Abstract
OBJECTIVE: This study was performed to clarify the influence of nonalcoholic fatty liver disease (NAFLD) on liver stiffness measurement (LSM) and establish a new diagnostic model. METHODS: A retrospective cohort of 601 patients with chronic hepatitis B (CHB) was enrolled as the derivation group, and a prospective cohort of 30 patients with concurrent CHB and NAFLD was enrolled as the validation group. RESULTS: The area under the receiver operating characteristic curve of LSM in patients with CHB without NAFLD (0.792) was higher than that in patients with concurrent CHB and NAFLD (0.720) in diagnosing significant liver fibrosis. Patients with concurrent CHB and NAFLD had significantly higher LSM values than those without NAFLD among the overall F0-F1 patients (6.88 vs. 5.80). The LSM value in the higher controlled attenuation parameter (CAP) quartile was significantly higher than that in the normal CAP quartile among F0-F1 patients (6.80 vs. 5.74). The efficacy of our new diagnostic model for liver fibrosis (Fibro-NAFLD) was higher than that of LSM in both study groups. CONCLUSION: NAFLD with a high CAP value increases the risk of false-positive diagnosis of significant fibrosis. The Fibro-NAFLD model improves the diagnostic efficacy of LSM in patients with concurrent CHB and NAFLD.