One Blood Test, 50 Cancers: How AI-Powered Screening Is Changing Early Detection — And Why It’s Not FDA-Approved Yet

Introduction

Imagine a single blood draw that could flag signals for more than 50 types of cancer at once — long before symptoms appear. That’s the promise behind a new generation of multi-cancer early detection (MCED) tests, led by three names increasingly showing up in doctors’ offices and biotech headlines: Galleri, from GRAIL; Cancerguard, from Abbott (via its 2026 acquisition of Exact Sciences); and SPOT-MAS 10, from the Vietnam-based biotech Gene Solutions.

These tests use machine learning to scan a blood sample for faint traces of tumor DNA, then try to pinpoint where in the body a cancer signal might be coming from. It’s a genuinely different approach to screening — one blood draw instead of a dozen separate exams.

But there’s a fact that rarely makes the headlines: none of these three tests currently has FDA clearance or approval. All three are offered as Laboratory Developed Tests (LDTs), a regulatory category that allows clinical labs to offer tests without full FDA marketing authorization. That distinction matters enormously for how patients, doctors, and health systems should think about this technology.

This article walks through what these tests actually do, what the clinical evidence shows, where the real benefits lie, and where the genuine limitations are — without the hype.

What Is Multi-Cancer Early Detection Test Technology?

MCED tests are blood-based liquid biopsies designed to detect a shared “cancer signal” across many tumor types from a single sample, rather than requiring one dedicated test per cancer type.

Here’s how the three leading tests compare on the basics:

A few things stand out immediately. Galleri and Cancerguard both target a broad panel of 50-plus cancers. SPOT-MAS 10 is narrower by design, focused on 10 cancer types including breast, lung, liver, colorectal, gastric, ovarian, pancreatic, esophageal, endometrial, and head-and-neck cancers. That’s not a weakness — it’s simply a different design choice, and conflating the two scopes would misrepresent what each product actually claims to do.

How Does It Work?

The underlying science is called liquid biopsy. When tumors grow, they shed tiny fragments of their own DNA into the bloodstream — known as circulating cell-free DNA, or cfDNA. Healthy cells shed cfDNA too, so the challenge is telling the two apart in a sea of genetic noise.

This is where the AI comes in — and it does two distinct jobs:

1. Signal detection. Machine learning models are trained to recognize patterns in cfDNA that are more common in cancer cells than in normal ones. Cancer DNA tends to have distinctive methylation patterns (chemical tags that switch genes on or off) and different fragmentation patterns (the way DNA breaks apart). The AI model scans for these tell-tale signatures.

• Galleri leans primarily on methylation pattern analysis.
• Cancerguard combines ctDNA analysis with additional protein biomarkers.
• SPOT-MAS 10 integrates genetic, epigenetic, and fragmentomic signals together.

2. Tissue-of-origin prediction. If a cancer signal is detected, a second AI model tries to predict which organ or tissue it likely came from — steering the follow-up workup toward the right imaging or diagnostic test, rather than leaving doctors guessing.

Think of it like sifting through a massive pile of sand for a handful of specific-colored grains, then trying to guess which beach they came from. That’s roughly what these algorithms are doing, at a molecular scale, across billions of DNA fragments per sample.

What Healthcare Problem Does It Solve?

Cancer screening today is fragmented. The U.S. Preventive Services Task Force (USPSTF) currently recommends routine screening for only four cancers — breast, cervical, colorectal, and lung (the latter only in high-risk individuals). Prostate screening exists in a more individualized, shared-decision framework.

That leaves a major gap. Independent research — including a peer-reviewed analysis co-authored by GRAIL scientists — has found that roughly 70% of U.S. cancer deaths occur in cancer types that currently have no recommended screening test at all: ovarian, pancreatic, liver, and many others that are often caught only after symptoms appear, at a later and less treatable stage.

MCED tests are positioned to address exactly this gap — not by replacing mammograms or colonoscopies, but by adding a layer of detection for cancers that otherwise go unscreened until it’s too late.

