Previous Article in Journal
Albumin–Bilirubin Grade and HIV Status in Hepatocellular Carcinoma as Predictors of Survival in Zimbabwe
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Livers
Volume 6
Issue 3
10.3390/livers6030050
livers-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Integrative Profiling of Metabolic CYP Expression, DNA Mutation Rates, and Immune Cell Infiltration for Survival Prognosis in Hepatocellular Carcinoma

by
Mona Dawood
1,2,
Axel Guthart
1,
Ednah Ooko
3,4,
Ralf Weiskirchen
5,
Thomas Efferth
1 and
Joelle C. Boulos
1,*
1
Department of Pharmaceutical Biology, Institute of Pharmaceutical and Biomedical Sciences, Johannes Gutenberg University, Staudinger Weg 5, 55128 Mainz, Germany
2
Department of Molecular Biology, Faculty of Medical Laboratory Science, Al-Neelain University, Khartoum 12702, Sudan
3
Laboratory of Cell Biology, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892, USA
4
Department of Biological Sciences, School of Natural and Applied Sciences, Masinde Muliro University of Science and Technology, Kakamega 50100, Kenya
5
Institute of Molecular Pathobiochemistry, Experimental Gene Therapy and Clinical Chemistry (IFMPEGKC), RWTH University Hospital Aachen, D-52074 Aachen, Germany
*
Author to whom correspondence should be addressed.
Livers 2026, 6(3), 50; https://doi.org/10.3390/livers6030050 (registering DOI)
Submission received: 12 March 2026 / Revised: 14 May 2026 / Accepted: 22 May 2026 / Published: 9 June 2026
Browse Figures
Versions Notes

Abstract

Background/Objectives: Hepatocellular carcinoma (HCC) is challenging to treat with chemotherapy. Immunotherapy has shown moderate responses in inflammatory and immunosuppressive tumor environments. Hepatic cytochrome P450 monooxygenases (CYPs) play a crucial role in xenobiotic and drug metabolism, as well as lipid and steroid metabolism. We aimed to investigate whether CYP expression and various parameters of the innate and adaptive immune system are prognostic factors for the survival of HCC patients. Methods: HCC biopsies (n = 370) from The Cancer Genome Atlas (TCGA) database were analyzed using Kaplan–Meier statistics and the KMPlotter algorithm. Parameters such as immune cell infiltration, DNA mutation rates, and neoantigen load were selected for survival analysis and subjected to hierarchical cluster analysis. The expression of candidate CYP genes in tumors was compared to that in normal liver tissues. Furthermore, tumor infiltration of innate immune cells (basophilic and eosinophilic granulocytes, natural killer cells), adaptive immune cells (CD4+ memory and CD8+ cytotoxic T cells, regulatory T cells, type 1 and type 2 helper T cells), and mesenchymal stem cells was examined. Results: High expression of CYP19A1 and CYP26B1 was associated with shorter survival, whereas high expression of CYP3A5, CYP3A43, CYP7A1, and CYP27A1 was linked to longer survival. Mutation rates combined with CYP expression showed a correlation with five out of six CYP genes, while a high neoantigen load produced less definitive results. A specific cluster exhibiting high CYP expression and immune cell counts or mutation/neoantigen rates was associated with shorter survival, while another cluster was linked to longer survival. Conclusions: CYPs involved in the metabolic regulation of HCC, including CYP3A5, CYP3A43, CYP7A1, CYP19A1, CYP26B1, and CYP27A1, were found to have prognostic value for patient survival. Combined signatures that include CYP expression, mutational rates, and immune cell infiltration into tumors further enhanced the prognostic value for patient survival. This suggests that CYPs may influence the creation of a tumor-specific metabolic microenvironment that impacts immune functions. These combined signatures could be utilized for patient stratification to personalize tumor treatment and develop novel combination therapies aimed at optimizing treatment outcomes, such as combining transarterial chemoembolization (TACE) with immune checkpoint inhibitors.

Graphical Abstract

1. Introduction

Hepatocellular carcinoma (HCC) is the most common primary malignant tumor of the liver, accounting for approximately 85–90% of cases. Worldwide, HCC ranks as the sixth to seventh most common cancer, and unfortunately is the third to fourth most common cause of tumor-related death. There is a significant geographical concentration, especially in East Asia (particularly China), Southeast Asia, and sub-Saharan Africa, which may be related to environmental factors. It is not known why men are more frequently affected than women epidemiologically. As a central organ of metabolism in the human body, the prognosis of tumors deriving from the liver is poor, although patients with low-stage tumors live longer than those with higher stages [1,2]. Unfortunately, the outcome is still not satisfactory for the majority of patients despite considerable progress that has been made in the management of HCC during the past years.
In the vast majority of cases, HCC develops based on existing liver cirrhosis. The most crucial etiological risk factors include chronic hepatitis B virus (HBV) and hepatitis C virus (HCV) infections. Although hepatitis virus infections are among the most frequently found causes, other causes include alcoholic cirrhosis, non-alcoholic fatty liver disease (NAFLD), and non-alcoholic steatohepatitis (NASH), which are frequently underestimated. Furthermore, exposure to aflatoxin B1 from mold also plays an important role in other regions [3]. Additionally, genome-wide association analysis (GWAS) revealed that genetic factors predisposing to the metabolic syndrome also increase the risk of HCC [4,5].
At the molecular level, hepatocarcinogenesis is characterized by aberrant signal transduction pathways due to mutations in key genes, e.g., TP53 and β-catenin mutations and deregulated signaling pathways such as PI3K/AKT/MTOR, RAS/MAPK, IGF, HGF/MET, VEGF, EGFR, and PDGF [6,7]. Immune escape due to dysregulated interleukin and transforming growth factor β (TGF-β) signaling represents another factor for liver cancer development [8]. These factors are considered both as diagnostic biomarkers and targets for targeted therapies and immuno-oncological treatment approaches.
Classic systemic intravenous chemotherapy is hardly effective in HCC due to inherent chemoresistance. Curative treatment attempts are only applied in the early stages of the disease. These include surgical liver resection, local ablation procedures (radiofrequency ablation (RFA) and microwave ablation (MWA)), and—in selected cases—liver transplantation. However, which options are available when curative measures are no longer successful? Possible approaches are the locoregional treatment methods. In transarterial chemoembolization (TACE), chemotherapeutic agents (doxorubicin, epirubicin, or cisplatin) are selectively given via the hepatic artery, followed by vessel embolization as postoperative adjuvant therapy. The advantage of this technique compared to standard chemotherapy is that the ischemic effect reached by vessel embolization enhances the cytotoxic effects of the chemotherapeutic drugs. Another method is transarterial radioembolization or selective internal radiotherapy (TARE/SIRT), where yttrium-90-loaded microspheres are introduced into the tumor area for local radiation applications. Systemic therapy is reserved for advanced or locally untreatable diseases [9,10]. In first-line treatment, the combination of the PD-L1 antibody atezolizumab and the VEGF antibody bevacizumab is considered standard therapy [11]. Alternatively, tyrosine kinase inhibitors such as lenvatinib or sorafenib are used. Additional targeted substances like regorafenib, cabozantinib, and the VEGF receptor 2 antibody ramucirumab are available as second-line treatments [12,13,14]. For the majority of patients, these more individualized treatment strategies result in a noticeable slowing of the progression of their disease.
DNA-damaging chemotherapeutic agents can increase somatic mutations and generate new neoantigens. This should theoretically enhance the immunogenicity of a tumor, although this has not always been convincingly demonstrated in clinical practice. Drug-induced DNA damage with an increase in neoantigens creates an interface between chemotherapy and immunotherapy. Some drugs, including anthracyclines and platinum compounds, are known to induce immunogenic cell death by causing DNA double-strand breaks, endoplasmic reticulum stress, and reactive oxygen species (ROS), leading to the release of tumor-associated antigens and the exposure of danger-associated molecular patterns (DAMPs). These signals can trigger antitumor immune responses [15].
HCC cells express specific immunogenic cell death markers following TACE with doxorubicin [16,17]. However, there is currently no clear clinical evidence that anthracycline-induced immunogenic cell death is primarily responsible for the therapeutic success in HCC. A similar effect is observed with platinum derivatives. Oxaliplatin-induced intra- and interstrand crosslinks, defective DNA repair, and mutagenic replication increase tumor antigen presentation and induce immunogenic cell death [18].
One reason for the modest response to immunotherapy alone is the specific tumor microenvironment of HCC, characterized by chronic inflammation and immunosuppression. This environment leads to T cell exhaustion, preventing immunogenic cell death-associated danger signals from triggering an effective adaptive immune response. In response, combination therapies involving loco-regional procedures and immunotherapy have been developed [19]. The hypothesis is that anthracyclines such as doxorubicin induce immunogenic cell death during TACE, while immune checkpoint inhibitors prevent T-cell exhaustion, enhancing the antitumor immune response.
Since standard chemotherapeutics are largely ineffective, immunotherapy inhibiting programmed death-1 (PD-1), programmed death-ligand 1 (PD-L1), and cytotoxic T lymphocyte antigen-4 (CTLA-4) has become a central component of HCC treatment in recent years [20]. Checkpoint inhibitors release key immune brake mechanisms, reactivating tumor-specific T cells. Although only a subset of patients benefits from immunotherapy, with objective response rates of approximately 20–30%, it has significant prognostic value. It provides long-term disease control for responders and offers substantial survival benefits compared to previous systemic therapy approaches.
CYP enzymes (cytochrome P450 monooxygenases) are a family of 57 isoenzymes that play a crucial role in the hepatic phase 1 metabolism of xenobiotics, including cancer drugs and other endogenous substances. Some CYP enzymes can activate carcinogenic substances, while others can degrade them [21]. Interestingly, the metabolism and pharmacokinetics of anthracyclines (e.g., doxorubicin) and tyrosine kinase inhibitors (e.g., sorafenib) are partially dependent on CYP enzymes. CYP enzymes produce ROS and endoplasmic stress, leading to DNA damage [22,23]. This connection between CYP-induced DNA damage, the mutation rate of liver tumors, and their neoantigen load raises the question of whether CYP-mediated metabolism of anticancer drugs increases neoantigen release and triggers immunogenic cell death. First experimental evidence supports this notion [24], although clinical evidence is currently limited and systematic analyses are not yet available. This question needs to be addressed in the future.
This study aimed to investigate the CYP expression in HCC and its significance for patient survival. To achieve this, 370 tumor biopsies sequenced as part of The Cancer Genome Atlas (TCGA) project were analyzed using Kaplan–Meier statistics. Additionally, we examined whether the expression of 35 CYP genes, along with mutation rate, neoantigen load, and cell counts of 11 immune cell types, could be prognostically relevant parameters for tumor infiltration. The results indicate that specific CYPs and immune cell infiltration patterns were significant for the overall survival of HCC patients. These findings suggest that these profiles could be considered in the development of novel diagnostic and treatment protocols for personalized therapy.

2. Materials and Methods

2.1. Kaplan–Meier Survival Analysis

We calculated the survival probabilities of patients using the KM Plotter algorithm (https:/www.kmplot.com/analysis) (accessed on 21 January 2026) as previously described [25,26]. The human genome encodes 57 functional CYP genes; however, only 35 of them are represented in the KMPlotter database. A total of 370 HCC cases from TCGA were included in the analysis. To avoid type I errors, we used a false discovery rate (FDR) cutoff of 5% [27].

