Abstract
<title>Abstract</title> <p>Background Assessment of systemic and local immune responses is critical for determining the efficacy of cancer interventions. Our findings indicate that baseline systemic immune-inflammation index (SII) and systemic inflammation response index (SIRI) are strongly and independently correlated with treatment response following neoadjuvant chemoimmunotherapy in patients with locally advanced esophageal squamous cell carcinoma (ESCC). Methods This study aimed to evaluate the predictive value of baseline SII and SIRI for the efficacy of neoadjuvant chemoimmunotherapy in locally advanced ESCC. From January 2021 to December 2024, this retrospective study enrolled 126 patients with ESCC, all of whom received 2–4 cycles of standard-dose immunotherapy combined with platinum-based chemotherapy (without radiotherapy) and underwent curative surgery within 4–6 weeks after the last neoadjuvant treatment. SII and SIRI were calculated based on the complete blood count results before the first treatment. The primary endpoint of this study was tumor regression grade (TRG), and secondary endpoints included tumor response assessment, residual tumor proportion (RTP), disease-free survival (DFS) and overall survival (OS). Ordinal logistic regression analysis was used to assess the association between these indicators and TRG. Receiver operating characteristic (ROC) curves were utilized to assess their predictive ability for poor tumor regression of patients with ESCC who received neoadjuvant chemoimmunotherapy. Results Increased SII and SIRI were significantly associated with a higher pathological TRG. Radiological assessment further revealed markedly lower objective response rates in patients with elevated SII or SIRI. ROC curves demonstrated SII and SIRI had a good predictive ability for poor tumor regression (SII: AUC = 0.769 and SIRI: AUC = 0.695, respectively). Conclusion Baseline SII and SIRI serve as independent and valuable biomarkers for predicting neoadjuvant therapy response in patients with locally advanced ESCC.</p>