A blood-based microRNA (miRNA) assay that integrates with carbohydrate antigen 19-9 (CA19-9) to form a composite score called PANXEON demonstrated improved sensitivity for detecting early-stage pancreatic ductal adenocarcinoma (PDAC), according to findings from an international, multicenter, observational study (NCT06388967) published in Nature Medicine.
What were the key findings from the PANXEON validation study?
The combined cell-free and exosomal miRNA signature achieved an area under the receiver operating characteristic curve (AUROC) of 98.4% for all-stage and stage I to II PDAC in the training cohort, with comparable replicability in the independent validation cohort, which had an AUROC of 96.9% for all-stage PDAC and 97.3% for stage I to II PDAC. In an external, independent testing cohort composed of high-risk patients, the miRNA signature achieved an AUROC of 88.6% for distinguishing PDAC or high-grade dysplasia (HGD) from all controls, with a sensitivity of 83.8% for early-stage PDAC while showing minimal cross-reactivity with other gastrointestinal cancers.
When combined with CA19-9 into the PANXEON composite score, sensitivity for stage I to II PDAC increased to 86.8%, with false-positive rates of 3.2% in low-risk controls and 15.6% in high-risk controls in the testing cohort. PANXEON also demonstrated potential for detecting HGD in 67 individuals who underwent surgery for clinically high-risk pancreatic cysts, correctly identifying 64.3% (n = 9/14) of pathologically confirmed HGD cases while remaining negative in 75.5% (n = 40/53) of cysts without HGD.
How was the PANXEON study designed?
Investigators conducted a population-based, observational, prospective biomarker study that initially assessed 1785 plasma samples for eligibility; 28 were excluded because of suboptimal RNA quality, leaving 1757 plasma samples from 1649 unique patients recruited across 12 institutions in the US, Japan, Italy, and South Korea. Participants were divided into 5 cohorts: a training cohort (n = 707), an independent validation cohort (n = 305), an external and independent testing cohort (n = 440), a pretreatment/posttreatment cohort (n = 137), and a cross-reactivity cohort (n = 168).
The final miRNA signature comprised 4 cell-free miRNAs and 6 exosomal miRNAs. The 4 cell-free miRNAs were hsa-miR-142-3p, hsa-miR-30c-5p, hsa-miR-335-5p, and hsa-miR-340-5p. The 6 exosomal miRNAs were hsa-miR-200b-3p, hsa-miR-1260b, hsa-miR-145-3p, hsa-miR-145-5p, hsa-miR-429, and hsa-miR-200a-3p. The investigators also stacked these panels to create a cell-free and exosomal miRNA signature. These were measured via quantitative real-time polymerase chain reaction and developed using the eXtreme Gradient Boosting machine-learning algorithm.
How did the assay perform in monitoring treatment response and recurrence?
In a cohort of 19 individuals followed for a median of 15.2 months, miRNA signature levels decreased during neoadjuvant chemotherapy and after surgery, and increased in the vicinity of a disease recurrence. A decrease greater than 0.71 in miRNA signature levels following neoadjuvant chemotherapy was associated with improved outcomes, including a 6-month recurrence-free survival rate of 85.7% compared with 41.7% among those with a smaller decrease (log-rank P = .035).
What is the clinical significance of these findings?
Current PDAC surveillance methods, such as radiology-based and endoscopy-based approaches, are invasive, costly, and primarily accessible to high-risk individuals; CA19-9, the biomarker most commonly associated with PDAC, is FDA approved solely for cancer management rather than diagnosis or screening because of its limited sensitivity and specificity. The study authors noted that a minimally invasive biomarker such as PANXEON is not intended to screen the general population, but rather to serve as a first-line triage test for individuals already recognized as being at elevated risk, including those with hereditary or familial predisposition, pancreatic cystic neoplasms, or chronic pancreatitis.
Using a Markov model simulation, the investigators estimated that PANXEON-based screening in a high-risk population could shift PDAC diagnoses toward earlier stages, with 81.2% of cases detected at stage I compared with 5.9% in the absence of screening, corresponding to a modeled improvement in estimated 5-year survival from 7.06% to 30.88%. The authors concluded that PANXEON may complement existing strategies for detecting early-stage PDAC and its precursors, and they noted that the findings warrant confirmation in larger, prospectively followed high-risk populations.
References
Xu C, Mannucci A, Han H, et al. Liquid biopsy for early detection of pancreatic ductal adenocarcinoma. Nat Med. Published online September 16, 2026. doi:10.1038/s41591-026-04625-x
