非小细胞肺癌患者血清铁蛋白表达水平及其对微血管侵犯预测价值研究

Study on Serum Ferritin Expression and Its Predictive Value for Microvascular Invasion in Patients with Non-Small Cell Lung Cancer

目的探讨血清铁蛋白对非小细胞肺癌(non-small cell lung cancer,NSCLC)患者微血管侵犯(microvascular invasion,MVI)的预测价值。方法纳入2020年1月~2021年12月在武警四川省总队医院接受根治性手术治疗的109例NSCLC患者作为研究对象,按术后病理检查结果是否并发MVI分为MVI组(n=46,42.20%)和N-MVI组(n=63,57.80%),比较两组患者一般临床病理资料以及血清铁蛋白表达水平的差异,多因素Logistic回归分析评估血清铁蛋白对NSCLC患者并发MVI的预测价值,对血清铁蛋白与患者临床病理特征进行Person相关性分析。绘制血清铁蛋白以及其他独立预测因素预测NSCLC患者并发MVI的ROC曲线和决策曲线,进一步分析其预测效能和净收益率。结果MVI组患者肿瘤直径5.92(3.12,7.65)cm,癌胚抗原(carcinoembryonic antigen,CEA)73.91±9.25ng/ml,糖类抗原125(sugar antigen 125,CA125)68.92±10.21U/ml以及铁蛋白678.93±41.23μg/L水平高于N-MVI组[4.13(2.12,5.67)cm,64.22±11.37ng/ml,61.34±11.24U/ml,635.94±51.24μg/L],差异均有统计学意义(t/Z=2.765~5.662,均P<0.05)。MVI组中多发肿瘤(67.39%)、Ⅲa期TNM/T分期(73.01%)、血管内皮生长因子(vascular endothelial growth factor,VEGF)阳性(65.22%)以及淋巴结转移(70.00%)所占比例高于N-MVI组(36.51%,31.75%,34.78%,36.51%),差异具有统计学意义(χ^(2)=10.144~18.991,均P<0.05)。多因素Logistic回归分析结果显示,肿瘤直径、TNM/T分期、淋巴结转移以及铁蛋白4个指标是NSCLC患者并发MVI的独立预测因素(P<0.05),其中铁蛋白每增加1μg/L,并发MVI的风险增加1.234倍(OR:2.234,95%CI:1.456~3.889,P=0.013)。Person相关性分析显示,NSCLC患者血清铁蛋白表达水平与肿瘤直径、CEA,CA125,VGEF呈正相关(r=0.584~0.772,均P<0.05)。ROC分析结果显示,铁蛋白预测NSCLC患者并发MVI的曲线下面积(AUC)为0.734(0.718~0.867),最佳诊断截点为652μg/L,而4个指标的联合预测的价值最高,AUC为0.938(0.868~0.975)。决策曲线分析结果显示,在大多数合理阈值概率范围内,肿瘤直径、TNM/T分期、淋巴结转移以及铁蛋白4个指标预测NSCLC患者并发MVI均具有良好的净收益率,并且联合预测的净收益率高于单一指标。结论血清铁蛋白用于预测NSCLC患者并发MVI具有良好的预测效能和净收益率,并且结合血清铁蛋白的联合预测模型可提高诊断NSCLC患者并发MVI的准确性。

Objective To investigate the predictive value of ferritin on microvascular invasion(MVI)in patients with non-small cell lung cancer(NSCLC).Methods 109 NSCLC patients who received radical surgical treatment in Sichuan Provincial Corps Hospital of Armed Police from January 2020 to December 2021 were included as the study subjects.46 patients(42.20%)in MVI group and 63 patients(57.80%)in N-MVI group were divided according to whether they were complicated with MVI in postoperative pathological examination.The differences in general clinical pathological data and serum ferritin expression level of patients between the two groups were compared.Multivariate logistic regression analysis was used to evaluate the predictive value of serum ferritin in NSCLC patients with MVI,and Person correlation analysis was conducted between ferritin and clinical pathological characteristics of patients.Draw the ROC curve and decision curve of serum ferritin and other independent predictors to predict NSCLC patients with MVI,and further analyze their prediction efficiency and net return.Results The tumor diamete 5.92(3.12,7.65)cm,carcinoembryonic antigen(CEA)73.91±9.25ng/ml,carbohydrate antigen 125(CA125)68.92±10.21U/ml and ferritin 678.93±41.23μg/L in MVI group were higher than those in N-MVI group[4.13(2.12,5.67)cm,64.22±11.37ng/ml,61.34±11.24U/ml,635.94±51.24μg/L],the differences were statistically significant(t/Z=2.765~5.662,all P<0.05).The proportion of multiple tumors(67.39%vs 36.51%),stage IIIa TNM/T(73.01%vs 31.75%),positive vascular endothelial growth factor(VEGF)(65.22%vs 31.75%)and lymph node metastasis(70.00%vs 36.51%)in MVI group was higher than that in N-MVI group,the differences were statistically significant(χ^(2)=10.144~18.991,all P<0.05).Multivariate logistic regression analysis showed that tumor diameter,TNM/T stage,lymph node metastasis and ferritin were independent predictors of NSCLC patients with MVI(P<0.05),for every 1μg/L increase in ferritin,the risk of MVI increased by 1.234 times(OR:2.234,95%CI:1.456~3.889,P=0.013).Person correlation analysis showed that ferritin was significantly positively correlated with tumor diameter,CEA,CA125 and VEGF in patients with NSCLC(r=0.584~0.772,all P<0.05),ROC analysis showed that the area under the curve(AUC)of ferritin for predicting NSCLC patients with MVI was 0.734(0.718~0.867),and the best diagnostic cut-off point was 652μg/L.The value of combined prediction of the four indicators was the highest,with AUC of 0.938(0.868~0.975).The analysis results of decision curve showed that within most reasonable threshold probability ranges,tumor diameter,TNM/T stage,Lymph node metastasis and ferritin four independent prediction indexes had good net benefits in predicting NSCLC patients with MVI,and the overall net benefit of joint prediction was higher than that of single index.Conclusion Serum ferritin could be used to predict MVI in NSCLC patients with good prediction efficiency and net profit rate,and the combined prediction model combined with ferritin can improve the accuracy of the diagnosis of NSCLC patients with MVI.