New Algorithm for the Integration of Ultrasound Into Cystic Fibrosis Liver Disease Screening. Journal of pediatric gastroenterology and nutrition Sellers, Z. M., Lee, L. W., Barth, R. A., Milla, C. n. 2019; 69 (4): 404–10

Abstract

Liver nodularity occurs across the spectrum of cystic fibrosis liver disease (CFLD), from regenerative nodules to cirrhosis, and can occur without liver enzyme abnormalities. Our aims were to determine if incorporating abdominal ultrasound (US) with annual laboratory testing improves the detection of CFLD and establish CF-specific thresholds for liver screening labs.CF patients at least 6 years old who were exocrine pancreatic-insufficient had an US with Doppler and shear wave elastography. Patients were divided into Normal, Echogenic, or Nodular groups, based on US findings. Results were compared with aspartate aminotransferase (AST), alanine aminotransferase (ALT), platelets, AST to platelet ratio index (APRI), Fibrosis 4 (FIB-4), and gamma-glutamyl transferase (GGT) to platelet ratio (GPR). Receiver operator curve, sensitivity, specificity, positive predictive value, negative predictive value, and optimal cut-off with Youden Index were calculated.From 82 patients, incorporation of US identified more nodular livers than using labs alone. The Nodular group had significantly greater median AST (44), ALT (48), GGT (46), APRI (0.619), FIB-4 (0.286), GPR (1.431). Optimal cut-offs to detect liver nodularity in CF were AST >33, ALT >45, GGT >21, Platelets <230, APRI >0.367, FIB-4 >0.222, GPR >0.682. Using GGT, APRI, and GPR, we generated an algorithm to direct the use of US in CFLD screening.Using modified serum lab thresholds, addition of liver fibrosis indices, and/or abdominal US can increase detection of liver nodularity in CF. A combination of GGT, GPR, and APRI can help direct which CF children should undergo US evaluation. These tools may improve earlier identification of fibrosis and/or cirrhosis in CF patients.

View details for DOI 10.1097/MPG.0000000000002412

View details for PubMedID 31181020