非免疫缺陷患者肺真菌病的CT表现

CT findings of cases diagnosed with pulmonary fungal disease in non-immunosuppressed patients

目的分析总结3种肺真菌病的cT表现，以提高对此病的诊断水平。方法回顾性分析2011年1月至2015年10月肺曲霉菌病48例，肺隐球菌病45例，肺毛霉菌病16例共109例均经病理确诊的肺真菌病的CT表现。患者均行胸部MSCT检查，分析患者基础疾病、病变类型、病变数目分布及大小分布、CT伴随征象。基础疾病、病变类型及各类不同肺真菌病间的cT伴随征象间差异比较采用χ^2检验、Fisher确切概率法。病变数目分布、大小分布情况差异采用Kruskal-WallisH检验。采用ROC曲线分析联合征象预测真菌类型的预测概率。结果糖尿病伴发肺毛霉菌病的发生率高于曲霉菌、隐球菌（χ^2值分别为22．704、17．509，P值均〈0．017），无基础疾病患者伴发肺毛霉菌病的发生率低于曲霉菌、隐球菌（Fisher确切概率法，P值均〈0．017）。3种肺真菌病均以结节肿块为主要表现，实变型及磨玻璃影型表现少见。肺曲霉菌病结节肿块型病变CT伴随征象树芽征、空气支气管征、空洞或空腔、空气新月征、晕轮征、反晕轮征分别为18、10、19、10、23、0例；肺隐球菌病分别为4、14、6、0、11、0例；肺毛霉菌病分别为2、0、4、0、g、6例。结节肿块型病变中曲霉菌病与毛霉菌病的病变数目分布不同（P〈0．05）。毛霉菌病与曲霉菌病、隐球菌病的病变大小分布不同（P值均〈0．05）。结节肿块型病变中，肺曲霉菌病CT伴随征象树芽征、空洞或空腔、空气新月征发生率高于肺隐球菌病（χ^2值分别为9．972、21．841、8．047，P值均〈0．017），肺毛霉菌反晕轮征发生率均高于肺曲霉菌与肺隐球菌（χ^2连续校正值为23．936、21．374，P值均〈0．017）。联合征象预测真菌病类型的ROC曲线下面积（AUC）为0．819。结论肺曲霉菌病、肺隐球菌病及肺毛霉菌病均以结节肿块型为常见。CT表现、病变的大小分布及数目分布、CT伴随征象发生隋况对于肺真菌病有较高的诊断及鉴别诊断价值。

Objective To investigate the CT findings of three types of pulmonary fungal disease in non-immunosuppressed patients. Methods We retrospectively collected 109 cases diagnosed with pulmonary fungal disease with pathological evidence which between January 2011 and October 2015 in the First Affiliated Hospital of Zhengzhou University, among which pulmonary aspergillosis were 48 cases, pulmonary cryptococcosis 45 cases and pulmonary nmcormycosis 16 cases. Each patient underwent a chest MSCT scan. The data were analyzed that including underlying diseases, lesion types, lesion size, number, accompanying signs of CT scan, etc. X2 test or Fisher exact test were used when compared underlying disease, lesion types, various of CT signs ere between different types of pulmonary fungal disease. Kruskal-Wallis H test were used when compared the incidence rate of lesions numbers and diameter, The ROC curve was used to analyze probability of predicting the fungal types by combined signs. Results The incidence of pulmonary mucormycosis with diabetes was higher than pulmonary aspergillosis and pulmonary cryptococcosis（χ^2=2.704, 17.509, P〈0.017） .The incidence of pulmonary mucormycosis with no underlying disease was lower than pulmonary aspergillosis and pulmonary cryptococcosis（Fisher test, P〈0.017）. Nodular or mass were main CT findings in the three pulmonary fungal disease. Consolidation, ground glass opacity and other manifestations were rare. Nodular or mass lesions in pulmonary aspergillosis accompany with CT signs like tree-in-bud, bronchogram sign, cavities, the air crescent sign, halo sign, the reversed-halo sign were rspectively 18, 10, 19, 10, 23, 0 cases; To pulmonary cryptococcosis were 4, 14, 6, 0, 11, 0 cases. To pulmonary mucormycosis were 2, 0, 4, 0, 4, 6 cases. Lesion number＇s distribution had a statistically difference between pulmonary aspergillosis and pulmonary cryptococcosis in nodular or mass lesions （P〈 0.05）. Lesions size＇s distribution had statistically differences respectively between pulmonary mucormycosis and pulmonary aspergillosis, pulmonary cryptococcosis （P〈0.05）. Nodular lesions of pulmonary aspergillosis had higher incidence of tree-in-bud, cavity and air-crescent compared to pulmonary cryptococcosis （χ^2=9.972, 21.841,8.047, P〈0.017）. Reversed halo sign was mostly seen in pulmonary mucormycosis than in pulmonary aspergillosis and pulmonary cryptococcosis （Continuous correction χ^2=23.936,21.374, P〈0.017）. The area under the ROC curve for the combined signs predictive fungal disease type was 0.819. Conclusions Nodules or masses are the most common image findings in pulmonary aspergillosis, pulmonary cryptococcosis and pulmonary mucormycosis. The distribution of size and numbers and CT accompanied signs have a higher diagnostic efficacy in diagnosing and distinguishing pulmonary fungal disease.