Characteristics of UDCA-treated and untreated patients with FALD

Of the 220 patients who underwent the Fontan procedure, 113 were treated with UDCA at a dose of 150 to 600 mg/day due to liver or biliary enzyme abnormalities (Table 1). A flowchart of UDCA-treated and untreated patients with FALD and their outcomes is shown in Figure 1. The median age at the time of Fontan surgery and UDCA treatment was 26.7 years (range, 12.3–57.2 years) and 5.6 years (range, 0.9–37.5 years), respectively, in UDCA-treated patients and 25.9 years (range, 9.3–47.1 years) and 4.6 years (range, 0.0–29.7 years) in untreated patients. UDCA-treated patients tended to be older at the time of treatment and Fontan surgery than untreated patients (both p = 0.07). Men were significantly more likely to receive UDCA treatment (p < 0.01). Underlying cardiac disease and complications such as asplenia, polysplenia, visceral inversion/confusion, PLE, and hepatitis viruses were not significantly different between the two groups. HCC was more common in patients who did not receive UDCA treatment (17 [15.9%] vs. 6 [5.3%], p = 0.01). In terms of the characteristics of FALD-HCC, maximum diameter was 36 (11-80) mm in UDCA (-), 19 (15-25) in UDCA treatment, number of nodules of 1, 2, and ≥ 3 was 10/3/4 and 5/0/1, respectively (Table 2). The stage of HCC (IA/IB/II/IIIA/IIIB/IVA/IVB) was 5/5/5/1/1/0/0 and 2/2/0/0/1/0/1, respectively (p = 0.19). Initial treatment for HCC included liver resection in 2 cases, transcatheter arterial chemo-embolization and proton beam therapy in 6 cases, and radiation in 5 cases, and chemotherapy in 1 case; however, 3 cases received best supportive care due to cardiac disease. There was no significant difference between UDCA-treated and non-treated patients regarding the HCC status. The use of antiplatelet agents was similar between the groups but anticoagulant use was significantly higher in UDCA-treated patients (p = 0.02). Mild diarrhea was the only side effect observed in patients receiving UDCA.

Table 1 Characteristics of patients with FALDTable 2 Characteristics of FALD-HCC

In the initial blood tests conducted at the time of UDCA treatment or the first visit to our department (Table 3), GGT levels were significantly higher in UDCA-treated patients (114 vs. 60 U/L, p < 0.01). Type IV collagen 7S levels were slightly elevated in UDCA-treated patients (7.5 vs. 8.0 ng/mL, p = 0.07). There were no significant differences in other laboratory data, including albumin levels, BNP levels, and MELD-XI scores.

Table 3 Laboratory data of UDCA-treated and untreated patients with FALDEfficacy of UDCA treatment evaluated according to laboratory data

We monitored changes in liver and biliary enzyme levels 3, 6, and 12 months after UDCA treatment. The mean serum levels of T-BIL were slightly lower at 6 and 12 months but the change was not statistically significant compared to the pretreatment levels (Fig. 2). By contrast, the AST, ALT, and GGT levels were significantly lower at all time points after treatment. The liver enzyme levels remained near normal for several years thereafter. The platelet counts were slightly higher after 1 year of treatment but the change was not statistically significant.

Fig. 2

Changes in liver and biliary enzyme levels and platelet count following UDCA treatment. a Mean T-BIL levels were slightly decreased at 6 and 12 months after UDCA treatment; however, the change was not statistically significant. There were significant decreases in (b) AST levels, c ALT levels, and d GGT levels following UDCA treatment. e Platelet counts slightly increased 12 months after treatment but the change was not statistically significant. Data are presented as means with standard deviations. ALT alanine transaminase, AST aspartate aminotransferase, GGT gamma-glutamyl transferase, M months, T-BIL total bilirubin, PLT platelet count, UDCA ursodeoxycholic acid, Y year

Comparison of UDCA treatment in propensity-matched conditions

Given the differences in patient backgrounds, a propensity-matched analysis was performed to ensure comparability between the UDCA-treated and untreated groups. Matching was based on age at Fontan surgery, sex, GGT level, MELD-XI score, and warfarin use (Tables 1 and 3). After matching, the age and laboratory data were similar between UDCA-treated and untreated patients.

