Abstract
Colorectal cancer (CRC) remains a global health burden, with rectal cancer comprising approximately one-third of these cases. The standard treatment for locally advanced rectal cancer is neoadjuvant chemoradiotherapy (nCRT) followed by surgery, yet patient responses are highly variable; around 20% achieve a pathological complete response (pCR), 40% have partial response, and some show resistance. The mechanisms driving these differences remain poorly understood, complicating efforts to predict outcomes reliably. Given the significant toxicities of nCRT, identifying predictive biomarkers is crucial to tailor treatment, avoid unnecessary toxicity, and improve outcomes. Although many potential biomarkers have been explored, none have yet been validated for routine clinical use. Emerging research highlights the complex interactions between the microbiome and tumour microenvironment (TME) as important factors in treatment response, but studies specifically investigating these interactions in nCRT remain limited.
To investigate the microbiome’s influence on tumour response to therapy, a co-culture model was developed using CRC cell lines and the commensal, Bacteroides thetaiotaomicron. DLD-1 cells exposed to 2 Gy radiation dose, in the presence of B. thetaiotaomicron, showed non-significant transcriptional shifts of genes related to immune evasion, inflammation, tissue remodelling, and survival. This suggests a complex interaction between radiation-induced stress and bacterium-mediated immune modulation, potentially driven by its rough LPS activating TLR2-dependent inflammatory pathways. Gene expression patterns varied at higher radiation doses and across different colorectal cancer cell lines, highlighting the complexity of tumour-microbiome interactions in radiotherapy response that is unlikely to be captured in 2D tumour-only systems lacking immune components.
Using the MetaFunc computational pipeline, integrated microbial and host genomic analyses of pre-and post-nCRT tumour and matched normal tissues showed that although phylum-level microbiome composition remained stable post-nCRT, species-level changes in tumours included enrichment of opportunistic pathogens, implicated in inflammation and resistance, alongside beneficial commensals in normal mucosa. nCRT induced broader microbial shifts, increasing diversity and altering taxa involved in immune modulation. Concurrent host transcriptomic profiling revealed changes in pathways related to extracellular matrix remodelling, growth factor binding, and ion channel activity post-treatment, possibly driving resistance. Pre-treatment tissues showed enrichment of genes linked to ribosome biogenesis and tumour progression. These results demonstrate dynamic, interconnected microbial and molecular responses to nCRT.
Spatial and molecular profiling with NanoString GeoMx revealed that complete responders have a distinct TME marked by lower immune checkpoint protein expression, increased apoptotic signalling, via cleaved caspase 9, and unique immune cell cluster patterns. The presence of tertiary lymphoid structures (TLSs) correlated with favourable outcomes; altered expression of CD56, GBA and EpCAM within TLSs indicates roles that require further study. In tumour areas lacking TLSs, complete responders showed higher STING activation and decreased immunosuppressive markers (ARG1, CTLA-4, CD68) in B cell-rich sections, reflecting a more immune-permissive microenvironment.
Together, these findings demonstrate that favourable responses to nCRT are associated with a balanced microbiome, active immune state, and enhanced apoptosis, while resistance corresponds to immunosuppressed and dysbiotic environments. Integrating microbiome analysis with molecular and spatial profiling holds promise for identifying predictive biomarkers and therapeutic targets. This integrated approach may guide future clinical strategies combining nCRT with microbiome modulation or immune-targeted treatments, aiming to improve precision and efficacy for rectal cancer patients.