Prenatal discovery of hereditary hemorrhagic telangiectasia-associated hepatic arteriovenous malformations by multimodal ultrasound and whole-exome sequencing: a rare case description and literature analysis

Introduction

Hereditary hemorrhagic telangiectasia (HHT), also known as Osler-Weber-Rendu disease, is a rare autosomal dominant disorder characterized by the absence of capillaries between arteries and veins, leading to vascular lesions such as telangiectasias and arteriovenous malformations (AVMs) (1). The condition affects approximately 1 in 5,000 individuals and is primarily attributed to mutations in the endoglin gene (associated with HHT1) and the activin A receptor-like type 1 gene (ACVRL1 or ALK-1), which is linked to HHT2. Liver involvement is predominantly associated with mutations in the ACVRL1 gene (1). Current research indicates that vascular malformations in HHT develop as focal lesions in discrete anatomic locations rather than systemic defects, suggesting that primary genetic mutation alone cannot fully account for HHT pathogenesis. The “second-hit” hypothesis has been proposed, suggesting that the initial germline mutation in HHT genes leads to haploinsufficiency in endothelial cells (first hit), while a second hit from somatic mutations, inflammation, hypoxia, or vascular injury further drives disease expression and may act synergistically (2). Hepatic arteriovenous malformations (HAVMs) are found in 41% to 78% of adults with HHT but are rare in neonates (3). To date, no prenatal cases of HHT-associated HAVMs in fetuses have been reported. This report presents the first prenatal diagnosis of HHT-associated HAVMs in a fetus, identified via ultrasound and confirmed by whole-exome sequencing (WES), which revealed an ACVRL1 mutation.

Case presentation

A 29-year-old primigravida (G1P0) with no family history of genetic disorders underwent a routine prenatal ultrasound examination at 34 weeks of gestation. The pregnancy had been uncomplicated up to this point, with both the 24-week and 28-week routine prenatal ultrasounds showing no fetal anomalies. However, the 34-week two-dimensional ultrasound demonstrated enlargement of the right hepatic lobe. The common hepatic artery and its branches, including the left and right hepatic arteries, were markedly dilated. The maximum diameter of the extrahepatic hepatic artery was 0.40 cm, while the maximum diameter of the intrahepatic hepatic artery was 0.35 cm. Tortuous and dilated intrahepatic artery branches indicative of intrahepatic hypervascularization were predominantly observed in the right lobe. Dilated arterial branches were also identified in the subcapsular regions of the right lobe, suggesting peripheral hypervascularization (Figure 1A). Doppler ultrasound (DUS) identified significant dilation in both the hepatic arteries and veins, accompanied by multiple arteriovenous shunts between them (Figure 1B). The dilated hepatic artery exhibited high-velocity, low-resistance blood flow, with a peak velocity of 116 cm/s and a resistance index (RI) of 0.38. Moreover, the portal venous sinus demonstrated reversed blood flow, characterized by a pulsatile blood flow spectrum (Figure 1C,1D). Furthermore, four-dimensional spatiotemporal image correlation (4D STIC) identified notable dilation of the extrahepatic hepatic arteries, intrahepatic and peripheral hypervascularization, and multiple arteriovenous shunts (Figure 1E). These ultrasound findings were consistent with excessive formation and extensive dilation of the hepatic arteries, accompanied by hepatic artery-hepatic vein fistulas and hepatic artery-portal vein fistulas. Echocardiography showed mild right heart enlargement with a cardiothoracic area ratio of 0.38 and a fetal cardiovascular performance score (CVPS) of 9, likely secondary to the increased cardiac preload associated with the HAVMs.

Figure 1.

