18F-FLT和18F-FDG PET/CT SUVmax鉴别诊断肺结节的价值

目的 探讨18F-脱氧胸腺嘧啶核苷(FLT)和18F-脱氧葡萄糖(FDG)最大标准摄取值(SUVmax)在肺结节良恶性定性诊断上的价值.方法 自2006年1月至2007年6月,6个PET/CT中心对肺结节样病变开展了18F-FLT和18F-FDG PET/CT显像的多中心临床研究.共有55例患者被纳入该研究[男33例,女22例,年龄(53.3±17.1)岁],所有病例均行2种显像剂显像,并通过病理检查或随访确诊.对2种PET/CT显像结果分别进行视觉分析和半定量分析.采用SPSS13.0软件进行统计学处理.结果 (1)良性病变39例,恶性病变16例.(2)以病理检查或随访结果作为肺癌确诊的"金标准",与"金标准"进行比较,单独以18F-FDG的SUVmax＞2.5和18F-FLT的SUVmax＞1.35为阈值,对肺癌诊断的灵敏度、特异性、准确性分别为93.8%(15/16)、25.6%(10/39)、45.5%(25/55)和93.8%(15/16)、41.O%(16/39)、56.4%(31/55);18F-FDG及18F-FLT双盲和集体阅片诊断肺癌的灵敏度、特异性、准确性分别为87.5%(14/16)、59.0%(23/39)、67.3%(37/55)和68.8%(11/16)、76.9%(30/39)、74.5%(41/55);18F-FDG+18F-FLT双盲和集体阅片诊断肺癌的灵敏度、特异性、准确性分别为81.3%(13/16)、87.2%(34/39)、85.5%(47/55).(3)18F-FDG诊断肺癌的受试者工作特征(ROC)曲线下面积为0.744;18F-FLT诊断肺癌的ROC曲线下面积为0.783.(4)55例病例的18F-FLT、18F-FDG的SUVmax Pearson相关系数(r)=0.614,P＜0.05,表明二者呈线性相关.结论 18F-FDG诊断肺癌的灵敏度较18F-FLT高,但特异性低;18F-FLT的诊断能力较18F-FDG好;二者具有一定的相关性;二者联合诊断可以明显提高诊断准确性.

A multicentre clinical trial on differential diagnosis of pulmonary nodules with PET/CT using 18F-FLT and 18F-FDG

Objective Although 18F-fluorodeoxyglucose (FDG) PET/CT is recognized as a diagnosis of higher sensitivity in lung nodules than other conventional imaging methods, its false positive uptake by some benign lesions handicaps its value in the differential diagnosis of pulmonary nodules. A multicentre clinical trial was recently conducted in order to explore the feasibility of using a dual-tracer 18F-FDG/18F-fluorothymidine (FLT) PET/CT imaging to enhance the diagnosis of lung nodules. Methods Six medical centers participated in the prospective, randomized and blind clinical trial. The studied subjects underwent PET/CT imaging twice using 18F-FLT and 18F-FDG in a random order. The acquisition, collection and process of data were restrictively standardized according to the principle of good clinical practice (GCP). The uptake of the two radiotracers was measured as the maximum standardized uptake value (SUVmax) and the images were interpreted by a collective of readers in different reading strategies. The diagnostic performance of each interpretation strategy was comparatively analysed with the confirmed diagnosis from either pathological evidence or 12-month follow-up. SPSS 13.0 was used to analyse the data. Results From January 2006 to June 2007, 55 patients [33 men, 22 women, mean age (53.3±17.1) year] were collected. In 16 malignant and 39 benign lesions, the sensitivity, specificity, accuracy of 18F-FDG (cutoff SUVmax=2.5) alone were 93.8% (15/16), 25.6% (10/39), 45.5% (25/55), respectively. The corresponding figures of 18F-FLT (cutoff SUVmax=1.35) alone were 93.8%(15/16),41.0%(16/39),56.4%(31/55), respectively. The sensitivity, specificity, accuracy of 18F-FDG in combination of 18F-FLT were 81.3%(13/16),87.2%(34/39),85.5%(47/55), respectively. The areas under the receiver operating characteristic (ROC) curves were 0. 744 for 18F-FDG and 0.783 for 18F-FLT. The 18F-FLT uptake correlated moderately with that of 18F-FDG (r=0.614,P＜0.05). Conclusions The diagnostic specificity of 18F-FLT is superior but sensitivity inferior to that of 18F-FDG in diagnosing lung nodules. Combing the two may increase the accuracy.