Research by LSMU and Taiwanese Scientists Opens New Opportunities for Early Diagnosis of Pancreatic Cancer

Pancreatic cancer is one of the most aggressive oncological diseases – out of 100 patients diagnosed with it, approximately 20-30 are still alive after one year, and only about 8-10 after five years. One of the main reasons is that the disease is most often detected too late.

This year, the results of a project conducted by scientists from National Taiwan University and the Institute for Digestive Research at Lithuanian University of Health Sciences (LSMU) were published in the prestigious journal Nature Communications, promising a breakthrough in the early diagnosis of pancreatic cancer.

The authors of the study in the initial phase examined 478 patients with pancreatic cancer and 424 individuals at high risk of developing the disease. The results showed that by using only three clinical-biochemical parameters – the patient’s age, the concentrations of the biomarkers CA19-9 and activin A – along with an analysis of the blood serum metabolomic profile (including small-molecule metabolites and lipoproteins), an artificial intelligence (AI) model was able to detect the disease almost without error, achieving an AUC score of 0.99. In the Lithuanian patient cohort used to validate the molecular tool, the model also maintained very high diagnostic accuracy (0.93).

AUC indicates how well a diagnostic test or AI model distinguishes between diseased and non-diseased individuals – the closer the number is to 1, the more reliable the test. In other words, the system correctly identifies the disease in nearly 99 out of 100 cases.

“The implementation of blood serum metabolomic profiling in clinical practice, combined with AI analysis, would provide a real opportunity in the future to save more lives by detecting pancreatic cancer at a stage when treatment can still be effective” – sayes the project leader in Lithuania Professor of Gastroenterology Juozas Kupčinskas.

Prof. J. Kupčinskas emphasizes that the project is being further continued through collaboration between LSMU and Taiwanese scientists, integrating complex protein profiling (proteomics) data in order to further enhance the effectiveness of the diagnostic model.

The project is funded by the Research Council of Lithuania and the National Science and Technology Council of Taiwan under a bilateral cooperation program. More information about the study here.