Clinical validation of a cell-free DNA fragmentome assay for augmentation of lung cancer early detection.

Lung cancer screening via annual low-dose computed tomography (LDCT) has poor adoption. We conducted a prospective case-control study among 958 individuals eligible for lung cancer screening to develop a blood-based lung cancer detection test that when positive is followed by an LDCT. Changes in genome-wide cell-free DNA (cfDNA) fragmentation profiles (fragmentomes) in peripheral blood reflected genomic and chromatin characteristics of lung cancer. We applied machine learning to fragmentome features to identify individuals who were more or less likely to have lung cancer. We trained the classifier using 576 cases and controls from study samples, and then validated it in a held-out group of 382 cases and controls. The validation demonstrated high sensitivity for lung cancer, and consistency across demographic groups and comorbid conditions. Applying test performance to the screening eligible population in a five-year model with modest utilization assumptions suggested the potential to prevent thousands of lung cancer deaths.

Cell-Free DNA Fragmentomes in the Diagnostic Evaluation of Patients With Symptoms Suggestive of Lung Cancer.

BACKGROUND: The diagnostic workup of individuals suspected of having lung cancer can be complex and protracted because conventional symptoms of lung cancer have low specificity and sensitivity. RESEARCH QUESTION: Among individuals with symptoms of lung cancer, can a blood-based approach to analyze cell-free DNA (cfDNA) fragmentation (the DNA evaluation of fragments for early interception [DELFI] score) enhance evaluation for the possible presence of lung cancer? STUDY DESIGN AND METHODS: Adults were referred to Bispebjerg Hospital (Copenhagen, Denmark) for diagnostic evaluation of initial imaging anomalies and symptoms consistent with lung cancer. Numbers and types of symptoms were extracted from medical records. cfDNA from plasma samples obtained at the prediagnostic visit was isolated, sequenced, and analyzed for genome-wide cfDNA fragmentation patterns. The relationships among clinical presentation, cancer status, and DELFI score were examined. RESULTS: A total of 296 individuals were analyzed. Median DELFI scores were higher for those with lung cancer (n = 98) than those without cancer (n = 198; 0.94 vs 0.19; P < .001). In a multivariate model adjusted for age, smoking history, and presenting symptoms, the addition of the DELFI score improved the prediction of lung cancer for those who demonstrated symptoms (area under the receiver operating characteristic curve, 0.74-0.94). INTERPRETATION: The DELFI score distinguishes individuals with lung cancer from those without cancer better than suspicious symptoms do. These results represent proof-of-concept support that fragmentation-based biomarker approaches may facilitate diagnostic resolution for patients with concerning symptoms of lung cancer.

Detection and characterization of lung cancer using cell-free DNA fragmentomes.

Non-invasive approaches for cell-free DNA (cfDNA) assessment provide an opportunity for cancer detection and intervention. Here, we use a machine learning model for detecting tumor-derived cfDNA through genome-wide analyses of cfDNA fragmentation in a prospective study of 365 individuals at risk for lung cancer. We validate the cancer detection model using an independent cohort of 385 non-cancer individuals and 46 lung cancer patients. Combining fragmentation features, clinical risk factors, and CEA levels, followed by CT imaging, detected 94% of patients with cancer across stages and subtypes, including 91% of stage I/II and 96% of stage III/IV, at 80% specificity. Genome-wide fragmentation profiles across ~13,000 ASCL1 transcription factor binding sites distinguished individuals with small cell lung cancer from those with non-small cell lung cancer with high accuracy (AUC = 0.98). A higher fragmentation score represented an independent prognostic indicator of survival. This approach provides a facile avenue for non-invasive detection of lung cancer.

Bayesian copy number detection and association in large-scale studies.

BACKGROUND: Germline copy number variants (CNVs) increase risk for many diseases, yet detection of CNVs and quantifying their contribution to disease risk in large-scale studies is challenging due to biological and technical sources of heterogeneity that vary across the genome within and between samples. METHODS: We developed an approach called CNPBayes to identify latent batch effects in genome-wide association studies involving copy number, to provide probabilistic estimates of integer copy number across the estimated batches, and to fully integrate the copy number uncertainty in the association model for disease. RESULTS: Applying a hidden Markov model (HMM) to identify CNVs in a large multi-site Pancreatic Cancer Case Control study (PanC4) of 7598 participants, we found CNV inference was highly sensitive to technical noise that varied appreciably among participants. Applying CNPBayes to this dataset, we found that the major sources of technical variation were linked to sample processing by the centralized laboratory and not the individual study sites. Modeling the latent batch effects at each CNV region hierarchically, we developed probabilistic estimates of copy number that were directly incorporated in a Bayesian regression model for pancreatic cancer risk. Candidate associations aided by this approach include deletions of 8q24 near regulatory elements of the tumor oncogene MYC and of Tumor Suppressor Candidate 3 (TUSC3). CONCLUSIONS: Laboratory effects may not account for the major sources of technical variation in genome-wide association studies. This study provides a robust Bayesian inferential framework for identifying latent batch effects, estimating copy number, and evaluating the role of copy number in heritable diseases.

PatternMarkers & GWCoGAPS for novel data-driven biomarkers via whole transcriptome NMF.