Clinical Evidence

This is where the details matter most, and where evidence quality varies meaningfully between the three tests.

SPOT-MAS 10 — the strongest peer-reviewed evidence base

The K-DETEK study, published in the peer-reviewed journal BMC Medicine, followed 9,024 asymptomatic participants in Vietnam. It reported:

• Sensitivity: ~78.1%
• Specificity: ~99.8%
• Tissue-of-origin accuracy: ~84%

In that study, only 43 participants tested positive; of those, 17 were confirmed to have cancer — and 12 of those 17 were caught early enough for curative treatment. A separate real-world cohort of 12,281 people across Southeast Asia reported similar numbers (78.2% sensitivity, 99.8% specificity).

Galleri — large-scale, company-reported and peer-reviewed data

The original PATHFINDER study (part peer-reviewed, part company-reported) found:

• Positive predictive value (PPV): 43%
• Specificity: 99.5%
• Cancer signal origin accuracy: 88%

A larger follow-up, PATHFINDER 2 (25,490 participants), reported specificity of 99.6% and episode sensitivity of 73.7% for the 12 cancers responsible for two-thirds of U.S. cancer deaths — though sensitivity across all cancer types was notably lower, at 40.4%. These PATHFINDER 2 topline figures come from a GRAIL press release; full peer-reviewed publication was pending at the time of this reporting.

Cancerguard — SEC-filed clinical data

Per Exact Sciences’ own regulatory filings, the ASCEND-2 study reported 60% overall sensitivity at 98.5% specificity when excluding site-specific cancer organ analysis.

The evidence gap that matters most

No MCED test — Galleri, Cancerguard, or SPOT-MAS — has yet demonstrated a reduction in cancer-specific mortality in a completed, published randomized controlled trial. The NHS-Galleri trial, enrolling 140,000 participants in the UK, is the largest such study to date and has completed its first screening round, but final mortality outcomes have not yet been reported. This is the evidence that will ultimately determine whether MCED testing measurably saves lives — and it doesn’t exist yet.

Benefits

• Broader coverage. A single blood draw can screen for cancer types that currently have no standard screening pathway at all.
• High specificity. All three tests report specificity above 99% in their study populations, meaning false positives were relatively uncommon in those cohorts.
• Earlier detection in real cases. In SPOT-MAS’s K-DETEK study, the majority of cancer-confirmed participants were caught early enough for curative intervention.
• Convenience. One blood draw versus multiple separate screening procedures across different specialists and imaging centers.
• Tissue-of-origin guidance. Rather than leaving a positive result as a vague alarm, the AI attempts to direct follow-up testing toward the most likely source.

Challenges

• No FDA clearance or approval. All three tests are currently Laboratory Developed Tests under CLIA regulation — not FDA-approved medical devices. Galleri has a Breakthrough Device Designation (since 2018) and a pending Premarket Approval application; SPOT-MAS 10 received Breakthrough Device Designation in May 2026. Neither designation guarantees eventual approval.
• No USPSTF recommendation, no national Medicare coverage determination. This keeps out-of-pocket costs high ($689–$949) and limits insurance reimbursement.
• Lower sensitivity for early-stage disease. Detection rates are generally higher for later-stage cancers than for the earliest, most treatable stages — precisely where screening matters most.
• A positive result is not a diagnosis. It triggers further imaging and diagnostic workups, which carry their own costs, anxiety, and risk of unnecessary procedures if the signal turns out to be a false positive.
• Inconsistent published figures. Different studies of the same test (analytical validation vs. prospective trial vs. real-world cohort) report different sensitivity numbers — anywhere from roughly 70% to 78% for SPOT-MAS alone, depending on the specific study cited.
• Overdiagnosis risk. Independent clinical policy literature flags overdiagnosis and overtreatment as a recognized theoretical concern with population-level cancer screening generally, MCED tests included.

