Liquid Biopsy Index

Methods & Provenance

This index is built to be auditable: every figure traces to a source record, and the retrieval and classification are locked and versioned. Three tracks are published: research papers, patents and clinical trials. They are counted separately, in different units.

Research Papers

The Field Definition

A locked, versioned PubMed query (q3-2026-10-09) retrieves records involving a biofluid-derived sample and a recognized analyte class (ctDNA/cfDNA, CTCs, EVs/exosomes, cfRNA, circulating proteins/metabolites), across oncology, prenatal, transplant, and other applications. Before its current vocabulary existed the field wrote plasma DNA, circulating nucleic acids, fetal DNA in maternal plasma, tumor cells in the blood and circulating RNA; the query includes those terms for every year, and the numbers published under the earlier query are preserved in the method record.

Classification

Each record is scope-gated and tagged against a locked rubric (v4.1-2026-08-18) by Claude Haiku at temperature 0, along four independent axes: application, oncology breadth, analyte, and study purpose. The application axis has six categories ordered by the care pathway: screening, diagnosis, prognosis, therapy selection, treatment-response monitoring, and minimal residual disease. Oncology breadth has three values: one named cancer, two or more named cancers studied together, or a pan-cancer / tissue-agnostic approach.

Two rules are enforced outside the model. Every record tagged diagnosis gets a second, focused call that asks only which population the abstract describes; the tag is kept only for diagnostic workup and review articles, and dropped on case-control and other study designs. And the rule that application tags attach only to oncology records is applied in code after the model answers, not left to the prompt. Records the classifier is unsure about are flagged for review rather than silently kept or dropped. Before any bulk run, a candidate prompt must pass two mandatory gates: a per-record calibration gate on a hand-labeled gold set, and a population-level check that compares tag rates on a pinned sample against thresholds, because a prompt can be right record by record and still wrong in aggregate.

Trend lines count review articles and primary research papers together, for every tag. The mix varies by tag, and it runs most review-heavy for diagnosis: under the strict definition, primary studies of diagnostic workup are rare, so much of what carries that tag is review literature discussing the use.

Coverage

Coverage is complete: every record the query returns is classified individually (58,364 records, of which 40,475 are in scope; the earliest in-scope paper is from 1960). The counts need no estimation or weighting: each paper is counted once, in the year it first appeared. The research charts start in 1990, the first year with at least 5 in-scope papers; the 85 earlier in-scope papers are in Explore.

Reading The Research Figures

Geography, impact and industry affiliation are joined from OpenAlex. A paper's country is that of its corresponding or first author. When an author lists more than one institution, the index credits the one named first in that author's own affiliation. Impact uses the field-weighted citation impact (FWCI), which normalizes citations for field and age. The methods share is a lower bound: its stability over time is the reliable signal, not the exact level.

Patents

How Patents Are Retrieved

Patents are not found by keyword. Retrieval runs on a frozen family of CPC classification codes (v1.1, edition 2026.01) applied to the USPTO's bulk patent data, which pulls 46,410 candidate US-granted patents. The code family is deliberately a recall filter: it decides what gets read, not what counts.

The Same Scope Test As The Research Track

Every candidate is then read by the same Claude Haiku classifier at temperature 0, against a patent-side rubric (trackB-3part-v1.4-2026-08-17) that applies the two tests the paper rubric applies: is the analyte circulating or cell-free, and is it measured as a disease biomarker? A patent passes on what it claims, not on the CPC code that surfaced it. Result: 3,157 in-scope US documents. The same calibration gate is mandatory here before any run.

The Counting Unit Is An Invention, Not A Document

One invention is typically filed many times: continuations, divisionals, and parallel filings in other patent offices. Counting documents would therefore count the same idea repeatedly. In-scope documents are collapsed onto their INPADOC patent family, so a continuation chain or a multi-office filing counts once: 3,157 documents become 2,195 inventions (30.5% collapsed). Inventions are dated by earliest priority year (when the idea was first filed, not when it granted).

Reading The Patent Figures

The candidate patents come from the USPTO's PatentsView bulk product, not from a search service. Impact is age-normalized: the highest-impact inventions are ranked by forward citations per year since grant, never by raw counts, which would rank the oldest patents first.

Who Owns An Invention

A patent can name several owners, and many do: university with hospital, company with the lab whose work it licensed. Every organization named on any document in the family is kept, so 368 of the 2,195 US inventions have more than one owner, and 1,203 distinct organizations appear across 2,579 ownership links. A co-owned invention is therefore counted once under each of its owners: an owner ranking adds up to more than the number of inventions, because it measures share of ownership, not a division of the total. 92 inventions are held by individuals rather than an organization.

