Investors are pouring billions into companies claiming they can analyse DNA to find the disease early. But some scientists question if they really work.
Valerie Caro is perhaps unusual in that she considers herself fortunate to have been diagnosed with cancer.
The real estate broker from Flagstaff, Arizona, spent US$950 ($1,500) last year on an experimental blood test called Galleri which detected gallbladder cancer at an early stage. Following scans to verify the diagnosis, Caro had surgery to remove a tumour and underwent chemotherapy. She is now in recovery.
“Every single doctor I spoke with said, ‘You are so lucky to have caught it’ because it goes gallbladder, liver, lungs, like that. This literally saved my life,” says Caro, who is in her mid-fifties.
Caro is among a pioneering group of people in the US and UK who have elected to take part in a new form of cancer screening known as multi-cancer early detection tests (MCED). The tests use gene sequencing or other novel technologies to detect fragments of DNA expelled by cancerous cells which circulate in people’s blood, allowing the identification of multiple types of cancer from a single blood draw. They have been hailed as “revolutionary” and “cutting edge” by British and US health chiefs.
Health bodies in both nations have set up MCED clinical trials in the hope that the tests can be rolled out to the population at large; cancer is responsible for one in six deaths globally. The UK’s National Health Service is participating in a clinical trial of the Galleri test involving 140,000 patients. If the trial is successful, the NHS may buy 1mn units from Grail, the US company that developed Galleri.
Individuals like Caro who are eager to take the tests do not have to enrol in a clinical trial. Following success in some small initial trials, Grail and an Indian company, Datar Cancer Genetics, are already selling MCED tests to the public. Datar’s test is available in the UK with prices at various health clinics cited above £1,000 ($2,000). In the US, Galleri tests are sold as “laboratory-developed” because they have not yet been approved by the US Food and Drug Administration.
At least a dozen rivals are also developing similar products for a sector that analysts forecast could expand from around US$1.5 billion in revenues this year to US$23 billion by 2031. In March, following intense lobbying by industry, US lawmakers introduced bipartisan legislation in Congress aimed at fast-tracking funding for the tests for people in US government health schemes such as Medicare.
But not everyone is convinced the tests live up to the hype. Several health experts and scientists told the Financial Times that the tests could harm rather than help some patients due to risks associated with misdiagnosis, over-diagnosis and over-treatment.
Bold claims about the potential of “revolutionary” blood tests have made investors nervous following the recent Theranos scandal, which led to the sentencing of founder Elizabeth Holmes last year to 11 years in prison for fraud. Holmes raised billions of dollars of funding on claims Theranos had developed machines capable of running hundreds of tests — from cholesterol to complex genetic testing — from a single pinprick of blood. The technology didn’t work.
Critics of MCEDs also warn that a deluge of test results could put additional pressure on already overstretched health services. A Galleri pilot screening programme in Arizona for firefighters missed many instances of cancer, raising questions about accuracy. Companies might be putting profits above patient welfare by selling the tests before they complete the randomised trials required to prove they work and can save lives, they say.
“It is a disgrace that companies are selling these tests without knowing what the benefits and the harms are,” says Paul Pharoah, a professor of cancer epidemiology who works at Cedars-Sinai, a non-profit healthcare organisation in Los Angeles.
“False positives are a problem. False negatives are a problem. We don’t know how big those problems are, and we don’t know whether these tests make a meaningful difference to cancer mortality.”
Prevention: better than cure
Screening has been an important tool in the fight against cancer for more than half a century. It aims to reduce mortality or morbidity through early detection and treatment, for instance by identifying disease precursors such as patches of abnormal cells that can develop into cervical cancer or polyps that cause bowel cancer.
National screening programmes typically target a handful of common cancers, such as cervical, breast, bowel and lung, where there is evidence that early detection can improve outcomes for patients. Studies show that the rollout of pap tests, which detect abnormalities in cervical cells, is mainly responsible for a 75 per cent fall in the incidence and mortality of the disease in the US.
At Grail’s headquarters in Menlo Park, California, executives hope to recreate this success for more than 50 different cancers for which there are no screening tests available, including tumours in the pancreas, oesophagus, neck and gallbladder. About three-quarters of new cancer cases and deaths in the US are due to diseases without established screening programmes. Grail is pitching the Galleri test as a key tool that can help the Biden administration deliver its “cancer moonshot” initiative — a suite of policies aimed at reducing the cancer death rate by half within 25 years.
