Perivascular epithelioid cell tumors (PEComas) are a rare group of mesenchymal neoplasms characterized by distinctive histological and immunohistochemical features. This tumor family was first described by Bonetti et al (1) in 1992, initially encompassing renal angiomyolipoma and the clear cell ‘sugar’ tumor of the lung.

Subsequently, additional neoplasms were incorporated into the PEComa family based on their characteristic co-expression of melanocytic markers (HMB-45, Melan-A) and smooth muscle markers (SMA), including lymphangioleiomyomatosis (2). The same research group later expanded the category to include clear cell myomelanocytic tumors of the falciform ligament (ligamentum teres) and other clear cell tumors across multiple organs (3), such as the urinary bladder, prostate, uterus, ovary, vulva, vagina, lung, pancreas and liver (hepatic angiomyolipoma). Molecular studies have established a strong link between PEComas and the tuberous sclerosis complex (TSC), an autosomal dominant disorder caused by mutations or deletions of the TSC1 (9q34) or TSC2 (16p13.3) genes (4,5). Clinically, TSC is associated with intellectual disability, seizures, and various neoplasms, including angiomyolipomas, subependymal giant cell astrocytomas, cutaneous angiofibromas, cardiac rhabdomyomas, lymphangioleiomyomatosis, and multifocal micronodular pneumocyte hyperplasia. Further research has demonstrated that TSC1/TSC2 loss results in activation of Rheb/mTOR/p70S6K signaling pathway, a key driver of tumorigenesis in PEComas and the biological rationale for the use of mTOR inhibitors in selected cases (6-8). It is noteworthy, however, that PEComas do not occur more frequently in patients with clinical TSC. Instead, most PEComas harbor somatic mutations in TSC1 or, more commonly, TSC2 without affected individuals meeting the diagnostic criteria for systemic TSC. This distinction highlights the role of TSC gene alterations as driver mutation in PEComa tumorigenesis, rather than as manifestations of germline disease.

In 2005, Folpe et al (9) reported on 26 cases of soft tissue and gynecologic PEComas, have proposed histological criteria for risk classification into benign, uncertain malignant potential, and malignant categories. Parameters associated with aggressive behavior include tumor size >5 cm, infiltrative growth, high nuclear grade, necrosis, and mitotic activity >1/50 high-power fields. Surgical excision remains the standard treatment, particularly for tumors with high-risk features.

Yamasaki et al (10) have reported the first description of hepatic PEComa. in 2000. Since then, additional reports, including malignant variants, have been described, with a total of 224 primary hepatic PEComas described across 75 publications (11,12).

Here, we present an unusual case of hepatic PEComa diagnosed in the hepatobiliary surgery center of the University Hospital Münster, Germany. With a particular focus on histopathological evaluation, this case highlights the importance of distinguishing PEComas from morphologically overlapping hepatic neoplasms, particularly hepatocellular carcinoma, in order to ensure accurate diagnosis and appropriate treatment.

A 43-year-old man with no known comorbidities and in good health (height: 188 cm, weight: 75 kg, BMI: 21.2 kg/m²) presented with abdominal discomfort and right upper quadrant tenderness. Initial laboratory tests were unremarkable. The alpha-fetoprotein (AFP) was within the reference range (2.4 ng/ml; reference <7 ng/ml). During admission, fluctuating elevations of GPT and GOT were noted. Dynamic contrast-enhanced CT multi-slice (MSCT) revealed a large tumor in the hepatic left lobe measuring 11.6x10.3x6.8 cm (Fig. 1A-D). The lesion was inhomogeneous, peripherally hypervascular, and showed partial washout in the portal venous phase. Based on imaging, a liver adenoma was suspected, and the interdisciplinary tumor board recommended resection.

The patient underwent robotic-assisted left hemihepatectomy. Postoperative recovery was uneventful, and he was discharged on day 5 with negative resection margins.

Gross pathology revealed a liver specimen weighing 604 g containing a lobulated, heterogeneous, pseudocapsulated tumor measuring 12x8x5.5 cm, located 2 mm from the resection margin (Fig. 2).

