Working Physician Notes Evidence-Based Published 2026-05-15

Multi-cancer early detection blood tests — a data-driven guide

A new class of cancer screening tests — multi-cancer early detection (MCED) blood tests like GRAIL's Galleri — claims to detect more than 50 cancer types from a single blood draw. The technology is real, the published data is more interesting than the marketing suggests, and the question of whether a particular patient should take one is more nuanced than yes-or-no. This article walks through what the tests actually do, what the evidence shows, who benefits, and how to integrate them into a sensible cancer-screening plan.

Book a Free 15-Min Call Reserve My Spot

What MCED tests actually measure

In plain English: cancer cells leave traces in your blood. MCED tests look for those traces.

More precisely: cancer cells shed small fragments of DNA into the bloodstream as they grow and die. This cell-free DNA (cfDNA) is detectable in any blood sample. What distinguishes cancer-derived cfDNA from normal cfDNA is its methylation pattern — the chemical modifications on the DNA that vary by cell type and by the abnormal regulatory programs running in cancer cells.

The current generation of multi-cancer early detection tests — Galleri being the most prominent, with others like Cancerguard (Exact Sciences) and IvyGene entering the field — uses high-throughput sequencing of cfDNA combined with machine-learning analysis of methylation patterns to do two things:

  1. Detect whether a cancer signal is present in the blood sample.
  2. Localize the probable tissue of origin (which organ or tissue type the cancer is likely coming from), to direct downstream diagnostic workup.

Galleri specifically reports detection across more than 50 cancer types, with a cancer signal origin (CSO) prediction designed to guide the workup if a signal is detected. Other MCED tests use different combinations of methylation, fragmentomics, and protein markers, with somewhat different performance characteristics.

The technology is not FDA-approved. Galleri operates under CLIA as a laboratory-developed test (LDT). Recent FDA actions have moved toward more direct oversight of LDTs, but as of May 2026, MCED tests remain in this regulatory category. They are not covered by Medicare or most commercial insurance.

What the published evidence shows

Three studies drive most of the published evidence on Galleri specifically; broader trials are underway:

The CCGA studies (Circulating Cell-free Genome Atlas) — case-control studies in cancer patients vs. matched non-cancer controls. The 2021 CCGA Sub-Study 3 reported overall test specificity of 99.5 percent (false positive rate ~0.5 percent) and overall sensitivity of 51.5 percent across all stages. Sensitivity varied substantially by stage:

  • Stage I: 16.8 percent
  • Stage II: 40.4 percent
  • Stage III: 77.0 percent
  • Stage IV: 90.1 percent

Cancer signal origin localization accuracy was approximately 88 percent (the test correctly identified the tissue of origin in cases where a signal was detected).

PATHFINDER (published in Lancet 2023) — a prospective interventional study of approximately 6,600 adults aged 50 and older with no known active cancer. Test results were returned to participants and clinicians; downstream workup was tracked.

  • A cancer signal was detected in 1.4 percent (92 of 6,621) participants.
  • Among those with a signal, positive predictive value (PPV) was 38 percent — meaning 38 percent of "cancer signal detected" results led to a confirmed cancer diagnosis.
  • 71 percent of the cancers detected were not amenable to conventional screening (i.e., cancer types or sites that don't have established screening programs like mammography, colonoscopy, low-dose lung CT).
  • Median time from positive result to diagnostic resolution was 79 days.

NHS-Galleri — the first randomized, controlled trial of MCED screening, enrolling approximately 142,000 NHS participants aged 50 to 77 across three annual screening rounds in England. Headline results were announced February 19, 2026, with detailed analyses scheduled for presentation at the ASCO 2026 Annual Meeting in June. The findings were mixed and clinically interesting:

  • The primary endpoint — a statistically significant reduction in Stage III-IV cancer diagnoses across all cancers — was not met. This is the headline number that matters for regulatory adoption and payer coverage decisions, and missing it is genuinely significant.
  • In the prespecified subgroup of 12 deadly cancers, Stage IV diagnoses fell by more than 20 percent in years 2 and 3 of sequential screening — clinically meaningful for cancers with the highest baseline mortality.
  • Annual Galleri screening combined with standard screening produced approximately a four-fold higher cancer detection rate (for breast, colorectal, cervical, and high-risk lung cancers) compared with standard screening alone.
  • A substantial reduction in cancers detected clinically through emergency presentation — which is associated with significantly higher mortality and treatment cost than screening-detected cancer.

