancers.

Targeted RNAseq assay incorporating unique molecular identifiers for improved quantification of gene expression signatures and transcribed mutation fraction in fixed tumor samples

BACKGROUND

Our objective was to assess whether modifications to a customized targeted RNA sequencing (RNAseq) assay to include unique molecular identifiers (UMIs) that collapse read counts to their source mRNA counts would improve quantification of transcripts from formalin-fixed paraffin-embedded (FFPE) tumor tissue samples. The assay (SET4) includes signatures that measure hormone receptor and PI3-kinase related transcriptional activity (SETER/PR and PI3Kges), and measures expression of selected activating point mutations and key breast cancer genes.

METHODS

Modifications included steps to introduce eight nucleotides-long UMIs during reverse transcription (RT) in bulk solution, followed by polymerase chain reaction (PCR) of labeled cDNA in droplets, with optimization of the polymerase enzyme and reaction conditions. We used Lin's concordance correlation coefficient (CCC) to measure concordance, including precision (Rho) and accuracy (Bias), and nonparametric tests (Wilcoxon, Levene's) to compare the modified (NEW) SET4 assay to the original (OLD) SET4 assay and to whole transcriptome RNAseq using RNA from matched fresh frozen (FF) and FFPE samples from 12 primary breast cancers.

RESULTS

The modified (NEW) SET4 assay measured single transcripts (p< 0.001) and SETER/PR (p=0.002) more reproducibly in technical replicates from FFPE samples. The modified SET4 assay was more precise for measuring single transcripts (Rho 0.966 vs 0.888, p< 0.01) but not multigene expression signatures SETER/PR (Rho 0.985 vs 0.968) or PI3Kges (Rho 0.985 vs 0.946) in FFPE, compared to FF samples. It was also more precise than wtRNAseq of FFPE for measuring transcripts (Rho 0.986 vs 0.934, p< 0.001) and SETER/PR (Rho 0.993 vs 0.915, p=0.004), but not PI3Kges (Rho 0.988 vs 0.945, p=0.051). Accuracy (Bias) was comparable between protocols. Two samples carried a PIK3CA mutation, and measurements of transcribed mutant allele fraction was similar in FF and FFPE samples and appeared more precise with the modified SET4 assay. Amplification efficiency (reads per UMI) was consistent in FF and FFPE samples, and close to the theoretically expected value, when the library size exceeded 400,000 aligned reads.

CONCLUSIONS

Modifications to the targeted RNAseq protocol for SET4 assay significantly increased the precision of UMI-based and reads-based measurements of individual transcripts, multi-gene signatures, and mutant transcript fraction, particularly with FFPE samples.

Multi-Omics Investigation of Innate Navitoclax Resistance in Triple-Negative Breast Cancer Cells

Cancer cells employ various defense mechanisms against drug-induced cell death. Investigating multi-omics landscapes of cancer cells before and after treatment can reveal resistance mechanisms and inform new therapeutic strategies. We assessed the effects of navitoclax, a BCL2 family inhibitor, on the transcriptome, methylome, chromatin structure, and copy number variations of MDA-MB-231 triple-negative breast cancer (TNBC) cells. Cells were sampled before treatment, at 72 h of exposure, and after 10-day drug-free recovery from treatment. We observed transient alterations in the expression of stress response genes that were accompanied by corresponding changes in chromatin accessibility. Most of these changes returned to baseline after the recovery period. We also detected lasting alterations in methylation states and genome structure that suggest permanent changes in cell population composition. Using single-cell analyses, we identified 2350 genes significantly upregulated in navitoclax-resistant cells and derived an 18-gene navitoclax resistance signature. We assessed the navitoclax-response-predictive function of this signature in four additional TNBC cell lines in vitro and in silico in 619 cell lines treated with 251 different drugs. We observed a drug-specific predictive value in both experiments, suggesting that this signature could help guiding clinical biomarker studies involving navitoclax.

Abstract 6333: Genomic, transcriptomic, and epigenetic profiling of triple-negative breast cancer cells after Navitoclax treatment

Cancer cells possess various anti-apoptotic defense mechanisms that can protect them from drug induced cell death. We investigated the effects of treatment with the bcl-2/bcl-XL inhibitor Navitoclax on transcriptome, methylome, and genome of MDA-MB-231 cells. Two biological replicates were treated simultaneously and profiled using 10x Genomics single-cell RNA sequencing (scRNAseq; gene expression), bulk RNA sequencing (bRNAseq; gene expression), ATAC sequencing (ATAC; chromatin accessibility), bisulphite targeted sequencing (DNAm; DNA methylation) and shallow whole-genome sequencing (WGS; copy number variants) at baseline before treatment, after 3 days of 10µM navitoclax treatment, and after 10 days of cell recovery from treatment. Variance between biological replicates on transcriptome level was minimal before the treatment (R&gt;0.99), but we noticed divergence in gene expression after recovery from the treatment (both on bulk and single-cell level), mostly driven by cell cycle genes. On other platforms, biological replicates were similar in all 3 phases. After treatment, we observed more genes with decreased expression (n=151) in comparison to baseline, but after 10 days of recovery there were more up-regulated genes (n=655) in comparison to samples after treatment. Chromatin accessibility for most genomic regions increased after treatment in comparison to baseline, suggesting an acute response to the stress caused by drug treatment, and then returned to baseline level after recovery period. DNA methylation patterns showed more regions with decreased methylation after treatment (n=529) in comparison to baseline which remained detectable even after the recovery period. In WGS data, we found 752 genes with deletions and amplifications only in baseline samples representing genomic background of drug sensitive cells, mostly enriched in epithelial mesenchymal transition pathway. On single-cell level, we identified a subset of cells that were resistant to treatment and discovered 2,324 genes significantly up-regulated in these cells, that could be potential markers of resistance. In gene set analysis, these markers were enriched in MYC and E2F target gene sets, and were involved in angiogenesis and JAK-STAT pathway. We measured the expression of 16 top up-regulated markers of resistance in 4 different TNBC cell lines using qPCR and found that 5 were significantly enriched after treatment in 3 cell lines, 5 in 2 cell lines and 5 in a single cell line. Summarizing, we thoroughly described molecular effects of Navitoclax treatment and showed that most cells return to the basal transcriptional state after the drug recovery period, but bulk genome and methylome are permanently changed. Finally, we provided a list of new markers of resistance that may be useful in the studies of combinational therapies with other drugs. This work was funded by Breast Cancer Research Foundation.

