ESMO Congress 2023 | OncologyPRO

Abstract 156P

Background

Abnormal collagen fiber architectures in tumor fibrosis affect prognosis and treatment response in pancreatic cancer (PC) patients. Increased fibrotic activity releases collagen metabolites into the bloodstream. The aim was to explore these metabolites to cluster PC patients into tumor fibrosis endotypes and evaluate overall survival (OS).

Methods

Serum was collected from patients with PC prior to chemotherapy or resection and included in the BIOPAC Study (ID: NCT03311776) (stage I (n=15), Stage II (n=201), stage III (n=164) and stage 4 (n=434)). Principal component analysis was performed on the levels of five collagen metabolites (Table) measured by ELISAs. K-means clustering was used to discover clusters and Kaplan-Meier curves and the cox proportional hazards model used to dicover associations with OS. Table: 156P

Biomarker name	collagen metabolite
C3M	Neo-epitope of MMP-9 mediated degradation of type III collagen
PRO-C3	Released N-terminal pro-peptide of type III collagen
PRO-C5	Released C-terminal pro-peptide of type V collagen
PRO-C6	C-terminal of released C5 domain of type VI collagen α3 chain (endotrophin)
PRO-C11	Released N-terminal pro-peptide of type XI collagen

Results

Three putative endotypes (cluster A, B, C) were identified. Patients in cluster B had higher C3M, PRO-C5 and PRO-C11 levels compared to cluster A, whereas patients in cluster C had higher PRO-C3 and PRO-C6 levels compared to cluster A. Patients with endotype A had a median OS of 332 days vs. 185 and 144 days for endotype B and C (p<0.001). Compared to endotype A, both endotype B (HR=1.56, 95% CI 1.28–1.91, p<0.001) and endotype C (HR=1.69, 95% CI 1.26–2.26, p<0.001) were predictors of poor OS when corrected for age, sex, performance status, body mass index, diabetes, cachexia, stage, number of metastatic sites, and CA19-9.

Conclusions

Clustering of patients with PC based on five collagen-derived metabolites measured in serum identifies 3 distinct endotypes that were associated with poor OS independent of other common risk factors. These findings indicate that non-invasive collagen biomarkers can be used to identify fibrotic endotypes in patients with PC and that such endotypes may inform on patient prognostication.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

The Danish Research Foundation.

Disclosure

J. Thorlacius-Ussing, N.I. Nissen, C. Jensen, M. Karsdal, N. Willumsen: Financial Interests, Institutional, Full or part-time Employment: Nordic bioscience A/S. All other authors have declared no conflicts of interest.