Microscopy (Fig. 3A-E) showed epithelioid, polyhedral cells with central to eccentric nuclei and abundant eosinophilic cytoplasm, arranged in nests and separated by a delicate capillary network. Central necrosis (Fig. 3B; arrow) was present, and, at the periphery, the tumor infiltrated its pseudocapsule with vascular invasion (Fig. 3D and E; arrow). Nuclei displayed irregular contours with focal nucleolar prominence, and the mitotic rate was up to 6/10 HPF. Based on these features and the epithelioid morphology, the initial histopathological differential diagnosis was moderately differentiated hepatocellular carcinoma (HCC).

Immunohistochemistry (IHC) demonstrated complete negative for HepPar-1, Arginase-1, Glypican-3, and Glutamine Synthetase. HSP-70 showed weak, non-specific positivity. The Ki-67 proliferation index was <5% and CD34 staining highlighted the peritumoral capillary network.

To refine these unexpected findings, an extended IHC panel was applied. The tumor was negative for β-catenin, Serum Amyloid A (SAA), L-FABP, CD10, CK7, CK20, and pancytokeratin, effectively excluding HCC and hepatocellular adenoma. Further testing revealed no expression of SOX10, S100, PAX8, Chromogranin A, Vimentin, CD117 (c-Kit), Desmin, H-Caldesmon, Myogenin, MyoD1, OCT-4, and SALL4. BRG1 and INI-1 expression were preserved. Notably, tumor cells showed strong, diffuse expression of Melan-A and HMB-45, with focal weak positivity for SMA (Fig. 4A-D), consistent with a perivascular epithelioid cell tumor (PEComa).

Molecular analysis supported this diagnosis. NGS revealed truncating mutations in exons 10 and 27 of the TSC2 gene (Table I). FISH analysis excluded a TFE3 rearrangement at Xp11.23, ruling out TFE3-associated PEComa.

The final pathology report summarized the histological and molecular features, highlighting adverse prognostic indicators: large tumor size, infiltrative growth, necrosis, and mitotic activity (7). The case was reviewed at the multidisciplinary tumor board, and close clinical follow-up was recommended. The patient has been treated surgically to date, and follow-up was recommended. The most recent whole-body imaging, performed five months after the left hemihepatectomy, showed no metastases or residual lesions.

The Ethics Committee of University Hospital Münster (Approval No. 2019-636-f-S) approves this case.

Most hepatic PEComas reported to date are benign. However, histological and clinical features predictive of aggressive behavior have been defined. Yoo et al (12) have identified ‘worrisome features’, including tumor size ≥7 cm, infiltrative borders, mitotic activity >1/10 mm², necrosis, vascular invasion, and classification as PEComa not otherwise specified (NOS). Risk stratification is based on these features: high-risk if ≥3 are present, intermediate-risk if 1-2 are present, and low-risk if none are identified. In the present case, multiple adverse features indicated high-risk disease.

Molecular findings further supported this assessment. The identified truncating mutations in TSC2 are typically loss-of-function variants that drive mTOR pathway activation and are more frequently associated with aggressive clinical behavior, whereas missense variants may have variable consequences depending on their functional domains. Pan et al (7) have demonstrated that TSC2 alterations correlate with poor prognosis, highlighting the biological relevance of mutation type.

Despite their malignant potential, PEComas generally have a low recurrence (3.1%) and metastasis rate (2.7%), as reported by Kvietkauskas et al (13). Complete surgical resection with negative margins remains the standard treatment. In selected cases, neoadjuvant mTOR inhibitors have been shown to reduce tumor size and resection without complications (14).

Histological subtypes may also guide diagnostic interpretation. Bennett et al (15) have emphasized that epithelioid PEComas, such as in our patient, often lack strong smooth muscle marker expression, in contrast to spindle cell-dominant PEComas, which generally stain strongly positive. Recognition of this pattern is essential to avoid misdiagnosis.

In summary, we report a rare case of hepatic PEComa with multiple high-risk features and pathogenic TSC2 mutations. This underscores the importance of a thorough immunohistochemical and molecular workup in epithelioid liver tumors to differentiate PEComa from morphologically similar entities such as hepatocellular carcinoma Awareness of this diagnostic pitfall is crucial for pathologists and clinicians to ensure accurate diagnosis, appropriate patient management, and optimal therapeutic decision-making.