韩国国家肺癌筛查项目中S修饰符的患病率、共现模式及预后意义
该回顾性研究纳入2019年8月至2020年12月韩国国家肺癌筛查项目的125600名基线筛查参与者,评估七种标准化S修饰符的患病率、全因死亡关联及共现模式。冠脉钙化最常见(15.07%),其次为肺气肿(13.77%)和间质性肺异常(2.65%);所有S修饰符均与全因死亡升高相关,潜类别分析识别出四种死亡风险逐步升高的类别。
要点
- 冠脉钙化(15.07%)与肺气肿(13.77%)是低剂量CT肺癌筛查中最常见的S修饰符。
- 胸膜和/或心包积液调整后HR为8.28(95%CI:5.48,12.51),与全因死亡关联最强。
- 潜类别分析从孤立性肺气肿(调整后HR 1.22)到高风险修饰符(调整后HR 5.35)呈阶梯式死亡风险升高。
核心结论
- 在全国性肺癌筛查人群中,使用七种标准化S修饰符的结构化报告揭示了其患病率及与全因死亡的关联,验证了其在识别临床显著异常方面的临床效用。
研究设计与样本
- 回顾性研究,纳入2019年8月至2020年12月韩国国家肺癌筛查项目的125600名基线筛查参与者(平均年龄62岁±5.3,123331名男性),采用多变量Cox回归评估全因死亡预后价值,并行潜类别分析识别共现模式。
关键影像表现/征象
- 冠脉钙化最常见,患病率15.07%(18892/125366)。
- 肺气肿患病率13.77%(17300/125600),间质性肺异常患病率2.65%(3324/125600),肺部感染患病率0.90%(1123/124477)。
- 肺外恶性肿瘤(74/125257)、主动脉瘤(78/125256)、胸膜和/或心包积液(75/125253)各自观察到不足0.1%。
诊断标准与关键数值
- 胸膜和/或心包积液调整后HR为8.28(95%CI:5.48,12.51),肺外恶性肿瘤为3.58(95%CI:1.97,6.49),主动脉瘤为3.28(95%CI:1.71,6.32)。
- 间质性肺异常调整后HR为2.16(95%CI:1.89,2.47),冠脉钙化为1.41(95%CI:1.30,1.53),肺气肿为1.15(95%CI:1.05,1.25),除肺气肿P=0.002外其余均P<0.001。
- 潜类别分析识别四类,从孤立性肺气肿(调整后HR 1.22;95%CI:1.10,1.36;P<0.001)到高风险修饰符(调整后HR 5.35;95%CI:3.40,8.41;P<0.001)死亡风险阶梯式升高。
临床应用与阅片要点
- 韩国国家肺癌筛查项目采用结构化报告记录七种标准化S修饰符。
- 中位随访3.7年期间,2.69%(3380例)参与者死亡。
局限性
- 原文未报告局限性。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
Background The documentation of clinically significant incidental findings (S modifiers) in low-dose CT lung cancer screening varies among radiologists. Although the Korean National Lung Cancer Screening Program adopted structured reporting for seven standardized S modifiers, the prognostic value of standardized S modifiers has not been evaluated comprehensively. Purpose To evaluate the implementation of structured reporting for prespecified S modifiers by analyzing their prevalence, mortality associations, and co-occurrence patterns. Materials and Methods This retrospective study included baseline screening participants from the Korean National Lung Cancer Screening Program between August 2019 and December 2020. The prevalence of seven S modifiers was calculated, and their prognostic value for all-cause mortality was assessed using multivariable Cox regression. Latent class analysis (LCA) was performed to identify co-occurrence patterns, which were analyzed for mortality risk stratification. Results Among 125 600 participants (mean age ± SD, 62 years ± 5.3; 123 331 men), 2.69% (n = 3380) died over a median follow-up of 3.7 years. Coronary artery calcification was most prevalent (15.07% [18 892 of 125 366 participants]), followed by emphysema (13.77% [17 300 of 125 600 participants]), interstitial lung abnormalities (ILAs) (2.65% [3324 of 125 600 participants]), and pulmonary infection (0.90% [1123 of 124 477 participants]). Extrapulmonary malignancy (74 of 125 257 participants), aortic aneurysm (78 of 125 256 participants), and pleural and/or pericardial effusion (75 of 125 253 participants) were each observed in less than 0.1% of participants. All S modifiers were associated with increased all-cause mortality, with adjusted hazard ratios (HRs) of 8.28 (95% CI: 5.48, 12.51) for pleural and/or pericardial effusion, 3.58 (95% CI: 1.97, 6.49) for extrapulmonary malignancy, 3.28 (95% CI: 1.71, 6.32) for aortic aneurysm, 2.16 (95% CI: 1.89, 2.47) for ILAs, 1.41 (95% CI: 1.30, 1.53) for coronary artery calcification, and 1.15 (95% CI: 1.05, 1.25) for emphysema (P < .001 for all except for emphysema, with P = .002). LCA helped identify four distinct classes with a stepwise increase in mortality from isolated emphysema (adjusted HR, 1.22; 95% CI: 1.10, 1.36; P < .001) to high-risk modifiers (adjusted HR, 5.35; 95% CI: 3.40, 8.41; P < .001). Conclusion In a nationwide lung cancer screening group, structured reporting using seven standardized S modifiers revealed both their prevalence and associations with all-cause mortality, validating their clinical utility in identifying clinically significant abnormalities. © RSNA, 2026 Supplemental material is available for this article. See also the editorial by White and Gierada in this issue.
引用
- Kim H, Jo E, Kim J, et al. Prevalence, Co-occurrence, and Prognostic Implications of S Modifiers in the Korean National Lung Cancer Screening Program. Radiology. 2026. PMID: 41493277. doi:10.1148/radiol.251656
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。