Morphological classification of pancreatic ductal adenocarcinoma that predicts molecular subtypes and correlates with...

Area of Science:

• Oncology
• Pathology
• Genomics

Background:

• Pancreatic ductal adenocarcinoma (PDAC) exhibits molecular subtypes with prognostic and therapeutic implications.
• Clinicomorphological correlation of these subtypes is lacking.
• Current grading systems may not fully capture PDAC heterogeneity.

Purpose of the Study:

• To identify morphological patterns in PDAC that correlate with known molecular subtypes.
• To investigate the biological significance of these morphological patterns.
• To re-evaluate the current PDAC grading system.

Main Methods:

• Analysis of 86 primary, chemotherapy-naive PDAC resection specimens with RNA-Seq data.
• Identification of reproducible morphological patterns.
• Differential gene expression analysis based on morphology.
• Comparison of gene signatures with published molecular subtypes.
• Correlation of overall survival (OS) with morphological and molecular subtypes.

Main Results:

• Four morphological patterns were identified, segregating into 'gland forming' and 'non-gland forming' components.
• A cut-off of ≥40% 'non-gland forming' defined two groups (A and B) with gene signatures correlating to molecular subtypes.
• Significant differences in OS were observed between morphological groups.
• Morphological classification remained prognostic within moderately differentiated and 'classical' PDAC.

Conclusions:

• PDAC can be morphologically classified into distinct, biologically relevant categories that predict molecular subtypes.
• This morphological classification provides a basis for an improved PDAC taxonomy.
• The findings may inform future treatment strategies and deep learning model development.