Pancreatic cancer is one of the most common solid tumours and pancreatic ductal adenocarcinomas (PDACs) are the most common type of pancreatic cancer, accounting for 95% of cases. The aim of this study was not to analyze the risk factors that may be associated with PDAC, the most common type of pancreatic cancer, but to determine how these factors, taken together, affect the diagnosis of the disease. An open-access dataset containing PDAC and associated factors was used in the study. Mann Whitney U test was used in data analysis. ROC analysis was used to determine the power of the variables individually and together, in discriminating PDAC. According to the results of ROC analysis for control-PDAC and benign hepatobiliary disease-PDAC groups, the discriminatory power of the creatinine variable between the groups was found to be quite low (AUC control-PDAC=0.531, AUC benign hepatobiliary disease-PDAC=0.514, respectively). The AUC values obtained from LYVE1, REG1B, TFF1 variables for control-PDAC and benign hepatobiliary disease-PDAC comparisons are high and the discriminative power of the variables is said to be high. The results of ROC analyses performed with variables obtained from binary and triple combinations of these variables were also found to be quite high. Based on the findings presented in the article, it can be concluded that the individual components connected with the disease, when utilized in isolation for diagnostic purposes, exhibit low impact. However, when these aspects are collectively included, there is an observed enhancement in diagnostic accuracy. Hence, it is advisable to collectively evaluate the elements believed to be linked with the disease during the diagnostic process, as this approach may yield more precise outcomes.
Key words: Pancreatic cancer, pancreatic ductal adenocarcinoma, variable combination, diagnosis
|
