Gender Differences in the Livers of Patients with Hepatocellular Carcinoma and Chronic Hepatitis C InfectionNishida N.a, b · Arizumi T.a · Hayaishi S.a · Takita M.a · Kitai S.a · Yada N.a · Hagiwara S.a · Inoue T.a · Minami Y.a · Ueshima K.a · Sakurai T.a · Ikai I.c · Kudo M.a
aDepartment of Gastroenterology and Hepatology, Faculty of Medicine, Kinki University, Osaka, bDepartment of Gastroenterology and Hepatology, Kyoto University Graduate School of Medicine, Kyoto, and c Department of Surgery, National Hospital Organization Kyoto Medical Center, Kyoto, Japan
Objectives: A unique causative aspect of hepatocellular carcinoma (HCC) is a gender difference in its incidence. To determine the specific factors that contribute to a male predominance, we analyzed the clinicopathological factors, and genetic and epigenetic alterations of HCCs in male and female patients. Methods: We retrospectively analyzed three cohorts of patients: the first cohort consisted of 547 patients identified with the first event of HCC, the second cohort included 176 HCC patients, and the third 127 patients with chronic hepatitis C (CHC). Results: Male patients were found to have HCC more frequently than female patients in cases of non-cirrhotic liver (p = 0.0030 by the χ2 test), especially in hepatitis C-positive cases. However, there were no gender-specific differences in the genetic and epigenetic alterations of cancer-related genes. Deposition of iron was more severe in male CHC patients than in female patients. Conclusions: Male patients with CHC develop HCC more frequently when they have a non-cirrhotic liver than do female patients. This gender difference could be, at least partially, attributed to a different degree of iron deposition, which contributes to the development of HCC in the absence of liver cirrhosis in men with CHC.
Copyright © 2012 S. Karger AG, Basel
Hepatocellular carcinoma (HCC) is a common malignancy worldwide and the overall prevalence of the at-risk population is expected to grow with time. HCC is causatively associated with several distinct risk factors, such as chronic infection with hepatitis B virus (HBV) or hepatitis C virus (HCV) . Another unique causative aspect of this tumor is a gender-specific difference in its incidence. Men have a higher prevalence of HCC than women, and the ratio of affected men to affected women varies between 2:1 and 4:1 [1,2]. Although the reasons for the difference in incidence between men and women are not fully understood, it is possible that they include several environmental factors such as a higher prevalence of viral hepatitis, alcohol intake, and smoking in men compared to women . Recently, Naugler et al.  reported that hormonal factors might also play a role in the gender difference of the incidence of HCC development.
The male predominance of HCC is further supported by the clinical observation that chronic hepatitis progresses more rapidly to cirrhosis in men than in women. In fact, cirrhosis that leads to HCC development is considered to be more common in men and postmenopausal women, suggesting that sex hormones might play a role in the gender difference in the incidence of progression of liver fibrosis [5,6]. In this study, we retrospectively analyzed three cohorts of patients to determine which factors contribute to the gender difference in the incidence of HCC. The first cohort consisted of 547 HCC patients who were diagnosed as the first event of HCC, and we tried to identify the background factors associated with the difference between male and female patients. The second cohort included 176 HCC patients in whom the genetic and epigenetic alterations in HCC tissues had been examined previously, and we analyzed the gender differences of genomic alterations in this group [7,8,9]. The third cohort consisted of 127 patients with chronic hepatitis C (CHC), and we analyzed the gender differences in liver tissue biopsies with special focus on iron deposition.
In this report, we show that male patients with chronic viral hepatitis, especially in HCV-positive cases, tend to develop HCC more frequently in a background of non-cirrhotic liver with mild fibrosis than do female patients. This gender difference could be, at least partially, the result of a different degree of iron deposition, which contributes to the development of HCC in male CHC patients without severe fibrosis.
