光学相干断层扫描结合组织型转谷氨酰胺酶lgA抗体检测对乳糜泻的诊断价值

Diagnostic value of optical coherence tomography combined with tissue transglutaminase lgA antibody detection for celiac disease

摘要:

目的 探讨光学相干断层扫描（ optical coherence tomography，OCT）结合组织型转谷氨酰胺酶lgA抗体（tissue transglutaminase IgA antibody，tTGA）检测对乳糜泻的诊断价值，构建相关列线图模型并验证其预测患乳糜泻的效能。方法 选取2019年1月至2022年8月在新疆维吾尔自治区人民医院就诊的109例乳糜泻筛查评分≥3分患者作为研究对象，按是否病理确诊为乳糜泻分为乳糜泻组和非乳糜泻组，比较两组患者一般临床资料、tTGA以及OCT表现，将单因素分析中有统计学意义的指标进一步采用多因素logistic回归模型分析，获得独立预测因素，基于独立预测因素构建预测患乳糜泻风险的列线图模型，绘制列线图模型的校准曲线，评价列线图模型的校准能力，绘制独立预测因素的受试者工作特征（receiver operating characteristic， ROC）曲线，进一步分析其预测效能。结果 乳糜泻组体重指数［（20.2±1.4）kg/m2比（23.3±1.5）kg/m2，t=2.459，P=0.023］以及血红蛋白含量［（79.8±10.7）g/L比（88.7±10.9）g/L，t=3.487，P<0.001］明显低于非乳糜泻组，乳糜泻筛查评分明显高于非乳糜泻组［（7.0±1.0）分比（4.0±1.0）分，t=8.157，P<0.001］，tTGA阳性所占比例（16/24比20/85，χ2=5.462，P=0.024）以及OCT小肠绒毛萎缩所占比例（20/24比10/85，χ2=9.255，P<0.001）明显高于非乳糜泻组；多因素logistic回归分析结果显示，tTGA阳性（OR=2.687，95%CI：1.496~7.289，P=0.011）、乳糜泻筛查评分（OR=2.336，95%CI：1.254~7.875，P=0.017）以及OCT小肠绒毛萎缩（OR=5.635， 95%CI：1.534~12.009，P<0.001）为乳糜泻的独立预测因素。以tTGA、乳糜泻筛查评分以及OCT小肠绒毛萎缩结合其影响权重建立列线图模型，校准曲线显示，乳糜泻患病风险的预测值与实际观测值符合度良好（P>0.05）。ROC分析结果显示，tTGA预测乳糜泻的曲线下面积（area under curve，AUC）为0.756（95%CI： 0.721~0.826）；乳糜泻筛查评分预测乳糜泻的AUC为0.789（95%CI： 0.751~0.854），最佳诊断截点为8分； OCT小肠绒毛萎缩预测乳糜泻的AUC为0.819（95%CI： 0.783~0.872），三者联合预测的AUC为0.913（95%CI： 0.867~0.954）。结论 基于tTGA、乳糜泻筛查评分以及OCT小肠绒毛萎缩建立的列线图模型能用于准确预测乳糜泻患病风险。

Abstract:

Objective To evaluate the diagnostic value of optical coherence tomography (OCT) combined with tissue transglutaminase lgA antibody (tTGA) detection for celiac disease, and to construct a correlation nomogram model verify its efficacy in predicting celiac. Methods One hundred and nine patients with celiac disease screening score ≥ 3 who visited People''s Hospital of Xinjiang Uygur Autonomous Region from January 2019 to August 2022 were selected as the subjects. They were divided into celiac disease group and non-celiac disease group according to pathological diagnosis. The general clinical data, tTGA and OCT performance of patients between the two groups were compared. The significant indicators in the single factor analysis were further analyzed by using the logistic multivariate regression model to obtain independent predictors, and a nomogram model was then established to predict the risk of celiac disease based on independent predictors. The calibration curve of the nomogram model was drawn, and the calibration ability was evaluated. Finally, the receiver operating characteristic (ROC) curve of independent prediction factors was drawn, and their prediction effectiveness was further analyzed. Results The body mass index (20.2±1.4 kg/m2 VS 23.3±1.5 kg/m2, t=2.459, P=0.023) and hemoglobin (79.8±10.7 g/L VS 88.7±10.9 g/L, t=3.487, P<0.001) in the celiac disease group were significantly lower than those in the non-celiac disease group. The screening score of celiac disease (7.0±1.0 VS 4.0±1.0, t=8.157, P<0.001) in the celiac disease group was significantly higher than that of the other group. The proportion of tTGA positive (16/24 VS 20/85, χ2=5.462, P=0.024) and small intestinal villus atrophy on OCT (20/24 VS 10/85, χ2=9.255, P<0.001) in the celiac disease group were significantly higher than those in the other group. The results of logistic multivariate regression analysis showed that tTGA positive (OR=2.687, 95%CI: 1.496-7.289, P=0.011), the screening score of celiac disease (OR=2.336, 95%CI: 1.254-7.875, P=0.017) and OCT small intestinal villus atrophy (OR=5.635, 95%CI:1.534-12.009, P<0.001) were independent predictors for celiac disease. The nomogram model was established based on the tTGA, the screening score of celiac disease and OCT small intestinal villus atrophy combined with their influential weights. The calibration curve showed that the predicted risk of celiac disease was in good agreement with the observed value (P>0.05). ROC analysis results showed that the area under the curve (AUC) of positive tTGA predicting celiac disease was 0.756 (95%CI: 0.721-0.826). The AUC of celiac disease screening score was 0.789 (95%CI: 0.751-0.854), and the best cut-off point for diagnosis was 8 points. The AUC of small intestinal villus atrophy in OCT was 0.819 (95%CI: 0.783-0.872). The AUC jointly predicted by the above three was 0.913 (95%CI: 0.867-0.954). Conclusion The nomogram model based on tTGA, celiac disease screening score and OCT small intestinal villus atrophy can be used to accurately predict the risk of celiac disease.