（18）F-氟脱氧葡萄糖在162例无恶性肿瘤病史患者偶然发现的小肺结节中的诊断效能

Diagnostic Performance of (18)F-Fluorodeoxyglucose in 162 Small Pulmonary Nodules Incidentally Detected in Subjects Without a History of Malignancy.

作者信息

Calcagni Maria Lucia, Taralli Silvia, Cardillo Giuseppe, Graziano Paolo, Ialongo Pasquale, Mattoli Maria Vittoria, Di Franco Davide, Caldarella Carmelo, Carleo Francesco, Indovina Luca, Giordano Alessandro

机构信息

Institute of Nuclear Medicine, Università Cattolica del Sacro Cuore, Rome, Italy.

出版信息

Ann Thorac Surg. 2016 Apr;101(4):1303-9. doi: 10.1016/j.athoracsur.2015.10.072. Epub 2016 Jan 26.

DOI:10.1016/j.athoracsur.2015.10.072

PMID:26822348

Abstract

BACKGROUND

Solitary pulmonary nodule (SPN) still represents a diagnostic challenge. The aim of our study was to evaluate the diagnostic performance of (18)F-fluorodeoxyglucose positron emission tomography-computed tomography in one of the largest samples of small SPNs, incidentally detected in subjects without a history of malignancy (nonscreening population) and undetermined at computed tomography.

METHODS

One-hundred and sixty-two small (>0.8 to 1.5 cm) and, for comparison, 206 large nodules (>1.5 to 3 cm) were retrospectively evaluated. Diagnostic performance of (18)F-fluorodeoxyglucose visual analysis, receiver-operating characteristic (ROC) analysis for maximum standardized uptake value (SUVmax), and Bayesian analysis were assessed using histology or radiological follow-up as a golden standard.

RESULTS

In 162 small nodules, (18)F-fluorodeoxyglucose visual and ROC analyses (SUVmax = 1.3) provided 72.6% and 77.4% sensitivity and 88.0% and 82.0% specificity, respectively. The prevalence of malignancy was 38%; Bayesian analysis provided 78.8% positive and 16.0% negative posttest probabilities of malignancy. In 206 large nodules (18)F-fluorodeoxyglucose visual and ROC analyses (SUVmax = 1.9) provided 89.5% and 85.1% sensitivity and 70.8% and 79.2% specificity, respectively. The prevalence of malignancy was 65%; Bayesian analysis provided 85.0% positive and 21.6% negative posttest probabilities of malignancy. In both groups, malignant nodules had a significant higher SUVmax (p < 0.0001) than benign nodules. Only in the small group, malignant nodules were significantly larger (p = 0.0054) than benign ones.

CONCLUSIONS

(18)F-fluorodeoxyglucose can be clinically relevant to rule in and rule out malignancy in undetermined small SPNs, incidentally detected in nonscreening population with intermediate pretest probability of malignancy, as well as in larger ones. Visual analysis can be considered an optimal diagnostic criterion, adequately detecting a wide range of malignant nodules with different metabolic activity.

摘要

背景

孤立性肺结节（SPN）仍是一个诊断难题。我们研究的目的是在最大样本量的小SPN中评估（18）F - 氟脱氧葡萄糖正电子发射断层扫描 - 计算机断层扫描的诊断性能，这些小SPN是在无恶性肿瘤病史的受试者（非筛查人群）中偶然发现的，且在计算机断层扫描中未明确诊断。

方法

回顾性评估了162个小（>0.8至1.5 cm）结节，并选取206个大结节（>1.5至3 cm）作为对照。以组织学或影像学随访作为金标准，评估（18）F - 氟脱氧葡萄糖视觉分析、最大标准化摄取值（SUVmax）的受试者操作特征（ROC）分析以及贝叶斯分析的诊断性能。

结果

在162个小结节中，（18）F - 氟脱氧葡萄糖视觉分析和ROC分析（SUVmax = 1.3）的敏感性分别为72.6%和77.4%，特异性分别为88.0%和82.0%。恶性肿瘤的患病率为38%；贝叶斯分析得出恶性肿瘤的阳性和阴性检验后概率分别为78.8%和16.0%。在206个大结节中，（18）F - 氟脱氧葡萄糖视觉分析和ROC分析（SUVmax = 1.9）的敏感性分别为89.5%和85.1%，特异性分别为70.8%和79.2%。恶性肿瘤的患病率为65%；贝叶斯分析得出恶性肿瘤的阳性和阴性检验后概率分别为85.0%和21.6%。在两组中，恶性结节的SUVmax均显著高于良性结节（p < 0.0001）。仅在小样本组中，恶性结节显著大于良性结节（p = 0.0054）。