OBJECTIVE: The aim of this study is to identify biomarkers that predict efficacy of preoperative therapy and survival for esophageal squamous cell carcinoma (ESCC).
BACKGROUND: It is essential to improve the accuracy of preoperative molecular diagnostics to identify specific patients who will benefit from the treatment; thus, this issue should be resolved with a large-cohort, retrospective observational study.
METHODS: A total of 656 patients with ESCC who received surgery after preoperative CDDP + 5-FU therapy, docetaxel + CDDP + 5-FU therapy or chemoradiotherapy (CRT) were enrolled. Immunohistochemical analysis of TP53, CDKN1A, RAD51, MutT-homolog 1, and programmed death-ligand 1 was performed with biopsy samples obtained before preoperative therapy, and expression was measured by immunohistochemistry.
RESULTS: In all therapy groups, overall survival was statistically separated by pathological effect (grade 3 > grade 2 > grade 0, 1, P < 0.0001). There was no correlation between TP53, CDKN1A, MutT-homolog 1, programmed death-ligand 1 expression, and pathological effect, whereas the proportion of positive RAD51 expression (≥50%) in cases with grade 3 was lower than that with grade 0, 1, and 2 (P = 0.022). In the CRT group, the survival of patients with RAD51-positive tumor was significantly worse than RAD51-negative expressors (P = 0.0119). Subgroup analysis of overall survival with respect to positive RAD51 expression indicated preoperative chemotherapy (CDDP + 5-FU or docetaxel + CDDP + 5-FU) was superior to CRT.
CONCLUSIONS: In ESCC, positive RAD51 expression was identified as a useful biomarker to predict resistance to preoperative therapy and poor prognosis in patients who received preoperative CRT. Administration of preoperative chemotherapy may be warranted for patients with positive RAD51 expression.