The product candidate
One blood draw, read for the methylation patterns cancer leaves behind.
Tumours shed fragments of DNA into the bloodstream. Those fragments carry chemical marks — methylation — in patterns that differ from healthy tissue, and that differ again depending on which organ the fragment came from. The EarlySignal test is being designed to read those patterns.
Intended workflow
Five steps, from order to report.
This is the workflow the test is being built around. Each step is a design intention; none has been validated.
Order
A licensed healthcare provider orders the test, either directly or through a telemedicine partner.
01
Collect
A single standard venipuncture blood draw, taken through national phlebotomy networks rather than a dedicated collection footprint.
02
Sequence
Targeted methylation sequencing of informative CpG sites in cell-free DNA, performed in a central high-complexity laboratory.
03
Classify
A machine-learning classifier looks for a coherent cancer signal across those sites and, where one is found, predicts the likely tissue of origin.
04
Report
Results return through a provider portal, with a plain-language summary written for the patient. A signal is a reason to investigate further, not a diagnosis.
05
Design intentions
Where we are trying to be different.
Four choices define the programme. Each is an objective we are working toward — not a result we have achieved or measured.
| Element | Approach | What we are aiming for |
|---|---|---|
| Methylation panel | A broader panel of methylation markers than certain first-generation tests | Better early-stage sensitivity in the cancers with the least screening today |
| Classifier | An architecture intended to capture long-range methylation patterns rather than isolated sites | Stronger detection and a more reliable tissue-of-origin call |
| Sequencing chemistry | A contemplated licensing arrangement with a chemistry provider | A lower cost per test at scale |
| Access | Self-insured employers, health systems and telemedicine ordering first | Reaching people ahead of broad insurance coverage |
Status of these items
The sequencing-chemistry arrangement is contemplated and has not been entered into on a definitive basis. Any figure we use internally for sensitivity, specificity, tissue-of-origin accuracy, turnaround or cost is a design target or is drawn from third-party published work. It is not a result our test has produced.
The honest part
What the field has and has not shown.
Multi-cancer early detection is new, and whether it helps patients live longer is still being argued out in the literature. The largest randomised trial run so far, NHS-Galleri in the United Kingdom, enrolled roughly 142,000 participants. Its full results were presented at the ASCO Annual Meeting in May 2026.
That trial did not meet its primary endpoint: it did not show a statistically significant reduction in combined late-stage incidence across the twelve cancers it was designed around. Some secondary measures were more encouraging — fewer Stage IV diagnoses, and a higher overall detection rate when screening was added to usual care — but the primary result stands.
As things stand, no multi-cancer test has been approved by the FDA, none is covered by Medicare, and none appears in United States practice guidelines. Anyone evaluating this field, including us, has to hold both things at once: a real and well-documented gap in screening, and an unproven answer to it.
Regulatory path
Five steps, none of them taken yet.
This is the sequence we intend to follow, with an accurate status against each step.
| Step | What it involves | Status |
|---|---|---|
| Laboratory-developed test | Initial launch as an LDT performed in a CLIA-certified high-complexity laboratory | Not commenced |
| Breakthrough Device Designation | Application to the FDA for designation of the product candidate | Not obtained |
| Pre-submission | Engaging the FDA on the design of a pivotal clinical study | Not commenced |
| Pivotal study | A large interventional study with extended follow-up, intended to support a premarket approval | Not commenced |
| Premarket approval | FDA review and approval of the test | Not obtained |
On the rules themselves
How laboratory-developed tests are regulated is currently unsettled. The FDA issued a final rule in 2024 that would have regulated LDTs as medical devices; a federal district court vacated that rule in 2025 and the FDA subsequently rescinded it. The position could change again through rulemaking, litigation or legislation.
The laboratory
Buying certification rather than building it.
A central element of the plan is to acquire a laboratory that already holds CLIA certification instead of standing up a high-complexity laboratory from nothing.
Certification, accreditation, state licensure in places like New York and California, trained personnel and inspection history all take time that an early company does not have much of. Acquiring them together is intended to shorten the path to a validated, orderable test.
No laboratory has been identified or agreed. Any acquisition would depend on suitable candidates, definitive agreements, diligence, the transfer or reissue of certification and licences, and sufficient capital.