The right way to read the NHS-Galleri data: the test is not ready to replace existing population screening, and the field cannot yet claim it reduces all-cancer mortality. At the same time, the secondary findings — particularly the Stage IV reduction in the deadliest cancers and the four-fold improvement in finding cancers that screening would have caught — are real signals of clinical utility for specific patient situations.

SYMPLIFY — a study of Galleri in patients with symptoms suggestive of cancer (rather than asymptomatic screening). PPV was substantially higher (~75 percent) in this population because pre-test probability of cancer was higher.

The interpretation gap — what these numbers actually mean

The MCED data has two faces. The face the marketing shows is "detects 50+ cancer types from a blood draw." The face the published data shows is more careful.

For early-stage cancer, sensitivity is modest. A 16.8 percent stage I sensitivity means the test misses roughly five out of six stage I cancers. The headline "51 percent overall sensitivity" is dominated by easier-to-detect later-stage disease. If the goal is "catch cancer at the earliest curable stage," the test as currently calibrated catches one in six at that stage.

Specificity is high but absolute false positives are not zero. A 99.5 percent specificity means roughly 1 in 200 healthy people who take the test will get a "cancer signal detected" result that isn't actually cancer. In a 50-year-old asymptomatic adult with a 0.5 to 1 percent pre-test probability of cancer, a positive result is more likely to be a false positive than a true positive — basic Bayesian math.

The PPV problem is real. A 38 percent PPV in PATHFINDER means 62 percent of people who got a "cancer signal detected" result did not have cancer. Each of those false positives triggered a workup — often imaging, sometimes biopsy, sometimes endoscopy — with the associated cost, risk, and patient anxiety.

The mortality question is partially answered, but incompletely. The NHS-Galleri trial (results announced February 2026) did not meet its primary endpoint — a statistically significant overall reduction in Stage III-IV cancer diagnoses. Mortality benefit, the ultimate clinical question, has not been demonstrated in any randomized trial. At the same time, NHS-Galleri did show meaningful reductions in Stage IV diagnoses for the most lethal cancers and substantial increases in early-stage detection for several screenable cancers. The data supports a "specific use cases, not population screening" interpretation rather than either "doesn't work" or "ready for prime time."

None of this argues against the technology. It argues for clear expectations about what a single MCED test does and doesn't do.

Where MCED genuinely changes the math

The limitations are real. But the case for utility, in the right clinical situation, isn't theoretical. There are specific places where MCED tests genuinely change outcomes — and they're worth naming directly.

1. Cancers without screening programs

The single most important finding from PATHFINDER, stated again with emphasis: 71 percent of the cancers MCED detected had no standard screening alternative. This is not a niche use case. It's the entire point of the technology.

Asymptomatic adults currently have no recommended screening for:

  • Pancreatic cancer
  • Ovarian cancer
  • Esophageal cancer
  • Hepatobiliary cancers (liver, gallbladder, bile duct)
  • Gastric cancer
  • Most kidney, small bowel, and head and neck cancers

These cancers are disproportionately responsible for cancer mortality precisely because they're diagnosed too late. Pancreatic cancer has a five-year survival of approximately 12 percent, and roughly 80 percent of diagnoses occur at stage IV. Ovarian cancer follows the same pattern — about 70 percent of diagnoses present at stages III or IV. These cancers aren't untreatable in early stages; they're typically detected after early stages have passed.

For these specific cancers, "any detection before the typical incidental diagnosis" is real clinical value. MCED's absolute sensitivity for stage I pancreatic cancer is modest, but it's infinitely better than zero — which is what asymptomatic screening currently offers for this disease.

2. Stage shifting saves lives, even imperfectly

Cancer mortality is heavily stage-dependent. Moving some fraction of stage IV diagnoses to stage III, or stage III to stage II, has meaningful population impact even if early-stage sensitivity is modest.

