Oxidative stress pathways in noncancerous human liver tissue to predict hepatocellular carcinoma recurrence: A prospective, multicenter study


  • Potential conflict of interest: Nothing to report.

  • This work was supported by a Health and Labour Sciences Research Grant (H20-Kanen-Ippan-001) from the Ministry of Health, Labour and Welfare of Japan and by Grant-in Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan.


The prediction of cancer recurrence holds the key to improvement of the postoperative prognosis of patients. In this study, the recurrence of early-stage hepatocellular carcinoma (HCC) after curative hepatectomy was analyzed by the genome-wide gene-expression profiling on cancer tissue and the noncancerous liver tissue. Using the training set of 78 cases, the cytochrome P450 1A2 (CYP1A2) gene in noncancerous liver tissue was identified as the predictive candidate for postoperative recurrence (hazard ratio [HR], 0.447; 95% confidence interval [CI], 0.249-0.808; P = 0.010). Multivariate analysis revealed the statistically significant advantage of CYP1A2 down-regulation to predict recurrence (odds ratio, 0.534; 95% CI, 0.276-0.916; P = 0.036), and the expression of CYP1A2 protein was confirmed immunohistochemically. An independently multi-institutional cohort of 211 patients, using tissue microarrays, validated that loss of expression of CYP1A2 in noncancerous liver tissue as the only predictive factor of recurrence after curative hepatectomy for early-stage HCC (HR, 0.480; 95% CI, 0.256-0.902; P = 0.038). Gene set-enrichment analysis revealed close association of CYP1A2 down-regulation with oxidative stress pathways in liver tissue (P < 0.001, false discovery rate [FDR] = 0.042; P = 0.006, FDR = 0.035). Our results indicate these pathways as the molecular targets to prevent recurrence, as well as the potential prediction of the super high-risk population of HCC using liver tissue. (HEPATOLOGY 2011;54:1273–1281)

Hepatocellular carcinoma (HCC) is one of the most common malignancies, accounting for nearly 700,000 deaths per year, and the incidence is still increasing worldwide.1 A major obstacle in treatment is the high frequency of tumor recurrence that is mostly limited to liver tissue, even after curative resection.2 There have been a number of studies reporting that advanced tumor factors, including size, number, and vascular invasion of cancer, were significantly associated with HCC recurrence.3 Genome-wide gene-expression analysis by DNA microarray offers a systematic approach to unfold comprehensive information regarding transcription profiles.4 Furthermore, such studies should potentially lead to the development of a novel, molecular-targeting therapy of HCC.5 We have previously analyzed the genome-wide gene expression of advanced HCC with recurrence exceeding Milan criteria6 (solitary, ≤5 cm or up to three nodules ≤3 cm, without major vascular invasion or distant metastasis)7 and identified novel molecules as therapeutic targets of HCC.8 Using a prediction system obtained from studies based on comprehensive genetic analysis, the selected genes may represent different biological characters that lead to HCC recurrence.

On the other hand, there has been little understanding of the mechanisms of recurrence from the early stage of HCC.9 It has been reported that gene-expression profiling with DNA microarray of noncancerous liver tissue was closely related to the prognosis in patients with early-stage HCC.10 Additionally, the translation of such profiling data to the bedside requires that gene expression be validated as protein expression, and that annotated clinical samples be available for prospective studies to assess the clinical context and usefulness of putative biomarkers. Tissue microarrays are promising tools of array-based technology in cancer research, and their importance in pathology is increasing because of their role in the clinical validation of DNA microarrays.11 Tissue-microarray technology allows researchers to examine the expression and location of protein on hundreds of tissue samples while preserving morphology. This increased throughput accelerates the discovery of important biologic markers, compared to traditional marker studies, using whole slide sections and has made this technology an essential tool in human protein profiling.12

The present study aimed at seeking a biomarker predictive for the recurrence after curative resection for early-stage HCC,3 from both of the cancerous and noncancerous gene-expression profiles in DNA microarray data.7 Using a prediction system obtained from studies based on training analysis, the biomarker was independently validated by the prospective, multicenter analysis on tissue microarray. Our training and validation studies might indicate this molecule as a novel biomarker predictive for the postoperative recurrence of the patients with early-stage HCC. It might also be potentially useful for the screening of the super high-risk group of HCC using liver tissue.


