Development and validation of pFIB scores for exclusion of significant liver fibrosis in pediatric MASLD.

Abstract

Metabolic dysfunction-associated steatotic liver disease (MASLD) is the most prevalent pediatric liver disease, yet accurate risk scores for referral of children/adolescents with suspected clinically significant liver fibrosis are currently lacking. Clinical and biochemical variables were collected in a prospective cohort of 327 children and adolescents with severe obesity, in whom liver fibrosis was evaluated by transient elastography. Logistic regression was performed to establish continuous (pFIB-c) and simplified (pFIB-6) diagnostic scores that accurately exclude significant (≥F2) fibrosis. Performance for each was compared to established non-invasive fibrosis scores. These scores were validated in elastography (n=504) and multiple biopsy-proven MASLD (n=261) cohorts. Patient sex, ethnicity, weight z-score, HOMA-IR index, ALT, and presence of hypertension were included in the scores. The pFIB-c and pFIB-6 exhibited good discriminatory capacity (c-statistic of 0.839 and 0.826), outperforming existing indices. Negative predictive values (NPV) were >90% for both scores in the derivation and elastography validation cohorts. Performance in the histological cohorts varied (AUROCs for the pFIB-c between 0.710 and 0.770), as the scores were less accurate when applied to populations in tertiary referral centers characterized by a high prevalence of significant fibrosis and high ALT levels. Analyzing several cohorts totaling approximately 1100 children and adolescents, we developed novel risk scores incorporating readily available clinical variables. In accordance with the aim of excluding pediatric MASLD-associated fibrosis, the scores performed better in non-selected cohorts of children and adolescents living with obesity than in patients referred to tertiary liver units.