超声造影表现与胰腺神经内分泌肿瘤病理分级相关
该回顾性研究纳入93例经手术病理确诊的胰腺神经内分泌肿瘤(PNET),分析常规超声及超声造影表现与WHO 2010病理分级的关系。多因素有序回归显示,肿瘤≥2 cm、主胰管扩张、肝转移及动脉期高增强与病理分级显著相关,模型识别G1、G2、G3的准确率分别为77.4%、67.7%、90.3%。提示超声特征有助于术前评估PNET病理分级。
要点
- 动脉期高增强与较低病理分级独立相关(OR 4.676),提示富血供肿瘤分化更好。
- 肿瘤≥2 cm、主胰管扩张、肝转移与更高级别PNET独立相关。
- 联合上述超声特征构建的有序回归模型对G3识别准确率达90.3%。
关键影像表现/征象
- 动脉期高增强与较低病理分级相关(OR 4.676,95%CI 1.656-13.206,p=0.004)。
- 肿瘤≥2 cm与更高级别PNET相关(OR 0.110,95%CI 0.020-0.606,p=0.011)。
- 主胰管扩张与更高级别PNET相关(OR 0.103,95%CI 0.025-0.430,p=0.002)。
- 肝转移与更高级别PNET相关(OR 0.250,95%CI 0.072-0.869,p=0.029)。
诊断标准与关键数值
- 有序logistic回归模型识别G1、G2、G3的准确率分别为77.4%、67.7%、90.3%。
- 多因素有序回归分析中,肿瘤≥2 cm、主胰管扩张、肝转移及动脉期高增强为与病理分级独立相关的超声特征。
临床应用与阅片要点
- 阅片时关注肿瘤大小、主胰管是否扩张、有无肝转移及动脉期增强程度,综合判断PNET可能分级。
- 动脉期高增强提示低级别可能大,而较大肿瘤伴胰管扩张或肝转移应警惕高级别。
- 超声造影动脉期增强水平可作为术前无创分级参考,但需结合其他影像及临床信息。
局限性
- 单中心回顾性研究,样本量有限(尤其G3仅12例),结论需外部验证。
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)。
英文原文(摘要)· 点击展开
The correlation of sonographic findings with pathologic grades of pancreatic neuroendocrine tumors (PNETs) remains unclear. This study aimed to evaluate the usefulness of sonographic features in diagnosing the pathologic grade of PNETs. Conventional and contrast-enhanced ultrasonography findings of PNETs diagnosed by surgical pathology from July 2010 to June 2020 were retrospectively reviewed. Sonographic features were compared among three pathologic grades of PNETs according to the World Health Organization 2010 classification. Ordinal regression models were constructed to evaluate the usefulness of the sonographic features in diagnosing the pathologic grade of PNETs. This study enrolled 93 participants with PNETs: 50 grade 1, 31 grade 2 and 12 grade 3. Multivariate ordinal regression analysis suggested that tumor size ≥2 cm (odds ratio [OR], 0.110; 95% confidence interval [CI], 0.020-0.606; p = 0.011), dilation of the main pancreatic duct (OR, 0.103; 95% CI, 0.025-0.430; p = 0.002), hepatic metastases (OR, 0.250; 95% CI, 0.072-0.869; p = 0.029) and hyper-enhancement in arterial phase (OR, 4.676; 95% CI, 1.656-13.206; p = 0.004) were significantly associated with the pathologic grades of PNETs. The accuracy of the ordinal logistic regression model in identifying grade 1, 2 and 3 PNETs was 77.4%, 67.7% and 90.3%, respectively. The findings suggest that sonographic features, including tumor size, pancreatic duct dilation and hepatic metastasis, as well as the enhancement level in arterial phase, may help identify different pathologic grades of PNETs.
引用
- Huang J, Chen J, Xu M, et al. Contrast-Enhanced Ultrasonography Findings Correlate with Pathologic Grades of Pancreatic Neuroendocrine Tumors. Ultrasound Med Biol. 2021. PMID: 33934943. doi:10.1016/j.ultrasmedbio.2021.02.013
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。