Elevated central venous pressure in those with Fontan circulation causes liver congestion and hepatomegaly. We assessed if liver volume by magnetic resonance imaging (MRI) is associated with adverse cardiovascular outcomes. Retrospective study of 122 patients with Fontan circulation who were >10 years old and had a liver MRI with magnetic resonance elastography. Liver volume (ml) was measured by manual segmentation from axial T2-weighted images and was indexed to body surface area. The composite outcome included death, heart transplant, ventricular assist device placement, or nonelective cardiovascular hospitalization. The median age at the time of MRI was 18.9 (interquartile range 15.8 to 25.9) years, and 47% of the patients were women. The mean indexed liver volume was 1,133 ± 180 ml/m 2 . Indexed liver volume was not significantly associated with age, years since Fontan, or with liver stiffness ( r = 0.15, p = 0.10), but was positively correlated with Fontan pressure ( r = 0.32, p = 0.002). Over a median follow-up of 2.1 (0.8 to 4.2) years, 32 patients (26%) experienced the composite outcome. Higher indexed liver volume was associated with a greater hazard for the composite outcome (hazard ratio per 1 SD increase = 1.74, 95% confidence interval 1.27 to 2.35, p = 0.0004) but increased liver stiffness was not significantly associated with the composite outcome (hazard ratio per 1 SD increase 1.44, 95% confidence interval 0.90 to 2.21, p = 0.11). In conclusion, greater liver volume indexed to body surface area is associated with unfavorable hemodynamics and adverse outcomes in patients with Fontan circulation. Liver volume may be a useful, simple imaging biomarker in adolescents and adults with Fontan circulation.
Fontan-associated liver disease (FALD) affects nearly all patients with Fontan circulation and may progress over time. Congestion is a hallmark of FALD and, as a consequence, both liver enlargement and increased liver stiffness are common. Eventually, liver fibrosis develops with some patients progressing to cirrhosis. Liver imaging is a standard component of Fontan surveillance to monitor the progression of FALD and portal hypertension, screen for hepatocellular carcinoma, and as an imaging biomarker of Fontan circulation hemodynamics. Liver stiffness, measured by magnetic resonance elastography (MRE), may serve as a biomarker of Fontan circulation function and may identify patients at highest risk of poor outcomes. Liver volume can also be measured by magnetic resonance imaging (MRI), although this straightforward metric has not been investigated as an imaging biomarker of Fontan hemodynamics. This retrospective study aims to evaluate associations between liver volume and clinical factors and cardiovascular outcomes in patients with a Fontan circulation with the hypothesis that liver volume and liver stiffness provide independent, additive information to identify patients at increased risk for adverse outcomes.
The study cohort included patients with a Fontan circulation who were >10 years old and underwent liver MRI with MRE performed at our institution between January 1, 2010 and February 1, 2020. Patients were excluded (n = 3) if they had imaging artifact that precluded the measurement of liver volume or if they did not receive further follow-up at our institution after liver MRI (n = 1). For patients with more than 1 liver MRI/MRE, we utilized data from the first MRI. Clinical surveillance for adolescents and adults with a Fontan at our institution includes liver MRI/MRE in the absence of an MRI contraindication. This retrospective study was approved by the Cincinnati Children’s Hospital Institutional Review Board and was performed in compliance with the Health Insurance Portability and Accountability Act. The requirement for informed consent/assent was waived because of the retrospective nature of the study.
Liver MRI and MRE were performed on 1.5-T MRI scanners (Ingenia; Philips Healthcare, Best, and Netherlands; and Signa HDx or Optima MR450w; GE Healthcare, Waukesha, Wisconsin). Liver volume was measured by assisted manual segmentation from axial T2-weighted images (Vitrea, Vital Images) ( Figure 1 ). To adjust for differences in body size, liver volume was indexed to body surface area (BSA) calculated by the Du Bois method. Complex chemical shift-encoded MRI was used for determination of MRI proton density fat fraction. Hepatic steatosis was defined as percentage of liver fat ≥5%. For elastography, both two-dimensional gradient-recalled echo and spin-echo echo-planar imaging techniques were used during the study period which have been shown to yield comparable results. For each MRE examination, 4 axial images were obtained through the mid liver, avoiding the superior-most and inferior-most portions of the liver. Regions of interest were manually drawn on each image to measure shear stiffness, guided by phase and magnitude images. Regions of interest included the right hepatic lobe and segment IV of the left lobe, excluding visible blood vessels and avoiding areas of artifact and tissue within 1 cm of the liver capsule. To evaluate the combination of liver volume and liver stiffness, the indexed liver volume was multiplied by the liver stiffness to derive the size-stiffness product.
