Search

Purpose: To develop a deep learning-based algorithm that automatically and accurately classifies patients as either having pulmonary emboli or not in CT pulmonary angiography (CTPA) examinations. Materials and methods: For model development, 700 CTPA examinations from 652 patients performed at a single institution were used, of which 149 examinations contained 1497 PE traced by radiologists. The nnU-Net deep learning-based segmentation framework was trained using 5-fold cross-validation. To enhance classification, we applied logical rules based on PE volume and probability thresholds. External model evaluation was performed in 770 and 34 CTPAs from two independent datasets. Results: A total of 1483 CTPA examinations were evaluated. In internal cross-validation and test set, the trained model correctly classified 123 of 128 examinations as positive for PE (sensitivity 96.1 %; 95 % C.I. 91–98 %; P

Summary: Background: Loss of heterozygosity (LOH) diminishes genetic diversity within cancer genomes. A tumour arising in an individual heterozygous for a functional and a loss-of-function (LoF) allele of a gene occasionally retain only the LoF allele. This can result in deficiency of specific protein activities in cancer cells, creating unique differences between tumour cells and normal cells of the individual. Such differences may constitute vulnerabilities that can be exploited through allele-specific therapies. Methods: To discover frequently lost genes with prevalent LoF alleles, we mined the 1000 Genomes dataset for SNVs causing protein truncation through base substitution, indels or splice site disruptions, resulting in 60 LoF variants in 60 genes. From these, the variant rs3892097 in the liver enzyme CYP2D6 was selected because it is located within a genomic region that frequently undergoes LOH in several tumor types including hepatocellular cancers. To evaluate the relationship between CYP2D6 activity and the toxicities of anticancer agents, we screened 525 compounds currently in clinical use or undergoing clinical trials using cell model systems with or without CYP2D6 activity. Findings: We identified 12 compounds, AZD-3463, CYC-116, etoposide, everolimus, GDC-0349, lenvatinib, MK-8776, PHA-680632, talazoparib, tyrphostin 9, VX-702, and WZ-3146, using an engineered HEK293T cell model. Of these, talazoparib and MK-8776 demonstrated consistently heightened cytotoxic effects against cells with compromised CYP2D6 activity in engineered hepatocellular cancer cell models. Moreover, talazoparib displayed CYP2D6 genotype dependent effects on primary hepatocellular carcinoma organoids. Interpretation: Exploiting the loss of drug-metabolizing enzyme gene activity in tumor cells following loss of heterozygosity could present a promising therapeutic strategy for targeted cancer treatment. Funding: This work was funded by Barncancerfonden (T.S, PR2022-0099 and PR2020-0171, X.Z, TJ2021-0111), Cancerfonden (T.S, 211719Pj and D.G, 222449Pj), Vetenskapsrådet (T.S, 2020-02371 and D.G, 2020-04707), and the Erling Persson Foundation (T.S, 2020-0037 and T.S, 2023-0113).

Abstract Background Ephrin (EPH) receptors have been implicated in tumorigenesis and metastasis, but the functional understanding of mutations observed in human cancers is limited. We previously demonstrated reduced cell compartmentalisation for somatic EPHB1 mutations found in metastatic colorectal cancer cases. We therefore integrated pan-cancer and pan-EPH mutational data to prioritise recurrent EPHB1 mutations for functional studies to understand their contribution to cancer development and metastasis. Methods Here, 79,151 somatic mutations in 9,898 samples of 33 different tumour types were analysed with a bioinformatic pipeline to find 3D-mutated cluster pairs and hotspot mutations in EPH receptors. From these, 15 recurring EPHB1 mutations were stably expressed in colorectal cancer followed by confocal microscopy based in vitro compartmentalisation assays and phospho-proteome analysis. Results The 3D-protein structure-based bioinformatics analysis resulted in 63% EPHB1 mutants with compartmentalisation phenotypes vs 43% for hotspot mutations. Whereas the ligand-binding domain mutations C61Y, R90C, and R170W, the fibronectin domain mutation R351L, and the kinase domain mutation D762N displayed reduced to strongly compromised cell compartmentalisation, the kinase domain mutations R743W and G821R enhanced this phenotype. While mutants with reduced compartmentalisation also had reduced ligand induced receptor phosphorylation, the enhanced compartmentalisation was not linked to receptor phosphorylation level. Phosphoproteome mapping pinpointed the PI3K pathway and PIK3C2B phosphorylation in cells harbouring mutants with reduced compartmentalisation. Conclusions This is the first integrative study of pan-cancer EPH receptor mutations followed by in vitro validation, a robust way to identify cancer-causing mutations, uncovering EPHB1 mutation phenotypes and demonstrating the utility of protein structure-based mutation analysis in characterization of novel cancer genes. Video Abstract

