549P - Combined morphometric immune signatures define the prognosis of patients with resectable colorectal liver metastases

Background

During colorectal cancer (CRC) progression, half of all patients develop CRC liver metastasis (CRLM). Perioperative chemotherapy plus targeted therapy translates to high resection and improved survival rates. Still, CRLM are dependent on pathologic tumor staging according UICC System of primary CRC. The genetic and epigenetic background of CRLM defined by microsatellite stability, KRAS/BRAF, CpG-Island-methylator phenotype and immune infiltrates is still under debate as prognostic factors for overall survival (OS). Our aim was to assess the prognostic impact of immune responses in CRLM analyzed by complex immune profiling, morphometry and various sequencing techniques.

Methods

227 consecutive patients with FFPE of CRLM resections (2008-2018) and follow-up data from the clinical database were prospectively collected until 2023. Next-generation sequencing (NGS), targeted RNAseq, immunohistochemical staining (IHC) and computerized morphometric models evaluated tumor material for different immune cell types. In-silico analyses were done by network associations. Chemotherapy-naive and treated patients were compared to their outcome.

Results

All CRLM patients with preoperative chemotherapy showed a significant worse OS vs therapy-naïve patients after CRLM surgery (p<0,0001). 113 patients with pretreatment had higher CD8+ Tcell infiltrates (p=0,001) and M2 macrophages (p<0,0001). In the perioperative group, patients with higher M1 macrophage infiltration versus no infiltration (p=0,024) as well as with CD8+ Tcell low-distance to tumor cells versus high distances (p=0,033) had significant better OS. In-silico gene expression analyses of 172 CD8 Tcell+ CRLM had downregulation of cancer-associated genes, and the quality of immune cells was regulated by IFNg immune defense pathways.

Conclusions

In CRLM patients, we defined immune cell infiltrate clusters in responders and non-responders after preoperative chemotherapy. Combined morphometric immune signatures are new prognostic and predictive biomarkers, to identify new specific targets for perioperative therapy, to optimize postoperative follow-up or to plan individualized vaccination strategies.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

MSD.

Disclosure

M. Kloth: Financial Interests, Institutional, Funding: MSD. All other authors have declared no conflicts of interest.