论文部分内容阅读
目的:探讨基于n 18F-脱氧葡萄糖(FDG)PET/CT肿瘤原发灶代谢参数对非小细胞肺癌(NSCLC)隐匿性淋巴结转移(OLM)的预测价值。n 方法:回顾性分析2013年1月至2018年12月于苏州大学附属第三医院行n 18F-FDG PET/CT检查并于术前诊断为临床N0期(cN0)的183例患者[男72例，女111例，年龄(61.5±8.4)岁]，所有患者于检查后3周内于本院行原发灶切除术及系统性淋巴结清扫术，依据术后有无淋巴结转移分为OLM阳性(OLM+)组及OLM阴性(OLM-)组。分析获得n 18F-FDG PET/CT原发灶相关参数：原发灶最大径(Dn max)、原发灶位置、形态学特征、原发灶最大标准摄取值(SUVn max)、平均标准摄取值(SUVn mean)、肿瘤代谢体积(MTV)、病灶糖酵解总量(TLG)、原发灶SUVn max-肝SUVn mean比值(TLRn max)、原发灶TLG-肝SUVn mean比值(TLRn TLG)。运用Mann-Whitney n U检验、n χ2检验对各参数进行组间比较，采用logistic回归分析OLM的独立危险因素，采用受试者工作特征(ROC)曲线分析评估不同参数的诊断效能。n 结果:183例患者中，25例(13.7%，25/183)存在OLM，共46枚淋巴结转移(N1淋巴结15枚，N2淋巴结31枚)。OLM+组原发灶Dn max[2.9(2.3，3.7)与2.3(1.7，2.8) cm]、分叶征[76.0%(19/25)与37.3%(59/158)]、原发灶SUVn max[11.1(7.9，17.7)与4.7(2.3，9.2)]、TLG[41.5(10.2，91.1)与15.6(6.5，23.8) ml]、TLRn max[4.7(3.5，7.6)与2.1(0.9，4.0)]、TLRn TLG[18.1(5.0，44.3)与6.1(3.0，11.4) ml]均高于OLM-组(n z值：-4.709～-3.247, n χ2＝13.190，均n P<0.05)。多因素logistic回归分析显示TLRn max[比值比(n OR)＝15.145，95% n CI：3.381～67.830，n P<0.001]和Dn max(n OR＝3.220，95% n CI：1.192～8.701，n P＝0.021)是OLM的独立危险因素。TLRn max预测OLM ROC曲线下面积(AUC)最大(0.794)，其阈值为3.12时诊断OLM的灵敏度、特异性、准确性、阳性预测值和阴性预测值分别为92.0%(23/25)、63.3%(100/158)、67.2%(123/183)、28.4%(23/81)和98.0%(100/102)。n 结论:18F-FDG PET/CT原发灶代谢参数TLRn max是NSCLC患者OLM的独立危险因素，其可灵敏预测NSCLC患者的OLM。n “,”Objective:To investigate the predictive value of n 18F-fluorodeoxyglucose (FDG) PET/CT metabolic parameters for occult lymph node metastasis (OLM) in non-small cell lung cancer (NSCLC).n Methods:A total of 183 patients (72 males, 111 females; age (61.5±8.4) years) who underwent n 18F-FDG PET/CT and preoperatively diagnosed with clinical N0 stage (cN0) in Third Affiliated Hospital of Soochow University from January 2013 to December 2018 were retrospectively enrolled. All patients underwent anatomical pulmonary resection with systematic lymph node dissections within 3 weeks after n 18F-FDG PET/CT examinations. According to the presence or absence of lymph node metastasis, patients were divided into OLM positive (OLM+ ) group and OLM negative (OLM-) group. Parameters of primary lesions, such as the maximum diameter (Dn max), tumor sites, morphological features, maximum standardized uptake value (SUVn max), mean standardized uptake value (SUVn mean), metabolic total volume (MTV), total lesion glycolysis (TLG), tumor SUVn max to liver SUVn mean (TLRn max), tumor TLG to liver SUVn mean (TLRn TLG) were analyzed. Mann-Whitney n U test and n χ2 test were used to compare the parameters between groups. Multivariable logistic regression was used to analyze the independent risk factors for OLM. Receiver operating characteristic (ROC) curve analysis was used to evaluate the diagnostic value of different parameters.n Results:Among 183 patients, 25 (13.7%, 25/183) of them were diagnosed as OLM. In OLM+ group, 46 lymph nodes were pathologically positive for metastasis, including 15 N1 disease and 31 N2 disease. Dn max (2.9(2.3, 3.7) n vs 2.3(1.7, 2.8) cm), lobulation ((76.0%(19/25) n vs 37.3%(59/158)), SUVn max (11.1(7.9, 17.7) n vs 4.7(2.3, 9.2)), TLG (41.5(10.2, 91.1) n vs 15.6(6.5, 23.8) ml), TLRn max (4.7(3.5, 7.6) n vs 2.1(0.9, 4.0)) and TLRn TLG (18.1(5.0, 44.3) n vs 6.1(3.0, 11.4) ml) of the primary lesions in OLM+ group were significantly higher than those in OLM-group (n z values: from -4.709 to -3.247, n χ2=13.190, all n P<0.05). Multivariable logistic regression analysis showed that TLRn max (odds ratio (n OR)=15.145, 95% n CI: 3.381-67.830, n P<0.001) and Dn max (n OR=3.220, 95% n CI: 1.192-8.701, n P=0.021) were independent risk factors for OLM. TLRn max yielded the highest area under curve (AUC; AUC=0.794) with the threshold of 3.12, and the sensitivity, specificity, accuracy, positive predictive value and negative predictive value for predicting OLM were 92.0%(23/25), 63.3%(100/158), 67.2%(123/183), 28.4%(23/81) and 98.0%(100/102), respectively.n Conclusions:TLRn max of tumor is the independent risk factor for OLM in NSCLC patients. TLRn max can sensitively predict OLM preoperatively in patients with NSCLC.n