Hepatocellular carcinoma in Fontan-associated liver disease: Incidence, risk stratification, and surveillance implications

Abstract

Background:

Hepatocellular carcinoma (HCC) is a serious complication of Fontan-associated liver disease (FALD). However, data on its incidence, risk factors, and outcomes in this rare population are limited. We aimed to assess the incidence and predictors of HCC and evaluate the utility of surveillance in patients with FALD.

Methods:

We conducted a retrospective cohort study on patients with FALD who were referred to our gastroenterology department. The cumulative incidence of HCC after FALD diagnosis was calculated, and multivariate analysis was used to identify independent risk factors. Overall survival after FALD diagnosis was compared between patients who developed HCC and those who did not.

Results:

A total of 120 patients with FALD were followed up for a median period of 50.0 months. The cumulative incidence rates of HCC at 1, 2, and 3 years after FALD diagnosis were 0.9%, 2.7%, and 3.9%, respectively. Elevated alpha-fetoprotein (AFP) at FALD diagnosis was independently associated with HCC development (hazard ratio, 1.29; 95% confidence interval, 1.04–1.60; p=0.01). There was no significant difference in overall survival between patients with and without HCC in this cohort (log-rank test, p=0.50).

Conclusions:

Patients with FALD face a non-negligible risk of developing HCC, thus supporting the need for surveillance. Elevated AFP levels at the time of FALD diagnosis may aid in risk stratification, which enables the targeted monitoring and early detection of HCC. Furthermore, routine surveillance contributes to the excellent prognosis of patients with FALD who develop HCC, thus making their prognosis comparable to that of patients without HCC.

INTRODUCTION

The Fontan procedure is the standard palliative treatment for functional single-ventricle congenital heart disease. It has significantly improved patient survival, with 30-year cardiac survival rates exceeding 80%.¹,² Consequently, increased attention has been paid to various late end-organ complications. Fontan-associated liver disease (FALD) is one such complication. In 2023, the European Association for the Study of the Liver (EASL) and the European Reference Network on Rare Liver Diseases defined FALD as follows: FALD encompasses a broad spectrum of structural, functional, and clinical liver alterations secondary to Fontan hemodynamic changes.³ Among this wide range of conditions, hepatocellular carcinoma (HCC) that arises in the setting of FALD (hereafter referred to as FALD-HCC) is a significant complication that has serious effects on survival.⁴,⁵

In 2005, Ghaferi and Hutchins⁶ reported the first case of FALD-HCC in a patient who died 18 years after undergoing the Fontan procedure. Several case reports have demonstrated the occurrence of FALD-HCC over the past 2 decades, and studies have recently been conducted to elucidate the risk factors associated with FALD-HCC.⁷,⁸,⁹ However, most of these studies were cross-sectional and analyzed risk factors by using blood test values obtained close to the final imaging assessment. Researchers have also examined the cumulative incidence of HCC after the Fontan procedure. Owing to their cross-sectional nature and the timing of data collection, existing studies have a limited ability to evaluate liver-specific cancer risk and inform HCC surveillance strategies for patients with FALD. Therefore, longitudinal cohort data of patients with FALD are needed to better define the incidence and risk stratification.

Given that FALD begins insidiously after the Fontan procedure, the time from Fontan operation and the patient’s age are clinically meaningful metrics for HCC development. However, our primary aim was to evaluate liver-specific cancer risk in the clinical context of the transition from pediatric to adult care, using the time of clinically recognized FALD as an anchor point for adult hepatology practice. Therefore, this retrospective cohort study sought to (1) estimate the cumulative incidence of HCC after FALD diagnosis, (2) identify the predictors of HCC by using clinical data available at the time of FALD diagnosis, and (3) assess the clinical outcomes, thus informing HCC surveillance strategies for patients with clinically diagnosed FALD who do not have HCC.

