Response to neoadjuvant chemoradiotherapy in locally advanced rectal cancer (LARC) varies considerably. To quantitatively assess treatment response, the Neoadjuvant Rectal (NAR) score has been established, which measures tumour downstaging and nodal status, correlating positively with survival outcomes. This study investigates if low and high NAR scores denote biologically distinct groups prior to treatment.
Using transcriptome microarrays, the study analyzed RNA expression from pre-treatment biopsies of LARC patients extracted from two publicly available datasets: GSE87211 and the S:CORT consortium Grampian cohort. A validation cohort from the Beatson West of Scotland Cancer Centre was assessed via multiplex immunofluorescence, focusing on tumour-stromal ratios.
The analysis revealed that 17 MSigDB Hallmarks were significantly enriched in high NAR tumours, with the 'Epithelial-Mesenchymal Transition' hallmark showing marked enrichment (padj < 0.001). In contrast, signatures linked to immune activity were prevalent in low NAR tumours (p < 0.25). Additionally, higher densities of Cytotoxic Lymphocytes were noted in low NAR tumours (p < 0.05), while high NAR tumours exhibited an increase in stromal cells such as Endothelial Cells and Fibroblasts (p < 0.05). This was further corroborated by higher CD8 cell densities, indicating increased immune infiltration in low NAR tumours.
The study concludes that pre-treatment tissue analyses reveal distinct microenvironmental characteristics between low and high NAR tumours. Although there are notable differences in immune and stromal features based on NAR classification, these factors alone do not serve as adequate predictive biomarkers for treatment response in LARC.
Neoadjuvant radiation treatments for locally advanced rectal cancer have become increasingly standard, yet patient responses frequently differ. Around 20% of patients achieve a complete response post-treatment, while approximately 10% show minimal to no regression. Identifying reliable response biomarkers is crucial for targeting treatments to patients who are more likely to benefit while avoiding unnecessary interventions in less responsive cases.
The NAR score, conceptualized by George et al. in 2015, assesses treatment effects by focusing on clinical downstaging. A NAR score under 8 indicates a favorable response and correlates with improved patient outcomes, while scores over 16 suggest poor responses and are associated with worse long-term results. Validated across diverse neoadjuvant therapy regimens, the NAR score serves as a significant predictor for survival and recurrence rates.
Despite its relevance, a comprehensive investigation into the biological distinctions associated with varying NAR scores, particularly between high and low scorers, has been lacking. This study utilizes multiple cohort data to probe into these differences, specifically examining the biological underpinnings in extreme response categories.
The study included patients from the Beatson West of Scotland Cancer Centre, the S:CORT Grampian cohort, and the GSE87211 database, all diagnosed with rectal adenocarcinomas. Following informed consent, all participants received neoadjuvant radiotherapy followed by surgical intervention, and NAR scores were computed based on pre-treatment clinical T-stage viewed through MRI combined with post-treatment pathological assessments.
The cohort from Beatson involved 57 patients treated between 2008-2016 with standard radiotherapy alongside either capecitabine or 5-fluorouracil. Ethical approval for this cohort was granted by the NHS Greater Glasgow and Clyde Research Ethics Committee.
The S:CORT consortium included 193 patients receiving either 50 Gy of radiotherapy across 25 fractions or a shorter 25 Gy regimen, with some also receiving supportive chemotherapy post-radiation.
GSE87211 involved 200 patients enrolled in clinical trials, with carefully curated transcriptomic microarray data available for analysis.
At the Beatson centre, multiplex immunofluorescence utilized advanced imaging technology to measure immune infiltration. Tumour epithelium and stroma were distinctly identified using specific antibodies, highlighting differences in cellular composition. This was followed by detailed quantification focusing on CD8-positive lymphocyte density across tumor samples.
The TSP was evaluated through histological scoring, distinguishing samples with greater than or equal to 50% of stroma from those with less, providing insights into the tumour microenvironment's composition.
Tumour purity was assessed utilizing the ESTIMATE algorithm, which computes stromal and immune scores from gene expression data, offering a quantitative measure of the tumour microenvironment.
Subsequent analyses included archival biopsies subjected to the Almac Diagnostics XCEL microarray, which provided detailed expression profiles necessary for examining the biological characteristics of LARC tumours and their responses to neoadjuvant therapies.