Pancreatology : official journal of the International Association of Pancreatology (IAP) ... [et al.] · 2025
Serum autoantibody signatures enable non-invasive early detection of pancreatic cancer.
СтатьяНазвание и аннотация: Перевод пока не подготовлен
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Sun Li, Liu Yuqi, Yang Qian, Zhu Jicun, Li Tiandong, Jiang Donglin, Lu Yin, Liu Ling, Wang Yue, Zhang Xiaodan, Wang Zihan, Ma Hanke, Shi Jianxiang, Wang Keyan, Wang Peng, Song Chunhua, Wang Kaijuan, Ye Hua
Аннотация
BACKGROUND: Autoantibodies against tumor-associated antigens (TAAs) are promising biomarkers for early immunodiagnosis of cancers. This study was designed to screen and verify tumor-associated autoantibodies (TAAbs) in sera as diagnostic biomarkers for pancreatic cancer (PC).
METHODS: Bioinformatics-based data mining and the customized proteome microarray based on cancer driver genes were used to identify the potential TAAs. Enzyme-linked immunosorbent assay (ELISA) was adopted to assess the expression levels of the corresponding autoantibodies in 457 serum samples, followed by a diagnostic model was then constructed using machine learning.
RESULTS: Eleven candidate TAAs (EDNRA, OLR1, SEMA3C, AHNAK2, DHRS9, TMPRSS4, CCL20, SERPINB3, GPRC5A, TMC5 and MBOAT2) were identified through bioinformatics analysis and human protein chips, and further validated by ELISA. The titers of five TAAbs (anti-AHNAK2, anti-CCL20, anti-DHRS9, anti-OLR1, and anti-SERPINB3) that exhibited significant differences between PC and control subjects (P < 0.05), with AUC values ranging from 0.61 to 0.71. The random forest model developed using these five TAAbs exhibited AUCs of 0.83 (48.36 % sensitivity, 90.16 % specificity) and 0.76 (40.74 % sensitivity, 91.60 % specificity) for discriminating PC from healthy controls. Additionally, the model yielded AUCs of 0.86 (48.78 % sensitivity, 90.16 % specificity) and 0.80 (38.89 % sensitivity, 95.06 % specificity) for distinguishing early stage PC from controls in the training and test set, respectively. Notably, combining the random forest model with CA19-9 increased the positive rate to 88.97 %.
CONCLUSION: Our findings indicated that the rondom froest model based on five autoantibodies might help identify preclinical and early-stage PC.
METHODS: Bioinformatics-based data mining and the customized proteome microarray based on cancer driver genes were used to identify the potential TAAs. Enzyme-linked immunosorbent assay (ELISA) was adopted to assess the expression levels of the corresponding autoantibodies in 457 serum samples, followed by a diagnostic model was then constructed using machine learning.
RESULTS: Eleven candidate TAAs (EDNRA, OLR1, SEMA3C, AHNAK2, DHRS9, TMPRSS4, CCL20, SERPINB3, GPRC5A, TMC5 and MBOAT2) were identified through bioinformatics analysis and human protein chips, and further validated by ELISA. The titers of five TAAbs (anti-AHNAK2, anti-CCL20, anti-DHRS9, anti-OLR1, and anti-SERPINB3) that exhibited significant differences between PC and control subjects (P < 0.05), with AUC values ranging from 0.61 to 0.71. The random forest model developed using these five TAAbs exhibited AUCs of 0.83 (48.36 % sensitivity, 90.16 % specificity) and 0.76 (40.74 % sensitivity, 91.60 % specificity) for discriminating PC from healthy controls. Additionally, the model yielded AUCs of 0.86 (48.78 % sensitivity, 90.16 % specificity) and 0.80 (38.89 % sensitivity, 95.06 % specificity) for distinguishing early stage PC from controls in the training and test set, respectively. Notably, combining the random forest model with CA19-9 increased the positive rate to 88.97 %.
CONCLUSION: Our findings indicated that the rondom froest model based on five autoantibodies might help identify preclinical and early-stage PC.