Abstract Mediastinal structure measurements are important for the radiologist’s review of computed tomography pulmonary angiography (CTPA) examinations. In the reporting process, radiologists make measurements of diameters, volumes, and organ densities for image quality assessment and risk stratification. However, manual measurement of these features is time consuming. Here, we sought to develop a time-saving automated algorithm that can accurately detect, segment and measure mediastinal structures in routine clinical CTPA examinations. In this study, 700 CTPA examinations collected and annotated. Of these, a training set of 180 examinations were used to develop a fully automated deterministic algorithm. On the test set of 520 examinations, two radiologists validated the detection and segmentation performance quantitatively, and ground truth was annotated to validate the measurement performance. External validation was performed in 47 CTPAs from two independent datasets. The system had 86–100% detection and segmentation accuracy in the different tasks. The automatic measurements correlated well to those of the radiologist (Pearson’s r 0.68–0.99). Taken together, the fully automated algorithm accurately detected, segmented, and measured mediastinal structures in routine CTPA examinations having an adequate representation of common artifacts and medical conditions.

Abstract Colorectal cancer (CRC) is the third most common cancer and the second leading cause of death in the world. In most cases, drug resistance and tumor recurrence are ultimately inevitable. One obstacle is the presence of chemotherapy-insensitive quiescent cancer cells (QCCs). Identification of unique features of QCCs may facilitate the development of new targeted therapeutic strategies to eliminate tumor cells and thereby delay tumor recurrence. Here, using single-cell RNA sequencing, we classified proliferating and quiescent cancer cell populations in the human colorectal cancer spheroid model and identified ATF3 as a novel signature of QCCs that could support cells living in a metabolically restricted microenvironment. RNA velocity further showed a shift from the QCC group to the PCC group indicating the regenerative capacity of the QCCs. Our further results of epigenetic analysis, STING analysis, and evaluation of TCGA COAD datasets build a conclusion that ATF3 can interact with DDIT4 and TRIB3 at the transcriptional level. In addition, decreasing the expression level of ATF3 could enhance the efficacy of 5-FU on CRC MCTS models. In conclusion, ATF3 was identified as a novel marker of QCCs, and combining conventional drugs targeting PCCs with an option to target QCCs by reducing ATF3 expression levels may be a promising strategy for more efficient removal of tumor cells.

Abstract A comprehensive characterization of blood proteome profiles in cancer patients can contribute to a better understanding of the disease etiology, resulting in earlier diagnosis, risk stratification and better monitoring of the different cancer subtypes. Here, we describe the use of next generation protein profiling to explore the proteome signature in blood across patients representing many of the major cancer types. Plasma profiles of 1463 proteins from more than 1400 cancer patients are measured in minute amounts of blood collected at the time of diagnosis and before treatment. An open access Disease Blood Atlas resource allows the exploration of the individual protein profiles in blood collected from the individual cancer patients. We also present studies in which classification models based on machine learning have been used for the identification of a set of proteins associated with each of the analyzed cancers. The implication for cancer precision medicine of next generation plasma profiling is discussed.

