External validation of plasma CSF1 as a preoperative prognostic marker in patients with resectable intrahepatic cholangiocarcinoma

Antonio Akiki,Hanna Jacobsson,Ghada Nouairia,Martin Cornillet,Niklas K Björkström,Ernesto Sparrelid,Helena Taflin,Hannes Jansson
DOI: https://doi.org/10.1101/2024.11.27.24318028
2024-11-28
Abstract:Background & Aims: Long-term prognosis after resection for intrahepatic cholangiocarcinoma (iCCA) remains poor and the preoperative risk assessment is difficult. A previous single-center study indicated two immune system-related proteins in plasma, colony stimulating factor 1 (CSF1) and TNF-related apoptosis-inducing ligand (TRAIL), as preoperative prognostic factors in iCCA. This study aimed to externally validate CSF1 and TRAIL as prognostic markers for patients with resectable iCCA. Methods: Preoperative plasma CSF1 and TRAIL concentrations (pg/mL) were determined from prospectively collected biobank samples using multiplex immunoanalysis (Proximity Extension Assay), from patients operated for iCCA at two tertiary referral centers, Karolinska (2010-2020) and Sahlgrenska (2019-2023) university hospi-tals. The primary outcome was overall survival (OS), analyzed by Kaplan-Meier method and Cox regression. Secondary outcome was disease-free survival (DFS). Results: Sixty-one patients with resection for iCCA were included. CSF1 above median was associated with lymph node metastasis (P=0.03). CSF1 was associated with both OS (hazard ratio [HR] 1.03, 95% confidence interval [CI] 1.01-1.05) and DFS (HR 1.02, 95% CI 1.00-1.04). The median OS was eight months for patients with CSF1 values in the upper quartile (≥158 pg/mL), compared to an overall median OS of 47 months. While TRAIL was not significantly associated with OS (P=0.22), values in the lower quartile (≤256 pg/mL) were associated with short DFS (P<0.01). Multivariable analyses confirmed the independent prognostic signifi-cance of CSF1. The C-index of CSF1 for OS was 0.70, with excellent calibration for three- and five-year OS. Conclusion: Plasma CSF1 was validated as a novel independent, well-calibrated preoperative predictor of poor survival in resectable iCCA, which could assist the preoperative risk assessment. Low plasma TRAIL was associated with early recurrence.
Oncology
What problem does this paper attempt to address?