| 摘要: |
| 目的 探讨超声声学结构定量（ASQ）诊断慢性乙型肝炎（CHB）肝纤维化的准确性及其与病理分期的相关性。方法 选择2019年8月-2021年2月于我院就诊的CHB患者122例纳入研究。记录纳入研究患者年龄、性别、体重和身高并计算体重指数（BMI），并检测血生化指标。所有患者均行ASQ检查，用ASQ分析软件进行分析得到参数红蓝曲线下面积（FD）比值。通过超声引导穿刺或部分肝叶切除活检，采用METAVIR评分（F0-F4）对肝纤维化病理组织学分期进行评价。相关性分析采用Spearman秩相关。敏感性和特异性相加减1即为约登指数，指数越大说明实验效果越好、真实性越大，根据每个肝纤维化阈值的最高约登指数：F0-F1 vs F2-F4（≥F2）、F0-F2 vs F3-F4（≥F3）、F0-F3 vs F4（≥F4）来确定预测的最佳截断值。采用ROC曲线和曲线下面积（AUC）评估ASQ参数的诊断性能。采用Woolf方法计算95%置信区间（CI）。采用Logistic回归分析确定FD比值的独立危险因素。结果 CHB患者的FD比值的范围为0.08-1.88，F0-F1、F2、F3、F4的FD比值分别为（0.23±0.07）、（0.31±0.09）、（0.50±0.11）、（0.74±0.12）。F0-F1与F2、F2与F3、F3与F4的FD比值对比差异有统计学意义（t=10.672，P＝0.012；t=12.084，P＝0.004；t=29.336，P＝0.000）。，随着肝纤维化病理组织学分期严重程度的增加FD比值呈逐步增加趋势，差异有统计学意义（F＝37.629，P＝0.000）。Spearman秩相关分析显示FD比值与肝纤维化病理组织学分期呈正相关（r＝0.71，P＝0.000），与脂肪变性程度呈负相关（r＝-0.44，P＝0.000），与炎症活动度没有显著相关性（r＝-0.15，P＝0.172）。FD比值诊断肝纤维化病理组织学分期≥F2、≥F3、≥F4的AUC值分别为0.796、0.805、0.832，AUC值均＞0.75，具有中等诊断价值，其中≥F4的诊断价值最高。FD比值预测肝纤维化病理组织学分期≥F2、≥F3、≥F4的截断值分别为0.27、0.36、0.59。Logistic回归分析结果显示：肝纤维化病理组织学分期是FD比值的独立危险因素（β＝1.562，OR＝7.381，95%CI＝4.097-11.053，P＝0.000）。结论 FD比值可较准确评估肝纤维化的严重程度，是预测肝纤维化病理分期的一个有效的无创影像学生物标志物。 |
| 关键词: 超声 声学结构量化 慢性乙型肝炎 肝纤维化 病理分期 |
| DOI： |
| 投稿时间：2021-03-17修订日期：2021-03-17 |
| 基金项目:北京市中国中医科学院自由探索项目(ZZ0708085)； |
|
| Accuracy of quantitative diagnosis of liver fibrosis by ultrasonic acoustic structure and its correlation with pathological stage in chronic hepatitis B |
| Zhang Ding-ding |
| () |
| Abstract: |
| Objective：To investigate the accuracy of ultrasonic acoustic structure quantification (ASQ) in the diagnosis of liver fibrosis in chronic hepatitis B (CHB) and its correlation with pathological stage. Methods：A total of 122 patients with CHB admitted to our hospital from August 2019 to February 2021 were included in this study. Age, sex, weight, and height were recorded and body mass index (BMI) was calculated. Blood biochemical indicators were measured. All patients received ASQ examination, and the parameter area under the red and blue curve （FD）ratio was obtained by using ASQ analysis software. METAVIR score (F0-F4) was used to evaluate the histopathological staging of liver fibrosis by ultrasound-guided puncture or partial lobectomy biopsy. Spearman rank correlation was used for correlation analysis. The addition or subtraction of 1 between sensitivity and specificity is Youden index. The greater the index is, the better the experimental effect is and the greater the authenticity is, and the optimal truncation value for prediction was determined according to the highest Yoden index of each liver fibrosis threshold: F0-F1 vs F2-F4 (≥F2), F0-F2 vs F3-F4 (≥F3), and F0-F3 vs F4 (≥F4). The diagnostic performance of ASQ parameters was evaluated using ROC curve and area under curve (AUC). The 95% confidence interval (CI) was calculated using the Woolf method. Logistic regression analysis was used to determine the independent risk factors for FD ratio. Results：The FD ratio of CHB patients ranged from 0.08 to 1.88, and the FD ratios of F0-F1, F2, F3, and F4 were (0.23±0.07), (0.31±0.09), (0.50±0.11), and (0.74±0.12), respectively. FD ratio between F0-F1 and F2, between F2 and F3, and between F3 and F4 was statistically significant (t=10.672, P = 0.012;T =12.084, P = 0.004;T =29.336, P = 0.000). The ratio of FD increased gradually with the increase of the severity of pathological stage of liver fibrosis, and the difference was statistically significant (F = 37.629, P = 0.000).. Spearman rank correlation analysis showed that the ratio of FD was positively correlated with the hiopathological stage of liver fibrosis (r = 0.71, P = 0.000), and negatively correlated with the degree of steatosis (r = -0.44, P = 0.000), but not significantly correlated with the degree of inflammatory activity (r = -0.15, P = 0.172). The AUC values of ≥F2, ≥F3 and ≥F4 in the diagnosis of hiopathological stages of liver fibrosis were 0.796, 0.805 and 0.832, respectively. The AUC values of ≥F4 were > 0.75, which had medium diagnostic value, and ≥F4 had the highest diagnostic value. The cut-off values of ≥F2, ≥F3 and ≥F4 in predicting the histopathological stage of liver fibrosis were 0.27, 0.36 and 0.59, respectively. Logistic regression analysis showed that hiopathological stage of liver fibrosis was an independent risk factor for FD ratio (β = 1.562, OR = 7.381, 95%CI = 4.097-11.053, P = 0.000). Conclusion：FD ratio can accurately evaluate the severity of liver fibrosis, and is an effective non-invasive imaging biomarker to predict the pathological stage of liver fibrosis. |
| Key words: ultrasound Acoustic structure quantization chronic hepatitis B hepatic fibrosis pathological stage |
