Abstract
BACKGROUND: The authors hypothesized that small ribonucleic acid (sRNA) obtained from blood samples after neoadjuvant therapy from patients treated with neoadjuvant chemoradiation therapy (NACRT) could serve as a novel biomarker for predicting pathologic complete response (pCR). METHODS: This study included 99 patients treated with esophagectomy after NACRT between March 2010 and October 2021 whose blood samples were collected between the end of NACRT and surgery. Next-generation sequencing (NGS) was used to analyze sRNAs from the blood samples. A predictive model for pCR comprising micro-RNA isoforms (isomiR), transfer RNA (tRNA)-derived sRNAs (tsRNAs), and clinical factors was constructed using cross-validation. RESULTS: Of the 99 patients, pCR was diagnosed for 30 and non-pCR for 69 of the patients. Among sRNAs, the isomiRs of let-7b and miR-93 and the tsRNA group derived from tRNA-Gly-CCC/GCC were identified as predictive factors. The clinical factors included a decrease in the maximum standardized uptake value (SUVmax) at the primary site, clinical complete response post-NACRT, preoperative biopsy, and post-NACRT carcinoembryonic antigen levels. The combined predictive model for pCR (C-PM) was established using the three sRNAs and four clinical factors. The area under the curve for the C-PM was 0.84, which was a significant factor in the multivariate analysis (odds ratio, 89.41; 95 % confidence interval 8.1-987.5; p < 0.001). CONCLUSIONS: Pathologic complete response after NACRT can be predicted by a predictive model constructed from preoperative clinical factors obtained via minimally invasive procedures and sRNA identified by NGS. Preoperative pCR prediction may influence treatment decision-making after NACRT.