Citation Format: Michal Marczyk, Vignesh Gunasekharan, Jun Zhao, Rihao Qu, Xiaotong Li, Gauri A. Patwardhan, Vikram B. Wali, Abhishek K. Gupta, Manoj M. Pillai, Yuval Kluger, Christos Hatzis, Lajos Pusztai. Genomic, transcriptomic, and epigenetic profiling of triple-negative breast cancer cells after Navitoclax treatment [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 6333.

Immunological Differences Between Immune-Rich Estrogen Receptor-Positive and Immune-Rich Triple-Negative Breast Cancers

PURPOSE

A subset of estrogen receptor–positive (ER-positive) breast cancer (BC) contains high levels of tumor-infiltrating lymphocytes (TILs), similar to triple-negative BC (TNBC). The majority of immuno-oncology trials target TNBCs because of the greater proportion of TIL-rich TNBCs. The extent to which the immune microenvironments of immune-rich ER-positive BC and TNBC differ is unknown.

PATIENTS AND METHODS

RNA sequencing data from The Cancer Genome Atlas (TCGA; n = 697 ER-positive BCs; n = 191 TNBCs) were used for discovery; microarray expression data from Molecular Taxonomy of Breast Cancer International Consortium (METABRIC; n = 1,186 ER-positive BCs; n = 297 TNBCs) was used for validation. Patients in the top 25th percentile of a previously published total TIL metagene score distribution were considered immune rich. We compared expression of immune cell markers, immune function metagenes, and immuno-oncology therapeutic targets among immune-rich subtypes.

RESULTS

Relative fractions of resting mast cells (TCGA P_adj = .009; METABRIC P_adj = 4.09E-15), CD8⁺ T cells (TCGA P_adj = .015; METABRIC P_adj = 0.390), and M2-like macrophages (TCGA P_adj= 4.68E-05; METABRIC P_adj = .435) were higher in immune-rich ER-positive BCs, but M0-like macrophages (TCGA P_adj = 0.015; METABRIC P_adj = .004) and M1-like macrophages (TCGA P_adj = 9.39E-08; METABRIC P_adj = 6.24E-11) were higher in immune-rich TNBCs. Ninety-one immune-related genes (eg, CXCL14, CSF3R, TGF-B3, LRRC32/GARP, TGFB-R2) and a transforming growth factor β (TGF-β) response metagene were significantly overexpressed in immune-rich ER-positive BCs, whereas 41 immune-related genes (eg, IFNG, PD-L1, CTLA4, MAGEA4) were overexpressed in immune-rich TNBCs in both discovery and validation data sets. TGF-β pathway member genes correlated negatively with expression of immune activation markers (IFNG, granzyme-B, perforin) and positively with M2-like macrophages (IL4, IL10, and MMP9) and regulatory T-cell (FOXP3) markers in both subtypes.

CONCLUSION

Different immunotherapy strategies may be optimal in immune-rich ER-positive BC and TNBC. Drugs targeting the TGF-β pathway and M2-like macrophages are promising strategies in immune-rich ER-positive BCs to augment antitumor immunity.

BatchI: Batch effect Identification in high-throughput screening data using a dynamic programming algorithm

Motivation

In contemporary biological experiments, bias, which interferes with the measurements, requires attentive processing. Important sources of bias in high-throughput biological experiments are batch effects and diverse methods towards removal of batch effects have been established. These include various normalization techniques, yet many require knowledge on the number of batches and assignment of samples to batches. Only few can deal with the problem of identification of batch effect of unknown structure. For this reason, an original batch identification algorithm through dynamical programming is introduced for omics data that may be sorted on a timescale.

Results

BatchI algorithm is based on partitioning a series of high-throughput experiment samples into sub-series corresponding to estimated batches. The dynamic programming method is used for splitting data with maximal dispersion between batches, while maintaining minimal within batch dispersion. The procedure has been tested on a number of available datasets with and without prior information about batch partitioning. Datasets with a priori identified batches have been split accordingly, measured with weighted average Dice Index. Batch effect correction is justified by higher intra-group correlation. In the blank datasets, identified batch divisions lead to improvement of parameters and quality of biological information, shown by literature study and Information Content. The outcome of the algorithm serves as a starting point for correction methods. It has been demonstrated that omitting the essential step of batch effect control may lead to waste of valuable potential discoveries.

Availability and implementation

The implementation is available within the BatchI R package at http://zaed.aei.polsl.pl/index.php/pl/111-software.