A useful comparison: low-dose chest CT screening for current and former heavy smokers reduces lung cancer mortality by 20 to 25 percent — and it does so by shifting the stage distribution at diagnosis downward, not by catching every cancer at stage I. The mortality benefit doesn't require perfect detection; it requires moving the stage curve.

The case for MCED follows the same logic. The technology doesn't need to be perfect to matter. It needs to move some patients from a worse stage at diagnosis to a better one. For the cancer types listed above — where the current alternative is "diagnosed at advanced stage symptomatically" — even a partial shift is clinically meaningful.

3. The diagnostic-odyssey shortcut

The SYMPLIFY study evaluated Galleri in patients with nonspecific symptoms suggestive of cancer — unexplained weight loss, fatigue, anemia, lymphadenopathy — rather than in asymptomatic screening. In that population, the positive predictive value rose to approximately 75 percent, and the test's tissue-of-origin prediction substantially compressed the workup.

This isn't a screening use case. It's a diagnostic acceleration use case. The current pathway for a patient presenting with 20 pounds of unexplained weight loss is months of cycling through GI workup, pulmonary workup, endocrine workup, oncology consult — often three to five specialists before a primary site is identified. An MCED test with tissue-of-origin localization can collapse that timeline into days, directing the workup to the correct organ system at the start rather than after a long process of elimination.

For patients experiencing the diagnostic odyssey, this use of MCED can be transformative even when the absolute test performance is moderate.

4. Lung cancer in never-smokers

Current screening for lung cancer — low-dose chest CT — is indicated only for current or former heavy smokers. PATHFINDER and follow-on studies have detected lung cancers in never-smokers who would not have been eligible for LDCT screening at all.

Rates of lung cancer in never-smokers have been rising for two decades and now account for approximately 20 percent of new lung cancer diagnoses, with an unclear set of contributors (air pollution, second-hand smoke, occupational exposures, possibly radon). This is a real and growing coverage gap in standard cancer screening, and MCED is currently one of the few tools that addresses it.

5. The genetically-elevated risk patient

Patients with known cancer-predisposing genetic mutations have higher pre-test probability of cancer, which substantially improves the Bayesian calculation. A BRCA1 or BRCA2 carrier with elevated pancreatic and breast cancer risk, a Lynch syndrome patient with elevated risk across multiple GI cancers, a TP53 carrier (Li-Fraumeni) — these patients have lifetime cancer risk substantially above the general population, and MCED's positive predictive value in this population is correspondingly higher.

For high-risk patients who have exhausted their available targeted screening options, MCED adds a layer of surveillance that wasn't previously available.

The technology isn't ready to replace established screening, and it shouldn't try to. But framing MCED as "interesting but not yet useful" misreads the data. For the specific clinical situations above — and the right patient profile in each — MCED can produce clinical benefit that isn't otherwise achievable. The clinical question isn't whether MCED is good or bad in the abstract. It's whether the individual patient sits in a situation where the technology adds real value.

Who benefits most

Across the published data and current expert consensus (American Cancer Society MCED Working Group, NCCN early commentary), the population most likely to benefit from MCED screening is:

  • Adults aged 50 to 79 with elevated baseline cancer risk relative to age-matched peers.
  • Patients who are already current on standard age-appropriate screening (colonoscopy, mammography, low-dose lung CT for smokers, cervical screening, PSA discussion where applicable) and looking for supplemental detection of cancers that don't have screening programs.
  • Patients with a strong family history of cancer — particularly cancers that lack standard screening (pancreatic, ovarian, esophageal, hepatobiliary).
  • Patients with elevated personal risk — known cancer-predisposing genetic mutations, prior cancer in remission, occupational exposures, etc.

The case is weakest — and the false-positive harm potentially largest — in:

  • Adults under 50 at average risk. Pre-test probability of cancer is low, making the math unfavorable for population-level screening.
  • Adults who are not current on standard screening. Spending money on Galleri while skipping a colonoscopy that the USPSTF recommends and that has decades of mortality benefit data is the wrong order of operations.
  • Adults whose health anxiety would not tolerate the false-positive workup. A "cancer signal detected" result that turns out to be nothing is psychologically demanding even when the workup eventually resolves benignly.

