Research report
A research report is the sourced material a domain
dossier is synthesized from — generated on a plan and a cadence, one topic per file.
A report carries no confidence tags. Its bracketed markers say who might have
an incentive to shade a line; none of them says anyone checked it. To reach a score, a line has
to be drafted onto a candidate as unverified and pass the Verifier or the
Corroborator, like everything else.
Section epidemiology · Version 2026-09-01 · Cadence annual · Evidence class 2 published · Sources 3 · Supersedes none
Domain: ctdna-mrd-assay-validation · Scope: Public primary sources reachable by this repo's connectors, WebSearch and WebFetch. Scoped per the domain dossier's own stated adaptation — knowledge-base/domains/ctdna-mrd-assay-validation.md §1 defines "epidemiology" for this RUO/laboratory-consumable domain as testing volume and market growth of tumor-informed ctDNA MRD assays, not disease prevalence, and that adaptation is carried forward here. Excludes CPT descriptors (AMA-licensed), non-US/non-EU jurisdictions (no connector reaches them), and any figure whose only source lacks an identifier from this fleet's closed vocabulary (PMID, NCT, K-number, CFR section, HCPCS code, DOI, CELEX, FR Doc number) — see Section 5 for what that excluded.
Sourcing: Rests on one US-based peer-reviewed decision-analytic model for eligible-population sizing and two international peer-reviewed reports (a single-vendor real-world MRD-testing cohort and a global oncology-provider practice survey) for testing-pattern context; each is a single, non-corroborated source, and the two market-leading US assay makers' own SEC-filed test-volume figures were read but could not be cited under this fleet's identifier rules.
This pass found little fleet-citable, quantitative evidence of testing-volume trend for tumor-informed ctDNA MRD assays. One US peer-reviewed decision-analytic model estimated that, per 1 million covered lives, only 35 commercial-plan members and 102 Medicare Advantage members would be eligible in a given year for the single use case it modeled (stage II colon cancer, adjuvant-chemotherapy decision-making) [1]. [single-source] A retrospective single-vendor series from Asia and the Middle East — outside this domain's US scope, read for context — found 22% first-test ctDNA positivity across 215 patients (279 samples), with test-ordering timing concentrated either shortly after resection (stage II) or after completing curative-intent treatment (stage III) [2]. [single-source] A global, mostly non-US survey of 393 oncology providers found ctDNA testing predominates among liquid-biopsy modalities in current practice (77% of respondents), with cost and test availability the most-cited implementation barriers [3]. [single-source] Natera's and Guardant Health's own SEC-filed quarterly earnings exhibits were read directly in this pass and describe several consecutive quarters of rising company-reported "oncology" test volume, but no figure from those documents appears below — see Section 5 for why.
_Baseline (v1). No prior version; this establishes the starting point for future diffs._
A peer-reviewed budget-impact analysis modeled how large the age- and stage-eligible population for tumor-informed ctDNA testing is in US commercial and Medicare Advantage populations. Verbatim: "In hypothetical plans with 1 million individuals covered, 35 commercial health plan members and 102 Medicare Advantage members aged 75 years and younger were eligible for ctDNA testing" [1]. [single-source] The same analysis used a 50% adoption rate as its base-case assumption for projecting cost impact — a modeling input, not an observed adoption rate [1]. [single-source] The model is a decision-analytic exercise populated with published age-specific colon-cancer incidence and adjuvant-chemotherapy-use rates, restricted to stage II colon cancer patients facing an adjuvant-chemotherapy decision — narrower than either Signatera's or Guardant Reveal's full marketed use, which extends to additional stages and tumor types [1].
A retrospective observational study reported the first 215 consecutive colorectal cancer patients (279 samples) tested with Guardant Reveal, a tissue-free MRD assay, across laboratories in Asia and the Middle East — not the US population this domain is otherwise scoped to, read here for testing-pattern context rather than a US volume estimate. Verbatim: "Overall, 22% of patients had ctDNA detected in their first MRD test, and the frequency of ctDNA positivity increased with increasing tumour stage... In patients with stage II CRC, 71% of tests were ordered within 12 weeks after tumour resection, while for patients with stage III disease, 69% of tests were ordered after completion of all curative-intent treatment" [2]. [single-source] The same report notes the assay version changed partway through the series (132 of 279 samples tested on an earlier genomic-plus-epigenomic version, the remaining 147 on a methylation-only version), meaning the cohort is not a single consistent assay's testing volume over time [2].
A cross-sectional web survey of 393 oncology providers across 59 countries (53% practicing in Europe) captured current liquid-biopsy use, not specific to tumor-informed MRD assays or to colorectal cancer. Verbatim: "Liquid biopsy use was variable: 83% of respondents used it in ≤50% of patients... Circulating tumor DNA (ctDNA) testing predominated (77%), while circulating tumor cell (CTC) use was uncommon (4% alone; 18% in combination with ctDNA)... Cost (67%), test availability (55%), difficulty interpreting results (48%), and turnaround time (47%) were the most frequently cited barriers" [3]. [single-source] The survey does not break results out for a US-only subgroup in the portion of the abstract read, and its subject is liquid biopsy broadly — which includes companion-diagnostic and therapy-selection testing alongside MRD — not tumor-informed MRD assays specifically [3].