2.2. Cox Regression Analysis

Univariable and multivariable Cox regression analyses were performed to further evaluate the prognostic relevance of the six CYP genes identified in the Kaplan–Meier analysis. Overall survival (OS) was the endpoint, and survival time was defined as days-to-death or to the last follow-up. The multivariable models included age, sex, pathological stage, and histological grade. Age was entered as a continuous variable, while the other covariates were treated as categorical variables. Pathological stage was categorized as stage I/II versus stage III/IV, and histological grade as G1/G2 versus G3/G4. Survival and clinical data for the TCGA-LIHC cohort were obtained from the National Cancer Institute Genomic Data Commons (NCI GDC) Data Portal [28], while the corresponding gene expression data were downloaded from UCSC Xena [29]. After merging the survival, clinical, and gene expression data and excluding cases with incomplete or inconsistent information, 325 HCC cases from TCGA cases were evaluated in the main Cox regression analysis. In an additional sensitivity analysis, cirrhosis/fibrosis status was considered using the Ishak fibrosis score as a surrogate. Therefore, Ishak 0–4 was compared to Ishak 5–6. This analysis included 191 evaluable cases. All statistical analyses were performed using IBM SPSS Statistics for Windows version 29.0.2.0 (IBM Corp., Armonk, NY, USA). Hazard ratios are reported with 95% confidence intervals (CI). Forest plots were created using GraphPad Prism 10.4.2 (GraphPad Software, Boston, MA, USA).

2.3. Kaplan–Meier Analysis of CYP Expression and Mutation Status

To investigate whether the prognostic relevance of CYP expression varies by mutational background, additional Kaplan–Meier analyses were performed for the genes TP53 and CTNNB1. Both were classified as wildtype or mutated based on the presence of non-synonymous mutations. Mutation status was obtained from UCSC Xena using the TCGA unified ensemble mutation calls [29,30]. The previously used TCGA-LIHC cohort (n = 325), was used, and mutation status was added. For each CYP, patients were divided into low- and high-expression groups based on the median. The CYP expression group was then combined with TP53 or CTNNB1 mutation status, resulting in four groups for each CYP/mutation pair. Kaplan–Meier curves were created in SPSS, and groups were compared using the log-rank test.

2.4. Evaluation of CYP Expression and Tumor Immune Cell Infiltration

Spearman correlation coefficient between CYP genes and immune-related markers was carried out using GraphPad Prism (GraphPad Software, San Diego, CA, USA). Normalized gene expression from the TCGA database was used. Then, a heatmap was generated using hierarchical clustering of the CIM miner software (https://discover.nci.nih.gov/cimminer/oneMatrix.do) (accessed on 5 May 2026).
Immune infiltration in TCGA does not use direct immunological assays, such as flow cytometry; instead, they apply multiple computational algorithms for the tumor transcriptome profiles. TIMER platform (http://timer.cistrome.org/) uses an R package called immunedeconv with integration of different algorithms such as TIMER, xCell, MCP-counter, CIBERSORT, EPIC, and quanTIseq to estimate the immune infiltration levels for TCGA [31].
These methods have several limitations: (1) TIMER counts immune cells indirectly using the bioinformatics tools that are based on bulk tumor transcriptome profiles. (2) It also lacks the sensitivity to determine the subsets of the immune cells, like T-regulatory subtypes [31].

2.5. Comparison of CYP mRNA Expression in HCC and Normal Liver

The National Cancer Institute (NCI) in the USA conducted the TCGA project, analyzing over 20,000 cancer samples and their corresponding normal tissues from 33 different cancer types. In this study, we examined the mRNA expression levels of six CYP genes in 371 HCC samples and 50 normal liver samples using Timer2.0 Cistrome (https://timer.cistrome.org) (accessed on 21 January 2026).

3. Results

3.1. Kaplan–Meier Survival Analyses

Of the 57 functional CYP genes in the human genome, the 35 available genes were included in the KMPlotter database (Table 1).
We analyzed the expression of these 35 CYP genes in 370 cases of HCC and correlated them with the overall survival times of the patients. For six CYP genes (CYP3A5, CYP3A43, CYP7A1, CYP19A1, CYP26B1, and CYP27A1), we found using Kaplan–Meier statistics that high gene expression was associated with survival time (p < 0.05; FDR ≤ 5%). Among these genes, high expression of CYP19A1 and CYP26B1 was linked to shorter survival times, while high expression of the other four genes was associated with longer survival times (Figure 1). As an additional conservative sensitivity check, a Bonferroni-corrected threshold was calculated for the 35 tested CYP genes (pBonferroni = 0.05/35 = 0.00143) [32]. A comparison of this threshold with the p-values reported in Figure 1 showed that CYP3A5, CYP7A1, CYP19A1, and CYP27A1 remained significant, whereas CYP26B1 and CYP3A43 did not pass the Bonferroni threshold. Since the Bonferroni correction controls the family-wise error rate and is highly conservative in exploratory multi-gene screening analyses, it may reduce sensitivity for detecting potentially relevant candidate genes [32]. Therefore, Benjamini–Hochberg FDR correction was used as the primary multiple-testing criterion [27].
Since standard therapy for HCC generally results in unsatisfactory response rates, we were intrigued to explore conditions that could help us develop new immunotherapeutic approaches. Therefore, we first investigated the connection between CYP expression and the mutation rates and neo-antigen loads of tumors. These two parameters are factors that promote immunogenic cell death and the success of immunotherapeutic approaches. In fact, we found that a high mutation rate combined with all CYP genes except CYP3A43 correlated with survival times (Table 2). When high CYP expression was combined with neoantigen load, the situation was not as clear: a high neoantigen load together with CYP7A1, CYP26B1, and CYP27A1 led to significant differences in survival time, while a low neoantigen load and high CYP19A1 expression were statistically significantly associated with survival time (Table 2).

3.2. Cox Regression Analyses

To further evaluate the survival associations identified in the Kaplan–Meier analysis, we performed univariable and multivariable Cox regression analyses (n = 325) (Table 3). The univariable model indicates that higher expression of CYP3A5, CYP3A43, CYP7A1, and CYP27A1 was associated with improved OS. In contrast, higher expression of CYP19A1 and CYP26B1 was associated with worse OS. All six CYP genes were statistically significant (p < 0.01) and consistent with the preceding Kaplan–Meier results.
In the multivariable model, age, sex, pathological stage, and histological grade were additionally adjusted (Table 3B). The overall pattern remained partly preserved. Higher expression of CYP3A5 (HR 0.907; p = 0.024) and CYP7A1 (HR 0.945; p = 0.032) remain associated with improved OS. Conversely, higher expression of CYP19A1 (HR 1.087; p = 0.009) and CYP26B1 (HR 1.098; p = 0.031) remained unfavorable for OS. For CYP3A43, the multivariable model showed borderline significance (p = 0.069), whereas CYP27A1 was no longer significant (p = 0.120). Full multivariable Cox regression results are provided in Supplementary Table S1.
In an additional sensitivity analysis restricted to cases with available Ishak fibrosis score data (n = 191), CYP19A1 (HR 1.151; 95% CI 1.038–1.275; p = 0.008) remained significantly associated with worse OS. The complete results are shown in Supplementary Table S2.

3.3. Survival Analysis Using Kaplan–Meier for CYP Expression and Mutational Status

In a further exploration analysis, the correlation between CYP mRNA expression and TP53 and CTNNB1 mutation status was examined. First, the mutation status alone was analyzed via Kaplan–Meier survival curves. Neither TP53 nor CTNNB1 mutation status was significantly associated with OS in the tested data set (p = 0.313 and p = 0.629, respectively).
Next, CYP expression status was combined with the mutation status of each gene. Overall, 12 CYP/mutation status combinations were analyzed. Nine of these combinations showed significant differences in OS, whereas the combinations CYP26B1/TP53, CYP26B1/CTNNB1, and CYP27A1/TP53 were not significant with p > 0.05 (Table 4 and Table 5).
The observed patterns were largely consistent with the survival trends described above for the individual CYPs. For CYP3A5, CYP3A43, CYP7A1, and CYP27A1, a higher expression was, in general, associated with a longer mean of survival within TP53- or CTNNB1-defined subgroups. This was most pronounced for CYP3A43 in combination with TP53 mutations (p = 0.002; Figure 2A and Table 4). In contrast, CYP19A1 and CYP26B1 showed the opposite trend, with higher expression being associated with shorter means of OS, particularly in mutation-positive subgroups. This trend was especially pronounced for CYP19A in combination with CTNNB1 mutation status (p = 0.023; Figure 2A and Table 5). Overall, these findings suggest that the relationship between CYP mRNA expression and OS may be influenced, at least in part, by the mutational context of the tumor.

3.4. CYP Expression and Tumor Immune Cell Infiltration

Then, we investigated the connection between CYP expression and immune cell infiltration. We took statistically significant results (p < 0.05; FDR ≤ 5%) of the six selected CYP genes combined with 11 different immune cell types and plotted them as a matrix. We assigned a value of −1 if the immune cell counts were decreased and a value of 1 if the cell counts were increased. If no correlation was found, we assigned a value of 0. In cases where both enriched and decreased immune cell counts correlated with survival times, no value was assigned, as we assumed no specific effects of the tumor infiltration on patient survival. Similarly, we also included the results of high or low mutation rates and neo-antigen loads.
This matrix was then subjected to hierarchical cluster analysis. Figure 3A shows the correlation of high CYP expression plus immune cell infiltration for shorter survival times, while Figure 3B shows the correlation for longer survival times. In both cluster image maps, two clusters appeared that separated the CYP genes from each other.
In Figure 3A, CYP26B1 and CYP19A1, combined with the infiltration of immune cells from the innate and adaptive immune systems, clustered together in Cluster 1, indicating that these CYP/immune cell combinations were relevant for the patients’ survival. On the other hand, the immune cell infiltration and expression of the other four CYP genes did not significantly affect survival times (Cluster 2). Interestingly, high mutation and neo-antigen load clustered along with the immune cell counts, and both parameters were also found in Cluster 1.
In Figure 3B, a similar analysis was conducted for patients with longer survival times. Cluster 1 consisted of a mixed type with most, but not all, combinations of CYP/immune cell counts being insignificant. Specifically, the genes CYP26B1, CYP19A1, CYP7A1, and CYP3A4 were grouped in this cluster. In Cluster 2, the majority of CYP/immune cell count pairs showed a significant correlation with longer survival, particularly for the genes CYP27A1 and CYP3A5. High mutation rates were clustered along with immune cell counts in Cluster 2, while high neo-antigen loads were found in both clusters.
Additional correlation analysis between the expression of the six CYPs and functional signatures of immune activation or exhaustion was performed. Correlation analysis showed an association between the expression of the selected CYPs and the immune genes. CYP19A1 and CYP26B (associated with a worse prognosis) exhibited positive correlations with immune exhaustion markers, including PDCD1, CTLA4, and HAVCR2, as presented in cluster AD (Figure 4) and negative correlations with activation markers such as GZMB, PRF1, and IFNG in cluster BD (Figure 4). The other CYPs (CYP3A5, CYP7A1, CYP3A43, and CYP27A1) showed the opposite pattern, displaying negative associations with exhaustion markers in cluster AC (Figure 4) and positive correlations with immune activation genes in cluster BC (Figure 4).
Hierarchical clustering further supported the separation of CYPs into two distinct CYP groups associated with immunosuppressive versus immune-activation states. These results revealed that CYP expression patterns are directly linked to the functional status of the tumor immune microenvironment.

