Latest Advances in Liquid Biopsy Technology

A biological miracle is happening in liquid biopsy diagnostics labs where a urine-based cell-free DNA methylation assay has demonstrated the ability to detect urological cancers including kidney cell carcinoma, bladder cancer, and upper tract urothelial cancer at Stage I disease, up to 4 years before conventional imaging detects any mass, with 97% sensitivity and 95% specificity in a 5,000-patient validation study — requiring nothing more than a routine urine sample collected at home. Cancerous cells shed methylated DNA fragments into the urine via the renal filtration system — fragments that carry cancer-specific epigenetic signatures invisible to conventional urinalysis but detectable by high-sensitivity methylation sequencing. The assay analyzes 4,500 cancer-associated CpG sites in a single urine sample, identifying which methylation patterns are present and computing a cancer probability score within 24 hours. Kidney cancer currently has a 93% 5-year survival rate if caught at Stage I — but 75% of cases are caught at Stage III or IV when survival drops below 10%. This test flips that statistic permanently. #UrineBiopsy #KidneyCancerDetection #LiquidBiopsy #EarlyDetection #drkevinramdhun

Could ‘priming’ improve ctDNA profiling?   On my recent travels, I had the chance to catch up on some of the literature and wanted to share what I thought was an interesting paper by Martin-Alonso et al., in Science, which could influence the future of liquid biopsies in cancer.   Liquid biopsies, including those that analyse cell-free DNA (cfDNA) from patients’ blood, are an important part of diagnosis, disease monitoring and molecular profiling. However, in oncology, the scarcity of circulating tumour DNA (ctDNA) means that enhancing liquid biopsy sensitivity is an important and necessary goal. It would improve the application of this technology if it were possible to get another 1-2 orders of magnitude improvement in sensitivity.   The authors of this paper hypothesized that reducing the causes of cfDNA loss – hepatic macrophage-mediated clearance and degradation by nucleases – would increase levels of ctDNA.   To test this hypothesis, they developed IV agents to target both mechanisms and gave them to healthy mice, in a process the authors called ‘priming’. Infusion of liposomal nanoparticles successfully curbed uptake of cfDNA by liver macrophages and the DNA-binding antibody, aST3, protected cfDNA from degradation by nucleases.   Priming of mice implanted with MC26 tumour cell lines boosted recovery of tumour-relevant ctDNA >10-fold. Priming also helped provide a more comprehensive tumour molecular profile and increased detection of small tumours from 75% in this mouse model.   The authors concluded that this approach could enhance ctDNA profiling, much like IV contrast agents have greatly improved the sensitivity of clinical imaging.   Clearly additional work is needed to translate these findings to clinical practice, including demonstration of safety of the agents used. The use of a priming strategy also complicates the testing procedure, but could be worth it for the improvement in sensitivity suggested. Current technology still has limitations on the sensitivity for small cancers which are exactly the ones we need to identify early via screening to make the biggest difference to mortality. Detection of minimal residual disease is another area where improvements in sensitivity would be desirable.   What do you think about this research? https://lnkd.in/gtsWhK9R 

Functional Liquid Biopsy Part 2: What if you didn't need a specialized assay? A week or so ago I talked about how #cfChIPSeq is transforming the liquid biopsy landscape by looking at active chromatin to infer tumor gene expression. But an incredible new preprint from Fred Hutch (Patton et al.) just pushed the boundary further. They’ve developed a deep learning framework capable of predicting single-gene expression directly from standard depth (~30-120x) WGS. By extracting fragmentomic and nucleosome positioning patterns across both promoters and gene bodies using a pair of tools called Triton & Proteus they can perform RNA-Seq in plasma. They apply this to SCLC subtyping, #ADCs, and resistance pathways e.g. Pluvicto neuroendocrine signatures that strongly correlated with rapid disease progression and resistance. Like other #FLBx tools, it’s likely to be best suited for advanced/metastatic disease rather than MRD, but this proves that the standard WGS data we are already collecting holds a massive, untapped functional playbook of the tumor. And opens the door to multiomics on WGS^2 (Whole Genome Squared) MRD assays like those from Veracyte, Inc., Foundation Medicine, Illumina, Inocras Inc. and Labcorp, bringing GX analysis to MRD when Tumor Fraction is high enough! Link to the full preprint, and my Substack, in the comments!