[1] Budget Impact Analysis of Circulating Tumor DNA Testing for Colon Cancer in Commercial Health and Medicare Advantage Plans — JAMA Health Forum (published 2024-05-03; accessed 2026-09-01). PMID 38819797 — https://pubmed.ncbi.nlm.nih.gov/38819797/ [peer-reviewed] [2] Utilization of tissue-free minimal residual disease testing in colorectal cancer patients from Asia and Middle East — Frontiers in Oncology (published 2024, month not given in the PubMed record; accessed 2026-09-01). PMID 39372864 — https://pubmed.ncbi.nlm.nih.gov/39372864/ [peer-reviewed] [3] Liquid biopsy in oncology practice: A global survey by the Young Committee of the International Society of Liquid Biopsy (ISLB) — The Journal of Liquid Biopsy (published 2026-09, day not given in the PubMed record; accessed 2026-09-01). PMID 42662884 — https://pubmed.ncbi.nlm.nih.gov/42662884/ [peer-reviewed]
Well established: Nothing in this report rests on two independent, non-affiliated sources reporting the same measured quantity — every quantitative line above comes from exactly one peer-reviewed document. There is no plainly-stated (unmarked) line in Section 3.
Thin: All three findings [1][2][3] carry [single-source] — each is exactly one peer-reviewed document, none independently corroborated by a second source measuring the same quantity.
Rescoped from class 3: None. This report answers descriptive, published-finding-shaped questions (what a model estimates, what a cohort measured, what a survey found) with no stakeholder-future question posed by the manifest row it fills.
Out of scope: Disease-population epidemiology of the underlying cancers (colorectal cancer incidence, stage distribution, or trend, including the independently well-documented rise in early-onset colorectal cancer) — per the domain dossier's own stated adaptation, "epidemiology" in this RUO/laboratory-consumable domain means testing volume and market growth, not disease prevalence, and that scoping decision is carried forward here rather than revisited. Non-US jurisdictions generally (no connector in this repo reaches non-US/non-EU laboratory or claims data; [2] and [3] are non-US findings read as adjacent context, not as US evidence).
A material, disclosed gap — SEC-filed company test-volume data read but not citable here: This pass directly fetched and read Natera, Inc.'s and Guardant Health, Inc.'s Form 8-K earnings-release exhibits for the four most recently reported quarters (Q4 2025 through Q2
companies' furnished exhibits describe several consecutive quarters of company-reported "oncology" test-volume growth (a category that includes, but does not isolate, their tumor-informed MRD assay), and neither company's own Form 10-Q discloses a unit or product-line breakdown at all — the figures exist only in the voluntarily furnished exhibit. None of that appears as a citable finding in Section 3, and no specific figure from those documents is stated anywhere in this report, because a SEC EDGAR accession number is not among this fleet's closed identifier vocabulary (PMID, NCT, K-number, CFR section, HCPCS code, DOI, CELEX, or Federal Register document number) enforced by check_research_reports.py. The domain dossier's own Section 1 already records specific figures from the same class of source under its own, less restrictive sourcing standard; this report does not repeat them, and records instead that they cannot currently be promoted into the research-fleet's more restrictive citation format.
Not searched vs. not found: A US-specific, claims-based or registry-based measurement of tumor-informed MRD testing volume or adoption rate over time was searched for directly, using more than a dozen distinct query formulations (including combinations of "SEER-Medicare", "national survey", "insurance claims", "utilization trend", "practice pattern", and "adoption", each paired with "circulating tumor DNA" and "minimal residual disease" or "colorectal cancer") across a 4-6 year PubMed window, and none was found. A PMID-bearing, US-specific real-world MRD testing cohort (as opposed to the ex-US cohort at [2]) was similarly searched for and not found. Whether a comparable figure exists in an NCCN or ASCO guideline update, a CAP proficiency-testing report, or a market-research house's licensed report was not searched this pass — the first two were queried in PubMed and returned no dedicated volume/adoption record; the third has no free, machine-queryable source per this repo's market connector and was not pursued through a paid source.
[inference] None in this report beyond what is already marked in Section 3's per-line markers; no additional cross-source synthesis is offered given how few independent findings this pass located.
| Proposition | Evidence class | Resolvable identifier | Dossier section |
|---|---|---|---|
| A decision-analytic budget-impact model estimated that, per 1 million commercially insured lives, 35 members aged 75 or younger were eligible for stage II colon cancer ctDNA testing in the modeled year, versus 102 among 1 million Medicare Advantage lives | 2 published | PMID 38819797 | 1 Condition & epidemiology |
| The same budget-impact model used a 50% adoption rate as its base-case assumption for projecting first-year cost savings of ctDNA testing in stage II colon cancer | 2 published | PMID 38819797 | 1 Condition & epidemiology |
| In a retrospective series of the first 215 consecutive colorectal cancer patients (279 samples) tested with Guardant Reveal in Asia and the Middle East, 22% had ctDNA detected on their first MRD test, with detection frequency increasing with tumor stage | 2 published | PMID 39372864 | 1 Condition & epidemiology |
| In the same series, 71% of stage II CRC patients' MRD tests were ordered within 12 weeks after tumor resection, while 69% of stage III patients' tests were ordered after completion of curative-intent treatment | 2 published | PMID 39372864 | 1 Condition & epidemiology |
| In a 393-respondent, 59-country survey of oncology providers, circulating tumor DNA testing predominated among liquid-biopsy modalities used (77%), while circulating tumor cell testing was uncommon (4% alone) | 2 published | PMID 42662884 | 1 Condition & epidemiology |