Cancer Rose, 3 octobre 2026

Sasieni P, Johnson P, Round T, Warwick J, Jones H, Dolly S, et al. Effect of Screening with Multicancer Early-Detection Test on Late-Stage Cancer Diagnosis. N Engl J Med. 22 sept 2026;395(13)
https://www.nejm.org/doi/10.1056/NEJMoa2505723?referrer=https%3A%2F%2Ft.co%2F

Effect of Screening with Multicancer Early-Detection Test on Late-Stage Cancer Diagnosis

A study co-led equally by Professor Richard D. Neal (University of Exeter) and Professor Charles Swanton (University College London and Francis Crick Institute), the latter being a world reference in the study of tumor evolution. The first listed author is Professor Peter Sasieni, a recognized expert in cancer epidemiology and screening from Queen Mary University of London. The team brings together individuals representing the British public healthcare system (NHS England) and the private biotechnology research sector, with the participation of scientists from the company GRAIL Bio UK (sponsor and funder of the trial).

Editorial

Written by David J. Hunter, an Australian epidemiologist, and Professor Richard Doll, an epidemiologist at Oxford Population Health. David Hunter was previously a professor in the Department of Epidemiology and the Department of Nutrition at Harvard University. He was an associate epidemiologist at the Channing Laboratory of Brigham and Women’s Hospital (Boston), where he participated in research programs on breast cancer, cancer epidemiology, and cancer genetics.

A Little Context

Despite the failures of future screenings (melanoma, thyroid, prostate, breast) to overcome cancer and its unfortunately most severe forms, we nevertheless retain the deeply held conviction that if we could detect every cancer cell, we could “defeat” the disease.

Therefore, MCED (Multi-Cancer Early Detection) tests have been recently developed. These are new blood tests capable of simultaneously identifying traces of several dozen different cancers (up to more than 50) at a very early stage, before the appearance of the first symptoms.

How do MCED tests work?
Here is the explanation from Colin Richard, professor of pathology at the School of Medicine, University of Washington:
All cells in the body, including tumor cells, release DNA into the bloodstream when they die. MCED tests look for traces of circulating tumor DNA in a blood sample. This circulating “cell-free” DNA contains information about the type of tissue it comes from and whether it is normal or cancerous.
MCED tests attempt to detect cancer at an early stage when tumor cells are not yet very numerous. Detecting these cancer cells can be difficult at an early stage because non-cancerous cells also shed DNA into the bloodstream. Since most of the DNA circulating in the blood comes from non-cancerous cells, detecting the presence of a few cancerous DNA molecules is like looking for a needle in a haystack.

To make things even more difficult, blood cells naturally shed abnormal DNA as people age, and these strands can be confused with circulating cancer DNA. This phenomenon, known as clonal hematopoiesis, baffled early attempts to develop MCED tests, resulting in too many false-positive results.

More recently developed tests are able to avoid interference from blood cells by focusing on a type of “molecular barcode” embedded in cancer DNA that identifies the tissue of origin.

In 2021, the biotechnology company GRAIL launched the first commercially available MCED test in the United States. Its Galleri test claims to be able to detect more than 50 different types of cancers. MCED tests are not yet covered by insurance and are not close to being covered, given the disappointing results of the study we are about to look at. The price of GRAIL’s Galleri test is currently $949, and for now, individuals must pay for it out of pocket.

Back in 2022, Gilbert Welch, an internist and senior researcher at the Center for Surgery and Public Health at Brigham and Women’s Hospital in Boston, and Barnett Kramer, an oncologist, member of the Lisa Schwartz Foundation for Truth in Medicine, and former director of the Cancer Prevention Division at the US National Cancer Institute, expressed reservations in an article published in the journal STAT.

The two American authors posed two fundamental questions at the time:

1. Do liquid biopsies work as advertised?
2. If liquid biopsies are effective, are they effective enough to make them worth it? (Read https://cancer-rose.fr/en/2022/09/15/liquid-biopsies-the-grail/)
   

For G. Welch, there was only one way to evaluate the effectiveness of liquid biopsies in the early detection of cancers, and that required a randomized trial in which participants would be divided into two groups. One would undergo regular screening, and the other would not. Participants would then be followed for about ten years, and the number of deaths in each group would be counted. Randomized controlled trials correspond to the “gold standard” of scientific studies; it is a proven method.

Coincidentally, the National Health Service (NHS) of England had then begun recruiting 140,000 people for a clinical trial. The most relevant outcome to measure would be the number of deaths in each group, according to G. Welch; in this case, the study we summarize below sought to determine whether the use of a blood screening test (developed by the company Grail) made it possible to reduce the number of cancer diagnoses discovered at an advanced stage (stages III or IV), compared to standard medical management.