SUMMARY: Non-negative Matrix Factorization (NMF) algorithms associate gene expression with biological processes (e.g. time-course dynamics or disease subtypes). Compared with univariate associations, the relative weights of NMF solutions can obscure biomarkers. Therefore, we developed a novel patternMarkers statistic to extract genes for biological validation and enhanced visualization of NMF results. Finding novel and unbiased gene markers with patternMarkers requires whole-genome data. Therefore, we also developed Genome-Wide CoGAPS Analysis in Parallel Sets (GWCoGAPS), the first robust whole genome Bayesian NMF using the sparse, MCMC algorithm, CoGAPS. Additionally, a manual version of the GWCoGAPS algorithm contains analytic and visualization tools including patternMatcher, a Shiny web application. The decomposition in the manual pipeline can be replaced with any NMF algorithm, for further generalization of the software. Using these tools, we find granular brain-region and cell-type specific signatures with corresponding biomarkers in GTEx data, illustrating GWCoGAPS and patternMarkers ascertainment of data-driven biomarkers from whole-genome data. AVAILABILITY AND IMPLEMENTATION: PatternMarkers & GWCoGAPS are in the CoGAPS Bioconductor package (3.5) under the GPL license. CONTACT: gsteinobrien@jhmi.edu or ccolantu@jhmi.edu or ejfertig@jhmi.edu. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Genome-Wide Association of Copy Number Polymorphisms and Kidney Function.

Genome-wide association studies (GWAS) using single nucleotide polymorphisms (SNPs) have identified more than 50 loci associated with estimated glomerular filtration rate (eGFR), a measure of kidney function. However, significant SNPs account for a small proportion of eGFR variability. Other forms of genetic variation have not been comprehensively evaluated for association with eGFR. In this study, we assess whether changes in germline DNA copy number are associated with GFR estimated from serum creatinine, eGFRcrea. We used hidden Markov models (HMMs) to identify copy number polymorphic regions (CNPs) from high-throughput SNP arrays for 2,514 African (AA) and 8,645 European ancestry (EA) participants in the Atherosclerosis Risk in Communities (ARIC) study. Separately for the EA and AA cohorts, we used Bayesian Gaussian mixture models to estimate copy number at regions identified by the HMM or previously reported in the HapMap Project. We identified 312 and 464 autosomal CNPs among individuals of EA and AA, respectively. Multivariate models adjusted for SNP-derived covariates of population structure identified one CNP in the EA cohort near genome-wide statistical significance (Bonferroni-adjusted p = 0.067) located on chromosome 5 (876-880kb). Overall, our findings suggest a limited role of CNPs in explaining eGFR variability.

Whole exome association of rare deletions in multiplex oral cleft families.

By sequencing the exomes of distantly related individuals in multiplex families, rare mutational and structural changes to coding DNA can be characterized and their relationship to disease risk can be assessed. Recently, several rare single nucleotide variants (SNVs) were associated with an increased risk of nonsyndromic oral cleft, highlighting the importance of rare sequence variants in oral clefts and illustrating the strength of family-based study designs. However, the extent to which rare deletions in coding regions of the genome occur and contribute to risk of nonsyndromic clefts is not well understood. To identify putative structural variants underlying risk, we developed a pipeline for rare hemizygous deletions in families from whole exome sequencing and statistical inference based on rare variant sharing. Among 56 multiplex families with 115 individuals, we identified 53 regions with one or more rare hemizygous deletions. We found 45 of the 53 regions contained rare deletions occurring in only one family member. Members of the same family shared a rare deletion in only eight regions. We also devised a scalable global test for enrichment of shared rare deletions.

CoGAPS matrix factorization algorithm identifies transcriptional changes in AP-2alpha target genes in feedback from therapeutic inhibition of the EGFR network.

Patients with oncogene driven tumors are treated with targeted therapeutics including EGFR inhibitors. Genomic data from The Cancer Genome Atlas (TCGA) demonstrates molecular alterations to EGFR, MAPK, and PI3K pathways in previously untreated tumors. Therefore, this study uses bioinformatics algorithms to delineate interactions resulting from EGFR inhibitor use in cancer cells with these genetic alterations. We modify the HaCaT keratinocyte cell line model to simulate cancer cells with constitutive activation of EGFR, HRAS, and PI3K in a controlled genetic background. We then measure gene expression after treating modified HaCaT cells with gefitinib, afatinib, and cetuximab. The CoGAPS algorithm distinguishes a gene expression signature associated with the anticipated silencing of the EGFR network. It also infers a feedback signature with EGFR gene expression itself increasing in cells that are responsive to EGFR inhibitors. This feedback signature has increased expression of several growth factor receptors regulated by the AP-2 family of transcription factors. The gene expression signatures for AP-2alpha are further correlated with sensitivity to cetuximab treatment in HNSCC cell lines and changes in EGFR expression in HNSCC tumors with low CDKN2A gene expression. In addition, the AP-2alpha gene expression signatures are also associated with inhibition of MEK, PI3K, and mTOR pathways in the Library of Integrated Network-Based Cellular Signatures (LINCS) data. These results suggest that AP-2 transcription factors are activated as feedback from EGFR network inhibition and may mediate EGFR inhibitor resistance.