Why This Matters for Oncology Teams
Multicancer early detection (MCED) testing has moved from theoretical concept to active clinical deployment within the last three years. The Galleri test (GRAIL) is now available in the United States. Two large prospective studies — the NHS-Galleri trial (140,000+ participants) and SYMPLIFY (5,461 symptomatic patients) — have reported results. Real-world data from more than 111,000 tests have been published. The evidence is now sufficient to form a considered clinical view.
For oncology teams, MCED matters across several dimensions: as a technology that will increasingly appear in patients' medical histories, as a potential tool in clinical trial eligibility and biomarker strategies, and as a scientific framework — cfDNA methylation analysis — that is being applied in other contexts including treatment response monitoring and minimal residual disease detection.
This article covers the biology, the technology, the clinical evidence, and what it means for trial design and site operations.
The Biology of Circulating Tumour DNA
All nucleated cells continuously shed cell-free DNA (cfDNA) into the blood through apoptosis and necrosis. In healthy individuals, the majority of cfDNA originates from white blood cells (haematopoietic cells), with smaller contributions from other tissues. In patients with cancer, tumour cells also shed DNA — this tumour-derived fraction is called circulating tumour DNA (ctDNA).
The ctDNA fraction of total cfDNA is highly variable, driven primarily by tumour burden. In patients with advanced, high-burden disease, ctDNA can constitute 10–40% of total cfDNA. In patients with localised, early-stage cancer, it is typically 0.01–0.1% — a vanishingly small signal in a large background of normal cfDNA. This is why stage I sensitivity is fundamentally limited for any liquid biopsy test.
What Makes Methylation Unique
DNA methylation is an epigenetic modification: a methyl group (–CH₃) is covalently attached to cytosine bases at CpG dinucleotides, without altering the underlying DNA sequence. Methylation regulates gene expression — methylated promoters typically silence gene transcription, while unmethylated promoters allow transcription.
Cancer involves widespread epigenetic reprogramming. Two patterns occur simultaneously:
- Global hypomethylation — repetitive elements and proto-oncogene promoters that are normally methylated become unmethylated, leading to chromosomal instability and oncogene activation
- Focal hypermethylation — CpG islands at tumour suppressor gene promoters become methylated and silenced (e.g., MLH1 in microsatellite-unstable tumours, CDKN2A in multiple tumour types, BRCA1 in some breast cancers)
Critically, these methylation patterns are tissue-of-origin-specific. The methylation signature of a colon cancer cell differs from that of a lung cancer cell, a pancreatic cancer cell, and a normal haematopoietic cell. This tissue-specificity is what enables MCED tests to predict not just that a cancer signal is present, but which organ it is likely coming from.
How the Galleri Test Works
The Galleri test (GRAIL, now an Illumina subsidiary following regulatory divestiture) uses whole-genome bisulfite sequencing of cfDNA. Bisulfite treatment converts unmethylated cytosines to uracil (subsequently read as thymine during sequencing) while methylated cytosines are protected and read as cytosine. Comparing bisulfite-converted sequences against the human reference genome reveals the methylation status of individual CpG sites genome-wide.
The test processes approximately 3–4 mL of plasma from a standard blood draw. Machine learning models trained on methylation profiles from thousands of cancer and non-cancer samples analyse the resulting data to:
Cancer Signal Detection
The model determines whether the methylation pattern in the sample is consistent with a cancer signal — or with normal, non-cancer cfDNA. A binary output (signal detected / signal not detected) is produced.
Tissue of Origin Prediction
If a cancer signal is detected, a second model analyses which organ the methylation pattern most closely resembles. The output is a predicted tissue of origin with confidence level — guiding the clinician toward appropriate imaging or biopsy.
Result Reporting
Results are reported to the ordering physician as: cancer signal not detected; or cancer signal detected with predicted tissue of origin and recommended follow-up. The test does not provide a cancer diagnosis — that requires confirmatory workup.
Clinical Follow-Up
A positive result initiates diagnostic workup — typically imaging (CT, MRI, PET) guided by the predicted tissue of origin. Subsequent endoscopy, biopsy, or pathology provides or excludes a cancer diagnosis. A negative result does not exclude cancer.
Performance Metrics — What the Data Show
Three datasets now provide performance data for Galleri across different clinical populations.
