声学造影平均渡越时间和拟合度综合评分诊断肝癌的价值研究

Value of a novel score system composited from hepatic mean transit time and fitting degree through contrast ultrasound in differential diagnosis of liver cancer

目的建立经静脉声学造影测定平均渡越时间和拟合度综合评分的方法,探讨新的参数评价声学造影诊断原发性肝癌的价值。方法研究对象为58例肝脏占位性疾病的患者(男35,女23),平均年龄(55.6±11.8)岁。经手术证实的肝癌患者20例,其中内径小于3 cm者8例;其余病例经MRI检查或病理穿刺证实为非肝癌患者,其中包括肝血管瘤9例、肝硬化结节11例、局灶性增生结节4例、乙型肝炎9例和其他病变5例。经静脉声学造影所采用的造影剂为声诺维,超声造影动态影像经脱机分析获得肝肿瘤和周围肝实质感兴趣区平均渡越时间和拟合度,计算平均渡越时间和拟合度预测肝癌的风险指数、J点、敏感性和特异性,并进行综合评分。结果肝癌患者的平均渡越时间[(32.8±21.5)s]较其他肝脏肿块[(61.4±45.6)s]短(P<0.01),肝癌患者拟合度[(73.0±15.5)%],高于非肝癌患者[(50.7±21.1)%](P<0.05)。平均渡越时间、渡越时间参照系数、拟合度和拟合度参照系数用于预测肝癌的J点分别为≤49.4 s,≤-8.8 s,>48.3%和>62.1%。所有患者的平均综合评分为(1.71±1.43)分。肝癌组的平均综合评分为(3.05±1.10)分,非肝癌组为(1.00±1.04)分,两者差异有统计学意义(P<0.01)。综合评分≥2预测肝癌的敏感性和特异性分别为95%和74%(AUC=0.90),预测小肝癌的敏感性和特异性分别为75%和86%(AUC=0.88)。单元逻辑回归分析显示平均渡越时间、拟合度和拟合度参照系数的风险指数升高均有统计学意义(P<0.05);单元逻辑回归分析显示,综合评分≥2预测肝癌较综合评分<2的可能性大7倍(odds ratio=7.14,P<0.01)。结论肝声学造影平均渡越时间和拟合度综合评分用于预测原发性肝癌有一定的临床应用价值,有助于与肝良性病变进行鉴别诊断。

Objective To establish a new score system for differential diagnosis of primary liver cancer through assessment of mean transmit time（ MTT） and fitting degree by contrast ultrasound. Methods The study subjects were 58 patients with space occupying disease in the liver（ the number of patients,male = 35,female = 23; mean age 55. 6 ± 11. 8 years）. Of all the cases,20 were liver cancer patients confirmed through surgery,including 8 patients with internal diameters of tumors 〈3 cm. The remaining 38 cases with other liver diseases confirmed by MRI or biopsy,of which 9 cases were hepatic hemangioma,11 were cirrhotic nodules,4 were focal nodular hyperplasia,9 were hepatitis B,and 5 were others pathologic changes. SonoVue was used as an ultrasound contrast agent.MTT and fitting degree in the region of interest were measured by the software contained in the Siemens S2000 Ultrasound Diagnostic System. The risk factor,J point,sensitivity and specificity of liver cancer were predicted through the calculation of mean transit time and fitting degree. Results MTT in liver cancer patients（ 32. 8 ± 21. 5） s was shorter than that of the other patients with non-hepatic tumors（ 61. 4 ± 45. 6） s（ P〈0. 01）. Mean fitting degree in the liver cancer patients（ 73. 0 ± 15. 5） % was higher than that of the nonliver cancer patients（ 50. 7 ± 21. 1） %（ P〈0. 05）. Mean transit time of MTT,reference value of MTT,fitting degree and reference value of fitting degree used to predict J point of liver cancer were ≤ 49. 4s,≤-8. 8s, 〉48. 32 and 〉62. 1,respectively. Mean composite scores of all the patients were 1. 71 ± 1. 43. Mean composite scores of the patients in the liver cancer group were（ 3. 05 ± 1. 10）and the scores of the patients in the non-liver cancer group were（ 1. 00 ± 1. 04）,and very significant differences could be seen,when comparisons were made between them（ P〈0. 01）. The sensitivity and specificity scores ≥ 2 used to predict liver cancer were 95%and 74%（ AUC = 0. 90） respectively. The sensitivity and specificity scores ≥ 2 used to predict liver cancer were 75% and 86%（ AUC =0.88） respectively. Unit logic regression analysis indicated that the elevated scores of the risk factor associated with mean transit time of MTT,fitting degree and reference value of fitting degree were all statistically significant（ P〈0. 05）. Unit logic regression analysis also indicated that the odds ratio of the scores ≥ 2 used to predict liver cancer was 7. 14 and the prevalence of liver cancer associated with it might be as high as 7-folds,as compared to the scores 〈2（ P〈0. 01）. Conclusion Mean transit time and composite scores of fitting degree used to predict primary liver cancer had certain clinical value,which would help to differentiate liver cancer from benign hepatic pathological changes,such as hemangioma,cirrhotic nodules and focal nodular hyperplasia.