Ultrasensitive detection of circulating tumour DNA via deep methylation sequencing aided by machine learning

Evaluation and integration of cell-free DNA signatures for detection of lung cancer

Cell-free DNA (cfDNA) analysis has shown potential in detecting early-stage lung cancer based on non-genetic features. To distinguish patients with lung cancer from healthy individuals, peripheral blood were collected from 926 lung cancer patients and 611 healthy individuals followed by cfDNA extraction. Low-pass whole genome sequencing and targeted methylation sequencing were conducted and various features of cfDNA were evaluated. With our customized algorithm using the most optimal features, the ensemble stacked model was constructed, called ESim-seq (Early Screening tech with Integrated Model). In the independent validation cohort, the ESim-seq model achieved an area under the curve (AUC) of 0.948 (95 % CI: 0.915-0.981), with a sensitivity of 79.3 % (95 % CI: 71.5-87.0 %) across all stages at a specificity of 96.0 % (95 % CI: 90.6-100.0 %). Specifically, the sensitivity of the ESim-seq model was 76.5 % (95 % CI: 67.3-85.8 %) in stage I patients, 100 % (95 % CI: 100.0-100.0 %) in stage II patients, 100 % (95 % CI: 100.0-100.0 %) in stage III patients and 87.5 % (95 % CI: 64.6%-100.0 %) in stage IV patients in the independent validation cohort. Besides, we constructed LCSC model (Lung Cancer Subtype multiple Classification), which was able to accurately distinguish patients with small cell lung cancer from those with non-small cell lung cancer, achieving an AUC of 0.961 (95 % CI: 0.949-0.957). The present study has established a framework for assessing cfDNA features and demonstrated the benefits of integrating multiple features for early detection of lung cancer.

Development and validation of an integrated system for lung cancer screening and post-screening pulmonary nodules management: a proof-of-concept study (ASCEND-LUNG)

Background

In order to address the low compliance and dissatisfied specificity of low-dose computed tomography (LDCT), efficient and non-invasive approaches are needed to complement its limitations for lung cancer screening and management. The ASCEND-LUNG study is a prospective two-stage case-control study designed to evaluate the performance of a liquid biopsy-based comprehensive lung cancer screening and post-screening pulmonary nodules management system.

Methods

We aimed to develop a comprehensive lung cancer system called Peking University Lung Cancer Screening and Management System (PKU-LCSMS) which comprises a lung cancer screening model to identify specific populations requiring LDCT and an artificial intelligence-aided (AI-aided) pulmonary nodules diagnostic model to classify pulmonary nodules following LDCT. A dataset of 465 participants (216 cancer, 47 benign, 202 non-cancer control) were used for the two models' development phase. For the lung cancer screening model development, cancer participants were randomly split at a ratio of 1:1 into the train and validation cohorts, and then non-cancer controls were age-matched to the cancer cases in a 1:1 ratio. Similarly, for the AI-aided pulmonary nodules model, cancer and benign participants were also randomly divided at a ratio of 2:1 into the train and validation cohorts. Subsequently, during the model validation phase, sensitivity and specificity were validated using an independent validation cohort consisting of 291 participants (140 cancer, 25 benign, 126 non-cancer control). Prospectively collected blood samples were analyzed for multi-omics including cell-free DNA (cfDNA) methylation, mutation, and serum protein. Computerized tomography (CT) images data was also obtained. Paired tissue samples were additionally analyzed for DNA methylation, DNA mutation, and messenger RNA (mRNA) expression to further explore the potential biological mechanisms. This study is registered with ClinicalTrials.gov, NCT04817046.

Findings

Baseline blood samples were evaluated for the whole screening and diagnostic process. The cfDNA methylation-based lung cancer screening model exhibited the highest area under the curve (AUC) of 0.910 (95% CI, 0.869-0.950), followed by the protein model (0.891 [95% CI, 0.845-0.938]) and lastly the mutation model (0.577 [95% CI, 0.482-0.672]). Further, the final screening model, which incorporated cfDNA methylation and protein features, achieved an AUC of 0.963 (95% CI, 0.942-0.984). In the independent validation cohort, the multi-omics screening model showed a sensitivity of 99.2% (95% CI, 0.957-1.000) at a specificity of 56.3% (95% CI, 0.472-0.652). For the AI-aided pulmonary nodules diagnostic model, which incorporated cfDNA methylation and CT images features, it yielded a sensitivity of 81.1% (95% CI, 0.732-0.875), a specificity of 76.0% (95% CI, 0.549-0.906) in the independent validation cohort. Furthermore, four differentially methylated regions (DMRs) were shared in the lung cancer screening model and the AI-aided pulmonary nodules diagnostic model.

Interpretation

We developed and validated a liquid biopsy-based comprehensive lung cancer screening and management system called PKU-LCSMS which combined a blood multi-omics based lung cancer screening model incorporating cfDNA methylation and protein features and an AI-aided pulmonary nodules diagnostic model integrating CT images and cfDNA methylation features in sequence to streamline the entire process of lung cancer screening and post-screening pulmonary nodules management. It might provide a promising applicable solution for lung cancer screening and management.

Funding

This work was supported by Science, Science, Technology & Innovation Project of Xiongan New Area, Beijing Natural Science Foundation, CAMS Innovation Fund for Medical Sciences (CIFMS), Clinical Medicine Plus X-Young Scholars Project of Peking University, the Fundamental Research Funds for the Central Universities, Research Unit of Intelligence Diagnosis and Treatment in Early Non-small Cell Lung Cancer, Chinese Academy of Medical Sciences, National Natural Science Foundation of China, Peking University People's Hospital Research and Development Funds, National Key Research and Development Program of China, and the fundamental research funds for the central universities.

Non-invasive diagnosis of pulmonary nodules by circulating tumor DNA methylation: A prospective multicenter study

BACKGROUND

With the popularization of computed tomography, more and more pulmonary nodules (PNs) are being detected. Risk stratification of PNs is essential for detecting early-stage lung cancer while minimizing the overdiagnosis of benign nodules. This study aimed to develop a circulating tumor DNA (ctDNA) methylation-based, non-invasive model for the risk stratification of PNs.

METHODS

A blood-based assay ("LUNG-TRAC") was designed to include novel lung cancer ctDNA methylation markers identified from in-house reduced representative bisulfite sequencing data and known markers from the literature. A stratification model was trained based on 183 ctDNA samples derived from patients with benign or malignant PNs and validated in 62 patients. LUNG-TRAC was further single-blindly tested in a single- and multi-center cohort.

RESULTS

The LUNG-TRAC model achieved an area under the curve (AUC) of 0.810 (sensitivity = 74.4 % and specificity = 73.7 %) in the validation set. Two test sets were used to evaluate the performance of LUNG-TRAC, with an AUC of 0.815 in the single-center test (N = 61; sensitivity = 67.5 % and specificity = 76.2 %) and 0.761 in the multi-center test (N = 95; sensitivity = 50.7 % and specificity = 80.8 %). The clinical utility of LUNG-TRAC was further assessed by comparing it to two established risk stratification models: the Mayo Clinic and Veteran Administration models. It outperformed both in the validation and the single-center test sets.

CONCLUSION

The LUNG-TRAC model demonstrated accuracy and consistency in stratifying PNs for the risk of malignancy, suggesting its utility as a non-invasive diagnostic aid for early-stage peripheral lung cancer.

CLINICAL TRIAL REGISTRATION

www.

CLINICALTRIALS

gov (NCT03989219).

Plasma ofCS-modified CD44 predicts the survival of patients with lung cancer

Plasma levels of oncofetal chondroitin sulfate (ofCS)-modified CD44 have emerged as a promising biomarker for multi-cancer detection. Here, we explored its potential to predict the survival of patients with lung cancer. A prospective observational cohort was conducted involving 274 newly diagnosed patients with lung cancer at the Sun Yat-sen University Cancer Center from 2013 to 2015. The plasma levels of ofCS-modified CD44 were measured, and Cox regression analysis was performed to assess the association between plasma-modified CD44 levels and overall survival (OS) as well as other prognostic outcomes. Prognostic nomograms were constructed based on plasma ofCS-modified CD44 levels to predict survival outcomes for patients with lung cancer. Patients with high expression ofCS-modified CD44 exhibited significantly worse outcomes in terms of OS (HR = 1.61, 95%CI = 1.13-2.29, p = 0.009) and progression-free survival (PFS). These findings were consistent across various analyses. The concordance index of the prognostic nomogram for predicting OS in both the training set and validation set were 0.723 and 0.737, respectively. Additionally, time-dependent receiver operating characteristic (ROC) curves showed that the nomogram could serve as a useful tool for predicting OS in patients with lung cancer. Plasma ofCS-modified CD44 may serve as an independent prognosis marker for patients with lung cancer. Further validation of its predictive value could enhance prognostic assessment and guide personalized treatment strategies for patients with lung cancer.

