Eduardo Vilar-Gomez1,2, Luis Calzadilla-Bertot1,3, Scott L Friedman4, Bienvenido Gra-Oramas5, Licet Gonzalez-Fabian5, Sacha Lazo-Del Vallin6, Moises Diago7, Leon A Adams3. 1. Department of Hepatology, National Institute of Gastroenterology, Havana, Cuba. 2. Division of Gastroenterology and Hepatology, Indiana University School of Medicine, Indiana, USA. 3. School of Medicine and Pharmacology, The University of Western Australia, Nedlands, Australia. 4. Division of Liver Diseases, Icahn School of Medicine at Mount Sinai, New York, USA. 5. Department of Pathology, National Institute of Gastroenterology, Havana, Cuba. 6. Department of Radiology, National Institute of Gastroenterology, Havana, Cuba. 7. Liver Unit, Department of Gastroenterology, Valencia University General Hospital, Valencia, Spain.
Abstract
BACKGROUND & AIMS: The dynamic response of serum fibrosis biomarkers to histological changes within the liver following lifestyle intervention (LI) is unknown. We explored relationships between changes in serum biomarkers and liver fibrosis in NASH patients undergoing LI. METHODS: Paired liver biopsies were performed in 261 NASH patients to assess fibrosis change after 1 year of LI. We explored the utility of serum fibrosis markers to predict changes in hepatic fibrosis and developed and internally validated a model for predicting fibrosis improvement in patients with baseline fibrosis. RESULTS: Regression, stabilization and worsening of fibrosis occurred in 51 (20%), 165 (63%) and 45 (17%) patients respectively. By multivariable analysis, change in HbA1c (OR, 0.39, P<.01), platelets (OR, 1.22, P<.01) and NFS (OR, 0.27, P<.01), as well as ALT normalization (OR, 9.7, P<.01) were independently associated with fibrosis improvement, whereas change in platelets (OR, 0.96, P<.01), and NFS (OR, 1.8, P<.01) as well as ALT normalization (OR, 0.21, P<.01) were linked to fibrosis progression. A model, including change in HbA1c, platelet and ALT normalization, was significantly more accurate (AUC of 0.96, 95% CI, l0.94-0.99) than NFS, FIB-4 and APRI for predicting fibrosis improvement. Using a threshold of ≥0.497, positive and negative predictive values were 94% (95% CI, 84-98) and 91% (95% CI, 81-96) respectively. CONCLUSIONS: Change in NFS, platelets and ALT normalization are associated with change in liver fibrosis after 1 year of LI. A model including change in HbA1c, platelet and ALT normalization discriminated patients with fibrosis improvement significantly better than other biomarkers.
BACKGROUND & AIMS: The dynamic response of serum fibrosis biomarkers to histological changes within the liver following lifestyle intervention (LI) is unknown. We explored relationships between changes in serum biomarkers and liver fibrosis in NASHpatients undergoing LI. METHODS: Paired liver biopsies were performed in 261 NASHpatients to assess fibrosis change after 1 year of LI. We explored the utility of serum fibrosis markers to predict changes in hepatic fibrosis and developed and internally validated a model for predicting fibrosis improvement in patients with baseline fibrosis. RESULTS: Regression, stabilization and worsening of fibrosis occurred in 51 (20%), 165 (63%) and 45 (17%) patients respectively. By multivariable analysis, change in HbA1c (OR, 0.39, P<.01), platelets (OR, 1.22, P<.01) and NFS (OR, 0.27, P<.01), as well as ALT normalization (OR, 9.7, P<.01) were independently associated with fibrosis improvement, whereas change in platelets (OR, 0.96, P<.01), and NFS (OR, 1.8, P<.01) as well as ALT normalization (OR, 0.21, P<.01) were linked to fibrosis progression. A model, including change in HbA1c, platelet and ALT normalization, was significantly more accurate (AUC of 0.96, 95% CI, l0.94-0.99) than NFS, FIB-4 and APRI for predicting fibrosis improvement. Using a threshold of ≥0.497, positive and negative predictive values were 94% (95% CI, 84-98) and 91% (95% CI, 81-96) respectively. CONCLUSIONS: Change in NFS, platelets and ALT normalization are associated with change in liver fibrosis after 1 year of LI. A model including change in HbA1c, platelet and ALT normalization discriminated patients with fibrosis improvement significantly better than other biomarkers.