Combined Blood and Imaging Data Help Identify More Lung Cancer Cases

Lung cancer diagnosis often depends on tissue sampling, but biopsy can be time-consuming, burdensome, and inconclusive. Some patients are too frail for the procedure, while others may need repeated sampling when results do not provide clarity. Patients with suspected lung cancer typically undergo computed tomography (CT), which may offer additional diagnostic information. New findings demonstrate that combining blood-based tumor markers with CT scan data may improve recognition of lung cancer.

Catharina Hospital (Eindhoven, Netherlands) and Eindhoven University of Technology (TU/e) are evaluating a combined diagnostic approach that integrates blood tests with CT imaging information. The blood tests measure tumor markers, which are substances in the blood that can indicate cancer. The CT scan assessment includes radiologic evaluation of the size, shape, and other characteristics of suspicious lung lesions.

The researchers used data from 346 patients divided into two study groups. In the first group, blood tests alone identified approximately 62% of patients with lung cancer. When the radiologist’s CT assessment was added to the blood test results in a computer model, detection increased to approximately 78%.

The team also evaluated whether artificial intelligence (AI) could automatically analyze CT scans. A deep learning model was used to recognize patterns in medical images and combine that information with blood test results. The combination of blood tests and AI recognized approximately 76% of lung cancer cases, while the same approach reached 73% in a second independent patient group.

The research was presented in late October at the European Society for Medical Oncology (ESMO) Congress in Madrid by PhD candidate Hao Cao, who received an ESMO Merit Award for the work. The method is not intended to replace biopsy at this stage, and a negative result cannot rule out lung cancer. It also cannot yet determine the specific type of lung cancer, which remains important for treatment selection. Further validation in routine clinical practice is planned before the approach can be used to support medical decision-making.

“Using blood tests alone, we were able to identify around 62 per cent of patients with lung cancer. That is useful information, but it also means we still miss a proportion of patients. So we started looking at whether we could improve that figure by adding other information,” said Hao Cao, who works at the General Clinical Laboratory of Catharina Hospital.

“That is a clear improvement. The blood tells us something, but the images provide other information. By using them together, we get a more complete picture,” said Cao.

Related Links:
Catharina Ziekenhuis
Eindhoven University of Technology