METHODS

Study design

In 2014, Okayama University Hospital established its Adult Congenital Heart Disease Center, which comprises a multidisciplinary team. Patients who have undergone the Fontan procedure were referred to the gastroenterology department for FALD evaluation during transition from pediatric to adult care or when they exhibited failing Fontan physiology. During the initial visit, patients underwent blood tests, abdominal ultrasound (US), and liver stiffness measurements. The FALD diagnosis is based on abnormal blood test results, such as increased liver enzymes and gamma-glutamyl transpeptidase (GGT) levels, or abnormal imaging results, such as morphological changes or tumor formation in the liver. This operational definition is consistent with the criteria revealed in a nationwide epidemiological survey in Japan,¹⁰ reflecting the prevalent identification of FALD in routine clinical practice. Patients diagnosed with FALD were subsequently enrolled for HCC surveillance, and abdominal US and alpha-fetoprotein (AFP) measurements were conducted every 6 or 12 months. Patients with a new or enlarging liver nodule measuring above 10 mm in diameter on US or those with an elevated serum AFP value underwent diagnostic imaging with multiphase contrast-enhanced computed tomography or magnetic resonance imaging.

A total of 131 consecutive patients who were referred to the gastroenterology department and diagnosed with FALD between July 10, 2017, and September 3, 2024, were assessed for eligibility. All patients met the advocated diagnostic criteria for FALD, which are based on a nationwide epidemiological survey in Japan.¹¹ Among them, 11 patients were excluded (Figure 1). Patients diagnosed with HCC at the time of FALD diagnosis were excluded because the purpose of this study was to identify the incidence and predictors of HCC throughout the surveillance and to assess the outcomes of surveillance in patients without HCC at FALD diagnosis. The remaining 120 patients were included in the final analysis. To evaluate liver-specific cancer risk in patients with clinically recognized FALD, we calculated the cumulative incidence of HCC after FALD diagnosis and analyzed the risk factors for HCC development by using clinical information and blood test results obtained at the time of FALD diagnosis. In addition, we compared the survival differences after FALD diagnosis between 2 groups: those who subsequently developed HCC and those who did not. The date of data cutoff regarding the incidence of HCC and survival was June 3, 2025.

FIGURE 1.

Study flowchart. This diagram outlines the selection process for the study cohort. A total of 131 consecutive patients who were referred to the gastroenterology department at Okayama University and diagnosed with FALD between July 10, 2017, and September 3, 2024, were assessed for eligibility. After excluding 11 patients because of screening failure, 120 patients were included in the final analysis. Abbreviations: FALD, Fontan-associated liver disease; HCC, hepatocellular carcinoma.

The study was approved by the Human Ethics Review Committee of Okayama University Hospital (approval number: 2510-042) and was conducted in accordance with the Declaration of Helsinki. Given that participants were provided with study information via the institutional website and were given the opportunity to refuse participation, written informed consent was not required.

Data collection

The following demographic and clinical information at the time of FALD diagnosis was obtained from the patients’ medical records: sex, date of initial Fontan surgery, date of FALD diagnosis, date of HCC diagnosis or last imaging assessment, date of last confirmed survival or death, presence of situs inversus or polysplenia, history of warfarin use, blood test values, and liver stiffness values. The Child–Turcotte–Pugh score was not evaluated in this study because most patients were treated with warfarin. Therefore, a model for end-stage liver disease excluding the international normalized ratio (MELD-XI) score was assessed instead of the model for the end-stage liver disease score. Fibrosis-4 (FIB-4)¹² and MELD-XI¹³ scores were calculated as previously described. From 2017 to March 2022, liver stiffness was measured using vibration-controlled transient elastography (FibroScan 530 Compact; Echosens, Paris, France). From March 2022 onward, liver stiffness was measured using two-dimensional shear wave elastography (Aplio-i800; Canon Medical Systems, Tokyo, Japan).

HCC diagnosis

The Liver Imaging Reporting and Data System,¹⁴ recommended in the guidelines from the EASL and the American Association for the Study of Liver Diseases (AASLD),¹⁵,¹⁶ does not apply to cardiac cirrhosis. It is also difficult to differentiate HCC from benign liver nodules in patients with Fontan physiology owing to its unique hemodynamics. Given that an evidence-based algorithm for diagnosing HCC has not yet been established for patients with Fontan physiology, our institute has defined operational diagnostic criteria. These criteria consist of the following 3 main parts: (1) typical features on dynamic contrast-enhanced imaging, (2) threshold growth of liver nodules or nodules measuring above 10 mm in diameter, and (3) a marked elevation or progressive increase in serum AFP levels. Criteria (1) and (2) are consistent with the reported “suspicious of malignancy” imaging features of liver nodules in FALD.³ At our institute, we consider a pathological assessment via a US-guided biopsy when the 3 criteria are not fully met, owing to its potential bleeding risk in this population. The Barcelona Clinic Liver Cancer (BCLC) staging system was used to stage HCC.¹⁷ The imaging interpretations were performed as part of routine clinical practice, and were not blinded to AFP values and other clinical information.