Prenatal Doppler ultrasound findings at 34 weeks of gestation. (A) Significant dilation was observed in both the intrahepatic and extrahepatic arteries, with multiple arteriovenous shunts between the hepatic arteries and veins: (a) common hepatic artery; (b) left hepatic artery; (c) right hepatic artery; (d) intrahepatic and peripheral hypervascularization; (e) arteriovenous shunts between hepatic arteries and veins; (f) hepatic vein. (B) Dilation of the hepatic veins, with visible arteriovenous shunts between the hepatic arteries and veins: (a) hepatic vein; (b) hepatic artery; (c) arteriovenous shunts between the hepatic arteries and veins. (C) Reversed blood flow in the portal vein: (a) portal venous sinus; (b) proper hepatic artery; (c) umbilical vein; (d) ductus venosus. (D) The hepatic artery demonstrates a high-velocity, low-resistance flow spectrum, while the portal vein exhibits reversed blood flow with a pulsatile waveform. (E) 4D STIC reveals dilation of the hepatic veins, intrahepatic and peripheral hypervascularization, and multiple arteriovenous shunts: (a) hepatic vein; (b) intrahepatic and peripheral hypervascularization; (c) arteriovenous shunts between hepatic arteries and veins; (d) umbilical vein; (e) ductus venosus; (f) portal venous sinus. 4D STIC, four-dimensional spatiotemporal image correlation; ED, end diastole; HA, hepatic artery; PV, portal vein; PS, peak systole; RI, resistance index.

Fetal magnetic resonance imaging (MRI) revealed a low-signal intensity area on T1-weighted imaging (T1WI) and a slightly high-signal intensity area on T2-weighted imaging (T2WI) in the posterior right lobe of the fetal liver. The area measured about 2.55 cm × 3.45 cm × 2.97 cm, with no diffusion restriction on diffusion-weighted imaging (DWI). This area demonstrated increased and disordered small vessels, with a prominent vessel draining into the right hepatic vein. Multiple tortuous and fine vascular shadows were observed within the porta hepatis. Based on these findings, HHT was suspected, and amniocentesis for genetic testing was recommended.

Given the concern for a poor prognosis, the parents elected to terminate the pregnancy. WES of the tissue obtained from the induced labor revealed a c.232delA (p.Arg78Glyfs*44) mutation within the ACVRL1 gene, associated with HHT2 and inherited from the mother. However, the family declined a fetal autopsy. Follow-up surveillance for HHT in both the proband’s mother and the aunt, who also carries the same ACVRL1 mutation, revealed no findings consistent with HHT, including no visceral AVMs or other manifestations such as epistaxis. All procedures performed in this study were in accordance with the ethical standards of the institutional and/or national research committee(s) and the Declaration of Helsinki and its subsequent amendments. Written informed consent was obtained from the patient for publication of this article and accompanying images. A copy of the written consent is available for review by the editorial office of this journal.

Discussion

The clinical diagnosis of HHT is traditionally based on the Curacao criteria, which include: (I) spontaneous, recurrent epistaxis; (II) mucocutaneous telangiectasia; (III) visceral AVMs in organs such as the lungs, spinal cord, brain, or liver; and (IV) a family history of HHT (3). A definite diagnosis requires at least three criteria, while two suggest a suspected diagnosis. In fetuses, manifestations such as mucocutaneous telangiectasia or epistaxis remain undetectable due to the constraints of the intrauterine environment. However, visceral AVMs may emerge as the earliest detectable indicators during prenatal assessments. Recent advances in molecular genetics have enabled more precise diagnosis of HHT (4). Despite these advances, genetic testing still misses approximately 15% of patients who fulfill the clinical diagnostic criteria for HHT, underscoring the necessity of a combined clinical and genetic diagnostic approach (5).

Hepatic involvement in HHT can be evaluated using imaging modalities such as angiography, multislice computed tomography (MSCT), MRI, and DUS (6). Among these, angiography is considered the gold standard but is invasive. MSCT and MRI, while effective, are limited by cost, radiation exposure, and the need for contrast agents (7). DUS offers a non-invasive assessment of both morphological and hemodynamic changes, avoiding the limitations of cost, radiation exposure, and the need for contrast agents. This makes DUS the preferred first-line diagnostic tool (7). Several diagnostic criteria based on DUS for HHT-related liver involvement have been proposed. Caselitz et al. identified a common hepatic artery diameter >0.7 cm and intrahepatic hypervascularization as primary criteria, which demonstrate high specificity. Secondary criteria include a proper hepatic artery peak velocity >110 cm/s, RI <0.6, portal vein peak velocity >25 cm/s, and extrahepatic artery tortuosity or dilation. A definitive diagnosis requires meeting two primary criteria or one primary criterion plus at least two secondary criteria (8). Buscarini et al. further classified HAVMs according to severity, with grade 0+ representing the earliest stage, characterized by peripheral hypervascularization. Severity is based on the extent of intrahepatic hypervascularization: mild if only the common hepatic artery is dilated beyond 6 mm, moderate if both intrahepatic and extrahepatic arteries are dilated, and severe if there are complex changes such as tortuosity and rigidity in intrahepatic arteries, along with hepatic or portal vein dilation and abnormal flow spectra (7). Buonamico et al. emphasized intrahepatic hypervascularization and the “color-spot” sign, which correlates with Buscarini’s peripheral hypervascularization. Validation using MSCT demonstrated that the “color-spot” sign has a sensitivity of 95% (6). In summary, despite the variability in DUS criteria, hepatic artery dilation and intrahepatic hypervascularization consistently demonstrate high specificity for diagnosing HHT-related liver involvement, whereas peripheral hypervascularization exhibits high sensitivity.