How MCED fits with standard cancer screening

The right framing is that MCED tests supplement standard cancer screening rather than replacing it. Standard screening programs each have decades of mortality benefit data:

  • Colonoscopy reduces colorectal cancer mortality by approximately 60-70 percent.
  • Mammography reduces breast cancer mortality by approximately 20-30 percent in screened women aged 50 and older.
  • Low-dose chest CT in current/former heavy smokers reduces lung cancer mortality by approximately 20-25 percent.
  • Cervical screening (Pap and HPV) has reduced cervical cancer mortality dramatically — by more than 80 percent since widespread adoption.
  • Skin checks for high-risk patients.
  • PSA remains a discussion-based decision rather than a universal recommendation.

The right way to think about Galleri or other MCED tests is as coverage for cancers that don't have established screening programs — pancreatic, ovarian, esophageal, hepatobiliary, gastric, and others where there is currently no standard screening for asymptomatic adults. Detecting these cancers earlier is plausibly valuable; the MCED data suggests it's possible for a meaningful subset of these cancers (later-stage detection is much better than early-stage detection per the data, but late-stage detection of pancreatic cancer at presentation is still earlier than the typical incidental diagnosis).

Skipping colonoscopy because Galleri is "good enough" is a clinical mistake. Galleri has roughly 40 to 50 percent sensitivity for colorectal cancer overall, far below colonoscopy. The two tests address different things and shouldn't be substituted.

The false-positive workup — what actually happens

Patients considering an MCED test should understand what a positive result triggers, because the workup is non-trivial.

A "cancer signal detected" result with a cancer signal origin (CSO) prediction directs further diagnostic workup. Depending on the CSO, this can include:

  • Cross-sectional imaging (CT chest/abdomen/pelvis, MRI)
  • Endoscopy (upper GI, colonoscopy)
  • PET scan
  • Tissue biopsy
  • Specialist referrals (typically oncology and the appropriate organ specialist)
  • Repeat liquid biopsy or tumor marker testing

For a true positive, this workup leads to diagnosis and treatment, which is the entire point. For a false positive, the patient goes through some or all of this workup before reaching the conclusion that no cancer is present. The cost can run $3,000 to $15,000 even when nothing is found. The psychological burden of 79 days of workup uncertainty (the PATHFINDER median) is real.

This is not an argument against the test. It's an argument for taking it with eyes open. A patient who would be unable to manage the false-positive scenario is probably not the right patient for the test.

Cost and access

Galleri direct-pay through GRAIL is currently $949 (May 2026). HSA and FSA funds typically qualify. Most commercial insurance does not cover the test; Medicare does not cover it for asymptomatic adults under current policy. CalPERS and a small number of self-funded employer plans have begun covering MCED screening for eligible employees.

Other MCED tests have varying pricing and availability. Cancerguard (Exact Sciences) was at the time of writing in earlier-stage development. The competitive landscape will shift as additional MCED tests reach the market.

The cost-effectiveness analysis depends entirely on what the mortality benefit turns out to be — which remains an open question pending NHS-Galleri and other ongoing outcome studies. Treating MCED as a supplemental screening option for appropriate patients at this cost is defensible; treating it as a substitute for established screening at this cost is not.

How we approach MCED testing

Our approach for patients in the practice who are considering an MCED test:

  1. Confirm standard screening is current. If colonoscopy, mammography, lung CT, cervical screening, and skin checks are not up to date, those are the higher-yield interventions to address first.
  2. Honest conversation about risk profile. Family history, personal risk factors, prior cancer history, occupational exposures. The benefit-to-harm ratio of MCED varies substantially by individual risk profile.
  3. Honest conversation about the false-positive scenario. What the patient would experience if the test comes back positive but workup reveals no cancer. Some patients are appropriate candidates; some aren't.
  4. If proceeding, the test is ordered through the patient-direct-pay vendor pathway — typically GRAIL directly, with results returning to the practice for integration and interpretation.
  5. If the result is "cancer signal detected," we coordinate the workup. This is exactly the type of clinical situation where having an established physician relationship matters — the workup is fast, organized, and integrated with the rest of your care.
  6. If the result is negative, we discuss interval testing. A negative MCED test doesn't replace next year's screening; it provides additional information that the patient should reconsider periodically as risk factors evolve.

