Mangafodipir trisodium-enhanced MR imaging of pancreatic disease

• Published in: European radiology 2006-05, Vol.16 (5), p.988-997
• Main Authors: Boraschi, Piero, Donati, Francescamaria, Gigoni, Roberto, Caramella, Davide, Boggi, Ugo, Falaschi, Fabio, Bartolozzi, Carlo
• Format: Article
• Language: English
• Subjects: Adult; Aged; Aged, 80 and over; Biopsy; Cancer; Confidence; Contrast agents; Contrast Media - administration & dosage; Cysts; Disease; Edetic Acid - administration & dosage; Edetic Acid - analogs & derivatives; Female; Humans; Image acquisition; Image contrast; Image Enhancement; Image Interpretation, Computer-Assisted; Image processing; Laparoscopy; Lesions; Magnetic resonance imaging; Magnetic Resonance Imaging - methods; Male; Medical imaging; Middle Aged; Observer Variation; Observers; Pancreas; Pancreatic diseases; Pancreatic Diseases - diagnosis; Pancreatic Diseases - epidemiology; Pancreatic Diseases - pathology; Pancreatitis; Patients; Prospective Studies; Pyridoxal Phosphate - administration & dosage; Pyridoxal Phosphate - analogs & derivatives; Sensitivity and Specificity; Signal to noise ratio; Surgery
• ISSN: 0938-7994; 1432-1084
• DOI: 10.1007/s00330-005-0099-8

Our study aimed to assess the diagnostic capabilities of mangafodipir trisodium-enhanced MRI for the evaluation of pancreatic disease. Sixty-three patients suspected of having pancreatic disease underwent MRI with a 1.5-T device. After the acquisition of axial and coronal T2-weighted sequences, the MR protocol included T1-weighted fat-suppressed breath-hold SPGR images obtained before and 30 min after the infusion of Mn-DPDP (Teslascan). The detection of a focal pancreatic lesion and its intensity were evaluated in consensus by two observers, who also attempted to characterize each lesion as benign or malignant. The reviewers were blinded to patient identification and all clinical, laboratory and previous imaging findings. MR imaging results were correlated with surgery (n=37), laparoscopy (n=1), biopsy (n=2) and imaging follow-up (n=22). Sixty-two subjects were effectively included in our analysis because one patient was lost to follow-up; final malignant and benign diagnoses were determined in 22 (35%) and 29 (47%) of the patients, respectively. The level of confidence in the diagnosis of the pancreatic lesion was significantly increased by the administration of Mn-DPDP as demonstrated by ROC analysis of unenhanced and post-contrast image sets (P=0.009). Overall, on the basis of observers' readings, MR assessment of pancreatic disease resulted in 57 correct diagnoses (accuracy, 92%) and five (8%) incorrect diagnoses. The sensitivity, specificity, positive predictive value and negative predictive value of the reviewers for the detection of pancreatic lesions and for the differentiation between benign and malignant masses were 91% (95% CI: 84 and 98%), 93% (95% CI: 86 and 99%), 87% (95% CI: 79 and 95%) and 95% (95% CI: 89 and 100%), respectively. Mn-DPDP-enhanced MRI is an effective diagnostic tool for evaluating pancreatic disease.