3.5. CYP mRNA Expression Levels in HCC Versus Normal Liver Tissue

It was interesting to compare CYP expression in HCC with that in normal liver tissue. CYP3A5, CYP7A1, and CYP19A1 were expressed at higher levels in tumors than in normal liver tissue, although the overall expression of CYP19A1 was lower in both tumor and normal liver samples compared to the other two CYP genes. In contrast, CYP3A43, CYP27A1, and CYP26B1 were expressed at higher levels in normal liver biopsies than in tumors, although CYP26B1 was also lower overall in normal liver and tumor samples (Figure 5).

4. Discussion

In this study, a total of 35 CYP genes were analyzed for their prognostic relevance to the overall survival of HCC patients. The mRNA expression of six genes was significantly associated with survival time. The expression of CYP3A5, CYP3A43, CYP7A1, and CYP27A1 correlated with longer survival, while the expression of CYP19A1 and CYP26B1 was associated with shorter survival. These results suggest that there are two different functional groups of CYP genes with prognostic potential. The first group prolongs survival because these genes contribute to liver-specific physiology and are associated with lower tumor aggressiveness. The liver is rich in immune cells and is the main organ of lipid and bile acid metabolism in the body, which stimulates immune functions [33,34]. Due to the constant interaction between the intestine and liver (gut-liver axis), the liver is exposed to high antigen concentrations [35]. The liver is therefore a unique metabolic-immunological organ in which metabolic signaling pathways can be particularly relevant for tumors in terms of prognosis. CYP7A1 is involved in bile acid metabolism, while the CYP27A1, CYP3A5, and CYP3A43 enzymes are involved in lipid and steroid metabolism [36,37,38,39,40,41,42]. A metabolic influence on tumor immunity, which has a favorable effect on survival time, is therefore evident. The second group contains genes that are known to amplify proliferative signals, such as through estrogen synthesis (CYP19A1) and retinoic acid metabolism (CYP26B1) [43,44]. This can cause liver tumors to become more aggressive and grow faster.
A major limitation of this study is its reliance on The Cancer Genome Atlas cohort, which is exclusively composed of surgically resected, treatment-naive tumor samples and thus depicts a population with earlier-stage disease [45]. In clinical practice, most HCC patients are not candidates for surgery and are instead treated with locoregional therapies such as TACE (for intermediate stage disease), or systemic immunotherapies (for advanced stage disease) [46]. These groups are not represented in TCGA [45] and may have significantly different clinical and biological characteristics than CYP expression profiles, which may be changed by the severity of the disease [47], the immune microenvironment, which is influenced by the disease stage [48], and tumor burden and liver function, which systematically differ between resectable and non-resectable patients [46,49]. This restricts the generalizability of our results. Consequently, validation in independent cohorts, including HCC patients receiving systemic and locoregional therapy, is essential.
Given that CYP expression may be influenced by the aforementioned factors, we conducted univariable and multivariable Cox regression analyses. Notably, Cox regression analyses corroborated CYP3A5, CYP7A1, CYP19A1, and CYP26B1 as independent prognostic factors after age, sex, pathological stage, and histological grade were adjusted. However, CYP3A43 and CYP27A1 lost significance in the multivariable model. The prognostic value of CYP19A1 was also confirmed after fibrosis adjustment, though given the reduced sample size and the use of an Ishak-based surrogate rather than a uniformly documented cirrhosis variable, these findings should be interpreted as supportive sensitivity analyses rather than definitive evidence of fibrosis-independent effects.
Interestingly, we found no correlation between survival times and classic drug-metabolizing CYPs (e.g., CYP1A2, CYP3A4, CYP2C9, CYP2C19, and CYP2D6), which play a central role in the detoxification of xenobiotics and drug metabolism. This suggests that the breakdown of foreign substances by pharmacokinetic CYPs plays a minor role in tumor progression compared to the endogenous metabolic regulation of the tumor microenvironment, in which the metabolic signaling CYPs are involved. Our results suggest that it is not drug metabolism but hepatic lipid and steroid metabolism that are prognostically relevant for HCC.
Despite our results, previous studies revealed that the relationship between CYP expression and patient prognosis is likely multifactorial, reflecting not only their role in xenobiotic metabolism but also their contribution to other signaling pathways. For instance, the main function of some CYPs, such as CYP3A4, is to metabolize many chemotherapy drugs. Zangar et al. (2004) demonstrated that poor coupling of the P450 catalytic cycle leads to the accumulation of ROS, which affects cellular signaling pathways [50]. Another study showed the ability of the CYP1A1/CYP1A2 to induce ROS formation and disrupt the redox state within cancer cells as a result of aminoflavone metabolism. Most probably, aminoflavone is converted into an electrophilic species and a hydroxylated metabolite, which elevate the ROS levels within these cells [51]. Moreover, the cytochromes also metabolize arachidonic acid and convert it to hydroxyeicosatetraenoic acids (HETEs). The 20-HETE is a pro-inflammatory mediator that produces inflammatory cytokines/chemokines in endothelial cells, including IL-8, IL-13, IL-4, and prostaglandin E2 [52].
Another important finding was that high DNA mutation rates in combination with high expression of five of the six CYP genes correlated with survival time.
High mutation rates increase the likelihood of more neoantigens on the surface of tumor cells, and neoantigens are of central importance for immune recognition [53,54]. The coupling of CYP expression with the mutation rate may indicate the immunological activity of HCC. However, the neoantigen load alone is not sufficient as an explanation, since both high neoantigen load with high CYP expression (CYP7A1, CYP26B1, and CYP27A1) and low neoantigen load with high CYP19A1 expression were observed. At the very least, our results suggest complex interactions between tumor metabolism and tumor immunology and that CYP genes regulate this interface.
The exploratory analysis combining CYP expression with TP53 and CTNNB1 mutation status suggests that the association between CYP expression and overall survival may partly depend on the mutational context. Since TP53 and CTNNB1 alone were not significant, but the combined CYP/mutation groups showed additional survival differences, CYP expression might add prognostic information beyond mutation status alone.
For this reason, in the second part of our investigation, we focused on the significance of immune cell infiltration in connection with CYP expression. Using hierarchical cluster analysis, we found that, in the first cluster, high expression of CYP26B1 and CYP19A1, a high mutation rate, a high neoantigen burden, and strong infiltration by cells from the innate and adaptive immune systems were associated with shorter survival times. The fact that patients die earlier despite high immune cell infiltration suggests dysfunctional immune activation, possibly due to infiltrating immune cells being functionally exhausted.
On the other hand, in the second cluster analysis, CYP27A1 and CYP3A5 expression, immune cell infiltration, and a high mutation rate correlated with longer survival in cluster 2. Not only was there high immune cell infiltration but also functioning immune recognition. However, a high neoantigen load was observed in both clusters of the second cluster analysis, indicating relatively unstable prognostic value.
The question arises as to how CYP genes can causally influence immune status. The CYP genes we have identified regulate the metabolism of lipids, steroids, bile acids, and retinoids. These metabolites influence immune cell function, tumor differentiation, and inflammation in the tumor microenvironment. Therefore, a causal relationship between CYP expression and tumor immunity is plausible.
CYP19A1 has aromatase activity and converts androgens into estrogens. The CYP19A1-expression and CYP19A1-mediated aromatase activity are increased in progressive HCC [55,56]. Estrogens modulate T-cell activation, Treg differentiation, macrophage polarization, and control cytokine release [57,58,59,60]. These functions can promote an immunosuppressive tumor microenvironment.
CYP26B1 metabolizes retinoic acid, which regulates T cell differentiation, dendritic cell maturation, and cytotoxic CD8+ T cell functions [61,62,63]. Therefore, low levels of retinoic acid can lead to weak antitumor immune responses.
CYP27A1 links cholesterol to bile acid and oxysterol metabolism. CYP7A1 is a key enzyme in classical bile acid synthesis [64,65]. Bile acids and oxysterols affect immune cell migration, regulate macrophage polarization, affect natural killer cells, and modulate inflammatory FXR and TGR5-coupled signal transduction pathways [66,67,68,69,70,71,72]. Altered bile acid profiles exert immunomodulatory effects on the tumor microenvironment.
CYP3A5 and CYP3A43 have been less well studied as yet but are also involved in steroid and lipid metabolism and thus influence the immune response [40,73]. They may additionally reflect the differentiation status of HCC. Therefore, the association of high CYP expression levels could explain the better survival rates in less differentiated and less aggressive tumors.
There are initial indications that CYP genes function as metabolic immunomodulators, not only for HCC but also for other tumor types. In breast cancer, CYP19A1 has been shown to produce estrogen as aromatase in tumors, which stimulates immunomodulatory effects [74]. In endometrial carcinomas, CYP19A1 was associated with hormonal activity and tumor development [75]. In prostate carcinoma, CYP19A1 was also found to be prognostically relevant [76,77]. Furthermore, androgen levels and androgen deprivation therapy influence the immune system in prostate carcinoma [78,79]. Thus, the concept of CYP gene-driven metabolic immunomodulation seems to be valid not only for HCC but also for other tumor types as a new emerging concept. However, it still needs further establishment.
The comparison of CYP expression in normal liver and HCC revealed interesting changes. Some CYPs were expressed at higher levels in tumors than in normal tissue (CYP3A5, CYP7A1, and CYP19A1), while others showed the opposite pattern (CYP3A43, CYP27A1, and CYP26B1). CYP19A1 and CYP26B1 were expressed at much lower levels overall in both normal liver and tumor samples than the other four CYP genes. The tumor-specific up- and down-regulation of CYP genes reflects how HCC alters its metabolic regulation and adapts to the selection pressure during hepatocarcinogenesis and progression. This ultimately leads to the selection of aggressive cell populations and evasion of the immune response. Nevertheless, it should be noted that the differential expression of the CYP genes does not always reflect the tumor immunity. Therefore, functional studies in vitro and in animal models should be established.
Next, we were intrigued by the potential practical implications of the connection between CYP genes and immune cell infiltration for the clinical management of HCC. Standard therapy is often effective, but there is a need for the development of new diagnostic and therapeutic strategies. Despite the significant chemoresistance of HCC, immunotherapeutic approaches have shown some success [80]. Liver tumors usually arise from inflammatory liver conditions such as hepatitis infections, fibrosis, steatosis, steatohepatitis, or cirrhosis, leading to the activation of many pre-primed T cells by checkpoint inhibitors. Treatment with checkpoint inhibitors prevents T cell exhaustion, maintaining the systemic effect of the immune response against the tumor. This effect is further enhanced by neoangiogenesis inhibitors such as bevacizumab [11,81]. Normalizing tumor angiogenesis promotes T cell infiltration. However, immunotherapy alone cannot completely eradicate the disease, and patient survival rates are still below optimal levels. This raises the question of how therapeutic options can be further optimized.
Signatures from CYP expression could potentially serve as predictive biomarkers for the response to immunotherapies. By combining signatures from CYP expression, mutation rate, and immune status, patients could be stratified to select tumors with an immune-active (“hot”) tumor microenvironment for appropriate immunotherapies.
The combined analyses of CYP expression, mutation rate, and immune status not only suggest the potential benefits of immunotherapy but also indicate that combination therapy with classic anticancer drugs may result in additive or synergistic effects. Doxorubicin and cisplatin, commonly used for TACE, induce DNA strand breaks, leading to increased DNA mutation rates and neoantigen generation. This can enhance the body’s immune response to HCC, ultimately improving patient survival times [82]. This process is known as immunogenic cell death [83,84,85]. Given that metabolic CYP enzymes influence immune cell function, combining TACE with immunotherapies such as immune checkpoint inhibitors (ICIs) could result in improved treatment outcomes. Although the current investigation did not evaluate the relationship between CYP expression and responsiveness to combination therapy (TACE and ICIs), emerging studies support the biological validity of this hypothesis. The effectiveness of chemotherapeutic drugs used in TACE can be affected by CYP enzymes, especially CYP3A4, a crucial enzyme for hepatic drug metabolism, which has been demonstrated to influence chemoresistance to doxorubicin used in TACE, with its activity modulated by tumor microenvironmental elements like tissue stiffness and hypoxia, and thus contributes to interpatient variation in treatment response [86].
Additionally, molecular classification of HCC subtypes, based on zonated metabolic properties, has revealed a Wnt/β-catenin-high subtype distinguished by increased CYP3A4 expression and an immune-desert phenotype marked by low expression of immune checkpoint molecules, emphasizing the interplay of CYP-related metabolic heterogeneity and the immune microenvironment [87]. Taking into consideration that only a fraction of HCC patients respond to ICIs and that established prognostic biomarkers are still lacking in clinical practice [88], integrating CYP expression into patient stratification procedures could help distinguish those who presumably will benefit from the combination of TACE and ICIs. This hypothesis should be further investigated in future research studies and clinical cohorts. Besides TACE, CYP expression profiles may also influence the selection of systemic therapy. As sorafenib and regorafenib are mainly metabolized by CYP3A4/5 and their oxidative metabolism is significantly decreased in HCC tumor tissues, the prognostic significance of CYP3A5 expression reported in this study may guide dose individualization of these targeted cancer agents [89,90].
Furthermore, other drug combinations are conceivable. In tumors with high CYP19A1 expression, aromatase inhibition (e.g., with letrozole) could potentially alter the immunosuppressive tumor microenvironment and enhance the effectiveness of immune checkpoint inhibitors [91]. The regulation of bile acid and oxysterol levels by CYP7A1 and CYP27A1 could also be utilized for immunotherapy, as bile acids can enhance the immunogenicity of tumors [92,93]. CYP26B1 metabolizes retinoic acid, and the degradation of retinoic acid is linked to HCC development [94]. Retinoic acid metabolism blocking agents (RAMBAs) inhibit the breakdown of endogenous retinoic acid by CYP26B1 (and other CYP26 isoforms), leading to increased retinoic acid levels and resulting in anti-proliferative and differentiation-promoting effects in tumors [95]. Additionally, the immunomodulatory properties of retinoic acid have a beneficial impact on the immune response [63,96], including in HCC [97]. However, since CYP19A1 and CYP26B1 were only minimally expressed in our analyses, these combination strategies may be less effective compared to TACE combined with checkpoint inhibitors. Further investigation is needed in the future.

