The Promise and Peril of One Blood Draw to Rule Them All: Multi-Cancer Detection Tests Face Their Reckoning

A simple blood draw that screens for fifty cancers at once sounds like the kind of medical advance that ends arguments. For the roughly 2 million Americans diagnosed with cancer each year — many of them caught too late for curative treatment — the appeal is obvious. But the science behind multi-cancer early detection (MCED) tests is now colliding with clinical reality, and the results are more complicated than any press release suggests.

In the past six months alone, the largest randomized trial of an MCED test failed to hit its primary endpoint [1], a new $20 blood test from UCLA promised to detect cancers and liver diseases from a single sample [2], and GRAIL submitted a formal FDA approval application for its Galleri test [3]. Meanwhile, Exact Sciences launched CancerGuard, a competing product, without waiting for FDA clearance [4]. The field is moving fast. Whether it is moving in the right direction depends on whom you ask.

What These Tests Claim to Do

MCED tests analyze cell-free DNA (cfDNA) — tiny fragments of genetic material shed into the bloodstream when cells die. By reading chemical modifications called methylation patterns on that DNA, the tests attempt to identify the tissue of origin and whether a cancer signal is present.

GRAIL's Galleri test, the most commercially advanced product, claims to screen for more than 50 cancer types from a single blood draw [5]. Exact Sciences' CancerGuard, launched in September 2025, covers a similar range while excluding breast and prostate cancers [4]. UCLA's MethylScan, published in the Proceedings of the National Academy of Sciences in April 2026, detected liver, lung, ovarian, and stomach cancers as well as several non-cancer liver diseases [2]. And Peking University's cf-EpiTracing platform, published in Nature in March 2026, demonstrated disease detection from as little as 50 microliters of plasma — roughly one drop of blood [6].

The numbers that matter most are sensitivity (how many true cancers the test catches) and specificity (how many healthy people it correctly identifies as cancer-free). These figures vary sharply across tests, cancer types, and disease stages.

The Numbers: What the Data Actually Shows

The PATHFINDER 2 study — the largest prospective MCED trial to date, with 35,878 participants across the United States and Canada — reported that adding Galleri to standard recommended screenings increased the cancer detection rate more than seven-fold [7]. Galleri's specificity stood at 99.6%, and its cancer signal origin prediction was accurate 93.4% of the time in confirmed cases [5].

But sensitivity told a different story. Across all cancers, Galleri detected 40.4% of cases. For the 12 cancers responsible for two-thirds of U.S. cancer deaths, sensitivity reached 73.7% [7]. The gradient by stage was steep: 16.8% for Stage I, 40.4% for Stage II, 77.0% for Stage III, and 90.1% for Stage IV [7].

Source: GRAIL PATHFINDER 2 Trial Results

Data as of Dec 26, 2025CSV

That stage gradient is the central tension of the field. The tests are best at finding cancers that are already advanced — precisely the cases where early detection matters least. For the earliest-stage cancers, where intervention could most change outcomes, detection rates remain low.

UCLA's MethylScan reported 63% sensitivity across all stages and 55% for early-stage cancers at 98% specificity, based on a study of 1,061 participants [2]. For liver cancer surveillance among high-risk individuals, sensitivity reached nearly 80% at roughly 90% specificity [2]. CancerGuard showed 68% sensitivity for the deadliest cancers — pancreatic, lung, liver, esophageal, stomach, and ovarian — though full stage-stratified data has not been published [4].

For comparison, mammography detects approximately 87% of breast cancers in screened populations, and colonoscopy catches roughly 95% of colorectal cancers [8]. Standard screening tests for individual cancers, built on decades of validation, generally outperform MCED tests for the specific cancers they target. The MCED value proposition rests on breadth — covering cancers for which no standard screening exists, such as pancreatic, ovarian, and stomach cancers.

The Trial That Shook the Field

In February 2026, the NHS-Galleri trial — a randomized study of 142,000 adults aged 50-77 in England — reported that the Galleri test failed to achieve a statistically significant reduction in Stage III and IV cancer diagnoses, its primary endpoint [1]. The result sent GRAIL's stock down 47% in a single trading session [9].

The test did detect cancers at higher rates and showed a reduction in Stage IV diagnoses for a pre-specified subset of 12 cancers [10]. GRAIL's chief scientific officer acknowledged the miss but pointed to what the company called "compelling clinical benefits" in secondary endpoints [10]. A preprint posted to medRxiv argued that "stage slip" — delays in the NHS diagnostic pathway that allowed detected cancers to progress before confirmation — explained the gap, estimating that 84 cases were affected and that only 25 timely diagnoses would have pushed the composite endpoint to significance [11].

