Tumor Remission and tumor-infiltrating lymphocytes during chemoradiotherapy: predictive and prognostic markers in locally advanced esophageal squamous cell carcinoma
Menée auprès de 164 patients atteints d'un carcinome épidermoïde de l'œsophage de stade localement avancé traité par chimioradiothérapie, cette étude évalue la sensibilité et la spécificité d'un modèle, incorporant le degré de rémission de la tumeur et le niveau d'infiltration intratumorale des lymphocytes durant le traitement, pour prédire la réponse pathologique complète et la survie des patients
Purpose : Clinical tools are unavailable for accurately predicting the pathological responses to chemoradiotherapy (CRT) of patients with esophageal squamous cell carcinoma (ESCC) before surgery. Here we evaluated tumor remission and tumor-infiltrating lymphocytes (TILs) during CRT as predictors of pathological responses and prognostic markers for patients with locally advanced ESCC treated with neoadjuvant CRT (neo-CRT) or definitive CRT.
Methods : We analyzed patients with locally advanced ESCC (N = 164) who underwent neo-CRT (N = 48) or definitive CRT (N = 116). Patients underwent endoscopic ultrasonography and biopsies when induction CRT finished. Tumor remission characteristics were designated minor (–/+) to excellent remission (ER) (+++). TILs were determined in 10% increments. Tumor remission, TILs, or both, were associated with pathological complete response (pCR) and survival in the neo-CRT group, and then analyzed in the definitive CRT group.
Results : ER and lymphocyte-predominant ESCC (LPE, ≥60% TILs) were identified according to the pCR rate and patients’ disease-free survival (DFS). We built a prediction model for pCR incorporating ER and LPE. The area under the receiver operating characteristic curve was 0.877, and sensitivity and specificity were 86.7% and 90.9%, respectively. Furthermore, this model identified pathological response with an excellent calibration. DFS of patients with ER&LPE tumors was significantly longer than that of other patients.
Conclusions : When we included tumor remission and TILs during CRT, our model predicted pCR with high probability and helped stratify prognostic subgroups, thereby guiding future therapy decisions for patients with locally advanced ESCC. Validation of this model in larger, prospective, multicenter studies is essential
International Journal of Radiation Oncology • Biology • Physics , résumé, 2018