Next, we compared the changes in liver and biliary enzyme levels between the matched groups (UDCA-treated group, n = 63; untreated group, n = 55). The treated group received treatment for a minimum follow-up period of 6 months, with an overall mean follow-up duration of 5.8 years (range, 0.5–20.9 years). The T-BIL and AST levels did not significantly differ between the two groups but ALT and GGT did, being significantly lower in treated patients. Platelet counts were not significantly different between the two groups (Fig. 3).

Fig. 3

Comparison of liver and biliary enzyme levels between UDCA-treated and untreated patients with ≥6 months of follow-up. There were no significant differences in (a) T-BIL levels or (b) AST levels between UDCA-treated and untreated patients. There were significant reductions in (c) ALT levels and d GGT levels in treated patients vs. untreated patients. e Platelet counts did not significantly differ between the groups. Data are presented as means with standard deviations. ALT alanine transaminase, AST aspartate aminotransferase, GGT gamma-glutamyl transferase, T-BIL total bilirubin, PLT platelet count, UDCA ursodeoxycholic acid

Survival and HCC rates of patients with FALD stratified by UDCA treatment

In total, 10 patients (4.5%) died during the study period, with 2 deaths in the treated group and 8 in the untreated group. The causes of death were liver failure/HCC in 6 patients, sepsis in 2 patients, and heart failure/subarachnoid hemorrhage in 1 patient each. After propensity score matching, survival rates were analyzed using the Kaplan–Meier method (Fig. 4). The 5-year survival rate was 98.4% for treated patients and 90.7% for untreated patients, with a significant difference between the groups (p = 0.04).

Fig. 4.

Survival and HCC incidence rates in patients with FALD stratified by UDCA treatment. a Kaplan–Meier survival curve for the total cohort. The 5-year survival rates were significantly higher in UDCA-treated patients than in untreated patients. b Kaplan–Meier curve showing the incidence of HCC in the total cohort. The incidence was significantly lower in treated patients. c Kaplan–Meier curve for the incidence of HCC in patients with ≥6 months of follow-up. The difference in HCC incidence between the groups was not statistically significant. d Kaplan–Meier curve for the time to HCC occurrence following Fontan surgery. The occurrence of HCC was significantly delayed in treated patients relative to untreated ones. FALD Fontan-associated liver disease, HCC hepatocellular carcinoma, UDCA ursodeoxycholic acid

HCC was observed in 4 UDCA-treated patients and in 15 untreated patients (Table 1). The 5-year incidence of HCC was significantly higher in the untreated group (24.2% vs. 5.6%, respectively (Fig. 4). Among patients followed ≥ 6 months (Fig. 1), the incidence of HCC increased in the untreated group; however, there were no significant differences between the groups (Fig. 4). When considering the timing of HCC occurrence post-Fontan surgery, the time to HCC occurrence was significantly longer in treated patients. This suggests that UDCA may reduce or delay HCC development. We analyzed the incidence of HCC in anticoagulant-treated versus untreated patients. HCC was observed in 12 anticoagulant-treated patients (12.2%) and in 7 untreated patients (18.4%), with no significant difference (p = 0.35).

Furthermore, Cox regression analysis was conducted to estimate the HR and 95% CI for the impact of UDCA treatment on HCC development, adjusting for potential confounders such as age at Fontan surgery, sex, GGT level, MELD-XI score, and warfarin use. UDCA treatment was negatively associated with HCC development, with an HR of 0.24 (95% CI, 0.081–0.737, p = 0.01).

In our Cox analysis on overall survival, HCC was strongly associated with poor outcomes (HR 11.05; 95% CI, 2.483–49.154; p < 0.01). Other significant factors included the MELD-XI score (HR 1.166; 95% CI, 1.064–1.277; p < 0.01) and age at Fontan surgery (HR 1.111; 95% CI, 1.023–1.206; p = 0.01). Although UDCA was potentially associated with improved overall survival (HR 0.202; 95% CI, 0.035–1.150; p = 0.07), GGT was not selected as a significant factor.