Sensitivity by Stage: The Critical Table
Overall sensitivity statistics obscure the most clinically important finding: sensitivity is strongly stage-dependent, and this has direct implications for the test's utility in a screening context.
| Cancer Stage | Galleri Sensitivity (Asymptomatic) | Why It Matters | 5-Year Survival (Reference) |
|---|---|---|---|
| Stage I | 16.8% | Misses 83% of cancers at the stage where outcomes are best | ~90–99% (tumour-dependent) |
| Stage II | ~40% | Moderate; majority of potentially curable disease still missed | ~70–85% (tumour-dependent) |
| Stage III | ~77% | Good detection but many stage III cancers are not curable | ~40–65% (tumour-dependent) |
| Stage IV | 90.1% | Excellent — but stage IV disease is largely incurable | ~15–20% (tumour-dependent) |
| All Stages | 51.5% | Overall rate in asymptomatic validation cohort | — |
The SYMPLIFY Study — Symptomatic Patients
The SYMPLIFY study evaluated Galleri in 5,461 patients referred through the urgent suspected-cancer pathway in England and Wales. Because these are symptomatic patients with a higher pre-test probability of cancer, sensitivity figures are higher than in asymptomatic cohorts: 66.3% overall sensitivity, 98.4% specificity, 85.2% accuracy for tumour origin prediction.
The SYMPLIFY findings suggest a potential use case for MCED testing in the symptomatic diagnostic pathway — particularly for cancers that are difficult to diagnose rapidly through conventional workup (e.g., pancreatic cancer, ovarian cancer). Whether this application reduces time-to-diagnosis and improves outcomes is not yet established.
The NHS-Galleri RCT — What 140,000 Participants Showed
The NHS-Galleri trial randomised participants to receive annual Galleri screening (up to three rounds) or standard care. With 140,000+ participants, it is one of the largest RCTs of a cancer screening intervention ever conducted.
What Was Shown
- Increase in early-stage cancer diagnoses in screening arm
- Reduction in emergency cancer presentations
- High specificity maintained at population scale
- Tissue-of-origin prediction guided efficient workup
What Was Not Shown
- Primary endpoint NOT met: no significant reduction in stage III/IV incidence
- No mortality data yet (follow-up ongoing)
- No quality-of-life benefit demonstrated
- No cost-effectiveness data from RCT context
The failure to meet the primary endpoint — a statistically significant reduction in advanced-stage cancer diagnoses — is the most important finding to understand. The trial was powered to detect this difference if it existed. Its absence means that at the group level, in this UK population over the trial period, the screening intervention did not significantly shift the stage distribution of cancers at diagnosis.
There are several plausible explanations: insufficient follow-up duration (lead-time effects take years to manifest in stage distribution), stage migration may have occurred but not reached statistical significance, or the sensitivity at early stages is simply too low to meaningfully shift population-level stage distribution within the trial period. Longer follow-up — and eventually, mortality data — will clarify this.
"Although the concept is appealing, tests for detecting multiple types of cancer still require careful validation because confirmatory tests may be imperfect, the costs and resources required are substantial, and it is unclear whether screening reduces mortality when considering lead-time and duration biases."
— American Gastroenterological Association, Gastroenterology 2025 (PMID 40267995)Positive Predictive Value — The Operational Consequence
Positive predictive value (PPV) is the probability that a patient with a positive test result actually has the disease. Unlike sensitivity and specificity — which are properties of the test — PPV depends on the prevalence of disease in the tested population.
Even with a specificity of >99%, in a population where cancer prevalence is approximately 0.5–1% (the range seen in real-world MCED deployment), a meaningful proportion of positive results will be false positives. Published PPV for Galleri ranges from 43–75% across different study populations. In practical terms:
- Every positive Galleri result requires follow-up imaging, often CT or PET
- In many cases, endoscopy or biopsy is also required
- Between 25–57% of positive results will not yield a cancer diagnosis
- Patients experience the psychological burden of a positive result before that determination is made
- Workup carries its own risks (radiation exposure, procedure complications, incidental findings)
For clinical trial settings, this PPV profile has important implications for MCED-based eligibility screening. Trials requiring confirmed cancer for enrolment cannot use MCED positivity alone — tissue confirmation remains necessary. Trials designed to detect early cancers using MCED as a primary endpoint must build in adequate diagnostic confirmation steps and account for the false positive rate in their sample size calculations.
Regulatory Landscape
The FDA has granted Galleri Breakthrough Device Designation, which provides early and intensive FDA guidance during development and priority review for the eventual submission. However, the test is currently marketed in the US as a laboratory-developed test (LDT) — historically regulated by the Centers for Medicare and Medicaid Services (CMS) under CLIA, with less stringent FDA oversight than IVDs.
The regulatory landscape for LDTs has been evolving. The FDA has been working toward requiring LDTs to demonstrate analytical and clinical validity through the standard PMA or 510(k) pathways. For MCED tests to achieve full FDA PMA as IVDs — the standard required before Medicare coverage decisions — sponsors will need to demonstrate clinical utility, not just analytical performance. Clinical utility requires evidence that the test changes clinical management in a way that improves patient outcomes. Mortality benefit is the gold standard.