Supplementary information

Supplementary data are available at Bioinformatics online.

Germline variant burden in cancer genes correlates with age at diagnosis and somatic mutation burden

Cancers harbor many somatic mutations and germline variants, we hypothesized that the combined effect of germline variants that alter the structure, expression, or function of protein-coding regions of cancer-biology related genes (gHFI) determines which and how many somatic mutations (sM) must occur for malignant transformation. We show that gHFI and sM affect overlapping genes and the average number of gHFI in cancer hallmark genes is higher in patients who develop cancer at a younger age (r = -0.77, P = 0.0051), while the average number of sM increases in increasing age groups (r = 0.92, P = 0.000073). A strong negative correlation exists between average gHFI and average sM burden in increasing age groups (r = -0.70, P = 0.017). In early-onset cancers, the larger gHFI burden in cancer genes suggests a greater contribution of germline alterations to the transformation process while late-onset cancers are more driven by somatic mutations.

Molecular Profiling for Predictors of Radiosensitivity in Patients with Breast or Head-and-Neck Cancer

Nearly half of all cancers are treated with radiotherapy alone or in combination with other treatments, where damage to normal tissues is a limiting factor for the treatment. Radiotherapy-induced adverse health effects, mostly of importance for cancer patients with long-term survival, may appear during or long time after finishing radiotherapy and depend on the patient’s radiosensitivity. Currently, there is no assay available that can reliably predict the individual’s response to radiotherapy. We profiled two study sets from breast (n = 29) and head-and-neck cancer patients (n = 74) that included radiosensitive patients and matched radioresistant controls.. We studied 55 single nucleotide polymorphisms (SNPs) in 33 genes by DNA genotyping and 130 circulating proteins by affinity-based plasma proteomics. In both study sets, we discovered several plasma proteins with the predictive power to find radiosensitive patients (adjusted p < 0.05) and validated the two most predictive proteins (THPO and STIM1) by sandwich immunoassays. By integrating genotypic and proteomic data into an analysis model, it was found that the proteins CHIT1, PDGFB, PNKD, RP2, SERPINC1, SLC4A, STIM1, and THPO, as well as the VEGFA gene variant rs69947, predicted radiosensitivity of our breast cancer (AUC = 0.76) and head-and-neck cancer (AUC = 0.89) patients. In conclusion, circulating proteins and a SNP variant of VEGFA suggest that processes such as vascular growth capacity, immune response, DNA repair and oxidative stress/hypoxia may be involved in an individual’s risk of experiencing radiation-induced toxicity.

Abstract P3-09-05: Predictive markers of response to durvalumab concurrent with nab-paclitaxel and dose dense doxorubicin cyclophosphamide (ddAC) neoadjuvant therapy for triple negative breast cancer (TNBC)

Background: The goal of this analysis was to identify immune cell types and mRNA expression signatures that are associated with pathologic complete response (pCR: ypT0/N0) to neoadjuvant durvalumab concurrent with weekly nab-paclitaxel (100 mg/m2) x 12 followed by ddAC x 4 chemotherapy in stage I-III triple negative breast cancer (TNBC).

Methods: Pre-treatment core needle biopsies were obtained from 57 patients (n=25 pCR, n=32 residual disease (RD)) who participated in Phase I/II clinical trial (NCT02489448) for RNA sequencing and histology assessment. Formalin-fixed paraffin-embedded (FFPE) core biopsies were stained by immunofluorescence for CD8, CD68 and cytokeratin, images were acquired on the Vectra-Polaris system and analyzed using InForm software with the adaptive segmentation into three compartments: Tumor, Stroma, and All (Tumor + Stroma). Correlation between immune cell density in the tissue compartments and pCR was assessed. RNA was extracted from core biopsies collected into RNAlater and poly-A enriched mRNAs were sequenced on Illumina platform using NovaSeq paired-end, 100bp fragments, with a depth of 50 million reads. Six specimens failed QC due to insufficient tumor cells in the sample. Correlation between pCR and immune markers and previously published immune, proliferation and DNA damage response deficiency (DDRD) gene signatures were assessed using logistic regression. Due to high overlap of genes across signatures, no multiple testing correction was done. Differentially expressed genes were found using DESeq2 R package with Benjamini-Hochberg correction for multiple testing. Broadside, an interaction mining tool, was used to extract sets of combinations of genes associated with pathologic response.

Results: In cases with pCR, we detected significantly higher overall CD8 cell density and a trend for higher CD8 in tumor and stroma compartments. There was no significant difference in CD68 cell density between response groups. At the RNA level, high expression of DDRD, IFNγ, T-cell, B-cell, dendritic cell, M1 macrophage signatures and the tumor inflammation gene signature were significantly associated with higher rate of pCR. High expression of epithelial mesenchymal transition and TGFβ and IL8/VEGF gene signatures were associated with higher rate of residual disease. Proliferation gene signatures were not associated with response in this TNBC population. At individual gene level, the highest pCR predictive value was observed for the STAT1 + PCF11 and LAMP3 + SDR39U1 doublets.

Conclusion: High CD8 cell density, high expression of a broad range of immune gene expression signatures and DNA damage response deficiency are associated with greater sensitivity to neoadjuvant anti-PD-L1 and chemotherapy.

Citation Format: Kim RM Blenman, Xiaotong Li, Michal Marczyk, Tess O'Meara, Vesal Yaghoobi, Vignesh Gunasekharan, Tristen Park, David Rimm, Lajos Pusztai. Predictive markers of response to durvalumab concurrent with nab-paclitaxel and dose dense doxorubicin cyclophosphamide (ddAC) neoadjuvant therapy for triple negative breast cancer (TNBC) [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P3-09-05.