Statistical analysis

Quantitative variables are expressed as medians with interquartile ranges (IQRs). Categorical data are presented as frequency counts and percentages. To compare the baseline characteristics of the 2 study groups, which were classified according to the development of HCC during follow-up, the Mann–Whitney U test and the Fisher exact test were used for quantitative and categorical variables, respectively. The follow-up time for the incidence of HCC was calculated from the date of FALD diagnosis until the date of HCC diagnosis or the final imaging follow-up before June 3, 2025. The follow-up time for survival was calculated from the date of FALD diagnosis until the date of death, or last confirmed survival before June 3, 2025. We examined the cumulative incidence of HCC and cumulative survival by using the Kaplan–Meier method. We then used the log-rank test to estimate the differences in incidence and survival between the 2 study groups. Univariate and multivariate analyses were conducted using the Cox proportional hazards model to identify the risk factors for HCC development after FALD diagnosis. The results are presented as hazard ratios (HRs) with corresponding 95% confidence intervals (CIs). Owing to multicollinearity, platelet count (a component of FIB-4) and total bilirubin (included in MELD-XI) were excluded from the analysis to avoid redundancy and model overfitting. To examine the robustness of the findings in the main analysis using the Cox proportional hazards model, we conducted a sensitivity analysis limited to patients who had undergone Fontan surgery more than 15 years earlier. The Youden index was used to determine the optimal cutoff values for continuous variables associated with HCC development. Variables with p<0.1 in the univariate analysis were included in the multivariate model. A p-value <0.05 was considered statistically significant. All statistical analyses were performed using the JMP statistical software (version 18.0.2; SAS Institute Inc., Cary, NC, USA).

RESULTS

Patient characteristics

None of the patients had a history of chronic viral hepatitis, steatotic liver disease, or autoimmune liver disease. Table 1 presents a summary of the characteristics of the patients in this cohort. Among the initial Fontan connections, 14 (11.7%) were atriopulmonary, 45 (37.5%) were lateral tunnel, and 60 (50.0%) were extracardiac. The median age at Fontan surgery was 3 years (IQR, 2–5), and FALD was diagnosed at a median age of 20 years (IQR, 16–26), corresponding to a median interval of 16.4 years (IQR, 12.0–21.7). Most patients were in the postoperative period during which the risk of HCC increases, corresponding to the transition to adult care. The demographic characteristics of these patients were similar to those of patients with FALD reported in the national survey in Japan.¹⁰ Among them, 24 patients (20%) underwent conversion surgery to total cavopulmonary connection at a median interval of 19.1 years (IQR, 15.2–21.6) after the initial Fontan surgery.

TABLE 1.

Baseline characteristics of patients at the time of FALD diagnosis (stratified by HCC development)

