Cancer can leave traces in the bloodstream long before doctors can obtain a piece of the tumor itself.
Researchers in Israel are developing a blood test that tries to read those traces using light rather than conventional DNA sequencing.
Early results suggest the approach may help identify lung cancer and possibly track how tumors respond to treatment.
The technology was developed by a team led by Professor Yuval Ebenstein at Tel Aviv University, working with JaxBio Technologies, Bnai Zion Medical Center and Sheba Medical Center. Their study was published in the journal npj Precision Oncology. It included 51 people with lung cancer and 52 healthy participants.
Lung cancer remains the leading cause of cancer death worldwide. One major reason is timing: tumors are often found only after the disease has become harder to treat. A dependable test that helps detect cancer earlier could therefore have an important effect on survival.
CT scanning is currently one of the main tools used to screen people at high risk. These scans can reveal very small lung abnormalities, but an abnormal spot is not necessarily cancer. Inflammation, scars and other harmless changes can sometimes look suspicious, creating a need for further tests.
A blood test could potentially add another piece of evidence. Blood contains tiny fragments of DNA released by cells throughout the body, known as cell-free DNA. Cancer cells also release DNA, meaning a blood sample can contain small molecular clues from a tumor.
The difficulty is finding those clues among DNA from normal cells. Some modern liquid biopsies solve this problem by sequencing DNA and searching for mutations associated with cancer. While powerful, this process can require expensive equipment, detailed laboratory work and complex computer analysis.
The Tel Aviv team instead examined chemical changes attached to DNA. Cells can add chemical tags to DNA that influence how genes are used without changing the underlying genetic code. Cancer can alter these patterns, creating a molecular signature that differs from the pattern in healthy cells.
The researchers extracted cell-free DNA from blood and added fluorescent markers that produce light. They then attached the DNA to a specially designed chip and scanned it optically. The pattern of light revealed information about chemical marks in selected areas of the DNA.
Using part of their data, the researchers created a cancer signature involving 170 genomic regions. They then evaluated the system on a separate set of samples in a blinded test, meaning the researchers assessing the samples did not know which participants had cancer. This type of separation is important because it reduces the chance that a model simply memorizes the samples used to build it.
Among patients with stage 2, 3 or 4 lung cancer, the test reached a sensitivity of 93.1%. That means it correctly detected a high proportion of people who had cancer in the study. Its specificity was 90.3%, meaning it also correctly classified most healthy people as not having cancer.
Those numbers are encouraging, but they should be interpreted carefully. A diagnostic test needs to work not only in a small research group but also among thousands of people with a wide range of health conditions. Performance can change when a test moves from a carefully selected study into everyday medical practice.
The researchers found another interesting feature in the DNA patterns. The test could distinguish adenocarcinoma from squamous cell carcinoma, the two major types of non-small-cell lung cancer. Knowing the cancer type is important because tumor biology helps guide treatment.
The scientists also followed chemical DNA patterns in some patients during treatment. When treatment was working according to medical imaging, the blood pattern shifted toward the pattern seen in people without cancer. When patients were not responding, researchers did not see the same shift.
This raises the possibility that the test might eventually do more than detect cancer. Doctors could potentially use repeated blood samples to help monitor whether a tumor is responding to therapy. The researchers stress that this part of the work is preliminary and needs much more evidence.
The method may have practical advantages if future trials confirm its performance. The team estimates that testing currently costs around $60 per sample and takes two to three days. Avoiding DNA sequencing could make the technology simpler and more affordable for laboratories that do not have advanced sequencing systems.
However, the most important unanswered question is how well it detects very early cancer. The reported 93.1% sensitivity applies to stage 2–4 disease, while screening is especially valuable when cancer is still at stage 1. A future screening test would also need to distinguish cancer from infections, chronic lung disease and other conditions that may alter DNA patterns.
The study’s design has useful strengths, including blinded validation and a biological signal that may also reflect treatment response. But the total sample of 103 participants is far too small to establish a new clinical test. Larger studies will need to include people of different ages and backgrounds as well as patients with non-cancerous lung abnormalities.
It is also unlikely that one blood test would immediately replace CT scans. A more realistic future role could be using blood and imaging together, with each test providing different information. If a blood result could help doctors decide which suspicious CT findings deserve closer investigation, it might reduce some unnecessary procedures while still finding serious cancers.
The research therefore represents an early but interesting step rather than a finished diagnostic tool. Detecting a tumor’s chemical fingerprint with light could make liquid biopsy technology cheaper and simpler if the findings hold up in larger trials. The real test will be whether this approach can accurately find early lung cancer in the people who would actually receive screening.
Source: Tel Aviv University.