Visceral AVMs are characterized by abnormal artery-to-vein connections. Visualizing the fetal hepatic artery via ultrasound is usually difficult, but its visibility often indicates dilation. Hepatic artery dilation, along with any of the following signs, strongly suggests the presence of HAVMs: (I) the presence of a “vascular lake” within the liver; (II) observation of reversed flow direction in the portal vein, which typically flows in the same direction as the adjacent hepatic artery (6,9); (III) a pulsatile Doppler waveform within the portal vein; and (IV) dilated hepatic veins with increased flow velocity. A review of existing literature reveals that none of the previous seven reports on HAVMs involved WES, indicating a significant gap in understanding the genetic basis of this condition (9-15). In our cases, we observed not only HAVMs but also marked intrahepatic and peripheral hypervascularization. These prenatal observations prompted targeted genetic testing, which revealed an ACVRL1 mutation through WES, thereby confirming the diagnosis of HHT. Notably, similar to diagnostic criteria in adults, fetal intrahepatic and peripheral hypervascularization represent key diagnostic markers for HHT-associated hepatic involvement in fetuses. Upon identification of fetal HAVMs, a comprehensive evaluation should be performed to assess signs of intrahepatic and peripheral hypervascularization. Prenatal two-dimensional grayscale ultrasound and DUS provide essential diagnostic information, while 4D STIC enhances precise visualization of arteriovenous shunts, thereby enabling accurate prenatal diagnosis. The main differential diagnosis for HHT is capillary malformation-arteriovenous malformation (CM-AVM) syndrome, which features multiple capillary malformations. Distinct from HHT, its AVMs primarily involve the central nervous system, skin, muscles and bones, with minimal visceral involvement. Importantly, HAVMs are both rarer and milder than those in HHT. Through detection of RASA1 or EPHB4 mutations, genetic testing confirms the diagnosis (16).

Furthermore, neonates with HHT-related HAVMs often experience high-output heart failure (17,18). In this case, increased cardiac output led to mild right heart enlargement, with a cardiothoracic ratio of 0.38, which is at the upper limit of the normal range (≤0.35). Hepatic artery embolization is generally inadvisable in adults with HAVMs due to the high risk of ischemia and increased mortality. However, it can be a lifesaving intervention in neonates. As reported by Saleh et al., 3 of 4 neonates who underwent early embolization survived the neonatal period. Moreover, during follow-up, HAVMs exhibited progressive involution in two of these infants (17). This difference is attributed to fundamental vascular differences between neonates and adults. Neonatal livers have minimal hepatic arterial dependence and a robust collateral circulation. Additionally, their regenerative capacity far exceeds that of adults, which enables them to tolerate embolization better. In contrast, adult livers have a higher hepatic arterial blood supply, and the presence of porto-venous shunts leaves them vulnerable to ischemia when embolized, thereby increasing the risk of biliary or hepatic necrosis (19). Therefore, establishing an early and accurate prenatal diagnosis through ultrasound is critical for reducing postnatal complications and improving clinical outcomes.

In conclusion, prenatal multimodal ultrasound is vital for diagnosing HHT-related liver vascular malformations and plays a key role in perinatal management. Genetic testing confirmed HHT in the fetus and identified ACVRL1 mutations in both the mother and aunt, emphasizing the importance of genetic screening and highlighting the unique genetic implications of this case.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All procedures performed in this study were in accordance with the ethical standards of the institutional and/or national research committee(s) and the Declaration of Helsinki and its subsequent amendments. Written informed consent was obtained from the patient for publication of this article and accompanying images. A copy of the written consent is available for review by the editorial office of this journal.