AIC, Akaike information criterion; AUC, area under the ROC curve; CI, confidence interval; CNDP1, carnosine dipeptidase 1; CYP1A2, cytochrome P450 1A2; FDR, false discovery rate; GSEA, gene set enrichment analysis; HR, hazard ratio; HCC, hepatocellular carcinoma; NES, normalized enrichment score; OAT, ornithine aminotransferase; OR, odds ratio; ROC, receiver operating characteristic.

Patients and Methods

Patients for the Training Set.

One hundred eighty-seven patients underwent curative hepatectomy for HCC from 2004 to 2007 at Tokyo Medical and Dental University Hospital (Tokyo, Japan). Among them, 98 cases met Milan criteria,6 (Barcelona Clinic liver cancer tumor stage 0-A),3 and written informed consent from 78 patients, as well as institutional review board approval, was obtained. Cancer tissue of HCC and noncancerous liver tissue adjacent to HCC were separately divided into two specimens immediately after surgery: one was snap-frozen in liquid nitrogen and stored at −80°C for microarray analysis, and the other was fixed in 10% formaldehyde solution and embedded in paraffin for histopathological analysis. Patients were followed up with assays of serum level of alpha-fetoprotein and protein induced by vitamin K absence or antagonists-II every month and with ultrasonography, computed tomography, and magnetic resonance imaging every 3 months. Median observation time was 15.0 months (95% confidence interval [CI], 14.0-21.0 months).

Gene-Expression Analysis.

Total RNA was extracted from cancer and adjacent noncancerous tissues using the RNeasy kit (Qiagen, Hilden, Germany), and integrity of obtained RNA was assessed using the Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA). All samples had an RNA Integrity Number greater than 5.0. Contaminant DNA was removed by digestion with RNase-free DNase (Qiagen). Using 2 μg of total RNA, complementary RNA was prepared using one-cycle target labeling and a control reagents kit (Affymetrix, Santa Clara, CA). Hybridization and signal detection of HG-U133 Plus 2.0 arrays (Affymetrix) was performed after the manufacturer's instruction. A total of 127 microarray datasets were normalized using robust multiarray average method under R statistical software (version 2.12.0), together with the BioConductor package. Estimated gene-expression levels were obtained in log2-transformed values, and 62 control probe sets were removed for further analysis.

Selection of Predictive Genes for HCC Recurrence.

To identify candidate genes for prediction of recurrence in early-stage HCC, we applied the combination of criteria for selection of gene probe sets (Fig. 1). First, probe sets corresponding to known genes were selected based on the NetAffx annotation file, version 31 (available at: http://www.affymetrix.com/analysis/index.affx). Then, we selected probe sets marked as “present” by Gene Expression Console software version 1.1 (version 1.1; Affymetrix) for more than 70% of patients. Next, the univariate Cox proportional hazards regression model was used to estimate the relationship between a gene-expression pattern and tumor recurrence rate for each probe set. Separate analyses were conducted for the cancer tissues, and the adjacent noncancerous tissues. Probe sets that satisfy P < 0.01 by the likelihood ratio test and more than 2-fold change in mean expression values between recurrence and nonrecurrence groups were selected. Furthermore, probe sets that satisfied P < 0.01 by the Wilcoxon signed-rank test and more than 2-fold change between the paired cancer and adjacent noncancerous tissues were selected. To identify the set of genes that best explain the recurrence of HCC, a multivariate Cox regression analysis with a forward variable-selection procedure, based on Akaike information criterion (AIC), was performed as, essentially, described by Lu et al.13 At each step, a variable showing the lowest AIC value was added. This procedure was started with a null model (i.e., a model with only the intercept parameter) and terminated if there was no improvement in the AIC value.

