放射性肠炎（Radiation enteritis，RE）是盆腔肿瘤放疗后常见并发症,随着盆腔肿瘤放射治疗的广泛开展，RE发病率也有随之增加。放射性肠炎一旦发生，会直接影响到患者的治疗效果和生存质量。因此，预测和预防急性放射性肠炎的发生就非常重要。目前国内外研究主要集中于急性放射性肠炎发生的单因素分析，而且缺乏有效的预测模型。本研究对急性放射性肠炎发生的临床因素和剂量学因素进行单因素和多因素分析，期望获得急性放射性肠炎发生的相关危险因素，并在此基础上初步建立急性RE的预测模型。危险因素分析首先在Embase.com、Elsevier ScienceDirect和CNKI数据库中进行“放射性肠炎”、“Radiation enteritis”等检索放射性肠炎相关文献，然后根据Cosort2010和纽卡斯尔评分标准评估质量，采用Review Manager（Revman）5.3软件进行Meta分析。Meta分析内容包括异质性、稳定性、发表偏倚和效应量。进而，对第三军医大学第一附属医院2014年1月至2016年5月行盆腔IMRT放射治疗的240例患者根据RTOG胃肠道反应评分标准统计≥2级RE的发病率，采用spss 18.0软件进行分析。对各危险因素进行了单因素统计学分析；对有统计学意义的单因素进行相关性分析；通过二项logistic回归强制进入方式进行多因素分析。采用后退logistic回归方法建立了急性放射性肠炎的预测模型，并对模型进行检验。Meta分析表明发生≥2级急性RE临床危险因素有高龄（P=0.02）、盆腔手术史(P＜0.001)、同步放化疗（P＜0.001），剂量学危险因素有V15（P＜0.0001）、V20（P＜0.00001）和V30（P=0.0001）、V40（P＜0.00001），谷氨酰胺的使用是保护因素（P＜0.0001）。临床分析表明，同步放化疗可以增加急性放射性肠炎发生的风险（P=0.022），DVH因素中V15、V30、V40是≥2急性放射性肠炎发生的危险因素（P值均＜0.05）。建立的预测模型真阳性率为68.18%，真阴性率为95.90%，总体预测准确率90.83%。结果表明，同步放化疗是发生≥2级RE的危险因素。剂量学因素中V15、和V30、V40是≥2级RE的危险因素。根据临床资料建立的急性放射性肠炎预测模型，特异性较高（95.90%），敏感性较低（68.18%），阴性结果的预测能力较高，对急性放射性肠炎的预测有一定的临床参考价值。

Radiation enteritis (RE) is a common complication of pelvic tumor radiotherapy. With the extensive development of pelvic radiotherapy, the incidence of RE has also increased. Radiation enteritis will directly affect the patient's treatment and quality of life. Therefore, it is very important to predict and prevent the occurrence of acute radiation enteritis. At present at home and abroad for the prevention of acute radiation enteritis prevention of independent risk factors for the study, the lack of effective prediction method. In this study, the clinical factors and dosimetry factors of acute radiation enteritis were analyzed by single factor and multivariate analysis. It was expected that the predictive model of acute radiation enteritis should be established.Risk factor analysis The literature on radioactive enteritis was first searched in Embase.com, Elsevier ScienceDirect and CNKI databases, and Radiation enteritis was used to evaluate the quality of radiation enteritis. The quality was assessed according to the Cosort2010 and Newcastle scoring criteria. Review Manager (Revman) 5.3 software Meta analysis was performed. Meta analysis includes heterogeneity, stability, publication bias, and effect. Furthermore, 240 patients who underwent pelvic IMRT radiotherapy from January 2014 to May 2016 in the First Affiliated Hospital of the Third Military Medical University were enrolled according to the RTOG gastrointestinal response score. The incidence of ≥ 2 RE was measured using spss 18.0 software To analyze. A single factor statistical analysis was performed on each risk factor; a single factor was analyzed for statistical analysis; multivariate analysis was performed by binary logistic regression. The predictive model of acute radiation enteritis was established by back logistic regression method, and the model was tested.Meta-analysis showed that there were V15 (P <0.0001) in the risk factors of acute RE, and the clinical risk factors of acute RE were higher (P = 0.01), pelvic surgery (P <0.001), concurrent radiotherapy and chemotherapy (P < V20 (P <0.00001) and V30 (P = 0.0001), V40 (P <0.00001), the use of glutamine was a protective factor (P <0.0001). Clinical analysis showed that concurrent radiotherapy and chemotherapy could increase the risk of acute radiation enteritis (P = 0.022), and the risk factors for V15, V30, V40 and ≥2 acute radiation enteritis in the DVH factor (P <0.05). The true positive rate was 68.18%, the true negative rate was 95.90%, and the overall prediction accuracy was 90.83%.The results showed that concurrent radiotherapy and chemotherapy is a risk factor of ≥ 2 RE. Dose factors V15, and V30, V40 is ≥ 2 level RE risk factors. (95.90%), the sensitivity was low (68.18%), the predictability of true negative was higher, and the clinical reference value for the prediction of acute radiation enteritis was determined by the clinical data.