Critics were less forgiving. The trial's design had been structured around stage shift as the definitive proof that early detection translates to clinical benefit. Without it, the fundamental question remains unanswered: does finding cancer earlier with these tests actually help patients live longer or better?

The False Positive Problem

A 99.6% specificity rate sounds nearly perfect. But at population scale, the math changes. For every 100,000 people screened, a 0.4% false positive rate produces 400 people told they may have cancer when they do not [12]. About half of positive Galleri results are ultimately confirmed as cancer, meaning roughly 1 in every 100 people tested receives a positive result, and about half of those undergo further workup — biopsies, imaging, specialist referrals — for a cancer that does not exist [8].

A 2025 study published in the PMC estimated the downstream burden: each false positive triggers an average diagnostic workup costing thousands of dollars, with potential complications from invasive procedures, time lost from work, and significant psychological distress [12]. The American Cancer Society has noted that follow-up after a positive MCED result may lead to invasive procedures with no established clinical pathway for managing the process [8].

For standard screening, this problem is well-mapped. Mammography has a cumulative false positive rate of roughly 50% over ten years of annual screening, and colonoscopy carries a small but real perforation risk [8]. But those tests have decades of data calibrating the harms-to-benefits ratio. MCED tests do not.

Who Profits, Who Pays

GRAIL, the company behind Galleri, holds exclusive licenses to more than 230 granted patents globally and owns over 170 pending patent applications [13]. The company's technology is licensed in part from the Chinese University of Hong Kong, to which GRAIL pays single-digit royalty percentages on net sales [13]. GRAIL sold 185,000 tests in 2025, generating $136.8 million in revenue, at a list price of $949 per test [9].

The financial history is turbulent. Illumina acquired GRAIL in 2021 for approximately $8 billion — a price that plaintiffs in class action lawsuits allege was more than 90% above the company's actual worth, driven by insider financial motives [13]. The lawsuits further allege that executives misled investors with financial projections "up to quadruple the actual estimates shown secretly to Illumina's board" [13].

Regarding the NHS-Galleri trial, the published design paper states that none of the authors received explicit funding for their work, and that neither GRAIL nor the authors' employers had a role in writing the manuscript or the decision to submit it [14]. However, several GRAIL employees commented on an earlier draft [14]. A Health Affairs analysis criticized the taxpayer-funded GRAIL Galleri-Medicare study as having a flawed design [15].

Exact Sciences priced CancerGuard at $689 [4]. UCLA's MethylScan projects costs below $20 per test at current sequencing prices, though this reflects reagent costs alone and does not account for clinical infrastructure, regulatory compliance, or the diagnostic workups triggered by positive results [2].

The Regulatory and Coverage Landscape

As of April 2026, no MCED test has received FDA approval [16]. GRAIL submitted the final module of its Premarket Approval (PMA) application on January 29, 2026 [3]. The FDA designated Galleri as a Breakthrough Device in 2018, which accelerates review but does not guarantee approval [3]. CancerGuard launched as a laboratory-developed test (LDT), a regulatory category that has historically faced less FDA scrutiny [4].

Congress has shown interest in mandating coverage. The political momentum for MCED tests has grown, with bipartisan backing for legislation that would require Medicare to cover approved tests [17]. GRAIL has stated that FDA approval would be a "major trigger" for payer decisions [3].

For the uninsured and underinsured — a population already least likely to receive standard cancer screening — a $949 test is out of reach. Even at $689, CancerGuard requires out-of-pocket payment, though it may be eligible for health savings accounts [4]. UCLA's projected $20 per-test cost could change the equation, but the test remains in early-stage validation with a study population of just over 1,000 [2].

The access gap is not hypothetical. The clinical trials that generated the evidence for these tests enrolled populations with health insurance, access to cancer centers, and willingness to participate in research — demographics that do not reflect the populations with the highest cancer mortality rates [18].

What Independent Experts Say

At a September 2025 symposium at Fred Hutchinson Cancer Center, biostatistician Ruth Etzioni, PhD, offered a blunt assessment: "These tests may give a wrong answer" [18]. Pulmonologist Matthew Triplette, MD, MPH, noted: "We don't have a guideline-based test now" [18].

A systematic review in the Annals of Internal Medicine concluded that evidence was "insufficient to evaluate harms and accuracy" of MCED screening [18]. The review noted that most accuracy studies relied on high risk-of-bias case-control designs using clinically confirmed cancer cases, rather than measuring prediagnostic performance in asymptomatic populations [16].