Researchers at Johns Hopkins University have made a potentially game-changing discovery in cancer diagnostics. They found that fragments of tumor DNA, known as circulating tumor DNA (ctDNA), can be detected in the bloodstream as early as three years before a clinical cancer diagnosis. This breakthrough opens up an extended window of opportunity for early cancer detection, which is critical for improving survival rates and enabling less invasive treatment options. The discovery was part of ongoing research into liquid biopsy techniques, which aim to identify cancer through simple blood tests instead of traditional imaging or tissue biopsies. These ctDNA fragments are shed by tumors into the bloodstream and can carry mutations specific to different types of cancers. Detecting them early may lead to personalized screening strategies, especially for individuals at higher risk. A study published in Nature Communications (2020) supports this claim. It discussed how early traces of cancer-related genetic alterations can be detected years before conventional methods would identify the disease. Such research is the backbone of companies like GRAIL, which are developing multi-cancer early detection tests based on similar principles. While the detection of tumor DNA years in advance is promising, it's important to note that: Not all cancers shed enough ctDNA to be detected early. Further validation across larger and more diverse populations is needed. There's a risk of false positives or overdiagnosis that must be managed. Nevertheless, this innovation represents a revolutionary leap forward in the field of oncology, offering hope for earlier intervention and better patient outcomes.

I'm excited to announce the publication of our team's latest work in npj Precision Medicine! We've developed MetaCH, a machine learning framework that improves the interpretation of liquid biopsies in cancer care. The promise of circulating tumor DNA (ctDNA) lies in its potential for disease monitoring and early diagnosis, offering a less invasive approach than traditional tumor biopsies. However, an obstacle to unlocking this potential is distinguishing true tumor-derived mutations from those arising from clonal hematopoiesis (CH), or age-related mutations in blood cells. MetaCH tackles this challenge by accurately classifying CH variants using only cell-free DNA from plasma samples, bypassing the need for costly and time-consuming matched white blood cell sequencing. MetaCH achieves this through a unique three-stage process: 🧬 The Mutational Enrichment Toolkit (METk) generates context-aware representations of mutations by integrating sequence context, gene information, and cancer type, capturing a more comprehensive picture of the mutational landscape. 🤖🤖🤖 Base classifiers trained on both large-scale public cancer and blood genomic datasets and a smaller, more detailed matched cfDNA dataset allow us to leverage the breadth of general cancer knowledge alongside the specificity of matched samples to score the CH-likelihood of variants. 🎯 A meta-classifier integrates the scores from the base classifiers, providing a final prediction of variant origin (tumor vs. CH). 🚀 MetaCH surpasses current classification methods across multiple types of cancer datasets to improve the accuracy of liquid biopsy-based cancer diagnostics and monitoring. ➡️ Learn more about MetaCH and its potential to transform cancer diagnostics: https://lnkd.in/eMhxhwNt Thanks to all co-authors! Gustavo Arango, Marzieh Haghighi, Gerald Sun, Elizabeth Choe, Aleksandra Markovets, J.Carl Barrett, Zhongwu Lai #PrecisionMedicine #AI #MachineLearning #CancerResearch #LiquidBiopsy #AstraZeneca #Oncology #ctDNA

