Pancreatic ductal adenocarcinoma (PDAC) remains one of the deadliest cancers known. Unlike many other cancers where survival rates have steadily improved, pancreatic cancer mortality remains almost equal to its incidence — meaning most people diagnosed with it will die from it.
Surgical resection is the only potentially curative treatment, but 80% of patients are diagnosed when the cancer has already spread beyond the point where surgery is feasible. By the time symptoms appear, the disease is usually advanced.
A catastrophic increase in pancreatic cancer cases is expected in coming years, driven by rising rates of obesity, diabetes, and aging populations. Without better early detection strategies, the human cost will escalate dramatically.
Identifying who is truly at high risk for pancreatic cancer is the first step toward meaningful early detection. Established risk factors include new-onset diabetes (particularly after age 50), chronic pancreatitis, family history, known genetic mutations (BRCA1/2, ATM, PALB2), and Lynch syndrome.
Hereditary pancreatic cancer syndromes — particularly familial pancreatic cancer and Peutz-Jeghers syndrome — confer lifetime risks high enough to justify formal surveillance programs with MRI and EUS starting at age 50 or 10 years before the youngest affected relative.
New-onset diabetes represents a unique early detection opportunity because it can precede pancreatic cancer diagnosis by one to two years and is associated with a 5-8 fold increased cancer risk.
Liquid biopsy — detecting cancer-derived material in blood or other body fluids — is one of the most actively researched areas in early cancer detection. For pancreatic cancer, the main targets include circulating tumor DNA (ctDNA), circulating tumor cells (CTCs), exosomes, and microRNAs shed by cancer cells.
CA 19-9, the current blood marker used for pancreatic cancer, is poorly suited for early detection because it is often normal in early-stage disease and elevated in benign conditions like obstructive jaundice. Novel markers including methylation patterns and multi-analyte blood tests are under investigation.
Multi-marker blood tests combining genetic mutations in ctDNA, protein biomarkers, and epigenetic markers show more promise than any single marker alone, but are not yet validated for clinical screening use.
CT scan is the standard first-line imaging tool when pancreatic cancer is suspected, but it misses many small tumors. MRI with diffusion-weighted imaging offers superior soft tissue contrast and is now preferred for characterizing cystic lesions and surveillance of high-risk individuals.
Endoscopic ultrasound (EUS) remains the most sensitive modality for detecting small pancreatic lesions, particularly those below 2 cm. When combined with fine needle aspiration, it provides both imaging characterization and tissue diagnosis.
Artificial intelligence and machine learning are increasingly applied across all imaging modalities used in pancreatic cancer, from detecting subtle changes on CT to automatically classifying EUS images and analyzing liquid biopsy data.
Currently, formal surveillance is only recommended for populations with very high genetic risk — carriers of pathogenic variants in BRCA2, ATM, or PALB2, and those with familial pancreatic cancer. For these individuals, annual MRI with or without EUS is the standard of care.
For the broader high-risk population — including new-onset diabetes patients and those with chronic pancreatitis — there are no broadly adopted screening protocols yet. Clinical trials are underway to determine whether screening these groups is feasible and cost-effective.
Even in high-risk groups, pancreatic cancer is uncommon enough that a screening test must have very high specificity to avoid a flood of false positives leading to unnecessary invasive follow-up.
Despite some promising developments in liquid biopsy, imaging, and AI, there is no validated early detection strategy for pancreatic cancer in the general or even moderate-risk population as of 2022. The situation is truly discouraging given the scale of the human cost.
The authors emphasize that rising incidence projections make this urgency even more acute. Without better early detection, a larger number of patients will be diagnosed too late for surgical cure in the coming years.
The review serves as a call to action for greater investment in pancreatic cancer research — particularly multi-center prospective studies of emerging detection tools and population-level risk stratification models.