Depicting Characteristic Staghorn Vessels in Solitary Fibrous Tumor of the Liver With Contrast-Enhanced Ultrasound and Ultrasound Localized Microscopy: A Case Report

Introduction

Solitary fibrous tumor of the liver (SFTL) is an exceptionally rare tumor originating from fibroblastic mesenchymal tissues, comprising spindle cells and collagen [1]. The tumors are most commonly seen in middle-aged individuals, with no obvious sex predilection reported in the limited existing literature [2]. Meanwhile, SFTLs are asymptomatic and identified incidentally and are detected during routine physical examinations in the previous cases [2,3].

Although previous studies reported the B-ultrasound, magnetic resonance imaging (MRI), and contrast-enhanced ultrasound (CEUS) findings of SFTL, these findings lack specificity. In addition, most case reports describe only a single imaging modality, with a lack of reports on the multi-modality imaging characteristics in the same case [4,5]. An additional constraint is that most case reports are confined to a single imaging modality, with limited literature documenting multi-modality characteristics in the same patient [4,5]. While definitive diagnosis ultimately relies on histopathology, which reveals characteristic structures such as staghorn vessels, these thin vascular channels are too delicate to be clearly visualized on routine imaging modalities including computed tomography (CT), MRI, and CEUS. This combination of nonspecific imaging findings and limited multi modality data creates substantial challenges for accurate preoperative differential diagnosis from hepatocellular carcinoma (HCC), fibrolamellar HCC, and metastatic lesions [6].

Radical surgical resection with negative margins has been performed in previous cases of SFTL in which the tumor was confined to the liver and distant from major hepatic vessels [1]. However, when the tumor is adjacent to major hepatic vessels, direct primary resection carries high surgical risk and technical difficulty. Multi-modality imaging aids tumor characterization, provides pathological clues, and precisely delineates tumor-vessel relationships. Thus, preoperative imaging is critical for individualized treatment and risk reduction. Without reliable imaging, misdiagnosis is common, leading to delayed or overtreatment.

Ultrasound localized microscopy (ULM) is an advanced imaging modality built on the foundation of CEUS. By precisely identifying individual microbubble contrast agents and dynamically tracking their spatial displacements in serial CEUS image sequences [7], ULM achieves a high spatial resolution down to the tens of micrometers scale [8], thus allowing for the highly detailed networks of microvasculature. Zheng et al reported that ULM can visualize the morphological characteristics of microvasculature within breast lesions, which confers important diagnostic value for the differentiation of benign and malignant breast lesions [9]. However, the application of ULM in SFTL has been rarely reported to date.

This case report presents the microvascular characteristics of SFTL visualized by ULM, together with its multimodality imaging findings and treatment strategy. It provides new insights for the diagnosis and optimized treatment of SFTL.

Case Report

A 55-year-old female patient was incidentally found to have a liver mass on a chest CT performed as part of routine health screening, without abdominal pain. The patient had no systemic symptoms (weight loss, fatigue, or anorexia) or relevant risk factors (alcohol use and metabolic syndrome). Physical examination revealed no positive findings. Tumor markers, including CEA, AFP, FER, CA125, CA19-9, CA15-3, were tested, and none were elevated. The patient had no history of chronic liver disease or hepatitis B infection, and her liver function panel (aminotransferase levels, bilirubin, albumin, and international normalized ratio) was within the reference range (Table 1).

An abdominal ultrasound examination was conducted, which revealed a hepatic space-occupying lesion with undetermined pathological nature. It revealed an oval mass measuring approximately 58 × 48 mm in segments I/II/IV. The lesion demonstrated mixed echogenicity with heterogeneous internal echoes and scattered hyperechoic foci. Color Doppler imaging showed abundant intralesional and peripheral vascularity. The mass was observed to displace and compress the left portal vein and left hepatic vein, resulting in the loss of clear boundaries between the vessels and the tumor.

CEUS was subsequently performed to characterize the vascular perfusion patterns of the lesion. SonoVue contrast (1.5 mL) was administered through the antecubital vein, and CEUS images were obtained on a Mindray Resona A20 ultrasound unit for 2 minutes. The main mass exhibited hyperenhancement in the arterial phase, which progressed to hypoenhancement in the portal venous and delayed phases (Figure 1). Tortuous intratumoral microvessels and peripheral annular vessels were visualized during the arterial phase. This enhancement pattern definitively ruled out focal nodular hyperplasia (FNH) and hepatic metastases, narrowing the differential diagnosis spectrum. However, some features on CEUS and MRI, such as contrast washout and incomplete capsular enhancement, were highly similar to those of HCC, which increased the difficulty of differential diagnosis.