Can nanopore sequencing revolutionize early disease detection? Current liquid biopsy techniques utilizing cell-free DNA often struggle to identify early-stage diseases due to their limited sensitivity. A recent groundbreaking study by Vikas Peddu, Karen Miga, Rebecca Fitzgerald, Daniel Kim, and their team from the University of California, Santa Cruz, and the University of Cambridge, published on bioRxiv, showcases the potential of long-read nanopore sequencing to be a game-changer. The researchers analyzed full-length cell-free RNA (cfRNA) from plasma samples to differentiate between healthy individuals, those with precancerous Barrett’s esophagus, and patients with esophageal adenocarcinoma. They discovered 270,679 novel intergenic cfRNAs and developed a tailored transcriptome reference for precise classification of both precancerous and cancerous conditions. Additionally, they pinpointed potential therapeutic targets within metabolic, signaling, and immune checkpoint pathways. These results highlight the efficacy of nanopore-based long-read RNA liquid biopsy platforms in early disease detection and targeted treatment, surpassing the capabilities of conventional methods. The methods developed in this study should be adaptable to detect other types of cancers using cfRNA in plasma. Exciting developments lie ahead in the realm of precision oncology! To delve deeper into the study, access the paper here: https://lnkd.in/erGvy_Wk #LiquidBiopsy #cfRNA #NanoporeSequencing #EarlyDetection #CancerDiagnostics #PrecisionMedicine #OncologyResearch

As we wrap up 2025, we share our collective insights on liquid biopsies, emphasizing cell-free DNA biology, both established and emerging technologies, and their clinical utility.  Here's the key questions addressed: -How is cfDNA biology leveraged for the development of fit-for-purpose liquid biopsy assays? -What are clinically validated assays for early cancer detection, and how do single-cancer tests differ from multi-cancer early detection liquid biopsy assays, especially with respect to their clinical implementation? -What are the challenges with cancer early detection (different for fragmentation vs methylation-based assays)? -What is the role of #AI integration in overcoming those? -How is plasma comprehensive genomic profiling (pCGP) guiding therapy? -What are the technical and biological challenges with pCGP? -What is hybrid capture NGS, and what are the different flavors of tumor-informed MRD assays from a technical standpoint? -How about residual disease detection in the early and metastatic setting? -What have we learned re: clinical sensitivity of ctDNA residual disease in predicting clinical endpoints (pathological response, disease recurrence)? -What is the emerging role of ctDNA dynamics or landmark status assessment in predicting immunotherapy response? -What would a ctDNA-interventional clinical trial design look like?  -How is/will ctDNA guide therapy optimization across cancers, stages and therapy settings? Importantly, we aggregated patient-level data (where available) from clinical trials in early-stage cancers and specifically computed the concordance and discordance rates between ctDNA residual disease and clinical endpoints. These analyses showed that while individuals with detectable ctDNA residual disease most frequently recur, there is a large fraction of individuals with undetectable ctDNA residual disease that also recur, highlighting the issues with the analytical sensitivity of currently used liquid biopsies. Fortunate to work with Blair Landon Akshaya Annapragada Noushin Niknafs and Victor Velculescu! Johns Hopkins Kimmel Cancer Center The Johns Hopkins University School of Medicine Johns Hopkins Medicine https://lnkd.in/egyMrevZ

Exciting news in the field of oncology and liquid biopsy! I’m sharing the latest research published in Nature Cancer: "A pan-cancer compendium of 1,294 plasma cell-free DNA methylomes and fragmentomes enabling multicancer detection". This study introduces the largest compiled multicancer plasma cfMeDIP-seq dataset to date, featuring 1,294 samples across 11 major cancer types, healthy controls, and individuals with Li-Fraumeni syndrome. What makes this work a milestone? * Multimodal Integration: While many studies look at one feature, this study integrated cfDNA methylome and fragmentome features, including 5′ end motifs, fragment ratios, and nucleosome footprints. This synergy significantly enhanced the robustness and accuracy of cancer detection and classification. * Standardized Workflow: To overcome the challenge of inconsistent data processing in liquid biopsy, the authors developed a uniform computational workflow to mitigate technical and biological batch effects across different cohorts. * A Single Assay Solution: Study demonstrated that cfMeDIP-seq can capture both methylation and fragmentation signatures in a single assay, which could streamline clinical workflows and reduce costs for large-scale screening. * Proven Generalization: Findings were successfully validated in an independent set of 220 samples, including cancer types entirely absent from our primary dataset. This compendium serves as a foundational resource for the scientific community to advance early cancer detection and monitor disease evolution more effectively. Congratulations to the extensive collaborative team at the University Health Network, University of Toronto, and the Ontario Institute for Cancer Research! #CancerResearch #LiquidBiopsy #NatureCancer #Genomics #PrecisionMedicine #EarlyDetection #Oncology OncoDaily Shivani Mahajan

