Simpler Novel Non-Invasive Program Score for Stage-by-Stage Diagnosis of Liver Fibrosis in Untreated Patients with Chronic Hepatitis B: A Multicenter Prospective Study
Yongping Yang,Yan Chen,Yongping Chen,Zujiang Yu,Qin Li,Huanming Xiao,Ke-Qin Hu,Lin Tan,Dedong Xiang,Qinhua Shang,Chunliang Lei,Liang Chen,Xiaoyu Hu,Jing Wang,Huabao Liu,Wei Lu,Weilai Chi,Zheng Dong,Xiaodong Wang,Zhiqin Li,Da Chen,Wenlin Bai,Changjiang Zhang,Guangming Xiao,Xun Qi,Jing Chen,Li Zhou,Cuihong Zhang,Huiwei Sun,Minghua Deng,Xiaoling Chi,Xiaolong Qi,Zheng Zhang
DOI: https://doi.org/10.2139/ssrn.3204723
2018-01-01
Abstract:Background: Non-invasive evaluation for liver fibrosis is of great clinical interest and value, but current models fail to accurately stage and differentiate each stage of liver fibrosis, especially in patients with chronic hepatitis B (CHB), and never taking off to be used clinically. Methods: This multicenter, prospective study used unbiased penalized logistic regression, assessed 30 variables referencing to histologic fibrosis stage in a large training cohort (n=800), and established simpler novel non-invasive program score (SNNPS). An independent validation cohort (n=400) was used to validate the SNNPS. Finally, a mobile program was then developed for one-step detailed calculation of certain hepatic fibrosis staging. Findings: Five variables-liver stiffness measurement (LSM), platelet counts, age, serum hyaluronic acid and spleen diameter, were identified as independent predictors for fibrosis stage and developing SNNPS that has AUCs of 0.893, 0.897, and 0.909 for significant fibrosis, advanced fibrosis, and cirrhosis, respectively. Using sub-models of SNNPS, S1 score ≤ 2.875 was 86% specific and 82% sensitive for mild fibrosis, S2 score ≤ 4.06 was 85% specific and 86% sensitive for significant fibrosis, and S3 score ≤ 4.402 or > 4.402 was 93% specific and 91% sensitive for advanced fibrosis or cirrhosis. In validation set, AUCs for significant fibrosis, advanced fibrosis, and cirrhosis were 0.904, 0.912, and 0.897, respectively. The SNNPS with highest LR and lowest LR- was significantly more specific and sensitive than AAR, APRI, LSM alone and Hepascore, and can be calculated via a mobile program in one-step. Interpretation: Using sophisticated mathematic modelling and a large multicenter study data developed and validated SNNPS that is superior to previously reported models, and can be easily calculated via mobile program to accurately diagnose various hepatic fibrosis stages in CHB patients in a wider range. Trial Registration Number: ClinicalTrials.gov. Identifier: NCT65418 Funding Information: National Major Science and Technology Special Project of China (2013ZX10005002). Competing Interest Declaration: The authors declare that there is no conflict of interest in this study Ethical Approval Statement: The study meets the requirements of Declaration of Helsinki, and protocol was approved by the ethics committee of each participating institution, and a written informed consent was obtained from all patients.