	All patients N=120	Non-HCC N=115	HCC N=5	p
Sex (male)	64 (53.3%)	60 (52.1%)	4 (80.0%)	0.37
Age at FALD diagnosis, years	20 (16–26)	20 (16–26)	22 (21–31)	0.18
Age at Fontan procedure, years	3 (2–5)	3 (2–5)	4 (3–7)	0.28
Initial type of Fontan connection				0.82
Atriopulmonary	14 (11.7%)	14 (12.2%)	0 (0.0%)
Lateral tunnel	45 (37.5%)	42 (36.5%)	3 (60.0%)
Extracardiac	60 (50.0%)	58 (50.4%)	2 (40.0%)
Unknown	1 (0.8%)	1 (0.9%)	0 (0.0%)
Antithrombotic therapy				1.00
Warfarin	82 (68.3%)	78 (67.8%)	4 (80.0%)
Direct oral anticoagulant	7 (5.8%)	7 (6.1%)	0 (0.0%)
Aspirin	30 (25.0%)	29 (25.2%)	1 (20.0%)
None	1 (0.8%)	1 (0.9%)	0 (0.0%)
Situs inversus (yes)	26 (21.7%)	25 (21.7%)	1 (20.0%)	1.00
Polysplenia (yes)	21 (17.5%)	20 (17.4%)	1 (20.0%)	1.00
Asplenia (yes)	11 (9.2%)	11 (9.6%)	0 (0.0%)	1.00
Hepatic venous waveform (flat or monophasic)	46 (38.3%)	42 (36.5%)	4 (80.0%)	0.07
Liver stiffness, kPa	17.9 (12.7–24.7)	17.8 (12.6–24.4)	32.0 (16.1–41.4)	0.11
Blood test
PLT, ×104/μL	14.7 (11.3–18.7)	14.9 (11.8–19.1)	8.8 (8.0–9.9)	<0.01
T. Bil, mg/dL	1.14 (0.80–1.59)	1.12 (0.80–1.59)	1.49 (1.28–2.00)	0.11
Alb, g/dL	4.6 (4.1–4.7)	4.5 (4.1–4.7)	4.8 (4.4–5.0)	0.13
AST, U/L	23 (19–30)	23 (19–30)	25 (15–30)	0.67
ALT, U/L	20 (15–25)	20 (15–25)	21 (14–34)	0.59
GGT, U/L	66 (45–107)	65 (45–103)	137 (78–152)	0.07
BNP, pg/mL	26.8 (11.7–60.7)	26.7 (11.6–61.7)	30.2 (15.3–114.7)	0.83
Hyaluronic acid, ng/mL	33.7 (21.2–59.6)	34.1 (21.0–60.4)	31.9 (21.2–36.3)	0.46
Type IV collagen 7 S, ng/mL	7.6 (6.1–9.0)	7.7 (6.1–9.0)	6.6 (6.3–7.3)	0.28
M2BPGi, COI	0.39 (0.26–0.59)	0.38 (0.26–0.59)	0.41 (0.33–0.72)	0.51
AFP, ng/mL	3.2 (2.2–5.4)	3.2 (2.2–5.3)	5.2 (3.0–15.9)	0.17
FIB-4	0.74 (0.47–1.10)	0.72 (0.45–1.08)	1.48 (1.01–1.63)	0.01
MELD-XI	10 (9–12)	10 (9–12)	12 (11–14)	0.11

Quantitative variables are expressed as medians with interquartile ranges. Categorical variables are expressed as absolute values and percentages. The differences in quantitative and categorical variables were tested using the Mann–Whitney U test and the Fisher exact test, respectively.

Abbreviations: AFP, alpha-fetoprotein; Alb, albumin; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BNP, brain natriuretic peptide; FALD, Fontan-associated liver disease; FIB-4, fibrosis-4; GGT, gamma-glutamyl transpeptidase; HCC, hepatocellular carcinoma; M2BPGi, Mac-2-binding protein glycosylation isomer; MELD-XI, model for end-stage liver disease excluding the international normalized ratio; N, number; PLT, platelets; T. Bil, total bilirubin.

FALD-HCC characteristics

During a median follow-up period of 50.0 months (IQR, 29.2–65.1) after FALD diagnosis, HCC developed in 5 patients (4.2%). At the time of diagnosis, 4 patients had a single HCC nodule, and 1 had 2 nodules, thus totaling 6 lesions. All nodules were located on the periphery of the liver, and 4 exhibited exophytic growth beyond the liver surface. None of the patients exhibited macrovascular invasion or extrahepatic spread. According to the BCLC classification, 4 cases were categorized as BCLC 0, and 1 case as BCLC A. Histopathological confirmation was available in 1 case; the others were diagnosed on the basis of the operational diagnostic criteria. In all patients, serum AFP levels were significantly elevated from baseline at diagnosis (ranging from 20.3 to 163 ng/mL) and returned to baseline following treatments.

Table 1 shows a summary of the baseline characteristics of patients with and without HCC. Patients who developed HCC had a significantly higher FIB-4 index and lower platelet count. Higher total bilirubin levels, increased GGT levels, and abnormal hepatic venous waveforms (flat or monophasic) were also observed in this group.

Incidence and predictors of HCC following FALD diagnosis

The cumulative incidence rates of HCC were 0.9%, 2.7%, and 3.9% at 1, 2, and 3 years after the diagnosis of FALD, respectively (Figure 2).

FIGURE 2.