5. Conclusions

In conclusion, we found that six out of 35 CYP genes significantly correlated with the survival times of HCC patients (CYP3A5, CYP3A43, CYP7A1, CYP19A1, CYP26B1, and CYP27A1). These CYPs are involved in metabolic regulation rather than classical drug metabolism. Combined signatures consisting of CYP expression, mutational rates, and immune cell infiltration into tumors enhance the prognostic value for patient survival. This suggests that CYPs may create a tumor-specific metabolic microenvironment that modulates immune functions. These combined signatures could be utilized for patient stratification to personalize tumor treatment and for the development of novel combination treatments to optimize outcomes (e.g., TACE plus immune checkpoint inhibitors).

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/livers6030050/s1, Table S1: Multivariable Cox regression models for overall survival in TCGA-LIHC (n = 325). All models were adjusted for age, sex, pathological stage, and histological grade. Sex was coded as male versus female, pathological stage as stage III/IV versus stage I/II, and histological grade as G3/G4 versus G1/G2.; Table S2: Multivariable Cox regression models for overall survival in TCGA-LIHC, including an Ishak fibrosis/cirrhosis surrogate (n = 191). All models were adjusted for age, sex, pathological stage, and histological grade. Sex was coded as male versus female, pathological stage as stage III/IV versus stage I/II, histological grade as G3/G4 versus G1/G2, and cirrhosis status as Ishak 0–4 versus Ishak 5–6.

Author Contributions

Conceptualization, T.E.; validation, M.D., E.O., R.W., T.E. and J.C.B.; investigation, M.D. and A.G.; resources, T.E., R.W. and J.C.B.; data curation, M.D. and A.G.; writing—original draft preparation, J.C.B.; writing—review and editing, M.D., T.E. and R.W.; visualization, M.D.; supervision, T.E.; project administration, E.O. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding. Ednah Ooko is funded by the NIH Intramural Research Program of the National Institute of Health (NIH), National Cancer Institute (NCI), and Center for Cancer Research, Bethesda, MD, USA.

Institutional Review Board Statement

Ethical review and approval were not required because this study was based solely on data extracted from previously published mRNA expression data deposited in the TCGA and KMPlot databases and did not involve direct human participation or identifiable personal data.

Informed Consent Statement

Informed consent was not required because the study used only anonymized, publicly available data from previously published mRNA expression data deposited in the TCGA and KMPlot databases.

Data Availability Statement

The original contributions presented in this study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
CDCluster of differentiation
CYPCytochrome P450 monooxygenase
FDRFalse discovery rate
HBVChronic hepatitis B virus
HCCHepatocellular carcinoma
HCVHepatitis C virus
KMKaplan–Meier
MWAMicrowave ablation
NAFLDNon-alcoholic fatty liver disease
NASHNon-alcoholic steatohepatitis
PD-L1Programmed cell death1 ligand 1
RFARadiofrequency ablation
ROSReactive oxygen species
SIRTSelective internal radiotherapy
TACETransarterial chemoembolization
TARETransarterial radioembolization
TCGAThe Cancer Genome Atlas
VEGFVascular endothelial growth factor