Abstract P5-13-02: Neoadjuvant aromatase inhibitor therapy plus the mTOR inhibitor everolimus in postmenopausal women with hormone receptor positive/HER2 negative breast cancer and an oncotype Dx recurrence score (≤25)

Background: Breast cancer (BC) patients (pts) with Oncotype Dx recurrence scores (RS) ≤ 25 represent a large fraction of BC cases. The TAILORx study demonstrated that these pts did not receive significant benefit from adjuvant chemotherapy. Despite their low to moderate risk and sensitivity to endocrine therapy, a number of BC pts may benefit from neoadjuvant therapy. The goal of this study is to assess tumor response, utilizing the preoperative endocrine prognostic index (PEPI), to a combination of the mTOR inhibitor everolimus and an aromatase inhibitor (AI) in this pt population. Methods: This is a phase II study evaluating the efficacy and safety of neoadjuvant AI and everolimus in postmenopausal pts with hormone receptor positive (HR+)/HER2 negative clinical stage II-III BC with low risk RS (≤ 25). Patient enrollment initiated in November 2014 at the Yale Cancer Center/Smilow Cancer Hospital and Care Centers. Key inclusion criteria are ECOG 0-2, adequate organ function, a fasting cholesterol ≤ 300 mg/dl and triglycerides ≤ 2.5 x IULN. Eligible pts received daily AI therapy (anastrozole 1 mg, letrozole 2.5 mg, or exemestane 25 mg) and everolimus 10 mg daily for up to a total of 26 weeks. The primary objective of the study was to determine the percent of postmenopausal pts with clinical stage II-III HR+/HER2- BC and a RS ≤ 25 who achieve a PEPI score of 0 following neoadjuvant AI and everolimus. The secondary objectives are to assess the tolerability and side effect profile, and to identify biologic markers predictive of a PEPI 0. Simon’s optimal two-stage design was utilized with a planned sample size of 27 eligible pts. First, 15 pts will be enrolled, if 5 or more of 15 eligible patients achieve a PEPI 0, the plan is to enroll another 12 pts. If 10 or more of the 27 eligible patients achieve a PEPI 0, we will conclude that the regimen warrants further study. This design has a power of 80% and a one-sided significance level of 0.1. Results: Of the 17 pts initially enrolled, 15 were evaluable for response; 4 of 15 (26%) had a PEPI scores of 0 which did not meet the primary endpoint; 4 (26%) had a path CR and 6 (40%) had PR. Grade 3 toxicities determined to be possibly/probably study related included anemia, anorexia, hypertension, maculopapular rash and hyperglycemia. One pt developed grade 3 atrial flutter and grade 4 QT prolongation which required a dose delay. One pt required hospitalization for pneumonia. Gene expression analysis by RNA seq was performed on 9 baseline (6 pts with CR/PR vs 3 non-responders); 13 post treatment (8 pts with CR/PR vs 5 non-responders) samples with 8 matched pairs (5 pts with CR/PR vs 3 non-responders). Baseline samples from pts who had CR/PR have significantly lower expression of MYC targets and oxidative phosphorylation genes, and significantly higher expression of genes associated with epithelial mesenchymal transition and interferon signaling, in comparison to non-responders. The expression of CYP19A1 gene, that codes aromatase, is significantly increased after therapy in samples from pts with CR/PR (logFC=1.33), but not in other pts. Expression of mTORC1 signaling pathway genes is increased after therapy only in samples from pts with CR/PR. Conclusion: Although this trial did not meet the set primary endpoint, 26% of pts achieved a PEPI 0. The combination of an AI and everolimus was overall well tolerated. Our study suggested that baseline expression levels of key pathways were associated with response to neoadjuvant AI plus everolimus.

Citation Format: Maysa M Abu-Khalaf, Kimberly Aderhold, Michal Marczyk, Gina Chung, Erin Hofstatter, Tara Sanft, Andrea Silber, Michael DiGiovanna, Daniel Zelterman, Lajos Puzstai, Christos Hatzis. Neoadjuvant aromatase inhibitor therapy plus the mTOR inhibitor everolimus in postmenopausal women with hormone receptor positive/HER2 negative breast cancer and an oncotype Dx recurrence score (≤25) [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P5-13-02.

GaMRed-Adaptive Filtering of High-Throughput Biological Data

Data filtering based on removing non-informative features, with unchanged signal between compared experimental conditions, can significantly increase sensitivity of methods used to detect differentially expressed genes or other molecular components measured in high-throughput biological experiments. Criteria for data filtering can be stated on the basis of averages or variances of signal levels across samples. The crucial parts of feature filtering are selection of filter type and cut-off threshold, which are specific to the particular dataset. In this paper, we present an algorithm and a stand-alone application, GaMRed, for adaptive filtering insignificant features in high-throughput data, based on Gaussian mixture decomposition. We have tested the performance of our algorithm using datasets from three different high-throughput biological experiments. We estimated the number of differentially expressed features after applying multiple testing correction and performed functional analysis of obtained features using Gene Ontology terms. Also, we checked if the control of false discovery rate and family-wise error rate after applying feature filtering remains at appropriate level. GaMRed is fast, automatic, and does not require expert knowledge in parameter tuning. The algorithm increases sensitivity of methods used to find differentially expressed features and biological validity of the findings. The program can be downloaded from: http://zaed.aei.polsl.pl/index.php/pl/oprogramowanie-zaed.

