基于光谱CT的列线图术前预测局部进展期胃癌Lauren分型:一项前瞻性研究
本研究前瞻性纳入2023年6月至12月经术后病理诊断的局部进展期胃癌患者,行腹部三期增强光谱CT检查,基于性别、动脉期标准化碘浓度和延迟期能谱曲线斜率构建列线图,术前预测Lauren分型,曲线下面积0.841(95%CI 0.749-0.932),特异度85.3%,准确率76.4%,敏感度68.4%。
要点
- 性别、动脉期标准化碘浓度和延迟期能谱曲线斜率是Lauren分型的独立预测因子。
- 基于上述指标的列线图预测Lauren分型AUC为0.841,特异度85.3%,准确率76.4%,敏感度68.4%。
- 该列线图可能有助于术前无创评估Lauren分型,指导个体化治疗策略和风险分层。
核心结论
- 基于临床特征和光谱CT定量参数的列线图在局部进展期胃癌Lauren分型的术前无创评估方面展现出较大潜力。
研究设计与样本
- 前瞻性研究,纳入2023年6月至12月经术后病理诊断为局部进展期胃癌且行腹部三期增强光谱CT的患者,按Lauren分型分为肠型和弥漫型两组。
诊断标准与关键数值
- 列线图基于性别、动脉期标准化碘浓度和延迟期能谱曲线斜率构建,预测Lauren分型的AUC为0.841(95%CI 0.749-0.932),特异度85.3%,准确率76.4%,敏感度68.4%。
临床应用与阅片要点
- 作者认为基于临床特征和光谱CT定量参数的列线图在局部进展期胃癌Lauren分型的术前无创评估方面展现出较大潜力,可能有助于指导个体化治疗策略和风险分层。
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
OBJECTIVES: To develop a nomogram based on clinical features and spectral quantitative parameters to preoperatively predict the Lauren classification for locally advanced gastric cancer (LAGC).
METHODS: Patients diagnosed with LAGC by postoperative pathology who underwent abdominal triple-phase enhanced spectral computed tomography (CT) were prospectively enrolled in this study between June 2023 and December 2023. All the patients were categorized into intestinal- and diffuse-type groups according to the Lauren classification. Traditional characteristics, including demographic information, serum tumor markers, gastroscopic pathology, and image semantic features, were collected. Spectral quantitative parameters, including iodine concentration (IC), effective atomic number (Zeff), and slope of the energy spectrum curve from 40 keV to 70 keV (λ), were measured three times for each patient by two blinded radiologists in arterial/venous/delayed phases (AP/VP/DP). Differences in traditional features and spectral quantitative parameters between the two groups were compared using univariable analysis. Independent predictors of the Lauren classification of LAGC were screened using multivariable logistic regression analysis. Receiver operating characteristic (ROC) curve analysis was used to assess the discriminating capability. Ultimately, the nomogram, including clinical features and spectral CT quantitative parameters, was developed.
RESULTS: Gender, nIC in AP (APnIC), and λ in DP (λd) were independent predictors for Lauren classification. The nomogram based on these indicators produced the best performance with an area under the curve of 0.841 (95% confidence interval: 0.749-0.932), specificity of 85.3%, accuracy of 76.4%, and sensitivity of 68.4%.
CONCLUSION: The nomogram based on clinical features and spectral CT quantitative parameters exhibits great potential in the preoperative and non-invasive assessment of Lauren classification for LAGC.
KEY POINTS: Question Can the proposed nomogram, integrating clinical features and spectral quantitative parameters, preoperatively predict the Lauren classification in locally advanced gastric cancer (LAGC)? Findings The nomogram, based on gender, arterial phase normalized iodine concentration, and slope of the energy spectrum curve in the delayed phase showed satisfactory predictive ability. Clinical relevance The established nomogram could contribute to guiding individualized treatment strategies and risk stratification in patients by predicting the Lauren classification for LAGC before surgery.
引用
- Zhang J, Su C, Zhang Y, et al. Spectral CT-based nomogram for preoperative prediction of Lauren classification in locally advanced gastric cancer: a prospective study. Eur Radiol. 2024. PMID: 39532722. doi:10.1007/s00330-024-11163-y
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。