References

  1. Lafaro, K.J.; Demirjian, A.N.; Pawlik, T.M. Epidemiology of hepatocellular carcinoma. Surg. Oncol. Clin. 2015, 24, 1–17. [Google Scholar] [CrossRef] [PubMed]
  2. Seven, İ.; Bayram, D.; Arslan, H.; Köş, F.T.; Gümüşlü, K.; Aktürk Esen, S.; Şahin, M.; Şendur, M.A.N.; Uncu, D. Predicting hepatocellular carcinoma survival with artificial intelligence. Sci. Rep. 2025, 15, 6226. [Google Scholar] [CrossRef] [PubMed]
  3. Vogel, A.; Meyer, T.; Sapisochin, G.; Salem, R. Saborowski A Hepatocellular carcinoma. Lancet 2022, 400, 1345. [Google Scholar] [CrossRef]
  4. Chen, L.; Wang, Y.; Li, S.; Lu, S.; Luo, Y.; Ning, Y.; Chen, J.; Zhou, S. Metabolic syndrome-related gene signature for prognosis and immune microenvironment prediction in hepatocellular carcinoma. Sci. Rep. 2025, 15, 39048. [Google Scholar] [CrossRef]
  5. Tang, R.; Liu, H.; Yuan, Y.; Xie, K.; Xu, P.; Liu, X.; Wen, J. Genetic factors associated with risk of metabolic syndrome and hepatocellular carcinoma. Oncotarget 2017, 8, 35403–35411. [Google Scholar] [CrossRef]
  6. Cornellà, H.; Alsinet, C.; Villanueva, A. Molecular Pathogenesis of Hepatocellular Carcinoma. Alcohol. Clin. Exp. Res. 2011, 35, 821–825. [Google Scholar] [CrossRef] [PubMed]
  7. Sakamoto, M.; Effendi, K.; Masugi, Y. Molecular diagnosis of multistage hepatocarcinogenesis. Jpn. J. Clin. Oncol. 2010, 40, 891–896. [Google Scholar] [CrossRef]
  8. Wu, M.-Y.; Yiang, G.-T.; Cheng, P.-W.; Chu, P.-Y.; Li, C.-J. Molecular targets in hepatocarcinogenesis and implications for therapy. J. Clin. Med. 2018, 7, 213. [Google Scholar] [CrossRef]
  9. de Alcântara, J.; Götz, G. Transarterial Radioembolization with Yttrium-90 and SIRT Versus Conventional Transarterial Chemoembolization for Hepatocellular Carcinoma: A Systematic Review and Meta-analysis. Acad. Radiol. 2025, 32, 6739–6750. [Google Scholar] [CrossRef]
  10. Zane, K.E.; Makary, M.S. Locoregional Therapies for Hepatocellular Carcinoma with Portal Vein Tumor Thrombosis. Cancers 2021, 13, 5430. [Google Scholar] [CrossRef]
  11. Finn, R.S.; Qin, S.; Ikeda, M.; Galle, P.R.; Ducreux, M.; Kim, T.-Y.; Kudo, M.; Breder, V.; Merle, P.; Kaseb, A.O.; et al. Atezolizumab plus bevacizumab in unresectable hepatocellular carcinoma. N. Engl. J. Med. 2020, 382, 1894–1905. [Google Scholar] [CrossRef] [PubMed]
  12. Dai, S.; Chen, Y.; Cai, W.; Dong, S.; Zhao, J.; Chen, L.; Cheng, C.S. Combination immunotherapy in hepatocellular carcinoma: Synergies among immune checkpoints, TKIs, and chemotherapy. J. Hematol. Oncol. 2025, 18, 85. [Google Scholar] [CrossRef] [PubMed]
  13. Forner, A.; Reig, M.; Bruix, J. Hepatocellular carcinoma. Lancet 2018, 391, 1301–1314. [Google Scholar] [CrossRef]
  14. Ikeda, M.; Morizane, C.; Ueno, M.; Okusaka, T.; Ishii, H.; Furuse, J. Chemotherapy for hepatocellular carcinoma: Current status and future perspectives. Jpn. J. Clin. Oncol. 2018, 48, 103–114. [Google Scholar] [CrossRef]
  15. Garg, A.D.; More, S.; Rufo, N.; Mece, O.; Sassano, M.L.; Agostinis, P.; Zitvogel, L.; Kroemer, G.; Galluzzi, L. Trial watch: Immunogenic cell death induction by anticancer chemotherapeutics. Oncoimmunology 2017, 6, e1386829. [Google Scholar] [CrossRef]
  16. Pinato, D.J.; Murray, S.M.; Forner, A.; Kaneko, T.; Fessas, P.; Toniutto, P.; Mínguez, B.; Cacciato, V.; Avellini, C.; Diaz, A.; et al. Trans-arterial chemoembolization as a loco-regional inducer of immunogenic cell death in hepatocellular carcinoma: Implications for immunotherapy. J. Immunother. Cancer 2021, 9, e003311. [Google Scholar] [CrossRef]
  17. Zhang, Y.; Yang, J.; Xie, S.; Chen, H.; Zhong, J.; Lin, X.; Yu, Z.; Xia, J. A novel prognostic signature based on immunogenic cell death score predicts outcomes and response to transcatheter arterial chemoembolization and immunotherapy in hepatocellular carcinoma. J. Cancer Res. Clin. Oncol. 2023, 149, 11411–11429. [Google Scholar] [CrossRef]
  18. Zhu, H.; Shan, Y.; Ge, K.; Lu, J.; Kong, W.; Jia, C. Oxaliplatin induces immunogenic cell death in hepatocellular carcinoma cells and synergizes with immune checkpoint blockade therapy. Cell. Oncol. 2020, 43, 1203–1214. [Google Scholar] [CrossRef]
  19. Liu, Q.Q.; Wang, X.X.; Ji, H.; Dou, Q.Y.; Zhang, H.M. The efficacy and safety of PD-1 inhibitor combined with TACE in the first-line treatment of unresectable hepatocellular carcinoma. Med. Oncol. 2024, 41, 91. [Google Scholar] [CrossRef] [PubMed]
  20. Hao, L.; Li, S.; Ye, F.; Wang, H.; Zhong, Y.; Zhang, X.; Hu, X.; Huang, X. The current status and future of targeted-immune combination for hepatocellular carcinoma. Front. Immunol. 2024, 15, 1418965. [Google Scholar] [CrossRef]
  21. Abu-Bakar, A.; Tan, B.H.; Halim, H.; Ramli, S.; Pan, Y.; Ong, C.E. Cytochromes P450: Role in Carcinogenesis and Relevance to Cancers. Curr. Drug Metab. 2022, 23, 355–373. [Google Scholar] [CrossRef] [PubMed]
  22. Bagdasaryan, A.A.; Chubarev, V.N.; Smolyarchuk, E.A.; Drozdov, V.N.; Krasnyuk, I.I.; Liu, J.; Fan, R.; Tse, E.; Shikh, E.V.; Sukocheva, O.A. Pharmacogenetics of Drug Metabolism: The Role of Gene Polymorphism in the Regulation of Doxorubicin Safety and Efficacy. Cancers 2022, 14, 5436. [Google Scholar] [CrossRef]
  23. McGill, M.R.; Kaufman, Y.J.; LoBianco, F.V.; Schleiff, M.A.; Aykin-Burns, N.; Miller, G.P. The role of cytochrome P450 3A4-mediated metabolism in sorafenib and lapatinib hepatotoxicity. Livers 2023, 3, 310–321. [Google Scholar] [CrossRef]
  24. Lévesque, J.F.; Day, S.H.; Jones, A.N. Protocols of in vitro protein covalent binding studies in liver. Methods Mol. Biol. 2011, 691, 283–301. [Google Scholar] [CrossRef]
  25. Győrffy, B. Integrated analysis of public datasets for the discovery and validation of survival-associated genes in solid tumors. Innovation 2024, 5, 100625. [Google Scholar] [CrossRef] [PubMed]
  26. Ooko, E.; Efferth, T. Identification of Prognostic Gene Signatures for Survival of Patients With Phaeochromocytoma, Paraganglioma, and Other Tumor Types. Cancer Diagn. Progn. 2025, 5, 808–832. [Google Scholar] [CrossRef]
  27. Benjamini, Y.; Hochberg, Y. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. J. R. Stat. Soc. Ser. B (Methodol.) 1995, 57, 289–300. [Google Scholar] [CrossRef]
  28. Heath, A.P.; Ferretti, V.; Agrawal, S.; An, M.; Angelakos, J.C.; Arya, R.; Bajari, R.; Baqar, B.; Barnowski, J.H.B.; Burt, J.; et al. The NCI Genomic Data Commons. Nat. Genet. 2021, 53, 257–262. [Google Scholar] [CrossRef]
  29. Goldman, M.J.; Craft, B.; Hastie, M.; Repečka, K.; McDade, F.; Kamath, A.; Banerjee, A.; Luo, Y.; Rogers, D.; Brooks, A.N.; et al. Visualizing and interpreting cancer genomics data via the Xena platform. Nat. Biotechnol. 2020, 38, 675–678. [Google Scholar] [CrossRef]
  30. Ellrott, K.; Bailey, M.H.; Saksena, G.; Covington, K.R.; Kandoth, C.; Stewart, C.; Hess, J.; Ma, S.; Chiotti, K.E.; McLellan, M.; et al. Scalable Open Science Approach for Mutation Calling of Tumor Exomes Using Multiple Genomic Pipelines. Cell Syst. 2018, 6, 271–281.E7. [Google Scholar] [CrossRef] [PubMed]
  31. Li, T.; Fu, J.; Zeng, Z.; Cohen, D.; Li, J.; Chen, Q.; Li, B.; Liu, X.S. TIMER2.0 for analysis of tumor-infiltrating immune cells. Nucleic Acids Res. 2020, 48, W509–W514. [Google Scholar] [CrossRef] [PubMed]
  32. Noble, W.S. How does multiple testing correction work? Nat. Biotechnol. 2009, 27, 1135–1137. [Google Scholar] [CrossRef] [PubMed]
  33. Fiorucci, S.; Marchianò, S.; Urbani, G.; Di Giorgio, C.; Distrutti, E.; Zampella, A.; Biagioli, M. Immunology of bile acids regulated receptors. Prog. Lipid Res. 2024, 95, 101291. [Google Scholar] [CrossRef]
  34. Jia, W.; Li, Y.; Cheung, K.C.P.; Zheng, X. Bile acid signaling in the regulation of whole body metabolic and immunological homeostasis. Sci. China Life Sci. 2024, 67, 865–878. [Google Scholar] [CrossRef]
  35. Tilg, H.; Adolph, T.E.; Trauner, M. Gut-liver axis: Pathophysiological concepts and clinical implications. Cell Metab. 2022, 34, 1700–1718. [Google Scholar] [CrossRef]
  36. Cali, J.J.; Russell, D.W. Characterization of human sterol 27-hydroxylase. A mitochondrial cytochrome P-450 that catalyzes multiple oxidation reaction in bile acid biosynthesis. J. Biol. Chem. 1991, 266, 7774–7778. [Google Scholar] [CrossRef]
  37. Chambers, K.F.; Day, P.E.; Aboufarrag, H.T.; Kroon, P.A. Polyphenol Effects on Cholesterol Metabolism via Bile Acid Biosynthesis, CYP7A1: A Review. Nutrients 2019, 11, 2588. [Google Scholar] [CrossRef]
  38. Chiang, J.Y.L.; Ferrell, J.M. Up to date on cholesterol 7 alpha-hydroxylase (CYP7A1) in bile acid synthesis. Liver Res. 2020, 4, 47–63. [Google Scholar] [CrossRef] [PubMed]
  39. Escher, G.; Krozowski, Z.; Croft, K.D.; Sviridov, D. Expression of sterol 27-hydroxylase (CYP27A1) enhances cholesterol efflux. J. Biol. Chem. 2003, 278, 11015–11019. [Google Scholar] [CrossRef]
  40. Klyushova, L.S.; Perepechaeva, M.L.; Grishanova, A.Y. The Role of CYP3A in Health and Disease. Biomedicines 2022, 10, 2686. [Google Scholar] [CrossRef]
  41. Niwa, T.; Murayama, N.; Imagawa, Y.; Yamazaki, H. Regioselective hydroxylation of steroid hormones by human cytochromes P450. Drug Metab. Rev. 2015, 47, 89–110. [Google Scholar] [CrossRef]
  42. Pikuleva, I.A. Cholesterol-metabolizing cytochromes P450. Drug Metab. Dispos. Biol. Fate Chem. 2006, 34, 513–520. [Google Scholar] [CrossRef]
  43. Lee, Y.B.; Kim, C.; Hong, J.; Kim, D. Molecular insights into CYP19A1 mutations and their role in estrogen production. Arch. Biochem. Biophys. 2025, 772, 110573. [Google Scholar] [CrossRef]
  44. Topletz, A.R.; Thatcher, J.E.; Zelter, A.; Lutz, J.D.; Tay, S.; Nelson, W.L.; Isoherranen, N. Comparison of the function and expression of CYP26A1 and CYP26B1, the two retinoic acid hydroxylases. Biochem. Pharmacol. 2012, 83, 149–163. [Google Scholar] [CrossRef]