Gene set enrichment for reproducible science: comparison of CERNO and eight other algorithms

MOTIVATION

Analysis of gene set (GS) enrichment is an essential part of functional omics studies. Here, we complement the established evaluation metrics of GS enrichment algorithms with a novel approach to assess the practical reproducibility of scientific results obtained from GS enrichment tests when applied to related data from different studies.

RESULTS

We evaluated eight established and one novel algorithm for reproducibility, sensitivity, prioritization, false positive rate and computational time. In addition to eight established algorithms, we also included Coincident Extreme Ranks in Numerical Observations (CERNO), a flexible and fast algorithm based on modified Fisher P-value integration. Using real-world datasets, we demonstrate that CERNO is robust to ranking metrics, as well as sample and GS size. CERNO had the highest reproducibility while remaining sensitive, specific and fast. In the overall ranking Pathway Analysis with Down-weighting of Overlapping Genes, CERNO and over-representation analysis performed best, while CERNO and GeneSetTest scored high in terms of reproducibility.

AVAILABILITY AND IMPLEMENTATION

tmod package implementing the CERNO algorithm is available from CRAN (cran.r-project.org/web/packages/tmod/index.html) and an online implementation can be found at http://tmod.online/. The datasets analyzed in this study are widely available in the KEGGdzPathwaysGEO, KEGGandMetacoreDzPathwaysGEO R package and GEO repository.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

The impact of RNA extraction method on accurate RNA sequencing from formalin-fixed paraffin-embedded tissues

Background

Utilization of RNA sequencing methods to measure gene expression from archival formalin-fixed paraffin-embedded (FFPE) tumor samples in translational research and clinical trials requires reliable interpretation of the impact of pre-analytical variables on the data obtained, particularly the methods used to preserve samples and to purify RNA.

Methods

Matched tissue samples from 12 breast cancers were fresh frozen (FF) and preserved in RNAlater or fixed in formalin and processed as FFPE tissue. Total RNA was extracted and purified from FF samples using the Qiagen RNeasy kit, and in duplicate from FFPE tissue sections using three different kits (Norgen, Qiagen and Roche). All RNA samples underwent whole transcriptome RNA sequencing (wtRNAseq) and targeted RNA sequencing for 31 transcripts included in a signature of sensitivity to endocrine therapy. We assessed the effect of RNA extraction kit on the reliability of gene expression levels using linear mixed-effects model analysis, concordance correlation coefficient (CCC) and differential analysis. All protein-coding genes in the wtRNAseq and three gene expression signatures for breast cancer were assessed for concordance.

Results

Despite variable quality of the RNA extracted from FFPE samples by different kits, all had similar concordance of overall gene expression from wtRNAseq between matched FF and FFPE samples (median CCC 0.63–0.66) and between technical replicates (median expression difference 0.13–0.22). More than half of genes were differentially expressed between FF and FFPE, but with low fold change (median |LFC| 0.31–0.34). Two out of three breast cancer signatures studied were highly robust in all samples using any kit, whereas the third signature was similarly discordant irrespective of the kit used. The targeted RNAseq assay was concordant between FFPE and FF samples using any of the kits (CCC 0.91–0.96).

Conclusions

The selection of kit to purify RNA from FFPE did not influence the overall quality of results from wtRNAseq, thus variable reproducibility of gene signatures probably relates to the reliability of individual gene selected and possibly to the algorithm. Targeted RNAseq showed promising performance for clinical deployment of quantitative assays in breast cancer from FFPE samples, although numerical scores were not identical to those from wtRNAseq and would require calibration.

Abstract P4-03-01: Pathway level complementarity of germline and somatic events in breast cancer

Background: Progression from a normal cell state to cancer requires multiple genomic hits in key regulatory pathways. In the case of hereditary cancer syndromes, some of these hits occur in the germline, but additional somatic mutations are required for malignant transformation. We hypothesize that this paradigm could be extended to sporadic cancers as well. What somatic mutation function as a cancer driver event may be determined by the constellation of germline variants a person is born with. We propose that even rare, non-recurrent, high functional impact germline variants in genes involved in cancer-related pathways could influence the biological impact of somatic mutations in other cancer-related genes. The goal of the current analysis was to examine associations between pathway alterations caused by high functional impact germline variants or somatic mutations in the “hallmarks of cancer” pathways in breast cancer.

Methods: We obtained germline DNA sequencing and copy number variation (CNV) data from the breast cancer TCGA cohort. After population clustering with the HapMap cohort, we selected a homogeneous group of 796 patients of Western European ancestry and downloaded the matching somatic mutations (SNVs and INDELs) that were available for 750 cases, that comprise the current study population. Germline CNVs were classified as recurrent or rare losses or gains. Potentially pathogenic germline variants (SNPs) were obtained from the PanCancer Altas project. All germline or somatic mutations were mapped at the gene level to the 50 Cancer Hallmarks pathway collection. We designated a pathway mutated if at least 1 gene had a germline or a somatic mutation. Complementarity between pathway alterations by germline and somatic events were evaluated using the Fisher exact test adjusted for multiple comparisons.