The Clinical Trial Published in the NEJM September 2026

A total of 142,250 people aged 50 to 77, recruited in England, provided blood samples during three annual visits as part of this initial randomized, double-blind, controlled trial.

That is, for a group of 71,122 people (the intervention group), the blood was immediately tested with the MCED test. For a group of 71,128 people (the control group), the blood was simply stored without being analyzed.

Next, these two groups were compared to examine the incidence rate (the number of cancers) of stage III or IV cancers among 12 predefined cancer types. Thus, the goal was to see whether very early detection through the search for circulating tumor DNA in the tested “intervention” group could prevent these patients from developing advanced-stage cancers. Formulated differently, does detecting small amounts of tumor DNA allow for early medical management to prevent patients from developing serious cancers?

The NHS-Galleri trial focused primarily on 12 specific types of cancers that are aggressive, often fatal, and for which there is no standardized mass screening program: pancreas, esophagus, liver/bile ducts, stomach, anus, bladder, colorectal, head and neck (upper aerodigestive tract), ovary, lymphoma, and myeloma.

More common cancers, such as breast or cervical cancer, were deliberately excluded from this list, as they already have well-established screenings (mammography, Pap smear).

Results of the Clinical Trial

They are overall rather disappointing.

The incidence rate of stage III or IV cancers showed no significant difference between the two groups. In other words, the multi-cancer blood screening test did not reduce the number of stage III or IV cancer diagnoses after three years of follow-up.

After three screening rounds, the incidence rate ratio for stage IV cancers suggested a potential downward trend for the most advanced stages, meaning that perhaps a small reduction might still be observed in the longer term for very advanced stages. According to the authors, a longer follow-up would be needed to see how these data evolve and to confirm whether the difference observed specifically for stage IV would be maintained over time.

But globally, the test still lacks efficacy where it is most expected—namely, a drastic and noticeable reduction in the most serious cancers, and therefore in mortality.

The editorial suggests restricting the use of these tests for now to well-targeted situations:

• * For patients already presenting symptoms of cancer whose origin remains undetectable by standard examinations.
• * For the surveillance of individuals at very high genetic or familial risk for cancers that are difficult to screen (such as pancreatic or ovarian cancers).

Early detection of cancer through a simple blood test nevertheless remains (rightly so?) the “Holy Grail” of oncology. However, it is absolutely essential to avoid the pitfall of overdiagnosis by generalizing highly expensive tests too broadly and too quickly (a mistake of rapid and thoughtless generalization that occurred with mammographic screening). These tests risk finding circulating DNA in almost everyone, and it is prudent to guide future research more effectively.

In medicine, we now see before our eyes the disastrous consequences of the unnecessary over-detection of indolent, highly frequent cancers (overdiagnosis). This leads to absurd results, such as equally unnecessary overtreatment of healthy individuals, turning them into patients and plunging them into distress, at the cost of infinite medical expenses and diverting financial resources from more useful healthcare areas for the population.

Another Timely Publication

Rebbeck TR. Precision prevention and early detection: framework for a new clinical cancer control pathway. The Lancet. Published online September 18, 2026.

This is a conference article published by Dr. Timothy R. Rebbeck in The Lancet on September 18, 2026, which proposes a new clinical framework called PPED (Precision Prevention and Early Detection of cancer).

T. Rebbeck starts from the current observation: the failure of existing methods.

The stage at which a cancer is diagnosed remains the determining factor for patient survival (90% 5-year survival for stage I compared to only 15% for stage IV). Yet, in the United States, only 13% of cancers are detected through current routine screenings (mammograms, colonoscopies), and 20% to 50% of cancers are only discovered during an emergency room visit.

Why do screenings miss their target? Presumably because they are based on a false theory of cancer evolution, according to which advanced and serious cancers stem from small and less serious cancers, and that stopping indolent cancers will reduce advanced stages. However, aggressive cancers that escape tests or recur aggressively possess great genetic instability from their very inception. Even if a test detects 99% of the tumor at an early stage, the remaining 1% of cells can be intrinsically resistant and capable of generating rapid metastases. When screening manages to detect this type of poor-prognosis cancer (which is very often missed by screening), it only advances the date of diagnosis (known as lead-time bias) without changing the patient’s ultimate fate.

So, what is PPED (Precision Prevention and Early Detection)?

PPED aims to tailor cancer research and control to each individual’s risk profile (based on genetics, family history, and environmental exposures).
It relies on 4 key questions: Who is at high risk? What intervention is appropriate? When and with what intensity should action be taken? How and where should these services be delivered?