  45. Ally, A.; Balasundaram, M.; Carlsen, R.; Chuah, E.; Clarke, A.; Dhalla, N.; Holt, R.A.; Jones, S.J.M.; Lee, D.; Ma, Y.; et al. Comprehensive and Integrative Genomic Characterization of Hepatocellular Carcinoma. Cell 2017, 169, 1327–1341.E23. [Google Scholar] [CrossRef]
  46. Marina, R.; Forner, A.; Rimola, J. BCLC strategy for prognosis prediction and treatment recommendation: The 2022 update. J. Hepatol. 2021, 19, S0168–S8278. [Google Scholar]
  47. Brown, Z.J.; Ruff, S.M.; Pawlik, T.M. The effect of liver disease on hepatic microenvironment and implications for immune therapy. Front. Pharmacol. 2023, 14, 1225821. [Google Scholar] [CrossRef]
  48. Llovet, J.M.; Castet, F.; Heikenwalder, M.; Maini, M.K.; Mazzaferro, V.; Pinato, D.J.; Pikarsky, E.; Zhu, A.X.; Finn, R.S. Immunotherapies for hepatocellular carcinoma. Nat. Rev. Clin. Oncol. 2022, 19, 151–172. [Google Scholar] [CrossRef] [PubMed]
  49. Kinsey, E.; Lee, H.M. Management of hepatocellular carcinoma in 2024: The multidisciplinary paradigm in an evolving treatment landscape. Cancers 2024, 16, 666. [Google Scholar] [CrossRef] [PubMed]
  50. Zangar, R.C.; Davydov, D.R.; Verma, S. Mechanisms that regulate production of reactive oxygen species by cytochrome P450. Toxicol. Appl. Pharmacol. 2004, 199, 316–331. [Google Scholar] [CrossRef] [PubMed]
  51. McLean, L.; Soto, U.; Agama, K.; Francis, J.; Jimenez, R.; Pommier, Y.; Sowers, L.; Brantley, E. Aminoflavone induces oxidative DNA damage and reactive oxidative species-mediated apoptosis in breast cancer cells. Int. J. Cancer 2008, 122, 1665–1674. [Google Scholar] [CrossRef]
  52. Panigrahy, D.; Kaipainen, A.; Greene, E.R.; Huang, S. Cytochrome P450-derived eicosanoids: The neglected pathway in cancer. Cancer Metastasis Rev. 2010, 29, 723–735. [Google Scholar] [CrossRef] [PubMed]
  53. Li, Z.; Chen, G.; Cai, Z.; Dong, X.; He, L.; Qiu, L.; Zeng, Y.; Liu, X.; Liu, J. Profiling of hepatocellular carcinoma neoantigens reveals immune microenvironment and clonal evolution related patterns. Chin. J. Cancer Res. 2021, 33, 364–378. [Google Scholar] [CrossRef] [PubMed]
  54. Vitale, I.; Sistigu, A.; Manic, G.; Rudqvist, N.P.; Trajanoski, Z.; Galluzzi, L. Mutational and Antigenic Landscape in Tumor Progression and Cancer Immunotherapy. Trends Cell Biol. 2019, 29, 396–416. [Google Scholar] [CrossRef] [PubMed]
  55. Brodeur, C.M.; Thibault, P.; Durand, M.; Perreault, J.P.; Bisaillon, M. Dissecting the expression landscape of cytochromes P450 in hepatocellular carcinoma: Towards novel molecular biomarkers. Genes Cancer 2019, 10, 97–108. [Google Scholar] [CrossRef]
  56. Koh, W.P.; Yuan, J.M.; Wang, R.; Govindarajan, S.; Oppenheimer, R.; Zhang, Z.Q.; Yu, M.C.; Ingles, S.A. Aromatase (CYP19) promoter gene polymorphism and risk of nonviral hepatitis-related hepatocellular carcinoma. Cancer 2011, 117, 3383–3392. [Google Scholar] [CrossRef]
  57. Belboul, A.; Ashworth, J.; Fadel, A.; McLoughlin, J.; Mahmoud, A.; El Mohtadi, M. Estrogen induces the alternative activation of macrophages through binding to estrogen receptor-alpha. Exp. Mol. Pathol. 2025, 143, 104971. [Google Scholar] [CrossRef]
  58. Bengtsson, A.K.; Ryan, E.J.; Giordano, D.; Magaletti, D.M.; Clark, E.A. 17beta-estradiol (E2) modulates cytokine and chemokine expression in human monocyte-derived dendritic cells. Blood 2004, 104, 1404–1410. [Google Scholar] [CrossRef]
  59. Nadkarni, S.; McArthur, S. Oestrogen and immunomodulation: New mechanisms that impact on peripheral and central immunity. Curr. Opin. Pharmacol. 2013, 13, 576–581. [Google Scholar] [CrossRef]
  60. Polanczyk, M.J.; Hopke, C.; Vandenbark, A.A.; Offner, H. Estrogen-mediated immunomodulation involves reduced activation of effector T cells, potentiation of Treg cells, and enhanced expression of the PD-1 costimulatory pathway. J. Neurosci. Res. 2006, 84, 370–378. [Google Scholar] [CrossRef]
  61. Guo, Y.; Lee, Y.C.; Brown, C.; Zhang, W.; Usherwood, E.; Noelle, R.J. Dissecting the role of retinoic acid receptor isoforms in the CD8 response to infection. J. Immunol. 2014, 192, 3336–3344. [Google Scholar] [CrossRef]
  62. Manicassamy, S.; Pulendran, B. Retinoic acid-dependent regulation of immune responses by dendritic cells and macrophages. Semin. Immunol. 2009, 21, 22–27. [Google Scholar] [CrossRef]
  63. Raverdeau, M.; Mills, K.H. Modulation of T cell and innate immune responses by retinoic Acid. J. Immunol. 2014, 192, 2953–2958. [Google Scholar] [CrossRef] [PubMed]
  64. Farazuddin, M.; Ludka, N.; Friesen, L.; Landers, J.J.; O’Konek, J.J.; Kim, C.H.; Baker, J.R. Retinoic Acid Signaling Is Required for Dendritic Cell Maturation and the Induction of T Cell Immunity. ImmunoHorizons 2023, 7, 480–492. [Google Scholar] [CrossRef] [PubMed]
  65. Gilardi, F.; Mitro, N.; Godio, C.; Scotti, E.; Caruso, D.; Crestani, M.; De Fabiani, E. The pharmacological exploitation of cholesterol 7alpha-hydroxylase, the key enzyme in bile acid synthesis: From binding resins to chromatin remodelling to reduce plasma cholesterol. Pharmacol. Ther. 2007, 116, 449–472. [Google Scholar] [CrossRef]
  66. Chen, X.; Yang, D.; Shen, W.; Dong, H.F.; Wang, J.M.; Oppenheim, J.J.; Howard, M.Z. Characterization of chenodeoxycholic acid as an endogenous antagonist of the G-coupled formyl peptide receptors. Inflamm. Res. 2000, 49, 744–755. [Google Scholar] [CrossRef]
  67. Chiang, J.Y.L.; Ferrell, J.M. Bile acid receptors FXR and TGR5 signaling in fatty liver diseases and therapy. Am. J. Physiol. Gastrointest. Liver Physiol. 2020, 318, G554–G573. [Google Scholar] [CrossRef] [PubMed]
  68. Godlewska, U.; Bulanda, E.; Wypych, T.P. Bile acids in immunity: Bidirectional mediators between the host and the microbiota. Front. Immunol. 2022, 13, 949033. [Google Scholar] [CrossRef]
  69. Lin, B.S.; Yin, M.Y.; Xie, S.A.; Li, P.; Li, X. Pleiotropic immunoregulation by bile acids in pathophysiology. Front. Immunol. 2026, 17, 1719092. [Google Scholar] [CrossRef]
  70. Ma, C.; Han, M.; Heinrich, B.; Fu, Q.; Zhang, Q.; Sandhu, M.; Agdashian, D.; Terabe, M.; Berzofsky, J.A.; Fako, V.; et al. Gut microbiome-mediated bile acid metabolism regulates liver cancer via NKT cells. Science 2018, 360, eaan5931. [Google Scholar] [CrossRef]
  71. Wang, L.; Gong, Z.; Zhang, X.; Zhu, F.; Liu, Y.; Jin, C.; Du, X.; Xu, C.; Chen, Y.; Cai, W.; et al. Gut microbial bile acid metabolite skews macrophage polarization and contributes to high-fat diet-induced colonic inflammation. Gut Microbes 2020, 12, 1819155. [Google Scholar] [CrossRef] [PubMed]
  72. Wei, H.; Suo, C.; Gu, X.; Shen, S.; Lin, K.; Zhu, C.; Yan, K.; Bian, Z.; Chen, L.; Zhang, T.; et al. AKR1D1 suppresses liver cancer progression by promoting bile acid metabolism-mediated NK cell cytotoxicity. Cell Metab. 2025, 37, 1103–1118.E7. [Google Scholar] [CrossRef]
  73. Kivistö, K.T.; Niemi, M.; Schaeffeler, E.; Pitkälä, K.; Tilvis, R.; Fromm, M.F.; Schwab, M.; Eichelbaum, M.; Strandberg, T. Lipid-lowering response to statins is affected by CYP3A5 polymorphism. Pharmacogenetics 2004, 14, 523–525. [Google Scholar] [CrossRef]
  74. Huang, H.; Zhou, J.; Chen, H.; Li, J.; Zhang, C.; Jiang, X.; Ni, C. The immunomodulatory effects of endocrine therapy in breast cancer. J. Exp. Clin. Cancer Res. 2021, 40, 19. [Google Scholar] [CrossRef] [PubMed]
  75. Thompson, D.J.; O’Mara, T.A.; Glubb, D.M.; Painter, J.N.; Cheng, T.; Folkerd, E.; Doody, D.; Dennis, J.; Webb, P.M.; Gorman, M.; et al. CYP19A1 fine-mapping and Mendelian randomization: Estradiol is causal for endometrial cancer. Endocr.-Relat. Cancer 2016, 23, 77–91. [Google Scholar] [CrossRef]
  76. Grindstad, T.; Skjefstad, K.; Andersen, S.; Ness, N.; Nordby, Y.; Al-Saad, S.; Fismen, S.; Donnem, T.; Khanehkenari, M.R.; Busund, L.T.; et al. Estrogen receptors α and β and aromatase as independent predictors for prostate cancer outcome. Sci. Rep. 2016, 6, 33114. [Google Scholar] [CrossRef] [PubMed]
  77. Kanda, S.; Tsuchiya, N.; Narita, S.; Inoue, T.; Huang, M.; Chiba, S.; Akihama, S.; Saito, M.; Numakura, K.; Tsuruta, H.; et al. Effects of functional genetic polymorphisms in the CYP19A1 gene on prostate cancer risk and survival. Int. J. Cancer 2015, 136, 74–82. [Google Scholar] [CrossRef]
  78. Aragon-Ching, J.B.; Williams, K.M.; Gulley, J.L. Impact of androgen-deprivation therapy on the immune system: Implications for combination therapy of prostate cancer. Front. Biosci. 2007, 12, 4957–4971. [Google Scholar] [CrossRef]
  79. Sommer, U.; Ebersbach, C.; Beier, A.K.; Baretton, G.B.; Thomas, C.; Borkowetz, A.; Erb, H.H.H. Influence of Androgen Deprivation Therapy on the PD-L1 Expression and Immune Activity in Prostate Cancer Tissue. Front. Mol. Biosci. 2022, 9, 878353. [Google Scholar] [CrossRef]
  80. Dyhl-Polk, A.; Mikkelsen, M.K.; Ladekarl, M.; Nielsen, D.L. Clinical Trials of Immune Checkpoint Inhibitors in Hepatocellular Carcinoma. J. Clin. Med. 2021, 10, 2662. [Google Scholar] [CrossRef]
  81. Cheng, A.L.; Qin, S.; Ikeda, M.; Galle, P.R.; Ducreux, M.; Kim, T.Y.; Lim, H.Y.; Kudo, M.; Breder, V.; Merle, P.; et al. Updated efficacy and safety data from IMbrave150: Atezolizumab plus bevacizumab vs. sorafenib for unresectable hepatocellular carcinoma. J. Hepatol. 2022, 76, 862–873. [Google Scholar] [CrossRef]
  82. Guo, Y.; Pan, Z.; Kan, X.; Li, T.; Gong, B.; Li, Y.; Yang, L.; Zheng, C. Immunotherapy improved the efficacy of TACE or TACE plus MTTs in HCC patients: A meta-analysis. Int. Immunopharmacol. 2025, 147, 114006. [Google Scholar] [CrossRef] [PubMed]
  83. Barjij, I.; Meliani, M. Immunogenic Cell Death as a Target for Combination Therapies in Solid Tumors: A Systematic Review Toward a New Paradigm in Immuno-Oncology. Cureus 2025, 17, e85776. [Google Scholar] [CrossRef]
