Detailed Abstract
[E-poster]
[EP123] Pancreatic cancer risk prediction model using a multi-biomarker panel
Yoo Jin CHOI1, Jin-Yong JANG*1, Woongchang YOON2, Yoonhyeong BYUN1, Jae Seung KANG1, Hyeong Seok KIM1, Youngmin HAN1, Hongbeom KIM1, Wooil KWON1, Areum LEE1, Young-Ah SUH1, Jiyoung PARK3, Yonghwan CHOI4, Junghyun NAMKUM4, Sung Gon YI4, Hyunsoo KIM4, Jiyoung YU3, Yongkang KIM5, Min-Seok KIM5, Sun-Whe KIM6
1Department of Surgery, Seoul National University Hospital, Korea
2Research Institute of Basic Science, Seoul National University, Korea
3Department of Biomedical Science, Seoul National University College of Medicine, Korea
4Immunodiagnostics R&D Team, IVD Business Unit 5, SK Telecome, Korea
5Department of Statistics, Seoul National University, Korea
6Department of Surgery, National Cancer Center, Korea
Introduction : The pancreatic ductal adenocarcinoma (PDAC) has dismal survival rate due to late detection. Thus, many researches have been tried to discover diagnostic biomarkers for early detection of PDAC. This study was to discover a risk prediction model for pancreatic cancer with high accuracy for early diagnosis using a triple-marker panel that we developed in 2018 with the enzyme-linked immunosorbent assay (ELISA).
Methods : Previously, we (Park and Choi et al. 2018) reported that three biomarkers, leucine-rich alpha-2 glycoprotein (LRG1), transthyretic (TTR), and CA 19-9, improved the diagnostic performance in pancreatic cancer with sensitivity of 82.5%; specificity of 92.1%. In current study, LRG1, TTR, and CA 19-9 were combined into a single panel and tested with ELISA on 696 plasma samples, including PDACs (n=396) and normal samples (n=300). We proposed a risk prediction model with machine-learning method, logistic regression (LR), and compared it with support vector machine (SVM) and random forest (RF). To commercialize this model, we searched two optimal thresholds to distinguish three risk groups (high, intermediate, and low) that reliably satisfy four measurements, negative predictive value (NPV), positive predictive value (PPV), sensitivity (SEN), and specificity (SPE), simultaneously greater than 0.95%.
Results : The Pearson correlation between the triple marker panel examined with ELISA and the individual marker panel examined with ELISA (Park and Choi et al. 2018) was 0.884. The risk prediction model distinguished pancreatic cancer from normal individuals with AUC 0.935. The thresholds were 0.11 and 0.77, that satisfied NPV 95.15%, PPV 97.55%,
Conclusions : We first validated reproducibility of the performance of the
Methods : Previously, we (Park and Choi et al. 2018) reported that three biomarkers, leucine-rich alpha-2 glycoprotein (LRG1), transthyretic (TTR), and CA 19-9, improved the diagnostic performance in pancreatic cancer with sensitivity of 82.5%; specificity of 92.1%. In current study, LRG1, TTR, and CA 19-9 were combined into a single panel and tested with ELISA on 696 plasma samples, including PDACs (n=396) and normal samples (n=300). We proposed a risk prediction model with machine-learning method, logistic regression (LR), and compared it with support vector machine (SVM) and random forest (RF). To commercialize this model, we searched two optimal thresholds to distinguish three risk groups (high, intermediate, and low) that reliably satisfy four measurements, negative predictive value (NPV), positive predictive value (PPV), sensitivity (SEN), and specificity (SPE), simultaneously greater than 0.95%.
Results : The Pearson correlation between the triple marker panel examined with ELISA and the individual marker panel examined with ELISA (Park and Choi et al. 2018) was 0.884. The risk prediction model distinguished pancreatic cancer from normal individuals with AUC 0.935. The thresholds were 0.11 and 0.77, that satisfied NPV 95.15%, PPV 97.55%,
Conclusions : We first validated reproducibility of the performance of the
SESSION
E-poster
E-Session 7/27 ~ 7/29 ALL DAY