Department of Radiology, Seoul National University College of Medicine, and Institute of Radiation Medicine, Seoul National University Medical Research Center, 101 Daehangno, Jongno-gu, Seoul 110-744, Korea.
Radiology. 2010 Apr;255(1):242-51. doi: 10.1148/radiol.09090547. Epub 2010 Feb 19.
To retrospectively investigate clinical and computed tomographic (CT) features of transient part-solid nodules (PSNs) initially detected at screening thin-section CT for lung cancer and to determine predictive factors that may differentiate transient PSNs from persistent PSNs.
This study was approved by the institutional review board. From January 2006 to August 2008, 93 individuals with 126 PSNs were identified from among 16777 individuals who underwent chest CT. Clinical features and CT characteristics of PSNs were reviewed, and clinical and thin-section CT features were compared between transient and persistent PSNs. To identify predictive factors of transient PSNs and evaluate predictive performance, logistic regression analysis and C statistic analysis were performed.
Eighty-eight (69.8%) of 126 PSNs were transient. Between transient and persistent PSNs, there were significant differences (P < .05) in patient age, patient sex, risk of lung cancer, presence of eosinophilia, mode of detection, lesion size, lesion multiplicity, size of solid portion, and lesion border. Multivariate analysis revealed that young patient age, detection of the lesion at follow-up, blood eosinophilia, lesion multiplicity, large solid portion, and ill-defined border were significant (P < .05) independent predictors of transient PSNs. The performance in the discrimination of transient PSNs from persistent PSNs of the logistic regression model that incorporated both clinical and thin-section CT features was significantly higher than the performance of the models that incorporated clinical features or thin-section CT features alone.
A substantial proportion of PSNs detected at screening CT were transient. Transient PSNs could be predicted with high accuracy by using the features of young patient age, detection of the PSN at follow-up, blood eosinophilia, lesion multiplicity, large solid portion, and ill-defined lesion border.
回顾性分析在肺癌筛查性低剂量 CT 中首次发现的一过性部分实性结节（PSN）的临床和 CT 特征，并确定可能区分一过性 PSN 和持续性 PSN 的预测因素。
本研究经机构审查委员会批准。2006 年 1 月至 2008 年 8 月，从 16777 例行胸部 CT 检查的个体中，共发现 93 名个体存在 126 个 PSN。对 PSN 的临床特征和 CT 特征进行了回顾性分析，并比较了一过性和持续性 PSN 的临床和薄层 CT 特征。为了确定一过性 PSN 的预测因素并评估预测性能，进行了逻辑回归分析和 C 统计分析。
126 个 PSN 中 88 个（69.8%）为一过性。在一过性和持续性 PSN 之间，患者年龄、患者性别、肺癌风险、嗜酸性粒细胞增多、检出方式、结节大小、结节数量、实性部分大小、结节边界等方面存在显著差异（P<0.05）。多变量分析显示，患者年龄较小、在随访中检出病变、血嗜酸性粒细胞增多、结节数量较多、实性部分较大、边界模糊是一过性 PSN 的显著（P<0.05）独立预测因素。纳入临床和薄层 CT 特征的逻辑回归模型在区分一过性 PSN 和持续性 PSN 方面的表现明显优于仅纳入临床或薄层 CT 特征的模型。
在筛查性 CT 中发现的 PSN 中，相当一部分是一过性的。通过使用患者年龄较小、在随访中检出病变、血嗜酸性粒细胞增多、结节数量较多、实性部分较大、边界模糊等特征，可以高度准确地预测一过性 PSN。