Neural Network Enables High Accuracy for Hepatitis B Surface Antigen Detection with a Plasmonic Platform

The detection of hepatitis B surface antigen (HBsAg) is critical in diagnosing hepatitis B virus (HBV) infection. However, existing clinical detection technologies inevitably cause certain inaccuracies, leading to delayed or unwarranted treatment. Here, we introduce a label-free plasmonic biosensing method based on the thickness-sensitive plasmonic coupling, combined with supervised deep learning (DL) using neural networks. The strategy of utilizing neural networks to process output data can reduce the limit of detection (LOD) of the sensor and significantly improve the accuracy (from 93.1%-97.4% to 99%-99.6%). Compared with widely used emerging clinical technologies, our platform achieves accurate decisions with higher sensitivity in a short assay time (∼30 min). The integration of DL models considerably simplifies the readout procedure, resulting in a substantial decrease in processing time. Our findings offer a promising avenue for developing high-precision molecular detection tools for point-of-care (POC) applications.

Bioreaction-Compatible Bivariate Lanthanide MOF Sensor Enables Stimulus-Multiresponsive Platform for ctDNA On-Site Detection

Detection of circulating tumor DNA (ctDNA) in liquid biopsy is of great importance for tumor diagnosis but difficult due to its low amount in bodily fluids. Herein, a novel ctDNA detection platform is established by quantifying DNA amplification by-product pyrophosphate (PPi) using a newly designed bivariable lanthanide metal-organic framework (Ln-MOF), namely, Ce/Eu-DPA MOF (CE-24, DPA = pyridine-2,6-dicarboxylic acid). CE-24 MOF exhibits ultrafast dual-response (fluorescence enhancement and enzyme-activity inhibition) to PPi stimuli by virtue of host-guest interaction. The platform is applied to detecting colon carcinoma-related ctDNA (KARS G12D mutation) combined with the isothermal nucleic acid exponential amplification reaction (EXPAR). ctDNA triggers the generation of a large amount of PPi, and the ctDNA quantification is achieved through the ratio fluorescence/colorimetric dual-mode assay of PPi. The combination of the EXPAR and the dual-mode PPi sensing allows the ctDNA assay method to be low-cost, convenient, bioreaction-compatible (freedom from the interference of bioreaction systems), sensitive (limit of detection down to 101 fM), and suitable for on-site detection. To the best of our knowledge, this work is the first application of Ln-MOF for ctDNA detection, and it provides a novel universal strategy for the rapid detection of nucleic acid biomarkers in point-of-care scenarios.

Translation of Epigenetics in Cell-Free DNA Liquid Biopsy Technology and Precision Oncology

Technological advancements in cell-free DNA (cfDNA) liquid biopsy have triggered exponential growth in numerous clinical applications. While cfDNA-based liquid biopsy has made significant strides in personalizing cancer treatment, the exploration and translation of epigenetics in liquid biopsy to clinical practice is still nascent. This comprehensive review seeks to provide a broad yet in-depth narrative of the present status of epigenetics in cfDNA liquid biopsy and its associated challenges. It highlights the potential of epigenetics in cfDNA liquid biopsy technologies with the hopes of enhancing its clinical translation. The momentum of cfDNA liquid biopsy technologies in recent years has propelled epigenetics to the forefront of molecular biology. We have only begun to reveal the true potential of epigenetics in both our understanding of disease and leveraging epigenetics in the diagnostic and therapeutic domains. Recent clinical applications of epigenetics-based cfDNA liquid biopsy revolve around DNA methylation in screening and early cancer detection, leading to the development of multi-cancer early detection tests and the capability to pinpoint tissues of origin. The clinical application of epigenetics in cfDNA liquid biopsy in minimal residual disease, monitoring, and surveillance are at their initial stages. A notable advancement in fragmentation patterns analysis has created a new avenue for epigenetic biomarkers. However, the widespread application of cfDNA liquid biopsy has many challenges, including biomarker sensitivity, specificity, logistics including infrastructure and personnel, data processing, handling, results interpretation, accessibility, and cost effectiveness. Exploring and translating epigenetics in cfDNA liquid biopsy technology can transform our understanding and perception of cancer prevention and management. cfDNA liquid biopsy has great potential in precision oncology to revolutionize conventional ways of early cancer detection, monitoring residual disease, treatment response, surveillance, and drug development. Adapting the implementation of liquid biopsy workflow to the local policy worldwide and developing point-of-care testing holds great potential to overcome global cancer disparity and improve cancer outcomes.

Lung tumor discrimination by deep neural network model CanDo via DNA methylation in bronchial lavage

Bronchoscopic-assisted discrimination of lung tumors presents challenges, especially in cases with contraindications or inaccessible lesions. Through meta-analysis and validation using the HumanMethylation450 database, this study identified methylation markers for molecular discrimination in lung tumors and designed a sequencing panel. DNA samples from 118 bronchial washing fluid (BWF) specimens underwent enrichment via multiplex PCR before targeted methylation sequencing. The Recursive Feature Elimination Cross-Validation and deep neural network algorithm established the CanDo classification model, which incorporated 11 methylation features (including 8 specific to the TBR1 gene), demonstrating a sensitivity of 98.6% and specificity of 97.8%. In contrast, bronchoscopic rapid on-site evaluation (bronchoscopic-ROSE) had lower sensitivity (87.7%) and specificity (80%). Further validation in 33 individuals confirmed CanDo's discriminatory potential, particularly in challenging cases for bronchoscopic-ROSE due to pathological complexity. CanDo serves as a valuable complement to bronchoscopy for the discriminatory diagnosis and stratified management of lung tumors utilizing BWF specimens.

Accurate Early Detection and EGFR Mutation Status Prediction of Lung Cancer Using Plasma cfDNA Coverage Patterns: A Proof-of-Concept Study

Lung cancer is a major global health concern with a low survival rate, often due to late-stage diagnosis. Liquid biopsy offers a non-invasive approach to cancer detection and monitoring, utilizing various features of circulating cell-free DNA (cfDNA). In this study, we established two models based on cfDNA coverage patterns at the transcription start sites (TSSs) from 6X whole-genome sequencing: an Early Cancer Screening Model and an EGFR mutation status prediction model. The Early Cancer Screening Model showed encouraging prediction ability, especially for early-stage lung cancer. The EGFR mutation status prediction model exhibited high accuracy in distinguishing between EGFR-positive and wild-type cases. Additionally, cfDNA coverage patterns at TSSs also reflect gene expression patterns at the pathway level in lung cancer patients. These findings demonstrate the potential applications of cfDNA coverage patterns at TSSs in early cancer screening and in cancer subtyping.

Circulating tumor DNA (ctDNA): can it be used as a pan-cancer early detection test?

Circulating tumor DNA (ctDNA, DNA shed by cancer cells) is emerging as one of the most transformative cancer biomarkers discovered to-date. Although potentially useful at all the phases of cancer detection and patient management, one of its most exciting possibilities is as a relatively noninvasive pan-cancer screening test. Preliminary findings with ctDNA tests such as Galleri or CancerSEEK suggest that they have high specificity (> 99.0%) for malignancy. Their sensitivity varies depending on the type of cancer and stage of disease but it is generally low in patients with stage I disease. A major advantage of ctDNA over existing screening strategies is the potential ability to detect multiple cancer types in a single test. A limitation of most studies published to-date is that they are predominantly case-control investigations that were carried out in patients with a previous diagnosis of malignancy and that used apparently healthy subjects as controls. Consequently, the reported sensitivities, specificities and positive predictive values might be lower if the tests are used for screening in asymptomatic populations, that is, in the population where these tests are likely be employed. To demonstrate clinical utility in an asymptomatic population, these tests must be shown to reduce cancer mortality without causing excessive overdiagnosis in a large randomized prospective randomized trial. Such trials are currently ongoing for Galleri and CancerSEEK.

Implementable assay for monitoring minimum residual disease after radical treatment for colorectal cancer

Considering the cost and invasiveness of monitoring postoperative minimal residual disease (MRD) of colorectal cancer (CRC) after adjuvant chemoradiotherapy (ACT), we developed a favorable approach based on methylated circulating tumor DNA to detect MRD after radical resection. Analyzing the public database, we identified the methylated promoter regions of the genes FGD5, GPC6, and MSC. Using digital polymerase chain reaction (dPCR), we termed the "amplicon of methylated sites using a specific enzyme" assay as "AMUSE." We examined 180 and 114 pre- and postoperative serial plasma samples from 28 recurrent and 19 recurrence-free pathological stage III CRC patients, respectively. The results showed 22 AMUSE-positive of 28 recurrent patients (sensitivity, 78.6%) and 17 AMUSE-negative of 19 recurrence-free patients (specificity, 89.5%). AMUSE predicted recurrence 208 days before conventional diagnosis using radiological imaging. Regarding ACT evaluation by the reactive response, 19 AMUSE-positive patients during their second or third blood samples showed a significantly poorer prognosis than the other patients (p = 9E-04). The AMUSE assay stratified four groups by the altered patterns of tumor burden postoperatively. Interestingly, only 34.8% of cases tested AMUSE-negative during ACT treatment, indicating eligibility for ACT. The AMUSE assay addresses the clinical need for accurate MRD monitoring with universal applicability, minimal invasiveness, and cost-effectiveness, thereby enabling the timely detection of recurrences. This assay can effectively evaluate the efficacy of ACT in patients with stage III CRC following curative resection. Our study strongly recommends reevaluating the clinical application of ACT using the AMUSE assay.

