PROJECT ABSTRACT Liquid biopsies have made significant strides in cancer medicine, with applications spanning early cancer detection, assessment of therapy effectiveness and drug-resistance, monitoring tumor-dynamics and early detection of minimal residual disease, MRD. Yet, substantial questions remain, prohibiting translation to clinical practice on many occasions. For instance, the inability to detect tumor-circulating DNA (ctDNA) in a fraction of gliomas, gastric and renal cancers, even for advanced stages, has led to the notion that some tumors are ‘non- shedders’, thus postulating there are biological processes preventing some cancers from shedding DNA in the blood. If true, this would prohibit the use of ctDNA as a tumor biomarker that precedes imaging and clinical manifestation of disease. Similar questions apply to early-stage tumors that test negative for ctDNA, but patients do experience clinical recurrence: was ctDNA never there or was it simply too low to be detected? Yet, improved detection techniques applied to ‘non-shedder’ cancers have recently led to major increases in the ctDNA detection rate, converting previously ctDNA-negative samples to positive. This indicates that possibly all cancers shed DNA in the circulation, but some can only be detected with methods providing extremely high sensitivity. This R01 competing continuation builds on two technologies developed during the first grant cycle and can address the questions. (a) MAESTRO -minor allele enriched sequencing through recognition oligonucleotides- that enables extremely high sensitivity MRD detection. And (b) Duplex Repair (DR) that reduces noise during subsequent duplex sequencing. We aim to show that the combined technology, DS-DR- MAESTRO enables detection of ctDNA as sensitive as a single tumor-cell genome-equivalent circulating in the ~5l human blood pool at the time of blood collection, while retaining specificity. This sensitivity can be achieved following sequencing of the primary tumor and collecting total circulating-DNA (cfDNA) typically recovered from a standard 10-40ml blood collection. We will optimize the technology to the level required for subsequent application in clinical studies. Aims 1-2 will optimize the technical aspects of the technology using serial dilutions of cfDNA from tumor patients and normal volunteers. Aim 3 will apply the technology for detection of minimal residual disease in circulating DNA from esophageal cancer patients whose tumors tend to shed very little ctDNA. Successful technology development is anticipated to extend the use of cfDNA for detecting MRD in low-shedder cancers and significantly shrink the size of cancers detectable. It will also increase the predictive power for multiple endpoints, and assist pre/post-surgery ctDNA management decisions, thereby providing a window of opportunity for many additional patients and improving precision medicine.