  84. Dou, L.; Fang, Y.; Yang, H.; Ai, G.; Shen, N. Immunogenic cell death: A new strategy to enhancing cancer immunotherapy. Hum. Vaccines Immunother. 2024, 20, 2437918. [Google Scholar] [CrossRef] [PubMed]
  85. Li, C.; Qi, X.; Yan, M. Chemotherapy-induced immunogenic cell death in combination with ICIs: A brief review of mechanisms, clinical insights, and therapeutic implications. Front. Pharmacol. 2025, 16, 1572195. [Google Scholar] [CrossRef]
  86. Özkan, A.; Stolley, D.L.; Cressman, E.N.K.; McMillin, M.; DeMorrow, S.; Yankeelov, T.E.; Rylander, M.N. Tumor Microenvironment Alters Chemoresistance of Hepatocellular Carcinoma Through CYP3A4 Metabolic Activity. Front. Oncol. 2021, 11, 662135. [Google Scholar] [CrossRef]
  87. Aoki, T.; Nishida, N.; Kurebayashi, Y.; Sakai, K.; Fujiwara, N.; Tsurusaki, M.; Hanaoka, K.; Morita, M.; Chishina, H.; Takita, M.; et al. Molecular classification of hepatocellular carcinoma based on zoned metabolic feature and oncogenic signaling pathway. Clin. Mol. Hepatol. 2025, 31, 981–1002. [Google Scholar] [CrossRef]
  88. Qin, R.; Jin, T.; Xu, F. Biomarkers predicting the efficacy of immune checkpoint inhibitors in hepatocellular carcinoma. Front. Immunol. 2023, 14, 1326097. [Google Scholar] [CrossRef] [PubMed]
  89. Ye, L.; Yang, X.; Guo, E.; Chen, W.; Lu, L.; Wang, Y.; Peng, X.; Yan, T.; Zhou, F.; Liu, Z. Sorafenib metabolism is significantly altered in the liver tumor tissue of hepatocellular carcinoma patient. PLoS ONE 2014, 9, e96664. [Google Scholar] [CrossRef]
  90. Zhang, Y.; Wang, Z.; Wang, Y.; Jin, W.; Zhang, Z.; Jin, L.; Qian, J.; Zheng, L. CYP3A4 and CYP3A5: The crucial roles in clinical drug metabolism and the significant implications of genetic polymorphisms. PeerJ 2024, 12, e18636. [Google Scholar] [CrossRef]
  91. Liu, L.; Mo, M.; Chen, X.; Chao, D.; Zhang, Y.; Chen, X.; Wang, Y.; Zhang, N.; He, N.; Yuan, X.; et al. Targeting inhibition of prognosis-related lipid metabolism genes including CYP19A1 enhances immunotherapeutic response in colon cancer. J. Exp. Clin. Cancer Res. 2023, 42, 85. [Google Scholar] [CrossRef]
  92. Tong, Y.; Lou, X. Interplay between bile acids, gut microbiota, and the tumor immune microenvironment: Mechanistic insights and therapeutic strategies. Front. Immunol. 2025, 16, 1638352. [Google Scholar] [CrossRef]
  93. Varanasi, S.K.; Chen, D.; Liu, Y.; Johnson, M.A.; Miller, C.M.; Ganguly, S.; Lande, K.; LaPorta, M.A.; Hoffmann, F.A.; Mann, T.H.; et al. Bile acid synthesis impedes tumor-specific T cell responses during liver cancer. Science 2025, 387, 192–201. [Google Scholar] [CrossRef]
  94. Chen, F.; Cai, Y.; Zhou, J. Relationship between retinol metabolism and hepatocellular carcinoma: A comprehensive analysis of Mendelian randomization, prognostic characteristic and experiment. Discov. Oncol. 2025, 16, 513. [Google Scholar] [CrossRef]
  95. Nelson, C.H.; Buttrick, B.R.; Isoherranen, N. Therapeutic potential of the inhibition of the retinoic acid hydroxylases CYP26A1 and CYP26B1 by xenobiotics. Curr. Top. Med. Chem. 2013, 13, 1402–1428. [Google Scholar] [CrossRef]
  96. Bi, G.; Liang, J.; Bian, Y.; Shan, G.; Besskaya, V.; Wang, Q.; Zhan, C. The immunomodulatory role of all-trans retinoic acid in tumor microenvironment. Clin. Exp. Med. 2023, 23, 591–606. [Google Scholar] [CrossRef]
  97. Li, X.; Luo, X.; Chen, S.; Chen, J.; Deng, X.; Zhong, J.; Wu, H.; Huang, X.; Wang, C. All-trans-retinoic acid inhibits hepatocellular carcinoma progression by targeting myeloid-derived suppressor cells and inhibiting angiogenesis. Int. Immunopharmacol. 2023, 121, 110413. [Google Scholar] [CrossRef]
Livers 06 00050 g001
Figure 1. Prognostic relevance of CYP mRNA expression for overall survival of 370 HCC patients as determined by Kaplan–Meier statistics. (A) CYP3A5; (B) CYP3A43; (C) CYP7A1; (D) CYP19A1; (E) CYP26B1; (F) CYP27A1.
Figure 1. Prognostic relevance of CYP mRNA expression for overall survival of 370 HCC patients as determined by Kaplan–Meier statistics. (A) CYP3A5; (B) CYP3A43; (C) CYP7A1; (D) CYP19A1; (E) CYP26B1; (F) CYP27A1.
Livers 06 00050 g001
Livers 06 00050 g002
Figure 2. Kaplan–Meier analysis in HCC patients grouped by CYP expression and mutation status. Patients were divided into four groups based on low or high CYP mRNA expression and wild-type (WT) or mutated TP53/CTNNB1 status (mut). (A) CYP3A43 expression in combination with TP53 mutation status. (B) CYP19A1 expression in combination with CTNNB1 mutation status.
Figure 2. Kaplan–Meier analysis in HCC patients grouped by CYP expression and mutation status. Patients were divided into four groups based on low or high CYP mRNA expression and wild-type (WT) or mutated TP53/CTNNB1 status (mut). (A) CYP3A43 expression in combination with TP53 mutation status. (B) CYP19A1 expression in combination with CTNNB1 mutation status.
Livers 06 00050 g002
Livers 06 00050 g003
Figure 3. Hierarchical cluster analysis of Kaplan–Meier overall survival analyses of CYP mRNA expression combined with DNA mutation rates, neoantigen loads, and 11 immune cell types of the innate and adaptive immune system (p < 0.05; FDR ≤ 5%). (A) shows the cluster analysis related to short survival and (B) with long survival.
Figure 3. Hierarchical cluster analysis of Kaplan–Meier overall survival analyses of CYP mRNA expression combined with DNA mutation rates, neoantigen loads, and 11 immune cell types of the innate and adaptive immune system (p < 0.05; FDR ≤ 5%). (A) shows the cluster analysis related to short survival and (B) with long survival.
Livers 06 00050 g003
Livers 06 00050 g004
Figure 4. Correlation between CYP expression and functional signatures of immune activation or exhaustion. Spearman correlation coefficient was carried out using GraphPad Prism between the CYP genes (rows) and the immune activation/exhaustion genes (columns) for TGCA tumor samples. The red color represents the positive correlation, while the blue color indicates the negative correlation. Cluster A corresponds to inhibitory/exhaustion-associated immune markers (HAVCR2, CTLA4, PDCD1), cluster B presents cytotoxic/effector-associated immune markers (PRF1, GZMB, LAG3, GZMA, IFNG), cluster C and D represent the CYP genes.
Figure 4. Correlation between CYP expression and functional signatures of immune activation or exhaustion. Spearman correlation coefficient was carried out using GraphPad Prism between the CYP genes (rows) and the immune activation/exhaustion genes (columns) for TGCA tumor samples. The red color represents the positive correlation, while the blue color indicates the negative correlation. Cluster A corresponds to inhibitory/exhaustion-associated immune markers (HAVCR2, CTLA4, PDCD1), cluster B presents cytotoxic/effector-associated immune markers (PRF1, GZMB, LAG3, GZMA, IFNG), cluster C and D represent the CYP genes.
Livers 06 00050 g004
Livers 06 00050 g005
Figure 5. CYP mRNA expression in HCC compared to normal liver tissue. Red indicates HCC samples; blue indicates normal liver tissue.
Figure 5. CYP mRNA expression in HCC compared to normal liver tissue. Red indicates HCC samples; blue indicates normal liver tissue.
Livers 06 00050 g005
Table 1. CYP gene family members were analyzed in the present investigation for their prognostic value in renal cell carcinoma 1.
Table 1. CYP gene family members were analyzed in the present investigation for their prognostic value in renal cell carcinoma 1.
SymbolNameAdditional NameFunctionLocalization
CYP11A1Cytochrome P450, family 11, subfamily A, polypeptide 1Cholesterol side-chain cleavage enzyme, mitochondrialCatalyzes the conversion of cholesterol to pregnenolone in the synthesis of steroid hormones.Mit
CYP11B1Cytochrome P450, family 11, subfamily B, polypeptide 1Steroid 11-β-hydroxylaseConverts progesterone to cortisol in the adrenal cortex.Mit
CYP11B2Cytochrome P450, family 11, subfamily B, polypeptide 2Aldosterone-synthesizing enzymeSteroid 18-hydroxylase activity to synthesize aldosterone and 18-oxocortisol; steroid 11β-hydroxylase activity.Mit
CYP17A1Cytochrome P450, family 17, subfamily A, polypeptide 117-α-Hydroxyprogesterone aldolaseCorticoid and androgen biosynthesis.ER
CYP19A1Cytochrome P450, family 19, subfamily A, polypeptide 1Flavoprotein-linked monooxygenaseEstrogen biosynthesis.ER
CYP1A1Cytochrome P450, family 1, subfamily A, polypeptide 1Aryl hydrocarbon hydroxylaseMetabolizes PAHs to carcinogenic intermediates. Associated with lung cancer risk.ER
CYP1A2Cytochrome P450, family 1, subfamily A, polypeptide 2Cholesterol 25-hydroxylaseMetabolizes some PAHs to carcinogenic intermediates.ER
CYP1B1Cytochrome P450, family 1, subfamily B, polypeptide 1Hydroperoxy icosatetraenoate dehydrataseMetabolizes procarcinogens, e.g., polycyclic aromatic hydrocarbons and 17β-estradiol.ER
CYP20A1Cytochrome P450, family 20, subfamily A, polypeptide 1 Iron ion binding and oxidoreductase activity.N.D.
CYP21A2Cytochrome P450, family 21, subfamily A, polypeptide 2Steroid 21-hydroxylaseSynthesis of steroid hormones, e.g., cortisol and aldosterone.ER
CYP24A1Cytochrome P450, family 24, subfamily A, polypeptide 1Vitamin D 24-hydroxylaseDegradation of 1,25-dihydroxyvitamin D3.Mit
CYP26A1Cytochrome P450, family 26, subfamily A, polypeptide 1Retinoic acid 4-hydroxylase4-Hydroxylation and 18-hydroxylation activities on retinoids, including all-trans-retinoic acid.ER
CYP26B1Cytochrome P450, family 26, subfamily B, polypeptide 1Retinoic acid-metabolizing cytochromeInactivation of all-trans retinoic acid to hydroxylated forms.ER
CYP26C1Cytochrome P450, family 26, subfamily C, polypeptide 1 Inactivation of all-trans retinoic acid to hydroxylated forms.N.D.
CYP27A1Cytochrome P450, family 27, subfamily A, polypeptide 15-β-Cholestane-3-α,7-α,12-α-triol 26-hydroxylaseOxidizes cholesterol intermediates as part of the bile synthesis pathway.Mit
