Fontan venovenous collaterals and hepatic fibrosis

Objective We hypothesized that a relationship might exist between angiographically demonstrable, post‐Fontan venovenous collaterals, and hepatic fibrosis. Methods We analyzed data from post‐Fontan patients that underwent cardiac catheterization and transvenous‐hepatic biopsy procedures between March 2012 and March 2020. From innominate vein angiography, we determined those that either had or lacked venovenous collaterals. Additionally, we examined data from post‐Fontan patients that underwent hepatic ultrasound, shear‐wave elastography between January 2017 and March 2020. Results We identified 164 patients that met inclusion criteria. Of the 164, 101 (62%) had venovenous collaterals. Of the 101 with collaterals, average total fibrosis score (TFS) was 3.2 and the average rate of fibrosis progression was 0.28 vs an average TFS of 2.1 and an average fibrosis progression rate of 0.22 for those without collaterals (P = .00001 and P = .01, respectively). Of the 101 with collaterals, oxygen saturation was 91% ± 4% vs 93% ± 3% (P = .048) without collaterals. Of the 164, 86 (52%) underwent ultrasound shear‐wave elastography. Of the 86 patients undergoing elastography, 50 (58%) were performed in those with collaterals, and 36 (42%) in those without collaterals. For the 50 with collaterals, average elastography values were 13.3 vs 11.2 kPa for the 36 without collaterals (P = .006). We found no statistically significant differences for age at biopsy, Fontan duration, Fontan‐type, type of functional univentricle, laboratory, clinical, or hemodynamic values between those with or without collaterals. Conclusions The presence of angiographically demonstrated venovenous collaterals was associated with statistically, significantly more advanced liver fibrosis than those without collaterals.