BACKGROUND: Strong correlations have been shown between systemic oxidative stress (SOS) and the occurrence, metastasis, and prognosis of many types of cancers. It is yet unknown how SOS levels relate to the prognosis of esophageal squamous cell carcinoma (ESCC). The current research aims to explore the prognostic role of systemic oxidative stress index (SOSI) on ESCC receiving neoadjuvant immunochemotherapy (nICT). METHODS: Retrospective recruitment was used to identify 224 nICT-treated ESCC patients. In order to determine the integrative score of SOSI, logistic regression analyses were utilized to screen independent risk variables, with disease-free survival (DFS) serving as the dependent variable. Given the non-linear relationship between SOSI and DFS, the best threshold was determined using a restricted cubic spline (RCS) model. Independent variable determination was executed using a cox regression analysis. For prognostic prediction, a risk categorization method based on recursive partitioning analysis (RPA) was also created. RESULTS: Four SOS-related indicators, including albumin, creatinine, blood urea nitrogen, and direct bilirubin, were used to establish the SOSI. The ideal threshold of SOSI, shown by the non-linear relationship between DFS and SOSI (P<
0.001), was used to compare between two groups. As a potential prognostic factor for those nICT-treated ESCC patients, SOSI showed a strong correlation with both DFS and overall survival (OS). Patients with low SOSI had better DFS (55.1% vs. 85.5%, P<
0.001) and OS (72.6% vs. 79.1%, P=0.013). Then, a new staging that included TNM and SOSI based on RPA algorithms was produced. In terms of prognostication, the RPA model performed significantly better than TNM classification. CONCLUSION: SOSI is a simple and useful score based on available SOS-related indices. In ESCC receiving nICT, low SOSI is found to be an important factor of better prognosis.