Caroli disease is a rare congenital malformation of the ductal plate characterized by non-obstructive, cystic dilatation of the intrahepatic bile ducts; when congenital hepatic fibrosis or cirrhosis co-occurs, the phenotype is termed Caroli syndrome [1]. Biallelic variants in polycystic kidney and hepatic disease 1 (PKHD1), which encodes fibrocystin/polyductin expressed in hepatobiliary and renal tubular epithelia, anchor Caroli phenotypes within the hepatorenal fibrocystic disease spectrum. PKHD1 is the established cause of autosomal recessive polycystic kidney disease (ARPKD), in which biliary abnormalities and congenital hepatic fibrosis are frequent components [2, 3].

Across published cohorts, most PKHD1-related biliary presentations are compound heterozygous. By contrast, the fetal consequences of homozygous PKHD1 variants, especially regarding whether they can manifest as an isolated hepatobiliary phenotype without renal involvement, remain poorly defined. This uncertainty has practical consequences for prenatal care. When fetal imaging shows intrahepatic bile-duct dilatation with normal kidneys, clinicians must distinguish Caroli disease from mimics such as choledochal cyst and cystic biliary atresia, and interpret exome findings in support of accurate counseling, perinatal planning, and recurrence-risk management.

Here we describe a fetus with diffuse intrahepatic bile-duct dilatation and normal kidneys in whom trio-based whole-exome sequencing (WES-trio) identified a homozygous PKHD1 c.2507T$>$C (p.Val836Ala) variant. By integrating fetal ultrasound, magnetic resonance imaging (MRI), and trio-based WES, this study aimed to define the prenatal presentation of this specific genotype and establish a diagnostic workflow [4]. The findings have direct clinical implications for counseling on prognosis and for informing reproductive decisions, such as recurrence prevention through preimplantation genetic testing.

Variants in the PKHD1 gene are the primary cause of ARPKD. The PKHD1-encoded protein, fibrocystin/polyductin complex (FPC), interacts with NEDD4 ubiquitin ligase-interacting protein 2 (NDFIP2), which interacts with several members of the C2-WWW-HECT domain-containing E3 ubiquitin ligase family. These interactions alter the subcellular localization and function of these ligases, leading to increased activity of transforming growth factor $\beta$ (TGF-$\beta$) and the epithelial sodium channel (ENaC) in bile duct epithelial cells. These molecular changes may ultimately contribute to hepatic and renal fibrosis [5]. Some patients with ARPKD present with clinical features resembling those of Caroli disease [6]. These findings suggest that variants in the PKHD1 gene not only affect renal development, but also involve the intrahepatic biliary system. Epidemiologically, Caroli disease is exceptionally rare ($\approx$1 per 1,000,000 population), which likely contributes to the scarcity of systematic prenatal series [7]. Furthermore, Caroli disease/Caroli syndrome (CD/CS) lie within the fibropolycystic liver disease spectrum derived from ductal-plate malformations, with phenotype varying by the level and extent of embryologic involvement [8, 9].

Caroli disease and ARPKD frequently co-occur, with a reported co-occurrence rate of up to 71% [10]. In addition, studies have reported cases of Caroli disease associated with PKHD1 gene variants in certain families, even in the absence of overt renal disease [11, 12]. Imaging is central to diagnosis. Ultrasonography may reveal cystic dilatation of the intrahepatic bile ducts, particularly in complex cases of Caroli syndrome associated with congenital hepatic fibrosis. In the prenatal setting, the spectrum of isolated congenital hepatobiliary malformations (ICHM) most commonly includes choledochal cysts and cystic biliary atresia, but also encompasses congenital hepatic cysts, gallbladder agenesis/non-visualization, and, less frequently, the Caroli spectrum [13]. A concise, literature-based overview of ICHM entities, prenatal imaging clues, and typical outcomes is provided (Table 1, Ref. [14, 15, 16, 17, 18]). Abdominal computed tomography (CT) may demonstrate cystic dilatation with the “central dot sign”, or communication between the bile duct and small ductules “tadpole sign” [19, 20, 21]. MRI and magnetic resonance cholangiopancreatography (MRCP) are key diagnostic modalities for Caroli syndrome, providing detailed information regarding lesion localization, size, and the presence of intrahepatic bile duct stones [22, 23, 24]. Fetal MRI complements ultrasound by delineating the cyst’s relationship to the biliary tree and the hepatoduodenal ligament, improving anatomic confidence for choledochal cyst and other ICHM [15]. For suspected CD/CS, prenatal MRI can reveal the central dot sign—the portal fibrovascular bundle within a dilated intrahepatic duct—thereby increasing diagnostic confidence and has been demonstrated in fetuses with ARPKD [14]. A recent prenatal series further reported a consistent central-dot sign across multiple fetuses diagnosed with Caroli disease, underscoring its in-utero utility [25].

