Jie Chen1,2,3,4, Xiao-Long Li1,2,3, Yi-Feng Zhang1,2,3, Dan Wang1,2,3, Qiao Wang1,2,3, Chong-Ke Zhao1,2,3, Ming-Xu Li1,2,3, Qing Wei5, Guo Ji5, Hui-Xiong Xu1,2,3. 1. Department of Medical Ultrasound, Shanghai Tenth People's Hospital, Ultrasound Research & Education Institute, Tongji University School of Medicine, Shanghai 200072, PR China. 2. Thyroid Institute, Tongji University School of Medicine, Shanghai 200072, PR China. 3. Shanghai Center for Thyroid Disease, Shanghai 200072, PR China. 4. Department of Medical Ultrasound, Shanghai Chest Hospital, Shanghai Jiaotong University School of Medicine, Shanghai 200030, PR China. 5. Department of Pathology, Shanghai Tenth People's Hospital, Tongji University School of Medicine, Shanghai 200072, PR China.
Abstract
Aim: To compare the value of predictive power of the models for central cervical lymph node metastasis (CLNM) in papillary thyroid carcinomas (PTCs). Patients & methods: 220 PTCs were prospectively enrolled into the study with pathological examination. We established a new risk model with univariate and multivariate analyses and receiver-operating characteristic curves were plotted. Z-test was performed to compare the area under two curves and validated the predictive model for central CLNM in PTCs. The comparison of previous and new predictive model was analyzed. Results: Microcalcification, capsule contact or involvement, internal flow and BRAFV600E mutation were four independent risk factors for PTCs with central CLNMs. The area under the curves for the new and the previous model were 0.948 and 0.934 (p = 0.572), respectively. Conclusion: Two predictive models showed strong consistency in predicting central CLNM in PTCs. The predictive model may be helpful in selecting appropriate treatment method in PTCs.
Aim: To compare the value of predictive power of the models for central cervical lymph node metastasis (CLNM) in papillary thyroid carcinomas (PTCs). Patients & methods: 220 PTCs were prospectively enrolled into the study with pathological examination. We established a new risk model with univariate and multivariate analyses and receiver-operating characteristic curves were plotted. Z-test was performed to compare the area under two curves and validated the predictive model for central CLNM in PTCs. The comparison of previous and new predictive model was analyzed. Results: Microcalcification, capsule contact or involvement, internal flow and BRAFV600E mutation were four independent risk factors for PTCs with central CLNMs. The area under the curves for the new and the previous model were 0.948 and 0.934 (p = 0.572), respectively. Conclusion: Two predictive models showed strong consistency in predicting central CLNM in PTCs. The predictive model may be helpful in selecting appropriate treatment method in PTCs.