Cumulative incidence of HCC. The cumulative incidence rates of HCC at 1, 2, and 3 years after FALD diagnosis were 0.9%, 2.7%, and 3.9%, respectively. Abbreviations: FALD, Fontan-associated liver disease; HCC, hepatocellular carcinoma.

Candidate variables for univariate analysis were selected on the basis of previous literature and included clinical factors (interval between the Fontan procedure and FALD diagnosis, warfarin use, situs inversus, and polysplenia), biomarkers (GGT, brain natriuretic peptide, hyaluronic acid, type IV collagen 7 S, Mac-2-binding protein glycosylation isomer, and AFP), imaging-based measures (liver stiffness and hepatic venous waveform), and composite scores (FIB-4 and MELD-XI). Multivariate analysis was performed, including liver stiffness, AFP, hepatic venous waveform, and FIB-4 index, which were selected on the basis of clinical relevance and univariate results. As shown in Table 2, AFP at FALD diagnosis was identified as an independent predictor of HCC development (HR, 1.29; 95% CI, 1.04–1.60; p=0.01). To address concerns about overfitting in the multivariate model due to the small number of HCC events and to examine the robustness of this finding, we conducted a sensitivity analysis limited to 71 patients who underwent Fontan surgery more than 15 years earlier. In this cohort, serum AFP value at the time of FALD diagnosis remained the only independent predictor of HCC development (Supplemental Table S1, http://links.lww.com/HC9/C248), thus highlighting its potential role in the early risk stratification of HCC rather than as a stand-alone diagnostic test.

TABLE 2.

Risk factors for HCC development at the time of FALD diagnosis

	Univariable analysis		Multivariable analysis
	HR (95% CI)	p	HR (95% CI)	p
Fontan-to-FALD interval, years	1.06 (0.94–1.18)	0.25
Liver stiffness, kPa	1.04 (0.98–1.10)	0.09	1.03 (0.95–1.10)	0.31
GGT, U/L	1.00 (0.99–1.01)	0.18
BNP, pg/mL	1.00 (0.98–1.00)	0.76
Hyaluronic acid, ng/mL	0.97 (0.91–1.00)	0.29
Type IV collagen 7 S, ng/mL	0.76 (0.44–1.01)	0.28
M2BPGi, COI	1.43 (0.05–5.22)	0.72
AFP, ng/mL	1.27 (1.06–1.50)	<0.01	1.29 (1.04–1.60)	0.01
Hepatic venous waveform (flat or monophasic)	6.99 (0.78–62.7)	0.08	6.04 (0.57–63.1)	0.13
Warfarin medication (no)	0.48 (0.05–4.33)	0.51
Situs inversus (yes)	0.84 (0.09–7.56)	0.88
Polysplenia (yes)	1.14 (0.12–10.2)	0.90
FIB-4	2.50 (0.75–6.42)	0.07	0.99 (0.17–4.07)	0.99
MELD-XI	1.27 (0.85–1.84)	0.20

Univariate and multivariate analyses were conducted using the Cox proportional hazards model to identify the predictors of HCC development at the time of FALD diagnosis.

Abbreviations: AFP, alpha-fetoprotein; BNP, brain natriuretic peptide; FALD, Fontan-associated liver disease; FIB-4, fibrosis-4; GGT, gamma-glutamyl transpeptidase; HCC, hepatocellular carcinoma; M2BPGi, Mac-2-binding protein glycosylation isomer; MELD-XI, model for end-stage liver disease excluding the international normalized ratio.

Subgroup analysis of HCC incidence stratified by AFP value

To evaluate the clinical utility of AFP in risk stratification, we conducted a subgroup analysis of HCC incidence following FALD diagnosis. The optimal AFP cutoff value was determined to be 4.5 ng/mL by using receiver operating characteristic curve analysis (sensitivity, 80.0%; specificity, 67.8%; positive predictive value, 9.8%; negative predictive value, 98.7%). The cumulative incidence of HCC was significantly higher in patients with AFP ≥ 4.5 ng/mL than in those with lower levels. In the high AFP group, the 1-, 2-, and 3-year HCC incidence rates after FALD diagnosis were 2.5%, 5.1%, and 8.5%, respectively. By contrast, the respective rates were 0.0%, 1.4%, and 1.4% in the low AFP group (log-rank test: p=0.02; Figure 3). This finding supports the potential role of AFP levels in identifying patients with FALD who may benefit from closer monitoring for HCC.

