A new pancreatic cancer prediction model, utilizing routine blood markers, may reduce unnecessary biopsies by up to 6%. Researchers have validated this model, which assesses early-stage pancreatic cancer risk, and found it outperforms individual biomarkers. Decision curve analysis indicates a 6% reduction in unnecessary biopsies without missing early-stage cases. This development showcases efforts to employ advanced technology for early pancreatic cancer detection, complementing previous AI models designed to detect the disease before diagnosis, offering hope for improved outcomes in a typically aggressive cancer.
A recently developed pancreatic cancer prediction model, which utilizes common blood markers, has the potential to enhance early detection and reduce unnecessary biopsies by as much as six percent, according to a study published in JAMA Network Open.
The researchers have crafted and externally validated this prediction model, aiming to identify early-stage pancreatic cancer by examining routinely collected blood biomarkers. The study underscores the importance of precise risk prediction for pancreatic cancer, as it can expedite early detection and spare low-risk patients from unnecessary diagnostic procedures, such as biopsies.
Currently, diagnosing pancreatic cancer relies on a lengthy and invasive combination of methods, including clinical symptoms, serum levels of carbohydrate antigen 19-9 (CA19-9), radiological findings, and confirmation through fine-needle aspiration or brush cytology. Often, these tests yield inconclusive results, necessitating further confirmation through biopsy or tumor resection.
Due to the complexity and invasiveness of these methods, the researchers explored a more efficient approach to prevent unnecessary diagnostics and expedite treatment for pancreatic cancer patients. Routine blood-based biomarkers like CA19-9 and bilirubin levels, commonly measured in individuals screened for pancreatic cancer, may offer valuable insights into cancer risk.
To test this hypothesis, the research team developed a risk prediction model that leverages these biomarkers to distinguish between early-stage pancreatic cancer and benign periampullary diseases.
The researchers gathered data from adult patients with pancreatic cancer or benign periampullary disease treated between 2014 and 2022 at four academic hospitals in Italy, the Netherlands, and the United Kingdom. They evaluated serum levels of CA19-9 and bilirubin from each patient at the time of diagnosis and before the commencement of medical intervention.
The study included 249 patients in the development cohort and 296 in the validation cohort. They measured the model’s performance in terms of discrimination, using the area under the curve (AUC), and compared it to the performance of individual biomarkers.
During external validation, the model achieved an AUC of 0.89 when distinguishing between early-stage pancreatic cancer and benign periampullary diseases, surpassing the performance of both CA19-9 and bilirubin. In a subset of patients without elevated tumor marker levels, the model achieved an AUC of 0.84.
Moreover, at a risk threshold of 30 percent, decision curve analysis revealed that utilizing the prediction model to guide biopsies could reduce unnecessary biopsies by six percent, without missing any patients with early-stage pancreatic cancer.
Overall, these findings suggest that the model has the potential to assess the additional diagnostic and clinical value of new biomarkers while avoiding potentially unnecessary invasive diagnostic procedures for low-risk patients. This research highlights recent advances in using advanced technologies to predict pancreatic cancer risk.
In May, researchers from Harvard Medical School and the University of Copenhagen developed an artificial intelligence (AI) model capable of detecting pancreatic cancer up to three years before diagnosis, using routinely collected clinical data. This model has the potential to expand population-based screening for pancreatic cancer and improve the detection and treatment of this aggressive and deadly disease.
