Clinical validation of the 2017 international consensus guidelines on intraductal papillary mucinous neoplasm of the pancreas

PURPOSE: The 2017 international consensus guidelines (ICG) for intraductal papillary mucinous neoplasm (IPMN) of the pancreas were recently released. Important changes included the addition of worrisome features such as elevated serum CA 19-9 and rapid cyst growth (>5 mm over 2 years). We aimed to clinically validate the 2017 ICG and compare the diagnostic performance between the 2017 and 2012 ICG. METHODS: This was a retrospective cohort study. During January 2000–January 2017, patients who underwent complete surgical resection and had pathologic confirmation of branch-duct or mixed-type IPMN were included. To evaluate diagnostic performance, the areas under the receiver operating curves (AUCs) were evaluated. RESULTS: A total of 448 patients were included. The presence of mural nodule (hazard ratio [HR], 9.12; 95% confidence interval [CI], 4.60–18.09; P = 0.001), main pancreatic duct dilatation (>5 mm) (HR, 5.32; 95% CI, 2.67–10.60; P = 0.001), thickened cystic wall (HR, 3.40; 95% CI, 1.51–7.63; P = 0.003), and elevated CA 19-9 level (>37 unit/mL) (HR, 5.25; 95% CI, 2.05–13.42; P = 0.001) were significantly associated with malignant IPMN. Malignant lesions showed a cyst growth rate >5 mm over 2 years more frequently than benign lesions (60.9% vs. 29.7%, P = 0.012). The AUC was higher for the 2017 ICG than the 2012 ICG (0.784 vs. 0.746). CONCLUSION: The new 2017 ICG for IPMN is clinically valid, with a superior diagnostic performance to the 2012 ICG. The inclusion of elevated serum CA 19-9 level and cyst growth rate to the 2017 ICG is appropriate.

Review of statistical methods for survival analysis using genomic data

Survival analysis mainly deals with the time to event, including death, onset of disease, and bankruptcy. The common characteristic of survival analysis is that it contains “censored” data, in which the time to event cannot be completely observed, but instead represents the lower bound of the time to event. Only the occurrence of either time to event or censoring time is observed. Many traditional statistical methods have been effectively used for analyzing survival data with censored observations. However, with the development of high-throughput technologies for producing “omics” data, more advanced statistical methods, such as regularization, should be required to construct the predictive survival model with high-dimensional genomic data. Furthermore, machine learning approaches have been adapted for survival analysis, to fit nonlinear and complex interaction effects between predictors, and achieve more accurate prediction of individual survival probability. Presently, since most clinicians and medical researchers can easily assess statistical programs for analyzing survival data, a review article is helpful for understanding statistical methods used in survival analysis. We review traditional survival methods and regularization methods, with various penalty functions, for the analysis of high-dimensional genomics, and describe machine learning techniques that have been adapted to survival analysis.

Review of statistical methods for survival analysis using genomic data

Survival analysis mainly deals with the time to event, including death, onset of disease, and bankruptcy. The common characteristic of survival analysis is that it contains “censored” data, in which the time to event cannot be completely observed, but instead represents the lower bound of the time to event. Only the occurrence of either time to event or censoring time is observed. Many traditional statistical methods have been effectively used for analyzing survival data with censored observations. However, with the development of high-throughput technologies for producing “omics” data, more advanced statistical methods, such as regularization, should be required to construct the predictive survival model with high-dimensional genomic data. Furthermore, machine learning approaches have been adapted for survival analysis, to fit nonlinear and complex interaction effects between predictors, and achieve more accurate prediction of individual survival probability. Presently, since most clinicians and medical researchers can easily assess statistical programs for analyzing survival data, a review article is helpful for understanding statistical methods used in survival analysis. We review traditional survival methods and regularization methods, with various penalty functions, for the analysis of high-dimensional genomics, and describe machine learning techniques that have been adapted to survival analysis.