HAMBURG — The day that had promised to shake up cancer research had arrived, and Wolfgang Janni is sitting in a conference hall in Düsseldorf, waiting for something big to happen. A doctor from Heidelberg has just stepped up to the podium, ready to give a lecture. That morning, the Bild newspaper splashed the doctor’s photo across its front page — white coat, arms crossed — next to the headline: “World Sensation from Germany: Blood Test Detects Breast Cancer.”
The innovation it announces sounds groundbreaking — a real game-changer, even: a simple blood test that can detect a cancerous tumor before it even has a chance to do harm.
The Heidelberg doctor is about to share this “breakthrough” with his colleagues for the first time. But as Janni looks around the room, the “sensation” promised in the headlines is nowhere in sight. “There were a lot of big claims, but very little evidence,” Janni says. “Lofty words, but no solid data.”
Janni, who himself is researching similar tests, is unimpressed.
What had appeared to be a major discovery on February 21, 2019, soon devolved into a scandal, with the public prosecutor’s office getting involved. False promises, unproven methods. Key officials resigned, and more than five years later, the promised tests are still nowhere to be seen.
“Since that day, I’ve become skeptical whenever someone starts talking about the next big thing in cancer detection,” says Janni, a breast cancer researcher and director of the Women’s Clinic at Ulm University Hospital. The scandal revealed that patients were not the only ones hoping for a miracle test: researchers were also chasing personal breakthroughs while companies eyed a piece of the financial pie.
Now, however, the stakes are even higher. While the Heidelberg researchers once promised a test as reliable as a mammogram for detecting breast cancer, today’s tests are reportedly able to detect multiple types of cancer at once.
Holy Grail
One of the most high-profile players in this field is Grail, a U.S.-based biotech company with a name that sounds anything but modest: “Holy Grail”.
Their test, called Galleri, is touted as a game-changer. It can reportedly identify the “fingerprints” of over 50 different cancers in a blood sample, promising to “redefine early cancer detection,” according to the company’s website.
They’re already running large-scale studies in several countries. No other company researching multi-cancer tests has made it this far. But similar studies are underway in Germany, aiming to detect cancer earlier and more accurately than ever before.
Will all the traditional methods (MRI scans, biopsies, endoscopies…) soon be a thing of the past?
“One thing is clear: We’re much closer to a blood test for cancer than anyone could have imagined just a few years ago,” says Janni. So, will there soon be a blood test that your family doctor can administer to accurately detect cancer? Will all the traditional methods (MRI scans, biopsies, endoscopies…) soon be a thing of the past?
How it works
To understand whether such tests could work, one needs to understand the way they are supposed to detect cancer from just a few drops of blood.
Every cancer is a set of cells that starts to grow uncontrollably, often because the body’s mechanisms for keeping that in check fail. Doctors typically identify these rogue cells by examining tissue samples under a microscope or by looking for space-occupying lesions in X-rays or CT scans — signs that something abnormal is growing in the body.
But when it comes to blood tests, the search is far more precise. The tests being developed around the world aim to detect the tiny genetic markers left behind by tumor cells, and our blood is full of these markers: every cell releases fragments of its DNA when it dies, and even when it communicates with other cells. Which means that tiny bits of genetic material are constantly circulating in our bloodstream. The challenge for researchers is finding the right fragments, i.e. the ones that point to cancer, and identifying which organ the tumor might be located in.
The secret weapon here is artificial intelligence. Researchers use algorithms to analyze the genetic snippets in blood samples from healthy individuals and cancer patients alike. In simple terms, the AI sifts through all the gene fragments and looks for patterns unique to cancer, even specific types of cancer like breast, prostate, or colon cancer. This allows the AI to spot recurring patterns in the genetic data, without needing to know exactly what each gene does in the context of the disease’s spread.
If it misses these markers, the test could give a false sense of security.
Researchers are also hunting for specific mutations they already understand — genetic changes that turn healthy cells into cancer cells by disabling certain regulatory mechanisms.
For a test to be accurate, it has to pick out the tumor cell markers from the thousands of other genetic snippets in the blood. And it needs to do so with an extremely high level of precision. If it misses these markers, the test could give a false sense of security; on the other hand, if it mistakes harmless genetic traces for tumor DNA, patients might end up undergoing unnecessary tests, treatments, and anxiety.
So, can the company calling itself the “Holy Grail” pull it off? Their blood test costs 9, but they promise it will detect “many of the deadliest types of cancer” before any symptoms show. The test checks for over 50 different cancers at once, including some for which there are no current screening options, such as pancreatic, esophageal, and ovarian cancers —types that are often diagnosed too late for effective treatment.
The company guarantees those who take the test “more control” — and the peace of mind that they’ve done everything they could to catch cancer early. And who wouldn’t want that?