Development of new techniques and clinical applications of liquid biopsy in lung cancer management

Lung cancer is an exceedingly malignant tumor reported as having the highest morbidity and mortality of any cancer worldwide, thus posing a great threat to global health. Despite the growing demand for precision medicine, current methods for early clinical detection, treatment and prognosis monitoring in lung cancer are hampered by certain bottlenecks. Studies have found that during the formation and development of a tumor, molecular substances carrying tumor-related genetic information can be released into body fluids. Liquid biopsy (LB), a method for detecting these tumor-related markers in body fluids, maybe a way to make progress in these bottlenecks. In recent years, LB technology has undergone rapid advancements. Therefore, this review will provide information on technical updates to LB and its potential clinical applications, evaluate its effectiveness for specific applications, discuss the existing limitations of LB, and present a look forward to possible future clinical applications. Specifically, this paper will introduce technical updates from the prospectives of engineering breakthroughs in the detection of membrane-based LB biomarkers and other improvements in sequencing technology. Additionally, it will summarize the latest applications of liquid biopsy for the early detection, diagnosis, treatment, and prognosis of lung cancer. We will present the interconnectedness of clinical and laboratory issues and the interplay of technology and application in LB today.

Association between high or low-quality carbohydrate with depressive symptoms and socioeconomic-dietary factors model based on XGboost algorithm: From NHANES 2007-2018

BACKGROUND

Depressive symptoms are a serious public mental health problem, and dietary intake is often considered to be associated with depressive symptoms. However, the relationship between the quality of dietary carbohydrates and depressive symptoms remains unclear. Therefore, this study aimed to investigate the relationship between high and low-quality carbohydrates and depressive symptoms and to attempt to construct an integrated model using machine learning to predict depressive symptoms.

METHODS

A total of 4982 samples from the National Health and Nutrition Examination Survey (NHANES) were included in this study. Carbohydrate intake was assessed by a 24-h dietary review, and depressive symptoms were assessed using the Patient Health Questionnaire-9 (PHQ9). Variance inflation factor (VIF) and Relief-F algorithms were used for variable feature selection.

RESULTS

The results of multivariate linear regression showed a negative association between high-quality carbohydrates and depressive symptoms (β: -0.147, 95 % CI: -0.239, -0.056, p = 0.002) and a positive association between low-quality carbohydrates and depressive symptoms (β: 0.018, 95 % CI: 0.007, 0.280, p = 0.001). Subsequently, we used the XGboost model to produce a comprehensive depressive symptom evaluation model and developed a corresponding online tool (http://8.130.128.194:5000/) to evaluate depressive symptoms clinically.

LIMITATIONS

The cross-sectional study could not yield any conclusions regarding causality, and the model has not been validated with external data.

CONCLUSIONS

Carbohydrate quality is associated with depressive symptoms, and machine learning models that combine diet with socioeconomic factors can be a tool for predicting depression severity.

Advances in Machine Learning Processing of Big Data from Disease Diagnosis Sensors

Exploring accurate, noninvasive, and inexpensive disease diagnostic sensors is a critical task in the fields of chemistry, biology, and medicine. The complexity of biological systems and the explosive growth of biomarker data have driven machine learning to become a powerful tool for mining and processing big data from disease diagnosis sensors. With the development of bioinformatics and artificial intelligence (AI), machine learning models formed by data mining have been able to guide more sensitive and accurate molecular computing. This review presents an overview of big data collection approaches and fundamental machine learning algorithms and discusses recent advances in machine learning and molecular computational disease diagnostic sensors. More specifically, we highlight existing modular workflows and key opportunities and challenges for machine learning to achieve disease diagnosis through big data mining.

DeepNphos: A deep-learning architecture for prediction of N-phosphorylation sites

MOTIVATION

Phosphorylation, a prevalent post-translational modification, plays a crucial role in regulating cellular activities. This process encompasses O-phosphorylation (e.g., phosphoserine) and N-phosphorylation (e.g., phospho-lysine (pK), phospho-arginine (pR), and phospho-histidine (pH)). While significant research has focused on O-phosphorylation, resulting in the development of various algorithms for predicting O-phosphorylation sites with commendable performance, there has been a notable absence of models designed to predict N-phosphorylation sites. This study introduces an integrated model named DeepNphos, designed to predict N-phosphorylation sites. This model is developed based on the analysis of thousands of experimentally identified pK, pR and pH sites.

RESULTS

Observing that the Convolutional Neural Network (CNN) model, incorporating the One-Hot encoding feature, demonstrates favorable performance in comparison to other models when predicting pK, pR, and pH sites. Additionally, pK exhibits similarities to other lysine modification types, and integrating the CNN model with a deep-transfer learning (DTL) strategy based on tens of thousands of known lysine modification sites could enhance pK prediction performance. In contrast, pR exhibits little similarity to other arginine modification types, and the integration of DTL has minimal impact on pR prediction performance. Furthermore, the decision was made to refrain from incorporating the DTL strategy in predicting pH sites, given the scarcity of histidine modification sites beyond those associated with pH. The final classifiers for predicting pK, pR, and pH sites achieve AUC values of 0.856, 0.805 and 0.802 for ten-fold cross-validation, respectively. Overall, DeepNphos is the first classifier for predicting N-phosphorylation sites, accessible at https://github.com/ChangXulinmessi/DeepNPhos.

Intelligent quantitative recognition of sulfide using machine learning-based ratiometric fluorescence probe of metal-organic framework UiO-66-NH2/Ppix

Sulfides possess either high toxicity or play crucial physiological role such as gas transmitter dependent upon dosage, hence the significant for their rapid sensitive and selective concentration determination. Herein, a machine learning enhanced ratiometric fluorescence sensor was engineered for sulfide determination by incorporating the nanometal-organic framework (UiO-66-NH2) along with protoporphyrin IX (Ppix). The blue fluorescence at 431 nm originated from the moiety of UiO-66-NH2 by 365 nm excitation serves as an internal calibration reference signal, while the red fluorescence at 629 nm from the moiety of Ppix serves as the analytical signal, and the intensity is correlated to the amount of sulfides. The fluorescence color of the sensor gradually varies from blue to red upon sequential addition of copper and sulfide ions, resulting in RGB (Red, Green, Blue) feature values for corresponding sulfide concentrations, which facilities the advanced data processing techniques using machine learning algorithms. On the basis of fluorescence image fingerprint extraction and machine learning algorithms, an online data analysis model was developed to improve the precision and accuracy of sulfide determination. The established model employed Linear Discriminant Analysis (LDA) and was subjected to rigorous cross-validation to ensure its robustness. By analyzing the correlation between RGB feature values and sulfide concentrations, the study highlighted a significant positive relationship between the red feature values and sulfide concentrations. The application of machine learning techniques on the ratiometric fluorescence signal of the UiO-66-NH2/Ppix probe demonstrated its potential for intelligent quantitative determination of sulfides, offering a valuable and efficient tool for pollution detection and real-time rapid environmental monitoring.

Circulating cell-free DNA-based multi-cancer early detection

Patients benefit considerably from early detection of cancer. Existing single-cancer tests have various limitations, which could be effectively addressed by circulating cell-free DNA (cfDNA)-based multi-cancer early detection (MCED). With sensitive detection and accurate localization of multiple cancer types at a very low and fixed false-positive rate (FPR), MCED has great potential to revolutionize early cancer detection. Herein, we review state-of-the-art approaches for cfDNA-based MCED and their limitations and discuss both technical and clinical challenges in the development and application of MCED tests. Given the constant improvements in technology and understanding of cancer biology, we propose that a cfDNA-based targeted sequencing assay that integrates multimodal features should be optimized for MCED.

Liquid biopsy in lung cancer

Lung cancer is a highly prevalent malignancy worldwide and the primary cause of mortality. The absence of systematic and standardized diagnostic approaches for identifying potential pulmonary nodules, early-stage cancers, and indeterminate tumors has led clinicians to consider tissue biopsy and pathological sections as the preferred method for clinical diagnosis, often regarded as the gold standard. The conventional tissue biopsy is an invasive procedure that does not adequately capture the diverse characteristics and evolving nature of tumors. Recently, the concept of 'liquid biopsy' has gained considerable attention as a promising solution. Liquid biopsy is a non-invasive approach that facilitates repeated analysis, enabling real-time monitoring of tumor recurrence, metastasis, and response to treatment. Currently, liquid biopsy includes circulating tumor cells, circulating cell-free DNA, circulating tumor DNA, circulating cell-free RNA, extracellular vesicles, and other proteins and metabolites. With rapid progress in molecular technology, liquid biopsy has emerged as a highly promising and intriguing approach, yielding compelling results. This article critically examines the significant role and potential clinical implications of liquid biopsy in the diagnosis, treatment, and prognosis of lung cancer.