Summary: Background: Cancer immunity is based on the interaction of a multitude of cells in the spatial context of the tumour tissue. Clinically relevant immune signatures are therefore anticipated to fundamentally improve the accuracy in predicting disease progression. Methods: Through a multiplex in situ analysis we evaluated 15 immune cell classes in 1481 tumour samples. Single-cell and bulk RNAseq data sets were used for functional analysis and validation of prognostic and predictive associations. Findings: By combining the prognostic information of anti-tumoural CD8+ lymphocytes and tumour supportive CD68+CD163+ macrophages in colorectal cancer we generated a signature of immune activation (SIA). The prognostic impact of SIA was independent of conventional parameters and comparable with the state-of-art immune score. The SIA was also associated with patient survival in oesophageal adenocarcinoma, bladder cancer, lung adenocarcinoma and melanoma, but not in endometrial, ovarian and squamous cell lung carcinoma. We identified CD68+CD163+ macrophages as the major producers of complement C1q, which could serve as a surrogate marker of this macrophage subset. Consequently, the RNA-based version of SIA (ratio of CD8A to C1QA) was predictive for survival in independent RNAseq data sets from these six cancer types. Finally, the CD8A/C1QA mRNA ratio was also predictive for the response to checkpoint inhibitor therapy. Interpretation: Our findings extend current concepts to procure prognostic information from the tumour immune microenvironment and provide an immune activation signature with high clinical potential in common human cancer types. Funding: Swedish Cancer Society, Lions Cancer Foundation, Selanders Foundation, P.O. Zetterling Foundation, U-CAN supported by SRA CancerUU, Uppsala University and Region Uppsala.

Abstract Background Genes in the Ras pathway have somatic mutations in at least 60 % of colorectal cancers. Despite activating the same pathway, the BRAF V600E mutation and the prevalent mutations in codon 12 and 13 of KRAS have all been linked to different clinical outcomes, but the molecular mechanisms behind these differences largely remain to be clarified. Methods To characterize the similarities and differences between common activating KRAS mutations and between KRAS and BRAF mutations, we used genome editing to engineer KRAS G12C/D/V and G13D mutations in colorectal cancer cells that had their mutant BRAF V600E allele removed and subjected them to transcriptome sequencing, global proteomics and metabolomics analyses. Results By intersecting differentially expressed genes, proteins and metabolites, we uncovered (i) two-fold more regulated genes and proteins when comparing KRAS to BRAF mutant cells to those lacking Ras pathway mutation, (ii) five differentially expressed proteins in KRAS mutants compared to cells lacking Ras pathway mutation (IFI16, S100A10, CD44, GLRX and AHNAK2) and 6 (CRABP2, FLNA, NXN, LCP1, S100A10 and S100A2) compared to BRAF mutant cells, (iii) 19 proteins expressed differentially in a KRAS mutation specific manner versus BRAF V600E cells, (iv) regulation of the Integrin Linked Kinase pathway by KRAS but not BRAF mutation, (v) regulation of amino acid metabolism, particularly of the tyrosine, histidine, arginine and proline pathways, the urea cycle and purine metabolism by Ras pathway mutations, (vi) increased free carnitine in KRAS and BRAF mutant RKO cells. Conclusions This comprehensive integrative -omics analysis confirms known and adds novel genes, proteins and metabolic pathways regulated by mutant KRAS and BRAF signaling in colorectal cancer. The results from the new model systems presented here can inform future development of diagnostic and therapeutic approaches targeting tumors with KRAS and BRAF mutations.

Abstract Therapies targeting somatic bystander genetic events represent a new avenue for cancer treatment. We recently identified a subset of colorectal cancer (CRC) patients who are heterozygous for a wild-type and a low activity allele (NAT2*6) but lack the wild-type allele in their tumors due to loss of heterozygosity (LOH) at 8p22. These tumors were sensitive to treatment with a cytotoxic substrate of NAT2 (6-(4-aminophenyl)-N-(3,4,5-trimethoxyphenyl)pyrazin-2-amine, APA), and pointed to NAT2 loss being a therapeutically exploitable vulnerability of CRC tumors. To better estimate the total number of treatable CRC patients, we here determined whether tumor cells retaining also other NAT2 low activity variants after LOH respond to APA treatment. The prevalent low activity alleles NAT2*5 and NAT2*14, but not NAT2*7, were found to be low metabolizers with high sensitivity to APA. By analysis of two different CRC patient cohorts, we detected heterozygosity for NAT2 alleles targetable by APA, along with allelic imbalances pointing to LOH, in ~ 24% of tumors. Finally, to haplotype the NAT2 locus in tumor and patient-matched normal samples in a clinical setting, we develop and demonstrate a long-read sequencing based assay. In total, > 79.000 CRC patients per year fulfil genetic criteria for high sensitivity to a NAT2 LOH therapy and their eligibility can be assessed by clinical sequencing.