PPED introduces three major concepts:

• * Prioritizing high-risk populations: targeting individuals such as carriers of the BRCA1 mutation for ovarian cancer or people with Lynch syndrome for colorectal cancer. This optimizes the efficiency of examinations and improves the benefit-risk ratio.
• * De-escalation: reducing the frequency or intensity of exams for low-risk individuals.
• * Integration of new technologies: including the use of liquid biopsies and multi-cancer early detection blood tests (such as the Galleri test mentioned earlier). But applying Dr. Rebbeck’s logic, these tests should only be offered to the “right people”—meaning the Galleri test should never be offered to the general population in an undifferentiated manner. To become effective and financially reimbursable, it must be integrated into a precision pathway for individuals at specific risk.

For the author, the greatest challenge of the next ten years is not scientific (creating new tests), but organizational and human. The implementation of PPED faces several challenges:

• Lack of structure: the current care pathway is too fragmented between general medicine, mass screening programs, and specialists.
• Need for new professions: the emergence of a new clinical discipline is required, including “cancer preventionists” specifically trained to manage and monitor these personalized programs over the long term.
• An imperative for equity: currently, genetic risk calculation scores (such as polygenic risk scores) are constructed from data derived from populations of European origin. Applying them to populations of African, Asian, or Latino origin leads to clinical errors and risks worsening health inequalities.
  

Bringing These Two Publications Together

1. The Galleri test failed in its generalist approach

Dr. Rebbeck explains that screening only works if a high-risk population is targeted.
However, the NHS-Galleri trial tested 142,250 people from the general population (aged 50 to 77) without any specific targeting of their genetic risks or history. The test was unable to reduce the overall rate of stage III and IV cancers. Why? Because the test misses more than 83% of cancers at the most curable stage. And at the metastatic stage, the test detects almost everything (advanced tumors secrete much more circulating DNA), but at a stage where a cure is very difficult.
Furthermore, as seen above, the most aggressive cancers are often metastatic from the start; even when detected early, they have frequently already seeded, even if these seedings are not immediately visible in neighboring lymph nodes or distant organs.

2. Where the Galleri test could validate the PPED vision

Despite this failure on the primary endpoint, the clinical trial validated two fundamental pillars of PPED:

• 1-The false-positive rate is extremely low, at just 0.45%. Approximately 1 in 220 people receives a false alarm. This is a crucial point in mass screenings, as these events unnecessarily cause anxiety for tens of thousands of healthy people and saturate hospitals and offices with unjustified follow-up examinations (scans, biopsies).
• 2-The trial showed a potential drop in stage IV cancers. Applying Dr. Rebbeck’s logic, the Galleri test could be integrated in a targeted manner for the most at-risk individuals, within a precision pathway involving risk score calculations, and only if the patient is identified as being at “high risk.”

CONCLUSION

The Galleri trial shows that the technology works on a purely technical level (good specificity, good organ orientation), but we are not there yet in terms of effectively reducing advanced cancers. This is likely largely because the “nature” of cancer is too complex.

The Lancet article argues that human and clinical infrastructure is now needed to know whom to apply it to intelligently. According to the author, if PPED fails, it will not be due to a lack of technology, but rather due to a lack of healthcare infrastructure, trained personnel, and long-term patient follow-up.

We are more skeptical; the failure of mass screening is largely explained by our obsolete, linear, and “mechanistic” view of the disease.
For decades, cancer has been thought of like a clock: a cell becomes abnormal, then forms a small tumor (stage I), which grows (stage II), invades the lymph nodes (stage III), and finally sends out metastases (stage IV). According to this linear logic, it would be enough to arrive “earlier” to cure everything.
The biological reality is radically different and validates the hypothesis regarding the intrinsically poor characteristics of certain cancers.
The risk of these types of multiple detection tests is that they only detect aggressive cancers once they have already programmed their dissemination. This often happens right from the start, at the very moment of cancer genesis, from the very first cells, confirming that the intrinsic characteristics of the tumor dictate its detectability and danger.
“Small” cancers are not the origin of these most aggressive cancers, which stem from other cellular entities that are biologically very different. (Read https://cancer-rose.fr/en/2017/06/10/are-small-breast-cancers-good-because-they-are-small-or-small-because-they-are-good/)

Is everything lost then?

Perhaps this is where Dr. Rebbeck’s vision of PPED makes sense. If the natural history of cancer is not linear, the medical response must not be linear either.

The future is no longer about detecting an already formed tumor mass, but about targeted molecular interception. Instead of looking for a visible cancer, we look to identify inflammatory microenvironments or very early biological precursors in hyper-targeted patients. For now, we are facing a failure. Perhaps fundamental research in the future will bring a more precise understanding of cancer mechanics and more refined answers for patients most exposed to specific risks that make them prone to aggressive cancers.

Twists and Turns: Clinical Failure in the UK, but Green Light in the US!