Clinical and outcome measures were extracted from medical records. Demographic, laboratory, cardiac MRI, cardiac catheterization, and exercise testing data closest to the liver MRI were collected. Imaging and laboratory data were used to calculate the Model for End-stage Liver Disease Excluding INR international normalized ratio score, aspartate aminotransferase-to-platelet ratio index, , and VAST (Varices, Ascites, Splenomegaly, and Thrombocytopenia) score. The VAST score includes features of portal hypertension where 1 point is given for each varices (portosystemic shunts seen on liver MRI), ascites, splenomegaly, or thrombocytopenia. Glomerular filtration rate was estimated using Chronic Kidney Disease Epidemiology Collaboration equations using creatinine for patients aged >19 years and with the bedside Schwartz formula for patients aged <19 years. , Patients with missing clinical data were excluded from the analysis for that clinical variable.
Continuous variables are presented as mean ± SD for normally distributed variables and median (25th to 75th percentile) for non-normally distributed variables. Liver volume indexed to BSA (indexed liver volume) was categorized according to quartile to assess for associations with clinical variables. Liver volume, liver stiffness, and the liver size × stiffness product were also standardized to facilitate analysis of these continuous variables with an equivalent scale (i.e., per 1 SD change). Unadjusted linear regression was used to compare continuous variables between 2 groups, and either the chi-square or Fisher’s exact test was used to analyze categoric variables between groups. The outcomes of interest included death, heart transplant, ventricular assist device (VAD) placement, or nonelective cardiovascular hospitalization any time after the liver MRI and within 1 year of the MRI. The present study includes a cohort we previously studied (n = 70) but also includes younger patients (n = 44) and patients who had their first MRI performed since completion of the previous study (n = 11). Cox proportional hazards models were developed to estimate how liver stiffness and indexed liver volume affected the hazard for the composite outcome. Time-to-event was defined from liver imaging until the first occurrence of any component events in the composite outcome with censoring of event-free patients at the most recent clinical follow-up date when event status was known. A 2-sided p value <0.05 was considered statistically significant. Analyses used JMP (version 15, SAS Institute, Inc., Cary, North Carolina) and GraphPad Prism 9 (GraphPad Software Inc., San Diego, California).
The study included 122 patients who underwent liver MRI with liver volume measurement between 2010 and 2020. Demographic and clinical characteristics of the study population are shown in Table 1 . The mean liver volume was 1,940 ± 503 ml. There was a strong positive correlation between liver volume and weight ( r = 0.89, p <0.0001), height ( r = 0.60, p <0.0001), and BSA ( r = 0.78, p <0.0001). Liver volume was also associated with age ( r = 0.29, p =0.002) and time since the Fontan procedure ( r = 0.30, p = 0.0007). The mean indexed liver volume was 1,133 ± 180 ml/m 2 . Indexed liver volume was not significantly associated with weight, height, age, or years since the Fontan. Of the 102 patients where hepatic fat fraction was assessed, hepatic steatosis was present in 12 patients (12%). There was no difference in indexed liver volume in patients with versus patients without steatosis.
| Variable | Patients with liver volume (n=122) |
|---|---|
| Age at liver MRI (years) | 18.9 (15.8-25.9) |
| Time since Fontan (years) | 16.7 ± 6.9 |
| Women | 57 (47%) |
| White Black Asian Other race | 109 (89%) 8 (7%) 2 (1.5%) 3 (2.5%) |
| Weight (kg) | 66.3 ± 20.5 |
| Height (cm) | 163.3 ± 13.0 |
| Body surface area (m 2 ) | 1.70 ± 0.30 |
| Body mass index (kg/m 2 ) | 24.6 ± 6.0 |
| Dominant ventricular morphology Left ventricle Right ventricle Co-dominant | 67 (55%) 52 (43%) 3 (2%) |
| Echocardiogram (n=122): Moderate or greater ventricular dysfunction Moderate or greater atrioventricular valve regurgitation | 12 (10%) 16 (13%) |
| Cardiac Magnetic Resonance Imaging (n=80): End diastolic volume (ml/m 2 ) End systolic volume (ml/m 2 ) Systemic ventricular ejection fraction (%) | 90 (75 -115) 44 (36 – 60) 51 (45 – 56) |
| Cardiac catheterization (n=97): Fontan pressure (mmHg) Ventricular end diastolic pressure (mmHg) Cardiac index (L/min/m 2 ) Pulmonary vascular resistance (iWu) Aortic saturation (%) | 13 ± 3.7 10 ± 3.6 3.0 ± 0.8 1.3 (0.98-1.84) 92 (89-94) |
| Exercise test (n=107): Peak VO 2 (mL/kg/min) % predicted peak VO 2 | 23.2 ± 7.3 60 ± 15.7 |
There were 97 patients (80%) who had a clinically indicated cardiac catheterization. The median time between liver MRI and heart catheterization was 0.9 (0.2 to 2.4) years. In this subset, higher indexed liver volume was associated with higher Fontan pressure ( r = 0.32, p = 0.002) and higher ventricular end-diastolic pressure ( r = 0.25, p = 0.01). There was no association between indexed liver volume and cardiac index or pulmonary vascular resistance.
