Lab Invest. 2025 Feb 13. pii: S0023-6837(25)00017-0. [Epub ahead of print] 104107
Elise Pretzsch,
Christiane A Peschel,
Matjaz Rokavec,
Lucien Torlot,
Pan Li,
Heiko Hermeking,
Jens Werner,
Frederick Klauschen,
Jens Neumann,
Andreas Jung,
Jörg Kumbrink.
FOLFIRI, a combination of folinic acid, 5-Fluorouracil, and Irinotecan, is one of the recommended first-line chemotherapeutic treatments for metastatic colorectal cancer (mCRC). Unfortunately, acquired FOLFIRI resistance represents a common obstacle in the treatment of mCRC patients. Thus, we aimed to identify mechanisms, gene alterations and gene expression signatures contributing to acquired FOLFIRI resistance by mimicking this problem in a cell culture model and subsequent translation in clinical datasets. Three FOLFIRI resistant CRC cell lines were established by continuous FOLFIRI treatment. Comparative mutation screening (161 genes) and transcriptomics (pathway and differential expression analyses) were performed in parental and resistant cells. Data reconciliation was performed in GSE62322, a clinical FOLFIRI responder dataset (intrinsic resistance). Relapse-free survival (RFS) associations of identified differentially expressed genes (DEGs) and potential gene signatures were investigated in eight clinical CRC datasets. No mutual genetic alterations were found in FOLFIRI resistant derivatives. Resistant cell lines displayed activation of MAPK, immune response and EMT pathways. 12 DEGs, significantly differentially expressed in at least two of the three resistant cell lines, were identified. Comparison with GSE62322 and subsequent survival analyses revealed a five-gene FOLFIRI signature comprised of CAV2, TNC, TACSTD2, SERPINE2, and PERP that was associated with RFS in multiple datasets including TCGA CRC (Hazard ratio (HR)=2.634, p=4.53x10-6), in pooled samples of all datasets (all stages (N=1981): HR=1.852, p=6.44x10-13; stage IV (N=260): HR=2.462, p=5.22x10-9). A multivariate Cox regression analysis identified the five-gene signature as an independent prognostic factor in the TCGA dataset (HR=1.89, p=0.0202). Our analyses revealed a five-gene FOLFIRI resistance signature associated with RFS that may help to predict FOLFIRI resistance and thus avoid unnecessary ineffective treatment. Signature members might also represent targets to fight FOLFIRI resistance.
Keywords: FOLFIRI resistance; Metastatic colorectal cancer; biomarker; gene signature; survival