J Gastroenterol Hepatol Endosc | Volume 2, Issue 6 | Research Article | Open Access
Keshishian Jonathan1,2*, Kumar Ambuj3, Tsalatsanis Athanasios3, Malafa Mokenge4 and Vignesh Shivakumar5
1Department of Digestive Diseases and Nutrition, University of South Florida Morsani College of Medicine, USA
2Veterans Affairs Hospital, Digestive Diseases and Nutrition, USA
3Department of Internal Medicine, USF Health Program for Comparative Effectiveness Research and Evidence Based Medicine, USA
4Moffitt Cancer Center, USA
5SUNY Downstate Medical Center, New York, USA
*Correspondance to: Jonathan Keshishian
Fulltext PDFBackground: The management of pancreatic intraductal papillary mucinous neoplasia (IPMN) is dependent on their risk of malignant progression to invasive IPMN. Recently a 5- point scoring system has been proposed as a useful clinical tool to predict the malignancy of IPMNs; however, this tool has not been validated. Here, we investigate the ability of the 5 point scoring system to predict malignancy in patients with IPMNs who had surgical resection of their IPMNs.
Methods: A total of 179 patients underwent pancreatic resection for IPMN from 1998 to 2011 at our institute. Data was entered prospectively. Following review of scoring system from Shin et al. [4] (World J Surg 2010), primary data extracted from the database included the following information for each patient: age, presence of mural nodule, MPD dilation, CA19-9, history of pancreatitis, tumor size, duct communication (side, main, or mixed), and final pathology reports. This scoring system uses five independent variables: the presence of mural nodules, MPD diameter >6 mm, CA 19-9 >37 U/mL, history of pancreatitis, and age ≥ 60 years. One point was given to each variable. Validation of the scoring system was performed using a ROC analysis.
Results: Records of 98 patients included all five variables. Analysis showed that a cut-off of 3 points had the highest discriminating power. The associated risk ratio (RR) was 3.13 (95% CI: 1.51-6.49) and could predict IPMN malignancy with a sensitivity of 73.2% and a specificity of 80.7% (AUC: 0.81, 95% CI: 0.73-0.89). Additional analysis performed on side-branch variant of IPMN and mixed type (71 cases) for the cut-off of 3 points also showed the highest discrimination in predicting malignancy in IPMN. The RR was 6.49 (95% CI: 2.41-17.7) with a sensitivity of 84.6% and specificity of 86.7% (AUC: 0.90, 95% CI: 0.82-0.96).
Conclusion: The 5-point scoring system described by Shin et al. [4] was successfully validated and can be used to reliably predict malignancy in IPMN in both main branch and side/mixed cases of IPMN. This scoring system may assist clinicians in predicting malignancy in the preoperative patient with IPMN and is especially useful with the side branch and mixed variant.
Jonathan K, Ambuj K, Athanasios T, Mokenge M, Shivakumar V. Evaluating Utility of Scoring System to Predict Malignancy and Invasiveness of Intraductal Papillary Mucinous Neoplasms of the Pancreas at a Single Center. J Gastroenterol Hepatol Endosc. 2017;2(6):1034.