Current US Status
- Available by physician prescription
- Cost: ~USD 949 out-of-pocket
- Not covered by Medicare or most insurers
- LDT regulatory status (evolving)
- FDA Breakthrough Device Designation
Brazil Status (as of July 2026)
- Expected commercial availability end 2026
- Private clinics only; by physician prescription
- ANVISA registration not yet complete
- Not incorporated in public health system
- Cost-effectiveness review pending
MCED and Oncology Trial Design
As MCED technology matures, it intersects with clinical trial design in several ways that site teams and CRAs need to understand.
Eligibility Criteria and MCED Positivity
Sponsors designing early detection trials may use MCED positivity as a stratification factor or as part of enrichment strategies. However, because MCED positivity does not confirm cancer, protocols must specify whether tissue confirmation is required for enrolment. In interventional trials — particularly those involving toxic therapies — tissue confirmation should always be required. In observational or diagnostic accuracy studies, MCED positivity may itself be the index test being evaluated.
MCED as a Trial Endpoint
Several ongoing trials are using MCED testing as a primary or secondary endpoint — measuring whether an intervention (chemoprevention, lifestyle modification) alters MCED signal detection rates or changes the stage distribution of MCED-detected cancers. For these trials, the sensitivity limitations described above must be incorporated into statistical power calculations. A trial powered around a single specific cancer type detected via tissue biopsy will need significantly different sample size assumptions than one relying on MCED detection.
Minimal Residual Disease and Treatment Monitoring
Methylation-based ctDNA assays are also being applied in post-treatment MRD monitoring — detecting residual cancer signals after surgery or chemotherapy. Unlike mutation-based MRD assays (which require tumour genotyping to identify patient-specific mutations to track), methylation-based assays are tumour-agnostic and can be applied without prior tissue genotyping. This is a distinct use case from population screening, and the performance characteristics differ substantially — sensitivity in MRD monitoring is evaluated differently than in upfront cancer detection.
Site Staff Considerations
For trials incorporating MCED testing, site teams should be aware that:
- Blood draws must be processed into plasma within specified timeframes — cfDNA is degraded by white blood cell lysis if tubes are left at room temperature too long
- Specific cell-free DNA collection tubes (e.g., Streck Cell-Free DNA BCT) are typically required — these inhibit cfDNA degradation and white blood cell lysis
- Samples are typically shipped to central labs (CLIA-certified); local processing is not available
- A positive MCED result in a clinical trial participant requires protocol-specified follow-up procedures that may or may not be within the trial protocol — teams must know what to do
- Counselling support for participants receiving unexpected MCED-positive results is an ethical and operational requirement for any trial using MCED as a screening or monitoring tool
Which Cancer Types Does It Detect Best?
MCED sensitivity is not uniform across cancer types. The test performs best for cancers that shed substantial ctDNA (typically high-grade, high-burden, or biologically aggressive tumours) and worst for cancers that shed minimal ctDNA even at later stages.
| Cancer Category | Typically Higher Sensitivity | Typically Lower Sensitivity |
|---|---|---|
| Haematological | Multiple myeloma, lymphoma (high-burden) | Low-grade MDS, indolent lymphoma |
| Solid Tumour — High Shedding | Hepatocellular, pancreatic, ovarian, lung (SCLC) | — |
| Solid Tumour — Variable | Colorectal (stage III/IV), bladder | Breast (stage I/II), thyroid (well-differentiated) |
| No Established Screen | Galleri may offer first signal for ovarian, pancreatic, hepatocellular | Prostate (PSA overlap), skin (minimal cfDNA shedding) |
The highest-impact potential for MCED is in cancers with no current early detection pathway — ovarian, pancreatic, and hepatocellular cancers are diagnosed predominantly at late stages and have very poor prognoses. Even modest improvements in early detection rates for these tumour types could translate to meaningful survival benefit. The evidence that Galleri specifically delivers this benefit for these cancers at the population level is not yet established, but it represents the strongest theoretical justification for the technology.
Cost-Effectiveness and Access
At approximately USD 949 per test, with annual testing recommended for individuals aged 50 and older, MCED screening at population scale represents a substantial cost. Published cost-effectiveness analyses have found that MCED testing is unlikely to be cost-effective at current prices and sensitivity levels without demonstrated mortality benefit — consistent with standard health economic frameworks (ICER thresholds of USD 50,000–150,000 per quality-adjusted life year).
An ASCO 2026 analysis found that MCED testing becomes cost-effective only under optimistic assumptions about sensitivity improvement, follow-up costs, and mortality reduction. Current evidence does not yet support those assumptions.
For health systems considering coverage decisions — including Brazil's SUS and the US Medicare program — mortality benefit and cost-effectiveness data will be the gatekeeping requirements. Neither has been established as of July 2026.