The Future of AI in Cancer Screening

Several developments over the next one to three years will determine whether MCED testing becomes a standard part of preventive care:

• Galleri’s FDA decision. GRAIL expects to complete its modular Premarket Approval submission in the first half of 2026. A final FDA decision — approval, rejection, or request for more data — would be a pivotal moment for the entire category.
• NHS-Galleri mortality data. As the largest randomized controlled trial in this space, its eventual results on whether MCED screening actually reduces cancer deaths will be the evidence the field has been waiting for.
• Expanding AI sophistication. The tissue-of-origin prediction step — arguably where AI adds the most distinct value beyond simple signal detection — is likely to keep improving as training datasets grow.
• Regulatory clarity as the real bottleneck. Even strong clinical data won’t translate into widespread, insurance-covered adoption without a USPSTF recommendation or a national coverage determination — a process that has historically taken years for other screening technologies.

Key Takeaways

• Multi-cancer early detection tests use AI to scan blood for cancer-linked DNA signals and predict where in the body they originated.
• Galleri and Cancerguard cover 50+ cancer types; SPOT-MAS 10 covers 10 cancer types and ~75 subtypes — these are not directly equivalent in scope.
• SPOT-MAS 10’s K-DETEK study offers the strongest peer-reviewed evidence base among the three; Galleri and Cancerguard data are a mix of peer-reviewed and company-reported figures.
• None of the three tests is FDA-cleared or approved as of this writing; all are offered as Laboratory Developed Tests.
• No MCED test has yet shown it reduces cancer mortality in a completed randomized controlled trial.
• These tests are designed to complement, not replace, existing guideline-recommended cancer screening.

Conclusion

Multi-cancer early detection tests represent a genuinely interesting application of AI to a real and well-documented gap in cancer screening — the roughly 70% of cancer deaths tied to cancers with no routine screening test today. The technology is real, the science behind it is published and improving, and early real-world results are promising.

But this is a technology still in transition, not a finished product. Regulatory review is ongoing, mortality-outcome data doesn’t exist yet, and marketing claims sometimes outpace what the published evidence actually supports. Anyone considering one of these tests should talk to a healthcare provider first, understand exactly what a positive or negative result does and doesn’t mean, and treat the results as one input alongside — not instead of — standard screening.

The next 12 to 36 months, particularly Galleri’s FDA decision and the NHS-Galleri trial’s mortality data, will likely determine whether multi-cancer blood testing becomes a mainstream part of preventive medicine or remains a promising but unproven niche.

Frequently Asked Questions

1. What is a multi-cancer early detection (MCED) test? It’s a blood test that uses AI and DNA analysis to look for signals shared across many different cancer types from a single sample, rather than testing for one cancer at a time.

2. Is the Galleri test FDA approved? No. As of this writing, Galleri is offered as a Laboratory Developed Test. GRAIL holds an FDA Breakthrough Device Designation and has a Premarket Approval application in process, but full approval has not been granted.

3. Is the Cancerguard test FDA approved? No. Cancerguard is also a Laboratory Developed Test and has not been cleared or approved by the FDA, according to Exact Sciences’ own disclosures.

4. What is SPOT-MAS 10 and is it approved in the US? SPOT-MAS 10 is a multi-cancer blood test developed by Gene Solutions, primarily used in Asia. It received FDA Breakthrough Device Designation in May 2026, which is not the same as approval.

5. How many cancers can Galleri detect? Galleri is marketed to detect a shared cancer signal across more than 50 cancer types.

6. How many cancers can SPOT-MAS 10 detect? SPOT-MAS 10 targets 10 cancer types and approximately 75 subtypes, including breast, lung, liver, colorectal, gastric, ovarian, pancreatic, esophageal, endometrial, and head-and-neck cancers.

7. How accurate is the Galleri test? In the PATHFINDER study, Galleri showed a positive predictive value of 43% and specificity of 99.5%. In the larger PATHFINDER 2 study, specificity was 99.6%, with 73.7% sensitivity for the 12 cancers responsible for most U.S. cancer deaths.

8. How accurate is SPOT-MAS 10? Its peer-reviewed K-DETEK study reported approximately 78.1% sensitivity and 99.8% specificity in a cohort of over 9,000 people.

