Pancreatic ductal adenocarcinoma (PDAC) is a highly lethal cancer with a 5-year survival rate under 10%, making molecular stratification critical for improving outcomes. This study used publicly available data from The Cancer Genome Atlas (TCGA) to identify biologically distinct subtypes of pancreatic cancer through combined proteomics and genomics analysis.
The researchers applied proteogenomic analysis - integrating protein expression data (proteomics) with gene-level mutation and copy number data (genomics) - to a cohort of 150 PDAC patients. This integrative approach aims to reveal functional differences between tumors that might not be apparent from genomic data alone.
The central hypothesis is that distinct molecular subtypes of PDAC have different biological drivers, pathway dependencies, and clinical outcomes, which could guide future targeted therapy development and patient stratification in clinical trials.
The study used Reverse Phase Protein Array (RPPA) data from the TCGA PAAD cohort, which measures the expression levels of hundreds of proteins and phosphoproteins simultaneously. RPPA is a high-throughput antibody-based technology that enables functional proteomics at scale.
K-means clustering was applied to the RPPA protein expression profiles to partition patients into distinct groups. After evaluating multiple cluster numbers, k=2 yielded the most stable and biologically interpretable result, dividing the cohort into Subtype 1 and Subtype 2.
Downstream analyses included differential expression testing, pathway enrichment analysis using gene ontology and KEGG databases, somatic mutation frequency comparison between subtypes, and survival analysis using disease-free survival (DFS) as the primary endpoint.
Pathway enrichment analysis revealed that Subtype 1 was significantly enriched for the mTOR signaling pathway, which regulates cell growth, metabolism, and proliferation. This finding is clinically relevant because mTOR inhibitors (e.g., everolimus, rapamycin) are already approved for other cancers and could be repurposed for PDAC patients in Subtype 1.
Additional pathways enriched in Subtype 1 included PI3K-Akt and other growth factor receptor signaling cascades, suggesting a broadly activated oncogenic signaling network. In contrast, Subtype 2 showed enrichment for immune and stromal-related pathways, potentially indicating greater tumor microenvironment involvement.
These pathway differences at the protein level - rather than just the mRNA level - provide stronger evidence of functional pathway activation, since proteins are the direct effectors of cellular signaling and are more likely to reflect actual drug targets.
Across the full cohort, the most frequently mutated genes were consistent with known PDAC biology: KRAS was mutated in 91% of patients, TP53 in 71%, CDKN2A in 42%, and SMAD4 in 36%. These four genes represent the canonical driver mutation landscape of pancreatic cancer.
While these mutations were prevalent in both subtypes, their relative frequencies and co-occurrence patterns differed, contributing to the distinct molecular phenotypes identified by clustering. The high frequency of KRAS mutations (nearly universal) reinforces the challenge of targeting this oncogene directly, though KRAS G12C inhibitors have recently entered clinical trials for the small subset with that specific variant.
The identification of subtype-specific secondary genomic alterations beyond these four drivers may help explain differential pathway activity and could eventually guide selection of combination therapies tailored to each subtype.
Kaplan-Meier survival analysis showed a significant difference in disease-free survival (DFS) between the two subtypes. Subtype 1 patients had a median DFS of 17.1 months compared to 13.5 months for Subtype 2, a difference that reached statistical significance.
This counterintuitive result - where the subtype with more active growth signaling (mTOR pathway) has longer DFS - may reflect tumor heterogeneity, stage distribution, or the specific cohort composition. Alternatively, Subtype 2's immune/stromal enrichment may be associated with worse prognosis in PDAC, which has a notoriously immunosuppressive microenvironment.
These survival differences underscore the clinical importance of molecular subtyping in PDAC: patients with different subtypes may have genuinely different disease courses and may require different treatment strategies to achieve optimal outcomes.
This proteogenomic analysis demonstrates that protein-level data adds meaningful information beyond genomic profiling alone in PDAC subtyping. Integrating RPPA proteomics with genomic mutation data produces subtypes with distinct pathway enrichment and clinical outcomes, supporting the rationale for proteogenomic profiling in future pancreatic cancer studies.
The identification of mTOR pathway enrichment in a specific subtype provides a biologically motivated hypothesis for clinical trials targeting this pathway in molecularly selected PDAC patients. Future studies should validate these subtypes in independent cohorts and test whether mTOR inhibitor response correlates with Subtype 1 classification.
Limitations include the relatively small sample size from TCGA and the retrospective nature of the analysis. Prospective validation using fresh tumor tissue with standardized proteomics platforms will be necessary before molecular subtyping can be integrated into clinical decision-making for pancreatic cancer patients.