Related Experiment Video
Updated: May 18, 2026
Computer-assisted Large-scale Visualization and Quantification of Pancreatic Islet Mass, Size Distribution and Architecture
Published on: March 4, 2011
Integrated cytologic, biochemical, imaging, and molecular analysis of pancreatic cystic lesions using PancreaSeq: a
Jing Wang1, Wei Sun2, Tamas A Gonda3
1Department of Pathology, New York University Langone Health, New York City, New York; Department of Pathology, Moffitt Cancer Center, Bronx, New York.
Introduction:
Accurate preoperative evaluation of pancreatic cysts is essential. However, cytology and biochemical analysis are often limited by low cellularity, and risk stratification is critical for management. PancreaSeq Genomic Classifier (GC) analyzes cyst fluid for molecular alterations to aid diagnosis and risk assessment.
Materials And Methods:
We retrospectively analyzed 219 pancreatic cysts from 206 patients using PancreaSeq GC, integrating molecular findings with cytology, biochemical, imaging, surgical pathology, and follow-up.
Results:
PancreaSeq GC successfully analyzed 216/219 cysts (99%) and detected alterations in 182 (83%). Among cases with both cytology and molecular data (n = 201), concordance was high in cytologically mucinous neoplasms (94%) and atypical cases (95%). Notably, among cases reported as negative for malignancy or nondiagnostic on cytology (n = 128), PancreaSeq GC identified mucinous neoplasms in 82 cases (64%), demonstrating added value in limited samples. Surgical pathology correlation (n = 24) showed excellent performance for distinguishing mucinous from nonmucinous cysts (area under the curve [AUC] = 0.94, P < 0.001). Risk stratification for detection of any dysplasia yielded an AUC of 0.78 (P = 0.006), and for high-grade dysplasia an AUC of 0.74 (P = 0.046). PancreaSeq GC reliably predicted neuroendocrine tumors, but the sensitivity for focal high-grade dysplasia in mucinous neoplasms and serous cystadenoma was limited. Compared with carcinoembryonic antigen (CEA), cyst fluid glucose showed higher sensitivity but lower specificity for mucinous cyst detection.
Conclusions:
PancreaSeq GC provides significant diagnostic and risk-stratification value that complements cytological evaluation, particularly in indeterminate or nondiagnostic cytology specimens and when biochemical data are unavailable. Integration of molecular findings improves cyst classification and dysplasia risk assessment. Multidisciplinary assessment remains essential, given the assay's limited sensitivity for focal high-grade dysplasia and serous cystadenomas.
