摘要
目的 探究肝纤维化患者多模态超声弹性参数、生物学指标与肝纤维化进展的关系,并构建肝纤维化进展的列线图预测模型。方法 回顾性分析2022年11月至2023年12月商丘市第一人民医院收治的180例乙肝肝纤维化患者的临床资料,将轻度肝纤维化患者纳入对照组(n=120),中重度肝纤维化患者纳入研究组(n=60)。比较两组患者的一般资料、多模态超声弹性参数、生物学指标,应用Logistic回归分析乙肝肝纤维化进展至中重度的危险因素,根据危险因素、多模态超声弹性参数构建列线图预测模型,应用受试者工作特征(ROC)曲线、决策曲线分析法(DCA)分析列线图预测模型的预测价值。结果 研究组患者的天冬氨酸氨基转移酶(AST)/血小板比率指数(APRI)、血清乙肝病毒脱氧核糖核酸(HBV-DNA)、肿瘤坏死因子-α (TNF-α)、Nod样受体家族含pyrin结构域蛋白3(NLRP3)、五聚素蛋白3 (PTX3)、透明质酸(HA)、Ⅲ型前胶原(PCⅢ)、Ⅳ型胶原(PCⅣ)水平分别为1.16±0.37、(7.26±0.93) lgIU/mL、(68.25±7.13) pg/mL、(7.18±1.39)μg/L、(0.95±0.21) ng/mL、(471.26±152.68)μg/L、(172.38±31.74)μg/L、(106.52±28.19)μg/L,明显高于对照组的0.65±0.21、(6.02±0.71) lgIU/mL、(59.74±6.51) pg/mL、(6.05±1.14)μg/L、(0.79±0.17) ng/mL、(301.28±97.85)μg/L、(145.72±29.60)μg/L、(88.09±25.43)μg/L,差异均有统计学意义(P<0.05);研究组患者的杨氏模量值、肝纤维化指数(LF index)分别为(9.35±2.32) k Pa、1.31±0.35,明显大于对照组的(6.81±1.17) k Pa、0.92±0.31,差异均有统计学意义(P<0.05);多因素Logistic回归分析结果显示,APRI、血清HBV-DNA、TNF-α、NLRP3、PTX3、HA、PCⅢ、PCⅣ水平、杨氏模量值、LF Index均为乙肝肝纤维化进展至中重度的危险因素(P<0.05);基于上述危险因素构建的乙肝肝纤维化进展的列线图预测模型,ROC曲线显示新列线图预测模型AUC为0.907(95%CI:0.862~0.951),具有较高预测效能;DCA曲线显示新列线图模型具有正向净获益。结论 APRI、血清HBV-DNA、TNF-α、NLRP3、PTX3、HA、PCⅢ、PCⅣ水平、杨氏模量值、LF Index均是乙肝肝纤维化进展至中重度的危险因素,基于上述危险因素构建的风险预测模型敏感度较高,具有一定预测价值。
Objective To explore the relationship between multimodal ultrasonic elastography parameters,biological indicators and the progression of liver fibrosis in patients with liver fibrosis,and to construct a nomogram prediction model for the progression of liver fibrosis.Methods A retrospective analysis was conducted on the clinical data of 180 patients with hepatitis B liver fibrosis admitted to Shangqiu First People's Hospital from November 2022 to December 2023.Patients with mild liver fibrosis were assigned into the control group(n=120),while patients with moderate to severe liver fibrosis were assigned into the study group(n=60).The general information,multimodal ultrasonic elastography parameters,and biological indicators were compared between the two groups of patients.Logistic regression analysis was applied to identify the risk factors for the progression of hepatitis B liver fibrosis to moderate-severe stages.Based on the risk factors and multimodal ultrasonic elastography parameters,a nomogram prediction model was constructed.The predictive value of the nomogram prediction model was analyzed using receiver operating characteristic(ROC)curve and decision curve analysis(DCA).Results The levels of aspartate aminotransferase(AST)/platelet ratio index(APRI),serum hepatitis B virus deoxyribonucleic acid(HBV-DNA),tumor necrosis factor-α(TNF-α),Nod-like receptor family with pyrin domain-containing protein 3(NLRP3),pentraxin 3(PTX3),hyaluronic acid(HA),typeⅢprocollagen(PCⅢ),and typeⅣcollagen(PCⅣ)in the study group were 1.16±0.37,(7.26±0.93)logIU/mL,(68.25±7.13)pg/mL,(7.18±1.39)μg/L,(0.95±0.21)ng/mL,(471.26 ±152.68) μg/L, (172.38±31.74) μg/L, and (106.52 ±28.19) μg/L, respectively, significantly higher than 0.65±0.21, (6.02±0.71) lgIU/mL, (59.74±6.51) pg/mL, (6.05±1.14) μg/L, (0.79±0.17) ng/mL, (301.28±97.85) μg/L, (145.72±29.60) μg/L, and (88.09±25.43) μg/L in the control group (P<0.05). The Young's modulus and liver fibrosis index (LF index) in the study group were (9.35±2.32) kPa and 1.31±0.35, respectively, significantly higher than (6.81±1.17) kPa and 0.92±0.31 in the control group (P<0.05). Multivariate logistic regression analysis showed that APRI, serum HBV-DNA, TNF-α, NLRP3, PTX3, HA, PCⅢ, PCⅣ levels, Young's modulus value, and LF Index were all risk factors for the progression of hepatitis B liver fibrosis to moderate severe stages (P<0.05). Based on the above risk factors, a nomogram prediction model for the progression of hepatitis B liver fibrosis was constructed. ROC curve indicated that the AUC of the new no-mogram prediction model was 0.907 (95% CI: 0.862-0.951), demonstrating high predictive performance. DCA showed that the new nomogram model had a positive net benefit. Conclusion APRI, serum HBV-DNA, TNF-α, NLRP3, PTX3, HA, PCⅢ, PCⅣ, Young's modulus, and LF index are risk factors for progression to moderate and severe hepatitis B liver fi-brosis. The risk prediction model based on these risk factors is highly sensitive and has certain predictive value.
作者
孔莹
梁静
郭睿
李彦
KONG Ying;LIANG Jing;GUO Rui;LI Yan(Department of Ultrasound,Shangqiu First People's Hospital,Shangqiu 476000,Henan,CHINA)
出处
《海南医学》
CAS
2024年第22期3289-3294,共6页
Hainan Medical Journal
基金
2020年河南省医学科技攻关计划联合共建项目(编号:LHGJ20200925)。
关键词
乙型肝炎
肝纤维化
多模态超声弹性参数
预测模型
危险因素
血清学试验
列线图
Hepatitis B
Liver fibrosis
Multimodal ultrasonic elastography parameters
Prediction model
Risk factors
Serological test
Nomogram