Materials and Methods
In the first stage of the study, 547 HCC patients consisting of 418 men and 129 women were examined retrospectively to determine the differences in their clinical backgrounds. These patients were identified with the first event of HCC at the Kyoto University Hospital between 1998 and 2006. The details of the profiles of the male and female patients are listed in table 1. The liver fibrosis stage (F-stage) of the background liver of the HCC patients was determined using the METAVIR scoring system . Informed consent was obtained from patients, and the study was approved by the institutional review boards of the institutions involved.
|Table 1. Gender difference in characteristics of the patients with HCC|
Previously we analyzed 176 pairs of HCC and their corresponding non-cancerous liver for mutations in the p53 and β-catenin genes, promoter hypermethylation of tumor suppresser genes (TSGs), the methylation status of long interspersed nucleotide element (LINE-1), and the degree of chromosomal alteration [7,8,9]. Using the data of this HCC cohort (the second HCC cohort), we compared the profile of genetic alterations in HCCs from male and female patients to understand the difference in the molecular pathogenesis of HCC. Details of this patient cohort were described previously [7,8].
Among them, mutation of the p53 and β-catenin genes were analyzed in 79 tumors by using direct sequencing of exon 3 of the β-catenin gene and exons 5 through 8 of the p53 gene in our previous study . To determine the degree of hypermethylation of TSGs, we quantified methylation levels on the promoters of 8 TSGs (HIC1, GSTP1, SOCS1, RASSF1, CDKN2A, APC, RUNX3, and PRDM2), which should play an important role in the initial events of human hepatocarcinogenesis using combined bisulfite restriction analysis . We analyzed the methylation status of interspersed nucleotide repeats, LINE-1, as a surrogate marker of global DNA methylation level using MethyLight . A serial dilution of bisulfite-treated CpGenome™ Universal Methylated DNA (Chemicon International, Inc.) was used for a standard curve and as a reference. The methylation-independent consensus Alu sequence was used as an endogenous control . In order to determine the number of chromosomal alterations, we analyzed allelic imbalance using 400 microsatellite markers (ABI PRISM® Linkage Mapping Set MD-10; Applied Biosystems). Fractional allelic loss (FAL) scores were calculated as the number of microsatellite loci with allelic imbalance divided by the number of total informative loci and expressed as a percentage, which broadly represents an index of the degree of chromosomal alteration in HCC .
In all, 22 and 13 of the 79 HCCs had a mutation of in p53 and β-catenin genes, respectively. The mean FAL (95% CI) was 22.6% (19.9–25.2). As the Z score was calculated as the difference between the methylation level of each tissue and the mean methylation level divided by the standard deviation, the mean methylation level expressed as a Z score should be zero. The mean methylation level (95% CI) onLINE-1 was 0.37 (0.34–0.40).
For the evaluation of iron deposition, we performed Prussian blue staining using liver biopsy samples from 127 CHC patients without a prior history of HCC. These patients consist of 84 men and 43 women with a median age of 57 years and interquartile range (IQR) of 46–64 years. The numbers of patients with F0, F1, F2, F3, and F4 were 7, 38, 36, 22, and 24, respectively. The degree of iron deposition in hepatocytes was classified as follows. If the proportion of hepatocytes with positive staining was ≥25% of the total hepatocytes population, this was regarded as a severe deposit, a positive staining rate of 24–6% was classified as a moderate deposit, and positive staining of ≤5% was classified as a weak deposit. Among 127 biopsies, 10 had a severe deposit, 41 had a moderate deposit, and 76 had a weak deposit.
In order to compare the categorical variables for the detection of gender differences, we applied the Pearson’s χ2 test, whereas the Wilcoxon rank-sum test and Student’s t test were used for comparisons of continuous variables. We applied the Z score for normalization when determining the methylation level on promoters of eight different TSGs. In order to discriminate between HCCs with different degrees of chromosomal alteration, we classified tumors into two groups: HCC samples with an FAL score >20% and tumors with FAL ≤20%. All p values were two-sided and p < 0.05 was considered statistically significant. All statistical analyses were calculated using the JMP version 9.0 software (SAS Institute, Inc., Cary, N.C., USA).