MCED testing is included as a patient-elected, patient-pay option in our 360° Deepscan tier. It's available as an optional add-on for Concierge Internal Medicine members at the same patient-direct-pay vendor cost. Private MD does not mark up the test or receive compensation from GRAIL or any other MCED vendor; the test is ordered and billed directly between you and the vendor.

The bottom line

Multi-cancer early detection blood tests are a real and rapidly evolving technology. The published data shows they detect cancer in some asymptomatic patients earlier than they would otherwise have been detected, particularly for cancers without established screening programs. The data also shows substantial limitations: low early-stage sensitivity, real false-positive rates, no proven mortality benefit yet, and substantial downstream workup burden for both true and false positives.

For the right patient — 50 to 79, current on standard screening, elevated baseline risk, willing to navigate a possible false-positive workup — the test is a reasonable supplemental option. For the wrong patient, it's expensive anxiety with a low expected yield.

The clinical judgment of who's the right patient is the part that should never be outsourced to a website. The technology is impressive; the indication is individual.

Ready to talk?

Book a free 15-minute Discovery Call. We'll walk through your situation, talk about whether MCED testing makes sense for you given your risk profile and screening status, and figure out together whether this practice fits what you're looking for.

Book a Free 15-Min Call   Reserve My Spot

Or email drsahni@privatemd.clinic · (279) 258-5567

About the author

Nishant Sahni, MD is board-certified in Internal Medicine and the founder of Private MD, a small-panel direct primary care and concierge longevity practice serving Granite Bay, Folsom, El Dorado Hills, and the greater Sacramento area. Former faculty at Mayo Clinic and the University of Minnesota.

Sources: Klein EA et al., Circulating Cell-free Genome Atlas (CCGA) Sub-Study 3, Annals of Oncology 2021; Schrag D et al., PATHFINDER prospective interventional study of MCED, Lancet 2023; Nicholson BD et al., SYMPLIFY observational study in symptomatic patients, Lancet Oncology 2023; NHS-Galleri trial headline results announced February 19, 2026 (GRAIL Inc. press release and NHS-Galleri trial updates); detailed NHS-Galleri analyses scheduled for ASCO 2026 Annual Meeting; United States Preventive Services Task Force screening recommendations; American Cancer Society MCED Working Group commentary; published GRAIL Galleri product information including pricing as of May 2026. Cited figures reflect publicly available data as of May 2026. This article is for educational purposes and is not a substitute for individualized medical advice or screening recommendations.

Found this useful? Share it.

LinkedIn X / Twitter Email

Legal & Editorial Notes

Educational content only. This article is for educational and informational purposes. It does not constitute medical advice, diagnosis, or treatment recommendations for any specific individual. Reading this article does not establish a physician-patient relationship with Dr. Sahni or Private MD.

Not a substitute for individual evaluation. Medical decisions should be made in consultation with your own physician based on your specific clinical situation, history, exam findings, and goals. Information in this article may not apply to your circumstances. If you have specific symptoms or health concerns, consult a qualified medical professional.

Cited data and pricing. Studies, statistics, and prices cited reflect publicly available information as of the date of publication (2026-05-15). Information changes over time. Verify current data with primary sources or your physician before making decisions.

No financial relationships, no endorsement. Mention of specific products, companies, services, or commercial entities is for illustrative comparison only and does not constitute endorsement. Private MD receives no compensation, commission, kickback, or referral fee from any vendor, manufacturer, laboratory, or commercial entity mentioned in this article.

Test regulatory status. Multi-cancer early detection tests, including Galleri (GRAIL) and similar products, are currently laboratory-developed tests (LDTs) operating under CLIA. They are not FDA-approved at the date of publication, and the regulatory landscape continues to evolve. Test performance characteristics may vary by patient population and may change as the technology develops. Coverage by Medicare and commercial insurance is limited; check current coverage before ordering.