Non‐invasive algorithm of enhanced liver fibrosis and liver stiffness measurement with transient elastography for advanced liver fibrosis in chronic hepatitis B

Summary Background The accuracy of Enhanced Liver Fibrosis ( ELF ; ADVIA Centaur, Siemens Healthcare Diagnostics, Tarrytown, NY , USA ) in assessing liver fibrosis in chronic hepatitis B ( CHB ) is to be determined. Aim To derive and validate a combined ELF ‐liver stiffness measurement ( LSM ) algorithm to predict advanced fibrosis in CHB patients. Methods Using the data of a previously reported cohort of 238 CHB patients, an ALT ‐based LSM algorithm for liver fibrosis was used as a training cohort to evaluate the performance of ELF against liver histology. The best combined ELF ‐ LSM algorithm was then validated in new cohort of 85 CHB patients not previously reported. Results In the training cohort, LSM has better performance of diagnosing advanced (≥F3) fibrosis (area under the receiver operating characteristics curve [ AUROC ] 0.83, 95% confidence interval [ CI 0.76–0.91] than ELF ( AUROC 0.69, 95% CI 0.63–0.75). The optimal cut‐off values of ELF were 8.4 to exclude advanced fibrosis, and 10.8 to confirm advanced fibrosis. In the training cohort, an ELF  ≤ 8.4 had a sensitivity of 95% to exclude advanced fibrosis; an ELF  > 10.8 had a specificity of 92% to confirm advanced fibrosis. In the combined algorithm, low ELF or low LSM could be used to exclude advanced fibrosis as both of them had high sensitivity (≥90%). To confirm advanced fibrosis, agreement between high ELF and high LSM could improve the negative predictive value specificity (from 65% and 74% to 80%). Conclusions An Enhanced Liver Fibrosis ‐ liver stiffness measurement algorithm could improve the accuracy of prediction of either ELF or LSM alone. Liver biopsy could be correctly avoided in approximately 60% of patients.