鼻咽癌放射性口腔粘膜炎多因素分析及建立判别模型的临床价值

背景与目的口腔粘膜炎是鼻咽癌放疗中最常见且对患者影响最大的急性反应之一。本研究拟分析鼻咽癌患者在综合治疗中发生严重口腔粘膜炎的相关因素,筛查出发生严重口腔粘膜炎的高危因素,并建立数学判别模型,对其发生情况做出预测。方法对102例鼻咽癌患者进行研究,按口腔粘膜炎的严重程度分为严重组和轻度组,对17项临床及实验室等指标进行单因素分析、多因素Logistic分析及判别分析。结果多因素Logistic分析提示化疗、粘膜反应早期未应用抗生素、口腔卫生差、患病前有吸烟史为发生严重口腔粘膜炎的高危因素;根据高危因素建立对数优势线性判别函数模型,将102例患者的临床资料进行回代判别,判别灵敏度为83.0%,特异度为71.4%,阳性预测值为75.9%,阴性预测值为79.5%。结论化疗、口腔卫生差、有吸烟史是严重的口腔粘膜炎发生的高危因素;早期干预对其有一定预防作用。本研究建立的函数模型有助于预测严重口腔粘膜炎的发生。

Analysis of oral mucositis risk factors during radiotherapy for nasopharyngeal carcinoma patients and establishment of a discriminant model

BACKGROUND & OBJECTIVE: Oral mucositis caused by radiotherapy is the most familiar complication of nasopharyngeal carcinoma (NPC), and it has strong impact on the patients. This study was to screen high risk factors of oral mucositis, and establish a mathematical model for forecasting its incidence probability in NPC patients. METHODS: Clinical and experimental factors of 102 NPC patients with oral mucositis were analyzed by univariate analysis, multifactorial Logistic regression, and discriminant analysis. RESULTS: Multifactorial Logistic regression showed that chemotherapy, bad oral hygiene status, not using antibiotic at early stage mucositis, smoking history were high risk factors of serious mucositis. A discriminant function was set up using these 4 factors, with positive forecast value of 75.9%, negative forecast value of 79.5%, sensitivity of 83.0%, and specificity of 71.4%. CONCLUSIONS: Chemotherapy, bad oral hygiene status, not using antibiotic at early stage mucositis, smoking history are high risk factors of serious mucositis after radiotherapy for NPC; early treatment has some preventive effect on it. The discriminant function may help to predict the incidence of serious oral mucositis.