ICHM may be subtle in mild cases, and misclassification between choledochal cyst (CC) and cystic biliary atresia (CBA) is common; in a systematic review of prenatal “biliary cysts”, approximately 22% proved to be biliary atresia postnatally [26]. Similarly, among perinatally detected subhepatic cysts, 15.4% were ultimately CBA despite sonographic appearances overlapping with CC [27]. Simple biometric cues can assist—cyst width $\le$2.5 cm strongly favors CBA over CC in early infancy [24]. Beyond the ultrasound, MRI better depicts the cyst’s location and continuity with the biliary tree, which helps avoid misdiagnosis and informs perinatal planning (see Table 1). CT and MRCP are important for postnatal confirmation; the central-dot sign is highly suggestive (though not pathognomonic) of the Caroli spectrum [28]. When gallbladder visualization is absent or equivocal, fetal MRI can re-identify a small/contracted gallbladder and help distinguish agenesis from biliary atresia, reducing false-positive NVFGB counseling [18]. By contrast, congenital hepatic cysts typically present late in gestation as isolated unilocular lesions without biliary communication, with generally favorable outcomes [17].

The c.2507T$>$C variant is a missense variant that results in the substitution of valine (Val) with alanine (Ala) at codon 836 of the PKHD1 gene. This variant has been associated with Caroli disease. It was first reported in 2014 in a dizygotic twin family in China, in which both twins carried c.2507T$>$C and c.2341C$>$T in a compound-heterozygous state and were diagnosed with Caroli disease. The mutations were consistent with familial cosegregation. However, one twin (Patient A) presented with asymptomatic splenomegaly at 1 year of age, was diagnosed with Caroli disease at 5 years due to anorexia and an epigastric mass, and developed cirrhosis and polycystic kidney disease by 7 years. In contrast, the other twin (Patient B) exhibited only intrahepatic biliary dilatation [29]. Liu et al. [30] studied three patients with ARPKD. Patient 2 was a 4-year-old male child who presented with polycystic kidney disease splenomegaly at 1 year of age, was diagnosed with Caroli disease at 5 years due to anorexia and an epigastric mass, and developed cirrhosis and polycystic kidney disease by 7 years. In contrast, the other twin (Patient B) exhibited only intrahepatic biliary dilatation [29]. Liu et al. [30] studied three patients with ARPKD. Patient 2 was a 4-year-old male child who presented with polycystic kidney disease and intrahepatic biliary dilatation. Genetic analysis identified c.2507T$>$C and c.2278C$>$T compound heterozygous variants in the PKHD1 gene, which were inherited from each parent, respectively. Yang et al. [31] reported an 8-month-old child with no abnormalities detected on prenatal ultrasound. At 8 months of age, the child was diagnosed with polycystic kidney disease and intrahepatic bile duct dilatation due to hepatomegaly. Genetic analysis revealed compound heterozygous variants in the PKHD1 gene (c.2507T$>$C and c.9292G$>$A), which were inherited from clinically unaffected parents. Qiu et al. [32] conducted a retrospective study on 29 Chinese children with PKHD1 gene variants, among whom one child was identified with compound heterozygous variants c.2507T$>$C and c.5935G$>$A, clinically presenting with mild dilatation of the intrahepatic bile ducts. Ishiko et al. [33] studied 32 patients with ARPKD, among whom six carried the c.2507T$>$C variant, all of whom were compound heterozygotes for c.2507T$>$C and other PKHD1 variants. Among these six patients, two had isolated Caroli disease, one had Caroli disease with liver fibrosis, one had liver fibrosis alone, and two developed liver cysts (Table 2, Ref. [29, 30, 31, 32, 34, 35]).

The patient reported in this study is homozygous for the c.2507T$>$C variant, and to our knowledge, no similar case has been previously documented. In the patient’s first pregnancy, fetal bile duct dilatation was detected at 32 weeks of gestation, and bile duct dilatation was again observed at 22⁺⁵ weeks in the current pregnancy. Taken together with previous literature reports, the onset time and severity of disease vary significantly among patients carrying the same variant—even among twins. Twin A developed asymptomatic splenomegaly at 1 year of age and, two years after the diagnosis of Caroli disease due to liver cirrhosis, presented with hypersplenism, severe anemia, and polycystic kidney disease. In contrast, Twin B exhibited only intrahepatic biliary dilatation [29]. The clinical presentation of this homozygous case appears more severe than many phenotypes described in the literature for compound heterozygous states. A key distinction is the early disease manifestation, with bile duct abnormalities detected at 22⁺⁵ weeks of gestation in our fetus, whereas other cases typically present at a later age. At the molecular level, the c.2507T$>$C (p.Val836Ala) variant resides within the transmembrane domain of fibrocystin, a protein encoded by the PKHD1 gene, which plays a crucial role in maintaining ciliary structure and bile homeostasis in bile duct epithelial cells [36]. An in vitro functional study has demonstrated that the localization of fibrocystin to cilia is closely linked to its biological function. The Val836Ala variant may impair fibrocystin function, leading to its mislocalization away from cilia and disruption of ciliary signaling pathways [37]. Unlike compound heterozygous variants, which only partially impair fibrocystin function, homozygous variants may result in a more substantial reduction of biallelic fibrocystin activity. This dose-dependent effect may explain the early embryonic development of bile duct abnormalities in this case. Furthermore, the variant site is located adjacent to a phosphorylation site (Ser833), and the amino acid substitution may interfere with interactions between fibrocystin and other cilia-related proteins, such as hepatocyte nuclear factor 1-beta (HNF1-$\beta$) [38], thereby exacerbating the development of bile-duct plate malformations.