Figure 1.

Flowchart of gene-selection procedures in the training study.

Optimal Predictive Model for HCC Recurrence.

Clinicopathological factors associated with recurrence were examined by a univariate Cox regression analysis. Factors that satisfied P < 0.05 were subjected to further analysis. A multivariate Cox regression analysis with a forward variable-selection procedure was then performed in an identical manner to the gene-selection method described above using the candidate factors. To establish the optimal predictive model for HCC recurrence, expression levels of the candidate genes and clinicopathological factors were combined. Using selected genes and clinicopathological characteristics by the procedures described above, a multivariate logistic regression analysis with forward variable-selection procedure, based on AIC, was performed in a similar manner to that of Cox regression analysis. Diagnostic performance was examined by the receiver operating characteristic (ROC) curve and the area under the ROC curve (AUC).

Immunohistochemical Analysis.

Substantial evaluation of the expression of the candidate gene was assessed by immunohistochemical staining on tissue sections from the patients with HCC meeting Milan criteria. The immunohistochemical studies were performed using anti-CYP1A2 antibody (3B8C1: sc-53614; Santa Cruz Biotechnology, Inc., Santa Cruz, CA) at 1:500 dilution with phosphate-buffered saline containing 1% bovine serum albumin (Sigma-Aldrich, St. Louis, MO), with reaction in an automated immunostainer (Ventana XT System; Ventana Medical Systems, Inc., Tucson, AZ), using heat-induced epitope retrieval and a standard diaminobenzidine detection kit (Ventana). Positivity was defined as more than 25% of cells staining with anti-CYP1A2 antibody. Immunohistochemical staining was estimated under a light microscope by two independent investigators.

Validation Study on Tissue Microarrays.

To validate the clinical significance of the candidate molecule, it was assessed prospectively using a multicenter cohort from 2008 to 2009: Tokyo Medical and Dental University Hospital, The University of Tokyo Hospital, Tokyo Women's Medical University Hospital, Nihon University Hospital, and Juntendo University Hospital. All 211 enrolled patients with early-stage HCC meeting Milan criteria provided written informed consent, and the relevant institutional review board approved the study. Using the surgically resected samples, tissue microarrays were performed with an automated immunostainer (Ventana XT System). The immunohistochemical staining was evaluated under a light microscope by two independent investigators.

Gene Set Enrichment Analysis.

To investigate biological backgrounds correlated to a gene-expression pattern, we used gene set enrichment analysis (GSEA) version 2.0.7 with MSigDB gene sets version 3.0.14 Probe sets marked as present in more than 30% of patients were used for this analysis to reduce noise at low expression levels. Gene set category C5, which is based on the Gene Ontology database, was used. Gene sets satisfying both P < 0.05 and a false discovery rate (FDR) <0.05 were considered as significant.

Statistical Analysis.

All statistical analyses were performed using R statistical software (version 2.12.0), including the microarray analysis, as mentioned above. Fisher's exact test was used for analysis of categorical data, and an exact Wilcoxon rank-sum test and an exact Wilcoxon signed-rank test were performed using the wilcox_exact function provided by the “coin” package (The Comprehensive R Archive Network), and the significance level was set at 0.05.


Candidate Marker Genes for Prediction of HCC Recurrence.