Owner names come as the patent offices recorded them, and the same organization is sometimes written two ways on two documents of one family: a misspelling, a national subsidiary, or a personal name in the other word order. Where that happens the invention reads as co-owned when it has a single owner. A conservative screen finds this on about 24 US and 27 China inventions; the pairs are listed openly and are resolved by hand against the registry, never merged automatically, because a parent and its subsidiary look identical to any automatic rule and are not the same owner.

The Separate China View

The US-anchored set only contains an invention if it was filed in the United States. Inventions filed only in China are retrieved separately from Google Patents public data and classified with the same rubric: 102,157 patent families read, 8,689 in scope. This is published as its own view rather than added to the US number, because the two are not directly comparable: the China set includes pending applications as well as grants, its abstracts are machine-translated, and it carries no citation data, so it has no impact ranking.

Applications That Have Not Granted Yet

A patent takes years to grant, so the granted-only view runs out before the present: the recent priority years are not small, they are unfinished. US pre-grant publications (applications the USPTO publishes ~18 months after filing, whether or not they ever grant) fill that gap. 15,437 still-pending applications were read by the same classifier against the same rubric, giving 1,435 in-scope documents and 1,366 pending inventions once collapsed onto patent families. Applications that already became one of the granted patents above are removed by an exact USPTO crosswalk, so nothing is counted twice, and pending is dated by the same earliest-priority rule. Pending is always shown as its own band, never merged into the granted number.

Two filters were applied before spending anything on classification. Applications filed before 2020 that never granted are abandoned, not pending: grant rate by filing cohort stops moving after about 2018, so only 2020-and-later filings are treated as live. And a set defined by absence from a grant list inherits every other reason a record could be absent, so the absence was checked directly rather than assumed.

Patent counts look small next to a keyword search on a patent site, and are meant to. They count in-scope, de-duplicated inventions, not documents matching a search term.

Clinical Trials

Clinical trials are read from ClinicalTrials.gov registrations. Candidate registrations are retrieved by a locked vocabulary search plus two recall channels (company products, and trials cited by in-scope papers), and each record is classified against a locked trial rubric: in or out of scope by whether the registration itself describes a liquid-biopsy assay in the trial’s own conduct, then the trial type (a test trial, a treatment selection trial, an eligibility selection trial, or a correlative study trial carrying the measurement) and its analytes. A hand-labeled calibration gate and a population-level check must both pass before any run, and a review of sampled results follows. 5,289 registrations were read; 4,302 are in scope.

Counting rules: a trial is one registration, counted once, in the year its registration was first posted, from 2000 onward. The active-in-year and completion views use actual dates only; withdrawn trials stay in the registration count and out of the active count. Registrations that never describe their liquid-biopsy content stay out, even when a paper cites the trial: the index counts what registrations say.

Registry History

The trial timeline figures rest on archived copies of ClinicalTrials.gov, one per quarter from 2017 to 2023 and one per month from 2024, plus this index's own fetch of the current registry. Each copy keeps the start, primary completion and completion dates and whether each was estimated or actual, the enrollment figure, the status, and the outcome measures. A change is dated to the interval between the two copies in which it appears, and month-level dates are compared at the month. Summary figures use the median and quartiles, and trials whose first recorded primary completion date sat more than ten years after their start date are flagged and shown separately. The chart of how an expected date holds up over time follows one fixed group of trials: each trial's date is read from the first yearly copy that records it as an estimate still ahead of that copy's date and compared with the yearly copies one to five years later, rounded to the nearest year, and only trials with a copy at every one of those five points are included.

The Change Vocabulary

Consecutive copies of a record are compared field by field. A completion date that moves later is a slip, one that moves earlier is a pull-in, and one that turns from estimated to actual is an actualization, which is not a slip. Changes to the start date, enrollment figure, enrollment target and status are each recorded under their own name. Outcome measures are compared as sets after normalizing case, spacing and trailing punctuation, so a reworded measure counts and a reordered list does not, and a set that is empty in one copy is recorded as missing from that copy rather than as a change. The change figures compare outcome measures by their titles only; each trial's page also shows changes to a measure's time frame. A trial not yet registered in the older copy is skipped for that interval.

License And Citation

Free-tier data and text: CC BY 4.0. Cite as “Liquid Biopsy Index, liquidbiopsyindex.org, rubric v4.1-2026-08-18, accessed 2026-10-13”.

Known Limitations