Results: At the germline level, 2,057 genes were affected by CNVs (mean 30, range 3-151 genes/patient), and a total of 43 genes carried germline pathogenic SNPs that affected 13.8% of the patients. At the somatic level, we detected 40,881 high functional impact mutations (mean 54.3, range 1-3889 mutations/patient) in 13,080 genes (mean 50.8, range 1-3166 genes/patient). The 50 Cancer Hallmark pathways contained 4386 genes (mean 146.5, range 32-200 genes/pathway), and were mutated in the majority of the patients (85% germline, 93% somatic). Several pathways, such as HEME_METABOLISM, INTERFERON_ALPHA_RESPONSE, and KRAS_SIGNALING, were frequently affected by germline alterations, while the somatic mutations were most frequently involved in the COMPLEMENT, E2F_TARGET, and UV_RESPONSE_UP. Interaction analysis revealed co-occurrence between MYC_TARGETS_V1 (germline) and UV_RESPONSE_DN (somatic) or MTORC1_SINGALING (somatic) (p&lt;0.01), and TNFA_SIGNALING_VIA_NFKB (germline) and IL6_JAK_STAT3_SIGNALING (germline) with E2F_TARGETS (somatic) (p&lt;0.01). We also observed an exclusive relationship between germline alterations in BILE_ACID_METABOLISM and somatic mutations in COMPLEMENT pathway (p&lt;0.01).

Conclusions: Our results highlight the importance of pathway-level analysis of germline alterations in breast cancer, which might help to understand the interrelationship between germline and somatic alterations in breast cancer.

Citation Format: Qing T, Marczyk M, Wali V, Gunasekharan V, Patwardhan G, Pusztai L, Hatzis C. Pathway level complementarity of germline and somatic events in breast cancer [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P4-03-01.

Abstract P2-06-06: Targeting loss of isoenzyme diversity as a novel therapeutic strategy in breast cancer

Background: Several metabolic steps are mediated by distinct proteins or isoenzymes that catalyze the same reaction, providing redundancy of metabolic functions. Metabolic states are frequently altered in cancer to support survival and proliferation in hypoxic and otherwise hostile microenvironments, and metabolic re-wiring often involve loss of isoenzyme diversity. We hypothesize that targeting enzymes that have lost isoenzyme diversity in cancer, but not in normal cells, provides an opportunity to selectively target cancers. In this study, we assessed mRNA expression of all known human isoenzyme families in breast cancer and normal breast tissue and identified isoenzymes with loss of diversity within each breast cancer subtype.

Methods: We obtained RNAseq data from cancer and patient-matched normal breast tissues from the TCGA (N=66 HR+, N=24 HER2+, and N=15 TNBC tumors). We retrieved annotated human isoenzyme families from the ENZYME nomenclature database. We compared expression in cancer and matched normal samples from the same patient to identify isoenzymes that had i) same or increased expression of the target isoenzyme in cancer vs normal and ii) reduced expression of the complementary isoenzymes in cancer. We developed five scores that capture various elements of these characteristics and prioritized candidates as targets based on clustering and their combined ranking based on the five scores. We validated overexpression of the candidate isoenzymes relative to other isoforms in breast cancer microarray data from ArrayExpress (E-GEOD-76250: 33 TNBC, and E-GEOD-70951: 30 TNBC, 108 HR+, 10 HER2+).

Results: We identified 321 enzymes in the TCGA discovery cohort that correspond to 829 unique isoenzymes. Overall, 636, 483 and 429 isoenzymes were differentially expressed in HR+, HER2+ and TNBC cancers, respectively, compared to corresponding normal samples. Of these, 308 isoenzymes were differentially expressed relative to normal in all 3 subtypes. In all, 112 and 92, and 84 were selected as candidate isoenzyme therapeutic targets in HR+, HER2+ and TNBC, respectively. 23 isoenzymes prioritized in clustering step were further validated. Finally, 6 isoenzymes were validated in HR+ (ALDOA, GUSB, GYG1, MIF, P3H1, PCK2), 10 in HER2+ (ALDH1L2, ALDOA, GLYATL2, GUSB, GYG1, GYS1, MIF, P3H1, PCK2, PTGS1) and 12 in TNBC (ADSS, ALAS1, ALDH1L2, ALDOA, ART3, GLYATL2, GUSB, GYS1, HS3ST1, MIF, PCK2, SOAT1), as potential targets for breast cancer treatment. Of these, 5 potential isoenzyme targets (ALDH1L2, GUSB, GLYATL2, MIF, PCK2), which were mostly hydrolases and transferases, were further selected for ongoing experimental validation in the laboratory. Decreased expression of the complementary isoforms of these 5 targets were primarily due to DNA methylation of the genes in cancer.

Conclusions: We found that loss of isoenzyme diversity is a broad phenomenon in breast cancers that may be explored therapeutically. We identified several instances of “isoenzyme addiction” in which cancers depend exclusively on a single isoenzyme while downregulating via methylation the complementary isoenzymes, providing cancer-specific targeting opportunities. We are currently validating several of these targets in cell line models.

Citation Format: Marczyk M, Gunasekharan V, Wali VB, Shi W, Patwardhan G, Qing T, Pusztai L, Hatzis C. Targeting loss of isoenzyme diversity as a novel therapeutic strategy in breast cancer [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P2-06-06.

Abstract P4-08-20: Pre-analytical effects of FFPE extraction methods on targeted and whole transcriptome sequencing assays for endocrine sensitivity in metastatic breast cancer

Background: The clinical management of patients with metastatic HR-positive breast cancer is often uncertain due to decreased sensitivity to anti-estrogen therapy over time. Recently, we developed a targeted RNAseq based 18-transcript SET ER/PR assay of endocrine sensitivity from biopsies of metastatic cancer. In this work we assess the effect of pre-analytical factors, specifically RNA extraction methods for FFPE tissue samples, on the reliability of the targeted RNAseq assay.

