[18F]F-FAPI-FUSCC-07 PET/CT 鉴别孤立性肺结节:头对头比较
FAPI示踪剂TBR的AUC显著高于FDG(0.801 vs. 0.677,p<0.05)。联合FAPI摄取与CT分叶征的预测模型在训练集AUC为0.866,独立验证集准确率81.08%。
要点
- FAPI示踪剂TBR诊断效能显著优于FDG(AUC 0.801 vs. 0.677)。
- 联合TBRFAPI与分叶征的模型训练集AUC达0.866。
- 验证集模型灵敏度87.50%、特异度69.23%、准确率81.08%。
核心结论
- [18F]F-FAPI-FUSCC-07 PET/CT鉴别良恶性孤立性肺结节优于[18F]F-FDG PET/CT。
- 联合FAPI功能参数与CT形态特征的预测模型在训练集和独立验证集均表现出满意鉴别能力。
研究设计与样本
- 采用ROC分析比较两种示踪剂半定量参数(SUVmax和TBR)的诊断效能。
- 多因素logistic回归模型在训练队列(n=100)构建,在独立队列(n=37)验证。
主要发现
- 两种示踪剂PET/CT均能有效鉴别良恶性孤立性肺结节。
- FAPI示踪剂TBR的AUC为0.801,显著高于FDG的0.677(p<0.05)。
- 联合TBRFAPI与分叶征的模型公式为P=1/(1+e^(-x)),x=-1.223+0.502×TBRFAPI+1.959×lobulation。
- 模型训练集AUC为0.866;验证集灵敏度87.50%、特异度69.23%、准确率81.08%、阳性预测值84.00%、阴性预测值75.00%。
- 探索性分析显示浸润性黏液腺癌的[18F]F-FAPI-FUSCC-07摄取显著低于其他亚型。
关键数值与推荐
- FAPI示踪剂TBR的AUC为0.801,FDG示踪剂TBR的AUC为0.677(p<0.05)。
- 预测模型训练集AUC为0.866。
- 验证集灵敏度87.50%、特异度69.23%、准确率81.08%、阳性预测值84.00%、阴性预测值75.00%。
临床怎么用
- [18F]F-FAPI-FUSCC-07 PET/CT可作为鉴别良恶性孤立性肺结节的非侵入性影像工具。
- 联合TBRFAPI与CT分叶征的预测模型有助于提高诊断准确性。
- 浸润性黏液腺癌FAPI摄取较低,判读时需注意。
局限性
- 回顾性研究设计。
- 结论需前瞻性研究进一步验证。
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
OBJECTIVE: To evaluate the diagnostic efficacy of a novel FAP-targeted tracer, [¹⁸F]F-FAPI-FUSCC-07, for solitary pulmonary nodules (SPNs) and to develop a reliable prediction model by integrating PET functional parameters with CT morphological features. METHODS: One hundred and thirty-seven patients with SPNs who underwent both [¹⁸F]F-FAPI-FUSCC-07 and [¹⁸F]F-FDG PET/CT were retrospectively enrolled in this study. Diagnostic performance of semi-quantitative parameters (SUVmax and TBR) for both tracers was evaluated and compared using ROC analysis. A multivariate logistic regression model was constructed in a training cohort (n = 100) and validated in an independent cohort (n = 37). RESULTS: Both [¹⁸F]F-FAPI-FUSCC-07 and [¹⁸F]F-FDG PET/CT were able to discriminate between benign and malignant SPNs effectively. Compared with [¹⁸F]F-FDG PET/CT, [¹⁸F]F-FAPI-FUSCC-07 PET/CT demonstrated significantly higher diagnostic accuracy (AUC for TBR of both tracers: 0.801 vs. 0.677, p < 0.05). To better leverage the advantages of [¹⁸F]F-FAPI-FUSCC-07 PET/CT, a diagnostic model combining FAPI uptake and CT morphological features was constructed using logistic regression. The model was formulated as P = 1 / (1 + e^(-x)), where P represents the probability of malignancy, x = -1.223 + 0.502 × TBRFAPI + 1.959 × lobulation. The diagnostic model achieved superior performance with an AUC of 0.866. In the validation set, the sensitivity, specificity, accuracy, positive predictive value, and negative predictive value of the model were 87.50%, 69.23%, 81.08%, 84.00%, and 75.00%, respectively. Exploratory analysis revealed significantly lower [¹⁸F]F-FAPI-FUSCC-07 uptake in invasive mucinous adenocarcinoma than in other subtypes. CONCLUSION: [¹⁸F]F-FAPI-FUSCC-07 PET/CT is a superior imaging tool for discriminating benign from malignant SPNs compared to [¹⁸F]F-FDG PET/CT. A prediction model combining its functional parameters with CT morphological features achieved satisfactory discriminatory ability (AUC: 0.866) in the training set and maintained good accuracy (81.08%) in an independent validation set, providing a promising non-invasive diagnostic strategy that warrants further validation in prospective studies.
引用
- Cheng Y, Li N, Zhu B, et al. Diagnostic performance of [18F]F-FAPI-FUSCC-07 PET/CT in characterizing solitary pulmonary nodules: a head-to-head comparison. Eur J Nucl Med Mol Imaging. 2026. PMID: 41699281. doi:10.1007/s00259-026-07819-0
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。