Sarcomatoid hepatocellular carcinoma (sHCC) is a rare HCC subtype with an extremely poor prognosis due to its high risk of recurrence, metastasis, and the absence of effective therapies. Histologically, sHCC is distinguished by the proliferation of spindle and giant cancer cells and often arises as a consequence of repeated non-surgical treatments, such as transarterial chemoembolization or radiofrequency ablation, for conventional HCC (cHCC) [1]. However, sHCC can also develop in untreated cases. Being a rare tumor, there are limited reports analyzing its cancer-related genetic mutations, expression patterns, or tumor immune microenvironment (TME) [2].

In recent years, molecular targeted agents (MTAs) and immune checkpoint inhibitors (ICIs) have been introduced for the treatment of HCC, with evidence suggesting that their efficacy is associated with the genetic alterations and characteristics of the TME [3-5]. However, sHCC is often excluded from clinical trials of pharmacological therapies in HCC, and standard drugs specifically effective for sHCC have not been developed. Given this context, understanding the genetic and expression profiles of sHCC, along with the features of its TME through omics analyses, is essential for the development of MTA- and ICI-based therapeutic strategies [6].

To date, several studies, albeit limited to small cohorts, have analyzed the genetic mutations and TME of sHCC. Zhang et al. [7] reported a high prevalence of mutations in CDKN2A, EPHAS, FANCM, and MAP3K1, with druggable mutations in cell cycle pathway genes observed in 46.6% of sHCC cases. Their findings suggest that CDK4/6 inhibitors may represent promising therapeutic targets (Table 1). Additionally, they noted upregulated expression of genes associated with epithelial-mesenchymal transition (EMT) and inflammatory responses [7]. Similarly, Morisue et al. [8] identified activation of the EMT and inflammatory response pathways in sHCC. Another study demonstrated that high expression of c-Met correlates with poor recurrence-free survival (RFS) after surgery and identified expressions of heat shock protein 70, glutamine synthetase, and glypican-3 as independent prognostic factors [9]. Jia et al. [10] showed high mutation frequencies of TP53 (89.3%), TERT (64.3%), and KRAS (21.4%) in sHCC, with notable genetic abnormalities involving p53 (96%) and DNA damage repair pathways (21%). Some cases also exhibited mutations such as NTRK1 fusions and BRCA1/2 mutations, which are targetable with existing drugs (Table 1) [10]. Yoshuantari et al. [11] highlighted decreased expression of components of the switch/sucrose non-fermenting (SWI/SNF) nucleosome remodeling complex, including BAP1, ARID1A, ARID2, and PBRM1, alongside frequent TP53 and TERT mutations (Table 1).

These findings suggest that developing drugs targeting the cell cycle, DNA damage repair, chromatin remodeling, and inflammatory pathways holds promise for treating sHCC. Furthermore, certain sHCC cases exhibit actionable alterations such as high c-Met expression, NTRK1 fusions, BRCA1/2 mutations, and KRAS mutations, indicating the potential utility of oncogene panel testing for guiding pharmacological therapy in sHCC [9,10].

With respect to the TME, Luo et al. [12] reported frequent expression of PD-L1 and B7-H3 in tumor cells of sHCC. These cases often showed infiltration of CD8⁺, PD-1⁺, and LAG3⁺ T cells (Table 1). On the other hand, some tumors exhibited limited CD8⁺ T cell infiltration but were enriched with Foxp3⁺ or LAG3⁺ T cells, indicating the coexistence of distinct TMEs within sHCC [12]. Morisue et al. [8] also demonstrated sHCC exhibits unique transcriptomic and immunologic features compared to cHCC, with a relatively high prevalence of cases showing elevated PD-L1 expression and significant infiltration of CD4⁺ and CD8⁺ T cells in sHCC. Similarly, Yoshuantari et al. [11] reported frequent PD-L1 expression in tumor cells and tumor-infiltrating immune cells, further suggesting that PD-L1 expression in tumor tissues is a favorable prognostic factor for survival. In contrast, Zhou et al. [9] found no significant correlation between PD-L1 expression and overall survival (OS) or RFS after surgery.

Recently, Sun et al. [13] conducted a comprehensive omics analysis of 10 sHCC tissue samples, revealing a high frequency of ARID2 mutations (70%). They demonstrated a strong association between ARID2 alterations and poorer OS and PFS. Furthermore, ARID2 mutations were shown to promote EMT under hypoxic conditions, accelerating tumor growth and metastasis. Supporting this, Yoshuantari et al. [11] showed reduced expression of SWI/SNF nucleosome remodeling complex components, including ARID2, suggesting a crucial tumor-suppressive role for this complex in sHCC. Previous studies highlighted increased expression of EMT-related genes and activation of inflammatory response pathways, further reinforce the conclusion by Sun et al. that ARID2 loss in cHCC contributes to tumor progression via EMT induction [7,8,13]. Notably, Morisue et al. [8] also identified distinct transcriptomic and immunological characteristics of sHCC, suggesting that ARID2 mutations might contribute to the specific pathophysiology of this subtype of HCC.

Sun et al. [13] emphasize the unique genetic profile of sHCC, particularly the higher prevalence of ARID2 mutations compared to cHCC, and confirm its critical role in EMT-driven pathogenesis. Previous studies have consistently demonstrated that inflammatory response pathways are activated in a considerable proportion of sHCC and that a substantial proportion of tumors express PD-L1 while showing infiltration of CD8⁺, PD-1⁺, and LAG3⁺ T cells in tumor [8,11,12]. These findings suggest that, at least, a subset of sHCCs present immune-active or immune-exhausted TMEs, potentially making them responsive to ICIs targeting the PD-1/PD-L1 axis [14,15]. Although reports of ICI treatment in sHCC are scarce, cases of sHCC responding to atezolizumab plus bevacizumab or achieving complete response with nivolumab in PD-L1-high tumors have been documented [8,16-18].

Interestingly, in melanoma, ARID2 knockout has been shown to enhance infiltration of CD8⁺ T cells into tumors, increase tumor sensitivity to ICIs, and upregulate CXCL9, CXCL10, and CCL5 expression [19]. These findings suggest that the high frequency of ARID2 mutations in sHCC might directly contribute to the induction of an ICI-sensitive TME. Similarly, several studies demonstrated that inactivation of SWI/SNF complex components such as PBRM1, ARID2, and BRD7 is involved in enhancing sensitivity of tumor cells to interferonγ and recruiting effector T cells through chemokine induction [20,21]. Maxwell et al. [22] further showed that mutations in ARID1A, another SWI/SNF component, activate type I interferon pathways via stimulation of R-loops-mediated interferon genes (STING)-dependent signaling, and promote antitumor immunity. Given their shared role in chromatin remodeling, mutations in ARID1A and ARID2 may have overlapping effects on the TME.

As sHCC is a rare subtype of HCC and is often excluded from clinical trials for pharmacological therapies, selecting appropriate treatments remains challenging. Identifying actionable mutations through oncogene panel testing and tailoring individualized therapies based on mutation profiles is crucial. On the other hand, the high frequency of ARID2 mutations and the downregulation of SWI/SNF nucleosome complex components in sHCC are notable features that may be pivotal in EMT formation, a hallmark of sHCC. Furthermore, insights from research in melanoma suggest that ARID2 or SWI/SNF complex dysfunction promotes antitumor immunity via recruitment of cytotoxic T cells and increased sensitivity to ICIs. These findings are compelling for understanding the TME in sHCC. Through omics analyses, the identification of novel druggable mutations and the development of biomarkers for ICI-based therapies hold promise for advancing treatment strategies for sHCC.