New regulatory thinking is needed for AI-based personalised drug and cell therapies in precision oncology

Until recently the application of artificial intelligence (AI) in precision oncology was confined to activities in drug development and had limited impact on the personalisation of therapy. Now, a number of approaches have been proposed for the personalisation of drug and cell therapies with AI applied to therapy design, planning and delivery at the patient's bedside. Some drug and cell-based therapies are already tuneable to the individual to optimise efficacy, to reduce toxicity, to adapt the dosing regime, to design combination therapy approaches and, preclinically, even to personalise the receptor design of cell therapies. Developments in AI-based healthcare are accelerating through the adoption of foundation models, and generalist medical AI models have been proposed. The application of these approaches in therapy design is already being explored and realistic short-term advances include the application to the personalised design and delivery of drugs and cell therapies. With this pace of development, the limiting step to adoption will likely be the capacity and appropriateness of regulatory frameworks. This article explores emerging concepts and new ideas for the regulation of AI-enabled personalised cancer therapies in the context of existing and in development governance frameworks.

mHapBrowser: a comprehensive database for visualization and analysis of DNA methylation haplotypes

DNA methylation acts as a vital epigenetic regulatory mechanism involved in controlling gene expression. Advances in sequencing technologies have enabled characterization of methylation patterns at single-base resolution using bisulfite sequencing approaches. However, existing methylation databases have primarily focused on mean methylation levels, overlooking phased methylation patterns. The methylation status of CpGs on individual sequencing reads represents discrete DNA methylation haplotypes (mHaps). Here, we present mHapBrowser, a comprehensive database for visualizing and analyzing mHaps. We systematically processed data of diverse tissues in human, mouse and rat from public repositories, generating mHap format files for 6366 samples. mHapBrowser enables users to visualize eight mHap metrics across the genome through an integrated WashU Epigenome Browser. It also provides an online server for comparing mHap patterns across samples. Additionally, mHap files for all samples can be downloaded to facilitate local processing using downstream analysis toolkits. The utilities of mHapBrowser were demonstrated through three case studies: (i) mHap patterns are associated with gene expression; (ii) changes in mHap patterns independent of mean methylation correlate with differential expression between lung cancer subtypes; and (iii) the mHap metric MHL outperforms mean methylation for classifying tumor and normal samples from cell-free DNA. The database is freely accessible at http://mhap.sibcb.ac.cn/.

An artificial intelligence model for embryo selection in preimplantation DNA methylation screening in assisted reproductive technology

Embryo quality is a critical determinant of clinical outcomes in assisted reproductive technology (ART). A recent clinical trial investigating preimplantation DNA methylation screening (PIMS) revealed that whole genome DNA methylation level is a novel biomarker for assessing ART embryo quality. Here, we reinforced and estimated the clinical efficacy of PIMS. We introduce PIMS-AI, an innovative artificial intelligence (AI) based model, to predict the probability of an embryo producing live birth and subsequently assist ART embryo selection. Our model demonstrated robust performance, achieving an area under the curve (AUC) of 0.90 in cross-validation and 0.80 in independent testing. In simulated embryo selection, PIMS-AI attained an accuracy of 81% in identifying viable embryos for patients. Notably, PIMS-AI offers significant advantages over conventional preimplantation genetic testing for aneuploidy (PGT-A), including enhanced embryo discriminability and the potential to benefit a broader patient population. In conclusion, our approach holds substantial promise for clinical application and has the potential to significantly improve the ART success rate.

A Novel Tissue-Free Method to Estimate Tumor-Derived Cell-Free DNA Quantity Using Tumor Methylation Patterns

Estimating the abundance of cell-free DNA (cfDNA) fragments shed from a tumor (i.e., circulating tumor DNA (ctDNA)) can approximate tumor burden, which has numerous clinical applications. We derived a novel, broadly applicable statistical method to quantify cancer-indicative methylation patterns within cfDNA to estimate ctDNA abundance, even at low levels. Our algorithm identified differentially methylated regions (DMRs) between a reference database of cancer tissue biopsy samples and cfDNA from individuals without cancer. Then, without utilizing matched tissue biopsy, counts of fragments matching the cancer-indicative hyper/hypo-methylated patterns within DMRs were used to determine a tumor methylated fraction (TMeF; a methylation-based quantification of the circulating tumor allele fraction and estimate of ctDNA abundance) for plasma samples. TMeF and small variant allele fraction (SVAF) estimates of the same cancer plasma samples were correlated (Spearman's correlation coefficient: 0.73), and synthetic dilutions to expected TMeF of 10-3 and 10-4 had estimated TMeF within two-fold for 95% and 77% of samples, respectively. TMeF increased with cancer stage and tumor size and inversely correlated with survival probability. Therefore, tumor-derived fragments in the cfDNA of patients with cancer can be leveraged to estimate ctDNA abundance without the need for a tumor biopsy, which may provide non-invasive clinical approximations of tumor burden.

Circulating tumor DNA: current implementation issues and future challenges for clinical utility

Over the past decades, liquid biopsy, especially circulating tumor DNA (ctDNA), has received tremendous attention as a noninvasive detection approach for clinical applications, including early diagnosis of cancer and relapse, real-time therapeutic efficacy monitoring, potential target selection and investigation of drug resistance mechanisms. In recent years, the application of next-generation sequencing technology combined with AI technology has significantly improved the accuracy and sensitivity of liquid biopsy, enhancing its potential in solid tumors. However, the increasing integration of such promising tests to improve therapy decision making by oncologists still has complexities and challenges. Here, we propose a conceptual framework of ctDNA technologies and clinical utilities based on bibliometrics and highlight current challenges and future directions, especially in clinical applications such as early detection, minimal residual disease detection, targeted therapy, and immunotherapy. We also discuss the necessities of developing a dynamic field of translational cancer research and rigorous clinical studies that may support therapeutic strategy decision making in the near future.

Circulating tumor DNA as liquid biopsy in lung cancer: Biological characteristics and clinical integration

Lung cancer maintains high morbidity and mortality rate globally despite significant advancements in diagnosis and treatment in the era of precision medicine. Pathological analysis of tumor tissue, the current gold standard for lung cancer diagnosis, is intrusive and intrinsically confined to evaluating the limited amount of tissues that could be physically extracted. However, tissue biopsy has several limitations, including the invasiveness of the procedure and difficulty in obtaining samples for patients at advanced stages., there Additionally,has been no major breakthrough in tumor biomarkers with high specificity and sensitivity, particularly for early-stage lung cancer. Liquid biopsy has been considered a feasible auxiliary tool for tearly dianosis, evaluating treatment responses and monitoring prognosis of lung cancer. Circulating tumor DNA (ctDNA), an ideal biomarker of liquid biopsy, has emerged as one of the most reliable tools for monitoring tumor processes at molecular levels. Herein, this review focuses on tumor heterogeneity to elucidate the superiority of liquid biopsy and retrospectively discussdeciphersolution. We systematically elaborate ctDNA biological characteristics, introduce methods for ctDNA detection, and discuss the current role of plasma ctDNA in lung cancer management. Finally, we summarize the drawbacks of ctDNA analysis and highlight its potential clinical application in lung cancer.

Artificial intelligence-based risk stratification, accurate diagnosis and treatment prediction in gynecologic oncology

As data-driven science, artificial intelligence (AI) has paved a promising path toward an evolving health system teeming with thrilling opportunities for precision oncology. Notwithstanding the tremendous success of oncological AI in such fields as lung carcinoma, breast tumor and brain malignancy, less attention has been devoted to investigating the influence of AI on gynecologic oncology. Hereby, this review sheds light on the ever-increasing contribution of state-of-the-art AI techniques to the refined risk stratification and whole-course management of patients with gynecologic tumors, in particular, cervical, ovarian and endometrial cancer, centering on information and features extracted from clinical data (electronic health records), cancer imaging including radiological imaging, colposcopic images, cytological and histopathological digital images, and molecular profiling (genomics, transcriptomics, metabolomics and so forth). However, there are still noteworthy challenges beyond performance validation. Thus, this work further describes the limitations and challenges faced in the real-word implementation of AI models, as well as potential solutions to address these issues.

