Intratumoral genetic heterogeneity (ITGH) as a biomarker for NCRT in LARC: CNAS and SVS do not impact LARC outcomes and ITGH estimation

Neoadjuvant chemoradiotherapy (nCRT) is the standard treatment for locally advanced rectal cancer (LARC), aimed at shrinking tumors and facilitating surgical resection. Recently, organ preservation has become an appealing strategy for managing rectal cancer. However, clinical and radiological features are poor predictors of response. Intratumoral genetic heterogeneity (ITGH) has emerged as a potential biomarker for assessing tumor response, and our previous studies have shown that ITGH increases following nCRT. Additionally, the effects of copy number alterations (CNAs) and structural variations (SVs) on LARC outcomes and their impact on ITGH estimation remain underexplored. In this retrospective study, we evaluated 72 patients with LARC, with pre-treatment tumor biopsy collected and subjected to whole exome sequencing. All patients received long-course nCRT with either 5FU-based chemotherapy or Capecitabine. Patients were categorized into three groups based on clinical outcomes: nCRT-Responders (n = 16), nCRT-Not Responders (n = 38), and nCRT-Not Responders Metastatic (n = 18). We used (SNVs/Indels) data to measure ITGH using the MATH score, observing a significant association between low MATH scores and nCRT response (p≤0.05, Wilcoxon test), especially when comparing nCRT-Responders to non-metastatic nCRT-Not Responders. SV events, including translocations, duplications, deletions, and insertions, were identified using MANTA, while CNAs (gains and losses) were detected with the GATK pipeline. SVs and CNAs were not correlated with clinical outcomes (p>0.05, Wilcoxon test). We further investigated whether CNAs could affect the MATH score’s association with clinical outcomes. To do this, we recalculated the MATH score using SNVs within and outside CNA regions individually and found no change in its association with clinical outcomes. Finally, we analyzed patients with non-metastatic clinical outcomes (n = 53). Using ROC curve analysis, we determined the best MATH score threshold (34.04), with a sensitivity of 89% and specificity of 53% (AUC=0.67). Multivariate analysis with the clinical-radiological features shows that MATH-low is significantly associated with nCRT-R (OR:11.66; p =0.01). Our results show no correlation between CNAs and SVs events with clinical outcomes, that CNAs events do not impact ITGH, and suggest that the MATH score may help select patients for preoperative treatment, potentially achieving a complete response and enabling organ preservation.

Corresponding author: Vandeclécio Lira da Silva (e-mail: clecio.de@gmail.com).