Optical chip identifies lung cancer from blood without genetic sequencing

A chemical tag read by a light scanner could cut the cost and laboratory equipment needed to detect tumour DNA, according to a 103-person validation study.

When cancer cells grow and die, they shed tiny pieces of their genetic material into the bloodstream.¹ Finding these circulating fragments offers a way to spot malignant tumours without taking a physical biopsy from a patient. For years, however, the practical hurdle has been finding those rare signals. The bloodstream carries an ocean of ordinary genetic material released by healthy cells, and tumour fragments make up only a tiny fraction of the total volume.¹ Picking them out has typically meant running deep genetic sequencing, an exacting laboratory process that relies on complex chemical reagents, expensive machines, and specialized computational analysis to spell out individual molecular codes.¹

That requirement has largely confined liquid biopsy, the clinical method of identifying cancer using a simple blood draw, to well-funded urban research hospitals. Lung cancer accounts for more cancer deaths globally than any other malignancy, and detecting it early offers patients the best chance of survival.¹ Yet the vast majority of the world lacks access to the specialized infrastructure needed to sequence circulating tumour DNA routinely, creating a persistent barrier between an effective diagnostic concept and daily medical practice.¹

A study published on August 10, 2026, in the journal npj Precision Oncology suggests that reading the sequence may not be necessary.¹ A research team led by Yuval Ebenstein, a professor at Tel Aviv University who holds appointments in the School of Chemistry, the Department of Biomedical Engineering, and the Zimin Institute, has built an alternative detection system.¹ Working alongside researchers from JaxBio Technologies, Bnai Zion Medical Center, and Sheba Medical Center, the group designed an optical chip that identifies lung cancer from circulating blood samples using light patterns rather than genetic sequencing.¹

How can light identify cancer in a blood sample?

The system works by detecting distinct chemical modifications on circulating DNA molecules instead of reading every genetic letter along their strands.¹ In human biology, cells constantly attach small chemical tags to their DNA to regulate gene activity. Cancer cells carry different patterns of these chemical tags compared to healthy tissue, effectively leaving a recognizable surface fingerprint on the fragments they release into the blood.¹

To read that pattern, technicians extract free-floating DNA molecules directly from a patient blood sample.¹ They attach light-emitting chemical markers to the fragments, bind the labeled DNA onto a specialized physical chip, and place the chip into an ordinary optical scanner.¹ The scanner measures how light reflects off the chemical markers across specific zones on the surface, producing an optical reading that distinguishes cancer-derived material from background DNA shed by normal tissues.¹

Bypassing sequencing removes the most demanding components of the diagnostic pipeline. The authors report that the complete procedure takes two to three days to finish, with an estimated processing cost of roughly $60 per blood sample.¹ Because optical scanners are already common fixtures in standard clinical facilities, the test could theoretically run in general hospitals without the dedicated sequencing cores that conventional liquid biopsies demand.¹

“Our goal is to make blood tests for cancer diagnosis more accessible, simpler, and less expensive without compromising accuracy,” Ebenstein said in a statement accompanying the research.¹ “We have developed a new approach that does not require genetic sequencing but instead identifies the tumor’s chemical ‘fingerprint’ with light, using a technology that can be implemented in standard clinical laboratories.”¹

What did the clinical trial reveal about the test?

The experimental platform achieved an overall diagnostic accuracy above 90 percent when tested on blood samples taken from 103 study participants.¹ The trial cohort included 51 individuals diagnosed with lung cancer alongside 52 healthy control subjects.¹ The researchers isolated a 170-region genomic signature during an initial training phase, then evaluated its diagnostic performance against a separate held-out validation group using blinded analysis, ensuring that the laboratory team evaluated the final samples without knowing patient diagnoses in advance.¹

For patients with stage 2 to stage 4 lung cancer, the test demonstrated a sensitivity of 93.1 percent, meaning it detected disease in roughly 93 out of every 100 cancer cases within those stages.¹ Its specificity among the healthy control group stood at 90.3 percent, correctly clearing approximately 90 out of every 100 disease-free individuals.¹ Beyond broad detection, the chemical signatures on the chip separated adenocarcinoma from squamous cell carcinoma, two major subtypes of non-small cell lung cancer that demand different clinical therapies and normally require surgical tissue biopsies to tell apart.¹

The authors also recorded an initial result tracking patient responses to ongoing cancer therapy.¹ In patients who responded favorably to clinical treatment, their optical DNA profiles shifted progressively back toward patterns observed in healthy control participants.¹ By contrast, individuals who did not respond to treatment showed no meaningful change in their optical test results over time, suggesting that the blood test could potentially help doctors monitor whether a selected medication is having an effect.¹

What are the limits of the current evidence?

The trial represents a preliminary feasibility study rather than clinical proof that the test is ready for widespread medical use.¹ The 93.1 percent sensitivity figure applies strictly to stage 2, stage 3, and stage 4 lung cancer, entirely omitting stage 1 disease.¹ Early-stage detection is the setting where diagnostic testing provides the greatest clinical benefit, but stage 1 tumours shed far smaller amounts of DNA into circulation, and the paper reports no data on how reliably the optical scanner detects that lower volume.¹

The composition of the control group also creates a simpler diagnostic environment than real-world lung cancer screening. The study compared 51 known cancer patients directly against 52 healthy individuals.¹ In a real-world clinical clinic, a screening tool must evaluate people who smoke or suffer from unrelated respiratory ailments, including chronic obstructive pulmonary disease, active pulmonary infections, and benign tissue nodules, all of which can alter biological markers in the blood.¹

The test specificity of 90.3 percent presents a major mathematical hurdle for population screening.¹ That figure implies that roughly one out of every 10 healthy individuals would receive a positive result.¹ Because lung cancer has a low prevalence across the broad population, testing thousands of individuals would produce far more false alarms than true cancer diagnoses, exposing disease-free patients to anxiety and invasive diagnostic follow-ups.¹

Evaluating 103 participants leaves open major questions about how patient age, smoking intensity, ethnic background, and coexisting medical conditions affect the optical markers.¹ The research announcement also omitted a competing-interests disclosure, which is essential context given that JaxBio Technologies is a commercial enterprise whose staff worked on the investigation.¹ For patients weighing cancer risks today, decisions about screening require consultation with a personal physician, while the optical chip remains an experimental technology awaiting multi-thousand-patient clinical trials.¹

Save this story

The record

37 sentences trace to 1 source.

All 1Documents 1Records 0Reporting 0People 0

1. 1 S scientificinquirer Document · scientificinquirer.com · 8 Nov 2026 A Lung Cancer Blood Test That Reads DNA With Light Instead of Sequencing It See the source