FIGURE 3.

Cumulative incidence of HCC stratified by AFP levels at the time of FALD diagnosis. Patients were stratified on the basis of an AFP cutoff value of 4.5 ng/mL. The cumulative incidence of HCC after FALD diagnosis was significantly higher in patients with AFP ≥ 4.5 ng/mL than in those with AFP < 4.5 ng/mL. Levels of significance: p=0.02 (log-rank test). Solid line, AFP ≥ 4.5 ng/mL; dotted line, AFP<4.5 ng/mL. Abbreviations: AFP, alpha-fetoprotein; FALD, Fontan-associated liver disease; HCC, hepatocellular carcinoma.

Treatments and outcomes of FALD-HCC

All 5 patients diagnosed with HCC were asymptomatic at the time of detection. The treatments included liver resection (n=1), percutaneous radiofrequency ablation (RFA; n=1), and conventional transcatheter arterial chemoembolization (cTACE; n=3). The patients treated with RFA and 2 of the patients treated with cTACE did not experience recurrence during this study. One patient treated with cTACE underwent additional sessions for intrahepatic recurrence. The patient who underwent liver resection experienced recurrence at 5 months and required a second resection; this patient later underwent radiotherapy and systemic chemotherapy for intrahepatic and vertebral metastases. All patients survived during the study period, with a median follow-up of 41.6 months (IQR, 19.4–54.8) after HCC diagnosis. The 1- and 3-year cumulative survival rates after HCC diagnosis were both 100%. There was no significant difference in overall survival after FALD diagnosis between patients with and without HCC (log-rank test: p=0.50; Figure 4). This finding supports the idea that surveillance in this population enables the early detection of HCC and allows patients with HCC to expect an excellent prognosis comparable to that of patients without HCC.

FIGURE 4.

Cumulative survival after FALD diagnosis. The patients were stratified on the basis of the development of HCC during the follow-up period. No significant difference in survival was observed between patients with and without HCC. Levels of significance: p=0.50 (log-rank test). Solid line, patients with HCC development; dotted line, patients without HCC development. Abbreviations: FALD, Fontan-associated liver disease; HCC, hepatocellular carcinoma.

DISCUSSION

To the best of our knowledge, this is the largest retrospective cohort study conducted to determine the incidence and predictors of FALD-HCC and to clarify the outcomes of HCC surveillance after FALD diagnosis. This study revealed 3 significant findings.

First, patients diagnosed with FALD have a non-negligible risk of developing HCC, with cumulative incidence rates of 0.9% and 3.9% at 1 and 3 years, respectively. Nandwana et al. reported a similar incidence rate of ~5% in the Fontan population during a median follow-up of 3.05 years. However, their study included a wide variability in surveillance strategies as methods and intervals.¹⁸ Of note, the incidence rates are comparable to those observed in patients with hepatitis B virus (HBV) infection who do not have cirrhosis.¹⁹ The AASLD recommends HCC surveillance in patients with HBV infection who do not have cirrhosis on the basis of a cost-effectiveness analysis, indicating that surveillance is justified when the annual incidence exceeds 0.2%.¹⁶ For patients with FALD, the position paper currently recommends HCC surveillance every 6 months starting 10 years after Fontan surgery as a comprehensive agreement because of the lack of robust evidence.³ Our results support this management strategy on the basis of sufficient evidence.

