放化疗是局部晚期食管癌的重要治疗手段，早期疗效预测是成功的关键，也是临床难点。治疗前FDG PET图像纹理分析量化肿瘤内糖代谢异质性，在疗效预测方面优于PET传统指标，但尚无针对局部晚期食管癌的预测模型；纹理与传统参数的相关性和互补性如何，疗效预测的生物学基础均不明确。我们发现不同分期食管癌纹理参数有显著差异，以往研究未考虑分期对异质性的影响；肿瘤内不均质的FDG分布对应不同放化疗敏感性的细胞群体，可能是纹理分析预测疗效的生物学基础。据此，拟通过动物和临床研究：①验证肿瘤糖代谢异质性与对放化疗反应异质性的相关性，明确纹理分析预测疗效的生物学基础；②以诱导凋亡/抑制增殖率（动物）和病理缓解率（临床）为疗效评判标准，评价纹理、传统及联合应用的预测价值，构建最佳模型；③评估该模型早期预测疗效及判断预后的效能。本研究将充分利用FDG PET所提供的肿瘤生物异质性信息，为食管癌个体化治疗奠定基础。

Radio-chemotherapy is the mainstay for treating locally advanced esophageal cancer. It is very important to predict response to therapy early in the course of treatment, potentially allowing selection of optimal strategy and survival improvement. FDG PET textural features analysis can quantify the heterogeneity of glucose metabolism. Pilot studies showed that some baseline FDG PET textual features could predict response to concomitant radio-chemotherapy in esophageal cancer. However, its correlations with traditional PET parameters，its biological mechanism and the optimal prediction model is unknown. Our preliminary study found there were significant differences in PET textual features among variable staged esophageal cancer. Meanwhile, intra-tumor heterogeneous uptake of FDG could differentiate tumor cell population with different sensitivity to radio-chemotherapy. Based on it, this project is proposed to: 1) clarify the biological mechanism for the prediction model by validating the correlation between intra-tumor heterogeneity of glucose metabolism and different response to treatment; 2) define the best parameters and their thresholds, establish the optimal prediction model with textual features and traditional PET parameters, according to the treatment response criteria by induced apoptosis, proliferation inhibition for animal studies and pathological tumor regression grade for clinical studies; 3) evaluate the efficacy of established model to early predict radio-chemotherapy response in esophageal cancer in both basic and clinical studies. This project will utilize biological information of tumor heterogeneity from FDG PET images and provide the rationale of personalized treatment for patients with locally advanced esophageal cancer.