As explained earlier in the article, the Galleri test—a blood test allowing the simultaneous detection of several types of cancer—is a failure and finds itself at the heart of an incredible clash of views.
On one side, the large British clinical trial, according to the promoters themselves (the NHS as an operational partner and the company Grail as a major financial sponsor), shows no reduction in advanced cancer stages for the 12 types of cancers explored.
But on the other side, experts from the US FDA have just recommended its approval! (FDA: US federal agency responsible for protecting public health by controlling the safety of drugs, food products, and medical devices).
This is reported in a BMJ article of which we were able to obtain the full text.

Let’s briefly summarize the facts: The clinical trial conducted in the UK is one of the most ambitious of its kind: 142,000 participants aged 50 to 77 followed for 3 years. The results, published in the prestigious New England Journal of Medicine, are clear: the test failed to meet its primary objective, showing no reduction in the overall incidence of advanced-stage cancers (stages 3 and 4 combined) after three rounds of screening. The British NHS is pulling back sharply, even though tests were already on pre-order in the UK.

Why did the FDA say “yes” (though not completely)?

The day after this publication, the FDA’s Medical Devices Advisory Committee met on September 23, 2026, and voted in favor of the test by 6 votes to 4.
For the panel, there would be a potential slight downturn in the most advanced stage 4 cancers over a certain period, to be confirmed in the longer term according to those who voted “in favor.” Additionally, the test would allow for the detection of cancers for which no screening protocol currently exists (such as pancreatic cancer).
On the other hand, the experts demanded that the product no longer be marketed as an “early” screening tool, but simply as a “detection” tool.

Two researchers cited in the BMJ article expressed concern.

For Dr. Stanley Lipkowitz of the NIH (National Institutes of Health), we have no proof that this test saves lives. The true evaluation criterion, namely the reduction in overall mortality, was not the primary objective of the trial. Data on mortality are not expected to be known for another two years.

For the medical geneticist and British researcher Professor Houlston (ICR – The Institute of Cancer Research in London), the cost and the risk of overdiagnosis represent major critical points. The test ($949) raises fears of a financial abyss for healthcare systems and would subject a multitude of patients to false positives, resulting in significant physical and psychological damage.

It is important to specify that the FDA has not yet definitively validated the test. However, the FDA usually follows the recommendations of its expert panel most of the time. The final decision on commercialization is expected in the coming months.
The UK is biding its time, while in the United States, there is unfortunately much to fear regarding the commercialization of an innovation without formal proof of its efficacy and safety.

Jeanne Lenzer and Shannon Brownlee, investigative journalists, were already talking about the FDA’s drift in “Science for Sale.“

Science for Sale: How Drugmakers Captured the FDA

The pharmaceutical industry has spent decades manipulating the country’s most important drug regulatory authority to obtain automatic approval. The Trump administration is only making things worse. With the proposed automatic translation, you will easily have access to this excellent in-depth article.

But haven’t we seen the manipulation of scientific data in screening before? Alas, yes…

Long before the generalization of mammographic screening, independent health professionals and researchers, including Nordic researchers from the Cochrane Collaboration, shared their doubts about the ability of mammographic screening to be useful to women.

The validity of an intuitive framework that we all share—which assumes a linear, inevitable natural history including a sufficiently long phase during which it is possible to cure the disease—was questioned. In the 2000s (and even before), researchers began to sound the alarm (Cochrane Collaboration; Bernard Junod, a Swiss epidemiologist), most of whom were prevented from publishing or even censored.

Despite the disappointing results of pilot screening sites in France in the 1980s, and despite all the warnings, screening was generalized nationwide in France in 2004, and campaigns were established with heavy media backing across Western countries.
For decades, it has been increasingly recognized that there are harms associated with screening. However, many doctors, medical students, and patients continue to believe that, for a large proportion of the population, recommended screenings allow for early diagnosis and treatment, and prevent premature deaths.
It is a true conspiracy of hope* taking precedence over all hard scientific data, which nevertheless stand out with stark evidence before our eyes over 30 years of vain dreams of one day curbing cancer through early detection.
*Interviews of the author of Conspiracy of Hope, Renee Pellerin

The evidence allows for a much better understanding of the harms associated with screening, including overdiagnosis, false positives, and excessive examinations.
Yet the belief persists, public knowledge is kept to a minimum, and patients are unaware of these potential harms, even and especially in populations that undergo regular screening.

The innovation of multiple diagnostic tests offers a market far too promising not to attract the greed of companies that see it as an inexhaustible financial windfall… To be continued…

🛈 Nous sommes un collectif de professionnels de la santé, rassemblés en association. Nous agissons et fonctionnons sans publicité, sans conflit d’intérêt, sans subvention. Merci de soutenir notre action sur HelloAsso.
🛈 We are an French non-profit organization of health care professionals. We act our activity without advertising, conflict of interest, subsidies. Thank you to support our activity on HelloAsso.