Yuntai Cao1, Guojin Zhang2, Haihua Bao3, Shenghui Zhang4, Jing Zhang2, Zhiyong Zhao2, Wenjuan Zhang2, Weixia Li3, Xiaohong Yan5, Junlin Zhou6. 1. Second Clinical School, Lanzhou University, Lanzhou, China; Key Laboratory of Medical Imaging, Gansu Province, Lanzhou, China; Department of Radiology, Affiliated Hospital of Qinghai University, Xining, China; Department of Radiology, Lanzhou University Second Hospital, Lanzhou, China. 2. Second Clinical School, Lanzhou University, Lanzhou, China; Key Laboratory of Medical Imaging, Gansu Province, Lanzhou, China. 3. Department of Radiology, Affiliated Hospital of Qinghai University, Xining, China. 4. Department of Physics, University of Illinois at Chicago, Chicago, USA. 5. Department of Critical Medicine, Affiliated Hospital of Qinghai University, Xining, China. 6. Key Laboratory of Medical Imaging, Gansu Province, Lanzhou, China; Department of Radiology, Lanzhou University Second Hospital, Lanzhou, China. Electronic address: ery_zhoujl@lzu.edu.cn.
Abstract
PURPOSE: To develop a dual-energy spectral CT (DESCT) nomogram for the preoperative identification of KRAS mutation in patients with colorectal cancer (CRC). METHOD: One hundred and twenty-four patients who underwent energy spectrum CT pre-operatively were recruited and split into mutated KRAS group (n = 50) and wild-type KRAS group (n = 74). DESCT parameters, including monochromatic CT value, iodine concentration, water concentration, and effective atomic number were measured independently by two reviewers in the arterial, venous, and delayed phases. Normalized iodine concentration (NIC) and slope k of the spectral HU curve were calculated. Evaluate other imaging features such as ATL/LTL ratio, tumor gross pattern, pericolorectal fat invasion (PFI) was also performed by these reviewers. Independent predictors for KRAS mutation were screened out using logistic regression, and these predictors were presented as a nomogram. The receiver operating characteristic (ROC) curve, calibration curve and decision curve analysis (DCA) were used to evaluate the clinical usefulness of the nomogram. RESULTS: The slope k in the arterial phase, effective atomic number in the arterial phase, NIC in the venous phase, ATL/LTL ratio and PFI were significant independent predictors for KRAS mutation. Based on these independent predictors, a quantitative nomogram was developed to predict individual KRAS mutation probability. The nomogram had excellent performance with an AUC of 0.848 and excellent calibration. DCA showed that our nomogram has outstanding clinical utility. CONCLUSIONS: This study demonstrates that a DESCT based nomogram has potential value for individual preoperative identification of KRAS mutation in CRC patients.
PURPOSE: To develop a dual-energy spectral CT (DESCT) nomogram for the preoperative identification of KRAS mutation in patients with colorectal cancer (CRC). METHOD: One hundred and twenty-four patients who underwent energy spectrum CT pre-operatively were recruited and split into mutated KRAS group (n = 50) and wild-type KRAS group (n = 74). DESCT parameters, including monochromatic CT value, iodine concentration, water concentration, and effective atomic number were measured independently by two reviewers in the arterial, venous, and delayed phases. Normalized iodine concentration (NIC) and slope k of the spectral HU curve were calculated. Evaluate other imaging features such as ATL/LTL ratio, tumor gross pattern, pericolorectal fat invasion (PFI) was also performed by these reviewers. Independent predictors for KRAS mutation were screened out using logistic regression, and these predictors were presented as a nomogram. The receiver operating characteristic (ROC) curve, calibration curve and decision curve analysis (DCA) were used to evaluate the clinical usefulness of the nomogram. RESULTS: The slope k in the arterial phase, effective atomic number in the arterial phase, NIC in the venous phase, ATL/LTL ratio and PFI were significant independent predictors for KRAS mutation. Based on these independent predictors, a quantitative nomogram was developed to predict individual KRAS mutation probability. The nomogram had excellent performance with an AUC of 0.848 and excellent calibration. DCA showed that our nomogram has outstanding clinical utility. CONCLUSIONS: This study demonstrates that a DESCT based nomogram has potential value for individual preoperative identification of KRAS mutation in CRCpatients.