Methods: FFPE blocks and matched fresh frozen (FF) sections from 12 tumors were collected at MD Anderson Cancer Center. RNA from FFPE slides was extracted in duplicate using three kits (Norgen, Qiagen, Roche), and RNAseq libraries from all samples were prepared using Kapa Total RNAseq kit. Targeted RNA libraries were prepared using droplet-based PCR (RainDance), and also by transcriptome-wide RNAseq for comparison. Reads were mapped to genomic sequence using STAR and expression was quantified using RSEM. Expression data were normalized based on expression of 10 reference genes. The effect of FFPE RNA extraction kit on the reliability of the SET index was assessed using linear mixed effects model (LME) analysis, and agreement with FF was assessed using the concordance correlation coefficient (CCC).

Results: Analysis of the whole transcriptome RNAseq data confirmed minimal 3'-end transcript bias from FFPE samples, irrespective of transcript size or FFPE kit. All 18 genes included in the SET index had high overall concordance between FFPE and FF (median CCC percentile=98.8, range 57.2-99.9 for Norgen; similar for the other two kits) and relatively consistent bias across genes, as estimated by the random effects of the LME model. Furthermore, compared to random 18-gene indices, concordance in the SET index values between FF and FFPE was higher than 99.8% of the random samples, verifying the analytical reliability of the selected genes. For the targeted RNAseq assay, RNA from FFPE extracted with the Norgen kit showed the highest concordance compared to FF (CCC=0.956, 95%CI 0.871-0.985). In general, the analytical variation of SET from FFPE samples was greater than that from FF (1.71-2.71 fold greater), with the lowest variation associated with the Norgen kit. The SET index values from targeted RNAseq for both FF and FFPE samples were consistently lower compared to transcriptome-wide RNAseq but were highly correlated, with the Norgen kit having the highest correlation between targeted and transcriptome-wide RNAseq (rho=0.915).

Conclusions: All three FFPE RNA extraction kits have excellent analytical performance compared to FF samples. The Norgen kit may be marginally better yielding higher concordance with FF and lower analytical variation between replicates. All genes in the SET ER/PR showed very good analytical performance in comparison to random indices and individual genes. Targeted gene RNA sequencing appears very promising as a platform for clinical deployment of quantitative assays, showing only a small (fixable) bias compared to RNAseq.

Citation Format: Marczyk M, Fu C, Lau R, Du L, Trevarton AJ, Sinn BV, Gould RE, Symmans WF, Hatzis C. Pre-analytical effects of FFPE extraction methods on targeted and whole transcriptome sequencing assays for endocrine sensitivity in metastatic breast cancer [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P4-08-20.

Stage I non-small-cell lung cancer: long-term results of lobectomy versus sublobar resection from the Polish National Lung Cancer Registry

OBJECTIVES

Anatomical lobar resection and mediastinal lymphadenectomy remain the standard for the treatment of early stage non-small-cell lung cancer (NSCLC) and are preferred over procedures such as segmentectomy or wedge resection. However, there is an ongoing debate concerning the influence of the extent of the resection on overall survival. The aim of this article was to assess the overall survival for different types of resection for Stage I NSCLC.

METHODS

We performed a retrospective analysis of the results of the surgical treatment of Stage I NSCLC. Between 1 January 2007 and 31 December 2013, the data from 6905 patients who underwent Stage I NSCLC operations were collected in the Polish National Lung Cancer Registry (PNLCR) and overall survival was assessed. A propensity score-matched analysis was used to compare 3 groups of patients, each consisting of 231 patients who underwent lobectomy, segmentectomy, or wedge resection.

RESULTS

In the unmatched and matched patient groups, lobectomy and segmentectomy were associated with a significant benefit compared to wedge resection regarding overall survival (log-rank P  < 0.001 and P  = 0.001). The Cox proportional hazard ratio comparing segmentectomy and lobectomy to wedge resection was 0.54 [95% confidence interval (CI): 0.37-0.77) and 0.44 (95% CI: 0.38-0.50), respectively, indicating a significant improvement in survival. There was no difference in the 5-year survival of patients after lobectomy (79.1%; 95% CI: 77.7-80.4%) or segmentectomy (78.3%; 95% CI: 70.6-86.0%). The 30-day mortality rate was 1.6, 2.6 and 1.4% for lobectomy, segmentectomy and wedge resection, respectively. Wedge resection was associated with a significantly lower 5-year survival rate (58.1%; 95% CI: 53.6-62.5%) compared to segmentectomy (78.3%; 95% CI: 70.6-86.0%) and lobectomy (79.1%; 95% CI: 77.7-80.5%). The propensity score matched analysis confirmed most of the results of the comparisons of unmatched study groups.

CONCLUSIONS

Wedge resection was associated with significantly lower 3-year and 5-year survival rates compared to the other methods of resection. There was no significant difference in 3-year or 5-year survival rates between lobectomy and segmentectomy. Segmentectomy, but not wedge resection, could be considered an alternative to lobectomy in the treatment of patients with Stage I NSCLC.

Ranking metrics in gene set enrichment analysis: do they matter?

Background

There exist many methods for describing the complex relation between changes of gene expression in molecular pathways or gene ontologies under different experimental conditions. Among them, Gene Set Enrichment Analysis seems to be one of the most commonly used (over 10,000 citations). An important parameter, which could affect the final result, is the choice of a metric for the ranking of genes. Applying a default ranking metric may lead to poor results.