However, the results are based largely on a study involving 6,662 people over the age of 50. The test identified a possible cancer diagnosis 92 times, but two-thirds of those patients turned out to be cancer-free. This means roughly one in every 115 test takers will get a false positive.
How the Galleri multi-cancer early detection test works, Nov. 16, 2022.
GRAIL/YouTube
Lots of marketing
So what we’re looking at here is essentially a lot of false alarms, unnecessary worries, and a fair amount of uncertainty. But here’s the thing: existing screening methods aren’t immune to this problem. For example, mammograms have a false positive rate even higher than this: nearly one in 10 women who gets a mammogram for the first time is told she has cancer, only for the diagnosis to be ruled out later.
But there are two crucial differences between conventional screening methods (like mammograms or the PSA test for prostate cancer) and these new blood tests. First, if a traditional screening test raises the suspicion of cancer, doctors can follow up with more specific, targeted tests: They can biopsy the breast or prostate or take a tissue sample from the intestines, which can provide clarity quickly.
If Galleri comes back positive, though, a much more complicated search begins. The test points doctors in the direction of which of the 50 cancers the patient might have, but it doesn’t narrow it down enough to pinpoint the exact kind of cancer or its location. The study revealed that it took an average of 162 days for patients with false alarms to get the all-clear. That’s more than five months of living with the fear of having a terminal illness, all the while undergoing up to 10 hospital visits, 16 lab tests, and five imaging procedures like CT scans and MRIs.
A testicle was mistakenly removed.
In some cases, harmless cysts or nodules were found and needed further examination. In one extreme case, even a testicle was mistakenly removed.
The second difference is more critical: mammography has been proven to reduce breast cancer mortality, as shown in large-scale studies. Women who get regular mammograms survive breast cancer longer than those who don’t, because the cancer is detected early enough to be treated effectively.
A reliable early detection test must not only find cancer but, more importantly, help reduce the mortality rate of the cancers it’s designed to detect: in a word, it has to detect tumors early enough. Right now, it is unclear whether the Galleri test can achieve that. Large-scale studies involving 140,000 subjects are underway in the UK, but the results aren’t in yet. The data so far is raising doubts: Only one in five cancers detected by the test were at an early stage. And that might not be a coincidence, as tiny tumors leave fewer genetic traces, making them harder to catch early.
A healthy dose of skepticism
Biochemist Holger Sültmann, head of the Cancer Genome Research Department at the German Cancer Research Center, is skeptical. He’s familiar with the test, the data that has been published, and the claims the company is making. “You shouldn’t be fooled. It’s all marketing,” he says. And it’s easy to see why, with early cancer detection being such a massive market.
Sültmann points out that the test doesn’t detect all types of cancer equally well. For some cancers, the results are promising, but for others — like prostate or breast cancer — the test isn’t as reliable.
He believes that blood tests targeting just one specific cancer type, rather than 50, will be far more realistic in the near future. But even these will take time.
“We have to ask ourselves: Is all this marketing ultimately misleading and even unethical?” Especially when you consider that people in the U.S. can already buy the test, even though it hasn’t been officially approved by any regulatory body.
The advantage is also its weakness.
Wolfgang Janni agrees, but he is more optimistic about it. “We’re closer to having a blood test for cancer than ever before,” he says, although adding one important caveat: “I’m talking about tests for patients who’ve already had cancer and are at risk of relapse — not tests for healthy people.”
When doctors treat cancer patients, they analyze the tumor’s genetic makeup. They figure out which mutations are present and create a unique genetic fingerprint of that tumor. After treatment, they can then search the blood for traces of these mutations. If any cancer cells survived and began to regrow, they could be detected early. The earlier this is spotted, the better the chances for successful treatment.
The advantage of this method, says Janni, is also its weakness. AI can analyze huge amounts of data and look for general patterns of what breast cancer might look like genetically. But not all breast cancers are genetically identical. The patterns are similar, but they differ from patient to patient. So, looking for a general “template” instead of a precise match makes the test less accurate.
Evolution, not revolution
This year, the Survive study was launched, a large survey led by the Ulm clinic, where Janni works. It is intended to show whether a blood test for breast cancer follow-up can achieve the crucial goal of prolonging survival after a cancer diagnosis. All patients receive the follow-up examinations recommended to date after their treatment. For some patients, the doctors also use blood tests to look for very specific traces of cancer cells.
“I am cautiously positive that we could be successful with this,” says Janni. Previous studies have shown that the genetic traces of recurring tumors can be detected in blood samples with relative success.
It may not be what you initially imagined when you thought of a blood test for cancer. It is not a “Holy Grail” test that you take in five minutes at your GP’s office and detect all tumors early. Janni and Sültmann say that such tests will not be available in any general practitioner in the foreseeable future. However, tests for cancer patients that detect relapses early and with little effort could be a true revolution. For patients and doctors like Janni, this would constitute a milestone — and help save countless lives.
