Natural history of focal nodular hyperplasia on gadoxetic acid-enhanced MRI: a multicenter cohort study

To investigate the natural history and longitudinal imaging evolution of focal nodular hyperplasia (FNH) using gadoxetic acid-enhanced MRI (EOB-MRI) in a multicenter cohort.

This retrospective multicenter cohort study enrolled patients with FNH who underwent both baseline and follow-up EOB-MRI. Lesion evolution was assessed using three complementary metrics: absolute diameter change (threshold ≥ 0.5 cm), relative percentage change (threshold ≥ 10%), and annual diameter change (cm/year). Hepatobiliary phase (HBP) phenotype transitions from baseline to final follow-up were assessed using a predefined four-tier classification system (types A-D). Multilevel mixed-effects models were applied to evaluate associations between clinical variables and longitudinal changes in lesion size.

A total of 93 patients with 128 FNH lesions were included. Over a median follow-up of 39.2 months (range: 12.1-84.8 months), the majority of lesions remained stable (78/128, 60.9%) or decreased in size (37/128, 28.9%), whereas a minority exhibited interval enlargement (13/128, 10.2%). Among enlarging lesions, longitudinal HBP phenotype transitions were observed, including four from type A to type C, two from type D to type C, and one from type B to type D. At final follow-up, the mean lesion diameter was modestly but significantly smaller than at baseline (3.15 ± 1.52 cm vs. 3.31 ± 1.33 cm; P = 0.014), corresponding to a mean relative change of -6.2% and a mean annual change of -0.05 cm/year. Multilevel mixed-effects models revealed no independent associations between lesion enlargement or continuous diameter changes and any of the evaluated clinical variables, including sex, age, baseline body mass index, hepatic steatosis, oral contraceptive use, pregnancy history, or follow-up duration (all P > 0.05).

Longitudinal EOB-MRI follow-up demonstrates that FNH exhibits a largely indolent natural history, with predominant lesion stability or spontaneous regression. Clinically significant enlargement is infrequent and shows no association with hormonal or other clinical factors. These findings support conservative management and indicate that routine imaging surveillance may be unnecessary in asymptomatic patients with typical imaging features.