A simple blood test could diagnose at a very early stage one of the 8 common cancers with a 70% success rate. This is not the first test of its kind, nor is it a universal test, but its preliminary results are interesting.
This is not the first time this type of test has been proposed by researchers, but the originality of the CancerSEEK test is that it looks for both DNA segments and proteins of 8 cancers in the blood. It would allow the diagnosis of ovarian, liver, stomach, pancreas, esophagus, colon and rectum, lung or breast cancers at a very early stage. All this in a single blood test such as can be done in any biology laboratory and for a reasonable price.
Liquid biopsy at your fingertips
Research in recent years has shown that fragments of genetic material, or proteins, of cancer cells can circulate in the blood long before the cancer itself is clinically apparent. These DNA or protein fragments are released by cells as they die, which happens for all cells.
The search for these circulating DNA fragments or proteins in the blood is called "liquid biopsy", as opposed to the normal biopsy of surgically puncturing or removing a small piece of tumor for analysis.
This approach is particularly interesting from the perspective of treatment because it can intervene at a very early stage, where the tumor is only a few millimeters. However, the earlier the diagnosis of cancer, the fewer lesions on the surrounding tissues, the higher the chances of recovery and the less cumbersome and difficult the treatment.
The CancerSEEK test evaluates the levels of eight specific proteins and the presence of 16 mutated gene segments specific to these 8 types of cancer circulating in the blood and gives a positive result in 70% of the cases in a study of more than 1000 cancerous patients. in the journal Science.
Although this test does not detect all the possible cancers of the adult, it is indeed not a universal test, it would allow to detect at once the eight most common cancers. These account for more than 60% of cancer deaths in the United States, as in France. In addition, 5 of the cancers covered by this new test are currently not subject to any organized screening.
In this study, the combinations of proteins and genetic abnormalities sought make this test very specific (more than 99 percent) for the 8 cancers concerned. This very high specificity is essential in medicine because false positive results could be responsible for many unnecessary and expensive examinations, or even exploratory surgical operations to confirm the presence of a cancer suspected to be wrong.
The specificity of this test was evaluated on samples taken from 812 people who have no cancer (healthy controls) and produced only seven false positive results. The results are published online by the journal Science on January 18, 2018.
A sensitivity that varies according to the cancers
To evaluate its ability to detect small, early cancers, where there are not yet many malignant cells, and not to miss, which doctors call "sensitivity", the test was evaluated in 1 005 patients non-metastatic stage I to III cancers of the ovary, liver, stomach, pancreas, esophagus, colon and rectum, lung or breast.
The overall sensitivity of the test, ie its ability to find cancer in patients, is 70% and ranges from a high of 98% for ovarian cancer to a low of 33% for breast cancer (where there are other screening tests such as mammography).
For the five cancers that currently do not have screening tests (ovarian, liver, stomach, pancreas and esophageal cancers), sensitivity ranges from 69% to 98%, which constitutes real progress.
A pragmatic approach
The researchers wanted to limit themselves to a small number of mutated genes and cancer proteins to keep a certain simplicity, simplicity that is not accompanied by a result at a discount.
The researchers first explored several hundred genes and 40 protein markers of the cancers concerned, then reducing the number to only 16 gene segments and eight proteins. In the calculations made, this is what appears to be the optimal number. Indeed, researchers have found that by multiplying the analysis of circulating DNA beyond a certain number, there was rather a degradation of the quality of the results.
As it is, the test should improve the treatment of cancers of the ovary, liver, stomach, pancreas, esophagus, colon and rectum, lung, and to a lesser extent , breast. If confirmed, its effectiveness will help treat these cancers at a very early stage in people at risk, a stage largely accessible to current treatments.
Many of the most promising cancer treatments available to us today only really benefit a small minority of patients by curing them. These are major advances, but the too late diagnosis of cancer in many patients means that these treatments do not allow them to be cured or put into prolonged remission. It's a loss of luck for them.
If we want to progress in the treatment of cancers, we must diagnose them earlier and therefore start looking at screening in a more pragmatic way. Being well aware that no test, whatever it is, will allow to detect all the cancers. There will be no universal test!
The considerable advance of the liquid biopsy
"The use of a combination of selected blood biomarkers for the early detection of cancer can potentially change the way we screen for cancer, and it is based on the same logic of using combinations of drugs to treat cancers," explains Nickolas Papadopoulos, senior author and oncologist at Johns Hopkins University.
"The mutations carried by the circulating DNA of the tumors can be markers highly specific of a cancer and, to capitalize on their great specificity, we sought to develop a test based on a small number of biomarkers, but robust, and capable detect at least one mutation in the vast majority of cancers. It is indeed essential to keep a small assortment of mutations to minimize false positive results and to keep these screening tests affordable. "
Researchers point out that this molecular test is only for cancer screening and, therefore, is different from other tests performed on circulating DNA, which rely on the analysis of a large number of genes associated with a given cancer. to identify potential therapeutic targets. Targeted treatments directed against a mutation are in fact the most effective current treatments.
Contribution of artificial intelligence
"A novelty of our classification method is that it combines the probability of analyzing different DNA mutations in parallel with measuring the levels of several proteins in order to make a final synthesis," says Cristian Tomasetti, Associate Professor of Physiology. oncology and biostatistics.
"Another new aspect of our approach is that it uses machine intelligence AI to allow the test to accurately determine the location of a tumor on a small number of sites in the body in 83% of the world. patients ".
To identify the mutations and target proteins of their test, the research team based their research on the oldest database in the United States: data collected for more than three decades of research on cancer genetics and compiled at the Ludwig Center of Johns Hopkins. This is where the first genetic models of cancer were created.
A test close to the routine
This test, called CancerSEEK, opens the possibility of doing with a simple blood test screening for cancers of the ovary, liver, stomach, pancreas, esophagus, colon and rectum, lung or breast, which are the most common.
It can, in principle, be performed in any city biology laboratory, along with other routine blood tests, prescribed by the general practitioner. It should be marketed for less than 500 dollars (400 Euros) according to its promoters, making it an inexpensive and simpler test than mammography in breast cancer or the search for blood in the stool for cancer colorectal.
Early detection offers the potential to reduce cancer treatments and improve outcomes for patients. In the best case, the cancers would be detected early enough to be cured by surgery alone, but even early cancers that are not curable by surgery alone, will respond to chemotherapy we have most often.
Validation studies, more important of this test, are currently in progress. We will see if it keeps its promises because it is generally at this stage of the validation that the previous tests showed their limits.