Initially, clinical background and status of tumor progression were compared between the 418 male and 129 female patients of the first cohort (table 1). The median age of the patients was significantly lower in the male group compared to the female group (median age and IQRs were 65 and 56–71 years for men and 66 and 60.5–73 years for women, p = 0.0208 by the Wilcoxon rank-sum test; table 1), suggesting that HCC emerges earlier in men than in women, as reported previously . There was no gender difference with respect to the presence or absence of hepatitis virus, and the majority of patients were HCV-positive. There was, however, a marked difference in the distribution of F-stage values in background liver between male and female patients. The proportion of F0–F2 (mild fibrosis) was significantly higher in men compared to women (179 of 330 patients (60.0%) for men, and 36 of 97 (37.1%) for women; p = 0.0030 by the χ2 test; table 1). The median tumor diameter was greater in male HCC patients compared to female HCC patients (median and IQRs were 4.5 and 3–7.5 cm for male patients, and 3.8 and 2.5–6.75 cm for female patients, p = 0.0314 by the Wilcoxon rank-sum test; table 1). However, there was no difference in tumor factors, indicating that the tumor stage was similar between the two groups.
We also analyzed several genomic alterations in HCC tumors from male and female patients among the second cohort of 176 HCC patients. No significant difference was detected in the clinical features, such as tumor diameter, the proportion of cases with well-differentiated HCC, no portal vein thrombosis, and a solitary tumor between the male and female groups, suggesting that there is no difference in tumor stage between male and female HCC cohorts (data not shown). Therefore, we compared genetic and epigenetic alterations in male and female HCC patients in this cohort, because there was no bias arising from differences in tumor stages. The frequencies of mutation in the p53 and β-catenin genes and the proportion of tumors with an FAL greater than 20% were 29% (17/59), 19% (11/59), and 47% (36/76), respectively, in male HCC patients, and 25% (5/20), 10% (2/20), and 48% (16/33), respectively, in female HCC patients. With respect to epigenetic alterations, the degree of hypermethylation in TSG promoters and hypomethylation of LINE-1 (a surrogate marker of global hypomethylation) in HCC tumors were similar in male and female patients. The mean methylation level of TSGs expressed as a Z score and 95% CI were 0.01 and –0.06 to 0.07 for HCC tumors in men, and 0.02 and –0.07 to 0.115 for tumors in women. The mean methylation level of LINE-1 and 95% CI were 0.36 and 0.32–0.40 for HCC tumors in men, and 0.39 and 0.34–0.44 for HCC tumors in women (table 1). Therefore, no gender difference was present with respect to genetic (mutation or chromosomal alteration) and epigenetic (methylation levels of gene promoter and repetitive DNA sequence) alterations, even between groups with similar tumor stages.
Next, we analyzed the association between F-stage in the background liver of HCC patients and gender among HBV-positive, HCV-positive, and virus-negative cases (fig. 1).
|Fig. 1. Gender difference in fibrosis of background liver in newly diagnosed HCC patients. The distribution of fibrosis of background liver was compared between male and female patients with newly diagnosed HCC with HBV-positive, HCV-positive, and virus-negative cases. The vertical line shows the proportion of patients in each fibrosis stage. Among 99 HBV-positive cases where the F-stage was determined, the number of patients showing each fibrosis stage was as follows: 9 male patients were F0, 15 were F1, 18 were F2, 7 were F3, and 22 were F4. For the female patients, 3 were F0, 4 were F1, 5 were F2, 6 were F3, and 10 were F4. Similarly, among 232 HCV-positive cases the number of patients showing each fibrosis stage was as follows: 9 male patients were F0, 31 were F1, 50 were F2, 29 were F3, and 62 were F4. Among the female patients, 2 were F0, 4 were F1, 6 were F2, 12 were F3, and 27 were F4. For the 93 virus-negative cases, the number of patients showing each fibrosis stage was as follows: 13 male patients were F0, 24 were F1, 11 were F2, 8 were F3, and 19 were F4. Among female patients, 3 were F0, 8 were F1, 3 were F2, 2 were F3, and 2 were F4. The p values showing differences between male and female patients determined by the χ2 test were p = 0.5209, p = 0.0224, and p = 0.7360 for HBV-positive, HCV-positive, and virus-negative cases, respectively. The p values showing gender difference using two categories of F0–F2 and F3 or F4 were: p = 0.1424, p = 0.0007, and p = 0.2256 for HBV-positive, HCV-positive, and virus-negative cases, respectively. Both HBsAg- and HCVAb-positive cases were included in the HBV-positive group as well as the HCV-positive group.|
There was no apparent gender difference with respect to F-stage in HBV-positive and virus-negative HCC cases (among the cases with F0–F4; p = 0.5209 and 0.7363 for HBV-positive and virus-negative HCC cases, respectively, by χ2 test; fig. 1). In contrast, the proportion of patients with ‘mild’ liver fibrosis of F0–F2 was significantly higher in male than in female HCV-positive HCC patients. Among 232 HCV-positive cases, the distribution of fibrosis scores was F0, 9; F1, 31; F2, 50; F3, 29, and F4, 62 in men and F0, 2; F1, 4; F2, 6; F3, 12, and F4, 27 in women (p = 0.0224 by χ2 test; fig. 1). The p values showing gender difference using two categories of F0–F2 and F3 or F4 were p = 0.0007 in HCV-positive cases.