🚀🧬 PanGIA Biotech, Inc. #Reports 97.8% #Sensitivity in #Urine-#Based AI Liquid Biopsy for #Prostate #Cancer A strong signal for the future of non-invasive, machine-learning-driven cancer diagnostics. PanGIA Biotech, Inc. announced a peer-reviewed clinical validation study in Diagnostics showing its PanGIA Analysis System (PAS) achieved 97.8% sensitivity across all Gleason grades, 97.3% specificity for high-grade disease, and an AUC of 0.91 in detecting prostate cancer from urine. 🗣️ Key leaders: Holly Magliochetti, CEO and Co-Founder, PanGIA Biotech, Inc. @Obdulio Piloto, PhD, Chief Scientific Officer and Co-Founder, PanGIA Biotech, Inc. 📊 Key study highlights: 👥 283 participants 🏥 26 U.S. urology practices 🧪 Urine-based liquid biopsy 🤖 Machine learning biomolecular pattern recognition 🎯 97.8% sensitivity 🔬 97.3% specificity for high-grade cancers 📈 AUC 0.91 Why this matters 👇 1️⃣ 🚻 Urine becomes a serious oncology sample type A non-invasive urine workflow dramatically improves patient compliance, repeat monitoring, and screening scalability, especially in prostate cancer. 2️⃣ 🤖 Pattern-based ML may outperform single-marker logic Instead of relying on PSA-like reductionist biomarkers, PAS reads complex metabolite and biochemical signal signatures, potentially capturing earlier or harder-to-classify disease states. 3️⃣ 🏥 Major potential to reduce unnecessary biopsies This could meaningfully improve urology workflow triage, helping reduce invasive procedures while improving confidence in high-grade disease referral. 4️⃣ 🌍 Platform expansion opportunity The bigger story may be the platform architecture—urine + CSF + ML pattern recognition creates a highly expandable engine for multi-cancer and neurologic diagnostics. 💡 My takeaway This is exactly where liquid biopsy innovation is heading: from single analytes to multidimensional biological pattern intelligence. The most strategic value here is not only prostate cancer performance, but the validation that machine learning can translate urine’s biochemical complexity into clinically actionable decisions. That could become highly disruptive for screening, longitudinal monitoring, and decentralized diagnostics. #LiquidBiopsy #ProstateCancer #MachineLearning #AI #Diagnostics #PrecisionMedicine #Urology #Oncology  https://lnkd.in/guyxj8fe

I am proud to share Adela's new clinical validation paper in npj Precision Oncology to monitor patient response to immunotherapy: “Clinical validation of a tissue-agnostic, genome-wide methylome enrichment assay to monitor response to pembrolizumab.” In this blinded study using plasma samples from the INSPIRE pembrolizumab trial in advanced solid tumors, early on-treatment changes in methylated ctDNA were strongly associated with response, clinical benefit, and improved outcomes, including progression-free and overall survival. These findings reinforce the promise of blood-based monitoring to provide clinically meaningful insight earlier in the course of treatment. This is significant as it makes testing universally accessible. By monitoring patient response through a simple blood draw, without requiring tumor tissue, we open the door to broader access, easier integration into care, faster time to answers, and a less burdensome experience for patients. Our vision is to help make precision oncology more available in the real world — across more cancers, more treatment settings, and more patients. This is the second blinded clinical validation of our cfDNA methylome platform, following our tissue-free MRD clinical validation in head and neck cancer published in Annals of Oncology. Together, these studies support the broader potential of a single, genome-wide methylation assay to help transform cancer monitoring from a tissue-dependent paradigm to one that is more accessible, scalable, and patient-centered. We believe the future of oncology will be shaped by technologies that meet patients where they are, reduce barriers to access, and give clinicians better information when it matters most. This is an important step in that direction. https://lnkd.in/eqD_RQxV #PrecisionMedicine #CancerCare #PrecisionOncology #Immunotherapy #LiquidBiopsy #CancerMonitoring #PatientAccess