This series reports clinicopathological and molecular genetic findings from nine cases of fibrolamellar hepatocellular carcinoma (FL-HCC) diagnosed at Fudan University Shanghai Cancer Center between January 2018 and December 2024. The objective was to characterize clinical presentation, pathological features, immunophenotype, and underlying molecular alterations using conventional histology, immunohistochemistry (IHC), fluorescence in situ hybridization (FISH), and RNA-based next-generation sequencing (NGS).
Nine patients with a diagnosis of FL-HCC were retrospectively identified from institutional records over the study interval. Clinicopathological and prognostic data were collected. Tumor samples underwent hematoxylin and eosin (HE) staining and immunohistochemical analysis. Molecular testing included RNA-based NGS in a subset and FISH with a PRKACA break-apart probe in others. The authors also reviewed related literature to place findings in context.
The cohort comprised 3 males and 6 females, with a reported central tendency for age of 18.0 years (interquartile range 10.5–26.0). Serum alpha-fetoprotein (AFP) was slightly elevated in one patient. Two patients had active hepatitis B virus infection. Tumor anatomical distribution was recorded as five cases in the left hepatic lobe, three in the right lobe, and one case with multiple lesions involving both lobes. Tumor maximum diameters ranged from 4.0 to 21.4 cm.
Microscopically, tumors were characterized by large polygonal neoplastic cells arranged predominantly in sheets or nests. A defining stromal feature was dense collagenous stroma separating tumor cell aggregates; collagen frequently formed cord-like or parallel lamellar bundles, producing the laminated intratumoral fibrous pattern classically associated with FL-HCC. Tumor cells contained abundant granular, eosinophilic cytoplasm, large vesicular nuclei, and prominent nucleoli. These morphologic descriptors are consistent with the diagnostic features of FL-HCC.
Immunohistochemistry showed uniform expression of CK7, HepPar-1, and Arg-1 across all examined tumor cells in this cohort. These markers were positive in every case, supporting hepatocellular differentiation alongside the distinct histologic pattern.
Molecular analysis targeted the characteristic gene fusion and PRKACA gene status. RNA-based NGS identified the DNAJB1::PRKACA gene fusion in six cases. For the remaining three cases, FISH using a PRKACA break-apart probe demonstrated PRKACA gene rearrangement. Together, these results indicate that all nine tumors in this series harbored either the specific DNAJB1::PRKACA fusion (by NGS) or PRKACA rearrangement (by FISH), confirming the molecular hallmark commonly reported in FL-HCC.
Follow-up duration for the cohort ranged from 7 to 80 months. One patient was lost to follow-up immediately after diagnosis. Of the eight patients with available follow-up data, three showed no evidence of disease at last contact, one experienced disease recurrence with lung metastasis, and four patients died of disease. These outcomes demonstrate variable prognosis within this small series.
This case series reinforces several established clinicopathological and molecular features of FL-HCC. The tumors occurred in a young population (median age ~18 years), showed a predilection for either hepatic lobe, and could reach large sizes (up to 21.4 cm). Histologically, tumors displayed large polygonal eosinophilic cells separated by laminated collagenous stroma, and immunohistochemistry uniformly demonstrated CK7, HepPar-1, and Arg-1 expression. Molecular testing confirmed the presence of the DNAJB1::PRKACA fusion in the majority of cases by RNA-based NGS, with PRKACA rearrangement identified by FISH in the remainder.
The authors conclude that FL-HCC is a distinct, uncommon variant of hepatocellular carcinoma defined by its laminated intratumoral fibrous stroma and a characteristic DNAJB1::PRKACA fusion or other PRKACA rearrangements. Follow-up in this small cohort showed a range of outcomes from no recurrence to metastatic relapse and disease-related death. Details such as specific treatments applied, radiologic imaging features, or comprehensive survival statistics were not reported in the abstract and therefore are not summarized here.
(Conflict of interest: authors declared no conflicts of interest.)