The present analysis indicates that male CHC patients develop HCC more frequently when mild fibrosis is present in the liver. To clarify the factors that accelerate hepatocarcinogenesis in the absence of severe fibrosis, we examined the degree of iron deposition in the hepatocytes within CHC tumors by using 127 liver biopsies from the third cohort of CHC patients who had no prior history of HCC.
There was no difference in the F-stage between male and female patients in this CHC cohort (53 patients were regarded as F0–F2 and 29 were F3 or F4 among 84 male patients, and 28 were F0–F2 and 15 were F3–F4 among 43 female patients; p = 0.5936 by χ2 test). However, iron deposition in hepatocytes was more prominent in CHC tumors from male patients than in those from female patients. Among the male patients, 10 of 84 biopsies (11.9%) showed severe deposits, 27 of 84 (32.1%) had moderate deposits, and 47 of 84 (60.0%) had weak deposits, whilst in females patients, 14/43 (32.6%) had moderate deposits and 29/43 (67.4%) had weak deposits, but there were no cases of severe deposits (p = 0.0565 by Pearson’s χ2 test and p = 0.0118 by the likelihood ratio χ2 test). This trend of gender difference in the extent of iron deposition was present even among cases where mild fibrosis was present. 7 of 53 (13.2%) male CHC patients with F0–F2 showed severe iron deposits, 13 of 53 (34.0%) had moderate iron deposits and 28 of 53 (53.0%) had mild iron deposits. In the female patient group, however, 7 of 28 (25%) and 21 of 28 (75%) CHC cases with F0–F2 showed moderate and mild iron deposits, respectively, whereas there were no cases of severe iron deposits (p = 0.0589 by Pearson’s χ2 test and p = 0.0195 by the likelihood ratio χ2 test).
HCC is causatively associated with several distinct risk factors, and displays different types of genetic and epigenetic alterations [13,14]. It is now well established that HCC is more prevalent in men worldwide, with the reported male:female ratios ranging from 2:1 to 4:1, although the reasons for this are still unclear . In this retrospective study, we analyzed three cohorts of patients to determine the gender differences in HCC cases. We report that male patients with chronic viral hepatitis, especially in HCV-positive cases, tend to develop HCC more frequently in a non-cirrhotic background of mild liver fibrosis, compared to female patients.
Initially, we compared several clinical parameters between male and female patients with HCC by using the clinical records of a large number of HCC cases. Although we found no difference in tumor factors, male patients with HCC were on average significantly younger than female patients, and the fibrosis of background liver was significantly milder in male patients than in female patients. Previous reports have shown that epigenetic inactivation of TSGs as a result of promoter methylation is more prevalent in HCCs that develop in a cirrhotic liver of HCV-positive patients, and another pathway of hepatocarcinogenesis might exist where global DNA hypomethylation and profound chromosomal alteration took place [8,9,14,15]. Therefore, we expected that molecular pathways underlying hepatocarcinogenesis might differ between male and female patients. However, we did not find a gender difference in the mutation of specific genes in HCC, nor in the degree of methylation of TSGs, the degree of global DNA hypomethylation, and the extent of chromosomal alteration. This suggests that the molecular pathways underlying the progression of HCC might be similar in both male and female patients with hepatocarcinogenesis.