“We’ve been fighting a war on cancer for 50 years, and it’s not one we’re winning and why is that? It’s because we still are finding most cancers too late when they have already developed symptoms and already spread,” says Josh Ofman, president of Grail.
The scientific breakthrough that led to the development of Galleri occurred a decade ago when Meredith Halks-Miller, a former laboratory director at gene-sequencing company Illumina, spotted abnormal DNA fragments in the blood of expectant mothers. She had a hunch that cancer was the only thing that could cause the particular pattern of fractured chromosomes and began following the otherwise healthy patients.
“All the mothers whom we predicted to have a malignancy were diagnosed with cancer shortly after delivering their babies. It was a revelation,” says Halks-Miller, who has since retired.
Halks-Miller’s discovery fired the starting gun on the race to develop a diagnostic blood test. Illumina spun off a new company, Grail, to develop the test, raising more than US$2 billion from investors including Bill Gates, Jeff Bezos and the Chinese company Tencent.
Galleri tests work by using genetic sequencing and artificial intelligence to scan a certain type of DNA — cell-free DNA — found in people’s blood for changes caused by cancer cells. This analysis can provide a signal that cancer is present and give information about where in the body the signal originated from.
Grail published the results of a study of almost 6,700 participants aged 50 years or over in September last year. Of 92 participants flagged with a potential cancer signal, about 35 participants were diagnosed with cancer. All but 10 of these diagnoses were for cancers that did not have routine cancer screening programmes in place. Fifty-seven were false positives, almost a third of which prompted an invasive procedure to rule out a cancer diagnosis.
Promising results from early MCED test trials led to a rush of start-ups and investment. In 2020, Illumina bought Grail back for $8bn, citing its ability to “help transform cancer care”. The transaction attracted the attention of European and US antitrust authorities, which said the tests had the potential to “revolutionise” the fight against cancer. Last year the EU ordered Illumina to unwind the transaction, arguing it would dent competition in the MCED market since Illumina would have an incentive to cut off Grail’s rivals from using its sequencing technology.
Grail will soon face competition in the US from Datar Cancer Genetics, which inked a US$250 million deal last year with US-based diagnostic laboratory company Artemis DNA to fund the rollout of its own MCED test called TruCheck in the US and Vietnam.
Other competitors such as Exact Sciences, based in Wisconsin, and San Francisco’s Freenome, a start-up backed by Roche, are racing to launch their own tests.
“When interest rates were low and cash was easier to come by there was huge excitement about the potential for these tests and companies were able to raise billions of dollars of investment,” says Vijay Kumar, an analyst at Evercore ISI, an investment bank.
But their success rests on getting regulatory approval. In the US, because Galleri has not yet won approval from the FDA, it is not covered by US government health insurance schemes. “This [approval] remains an open question that will depend on the results in ongoing trials,” Kumar adds.
Illumina and Grail increased their US lobbying spending fivefold to US$14.6 million in 2021 and 2022 when compared to the previous two years. Illumina also expanded lobbying in Europe and the UK, where it hired former prime minister David Cameron as an adviser in 2019 shortly before winning a gene sequencing contract with a company owned by the Department of Health and Social Care. (Illumina says it no longer employs Cameron as a paid adviser.) In November 2020, the NHS struck its deal with Grail to pilot Galleri — a collaboration that is key to winning regulatory approval. Interim results are expected next year.
An uncertain diagnosis
The full-throttle corporate race to persuade governments to integrate MCED tests in national screening programmes is raising concerns among some cancer specialists.
Research shows there are drawbacks to screening for certain types of cancer, particularly for those where there is limited evidence that early detection and treatment saves lives. Tests can sometimes produce a cancer signal, even when the body is cancer-free. Such false-positive results can lead to more tests or treatment, which can carry potential risks and cause needless anxiety.
Screening can fail to detect cancers, which provides a false sense of security and may cause some patients to neglect symptoms. About 20-30 per cent of women with breast cancer have tumours that are missed by mammogram screening, according to research.
As an oncologist, I love the concept. Cancer is a pandemic and we need a safe way to screen younger patients.
Screening can also pick up slow-growing tumours that don’t cause symptoms during a person’s lifespan, a process known as over-diagnosis. “Over-diagnosed patients...can’t be helped because there’s nothing to fix, but they can be hurt in the process,” said H Gilbert Welch, a cancer epidemiologist at Brigham and Women’s Hospital in Boston, Massachusetts.