BackgroundKRAS mutations, present in over 40% of metastatic colorectal cancer (mCRC), are negative predictive factors for anti-EGFR therapy. Mutations in KRAS-G12C have a cysteine residue for which drugs have been developed. Published data on this specific mutation are conflicting; thus, we studied the frequency and clinical characteristics in a real-world and population-based setting.MethodsPatients from three Nordic population-based cohorts and the real-life RAXO-study were combined. RAS and BRAF tests were performed in routine healthcare, except for one cohort. The dataset consisted of 2,559 patients, of which 1,871 could be accurately classified as KRAS, NRAS, and BRAF-V600E. Demographics, treatments, and outcomes were compared using logistic regression. Overall survival (OS) was estimated with Kaplan–Meier, and differences were compared using Cox regression, adjusted for baseline factors.ResultsThe KRAS-G12C frequency was 2%–4% of all tested in the seven cohorts (mean 3%) and 4%–8% of KRAS mutated tumors in the cohorts (mean 7%). Metastasectomies and ablations were performed more often (38% vs. 28%, p = 0.040), and bevacizumab was added more often (any line 74% vs. 59%, p = 0.007) for patients with KRAS-G12C- vs. other KRAS-mutated tumors, whereas chemotherapy was given to similar proportions. OS did not differ according to KRAS mutation, neither overall (adjusted hazard ratio (HR) 1.03; 95% CI 0.74–1.42, reference KRAS-G12C) nor within treatment groups defined as “systemic chemotherapy, alone or with biologics”, “metastasectomy and/or ablations”, or “best supportive care”, RAS and BRAF wild-type tumors (n = 548) differed similarly to KRAS-G12C, as to other KRAS- or NRAS-mutated (n = 66) tumors.ConclusionsIn these real-life and population-based cohorts, there were no significant differences in patient characteristics and outcomes between patients with KRAS-G12C tumors and those with other KRAS mutations. This contrasts with the results of most previous studies claiming differences in many aspects, often with worse outcomes for those with a KRAS-G12C mutation, although not consistent. When specific drugs are developed, as for this mutation, differences in outcome will hopefully emerge.

Allelic losses occurring in cancer cells have been suggested as potential targets for therapy. Here, the authors show how recurring loss of heterozygosity of a drug metabolic gene in colorectal cancers can be exploited using a low molecular weight compound.

We show that N-acetyltransferase 2 (NAT2) loss of heterozygosity can be targeted in >4% of colorectal cancers with the use of a small molecule. We identify and describe the effect of a compound that impairs the growth of colorectal tumors with slow NAT2 activity by half when compared to wild-type.

Abstract Background The Ras pathway genes KRAS, BRAF, or ERBBs have somatic mutations in ~ 60% of human colorectal carcinomas. At present, it is unknown whether the remaining cases lack mutations activating the Ras pathway or whether they have acquired mutations in genes hitherto unknown to belong to the pathway. Methods To address the second possibility and extend the compendium of Ras pathway genes, we used genome-wide transposon mutagenesis of two human colorectal cancer cell systems deprived of their activating KRAS or BRAF allele to identify genes enabling growth in low glucose, a Ras pathway phenotype, when targeted. Results Of the 163 recurrently targeted genes in the two different genetic backgrounds, one-third were known cancer genes and one-fifth had links to the EGFR/Ras/MAPK pathway. When compared to cancer genome sequencing datasets, nine genes also mutated in human colorectal cancers were identified. Among these, stable knockdown of FOXO3, NCOA3, and TCF7L2 restored growth in low glucose but reduced MEK/MAPK phosphorylation, reduced anchorage-independent growth, and modulated expressions of GLUT1 and Ras pathway related proteins. Knockdown of NCOA3 and FOXO3 significantly decreased the sensitivity to cetuximab of KRAS mutant but not wild-type cells. Conclusions This work establishes a proof-of-concept that human cell-based genome-wide forward genetic screens can assign genes to pathways with clinical importance in human colorectal cancer.