Cancer signature ensemble integrating cfDNA methylation, copy number, and fragmentation facilitates multi-cancer early detection

Cell-free DNA (cfDNA) sequencing has demonstrated great potential for early cancer detection. However, most large-scale studies have focused only on either targeted methylation sites or whole-genome sequencing, limiting comprehensive analysis that integrates both epigenetic and genetic signatures. In this study, we present a platform that enables simultaneous analysis of whole-genome methylation, copy number, and fragmentomic patterns of cfDNA in a single assay. Using a total of 950 plasma (361 healthy and 589 cancer) and 240 tissue samples, we demonstrate that a multifeature cancer signature ensemble (CSE) classifier integrating all features outperforms single-feature classifiers. At 95.2% specificity, the cancer detection sensitivity with methylation, copy number, and fragmentomic models was 77.2%, 61.4%, and 60.5%, respectively, but sensitivity was significantly increased to 88.9% with the CSE classifier (p value < 0.0001). For tissue of origin, the CSE classifier enhanced the accuracy beyond the methylation classifier, from 74.3% to 76.4%. Overall, this work proves the utility of a signature ensemble integrating epigenetic and genetic information for accurate cancer detection.

Individualized tumor-informed circulating tumor DNA analysis for postoperative monitoring of non-small cell lung cancer

We report a personalized tumor-informed technology, Patient-specific pROgnostic and Potential tHErapeutic marker Tracking (PROPHET) using deep sequencing of 50 patient-specific variants to detect molecular residual disease (MRD) with a limit of detection of 0.004%. PROPHET and state-of-the-art fixed-panel assays were applied to 760 plasma samples from 181 prospectively enrolled early stage non-small cell lung cancer patients. PROPHET shows higher sensitivity of 45% at baseline with circulating tumor DNA (ctDNA). It outperforms fixed-panel assays in prognostic analysis and demonstrates a median lead-time of 299 days to radiologically confirmed recurrence. Personalized non-canonical variants account for 98.2% with prognostic effects similar to canonical variants. The proposed tumor-node-metastasis-blood (TNMB) classification surpasses TNM staging for prognostic prediction at the decision point of adjuvant treatment. PROPHET shows potential to evaluate the effect of adjuvant therapy and serve as an arbiter of the equivocal radiological diagnosis. These findings highlight the potential advantages of personalized cancer techniques in MRD detection.

Advances in novel strategies for isolation, characterization, and analysis of CTCs and ctDNA

Over the past decade, the detection and analysis of liquid biopsy biomarkers such as circulating tumor cells (CTCs) and circulating tumor DNA (ctDNA) have advanced significantly. They have received recognition for their clinical usefulness in detecting cancer at an early stage, monitoring disease, and evaluating treatment response. The emergence of liquid biopsy has been a helpful development, as it offers a minimally invasive, rapid, real-time monitoring, and possible alternative to traditional tissue biopsies. In resource-limited settings, the ideal platform for liquid biopsy should not only extract more CTCs or ctDNA from a minimal sample volume but also accurately represent the molecular heterogeneity of the patient's disease. This review covers novel strategies and advancements in CTC and ctDNA-based liquid biopsy platforms, including microfluidic applications and comprehensive analysis of molecular complexity. We discuss these systems' operational principles and performance efficiencies, as well as future opportunities and challenges for their implementation in clinical settings. In addition, we emphasize the importance of integrated platforms that incorporate machine learning and artificial intelligence in accurate liquid biopsy detection systems, which can greatly improve cancer management and enable precision diagnostics.

Circulating biomarkers in perioperative management of cancer patients

Owing to the advances in surgical technology, most solid tumours can be controlled by surgical excision. The priority should be tumour control, while some routine perioperative management might influence cancer progression in an unnoticed way. Moreover, it is increasingly recognized that effective perioperative management should include techniques to improve postoperative outcomes. These influences are elucidated by the different functions of circulating biomarkers in cancer patients. Here, circulating biomarkers with two types of clinical functions were reviewed: (i) circulating biomarkers for cancer progression monitoring, for instance, those related to cancer cell malignancy, tumour microenvironment formation, and early metastasis, and (ii) circulating biomarkers with relevance to postoperative outcomes, including systemic inflammation, immunosuppression, cognitive dysfunction, and pain management. This review aimed to provide new perspectives for the perioperative management of patients with cancer and highlight the potential clinical translation value of circulating biomarkers in improving outcomes.

Using cfDNA and ctDNA as Oncologic Markers: A Path to Clinical Validation

The detection of circulating tumor DNA (ctDNA) in liquid biopsy samples as an oncological marker is being used in clinical trials at every step of clinical management. As ctDNA-based liquid biopsy kits are developed and used in clinics, companies work towards increased convenience, accuracy, and cost over solid biopsies and other oncological markers. The technology used to differentiate ctDNA and cell-free DNA (cfDNA) continues to improve with new tests and methodologies being able to detect down to mutant allele frequencies of 0.001% or 1/100,000 copies. Recognizing this development in technology, the FDA has recently given pre-market approval and breakthrough device designations to multiple companies. The purpose of this review is to look at the utility of measuring total cfDNA, techniques used to differentiate ctDNA from cfDNA, and the utility of different ctDNA-based liquid biopsy kits using relevant articles from PubMed, clinicaltrials.gov, FDA approvals, and company newsletters. Measuring total cfDNA could be a cost-effective, viable prognostic marker, but various factors do not favor it as a monitoring tool during chemotherapy. While there may be a place in the clinic for measuring total cfDNA in the future, the lack of standardization means that it is difficult to move forward with large-scale clinical validation studies currently. While the detection of ctDNA has promising standardized liquid biopsy kits from various companies with large clinical trials ongoing, their applications in screening and minimal residual disease can suffer from lower sensitivity. However, researchers are working towards solutions to these issues with innovations in technology, multi-omics, and sampling. With great promise, further research is needed before liquid biopsies can be recommended for everyday clinical management.

Cross-platform comparisons for targeted bisulfite sequencing of MGISEQ-2000 and NovaSeq6000

BACKGROUND

An accurate and reproducible next-generation sequencing platform is essential to identify malignancy-related abnormal DNA methylation changes and translate them into clinical applications including cancer detection, prognosis, and surveillance. However, high-quality DNA methylation sequencing has been challenging because poor sequence diversity of the bisulfite-converted libraries severely impairs sequencing quality and yield. In this study, we tested MGISEQ-2000 Sequencer's capability of DNA methylation sequencing with a published non-invasive pancreatic cancer detection assay, using NovaSeq6000 as the benchmark.

RESULTS

We sequenced a series of synthetic cell-free DNA (cfDNA) samples with different tumor fractions and found MGISEQ-2000 yielded data with similar quality as NovaSeq6000. The methylation levels measured by MGISEQ-2000 demonstrated high consistency with NovaSeq6000. Moreover, MGISEQ-2000 showed a comparable analytic sensitivity with NovaSeq6000, suggesting its potential for clinical detection. As to evaluate the clinical performance of MGISEQ-2000, we sequenced 24 clinical samples and predicted the pathology of the samples with a clinical diagnosis model, PDACatch classifier. The clinical model performance of MGISEQ-2000's data was highly consistent with that of NovaSeq6000's data, with the area under the curve of 1. We also tested the model's robustness with MGISEQ-2000's data when reducing the sequencing depth. The results showed that MGISEQ-2000's data showed matching robustness of the PDACatch classifier with NovaSeq6000's data.

CONCLUSIONS

Taken together, MGISEQ-2000 demonstrated similar data quality, consistency of the methylation levels, comparable analytic sensitivity, and matching clinical performance, supporting its application in future non-invasive early cancer detection investigations by detecting distinct methylation patterns of cfDNAs.

DNA Methylation in the Fields of Prenatal Diagnosis and Early Detection of Cancers

The central objective of the metamorphosis of discovery science into biomedical applications is to serve the purpose of patients and curtail the global disease burden. The journey from the discovery of DNA methylation (DNAm) as a biological process to its emergence as a diagnostic tool is one of the finest examples of such metamorphosis and has taken nearly a century. Particularly in the last decade, the application of DNA methylation studies in the clinic has been standardized more than ever before, with great potential to diagnose a multitude of diseases that are associated with a burgeoning number of genes with this epigenetic alteration. Fetal DNAm detection is becoming useful for noninvasive prenatal testing, whereas, in very preterm infants, DNAm is also shown to be a potential biological indicator of prenatal risk factors. In the context of cancer, liquid biopsy-based DNA-methylation profiling is offering valuable epigenetic biomarkers for noninvasive early-stage diagnosis. In this review, we focus on the applications of DNA methylation in prenatal diagnosis for delivering timely therapy before or after birth and in detecting early-stage cancers for better clinical outcomes. Furthermore, we also provide an up-to-date commercial landscape of DNAm biomarkers for cancer detection and screening of cancers of unknown origin.

Individualized dynamic methylation-based analysis of cell-free DNA in postoperative monitoring of lung cancer

BACKGROUND

The feasibility of DNA methylation-based assays in detecting minimal residual disease (MRD) and postoperative monitoring remains unestablished. We aim to investigate the dynamic characteristics of cancer-related methylation signals and the feasibility of methylation-based MRD detection in surgical lung cancer patients.