To further depict the vascular characteristics of the tumor, ULM was performed to generate a microscopic blood flow map during the arterial phase. ULM was conducted by a radiologist with over 5 years of abdominal ultrasound experience using a Mindray Resona A20 ultrasound system (Mindray Bio-Medical Electronics Co, Ltd, Shenzhen, China) equipped with a C6-1 linear probe (1–6 MHz) in adult abdominal mode, with scanning depth set at 15 cm, mechanical index at 0.110, and frame rate ≥300 fps. Before examination, patients received breathing training to maintain steady respiration and 3 to 5 seconds of breath-holding to reduce respiratory artifacts and were placed in a supine position with arms abducted to fully expose the target area, which was horizontally aligned using soft pads. An additional 1.5 mL of SonoVue contrast agent was intravenously injected via the antecubital vein as a bolus, followed by a 5-mL normal saline flush. Dynamic ULM imaging was initiated immediately after injection, with data acquired specifically in the arterial phase (20 seconds after injection, corresponding to peak contrast enhancement), and raw data were continuously acquired for no less than 60 seconds and stored for subsequent analysis.

The ULM demonstration revealed relatively abundant blood flow signals inside and around the mass. Furthermore, it confirmed abundant microvasculature within the tumor. A circumferential blood flow pattern was observed around the tumor periphery, and multiple branching microvessels inside the lesion, some exhibiting double bifurcations, presented a morphology staghornshaped vascular (Figure 1). This distinctive vascular feature provided crucial evidentiary support for the definitive diagnosis of SFTL.

Contrast-enhanced MRI was performed. A well-defined mass (54×48×56 mm) in segments I/II/IV demonstrated hypointensity on T1-weighted imaging and hyperintensity on T2-weighted imaging. Dynamic contrast-enhanced MRI revealed marked heterogeneous enhancement in the arterial phase, followed by washout in most regions during portal venous and delayed phases, with partial persistent enhancement. An incomplete peripheral capsular enhancement was noted (Figure 2).

The tumor’s proximity to the inferior vena cava meant that direct surgical resection carried substantial technical challenges and risks. Given the non-standard nature of transarterial chemoembolization (TACE) for SFTL, the treatment strategy was determined through multidisciplinary discussion and informed consent, with TACE recommended as initial treatment in view of the tumor’s hypervascular nature in the arterial phase rendering this approach feasible. Under digital subtraction angiography guidance, the patient underwent right femoral artery puncture under local anesthesia, with a 4F P-RH catheter advanced into the celiac artery. Angiography revealed tortuous, disordered hepatic vessels and ill-defined tumor staining, and a microcatheter was selectively inserted into the tumor-feeding branch of the left hepatic artery to perform treatment with a mixture of 5 mL Lipiodol and 1 vial of pingyangmycin, followed by Embosphere microspheres. Post-procedural angiography showed significantly diminished tumor staining without contrast extravasation. After catheter removal and puncture site hemostasis, the patient was transferred to the ward.

In the meanwhile, liver biopsy was performed during TACE (Figure 3). Liver biopsy showed absent normal hepatic architecture, spindle-shaped atypical tumor cells, intracytoplasmic pigment, and stromal collagenization, highly suggestive of SFTL, with definitive subtype pending surgical specimen. Immunohistochemistry results showed CD34 (+), ERG (−), SOX10 (−), CK (−), Vimentin (+), Ki-67 (5%+), HMB45 (−), Melan-A (−), S100 (−), H3K27Me3 (+, no loss), DOG-1 (−), CD117 (−), STAT6 (+). Special staining showed iron staining (−) and reticulin staining (+, fibrous stromal tissue).

CT and MRI at 2 months after TACE treatment demonstrated that the tumor regressed to 41×36 mm, representing a volume reduction of 47% and making the complete surgical resection feasible (Figures 4, 5).

Then, the patient underwent left hemihepatectomy with caudate lobectomy and cholecystectomy. Histopathological examination revealed a tumor composed of short spindle or ovoid cells with scant cytoplasm and nuclear atypia, accompanied by collagenized stroma and characteristic staghorn-shaped vessels. Focal degenerative cellular changes and extensive necrosis were also observed. Immunohistochemistry showed positivity for CD34 and STAT6, findings consistent with a solitary fibrous tumor with post-embolization changes (Figure 6).

The patient was followed up at 1, 3, and 6 months postoperatively. No symptom was reported, and no signs of tumor recurrence were evident on imaging studies. Tumor markers also remained within normal limits. Beyond imaging and tumor markers, the patient’s postoperative recovery was uneventful. She was completely free of pain, had no limitation in physical activity, and required no readmission during 6 months of follow-up.

Discussion

Previous cases have reported the imaging findings of SFTL, and most have focused on CT and MRI characteristics. SFTL often present as a heterogeneously enhancing mass in CT images [1–3,5]. On MRI, it appears hypointense on T1-weighted images and hyperintense on T2-weighted images. Only 1 case reported the findings of CEUS [4]. The described CEUS features [4], which encompass peripheral arterial enhancement with rapid centripetal filling, peak-phase heterogeneity, and subsequent wash-out, were present in our case as well. However, these characteristics are not characteristic and are shared by many other liver tumors.

To the best of our knowledge, ULM has rarely been reported to visualize the intratumoral microvasculature of a solitary fibrous tumor. ULM is an advanced ultrasound technique that enables microvascular depiction at a resolution finer than the hundreds-of-microns scale [10]. While CEUS revealed tortuous vessels within the tumor, ULM delineated these subtle structures with far greater clarity, demonstrating staghornshaped vascular patterns in some areas. Furthermore, real-time CEUS during the arterial phase identified not only tortuous microvessels inside the tumor but also surrounding annular vessels.

