Magnetic resonance imaging-derived fat fraction predicts risk of malignancy in intraductal papillary mucinous neoplasm

Evrimler S., Yip-Schneider M. T. , Swensson J., Soufi M., Muraru R., Tirkes T., ...More

ABDOMINAL RADIOLOGY, 2021 (Peer-Reviewed Journal) identifier identifier identifier

  • Publication Type: Article / Article
  • Volume:
  • Publication Date: 2021
  • Doi Number: 10.1007/s00261-021-03146-0
  • Journal Indexes: Science Citation Index Expanded, Scopus


Purpose Assess the relationship between MRI-derived pancreatic fat fraction and risk of malignancy in intraductal papillary mucinous neoplasm (IPMN). Methods MRIs of patients with IPMN who underwent pancreaticoduodenectomy were analyzed. IPMN with low-grade dysplasia (n = 29) were categorized as low-risk while IPMN at high risk of malignancy consisted of those with high-grade dysplasia/invasive carcinoma (n = 33). Pancreatic fat-fraction (FFmean) was measured using the 2-point Dixon-method. Images were evaluated for the high-risk stigmata and worrisome features according to the revised 2017 Fukuoka consensus criteria. Data on serum CA19-9, Diabetes Mellitus (DM) status, body mass index (BMI), and histological chronic pancreatitis were obtained. Results A significant difference in FFmean was found between the high-risk IPMN (11.45%) and low-risk IPMN (9.95%) groups (p = 0.027). Serum CA19-9 level (p = 0.021), presence of cyst wall enhancement (p = 0.029), and solid mass (p = 0.008) were significantly associated with high-risk IPMN. There was a significant correlation between FFmean and mural nodule size (r = 0.36, p < 0.01), type 2 DM (r = 0.34, p < 0.01), age (r = 0.31, p < 0.05), serum CA 19-9 (r = 0.30, p < 0.05), cyst diameter (r = 0.30, p < 0.05), and main pancreatic duct diameter (r = 0.26, p < 0.05). Regression analysis revealed FFmean (OR 1.103, p = 0.035) as an independent predictive variable of high-risk IPMN. Conclusion FFmean is significantly associated with high-risk IPMN and an independent predictor of IPMN malignant risk. FFmean may have clinical utility as a biomarker to complement the current IPMN treatment algorithm and improve clinical decision making regarding the need for surgical resection or surveillance.