2021 Volume 64 Issue 3 Pages 249-259
Background The standard treatment for resectable advanced esophageal squamous cell carcinoma in Japan is surgery followed by neoadjuvant chemotherapy, and it is important to predict the effect of neoadjuvant chemotherapy before treatment. Therefore, this study aims to extract conventional blood examination data, such as tumor markers and/or inflammatory/nutritional index levels, that can predict the pathological response of patients with esophageal squamous cell carcinoma to neoadjuvant chemotherapy.
Methods We retrospectively analyzed the medical records of 66 patients with thoracic esophageal squamous cell carcinoma who received neoadjuvant chemotherapy, followed by curative esophagectomy at Tottori University Hospital between June 2009 and December 2019.
Results We demonstrated that the product of the platelet-to-lymphocyte ratio (PLR) multiplied by the cytokeratin-19 fragment (CYFRA) level, which was termed “PLR-CYFRA,” is the most accurate indicator that predicts the pathological response to neoadjuvant chemotherapy, with the highest area under the curve [0.795 (95% confidence interval: 0.665–0.925), P < 0.001] in receiver operating characteristic analyses. Therefore, we divided patients into the PLR-CYFRALow (< 237.6, n = 21) and PLR-CYFRAHigh (≥ 237.6, n = 45) groups and found that the percentage of PLR-CYFRALow was significantly higher in patients with a better pathological response (P < 0.001). Furthermore, patients with good pathological response had significantly better prognoses in terms of disease-specific survival (P = 0.014), recurrence-free survival (P = 0.014), and overall survival (P = 0.032). In the multivariate analysis, PLR-CYFRA was an independent predictor of the pathological response of patients with esophageal squamous cell carcinoma to neoadjuvant chemotherapy (P = 0.002).
Conclusion Pretreatment PLR-CYFRA might be a useful and simple tool that predicts the pathological effect of neoadjuvant chemotherapy in esophageal squamous cell carcinoma.