Regarding the prognosis of Caroli disease, some patients maintain good long-term health with appropriate medical or surgical management. For example, the literature describes a case in which the patient remained in good health for more than 21 years following surgical intervention [39]. Another study reported four familial cases, in which a male patient presented with acute cholangitis at 6 years of age, while the other three sisters were asymptomatic at diagnosis. All four patients remained alive and asymptomatic for over 22 years at the final follow-up [40]. These findings suggest that, in certain cases, the long-term prognosis for patients with Caroli disease can be favorable. However, Caroli disease can also lead to several serious complications. The most common clinical manifestation is recurrent cholangitis. Additionally, Caroli disease is associated with portal hypertension and cirrhosis, as intrahepatic bile duct dilatation may promote gallstone formation, which can subsequently cause biliary colic, jaundice, and acute pancreatitis [24]. There is also an increased risk of malignancy. A study has shown that approximately 7% of patients who are not candidates for radical surgery develop cancer [41].

In summary, patients with Caroli disease exhibit diverse clinical manifestations and genetic variants. Although all cases involve PKHD1 gene mutations, the phenotypic expression varies significantly depending on the specific variant location. Despite our effort to collect all reported cases carrying the c.2507T$>$C variant, no homozygous case has been previously documented. Therefore, our report provides crucial clinical insights by being the first to link a homozygous PKHD1 variant to an early-onset, isolated prenatal Caroli disease. This finding is significant because it expands the known phenotype and alerts clinicians to consider this diagnosis even with normal fetal kidneys. As the diagnosis was made prenatally and the pregnancy was terminated, long-term clinical follow-up was not possible. Both parents were identified as carriers of the c.2507T$>$C variant and showed no abnormalities on ultrasound examination. Ultimately, achieving a definitive molecular diagnosis offers immediate and tangible value for at-risk families by enabling precise recurrence counseling and providing a clear path to prevent recurrence through the use of PGT-M. Recent prenatal reports also emphasize that integrating fetal ultrasound/MRI with trio-based exome sequencing improves recognition of CD/CS and supports tailored counseling [42]. This study provides the first evidence that a homozygous c.2507T$>$C variant in the PKHD1 gene can cause isolated fetal Caroli disease.

This study reports the first known prenatal diagnosis of isolated Caroli disease caused by a homozygous PKHD1 c.2507T$>$C (p.Val836Ala) variant. It demonstrates how integrated fetal ultrasound, fetal MRI, and WES-trio can resolve challenging hepatobiliary differentials in mid-gestation. The early, liver-limited presentation with normal fetal kidneys refines the prenatal phenotype associated with PKHD1. The homozygous state appears to correlate with earlier and more severe ductal involvement, suggesting a potential dosage effect on fibrocystin function. These insights provide immediate clinical utility by enabling precise genetic counseling, guiding perinatal management, and offering families a definitive strategy for recurrence-risk reduction through PGT-M. The case expands PKHD1 genotype–phenotype correlations and offers practical imaging–genomic teaching points for fetal medicine and clinical genetics. Further functional and longitudinal studies are needed to validate mechanisms and refine prognostication.

The data that support the findings of this study are available from the corresponding author upon reasonable request. Due to ethical and privacy restrictions related to human genomic data, some raw sequencing data are not publicly available. The publicly accessible data have been deposited in Zenodo https://doi.org/10.5281/zenodo.18263668.

JZ and HW designed the case report study. HW, HC, ZZ, and HZ acquired and interpreted the clinical and imaging data. XZ and HP performed the genomic sequencing and interpreted the genetic results. YW performed the literature search and data synthesis. ZX interpreted the imaging findings and designed the figures. JZ critically revised the manuscript for important intellectual content. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

This study was carried out in accordance with the guidelines of the Declaration of Helsinki. This study was approved by the Ethics Review Committee of Wenzhou People’s Hospital (Approval No. KY-202411-009). The authors confirm that written consent for the submission and publication of this case report, including all associated images, clinical data, and other information contained in the manuscript, has been obtained from the patient.

We are grateful to the patient and her family for their participation and for generously consenting to the publication of this case report.

This study was funded by Wenzhou basic scientific research project (Y2023528). This study was funded by Wenzhou City Major Scientific and Technological Innovation Project (Research on New Technologies for Women, Children and Reproductive Diseases) (ZY2024023).

The authors declare no conflict of interest.

Supplementary material associated with this article can be found, in the online version, at https://doi.org/10.31083/CEOG45794.