Dynamic, tumor-agnostic ctDNA analysis for MRD detection and longitudinal monitoring in pleural mesothelioma: The Janus-seq academic workflow.

Abstract

e20065 Background: Pleural mesothelioma (PM) is an aggressive thoracic malignancy with poor prognosis and limited tools for sensitive, real-time disease monitoring. Tissue biopsies are often invasive, difficult to obtain, and may not reflect tumor heterogeneity, while radiological imaging frequently detects progression late, particularly in patients with minimal residual disease (MRD). Circulating cell-free DNA (cfDNA) offers a minimally invasive alternative to capture systemic tumor burden; however, most assays are tumor-informed and require matched tissue, limiting scalability in PM. Tissue-independent and longitudinal monitoring strategies are urgently needed. Methods: Janus-seq is an in-house developed, multimodal, tumor-agnostic liquid biopsy workflow based on shallow whole-genome sequencing (sWGS). We applied Janus-seq for longitudinal monitoring in 23 PM patients enrolled in the prospective ONC/OSS-02/2020 monocentric clinical trial. From a single plasma sample, tumor fraction (TF), somatic copy number alterations (SCNAs), and fragmentomic profiles (FP) were simultaneously assessed. Targeted sequencing of 178 cancer-related genes was performed on the same cfDNA preparation to detect emerging single nucleotide variants (SNVs) without tumor tissue. Serial samples were analyzed to evaluate dynamic ctDNA changes and their association with radiological progression. Results: TF was quantifiable in 97.6% of plasma samples. FP discriminated PM patients from healthy controls ( p=0.01 ) and remained informative in TF-negative samples, supporting its value in low-shedding disease. TF and FP were moderately correlated (r²=0.56), indicating complementary biological signals. Longitudinal analysis showed that TF increases ≥20% anticipated radiological progression by up to 13 months, enabling early MRD detection. Plasma–tissue concordance was limited for SCNAs (58.1%) and SNVs (30.5%), highlighting ctDNA’s ability to capture systemic tumor heterogeneity. Serial profiling identified emerging actionable alterations, including PTCH1 , FGFR4 , and NOTCH3 . Conclusions: Janus-seq provides a sensitive, non-invasive, and tissue-independent strategy for MRD detection and longitudinal monitoring in PM. By integrating TF, FP, SCNAs, and SNVs from a single blood draw, this workflow enables early progression detection, captures clonal evolution, and supports clinical decision-making. Key implications include tumor-agnostic MRD detection without tissue, anticipation of progression months before imaging, and identification of emerging actionable mutations. Overall, Janus-seq represents a robust academic liquid biopsy approach with potential to improve patient management and advance precision medicine in thoracic oncology.