These detailed imaging characteristics closely correspond to the underlying pathological features of the lesion. Histologically, SFTL is characterized by spindle to ovoid cells exhibiting a predominantly patternless growth, with abundant collagen in the stroma. This typically presents as a hyperechoic nodule on 2dimensional ultrasound. Classic histological features include fascicles of spindle cells arranged in a fibrous stroma, along with the characteristic staghorn-like vascular pattern. This vasculature is characterized by thin-walled, branching vessels which serve as a crucial diagnostic clue for SFTL [11]. Notably, this vascular architecture can be directly visualized as a staghornlike pattern on ULM.

The enhancement pattern and vascular features identified via CEUS and ULM are critical for differentiating SFTL from well-differentiated HCC, FNH, and angiomyolipoma (AML). The systematic comparison between these 4 tumors is presented in Table 2 [1,4–6,12–18]. Key imaging findings that support the diagnosis of SFTL include intrinsic and extensive intratumoral small-vessel proliferation, which is demonstrated by ULM/CEUS as characteristic staghorn vessels characterized by regular, tree-like branching without disorganization or cross-linking, along with arterial phase hyperenhancement driven by this intrinsic vascular network. Unlike HCC, which relies on external, disorganized feeding vessels (penetrating, tortuous, cross-linked) to expand arterial supply and shows characteristic delayed-phase washout [19], the current lesion lacked external feeding vessels and delayed washout. Hence, HCC was ruled out. For a female patient without hepatitis B infection, a lesion showing arterial phase hyperenhancement with washout requires differentiation of FNH [20]. However, in this case, the characteristic central scar and spoke-wheel enhancement of FNH were not observed, making FNH a less likely diagnosis. Patchy hyperechoic areas suggested the presence of fatty components, which is one of the key imaging features of AML. However, blood vessels in AML typically show a tortuous course, with heterogeneous enhancement observed within the tumor. In this case, the presence of fatty components and heterogeneous enhancement did not allow complete exclusion of AML. Nevertheless, the tumor vessels showed a distinct staghorn-like pattern, so AML was not regarded as the primary diagnostic consideration.

Our case also provides multi-modality imaging in 1 patient. Imaging modalities complemented each other in guiding treatment. CEUS clarified vascular details, CT defined spatial relationships for surgical planning, and MRI characterized tissue components (fat, fluid, soft tissue). Their integrated findings ensured accurate evaluation and optimized decisionmaking.

Complete surgical resection with negative margins remains the gold standard treatment for localized disease [1]. However, there is currently no established consensus regarding the management of SFTL adjacent to critical hepatic vascular structures in which surgical risks are prohibitively high. Given that SFTL is a hypervascular neoplasm whose growth and metastatic dissemination are fundamentally angiogenesis-dependent [21], the tumor in this case was large in volume and located adjacent to major intrahepatic vessels, resulting in an excessively high risk of intraoperative hemorrhage if direct surgical resection were performed. Therefore, TACE performed before surgery can reduce the risk of intraoperative bleeding by embolizing the tumor’s blood supply arteries; meanwhile, it can shrink the tumor size through local effects to improve the feasibility of surgical resection for larger or complex liver tumors[1,22,23]. Thus, preoperative TACE was performed. TACE successfully achieved significant tumor volume reduction, thereby facilitating subsequent curative resection. Hence, TACE may be considered in selected cases in which complete surgical resection is technically challenging. In our case, left lobectomy with caudate lobectomy was chosen over limited enucleation for hepatic SFTL. Dorsal location far from the hepatic capsule made enucleation technically difficult with incomplete resection risks, and preoperative biopsy only confirmed the diagnosis but lacked sufficient histological data to assess malignant potential due to limited sampling. Meanwhile, since the tumor involved hepatic segments I/II/IV and was adjacent to the gallbladder, cholecystectomy was also required. Therefore, radical resection was performed to avoid residual lesions in accordance with clinical principles.

Despite the encouraging findings, this study has several limitations. First, although our study reported that ULM can reveal the characteristic staghorn-shaped blood vessels in SFTL, the novelty of ULM should be tempered by noting the lack of comparative validation. Second, this is a case report, which inherently limits the generalizability of the results. Due to the low incidence of this disease, our findings will need to be validated in a larger number of future cases. Third, the follow-up period was only 6 months, which may not be sufficient to fully assess long-term outcomes, highlighting the need for extended monitoring in future research.

Conclusions

SFTL is a rare tumor. The multimodal imaging features were presented in this case, in which CEUS and ULM revealed tortuous vessels within the lesion, potentially offering a clue for differentiating SFTL. For tumors in challenging locations adjacent to the portal vein, TACE can serve as an effective bridge to surgery by reducing tumor volume and facilitating subsequent resection. However, this report is limited by its single-case design, disease rarity, and short follow-up. Further studies with larger cohorts and longer follow-up are warranted to validate these findings and optimize clinical management of SFTL.