18 September 2025
What if a simple blood test could detect cancer long before symptoms appear? That’s the challenge taken up by bioinformatics professor and researcher, and liquid biopsy specialist, Alexandre Pellan Cheng. His research team is working on transforming cancer diagnosis into something as simple as a routine blood test, in order to improve early detection, patients’ quality of life and, ultimately, their chances of survival.
A Less Invasive Approach Than Conventional Biopsies
Traditionally, cancer diagnosis involves a hospital biopsy. Using a needle, a healthcare professional removes a fragment of suspicious tissue—a painful, logistically demanding procedure requiring specialized medical expertise. Alexandre Pellan Cheng is proposing another approach: liquid biopsies to analyze traces of DNA released into the blood by cancer cells.
When a cell dies, it releases fragments of DNA into the bloodstream. This is a normal part of human metabolism. Similarly, tumour cells send their own “detritus” into the bloodstream, including the DNA mutations found in cancer. Pellan Cheng’s challenge is to detect these tiny amounts of tumour DNA in a sea of normal DNA.
Seeing Cancer Before It Shows
The smaller the tumour, the less DNA it releases. Yet it is precisely at this early stage (stage 0 or 1) that a diagnosis considerably increases the chances of remission. Likewise, when a treated cancer seems to have disappeared, it may reappear discreetly, with almost undetectable levels of tumour DNA. It’s in these critical situations that Professor Pellan Cheng wants to intervene, by spotting the weak signals, the minute mutations that indicate the presence or return of cancer.
But how to distinguish tumour-specific mutations from an individual’s own genetic DNA? One common approach is to sequence the tumour’s DNA after surgery, then compare this profile with the patient’s own blood or healthy tissue. The difference between the two reveals what is known as tumour mutations. First-gen tests based on this principle are already in use in some clinics. Alexandre Pellan Cheng’s laboratory, on the other hand, is developing next-gen tests capable of taking the analysis much further.
When Tumours Won’t Talk
Sometimes, tumours cannot be sequenced. This is especially the case when a patient receives neoadjuvant treatments, such as chemotherapy, prior to surgery: the tumour then shrinks in size. While this is good news from a medical standpoint, it also makes laboratory analyses difficult. This is why another line of research is needed: detection without direct information from the tumour, based solely on circulating DNA in the blood.
This strategy relies on the precise control of background noise: in a sample, distinguishing between a machine error and a genuine mutation is crucial. “We’re often within the margin of error,” explains Alexandre Pellan Cheng. Therefore, his team is working on lowering these margins to correct sequencing errors, and designing algorithms capable of making this distinction. To this end, they rely heavily on artificial intelligence.
Mountains of Data to Analyze
A simple blood test generates terabytes of data. For each patient, billions of DNA fragments must be analyzed, filtered, corrected and compared. A large part of the laboratory’s work is therefore based on developing bioinformatics tools to manage this mass of data. It also involves meticulous sample preparation: purifying often degraded DNA, making it “machine-readable,” and correcting systematic errors even before computer processing.
The laboratory doesn’t just analyze data: it also produces it. Equipped with state-of-the-art material, Alexandre Pellan Cheng and his team can generate their own databases, essential for training artificial intelligence models in recognizing cancer signatures.
A More Personalized, Humane Approach to Medicine
The ultimate goal? Integrating these tools into clinical studies, first retrospectively, then prospectively, in order to prove their real value in the medical context. The idea is to change medical practice, by giving clinicians a less invasive, more frequent and more informative method than traditional approaches.
For example, a series of liquid biopsies taken before, during and after treatment could provide real time indications of therapy efficacy. Are tumour DNA levels decreasing? Should treatment be adjusted, or should the patient be spared unnecessary side effects? Liquid biopsies thus become a personalized monitoring tool, for patient comfort as much as recovery.
The prospects are endless: early detection, post-operative monitoring, adaptation of treatments… Although Professor Pellan Cheng’s research is currently focused on cancer, the technologies he is developing could also benefit other fields, such as organ transplant monitoring.
With sequencing machines are becoming faster and more accessible, what was once science fiction is gradually becoming a clinical reality. Thanks to researchers like Alexandre Pellan Cheng, tomorrow’s medicine may well be contained in a single vial of blood.
