Department of Surgery, The Johns Hopkins University School of Medicine, Baltimore, MD 21287, USA; Department of Surgery, New York University Langone Hospital, New York City, New York 10016, USA.
Department of Radiology, The Johns Hopkins University School of Medicine, Baltimore, MD 21287, USA.
Diagn Interv Imaging. 2024 Jan;105(1):33-39. doi: 10.1016/j.diii.2023.08.002. Epub 2023 Aug 17.
The purpose of this study was to develop a radiomics-signature using computed tomography (CT) data for the preoperative prediction of grade of nonfunctional pancreatic neuroendocrine tumors (NF-PNETs).
A retrospective study was performed on patients undergoing resection for NF-PNETs between 2010 and 2019. A total of 2436 radiomic features were extracted from arterial and venous phases of pancreas-protocol CT examinations. Radiomic features that were associated with final pathologic grade observed in the surgical specimens were subjected to joint mutual information maximization for hierarchical feature selection and the development of the radiomic-signature. Youden-index was used to identify optimal cutoff for determining tumor grade. A random forest prediction model was trained and validated internally. The performance of this tool in predicting tumor grade was compared to that of EUS-FNA sampling that was used as the standard of reference.
A total of 270 patients were included and a fusion radiomic-signature based on 10 selected features was developed using the development cohort (n = 201). There were 149 men and 121 women with a mean age of 59.4 ± 12.3 (standard deviation) years (range: 23.3-85.0 years). Upon internal validation in a new set of 69 patients, a strong discrimination was observed with an area under the curve (AUC) of 0.80 (95% confidence interval [CI]: 0.71-0.90) with corresponding sensitivity and specificity of 87.5% (95% CI: 79.7-95.3) and 73.3% (95% CI: 62.9-83.8) respectively. Of the study population, 143 patients (52.9%) underwent EUS-FNA. Biopsies were non-diagnostic in 26 patients (18.2%) and could not be graded due to insufficient sample in 42 patients (29.4%). In the cohort of 75 patients (52.4%) in whom biopsies were graded the radiomic-signature demonstrated not different AUC as compared to EUS-FNA (AUC: 0.69 vs. 0.67; P = 0.723), however greater sensitivity (i.e., ability to accurately identify G2/3 lesion was observed (80.8% vs. 42.3%; P < 0.001).
Non-invasive assessment of tumor grade in patients with PNETs using the proposed radiomic-signature demonstrated high accuracy. Prospective validation and optimization could overcome the commonly experienced diagnostic uncertainty in the assessment of tumor grade in patients with PNETs and could facilitate clinical decision-making.
本研究旨在利用 CT 数据建立一种用于术前预测无功能胰腺神经内分泌肿瘤（NF-PNETs）分级的放射组学特征。
回顾性分析 2010 年至 2019 年间行切除术的 NF-PNETs 患者。从胰腺协议 CT 检查的动脉期和静脉期提取了 2436 个放射组学特征。对与手术标本中观察到的最终病理分级相关的放射组学特征进行联合互信息最大化，以进行分层特征选择和放射组学特征的开发。使用约登指数确定用于确定肿瘤分级的最佳截断值。内部训练和验证随机森林预测模型。该工具在预测肿瘤分级方面的性能与作为参考标准的 EUS-FNA 采样进行了比较。
共纳入 270 例患者，其中 201 例患者的开发队列中建立了基于 10 个选定特征的融合放射组学特征。149 例为男性，121 例为女性，平均年龄为 59.4±12.3（标准差）岁（范围：23.3-85.0 岁）。在对 69 例新患者的内部验证中，观察到较强的区分度，曲线下面积（AUC）为 0.80（95%置信区间[CI]：0.71-0.90），相应的灵敏度和特异性分别为 87.5%（95% CI：79.7-95.3）和 73.3%（95% CI：62.9-83.8）。在研究人群中，143 例患者（52.9%）接受了 EUS-FNA 检查。26 例患者的活检结果不可诊断（18.2%），42 例患者（29.4%）因样本不足而无法分级。在 75 例（52.4%）活检分级的患者中，放射组学特征的 AUC 与 EUS-FNA 相比无差异（AUC：0.69 比 0.67；P=0.723），但敏感性更高（即，准确识别 G2/3 病变的能力更高（80.8%比 42.3%；P<0.001）。
使用所提出的放射组学特征对 PNETs 患者进行肿瘤分级的无创评估显示出较高的准确性。前瞻性验证和优化可以克服在评估 PNETs 患者肿瘤分级时经常遇到的诊断不确定性，并有助于临床决策。
