Magnetic Resonance Imaging Radiomics‐Based Nomogram From Primary Tumor for Pretreatment Prediction of Peripancreatic Lymph Node Metastasis in Pancreatic Ductal Adenocarcinoma: A Multicenter Study

• Published in: Journal of magnetic resonance imaging Vol. 55; no. 3; pp. 823 - 839
• Main Authors: Shi, Zhenshan, Ma, Chengle, Huang, Xinming, Cao, Dairong
• Format: Journal Article
• Language: English
• Published: Hoboken, USA John Wiley & Sons, Inc 01.03.2022; Wiley Subscription Services, Inc
• Subjects: Adenocarcinoma; Aged; Calibration; Cancer; Computed tomography; Confidence intervals; Decision analysis; Feature extraction; Female; Field strength; Humans; Lymph nodes; lymphatic metastasis; Lymphatic Metastasis - diagnostic imaging; Magnetic resonance imaging; Magnetic Resonance Imaging - methods; Male; Medical imaging; Metastases; Metastasis; Middle Aged; Nomograms; Pancreas; Pancreatic cancer; Pancreatic Neoplasms; Patients; Pretreatment; Radiomics; Resonance; Retrospective Studies; Statistical analysis; Statistical tests; Training; Tumors; magnetic resonance imaging; nomograms; pancreatic neoplasms; lymphatic metastasis
• ISSN: 1053-1807; 1522-2586; 1522-2586
• DOI: 10.1002/jmri.28048

Background Determining the absence or presence of peripancreatic lymph nodal metastasis (PLNM) is important to the pathologic staging, prognostication, and guidance of treatment in pancreatic ductal adenocarcinoma (PDAC) patients. Computed tomography and MRI had a poor sensitivity and diagnostic accuracy in the assessment of PLNM. Purposes To develop and validate a 3 T MRI primary tumor radiomics‐based nomogram from multicenter datasets for pretreatment prediction of the PLNM in PDAC patients. Study Type Retrospective. Subjects A total of 251 patients (156 men and 95 women; mean age, 60.85 ± 8.23 years) with histologically confirmed pancreatic ductal adenocarcinoma from three hospitals. Field Strength and Sequences A 3.0 T and fat‐suppressed T1‐weighted imaging. Assessment Quantitative imaging features were extracted from fat‐suppressed T1‐weighted (FS T1WI) images at the arterial phase. Statistical Tests Normally distributed data were compared by using t‐tests, while the Mann–Whitney U test was used to evaluate non‐normally distributed data. The diagnostic performances of the preoperative and postoperative nomograms were assessed in the external validation cohort with the area under receiver operating characteristics curve (AUC), calibration curve, and decision curve analysis (DCA). AUCs were compared with the De Long test. A p value below 0.05 was considered to be statistically significant. Results The AUCs of magnetic resonance imaging (MRI) Rad‐score were 0.868 (95% confidence level [CI]: 0.613–0.852) and 0.772 (95% CI: 0.659–0.879) in the training and internal validation cohort, respectively. The preoperative and postoperative nomograms could accurately predict PLNM in the training cohort (AUC = 0.909 and 0.851) and were validated in both the internal and external cohorts (AUC = 0.835 and 0.805, 0.808 and 0.733, respectively). DCA indicated that the two novel nomograms are of similar clinical usefulness. Data Conclusion Pre−/postoperative nomograms and the constructed radiomics signature from primary tumor based on FS T1WI of arterial phase could serve as a potential tool to predict PLNM in patients with PDAC. Evidence Level 3 Technical Efficacy Stage 2