Abstract Background The disco-interacting protein 2 homolog C (DIP2C) gene is an uncharacterized gene found mutated in a subset of breast and lung cancers. To understand the role of DIP2C in tumour development we studied the gene in human cancer cells. Methods We engineered human DIP2C knockout cells by genome editing in cancer cells. The growth properties of the engineered cells were characterised and transcriptome and methylation analyses were carried out to identify pathways deregulated by inactivation of DIP2C. Effects on cell death pathways and epithelial-mesenchymal transition traits were studied based on the results from expression profiling. Results Knockout of DIP2C in RKO cells resulted in cell enlargement and growth retardation. Expression profiling revealed 780 genes for which the expression level was affected by the loss of DIP2C, including the tumour-suppressor encoding CDKN2A gene, the epithelial-mesenchymal transition (EMT) regulator-encoding ZEB1, and CD44 and CD24 that encode breast cancer stem cell markers. Analysis of DNA methylation showed more than 30,000 sites affected by differential methylation, the majority of which were hypomethylated following loss of DIP2C. Changes in DNA methylation at promoter regions were strongly correlated to changes in gene expression, and genes involved with EMT and cell death were enriched among the differentially regulated genes. The DIP2C knockout cells had higher wound closing capacity and showed an increase in the proportion of cells positive for cellular senescence markers. Conclusions Loss of DIP2C triggers substantial DNA methylation and gene expression changes, cellular senescence and epithelial-mesenchymal transition in cancer cells.

On-site diagnostics technologies allow for rapid, cost-effective diagnosis with a particular importance for remote communities. Here the authors demonstrate the use of mobile phone based microscopy for targeted DNA sequencing andin situpoint mutation detection in tumours.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Loss of heterozygosity (LOH) is a common genetic event in the development of cancer. In certain tumor types, LOH can affect more than 20% of the genome, entailing loss of allelic variation in thousands of genes. This reduction of heterozygosity creates genetic differences between tumor and normal cells, providing opportunities for development of novel cancer therapies. Here, we review and summarize (1) mutations associated with LOH on chromosomes which have been shown to be promising biomarkers of cancer risk or the prediction of clinical outcomes in certain types of tumors; (2) loci undergoing LOH that can be targeted for development of novel anticancer drugs as well as (3) LOH in tumors provides up-and-coming possibilities to understand the underlying mechanisms of cancer evolution and to discover novel cancer vulnerabilities which are worth a further investigation in the near future.

Tissue localization of gene expression is increasingly important for accurate interpretation of large scale datasets from expression and mutational analyses. To this end, we have (1) developed a robust and scalable procedure for generation of mRNA hybridization probes, providing >95% first-pass success rate in probe generation to any human target gene and (2) adopted an automated staining procedure for analyses of formalin-fixed paraffin-embedded tissues and tissue microarrays. The in situ mRNA and protein expression patterns for genes with known as well as unknown tissue expression patterns were analyzed in normal and malignant tissues to assess procedure specificity and whether in situ hybridization can be used for validating novel antibodies. We demonstrate concordance between in situ transcript and protein expression patterns of the well-known pathology biomarkers KRT17, CHGA, MKI67, PECAM1 and VIL1, and provide independent validation for novel antibodies to the biomarkers BRD1, EZH2, JUP and SATB2. The present study provides a foundation for comprehensive in situ gene set or transcriptome analyses of human normal and tumor tissues.

The genomic revolution in oncology will entail mutational analyses of vast numbers of patient-matched tumor and normal tissue samples. This has meant an increased risk of patient sample mix up due to manual handling. Therefore, scalable genotyping and sample identification procedures are essential to pathology biobanks. We have developed an efficient alternative to traditional genotyping methods suited for automated analysis. By targeting 53 prevalent deletions and insertions found in human populations with fluorescent multiplex ligation dependent genome amplification, followed by separation in a capillary sequencer, a peak spectrum is obtained that can be automatically analyzed. 24 tumor-normal patient samples were successfully matched using this method. The potential use of the developed assay for forensic applications is discussed.

Comprehensive identification of the acquired mutations that cause common cancers will require genomic analyses of large sets of tumor samples. Typically, the tissue material available from tumor specimens is limited, which creates a demand for accurate template amplification. We therefore evaluated whether phi29-mediated whole genome amplification introduces false positive structural mutations by massive mate-pair sequencing of a normal human genome before and after such amplification. Multiple displacement amplification led to a decrease in clone coverage and an increase by two orders of magnitude in the prevalence of inversions, but did not increase the prevalence of translocations. While multiple strand displacement amplification may find uses in translocation analyses, it is likely that alternative amplification strategies need to be developed to meet the demands of cancer genomics.