Identification of candidate genes for recurrence of HCC was performed using the gene-expression profiles obtained by the DNA microarray (Fig. 1). Of 54,613 probe sets included in the HG-U133 Plus 2.0 array, 41,751 probe sets corresponding to the known human genes were selected. Among them, the numbers of probe sets marked as “present” in >70% of samples derived from cancer tissue and adjacent noncancerous tissue were 17,303 and 16,431, respectively. Using cancer tissue samples, 11 probe sets satisfied the criteria by the Cox likelihood ratio test for recurrence free (P < 0.01) and more than 2-fold change between the recurrence and nonrecurrence groups (Supporting Table 1). Among them, the criteria by Wilcoxon signed-rank test (P < 0.01) and more than 2-fold change between the paired cancer and noncancerous tissues were satisfied by two probe sets: immunoglobulin kappa locus and family with sequence similarity 134, member B (Table 1). Using noncancerous tissue samples, five probe sets satisfied the criteria by Cox regression analysis and more than 2-fold change (Supporting Table 1). Among them, three probe sets, cytochrome P450 1A2 (CYP1A2), carnosine dipeptidase 1 (CNDP1), and ornithine aminotransferase (OAT), showed significant down-regulation in paired cancer tissue. It is of interest that both carnosine and ornithine are known as amino acids closely related to oxidative stress in the liver.15, 16 The multivariate Cox regression analysis with the forward variable-selection procedure revealed that a model containing only the gene-expression pattern of CYP1A2 (HR, 0.447; 95% CI, 0.249-0.808; P = 0.010) was the best predictive model for the recurrence of HCC (Table 1).

Table 1. Cox Regression Analysis of Recurrence-Free Survivals in 78 Training Cases
 Univariate AnalysisVariable Selection
VariablesHR (95% CI)P ValueHR (95% CI)P Value
  1. Abbreviations: HR, hazard ratio; CI, confidence interval; IGK, immunoglobulin kappa locus; FAM134B, family with sequence similarity 134, member B; CYP1A2, cytochrome P450 1A2; CNDP1, carnosine dipeptidase 1; OAT, ornithine aminotransferase; HBs Ag, hepatitis B surface antigen; HCV Ab, hepatitis C virus antibody; ALT, alanine aminotransferase; AST, aspartate aminotransferase; AFP, alpha-fetoprotein; PIVKA-II, protein induced by vitamin K absence or antagonists-II; sm, surgical margin.

Molecular factors    
 IGK (cancer)0.798 (0.671-0.949)0.009
 FAM134B (cancer)0.783 (0.655-0.937)0.006
 CYP1A2 (noncancer)0.662 (0.515-0.850)0.0010.662 (0.515-0.850)0.001
 CNDP1 (noncancer)0.695 (0.554-0.872)0.002
 OAT (noncancer)0.617 (0.457-0.834)0.002
Clinicopathological factors    
 Age1.033 (0.994-1.074)0.084  
 Sex (female versus male)1.067 (0.535-2.129)0.853  
 HBs Ag0.849 (0.353-2.039)0.709  
 HCV Ab2.095 (1.048-4.188)0.031
Background liver 0.199  
 Normal liver1   
 Chronic hepatitis2.842 (0.371-21.753)   
 Liver cirrhosis3.971 (0.535-29.475)   
ALT0.998 (0.991-1.004)0.443  
AST1.007 (0.998-1.016)0.133  
Albumin0.447 (0.248 -0.808)0.0100.447 (0.249-0.808)0.010
Total bilirubin1.555 (0.652-3.707)0.332  
Child-Pugh A versus B2.907 (0.996-8.482)0.070  
AFP (ng/mL)1.264 (0.915-1.746)0.171  
PIVKA-II (mAU/mL)0.946 (0.599-1.492)0.808  
Tumor size (cm)1.058 (0.745-1.501)0.753  
Tumor number (multiple)1.293 (0.456-3.670)0.640  
Histological differentiation 0.216  
 Moderate1.553 (0.589-4.099)   
 Poor3.497 (0.913-13.394)   
Portal vein invasion1.335 (0.629-2.834)0.464  
Hepatic vein invasion8.590 (1.074-68.697)0.118  
eg versus ig0.936 (0.283-3.091)0.913  
fc (−) versus (+)1.571 (0.652-3.785)0.290  
fc-inf (−) versus (+)1.131 (0.589-2.169)0.711  
sm (−) versus (+)0.774 (0.298-2.012)0.588  

Clinicopathological Factors Associated With HCC Recurrence.