In 2012, the US Preventive Services Task Force recommended against screening all men for prostate cancer using a prostate-specific antigen blood test, concluding the benefits did not outweigh the risks. The advice reflected concerns about side effects linked to follow-up testing via unnecessary biopsies and treatment, which include erectile dysfunction and incontinence.
Welch says the new generation of MCEDs are a remarkable technology. But it is a mistake to rush their rollout without the necessary long-term studies that prove they save lives. “My guess is this is a quagmire that costs a lot of money to governments and individuals. If you test all Americans aged 50 and above it costs US$100 billion a year,” he says.
“Such a policy would lead people to go through a lot more testing. It will make some people poor and make others excessively worried and I don’t think it’s really going to change mortality much, if at all,” he adds.
Some doctors already prescribing the Galleri tests to patients say they have reservations. Dr Vershalee Shukla, who runs a cancer screening programme in Scottsdale, Arizona, for firefighters — a high-risk group due to their regular exposure to environmental toxins — says the Galleri test missed the cancer in almost 50 individuals out of roughly 3,000 tests she conducted over two years. The test accurately detected three cancers and produced two false positives, she adds.
Firefighters in the nearby city of Mesa are offered additional screening options under a publicly funded programme, including a full-body MRI scan. One of the patients, Drew Mogalian, says that after getting a clear result from Galleri he only went for a follow-up MRI scan after a discussion with his wife about the death of a colleague from cancer.
“The full-body MRI came back and showed a brain tumour,” he says.
Mogalian had a biopsy in September which confirmed the diagnosis. He has recently completed a course of chemotherapy and is awaiting further results.
Dr Shukla says her experience testing the firefighters provoked mixed emotions. “As an oncologist, I love the concept. Cancer is a pandemic and we need a safe way to screen younger patients and blood tests are very safe. There is no radiation and it’s minimally invasive,” she says. “But it missed almost 50 cancers. This provides a false sense of security to patients, who may not want to do additional screening that could pick up their cancers.”
She says it is “upsetting” that Grail is still promoting its test to firefighters even when she presented the company with her evidence. Additional testing is needed in younger, high-risk populations if the test is going to be marketed to them, she adds.
Grail says that all cancer screening tests have both false positives and false negatives, many of which are at a higher rate than Galleri. One in eight mammograms produces a false negative result while the rate of colonoscopies can be between 7 per cent and 25 per cent, according to Grail. Screening tests run on small populations will see different results based on age and risk, says the company’s chief medical officer, Jeffrey Venstrom. “At the national level, our cancer signal detection rates are consistent with our clinical studies and epidemiological modelling,” he says, adding that Galleri is finding asymptomatic, potentially deadly cancers that would otherwise have gone undetected.
Ofman, Grail’s president, says that with the NHS trial the company is undertaking the largest randomised control trial ever for a MCED test. The trial would “once again validate the performance” of Galleri and measure how it can reduce the number of cancers detected in their late stages, he says, adding that he feels a sense of “urgency” and “responsibility” to get tests out to as many patients as possible, as quickly as possible.
But the NHS trial will not measure whether Galleri reduces mortality rates when used as a screening test — that would take a decade to produce definitive results.
Pharoah of non-profit Cedars-Sinai says the trial is “seriously flawed” because the first and foremost thing you should measure is whether a test that would be applied to a screening programme reduces mortality. “Just because you reduce late-stage disease does not mean you will reduce mortality from cancer or other causes,” he says.
Susan Bewley, a consultant obstetrician and emeritus professor of obstetric and women’s health at King’s College London, says the problem with MCED screening is that the more of this you do on healthy people the more errors you get.
“This sort of screening test could bankrupt the NHS, prioritise people who are well over those who are sick, and make people ill. This is ethically dubious,” says Bewley. “Before these tests are rolled out we need to follow hundreds of thousands of people over many years to determine if the treatments actually stop them from dying.”
Bewley is concerned that the way the NHS trial is structured means that Grail is likely to claim success even though many of the people with early-stage cancer may never have become ill.
“Screening is like a modern form of bloodletting with leeches: if you died it’s because we didn’t leech you early enough and if you didn’t die then the leeches saved you,” she says.
In numbers
- US$8 billion: Sum for which Illumina controversially bought back Grail in 2020
- 1 in 8: Number of mammograms that produce a false negative, according to Grail
- US$14.6 million: Amount that Illumina and Grail spent on US lobbying in 2021 and 2022
Written by: Jamie Smyth
© Financial Times