9. How accurate is Cancerguard? Per Exact Sciences’ SEC filings, the ASCEND-2 study showed 60% overall sensitivity at 98.5% specificity.

10. Do these tests replace regular cancer screening like mammograms or colonoscopies? No. All three companies explicitly state their tests are meant to complement, not replace, standard guideline-recommended cancer screening.

11. How much does the Galleri test cost? Approximately $949, and it is generally not covered by insurance.

12. How much does the Cancerguard test cost? Approximately $689.

13. Does insurance cover multi-cancer early detection tests? Generally, no. There is currently no national Medicare coverage determination for MCED testing, and coverage decisions are left to individual insurers or local Medicare carriers.

14. What happens if I get a positive result? A positive result is not a diagnosis. It typically leads to follow-up diagnostic imaging, such as a CT or PET-CT scan, and further clinical evaluation to confirm whether cancer is actually present.

15. Can these tests give false positives? Yes. While specificity is high (99%+ in study populations), false positives can occur and typically lead to additional testing and associated anxiety or cost.

16. Can these tests miss cancer (false negatives)? Yes. Sensitivity, particularly for early-stage cancers, is meaningfully lower than for later-stage disease, meaning some cancers — especially small, early ones — may not be detected.

17. What technology powers these tests? They analyze circulating cell-free DNA (cfDNA) in blood, using machine learning to detect cancer-associated methylation and fragmentation patterns, then predict the likely tissue of origin.

18. Has any MCED test been shown to reduce cancer deaths? Not yet. No MCED test has demonstrated a reduction in cancer-specific mortality in a completed, published randomized controlled trial. The ongoing NHS-Galleri trial is expected to eventually provide this data.

19. Who should consider taking an MCED test? This is a decision to make with a healthcare provider, who can explain what the test can and cannot tell you given your personal risk factors and the current state of the evidence.

20. What is the NHS-Galleri trial? It’s a randomized controlled trial in the UK enrolling 140,000 participants — the largest study of its kind for a multi-cancer early detection test. It has completed its first screening round but has not yet reported final outcomes.

21. What role does AI specifically play in these tests, versus the lab science? AI is used for two main tasks: classifying whether a cancer signal is present in the DNA data, and predicting which tissue or organ that signal likely originated from — both of which are the pattern-recognition steps that would be extremely difficult to perform manually across such large, complex datasets.

22. Are these tests available worldwide? Availability varies. Galleri and Cancerguard are primarily available in the United States. SPOT-MAS 10 is used mainly in Vietnam and other parts of Southeast Asia, with expansion into additional markets such as Türkiye.

References

• GRAIL, Inc. — Official press releases and “Our History” (grail.com)
• GRAIL — PATHFINDER and PATHFINDER 2 study results (grail.com/press-releases)
• Exact Sciences Corporation — SEC Form 10-K filing (materials.proxyvote.com)
• Abbott — Press release on completed acquisition of Exact Sciences (abbott.mediaroom.com)
• Cancerguard official patient site — Regulatory disclaimer (cancerguard.com)
• Gene Solutions — Press releases on SPOT-MAS clinical and real-world data (PRNewswire APAC)
• K-DETEK study — Published in BMC Medicine (via GenomeWeb reporting and medRxiv preprint)
• American Cancer Society — “Multi-cancer Detection (MCD) Tests” (cancer.org)
• Cancer Therapy Advisor — “MCED Testing for Cancer Diagnosis: Progress and Roadblocks”
• Blue Shield of California — Medical policy document on Multicancer Early Detection Testing
• USC Schaeffer Center — “Cancer-Related Technologies Have Changed a Lot. So Should Cancer Screening.”
• American Society of Clinical Oncology (ASCO) Educational Book — “Multicancer Early Detection Tests at a Crossroads”
• npj Precision Oncology — Peer-reviewed study on MCED tests detecting cancers lacking USPSTF-recommended screening

“Editorial note: This article reflects publicly available information as of August 2026. Regulatory status, clinical trial results, and product availability for these tests may change. Readers should consult a licensed healthcare provider before making decisions about cancer screening.“