Fred Hutch oncologist Eric Collisson, MD, described a "diagnostic no man's land" — the gap between a positive MCED result and a confirmed diagnosis, during which patients lack a clear clinical pathway [18]. No established protocol exists for managing positive results, and roughly 50 different companies are developing tests with varying approaches and performance standards, while keeping assay methodologies proprietary and limiting independent scrutiny [18].

The ASCO (American Society of Clinical Oncology) Educational Book published a 2025 assessment warning that MCED tests are at a "crossroads" with "commercial availability ahead of definitive evidence" [16].

The PSA Precedent

The comparison that haunts this field is prostate-specific antigen (PSA) testing. Introduced for routine prostate cancer screening in the late 1980s, PSA testing led to massive increases in prostate cancer diagnoses — and massive increases in treatments for cancers that would never have caused symptoms or death [19].

Data from 1988 through 1998 showed overdiagnosis rates of approximately 29% for white men and 44% for Black men [19]. The downstream harms were substantial: unnecessary surgeries causing urinary incontinence and erectile dysfunction, radiation therapy with bowel and bladder side effects, and the psychological burden of a cancer diagnosis for a disease that would not have shortened life [19]. It took roughly two decades for guidelines to catch up, with major medical organizations pulling back from routine PSA screening by 2012 [19].

The parallel is not exact. MCED tests screen for many cancers simultaneously, some of which — like pancreatic and ovarian — are genuinely lethal and lack any standard screening. But the core risk is the same: detecting conditions that do not benefit from earlier treatment, triggering cascades of intervention that cause net harm.

Source: OpenAlex

Data as of Jan 1, 2026CSV

Academic interest in liquid biopsy cancer detection has surged, with nearly 88,000 papers published since 2011 and a peak of 12,774 in 2024 [20]. The volume of research reflects both genuine scientific promise and the enormous commercial stakes.

How MCED Tests Would Interact with Existing Screening

Every major medical organization, including the American Cancer Society and MD Anderson Cancer Center, emphasizes that MCED tests do not replace standard screening — mammograms, colonoscopies, Pap smears, and low-dose CT scans remain necessary regardless of MCED results [8]. The NCI's Division of Cancer Prevention lists "Will people stop standard-of-care screening if they get a negative MCD test?" as an explicit open question [12].

The fear is rational. A patient who receives a negative MCED result — which, given the tests' 40-63% sensitivity for all cancers, will include many people who actually have cancer — may feel falsely reassured and skip their colonoscopy or mammogram. This substitution effect could paradoxically increase late-stage diagnoses for the cancers where standard screening works best.

On the other hand, MCED proponents argue that a single blood draw could improve screening compliance by reducing the burden of multiple appointments and uncomfortable procedures [21]. For cancers with no current screening — which account for roughly 70% of cancer deaths — any detection is better than none [7].

The specialist workforce implications are also unclear. If MCED tests gain traction, gastroenterologists could see shifts in colonoscopy volume, radiologists could face changes in mammography and CT referral patterns, and oncologists would need to manage a new category of patients: those with a positive blood test and no confirmed diagnosis [18].

Source: OpenAlex

Data as of Jan 1, 2026CSV

Research publication volume on multi-cancer early detection blood tests has grown from roughly 4,300 papers in 2011 to over 31,600 in 2024, reflecting the field's rapid expansion across genomics, oncology, and diagnostic medicine [20].

What Comes Next

The field faces a timing problem. The tests are commercially available now, but the evidence needed to know whether they help or harm at scale will not arrive for years. The NCI-funded Vanguard study, which aims to enroll 24,000 participants at nine sites, does not expect feasibility results until 2027 [18]. Long-term mortality data — the only metric that can definitively answer whether early detection saves lives — will take longer still.

In the meantime, an estimated 50 companies are building MCED products [18], Congress is considering coverage mandates [17], and patients are paying hundreds or thousands of dollars out of pocket for tests that no regulatory agency has approved [9]. The UCLA and Peking University advances suggest that costs could eventually fall dramatically, but both remain in early-stage research with small study populations [2][6].

The question is not whether blood-based cancer detection will eventually work. The biology is real, the technology is advancing, and academic output continues to grow. The question is whether the gap between commercial availability and clinical evidence will close before the harms of premature adoption — false positives, overdiagnosis, misallocated healthcare spending, and false reassurance from negative results — accumulate faster than the benefits.

As Ruth Etzioni of Fred Hutch put it: "These tests may give a wrong answer" [18]. The challenge now is building a system that accounts for that uncertainty before millions of people are screened.