Digital quantification of stroma percentage enhances prognostic stratification in pancreatic cancer.

BACKGROUND

Pancreatic ductal adenocarcinoma (PDAC) is characterized by a prominent desmoplastic stroma, which plays a crucial role in tumor biology and treatment resistance. While the stromal compartment is a defining histopathological feature of PDAC, its prognostic significance remains incompletely understood. This study aimed to quantify the stromal content in PDAC using digital pathology and evaluate its association with patient outcomes.

METHODS

Tissue microarrays (TMAs) were constructed from resected PDAC specimens (n = 142). Digital analysis of tumor stroma percentage (TSP) was performed on tissue sections labeled with CA19-9. Cases were stratified into low and high TSP groups based on an optimized threshold of 44.2%. Associations between TSP and clinicopathological variables were assessed, and survival outcomes were analyzed using Kaplan-Meier and Cox proportional hazards models.

RESULTS

Digital quantification revealed wide intertumoral variability in TSP. A total of 127 (89%) patients were categorized into the high TSP group (>44.2% stroma). A high TSP was significantly associated with anatomic location of the tumor in the head of the pancreas. Patients with high TSP exhibited significantly prolonged overall survival (median: 27.8 months vs 12 months, p < 0.001). In multivariable analysis, high TSP remained an independent predictor of favorable prognosis (HR = 0.26, 95% CI: 0.13-0.52, p < 0.001).

CONCLUSION

A high TSP is independently associated with improved survival in PDAC. These findings challenge traditional views of the stroma as purely tumor-promoting and suggest a potential protective role of the stromal compartment in certain contexts.