Methods and results

In this work 28 benchmark data sets were used to evaluate the sensitivity and false positive rate of gene set analysis for 16 different ranking metrics including new proposals. Furthermore, the robustness of the chosen methods to sample size was tested. Using k-means clustering algorithm a group of four metrics with the highest performance in terms of overall sensitivity, overall false positive rate and computational load was established i.e. absolute value of Moderated Welch Test statistic, Minimum Significant Difference, absolute value of Signal-To-Noise ratio and Baumgartner-Weiss-Schindler test statistic. In case of false positive rate estimation, all selected ranking metrics were robust with respect to sample size. In case of sensitivity, the absolute value of Moderated Welch Test statistic and absolute value of Signal-To-Noise ratio gave stable results, while Baumgartner-Weiss-Schindler and Minimum Significant Difference showed better results for larger sample size. Finally, the Gene Set Enrichment Analysis method with all tested ranking metrics was parallelised and implemented in MATLAB, and is available at https://github.com/ZAEDPolSl/MrGSEA.

Conclusions

Choosing a ranking metric in Gene Set Enrichment Analysis has critical impact on results of pathway enrichment analysis. The absolute value of Moderated Welch Test has the best overall sensitivity and Minimum Significant Difference has the best overall specificity of gene set analysis. When the number of non-normally distributed genes is high, using Baumgartner-Weiss-Schindler test statistic gives better outcomes. Also, it finds more enriched pathways than other tested metrics, which may induce new biological discoveries.

Electronic supplementary material

The online version of this article (doi:10.1186/s12859-017-1674-0) contains supplementary material, which is available to authorized users.

Mixture Modeling of 2-D Gel Electrophoresis Spots Enhances the Performance of Spot Detection

2-D gel electrophoresis is the most commonly used method in biomedicine to separate even thousands of proteins in a complex sample on a single gel. Even though the technique is quite known, there is still a need to find an efficient and reliable method for detection of protein spots on gel image. In this paper, a three-step algorithm based on mixture of 2-D normal distribution functions is introduced to improve the efficiency of spot detection performed by the existing algorithms, namely Pinnacle software and watershed segmentation method. Comparison of methods is based on using simulated and real data sets with known true spot positions and different number of spots. Fitting a mixture of components to gel image allows for achieving higher sensitivity in detecting spots, regardless the method used to find initial conditions for the model parameters, and it leads to better overall performance of spot detection. By using mixture model, location of spot centers can be estimated with higher accuracy than using the Pinnacle method. An application of spot shape modeling gives higher sensitivity of obtaining low-intensity spots than the watershed method, which is crucial in the discovery of novel biomarkers.

Serum mass profile signature as a biomarker of early lung cancer

OBJECTIVES

Circulating molecular biomarkers of lung cancer may allow the pre-selection of candidates for computed tomography screening or increase its efficacy. We aimed to identify features of serum mass profile distinguishing individuals with early lung cancer from healthy participants of the lung cancer screening program.

METHODS

Blood samples were collected during a low-dose computed tomography (LD-CT) screening program performed by one institution (Medical University of Gdansk, Poland). MALDI-ToF mass spectrometry was used to characterize the low-molecular-weight (1000-14,000Da) serum fraction. The analysis comprised 95 patients with early stage lung cancer (including 30 screen-detected cases) and a matched group of 285 healthy controls. The cases were split into two independent cohorts (discovery and validation), analyzed separately 6 months apart.

RESULTS

Several molecular components of serum (putatively components of endogenous peptidome) discriminating patients with early lung cancer from controls were identified in a discovery cohort. This allowed building an effective cancer classifier as a model tuned to maximize negative predictive value, with an area under the curve (AUC) of 0.88, a negative predictive value of 100%, and a positive predictive value of 48%. However, the classifier performed worse in a validation cohort including independent sample sets (AUC 0.73, NPV 88% and PPV 30%).

CONCLUSIONS

We developed a serum mass profile-based signature identifying patients with early lung cancer. Although this marker has insufficient value as a stand-alone preselecting tool for LD-CT screening, its potential clinical usefulness in evaluation of indeterminate pulmonary nodules deserves further investigation.

Signal Partitioning Algorithm for Highly Efficient Gaussian Mixture Modeling in Mass Spectrometry

Mixture - modeling of mass spectra is an approach with many potential applications including peak detection and quantification, smoothing, de-noising, feature extraction and spectral signal compression. However, existing algorithms do not allow for automated analyses of whole spectra. Therefore, despite highlighting potential advantages of mixture modeling of mass spectra of peptide/protein mixtures and some preliminary results presented in several papers, the mixture modeling approach was so far not developed to the stage enabling systematic comparisons with existing software packages for proteomic mass spectra analyses. In this paper we present an efficient algorithm for Gaussian mixture modeling of proteomic mass spectra of different types (e.g., MALDI-ToF profiling, MALDI-IMS). The main idea is automated partitioning of protein mass spectral signal into fragments. The obtained fragments are separately decomposed into Gaussian mixture models. The parameters of the mixture models of fragments are then aggregated to form the mixture model of the whole spectrum. We compare the elaborated algorithm to existing algorithms for peak detection and we demonstrate improvements of peak detection efficiency obtained by using Gaussian mixture modeling. We also show applications of the elaborated algorithm to real proteomic datasets of low and high resolution.