低剂量计算机断层扫描在结核病流行国家用于肺癌筛查:系统综述与Meta分析
该系统综述与Meta分析纳入44项研究、477424名个体(59.3%为男性),评估结核病高负担国家采用低剂量计算机断层扫描(LDCT)进行肺癌筛查的筛查阳性率(SPR)与肺癌检出率(LCDR)。合并SPR为18.36%(95%CI:14.6-22.1),合并LCDR为0.94%(95%CI:0.75-1.15);采用NLST标准者的SPR显著高于采用Lung-RADS标准者(25.6% vs 10.4%,p<0.0001),而LCDR相近(0.91% vs 0.95%,p=0.92)。作者认为Lung-RADS标准优于NLST标准,可减少假阳性率。
要点
- 结核病高负担国家LDCT筛查合并SPR为18.36%,合并LCDR为0.94%。
- NLST标准SPR 25.6% vs Lung-RADS标准10.4%(p<0.0001),LCDR无显著差异。
- 作者认为Lung-RADS标准优于NLST标准以减少假阳性率。
核心结论
- 结核病高负担国家LDCT肺癌筛查的SPR与LCDR与结核病发病率较低地区相当。
- 作者认为在结核病高负担国家进行肺癌筛查时,Lung-RADS标准优于NLST标准,以减少假阳性率。
研究设计与样本
- 系统综述与Meta分析,检索建库至2023年12月6日采用LDCT进行肺癌筛查并报告SPR的结核病高负担国家研究。
- 共纳入44项研究、477424名个体,其中59.3%为男性。
- 采用随机效应模型合并数据,并用多因素Meta回归分析影响SPR的因素。
主要发现
- 评估SPR最常用的标准为肺影像报告和数据系统(Lung-RADS)(31%,14项研究)与国家肺癌筛查试验(NLST)标准(非钙化结节≥4 mm;10项研究)。
- 合并SPR为18.36%(95%CI:14.6-22.1),合并LCDR为0.94%(95%CI:0.75-1.15)。
- NLST标准组SPR为25.6%,Lung-RADS标准组为10.4%,NLST标准组显著高于Lung-RADS标准组(p<0.0001)。
- NLST标准组LCDR为0.91%,Lung-RADS标准组为0.95%,两组相近(p=0.92)。
- 多因素Meta回归分析中,使用NLST标准的研究SPR更高。
关键数值与推荐
- Lung-RADS标准用于评估SPR:31%(14项研究)。
- NLST标准(非钙化结节≥4 mm)用于评估SPR:10项研究。
- 合并SPR 18.36%(95%CI:14.6-22.1),合并LCDR 0.94%(95%CI:0.75-1.15)。
- NLST标准SPR 25.6% vs Lung-RADS标准10.4%(p<0.0001);LCDR 0.91% vs 0.95%(p=0.92)。
临床怎么用
- 作者认为在结核病高负担国家进行肺癌筛查时,Lung-RADS标准优于NLST标准,以减少假阳性率。
局限性
- 该分析受显著统计学异质性和发表偏倚的限制。
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)。
原文
INTRODUCTION: Lung cancer screening (LCS) using low-dose computed tomography (LDCT) reduces mortality. Nevertheless, in high tuberculosis-burden countries (HTBC), there are concerns about high false-positive rates due to persistent lung lesions from prior tuberculosis (TB) infections. This study aims to evaluate the screen-positive rate (SPR) of LDCT screening in HTBC. METHODS: We conducted a systematic review and meta-analysis to identify studies utilizing LDCT for LCS in HTBC and reported SPR from inception to December 6, 2023. The primary outcome was the SPR, and the secondary outcome was the lung cancer detection rate (LCDR). The summary data was pooled using a random-effects model, and factors influencing the SPR were analyzed using multivariable meta-regression analysis. RESULTS: A total of 44 studies with 477,424 individuals (59.3% men) were included in the systematic review. Lung Imaging Reporting and Data System (Lung-RADS) (31%, 14 studies) and National Lung Screening Trial (NLST) criteria (non-calcified nodule ≥ 4 mm; 10 studies) were the most common criteria used for assessing SPR. The pooled SPR and LCDR were 18.36% (95% confidence interval [CI]: 14.6-22.1) and 0.94% (95% confidence interval: 0.75-1.15), respectively. Although SPR was significantly higher with NLST criteria than Lung-RADS criteria (25.6% versus 10.4%, p < 0.0001), the LCDR remained similar (0.91% versus 0.95%, p = 0.92). Studies using NLST criteria had a higher SPR in multivariable meta-regression analysis. Nevertheless, the analysis is limited by significant statistical heterogeneity and publication bias. CONCLUSION: Lung cancer screening by LDCT in HTBC demonstrates comparable SPR and LCDR to regions with lower TB incidence rates. Lung-RADS criteria are preferable to NLST criteria for LCS in HTBC to mitigate false-positive rates.
引用
- Damaraju V, Krushna Karri JK, Gandrakota G, et al. Low Dose Computed Tomography for Lung Cancer Screening in Tuberculosis Endemic Countries: A Systematic Review and Meta-Analysis. J Thorac Oncol. 2024. PMID: 39581379. doi:10.1016/j.jtho.2024.11.020
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。