Meeting News Coverage

High serum ferritin not predictive of advanced fibrosis in NAFLD

ORLANDO, Fla. — Elevated serum ferritin levels were associated with severe liver injury among patients with nonalcoholic fatty liver disease but were not predictive of advanced fibrosis, according to data presented at Digestive Disease Week.

In a multicenter cohort study, researchers assessed data from 1,014 patients with biopsy-confirmed nonalcoholic fatty liver disease (NAFLD). Serum ferritin (SF) levels were collected, with the upper limit of normal (ULN) defined as fewer than 300 mcg/L among men and fewer than 200 mcg/L among women. Three cutoff values for SF were established and compared for predictive value: more than ULN (n=331), more than 1.5 times ULN (n=189) and more than twice ULN (n=103).

Patients with elevated SF were more frequently non-Caucasian, with lower levels of HDL cholesterol and higher levels of AST, ALT, bilirubin and HOMA. NASH was significantly more common among these participants (45.9% of cases vs. 34.8%; P=.003), as was fibrosis of stage 3 or 4 (33.3% vs. 23.5%; P<.001). Multivariate analysis, adjusting for factors including age, sex, race, BMI, ALT levels and diabetes, indicated a significant association between SF and advanced fibrosis using all three cutoffs. Risk increased along with cutoff values.

AUROC analysis indicated low accuracy and sensitivity values, but high specificity, for SF alone at all cutoffs in distinguishing of advanced fibrosis, with an AUROC of 0.55 (0.51-0.59), sensitivity of 41% and specificity of 70% for above ULN; AUROC 0.56 (0.52-0.6), 27% sensitivity and 84% specificity for 1.5 times ULN, and AUROC 0.54 (0.5-0.58), 16% sensitivity and 92% specificity for more than twice ULN (95% CI for all). Adding SF to current noninvasive scoring systems for the prediction of fibrosis did not affect diagnostic accuracy.

“Increased SF is associated with the presence and severity of fibrosis based on multivariate analysis, as shown by previously published studies,” researcher Anna Christina Dela Cruz, MD, digestive diseases and nutrition division of the University of Kentucky Medical Center, said. “However, SF on its own lacks the accuracy to detect the presence and severity of liver fibrosis.”

Disclosure: Researcher Christopher P. Day reported a board membership at Abbott. Researcher Jacob George has served on advisory committees/review panels for Bristol-Myers Squibb, Gilead Sciences, MSD, Novartis Pharmaceuticals and Roche Pharma.

For more information:

Dela Cruz AC. #382: Diagnostic Relevance of Serum Ferritin in Patients with Nonalcoholic Fatty Liver Disease. Presented at: Digestive Disease Week 2013; May 18-21, Orlando, Fla.

ORLANDO, Fla. — Elevated serum ferritin levels were associated with severe liver injury among patients with nonalcoholic fatty liver disease but were not predictive of advanced fibrosis, according to data presented at Digestive Disease Week.

In a multicenter cohort study, researchers assessed data from 1,014 patients with biopsy-confirmed nonalcoholic fatty liver disease (NAFLD). Serum ferritin (SF) levels were collected, with the upper limit of normal (ULN) defined as fewer than 300 mcg/L among men and fewer than 200 mcg/L among women. Three cutoff values for SF were established and compared for predictive value: more than ULN (n=331), more than 1.5 times ULN (n=189) and more than twice ULN (n=103).

Patients with elevated SF were more frequently non-Caucasian, with lower levels of HDL cholesterol and higher levels of AST, ALT, bilirubin and HOMA. NASH was significantly more common among these participants (45.9% of cases vs. 34.8%; P=.003), as was fibrosis of stage 3 or 4 (33.3% vs. 23.5%; P<.001). Multivariate analysis, adjusting for factors including age, sex, race, BMI, ALT levels and diabetes, indicated a significant association between SF and advanced fibrosis using all three cutoffs. Risk increased along with cutoff values.

AUROC analysis indicated low accuracy and sensitivity values, but high specificity, for SF alone at all cutoffs in distinguishing of advanced fibrosis, with an AUROC of 0.55 (0.51-0.59), sensitivity of 41% and specificity of 70% for above ULN; AUROC 0.56 (0.52-0.6), 27% sensitivity and 84% specificity for 1.5 times ULN, and AUROC 0.54 (0.5-0.58), 16% sensitivity and 92% specificity for more than twice ULN (95% CI for all). Adding SF to current noninvasive scoring systems for the prediction of fibrosis did not affect diagnostic accuracy.

“Increased SF is associated with the presence and severity of fibrosis based on multivariate analysis, as shown by previously published studies,” researcher Anna Christina Dela Cruz, MD, digestive diseases and nutrition division of the University of Kentucky Medical Center, said. “However, SF on its own lacks the accuracy to detect the presence and severity of liver fibrosis.”

Disclosure: Researcher Christopher P. Day reported a board membership at Abbott. Researcher Jacob George has served on advisory committees/review panels for Bristol-Myers Squibb, Gilead Sciences, MSD, Novartis Pharmaceuticals and Roche Pharma.

For more information:

Dela Cruz AC. #382: Diagnostic Relevance of Serum Ferritin in Patients with Nonalcoholic Fatty Liver Disease. Presented at: Digestive Disease Week 2013; May 18-21, Orlando, Fla.

    See more from Digestive Disease Week