We examined the association between clinicopathological factors and recurrence-free survivals of 78 patients with early-stage HCC meeting the Milan criteria (Table 1; lower). Median recurrence-free survival time was 23.7 months. Univariate Cox regression analysis demonstrated that HCV infection (HR, 2.095; 95% CI, 1.048-4.188; P = 0.031) and low serum albumin level (HR, 0.447; 95% CI, 0.248-0.808; P = 0.010) correlated with the cumulative recurrence-free rate. Other clinicopathological factors were not statistically significant. We further performed the forward variable-selection procedure using the two factors, and the serum albumin level was identified as the only risk factor for recurrence after curative resection for early-stage HCC (Table 1).

Optimal Prediction Model for HCC Recurrence.

To evaluate the best predictive model for patients with HCC meeting the Milan criteria, we considered the combined model using the serum albumin level and the gene-expression pattern of CYP1A2 (Table 2). A multiple logistic regression analysis demonstrated that CYP1A2 (OR, 0.534; 95% CI, 0.276-0.916; P = 0.036) was the only factor that contributed to the prediction of the recurrence. Moreover, the variable-selection procedure by logistic regression analysis identified that CYP1A2 was the best predictor for the recurrence of HCC meeting the Milan criteria (OR, 0.493; 95% CI, 0.273-0.787; P = 0.008).

Table 2. Multivariate Logistic Regression Analysis of Recurrence-Free Survivals in 78 Training Cases
 Multivariate AnalysisVariable Selection
VariablesOR (95% CI)P ValueOR (95% CI)P Value
  1. Abbreviations: OR, odds ratio; CI, confidence interval; CYP1A2, cytochrome P450 1A2.

CYP1A20.534 (0.276-0.916)0.0360.493 (0.273-0.787)0.008
Albumin0.649 (0.126-3.181)0.594

Cumulative recurrence-free survivals were significantly associated with CYP1A2 expression (Fig. 2A). Lower expression of CYP1A2 was statistically related to the recurrence of early-stage HCC (P = 0.00993). The predictive accuracy of the CYP1A2 for the HCC recurrence was assessed by the ROC curve, and the AUC value was 0.747 (Fig. 2B). Protein expression of CYP1A2 was confirmed by immunohistochemical staining on adjacent liver tissues. The CYP1A2 protein localized to the membrane of the endoplasmic reticulum of hepatocytes (Fig. 2C).

Figure 2.

Training study. (A) Recurrence-free survivals of postoperative patients with early-stage HCC associated with expression of CYP1A2 gene in the noncancerous liver tissue. The median expression level for each gene was used as a cut-off value. Red and green lines indicate Kaplan-Meier curves for patients with relative overexpression and underexpression, respectively. The log-rank test was used to assess the statistical difference between the two groups (P = 0.00993). (B) Prediction accuracy of CYP1A2 for HCC recurrence using logistic regression. The AUC was 0.747. (C) Immunohistochemical analysis of CYP1A2 in adjacent noncancerous liver tissues from the training study. Upper panel: CYP1A2-high case; lower panel: CYP1A2-low case (magnification, ×200).

Validation Study on Tissue Microarrays.

To examine the predictive significance of the CYP1A2 expression, we prospectively conducted a multicenter validation study on 211 patients with HCC meeting the Milan criteria. Median observation time was 14.2 months (95% CI, 12.9-14.7) in the validation cases. As compared to that in the training cases (15.0 months), no significant difference was recognized (P = 0.108 by the Wilcoxon rank-sum test). Median recurrence-free survival time was 23.7 and 21.1 months in the training and validation cases, respectively; indicating no significant difference of recurrence (P = 0.583 by log-rank test; Supporting Fig. 1). According to the tissue microarray analysis of noncancerous liver tissues adjacent to HCC in the validation study (Fig. 3A), 15 of 211 patients were identified as CYP1A2 (−), and the cumulative recurrence-free rates of CYP1A2 (−) patients were significantly lower than CYP1A2 (+) patients (Fig. 3B; P = 0.020 by log-rank test). We also investigated the association between cumulative recurrence-free rates, clinicopathological factors, and by univariate Cox regression analysis (Table 3). Interestingly, recurrence was not correlated with any clinicopathological factors in the validation cohort, but only with the loss expression of CYP1A2 protein in noncancerous tissue (HR, 0.480; 95% CI, 0.256-0.902; P = 0.038). Further logistic regression analysis, using the 19 clinicopathological factors and CYP1A2 expression, also revealed that CYP1A2 (−) was the only significant factor by univariate (OR, 0.256; 95% CI, 0.069-0.778; P = 0.024) and multivariate assessments (OR, 0.247; 95% CI, 0.058-0.860; P = 0.038).