METHODS

Matched tumor, tumor-adjacent tissues, and longitudinal blood samples from a cohort (MEDAL) were analyzed by ultra-deep targeted sequencing and bisulfite sequencing. A tumor-informed methylation-based MRD (timMRD) was employed to evaluate the methylation status of each blood sample. Survival analysis was performed in the MEDAL cohort (n = 195) and validated in an independent cohort (DYNAMIC, n = 36).

RESULTS

Tumor-informed methylation status enabled an accurate recurrence risk assessment better than the tumor-naïve methylation approach. Baseline timMRD-scores were positively correlated with tumor burden, invasiveness, and the existence and abundance of somatic mutations. Patients with higher timMRD-scores at postoperative time-points demonstrated significantly shorter disease-free survival in the MEDAL cohort (HR: 3.08, 95% CI: 1.48-6.42; P = 0.002) and the independent DYNAMIC cohort (HR: 2.80, 95% CI: 0.96-8.20; P = 0.041). Multivariable regression analysis identified postoperative timMRD-score as an independent prognostic factor for lung cancer. Compared to tumor-informed somatic mutation status, timMRD-scores yielded better performance in identifying the relapsed patients during postoperative follow-up, including subgroups with lower tumor burden like stage I, and was more accurate among relapsed patients with baseline ctDNA-negative status. Comparing to the average lead time of ctDNA mutation, timMRD-score yielded a negative predictive value of 97.2% at 120 days prior to relapse.

CONCLUSIONS

The dynamic methylation-based analysis of peripheral blood provides a promising strategy for postoperative cancer surveillance.

TRIAL REGISTRATION

This study (MEDAL, MEthylation based Dynamic Analysis for Lung cancer) was registered on ClinicalTrials.gov on 08/05/2018 (NCT03634826). https://clinicaltrials.gov/ct2/show/NCT03634826 .

Circulating cell-free DNA sequencing for early detection of lung cancer

INTRODUCTION

Lung cancer is a leading cause of death in patients with cancer. Early diagnosis is crucial to improve the prognosis of patients with lung cancer. Plasma circulating cell-free DNA (cfDNA) contains comprehensive genetic and epigenetic information from tissues throughout the body, suggesting that early detection of lung cancer can be done non-invasively, conveniently, and cost-effectively using high-sensitivity techniques such as sequencing.

AREAS COVERED

In this review, we summarize the latest technological innovations, coupled with next-generation sequencing (NGS), regarding genomic alterations, methylation, and fragmentomic features of cfDNA for the early detection of lung cancer, as well as their clinical advances. Additionally, we discuss the suitability of study designs for diagnostic accuracy evaluation for different target populations and clinical questions.

EXPERT OPINION

Currently, cfDNA-based early screening and diagnosis of lung cancer faces many challenges, such as unsatisfactory performance, lack of quality control standards, and poor repeatability. However, the progress of several large prospective studies employing epigenetic features has shown promising predictive performance, which has inspired cfDNA sequencing for future clinical applications. Furthermore, the development of multi-omics markers for lung cancer, including genome-wide methylation and fragmentomics, is expected to play an increasingly important role in the future.

Current and emerging applications of liquid biopsy in pan-cancer

Cancer morbidity and mortality are growing rapidly worldwide and it is urgent to develop a convenient and effective method that can identify cancer patients at an early stage and predict treatment outcomes. As a minimally invasive and reproducible tool, liquid biopsy (LB) offers the opportunity to detect, analyze and monitor cancer in any body fluids including blood, complementing the limitations of tissue biopsy. In liquid biopsy, circulating tumor cells (CTCs) and circulating tumor DNA (ctDNA) are the two most common biomarkers, displaying great potential in the clinical application of pan-cancer. In this review, we expound the samples, targets, and newest techniques in liquid biopsy and summarize current clinical applications in several specific cancers. Besides, we put forward a bright prospect for further exploring the emerging application of liquid biopsy in the field of pan-cancer precision medicine.

DNA Methylation Analysis in Monozygotic Twins Discordant for ALS in Blood Cells

ALS is a fatal motor neuron disease that displays a broad variety of phenotypes ranging from early fatal courses to slowly progressing and rather benign courses. Such divergence can also be seen in genetic ALS cases with varying phenotypes bearing specific mutations, suggesting epigenetic mechanisms like DNA methylation act as disease modifiers. However, the epigenotype dictated by, in addition to other mechanisms, DNA methylation is also strongly influenced by the individual's genotype. Hence, we performed a DNA methylation study using EPIC arrays on 7 monozygotic (MZ) twin pairs discordant for ALS in whole blood, which serves as an ideal model for eliminating the effects of the genetic-epigenetic interplay to a large extent. We found one CpG site showing intra-pair hypermethylation in the affected co-twins, which maps to the Glutamate Ionotropic Receptor Kainate Type Subunit 1 gene (GRIK1). Additionally, we found 4 DMPs which were subsequently confirmed using 2 different statistical approaches. Differentially methylated regions or blocks could not be detected within the scope of this work. In conclusion, we revealed that despite a low sample size, monozygotic twin studies discordant for the disease can bring new insights into epigenetic processes in ALS, pointing to new target loci for further investigations.

A sensitive fluorescence assay based on aggregation-induced emission by copper-free click reaction for rapid ctDNA detection

A fluorescence assay was developed based on aggregation-induced emission for circulating tumor (ctDNA) rapid detection. AIE-based probe (TPE-DNA) was constructed via strain-promoted azide-alkyne cycloaddition reaction between the azide-functionalized TPE-N₃ and the dibenzocyclooctyne functionalized DNA. AIE-based assay displayed significant fluorescence signal enhancement for the ctDNA targets. The experimental results were verified and explained based on density functional theory simulations. As a proof-of-concept, the ctDNA as the target was detected to determine the feasibility of the assay. The detection could be done within 10 min with the limit of detection of 39 pM and good selectivity without the assistance of enzyme or nano-materials. Compared with traditional click reaction, the assay has advantages of high efficiency and good biocompatibility. It is promising in rapid diagnostic of cancer marker nucleic acid with further efforts to develop intelligent miniaturized equipment.

Bridging biological cfDNA features and machine learning approaches

Liquid biopsies (LBs), particularly using circulating tumor DNA (ctDNA), are expected to revolutionize precision oncology and blood-based cancer screening. Recent technological improvements, in combination with the ever-growing understanding of cell-free DNA (cfDNA) biology, are enabling the detection of tumor-specific changes with extremely high resolution and new analysis concepts beyond genetic alterations, including methylomics, fragmentomics, and nucleosomics. The interrogation of a large number of markers and the high complexity of data render traditional correlation methods insufficient. In this regard, machine learning (ML) algorithms are increasingly being used to decipher disease- and tissue-specific signals from cfDNA. Here, we review recent insights into biological ctDNA features and how these are incorporated into sophisticated ML applications.

A pyroptosis-related gene signature for prognosis prediction in hepatocellular carcinoma

Introduction

Hepatocellular carcinoma (HCC) is one of the most invasive cancers with a low 5-year survival rate. Pyroptosis, a specialized form of cell death, has shown its association with cancer progression. However, its role in the prognosis of HCC has not been fully understood.

Methods

In our study, clinical information and mRNA expression for 1076 patients with HCC were obtained from the five public cohorts. Pyroptotic clusters were generated by unsupervised clustering based on 40 pyroptosis-related genes (PRGs) in the TCGA and ICGC cohort. A pyroptosis-related signature was constructed using least absolute shrinkage and selection operator (LASSO) regression according to differentially expressed genes (DEGs) of pyroptotic clusters. The signature was then tested in the validation cohorts (GES10142 and GSE14520) and subsequently validated in the CPTAC cohort (n=159) at both mRNA and protein levels. Response to sorafenib was explored in GSE109211.

Results

Three clusters were identified based on the 40 PRGs in the TCGA cohort. A total of 24 genes were selected based on DEGs of the above three pyroptotic clusters to construct the pyroptotic risk score. Patients with the high-risk score showed shorter overall survival (OS) compared to those with the low-risk score in the training set (P&lt;0.001; HR, 3.06; 95% CI, 2.22-4.24) and the test set (P=0.008; HR, 1.61; 95% CI, 1.13-2.28). The predictive ability of the risk score was further confirmed in the CPTAC cohort at both mRNAs (P&lt;0.001; HR, 2.99; 95% CI, 1.67-5.36) and protein levels (P&lt;0.001; HR, 2.97; 95% CI 1.66-5.31). The expression of the model genes was correlated with immune cell infiltration, angiogenesis-related genes, and sensitivity to antiangiogenic therapy (P&lt;0.05).

Discussion

In conclusion, we established a prognostic signature of 24 genes based on pyroptosis clusters for HCC patients, providing insight into the risk stratification of HCC.