To further study gender differences in HCC formation, we focused on the status of the background liver, because a considerable number of HCCs emerged in the liver with mild fibrosis (stage F0–F2) in male cases. Thus, an additional mechanism for the progression of fibrosis accompanied by inflammation and necrosis might exist and thus contribute to hepatocarcinogenesis in men [4,6]. We analyzed the gender difference with respect to the progression of F-stage and infection with hepatitis virus. Interestingly, we found that gender differences in F-stage were present exclusively in HCV-related cases. However, although the fibrosis seemed to be milder in male than female patients in HBV-related HCC cases, this was not statistically significant. Recently, Naugler et al.  reported a protective effect of estrogen in hepatocarcinogenesis using a mouse model of HCC induced by diethylnitrosamine administration. Cellular damage induced by diethylnitrosamine led to the induction of interleukin-6 production in Kupffer cells via signaling of toll-like receptors, and estrogen was found to blunt this response . As an inflammatory cytokine, interleukin-6 could induce hepatic apoptosis and hepatocyte proliferation, and it could be a potential risk factor for developing HCC, especially in men [4,16]. It is also reported that an increased risk of HCC among men is associated with higher levels of androgen signaling leading to the induction of vascular endothelial growth factor or signal transducers and activator of transcription 3, especially in HBV-related HCC [16,17]. Therefore, differences in the hormonal status between men and women may contribute to the higher incidence of HCC in men, even when the liver is non-cirrhotic. However, inflammation-mediated carcinogenesis may well be accompanied by severe fibrosis via the activation of hepatic stellate cells by inflammatory cytokines [4,16]. Therefore, to clarify how HCC could emerge without severe fibrosis in men with CHC, we further analyzed the iron deposition in CHC patients because it could act as a carcinogen-promoting HCC development even in CHC patients. We studied a cohort of 127 CHC patients where there was no gender difference in F-stage. As expected, iron deposition was more severe in male CHC patients than in female patients, even when liver fibrosis was mild. It has been reported that iron could act as a mutagen because it can induce oxidative stress in several kinds of cells including hepatocytes through a Fenton reaction . Iron deposition was also associated with an increased amount of 8-OHdG in DNA, which in turn leads to genetic and epigenetic alterations of cancer-related genes [18,19]. We hypothesize that iron deposition contributes to the male predominance of HCC occurrence even among patients with mild fibrosis, especially in HCV-positive cases.
In this report, we have studied the gender difference in the background liver of HCC patients and concluded that HCC could develop from non-cirrhotic liver in the presence of mild fibrosis more often in male patients than in female patients, especially in HCV-positive cases. As we did not take the duration of chronic hepatitis into account , the data obtained here might, partially, be attributed to a bias of the retrospective analysis. Another limitation of the study was a lack of information regarding alcohol and tobacco consumption . However, as we selected only the patients with a first-time HCC diagnosis for the analysis, and severe iron deposit could, at least partially, explain the male predominance of HCC in mild fibrosis; thus, it could be possible that HCC develops earlier in the clinical course of male CHC patients rather than female CHC patients, in the background of a mild F-stage. Further evidence is required to confirm the association between gender and HCC development in the context of hepatitis without severe fibrosis.
This work was supported in part by a grant from the Smoking Research Foundation and a grant from the Osaka Cancer Research Foundation.
The authors have no conflicts of interest to disclose.
Naoshi Nishida, MD
Department of Gastroenterology and Hepatology
Kinki University School of Medicine, 377-2 Ohno-Higashi
Osaka-Sayama, Osaka 589-8511 (Japan)
Published online: December 13, 2012
Number of Print Pages : 7
Number of Figures : 1, Number of Tables : 1, Number of References : 20
Digestive Diseases (Clinical Reviews)
Vol. 30, No. 6, Year 2012 (Cover Date: December 2012)
Journal Editor: Malfertheiner P. (Magdeburg)
ISSN: 0257-2753 (Print), eISSN: 1421-9875 (Online)
For additional information: http://www.karger.com/DDI