Figure 3.

Validation study using a prospective multicenter cohort. (A) Immunohistochemical analysis of CYP1A2 in adjacent noncancerous liver tissues using tissue microarray. Left panels: CYP1A2 (+) cases; right panels, CYP1A2 (−) cases (magnification,, ×100). (B) Cumulative recurrence free rate and protein expression of CYP1A2. Green and red line indicate CYP1A2 (−) and (+) patients, respectively. Statistically poor prognosis was related to loss expression of CYP1A2 (P = 0.020).

Table 3. Cox Regression Analysis of Recurrence-Free Survivals in 211 Validation Cases
VariablesHR (95% CI)P Value
  1. Abbreviations: HR, hazard ratio; CI, confidence interval; CYP1A2, cytochrome P450 1A2; HBs Ag, hepatitis B surface antigen; HCV Ab, hepatitis C virus antibody; AFP, alpha-fetoprotein; PIVKA-II, protein induced by vitamin K absence or antagonists-II.

Molecular factor  
 CYP1A2 (noncancer)0.480 (0.256-0.902)0.038
Clinicopathological factors  
 Age1.003 (0.982-1.024)0.808
 Sex (female versus male)1.100 (0.664-1.823)0.710
 HBs Ag1.134 (0.684-1.879)0.630
 HCV Ab1.016 (0.667-1.547)0.942
 Background liver 0.677
  Healthy liver1 
  Chronic hepatitis0.734 (0.332-1.623) 
  Liver cirrhosis0.844 (0.376-1.895) 
 Albumin1.108 (0.742-1.654)0.615
 Total bilirubin0.760 (0.412-1.404)0.369
 Child-Pugh A versus B0.271 (0.038-1.947)0.101
 AFP (ng/mL)1.182 (0.945-1.479)0.152
 PIVKA-II (mAU/mL)0.917 (0.678-1.240)0.568
 Tumor size (cm)0.887 (0.728-1.079)0.223
 Tumor number (multiple)1.616 (0.973-2.683)0.077
 Histological differentiation 0.258
  Moderately0.682 (0.418-1.112) 
  Poorly0.962 (0.443-2.091) 
 Portal vein invasion1.218 (0.719-2.065)0.473
 Hepatic vein invasion0.450 (0.165-1.227)0.077
 Invasive growth1.214 (0.661-2.229)0.542
 Capsular formation1.227 (0.740-2.033)0.419
 Capsular invasion1.162 (0.751-1.797)0.497
 Surgical margin1.328 (0.487-3.624)0.595

GSEA Evaluation for CYP1A2 Expression in HCC.

To identify biological pathways related to CYP1A2 expression, GSEA was performed using the gene-expression profiles of the 49 noncancerous tissues.14 Because CYP1A2 is one of the most major enzymes for xenobiotic metabolism in the liver,17 it was reasonable that most of the gene sets were associated with hepatic metabolism (Supporting Table 2). It is noteworthy that gene sets suppressing oxidative stress, such as PEROXISOME (P < 0.001; FDR = 0.042; normalized enrichment score [NES] = 1.808) and OXIDOREDUCTASE_ACTIVITY (P = 0.006; FDR = 0.035; NES = 1.846) demonstrated significantly positive correlation with CYP1A2 expression (Fig. 4). Our GSEA evaluation indicated that CYP1A2 down-regulation may be associated with degree of oxidative damage in the background liver.

Figure 4.