Cell-free DNA methylation biomarker for the diagnosis of papillary thyroid carcinoma

BACKGROUND

Cell-free DNA (cfDNA) is being explored as biomarker for non-invasive diagnosis of cancer. We aimed to establish a cfDNA-based DNA methylation marker panel to differentially diagnose papillary thyroid carcinoma (PTC) from benign thyroid nodule (BTN).

METHODS

220 PTC- and 188 BTN patients were enrolled. Methylation markers of PTC were identified from patients' tissue and plasma by reduced representation bisulfite sequencing and methylation haplotype analyses. They were combined with PTC markers from literatures and were tested on additional PTC and BTN samples to verify PTC-detecting ability using targeted methylation sequencing. Top markers were developed into ThyMet and were tested in 113 PTC and 88 BTN cases to train and validate a PTC-plasma classifier. Integration of ThyMet and thyroid ultrasonography was explored to improve accuracy.

FINDINGS

From 859 potential PTC plasma-discriminating markers that include 81 markers identified by us, the top 98 most PTC plasma-discriminating markers were selected for ThyMet. A 6-marker ThyMet classifier for PTC plasma was trained. In validation it achieved an Area Under the Curve (AUC) of 0.828, similar to thyroid ultrasonography (0.833) but at higher specificity (0.722 and 0.625 for ThyMet and ultrasonography, respectively). A combinatorial classifier by them, ThyMet-US, improved AUC to 0.923 (sensitivity = 0.957, specificity = 0.708).

INTERPRETATION

The ThyMet classifier improved the specificity of differentiating PTC from BTN over ultrasonography. The combinatorial ThyMet-US classifier may be effective in preoperative diagnosis of PTC.

FUNDING

This work was supported by the grants from National Natural Science Foundation of China (82072956 and 81772850).

Unintrusive multi-cancer detection by circulating cell-free DNA methylation sequencing (THUNDER): development and independent validation studies

BACKGROUND

Early detection of cancer offers the opportunity to identify candidates when curative treatments are achievable. The THUNDER study (THe UNintrusive Detection of EaRly-stage cancers, NCT04820868) aimed to evaluate the performance of ELSA-seq, a previously described cfDNA methylation-based technology, in the early detection and localization of six types of cancers in the colorectum, esophagus, liver, lung, ovary and pancreas.

PATIENTS AND METHODS

A customized panel of 161,984 CpG sites was constructed and validated by public and in-house (cancer: n=249; non-cancer: n=288) methylome data, respectively. The cfDNA samples from 1,693 participants (cancer: n=735; non-cancer: n=958) were retrospectively collected to train and validate two multi-cancer detection blood test models (MCDBT-1/2) for different clinical scenarios. The models were validated on a prospective and independent cohort of age-matched 1,010 participants (cancer: n=505; non-cancer: n=505). Simulation using the cancer incidence in China was applied to infer stage-shift and survival benefits to demonstrate the potential utility of the models in the real world.

RESULTS

MCDBT-1 yielded a sensitivity of 69.1% (64.8%‒73.3%), a specificity of 98.9% (97.6%‒99.7%) and tissue origin accuracy of 83.2% (78.7%‒87.1%) in the independent validation set. For early stage (I‒III) patients, the sensitivity of MCDBT-1 was 59.8% (54.4%‒65.0%). In the real-world simulation, MCDBT-1 achieved the sensitivity of 70.6% in detecting the six cancers, thus decreasing late-stage incidence by 38.7%‒46.4%, and increasing 5-year survival rate by 33.1%‒40.4%, respectively. In parallel, MCDBT-2 was generated at a slightly low specificity of 95.1% (92.8%-96.9%) but a higher sensitivity of 75.1% (71.9%-79.8%) than MCDBT-1 for populations at relatively high risk of cancers, and also had ideal performance.

CONCLUSION

In this large-scale clinical validation study, MCDBT-1/2 models showed a high sensitivity, specificity, and accuracy of predicted origin in detecting six types of cancers.

Machine Learning Quantified Tumor-Stroma Ratio Is an Independent Prognosticator in Muscle-Invasive Bladder Cancer

Although the tumor-stroma ratio (TSR) has prognostic value in many cancers, the traditional semi-quantitative visual assessment method has inter-observer variability, making it impossible for clinical practice. We aimed to develop a machine learning (ML) algorithm for accurately quantifying TSR in hematoxylin-and-eosin (H&E)-stained whole slide images (WSI) and further investigate its prognostic effect in patients with muscle-invasive bladder cancer (MIBC). We used an optimal cell classifier previously built based on QuPath open-source software and ML algorithm for quantitative calculation of TSR. We retrospectively analyzed data from two independent cohorts to verify the prognostic significance of ML-based TSR in MIBC patients. WSIs from 133 MIBC patients were used as the discovery set to identify the optimal association of TSR with patient survival outcomes. Furthermore, we performed validation in an independent external cohort consisting of 261 MIBC patients. We demonstrated a significant prognostic association of ML-based TSR with survival outcomes in MIBC patients (p < 0.001 for all comparisons), with higher TSR associated with better prognosis. Uni- and multivariate Cox regression analyses showed that TSR was independently associated with overall survival (p < 0.001 for all analyses) after adjusting for clinicopathological factors including age, gender, and pathologic stage. TSR was found to be a strong prognostic factor that was not redundant with the existing staging system in different subgroup analyses (p < 0.05 for all analyses). Finally, the expression of six genes (DACH1, DEEND2A, NOTCH4, DTWD1, TAF6L, and MARCHF5) were significantly associated with TSR, revealing possible potential biological relevance. In conclusion, we developed an ML algorithm based on WSIs of MIBC patients to accurately quantify TSR and demonstrated its prognostic validity for MIBC patients in two independent cohorts. This objective quantitative method allows application in clinical practice while reducing the workload of pathologists. Thus, it might be of significant aid in promoting precise pathology services in MIBC.

Systematic and benchmarking studies of pipelines for mammal WGBS data in the novel NGS platform

BACKGROUND

Whole genome bisulfite sequencing (WGBS), possesses the aptitude to dissect methylation status at the nucleotide-level resolution of 5-methylcytosine (5-mC) on a genome-wide scale. It is a powerful technique for epigenome in various cell types, and tissues. As a recently established next-generation sequencing (NGS) platform, GenoLab M is a promising alternative platform. However, its comprehensive evaluation for WGBS has not been reported. We sequenced two bisulfite-converted mammal DNA in this research using our GenoLab M and NovaSeq 6000, respectively. Then, we systematically compared those data via four widely used WGBS tools (BSMAP, Bismark, BatMeth2, BS-Seeker2) and a new bisulfite-seq tool (BSBolt). We interrogated their computational time, genome depth and coverage, and evaluated their percentage of methylated Cs.

RESULT

Here, benchmarking a combination of pre- and post-processing methods, we found that trimming improved the performance of mapping efficiency in eight datasets. The data from two platforms uncovered ~ 80% of CpG sites genome-wide in the human cell line. Those data sequenced by GenoLab M achieved a far lower proportion of duplicates (~ 5.5%). Among pipelines, BSMAP provided an intriguing representation of 5-mC distribution at CpG sites with 5-mC levels > ~ 78% in datasets from human cell lines, especially in the GenoLab M. BSMAP performed more advantages in running time, uniquely mapped reads percentages, genomic coverage, and quantitative accuracy. Finally, compared with the previous methylation pattern of human cell line and mouse tissue, we confirmed that the data from GenoLab M performed similar consistency and accuracy in methylation levels of CpG sites with that from NovaSeq 6000.

CONCLUSION

Together we confirmed that GenoLab M was a qualified NGS platform for WGBS with high performance. Our results showed that BSMAP was the suitable pipeline that allowed for WGBS studies on the GenoLab M platform.

Integrated analysis of circulating cell free nucleic acids for cancer genotyping and immune phenotyping of tumor microenvironment

The circulating cell-free nucleic acids (ccfNAs) consist of a heterogenous cocktail of both single (ssNA) and double-stranded (dsNA) nucleic acids. These ccfNAs are secreted into the blood circulation by both healthy and malignant cells via various mechanisms including apoptosis, necrosis, and active secretion. The major source of ccfNAs are the cells of hematopoietic system under healthy conditions. These ccfNAs include fragmented circulating cell free DNA (ccfDNA), coding or messenger RNA (mRNA), long non-coding RNA (lncRNA), microRNA (miRNA), and mitochondrial DNA/RNA (mtDNA and mtRNA), that serve as prospective biomarkers in assessment of various clinical conditions. For, e.g., free fetal DNA and RNA migrate into the maternal plasma, whereas circulating tumor DNA (ctDNA) has clinical relevance in diagnostic, prognostic, therapeutic targeting, and disease progression monitoring to improve precision medicine in cancer. The epigenetic modifications of ccfDNA as well as circulating cell-free RNA (ccfRNA) such as miRNA and lncRNA show disease-related variations and hold potential as epigenetic biomarkers. The messenger RNA present in the circulation or the circulating cell free mRNA (ccf-mRNA) and long non-coding RNA (ccf-lncRNA) have gradually become substantial in liquid biopsy by acting as effective biomarkers to assess various aspects of disease diagnosis and prognosis. Conversely, the simultaneous characterization of coding and non-coding RNAs in human biofluids still poses a significant hurdle. Moreover, a comprehensive assessment of ccfRNA that may reflect the tumor microenvironment is being explored. In this review, we focus on the novel approaches for exploring ccfDNA and ccfRNAs, specifically ccf-mRNA as biomarkers in clinical diagnosis and prognosis of cancer. Integrating the detection of circulating tumor DNA (ctDNA) for cancer genotyping in conjunction with ccfRNA both quantitatively and qualitatively, may potentially hold immense promise towards precision medicine. The current challenges and future directions in deciphering the complexity of cancer networks based on the dynamic state of ccfNAs will be discussed.