Second, our findings emphasized the role of AFP levels in predicting the development of FALD-HCC. A previous report noted that the most recent serum AFP levels before the onset of HCC in patients who underwent Fontan surgery are significantly higher than those in patients without HCC.⁹ In addition to this finding, the results of the current study indicated that elevated AFP levels at the time of FALD diagnosis (cutoff: 4.5 ng/mL) were associated with a significantly increased risk of subsequent HCC, thus enabling risk stratification at the time of enrollment into surveillance. A similar scenario has been observed in patients with chronic hepatitis C who achieved a sustained virological response (SVR) through treatment with direct-acting antiviral agents. Some studies have shown that the cumulative incidence of HCC after achieving SVR is significantly higher in patients with higher post-treatment AFP levels, with cutoff values ranging from 5.0 to 6.0 ng/mL.²⁰,²¹,²² Our finding—a serum AFP threshold of 4.5 ng/mL in patients with FALD—is comparable to those reported in these studies. Along with the results of the current study, we speculate that there is a relationship between serum AFP levels and hepatic carcinogenesis in a pathogenesis characterized by the absence or minimal inflammation in the liver parenchyma. On the other hand, although serial AFP trends (eg, serial increases) may improve the predictive accuracy during HCC surveillance periods, our dataset is not powered to formally model dynamic changes. Our primary aim was risk stratification at the time of FALD diagnosis rather than diagnostic discrimination. Our findings suggest that a low AFP threshold of 4.5 ng/mL with a very high negative predictive value may help identify patients at particularly low risk, while those above the threshold may require closer surveillance. This information may be of great importance to clinicians involved in HCC surveillance in this population.

In addition, our results suggested a potential relationship between the hepatic venous waveform and the development of FALD-HCC. Studies have reported that the Doppler waveform of the hepatic vein can change in line with the progression of liver fibrosis, evolving from triphasic to monophasic and finally becoming flat.²³ Nakatsuka et al.²⁴ indicated that this finding also applies to patients with FALD and could serve as a useful, non-invasive indicator of liver fibrosis after Fontan surgery. This finding suggests that an association exists between liver fibrosis and hepatocarcinogenesis in this population. Further investigations are warranted.

Third, HCC surveillance in this cohort, which comprises periodic abdominal US and AFP measurements, may have prolonged the survival of patients who developed HCC. Several studies reported high mortality rates for FALD-HCC, with a 1-year survival rate of ~50%.⁴,⁵ Egbe et al.⁵ noted the possibility of the rapid occurrence of FALD-HCC; however, they also demonstrated the survival benefit of diagnosing HCC in the asymptomatic phase. In the present study, all patients were diagnosed with HCC in the asymptomatic phase and were classified as BCLC stage 0 or A, which has a better chance of being cured. Consequently, the 1-year survival rate after HCC diagnosis in this cohort was 100%. Furthermore, there was no statistically significant difference in survival between patients who developed HCC and those who did not, thus suggesting the utility of surveillance in this vulnerable population.

This study has a few limitations. It was conducted at a single institution and included only patients who were referred to the gastroenterology department, thus possibly introducing selection bias. Most patients were in a transition period from pediatric to adult care; however, the timing of referral and FALD diagnosis was not standardized. This may have biased the lead time for HCC detection and limited the generalizability of our findings. Although we enrolled all consecutive patients and applied a standardized surveillance protocol, the sample size remained limited owing to the rarity of FALD, and the smaller number of HCC events and limited follow-up duration may have affected the statistical robustness. In particular, the current statement regarding our survival findings should be interpreted cautiously. A longer follow-up period may reveal survival differences that were not captured in our current dataset. Furthermore, non-invasive indices, including platelet counts and FIB-4 index, may be confounded in this population owing to congenital anomalies, such as asplenia or polysplenia, which can affect the platelet count. Finally, we analyzed the HCC incidence from the time of FALD diagnosis to evaluate the liver-specific cancer risk, thus possibly underestimating the actual duration of carcinogenic risk because FALD itself develops insidiously after the Fontan procedure. We focused on patients in whom FALD had been clinically recognized in an adult practice setting. This pragmatic, clinically oriented definition does not capture subclinical, earlier-stage FALD, which may affect the generalizability of our findings. For these reasons, the results should be considered preliminary and validated in larger, longer-term cohorts. Collaborative, multicenter, prospective studies with standardized timing for observation are warranted.

The current expert opinion supports HCC surveillance for all adults with Fontan physiology, regardless of liver test or imaging abnormalities. Our study aimed to identify those who may benefit from closer or more intensive monitoring, rather than defining who should or should not be surveilled at all. In conclusion, this study demonstrated that patients with FALD have a clinically meaningful risk of developing HCC and that serum AFP levels of 4.5 ng/mL or more at the time of FALD diagnosis can aid in identifying high-risk individuals. Moreover, structured surveillance appears to enable the early diagnosis of HCC and favorable outcomes. These findings strongly support the implementation of routine HCC surveillance in this population and may guide clinicians in optimizing the long-term management of patients with FALD.