CYP27B1Cytochrome P450, family 27, subfamily B, polypeptide 125-Hydroxyvitamin D3 1α-hydroxylaseHydroxylates 25-hydroxyvitamin D3 at the 1α position.Mit
CYP27C1Cytochrome P450, family 27, subfamily C, polypeptide 1All-trans retinol 3,4-desaturaseIron ion binding and oxidoreductase activity.N.D.
CYP39A1Cytochrome P450, family 39, subfamily A, polypeptide 1Oxysterol 7-α-hydroxylaseSubstrates include the oxysterols 25-hydroxycholesterol, 27-hydroxycholesterol, and 24-hydroxycholesterol.ER
CYP3A4Cytochrome P450, family 3, subfamily A, polypeptide 4Glucocorticoid-inducible P450Metabolism of approximately half the drugs in use today, including acetaminophen, codeine, cyclosporin A, diazepam, erythromycin, and chloroquine. The enzyme also metabolizes some steroids and carcinogens.ER
CYP3A43Cytochrome P450, subfamily IIIA, polypeptide 43 Hydroxylates testosterone; role in aging mechanisms and cancer progression.N.D.
CYP3A5Cytochrome P450, family 3, subfamily A, polypeptide 5 Metabolizes drugs and steroid hormones (testosterone and progesterone).N.D.
CYP3A7Cytochrome P450, family 3, subfamily A, polypeptide 5Xenobiotic monooxygenaseHydroxylates testosterone and dehydroepiandrosterone 3-sulfate in estriol formation during pregnancy.N.D.
CYP46A1Cytochrome P450, Family 46, Subfamily A, Polypeptide 1Cholesterol 24S-hydroxylaseConverts cholesterol to 24S-hydroxycholesterol.ER
CYP4A11Cytochrome P450, family 4, subfamily A, polypeptide 11Long-chain fatty acid ω-monooxygenaseHydroxylates medium-chain fatty acids, e.g., laurate and myristate.ER
CYP4A22Cytochrome P450, family 4, subfamily A, polypeptide 22Long-chain fatty acid ω-monooxygenaseHydroxylates medium-chain fatty acids, e.g., laurate and myristate.ER
CYP4B1Cytochrome P450, family 4, subfamily B, polypeptide 1Cytochrome P450-HPMetabolizes specific carcinogens.ER
CYP4F11Cytochrome P450, family 4, subfamily F, polypeptide 11Phylloquinone ω-hydroxylaseMetabolizes fatty acids and catalyzes N- and O-demethylation of specific drugs.N.D.
CYP4F12Cytochrome P450, family 4, subfamily F, polypeptide 12 Catalyzes the epoxidation of 22:6n-3 and 22:5n-3 polyunsaturated long-chain fatty acids; oxidizes arachidonic acid.ER
CYP4F2Cytochrome P450, family 4, subfamily F, polypeptide 2Leukotriene-B(4) ω-hydroxylase 1Degrades leukotriene B4 (inflammation mediator).ER
CYP4F22Cytochrome P450, family 4, subfamily F, polypeptide 22Ultra-long-chain fatty acid ω-hydroxylaseRole in the 12(R)-lipoxygenase pathway (inflammation mediator).N.D.
CYP4F3Cytochrome P450, family 4, subfamily F, polypeptide 3Leukotriene B4 ω hydroxylaseDegrades leukotriene B4 (inflammation mediator).ER
CYP51A1Cytochrome P450, family 51, subfamily A, polypeptide 1Lanosterol 14-α demethylaseParticipates in the synthesis of cholesterol by catalyzing the removal of the 14α-methyl group from lanosterol.ER
CYP7A1Cytochrome P450, family 7, subfamily A, polypeptide 1Cholesterol 7α-hydroxylaseConverts cholesterol to bile acids.ER
CYP7B1Cytochrome P450, family 7, subfamily B, polypeptide 125-Hydroxycholesterol 7-α-hydroxylaseConverts cholesterol to bile acids.ER
CYP8B1Cytochrome P450, family 8, subfamily B, polypeptide 17-α-Hydroxycholest-4-en-3-one 12-α-hydroxylaseCatalyzes the conversion of 7 α-hydroxy-4-cholesten-3-one into 7-α,12-α-dihydroxy-4-cholesten-3-one.ER
1 Information was taken from www.genecards.org (accessed on 25 February 2026); Abbreviations used: ER, endoplasmic reticulum; Mit, mitochondria; N.D., not described.
Table 2. Relationship between CYP mRNA expression combined with high mutational rates or high neoantigen load, and overall survival times of HCC (p < 0.05 and FDR < 5%).
Table 2. Relationship between CYP mRNA expression combined with high mutational rates or high neoantigen load, and overall survival times of HCC (p < 0.05 and FDR < 5%).
GeneNumberHazard RatioLog Rank (p)FDRRelationship
High mutation rates
CYP3A51800.36 (0.2–0.62)0.000161%high expression—long survival
CYP7A11800.35 (0.21–0.61)8.1 × 10−51%high expression—long survival
CYP19A11802.71 (1.6–4.61)0.000131%high expression—short survival
CYP26B11802.41 (1.47–3.97)0.000351%high expression—short survival
CYP27A11800.37 (0.23–0.61)4.6 × 10−51%high expression—long survival
High neoantigen loads
CYP7A12460.41 (0.26–0.63)2.4 × 10−51%high expression—long survival
CYP26B12461.93 (1.26–2.97)0.00221%high expression—short survival
CYP27A12360.46 (0.3–0.71)3 × 10−43%high expression—long survival
Low neoantigen loads
CYP19A11042.61 (1.25–5.45)3.9 × 10−53%high expression—short survival
Table 3. Results and Forest plots of univariable and multivariable Cox regression analyses for overall survival in TCGA-LIHC, based on 325 evaluable cases. (A) Univariable Cox regression results for the six prognostic CYP genes. (B) Multivariable Cox regression results adjusted for age, sex, pathological stage, and histological grade. Hazard ratios (HR) below 1 indicate improved OS with higher gene expression, while HR above 1 indicate worse OS. In the forest plots, green squares indicate HR point estimates below 1, while red squares indicate HR point estimates above 1. Full results of the multivariable model are provided in Supplementary Table S1.
Table 3. Results and Forest plots of univariable and multivariable Cox regression analyses for overall survival in TCGA-LIHC, based on 325 evaluable cases. (A) Univariable Cox regression results for the six prognostic CYP genes. (B) Multivariable Cox regression results adjusted for age, sex, pathological stage, and histological grade. Hazard ratios (HR) below 1 indicate improved OS with higher gene expression, while HR above 1 indicate worse OS. In the forest plots, green squares indicate HR point estimates below 1, while red squares indicate HR point estimates above 1. Full results of the multivariable model are provided in Supplementary Table S1.
(A)
GeneUnivariable HR (95% CI)p-ValueForest Plot
CYP3A50.893 (0.823–0.969)0.007Livers 06 00050 i001
CYP3A430.900 (0.835–0.970)0.006
CYP7A10.921 (0.877–0.967)<0.001
CYP27A10.847 (0.755–0.951)0.005
CYP19A11.109 (1.045–1.176)<0.001
CYP26B11.126 (1.039–1.220)0.004
(B)
GeneMultivariable HR (95% CI)p-ValueForest Plot
CYP3A50.907 (0.833–0.987)0.024Livers 06 00050 i002
CYP3A430.931 (0.862–1.005)0.069
CYP7A10.945 (0.897–0.995)0.032
CYP27A10.905 (0.797–1.026)0.120
CYP19A11.087 (1.021–1.157)0.009
CYP26B11.098 (1.009–1.195)0.031
Table 4. Data of Kaplan–Meier analysis in HCC patients grouped by CYP mRNA expression and TP53 mutation status. Patients were divided into four groups based on low (↓) or high (↑) CYP mRNA expression and wild-type (WT) or mutated TP53 status (mut). Mean OS is shown in months. Groups were compared using the log-rank test.
Table 4. Data of Kaplan–Meier analysis in HCC patients grouped by CYP mRNA expression and TP53 mutation status. Patients were divided into four groups based on low (↓) or high (↑) CYP mRNA expression and wild-type (WT) or mutated TP53 status (mut). Mean OS is shown in months. Groups were compared using the log-rank test.
CYPCYP ↓ TP53_WTCYP ↑ TP53_WTCYP ↓ TP53_MutCYP ↑ TP53_MutLog-Rank p
No. of PatientsMean OS (Months)No. of PatientsMean OS (Months)No. of PatientsMean OS (Months)No. of PatientsMean OS (Months)
CYP3A51145912073484643610.040
CYP3A431085912674544337680.002
CYP7A11175711773454546590.016
CYP19A11336810162296462450.031
CYP26B11246911062385853460.079
CYP27A11026113270604831600.066
Table 5. Data of Kaplan–Meier analysis in HCC patients grouped by CYP mRNA expression and CTNNB1 mutation status. Patients were divided into four groups based on low (↓) or high (↑) CYP mRNA expression and wild-type (WT) or mutated CTNNB1 status (mut). Mean OS is shown in months. Groups were compared using the log-rank test.
Table 5. Data of Kaplan–Meier analysis in HCC patients grouped by CYP mRNA expression and CTNNB1 mutation status. Patients were divided into four groups based on low (↓) or high (↑) CYP mRNA expression and wild-type (WT) or mutated CTNNB1 status (mut). Mean OS is shown in months. Groups were compared using the log-rank test.
CYPCYP ↓ CTNNB1_WTCYP ↑ CTNNB1_WTCYP ↓ CTNNB1_MutCYP ↑ CTNNB1_MutLog-Rank p
No. of PatientsMean OS (Months)No. of PatientsMean OS (Months)No. of PatientsMean OS (Months)No. of PatientsMean OS (Months)
CYP3A51215711773415446560.029
CYP3A431285311080344753570.007
CYP7A11255611374374150570.030
CYP19A11026813664606527340.023
CYP26B11057113359575830400.163
CYP27A1141609772214366600.041
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Dawood, M.; Guthart, A.; Ooko, E.; Weiskirchen, R.; Efferth, T.; Boulos, J.C. Integrative Profiling of Metabolic CYP Expression, DNA Mutation Rates, and Immune Cell Infiltration for Survival Prognosis in Hepatocellular Carcinoma. Livers 2026, 6, 50. https://doi.org/10.3390/livers6030050

AMA Style

Dawood M, Guthart A, Ooko E, Weiskirchen R, Efferth T, Boulos JC. Integrative Profiling of Metabolic CYP Expression, DNA Mutation Rates, and Immune Cell Infiltration for Survival Prognosis in Hepatocellular Carcinoma. Livers. 2026; 6(3):50. https://doi.org/10.3390/livers6030050

Chicago/Turabian Style

Dawood, Mona, Axel Guthart, Ednah Ooko, Ralf Weiskirchen, Thomas Efferth, and Joelle C. Boulos. 2026. "Integrative Profiling of Metabolic CYP Expression, DNA Mutation Rates, and Immune Cell Infiltration for Survival Prognosis in Hepatocellular Carcinoma" Livers 6, no. 3: 50. https://doi.org/10.3390/livers6030050

APA Style

Dawood, M., Guthart, A., Ooko, E., Weiskirchen, R., Efferth, T., & Boulos, J. C. (2026). Integrative Profiling of Metabolic CYP Expression, DNA Mutation Rates, and Immune Cell Infiltration for Survival Prognosis in Hepatocellular Carcinoma. Livers, 6(3), 50. https://doi.org/10.3390/livers6030050

Article Metrics

Back to TopTop