Expanding the Reach and Grasp of Lung Cancer Screening

Low-dose computer tomographic (LDCT) lung cancer screening reduces lung cancer-specific and all-cause mortality among high-risk individuals, but implementation has been challenging. Despite health insurance coverage for lung cancer screening in the United States since 2015, fewer than 10% of eligible persons have participated; striking geographic, racial, and socioeconomic disparities were already evident, especially in the populations at greatest risk of lung cancer and, therefore, most likely to benefit from screening; and adherence to subsequent testing is significantly lower than that reported in clinical trials, potentially reducing the realized benefit. Lung cancer screening is a covered health care benefit in very few countries. Obtaining the full population-level benefit of lung cancer screening will require improved participation of already eligible persons (the grasp of screening) and improved eligibility criteria that more closely match up with the full spectrum of persons at risk (the reach of screening), irrespective of smoking history. We used the socioecological framework of health care to systematically review implementation barriers to lung cancer screening and discuss multilevel solutions. We also discussed guideline-concordant management of incidentally detected lung nodules as a complementary approach to early lung cancer detection that can extend the reach and strengthen the grasp of screening. Furthermore, we discussed ongoing efforts in Asia to explore the possibility of LDCT screening in populations in whom lung cancer risk is relatively independent of smoking. Finally, we summarized innovative technological solutions, including biomarker selection and artificial intelligence strategies, to improve the safety, effectiveness, and cost-effectiveness of lung cancer screening in diverse populations.

Potential clinical utility of liquid biopsy in early-stage non-small cell lung cancer

BACKGROUND

Liquid biopsy has been widely researched for early diagnosis, prognostication and disease monitoring in lung cancer, but there is a need to investigate its clinical utility for early-stage non-small cell lung cancer (NSCLC).

METHODS

We performed a meta-analysis and systematic review to evaluate diagnostic and prognostic values of liquid biopsy for early-stage NSCLC, regarding the common biomarkers, circulating tumor cells, circulating tumor DNA (ctDNA), methylation signatures, and microRNAs. Cochrane Library, PubMed, EMBASE databases, ClinicalTrials.gov, and reference lists were searched for eligible studies since inception to 17 May 2022. Sensitivity, specificity and area under the curve (AUC) were assessed for diagnostic values. Hazard ratio (HR) with a 95% confidence interval (CI) was extracted from the recurrence-free survival (RFS) and overall survival (OS) plots for prognostic analysis. Also, potential predictive values and treatment response evaluation were further investigated.

RESULTS

In this meta-analysis, there were 34 studies eligible for diagnostic assessment and 21 for prognostic analysis. The estimated diagnostic values of biomarkers for early-stage NSCLC with AUCs ranged from 0.84 to 0.87. The factors TNM stage I, T1 stage, N0 stage, adenocarcinoma, young age, and nonsmoking contributed to a lower tumor burden, with a median cell-free DNA concentration of 8.64 ng/ml. For prognostic analysis, the presence of molecular residual disease (MRD) detection was a strong predictor of disease relapse (RFS, HR, 4.95; 95% CI, 3.06-8.02; p < 0.001) and inferior OS (HR, 3.93; 95% CI, 1.97-7.83; p < 0.001), with average lead time of 179 ± 74 days between molecular recurrence and radiographic progression. Predictive values analysis showed adjuvant therapy significantly benefited the RFS of MRD + patients (HR, 0.27; p < 0.001), while an opposite tendency was detected for MRD - patients (HR, 1.51; p = 0.19). For treatment response evaluation, a strong correlation between pathological response and ctDNA clearance was detected, and both were associated with longer survival after neoadjuvant therapy.

CONCLUSIONS

In conclusion, our study indicated liquid biopsy could reliably facilitate more precision and effective management of early-stage NSCLC. Improvement of liquid biopsy techniques and detection approaches and platforms is still needed, and higher-quality trials are required to provide more rigorous evidence prior to their routine clinical application.

Tumor fractions deciphered from circulating cell-free DNA methylation for cancer early diagnosis

Tumor-derived circulating cell-free DNA (cfDNA) provides critical clues for cancer early diagnosis, yet it often suffers from low sensitivity. Here, we present a cancer early diagnosis approach using tumor fractions deciphered from circulating cfDNA methylation signatures. We show that the estimated fractions of tumor-derived cfDNA from cancer patients increase significantly as cancer progresses in two independent datasets. Employing the predicted tumor fractions, we establish a Bayesian diagnostic model in which training samples are only derived from late-stage patients and healthy individuals. When validated on early-stage patients and healthy individuals, this model exhibits a sensitivity of 86.1% for cancer early detection and an average accuracy of 76.9% for tumor localization at a specificity of 94.7%. By highlighting the potential of tumor fractions on cancer early diagnosis, our approach can be further applied to cancer screening and tumor progression monitoring.

mHapTk: a comprehensive toolkit for the analysis of DNA methylation haplotypes

SUMMARY

Bisulfite sequencing remains the gold standard technique to detect DNA methylation profiles at single-nucleotide resolution. The DNA methylation status of CpG sites on the same fragment represents a discrete methylation haplotype (mHap). The mHap-level metrics were demonstrated to be promising cancer biomarkers and explain more gene expression variation than average methylation. However, most existing tools focus on average methylation and neglect mHap patterns. Here, we present mHapTk, a comprehensive python toolkit for the analysis of DNA mHap. It calculates eight mHap-level summary statistics in predefined regions or across individual CpG in a genome-wide manner. It identifies methylation haplotype blocks, in which methylations of pairwise CpGs are tightly correlated. Furthermore, mHap patterns can be visualized with the built-in functions in mHapTk or external tools such as IGV and deepTools.

AVAILABILITY AND IMPLEMENTATION

https://jiantaoshi.github.io/mhaptk/index.html.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Technological and computational advances driving high-throughput oncology

Engineering and computational advances have opened many new avenues in cancer research, particularly when being exploited in interdisciplinary approaches. For example, the combination of microfluidics, novel sequencing technologies, and computational analyses has been crucial to enable single-cell assays, giving a detailed picture of tumor heterogeneity for the very first time. In a similar way, these 'tech' disciplines have been elementary for generating large data sets in multidimensional cancer 'omics' approaches, cell-cell interaction screens, 3D tumor models, and tissue level analyses. In this review we summarize the most important technology and computational developments that have been or will be instrumental for transitioning classical cancer research to a large data-driven, high-throughput, high-content discipline across all biological scales.

Blood-based DNA methylation profiling for the detection of ovarian cancer

OBJECTIVES

Ovarian cancer is a fatal gynecological cancer due to the lack of effective screening strategies at early stage. This study explored the utility of DNA methylation profiling of blood samples for the detection of ovarian cancer.

METHODS

Targeted bisulfite sequencing was performed on tissue (n = 152) and blood samples (n = 373) obtained from healthy women, women with benign ovarian tumors, or malignant epithelial ovarian tumors. Based on the tissue-derived differentially-methylated regions, a supervised machine learning algorithm was implemented and cross-validated using the blood-derived DNA methylation profiles of the training cohort (n = 178) to predict and classify each blood sample as malignant or non-malignant. The model was further evaluated using an independent test cohort (n = 184).

RESULTS

Comparison of the DNA methylation profiles of normal/benign and malignant tumor samples identified 1272 differentially-methylated regions, with 49.4% hypermethylated regions and 50.6% hypomethylated regions. Five-fold cross-validation of the model using the training dataset yielded an area under the curve of 0.94. Using the test dataset, the model accurately predicted non-malignancy in 96.2% of healthy women (n = 53) and 93.5% of women with benign tumors (n = 46). For patients with malignant tumors, the model accurately predicted malignancy in 44.4% of stage I-II (n = 9), 86.4% of stage III (n = 59), 100.0% of stage IV tumors (n = 6), and 81.8% of tumors with unknown stage (n = 11). Overall, the model yielded a predictive accuracy of 89.5%.

CONCLUSIONS

Our study demonstrates the potential clinical application of blood-based DNA methylation profiling for the detection of ovarian cancer.