Liquid Biopsy Patents

Liquid biopsy patents cover cell-free-DNA (cfDNA) assay innovations; methylation and fragmentomics innovations; minimal-residual-disease (MRD) and tumor-informed innovations; and error-correction-sequencing and bioinformatics innovations — with IP held by multi-cancer-screening, therapy-selection, and MRD-monitoring companies. MAJOR LIQUID-BIOPSY PATENT HOLDERS: GRAIL (Illumina): Galleri, a multi-cancer early detection MCED test reading DNA-methylation patterns in cell-free DNA to detect many cancers (and predict tissue of origin) from a blood draw — methylation-classifier and assay IP. GUARDANT HEALTH: Guardant360 (ctDNA comprehensive genomic profiling for therapy selection), Shield (blood-based colorectal-cancer screening), and Reveal/MRD; deep ctDNA-assay and digital-sequencing IP. FOUNDATION MEDICINE (Roche): FoundationOne Liquid CDx (a companion-diagnostic ctDNA test). NATERA: Signatera — a TUMOR-INFORMED (personalized, bespoke per-patient panel designed from the patient's tumor) molecular-residual-disease test for recurrence monitoring. OTHERS: Exact Sciences (Cologuard, Oncotype, MRD), Freenome (multiomics screening), Delfi Diagnostics (fragmentomics — cfDNA fragment-size patterns), Caris, Tempus, and the foundational sequencing/error-correction holders. Methylation, fragmentomics, tumor-informed MRD, and error-suppressed sequencing are the core liquid-biopsy patent domains — and §101 (natural-correlation) is the dominant validity risk.

Natural-phenomenon §101 risk; assay-improvement claiming; specific-technique claiming; and method-of-treatment framing represent the central liquid-biopsy patentability problem — because detecting a naturally-occurring biomarker can be held an unpatentable natural phenomenon. THE §101 PROBLEM: in Ariosa v. Sequenom (Fed. Cir. 2015), claims to detecting cell-free FETAL DNA in maternal blood (a real, valuable discovery) were INVALIDATED — the court held that the cffDNA is a natural phenomenon and that detecting it with conventional techniques added nothing 'significantly more' (Mayo/Alice step 2). This is the existential liquid-biopsy patent risk: a claim to 'detect [naturally-occurring biomarker] in blood and correlate with [disease]' is vulnerable as a natural phenomenon/law of nature plus routine steps. SURVIVING §101: claims must recite a SPECIFIC, non-conventional technical improvement — a novel assay or sample-processing method, a specific error-correction/sequencing technique that wasn't routine, an unconventional combination, or a particular molecular method — not merely the correlation. METHOD-OF-TREATMENT FRAMING: claiming a method that USES the biomarker result to administer a specific treatment (a method of treatment, not just diagnosis) is more §101-durable (Vanda/CellzDirect lineage). DEVICE/COMPOSITION FRAMING: claiming specific probes, panels, reagents, or a particular machine/process avoids the bare-correlation trap. The lesson: patent the NOVEL TECHNIQUE (assay, chemistry, error correction), not the natural correlation — and consider trade-secret protection for classifiers.

Cell-free-DNA assay and library innovations; methylation and fragmentomics innovations; error-correction-sequencing innovations; and tumor-informed MRD and classifier innovations represent core liquid-biopsy patent domains — and these specific TECHNIQUES (not the correlations) are the durable IP. cfDNA-ASSAY PATENTS: blood collection/stabilization, cell-free-DNA extraction and library preparation optimized for the tiny, fragmented quantities of circulating tumor DNA, and targeted vs. whole-genome approaches. METHYLATION PATENTS: methylation-based cancer detection and tissue-of-origin (Grail) — bisulfite-free/enzymatic methylation sequencing, methylation-marker panels, and methylation classifiers (the specific assay/chemistry is patentable; the classifier is often trade-secret or claimed with the assay). FRAGMENTOMICS PATENTS: analyzing cfDNA fragment-size and end-position patterns (cancer changes fragmentation — Delfi), as a specific technical method. ERROR-CORRECTION SEQUENCING PATENTS: unique molecular identifiers UMIs, duplex/molecular barcoding, and computational error suppression to detect rare tumor variants at very low allele fraction (Guardant Digital Sequencing) — a specific, non-routine technical improvement (good §101 footing). MRD PATENTS: tumor-informed PERSONALIZED panels (designing a bespoke assay from the patient's tumor — Natera Signatera) versus tumor-naive fixed panels, and ultra-sensitive detection of residual disease. The specific assay chemistry, error-correction technique, and personalized-panel design are the highest-value, most-§101-durable liquid-biopsy IP.

