利用胸部CT扫描中的肱骨近端CT值预测骨质疏松与脆性骨折风险:一项开发与验证研究
这项多中心回顾性研究基于542例患者(训练队列397例、验证队列145例)的常规胸部CT,测量肱骨近端CT值并与临床变量联合构建列线图。临床+CT模型预测骨折风险的AUC在训练队列为0.922(95%CI 0.895-0.949)、验证队列为0.914(95%CI 0.861-0.967);诊断骨质疏松的AUC分别为0.804(95%CI 0.761-0.847)和0.800(95%CI 0.722-0.878),均显著优于临床模型。
要点
- 常规胸部CT可机会性测量肱骨近端CT值,无需额外费用或辐射。
- 临床+CT模型预测骨折风险验证队列AUC 0.914,诊断骨质疏松验证队列AUC 0.800。
- 决策曲线分析显示两个终点均有较高净临床获益。
核心结论
- 临床+CT模型整合临床因素与常规胸部CT所测肱骨近端CT值,可有效识别骨质疏松个体和高骨折风险人群,为机会性筛查提供简单有效的工具。
- 模型显示良好可靠性,决策曲线分析表明两个终点均具有较高净临床获益。
研究设计与样本
- 多中心回顾性研究,共分析三家医院542例患者,训练队列397例、验证队列145例。
- 先建立骨质疏松与骨折风险的临床模型作为基线,再从常规胸部CT图像测量肱骨近端CT值,经多变量逻辑回归与临床变量整合构建临床+CT联合模型并列线图。
- 采用受试者工作特征曲线分析、校准曲线和决策曲线分析进行评价。
诊断标准与关键数值
- 骨折风险预测:临床+CT模型训练队列AUC 0.922(95%CI 0.895-0.949),验证队列AUC 0.914(95%CI 0.861-0.967)。
- 骨质疏松诊断:临床+CT模型训练队列AUC 0.804(95%CI 0.761-0.847),验证队列AUC 0.800(95%CI 0.722-0.878)。
- 临床+CT模型在预测骨折风险和诊断骨质疏松方面均显著优于临床模型。
临床应用与阅片要点
- 可利用常规胸部CT图像测量肱骨近端CT值,与临床变量联合用于骨质疏松和脆性骨折风险的机会性筛查。
- 阅片时可关注肱骨近端骨密度CT值,作为机会性评估骨质疏松风险的补充信息。
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
RATIONALE AND OBJECTIVES: To develop a nomogram based on the proximal humerus computed tomography (CT) value derived from routine chest CT scans, enabling opportunistic screening for osteoporosis and assessment of fragility fracture risk without additional cost or radiation exposure.
MATERIALS AND METHODS: This multicenter retrospective study analyzed data from 542 patients across three hospitals (397 in the training cohort, 145 in the validation cohort). A clinical model for osteoporosis and fracture risk was first established as a baseline. Using routine chest CT images, the proximal humerus CT value was measured and integrated with clinical variables via multivariable logistic regression to construct a combined Clinical + CT model. The nomogram was evaluated using receiver operating characteristic (ROC) curve analysis, calibration curves, and decision curve analysis.
RESULTS: The Clinical + CT model significantly outperformed the clinical model in predicting fracture risk and diagnosing osteoporosis. For fracture risk prediction, the Clinical + CT model achieved the area under the curve (AUC) of 0.922 (95% CI: 0.895-0.949) in the training cohort and 0.914 (95% CI: 0.861-0.967) in the validation cohort. For osteoporosis diagnosis, the AUCs were 0.804 (95% CI: 0.761-0.847) and 0.800 (95% CI: 0.722-0.878), respectively.
CONCLUSION: The Clinical + CT model, which integrates clinical factors with proximal humerus CT values from routine chest CT scans, can effectively identify individuals with osteoporosis and high fracture risk, offering a simple and effective tool for opportunistic screening. The models demonstrate good reliability, and decision curve analysis indicates a high net clinical benefit for both endpoints.
引用
- Zhao CX, Wang RA, Feng X, et al. Prediction of Osteoporosis and Fragility Fracture Risk Using Proximal Humerus CT Value from Chest CT: A Development and Validation Study. Acad Radiol. 2026. PMID: 41887997. doi:10.1016/j.acra.2026.03.001
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。