GSEA evaluation of gene-expression profile associated with CYP1A2. (A) PEROXIOME gene set (P < 0.001; FDR = 0.042; NES = 1.808). (B) OXIDOREDUCTASE_ACTIVITY gene set (P = 0.006; FDR = 0.035; NES = 1.846). Left panels: enrichment plot; right panels: heatmap of genes in the gene sets.


In the present study of the prediction of recurrence, we focused on early-stage HCC cases meeting Milan criteria. According to our results, postoperative recurrence was not associated with any tumor factor that might regulate intrahepatic metastasis (Tables 1 and 2). Indeed, recurrence was clinicopathologically associated with two host factors, serum albumin levels and HCV infection in our training cases (Table 1), suggesting that multicentric recurrence was dominant for the patients with chronic liver damages.18 Therefore, the assessment of noncancerous background tissue should reflect clinical outcomes that are not restricted to tumor progression.19, 20 Our retrospective study indicated that the noncancerous gene expression of CYP1A2, CNDP1, and OAT was significantly associated with recurrence (Table 1). The variable-selection procedure revealed the noncancerous CYP1A2 gene as the best predictive model for the recurrence of HCC, but not including the cancer-derived genes (Table 1). Further prospective, multicenter study validated that noncancerous CYP1A2 expression was identified as a unique biomarker for the prediction of recurrence after the curative resection of early-stage HCC (Table 3). Using tissue microarrays, CYP1A2 showed significant negative correlation with the cumulative recurrence-free rates (Fig. 3).

CYP1A2 is a major form of hepatic cytochorme P450 oxidative system, which is involved in drug metabolism and cholesterol synthesis.17 Decreased expression of hepatic CYP1A2 was known to be significantly correlated with fibrotic progression of hepatitis C patients21 and pathological progress of nonalcoholic fatty liver disease.22 Barker et al. reported previously that CYP1A2 was down-regulated dramatically by oxidative stress in hepatocytes, indicating CYP1A2 as a specific surrogate marker of hepatic oxidative damage.23 According to knockout mice analysis by Shertzer et al., oxidative stress was significantly elevated in the liver microsomes of CYP1A2-knockout mice, compared to those of wild-type or CYP1A1-knockout mice.24 In this regard, CYP1A2 may be considered not only a biomarker of oxidative stress, but also an antioxidant enzyme. The other noncancerous candidates, CNDP1 and OAT, might also be associated with oxidative stress by the modulation of amino acids carnosine15 and ornithine.16 Oxidative stress is known to induce DNA damage, and accumulation of such genetic damage can eventually lead to hepatocarcinogenesis.25

To evaluate the biological pathways associated with CYP1A2 expression, we utilized GSEA on the gene-expression profiles of the noncancerous liver tissues.14 GSEA can directly analyze the changes of gene-expression levels as continuous variables.26 According to our GSEA assessment, the gene sets of peroxisome and oxidoreductase activity were significantly correlated with CYP1A2 expression levels (Fig. 4). The peroxisome is an organelle that participates not only in the generation of reactive oxygen species, but also in cell rescue from the damaging effects of such oxidative radicals.27 Because the loss of hepatic CYP1A2 might be a surrogate biomarker of oxidative damage related to hepatocarcinogenesis, the analysis of CYP1A2 might be useful for the potential screening of the super high-risk population of HCC by liver biopsy in patients with chronic hepatitis.1

In conclusion, we identified the down-regulation of CYP1A2 in noncancerous liver tissue as a predictive factor for the recurrence of early-stage HCC. The significance of noncancerous CYP1A2 was confirmed by a validation study using the prospective, multicenter cohort. Close association of CYP1A2 was implicated with the oxidative stress pathways in liver tissue. With respect to antioxidant agents for the prevention of hepatocarcinogenesis,28, 29 further investigation is necessary to verify the roles of CYP1A2-oxidative signaling in early-stage HCC recurrence and, also, in hepatocarcinogenesis.


The authors thank Hiromi Ohnari and Ayumi Shioya for clerical and technical assistance.