In patients who had a cardiac MRI (n = 80), there was no association between indexed liver volume and systemic ventricular morphology (left or right ventricle), ventricular ejection fraction, or ventricular volumes. Additionally, there was no association between indexed liver volume and percent predicted peak VO 2 on exercise testing (n = 107). Associations between indexed liver volume quartiles and clinical variables are shown in Table 2 .
| Indexed liver volume quartile | |||||
|---|---|---|---|---|---|
| Variable | 1 (<1021 mL/m 2 ) | 2 (1021-1103 mL/m 2 ) | 3 (1103-1226 mL/m 2 ) | 4 (>1226 mL/m 2 ) | P value |
| Age (years) | 17.9 (15.0-25.3) | 19.4 (17.2-28.3) | 20.6 (17.1-29.9) | 17.5 (13.6-22.8) | 0.12 |
| Time since Fontan (years) | 15.0 (10.7-21.2) | 16.3 (13.8-25.0) | 17.0 (12.9-24.4) | 14.8 (9.7-19.4) | 0.35 |
| Female | 17 (57%) | 16 (52%) | 12 (39%) | 13 (43%) | 0.50 |
| Fontan type Atriopulmonary Lateral tunnel Extracardiac conduit Other | 7 (23%) 7 (23%) 15 (50%) 1 (3%) | 2 (6%) 16 (52%) 12 (39%) 1 (3%) | 5 (16%) 13 (42%) 13 (42%) 0 | 1 (3%) 15 (50%) 14 (47%) 0 | 0.13 |
| Ventricular morphology Left Right | 16 (53%) 14 (47%) | 16 (52%) 15 (48%) | 17 (59%) 12 (41%) | 18 (62%) 11 (38%) | 0.85 |
| VAST score | 0.06 | ||||
| <2 ≥2 | 27 (90%) 3 (10%) | 25 (81%) 6 (19%) | 11 (35%) 20 (65%) | 20 (67%) 10 (33%) | |
| Ventricular EF (%) | 51 ± 7 | 52 ± 6 | 49 ± 9 | 48 ± 8 | 0.31 |
| Indexed EDV (mL/m 2 ) | 97 ± 39 | 89 ± 24 | 98 ± 30 | 109 ± 36 | 0.30 |
| Indexed ESV (mL/m 2 ) | 52 ± 28 | 43 ± 15 | 52 ± 24 | 59 ± 29 | 0.23 |
| Ventricular EDP (mmHg) | 8.5 ± 2.4 | 10.1 ± 2.9 | 12.0 ± 3.8 | 10.7 ± 4.0 | 0.003 |
| Fontan pressure (mmHg) | 11.6 ± 2.3 | 13.3 ± 1.7 | 14.5 ± 4.1 | 14.7 ± 4.8 | 0.008 |
| Cardiac index (L/min/ m 2 ) | 3.1 ± 0.7 | 3.2 ± 0.6 | 2.6 ± 0.6 | 3.2 ± 1.1 | 0.04 |
| PVRi (WU* m 2 ) | 1.2 ± 0.6 | 1.5 ± 0.7 | 1.7 ± 0.8 | 1.6 ± 1.0 | 0.20 |
| SVRi (WU*m 2 ) | 20.6 ± 5.2 | 20.5 ± 6.0 | 27.5 ± 12.6 | 21.2 ± 6.8 | 0.02 |
| Peak VO 2 (% predicted) | 65 ± 13 | 59 ± 14 | 58 ± 15 | 59 ± 19 | 0.33 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
