Blood tests have proven to be a promising tool for detecting and monitoring cancer. Researchers at Chalmers University of Technology and the University of Gothenburg in Sweden have developed a new method that allows them to analyze samples containing just 5 percent cancer DNA in blood, compared to the 15 to 20 percent currently required. This method could lead to better cancer treatment and improved monitoring of tumor progression.
The technique of using blood tests to analyze changes in tumor DNA is being studied in several clinical trials around the world. Current analytical methods work well when the amount of cancer DNA is relatively high, about 15 to 20 percent of the total DNA in the blood. However, levels of cancer DNA are often much lower than this. This may mean that the sample quality is not sufficient for detailed analysis.
“We wanted to develop a method that would be particularly effective in difficult cases where there is very little cancer DNA in the blood and a lot of what we would consider noise, i.e. mainly healthy DNA. So it works as we expected,” says Lotta Eriksson, a doctoral student in the Department of Mathematical Sciences at Chalmers and the University of Gothenburg.
Better monitoring and individually tailored treatment
Blood-based methods being tested in clinical trials are often used to determine whether cancer can be detected at all. Obtaining more detailed images is difficult, due in part to high costs and poor sample quality.
A new method, BayesCNA, can extract information previously hidden in low-quality samples and provide more detailed information about tumor composition. This helps us better understand how a patient’s cancer changes over time.
“When a treatment is effective, the amount of cancer DNA in the blood is significantly reduced, making it even more difficult to detect the cancer and to monitor how it changes. To get a clearer picture of how patients respond to treatment, it is important to be able to analyze samples containing low levels of cancer DNA,” said Esther Lakatos, assistant professor at Chalmers and the School of Mathematical Sciences at the University of Gothenburg.
Currently, tissue samples from the tumor itself are required to obtain detailed information about its composition. Being able to monitor tumor progression using blood tests could significantly improve the care of cancer patients.
“Patients may have one or two surgeries, but blood tests may be done just a few weeks apart during treatment. If we can get information about changes in the tumor from samples, we can better monitor progression and see what happens between treatment sessions. This helps doctors make more informed decisions, such as adjusting treatment to the composition of the tumor,” says Eszter Lakatos.
Statistical methods to amplify weak signals
The method was developed to analyze data known as low-pass whole-genome sequencing, a technique that provides a general overview of DNA structure. Although this technique has significant economic benefits, it provides limited information due to poor data quality.
“You can compare it to skimming a book instead of reading it properly. You get an overview of the DNA structure, but not a detailed picture,” says Estelle Lakatos.
New analysis methods use statistical algorithms to amplify the very weak signals present in this type of sample.
Machine learning is currently being used to solve many problems, but we were the first to try them out. But to our surprise, it turned out that classical statistics worked better in this case, which was especially pleasing to us mathematicians and statisticians. ”
Lotta Eriksson, University of Gothenburg
Aiming for clinical trials
The next step is to analyze the information about tumor composition that this method provides. Researchers are working hard to develop further ways to identify hidden characteristics of cancer that influence how patients respond to treatment.
“If we can demonstrate that this information is useful, we hope to see more collaboration within the research community and wider adoption of our method.” In the long term, we hope that the method we develop will be used in clinical trials and, with any luck, change the care of cancer patients, says Estelle Lakatos.
sauce:
Chalmers University of Technology
Reference magazines:
Eriksson, L., & Lakatos, E. (2026) Sensitive detection of copy number changes in low-pass liquid biopsy sequence data. Briefing session on bioinformatics. DOI: 10.1093/bib/bbag111. https://academic.oup.com/bib/article/27/2/bbag111/8524997.