Liquid biopsy startup IP strategy must navigate Grail methylation patents, Guardant digital-sequencing/ctDNA patents, Natera tumor-informed MRD patents, Foundation companion-diagnostic patents, the dominant §101 natural-phenomenon risk (Ariosa/Mayo — bare biomarker-detection claims are weak), CLIA/FDA regulatory pathways (LDT vs. PMA), and a landscape where the specific technique and clinical evidence/data matter as much as patents; understand that you CANNOT durably patent the natural correlation (a biomarker associates with a disease), so the durable IP is the NOVEL TECHNIQUE — assay chemistry, methylation/fragmentomics method, error-correction sequencing, personalized-panel design, and sample processing — plus method-of-treatment framing and trade-secret classifiers; identify whitespace in novel assay chemistries, error suppression, multi-analyte/multiomics, MRD sensitivity, and screening. LIQUID-BIOPSY STARTUP IP STRATEGY: PATENT THE TECHNIQUE, NOT THE CORRELATION (§101 IS EXISTENTIAL): Ariosa/Mayo make bare 'detect biomarker → diagnose disease' claims invalid — patent the specific, non-conventional assay chemistry, methylation/fragmentomics method, error-correction sequencing, and sample processing (a concrete technical improvement), and frame method-of-treatment claims where possible; ERROR-CORRECTION AND NOVEL ASSAY CHEMISTRY ARE THE STRONGEST IP: duplex/UMI error suppression and novel enzymatic-methylation/fragmentomics techniques are specific technical improvements with good §101 footing and real clinical value; CLASSIFIERS ARE OFTEN TRADE-SECRET: ML classifiers (the methylation/multiomics model + training data) are §101-vulnerable and best protected as trade secrets and via data scale, not patents; PERSONALIZED MRD AND MULTIOMICS ARE HIGHEST-VALUE WHITESPACE: tumor-informed bespoke MRD (Signatera-style) and multi-analyte (methylation + fragmentomics + protein) screening are differentiated, patentable techniques; CLINICAL EVIDENCE/DATA AND CLIA/FDA ARE PARALLEL MOATS: prospective validation, guideline inclusion, and regulatory clearance gate adoption as much as IP; WHEN TO PATENT: NOVEL TECHNIQUE WITH MEASURED PERFORMANCE: file once an assay shows measured results (sensitivity/specificity + limit of detection (variant allele fraction) + tissue-of-origin accuracy + MRD lead time) AS A SPECIFIC METHOD — measured LOD/sensitivity/specificity and the novelty of the technique are the critical liquid-biopsy IP metrics; KEY FTO CHECKLIST: Grail Galleri methylation MCED tissue-of-origin classifier; Guardant Digital Sequencing ctDNA UMI/error-suppression, Shield screening; Natera Signatera tumor-informed personalized MRD panel; Foundation Liquid CDx companion diagnostic; Delfi fragmentomics; enzymatic/bisulfite-free methylation sequencing; duplex/molecular barcoding error correction; Ariosa v. Sequenom / Mayo §101 natural-phenomenon (CENTRAL RISK); method-of-treatment framing; classifier trade-secret; CLIA LDT / FDA PMA.