Exposure to aflatoxins is a risk factor for hepatocellular carcinoma (HCC). Aflatoxins occur in peanut butter and are metabolized by genetically polymorphic enzymes such as glutathione-S-transferases encoded by glutathione-S-transferase μ 1 gene (GSTM1) and glutathione-S-transferase θ 1 gene (GSTT1) and microsomal epoxide hydrolase encoded by epoxide hydrolase gene (EPHX). The rate at which aflatoxins become activated or detoxified may depend on polymorphisms in the encoding genes. GSTM1 homozygous deletion was indeed found to modify the association between peanut butter consumption and HCC. In this study, we investigate possible roles of GSTT1 and EPHX polymorphisms in this relationship. From a Sudanese case-control study on HCC, we analyzed data of 112 incident cases and 194 controls. All participants were interviewed using a standardized questionnaire inquiring about social and demographic factors, peanut butter consumption, and other known HCC risk factors. Univariate analysis showed that GSTT1 polymorphism was not associated with HCC, whereas EPHX 113HH and 139HH genotypes increased the risk of HCC (Odds ratio, 3.10; 95% Confidence interval, 1.18–8.12). Adjustment for age and region of origin slightly attenuated this association (Odds ratio, 2.56; 95% Confidence interval, 0.83–7.95). Interestingly, unlike GSTM1, both GSTT1 and EPHX polymorphism did not modify the association between peanut butter consumption and HCC. In conclusion, these epidemiological findings do not suggest significant roles of GSTT1 and EPHX in aflatoxin metabolism, although EPHX polymorphism is possibly related to the increased risk of HCC. Further studies are needed to investigate mechanisms by which the EPHX polymorphism potentially modifies cancer risk.

HCC3 is one of the major cancer types in developing countries, where most significant risk factors are chronic hepatitis virus infection and exposure to aflatoxins (1). Aflatoxins are produced by Aspergillus fungi, which mainly occur in poorly stored maize and peanuts. In Sudan, aflatoxin exposure most likely occurs via consumption of peanut butter (2), which is a popular food. As we reported previously (3), peanut butter consumption is indeed related to the increased risk of HCC in Sudan. Genetic differences in aflatoxin metabolism may explain the observed differences between prevalence rates of HCC in populations with similar aflatoxin exposures and hepatitis infection rates (4).

The most potent mutagenic and carcinogenic of the aflatoxins is AFB1. Fig. 1 depicts the metabolism of AFB1, which is mainly metabolized by cytochrome P450 3A4 into the genotoxic metabolite AFB1-8,9-exo-epoxide. This metabolite can bind to DNA, causing G-to-T transversions (5) that may ultimately lead to cancer. Detoxification prevents formation of DNA adducts; the metabolite may be conjugated to glutathione by GSTs or may be hydrolyzed. Hydrolysis occurs spontaneously or is catalyzed by mEH (6, 7).

Two genetically polymorphic GSTs play a role in AFB1 detoxification: GST-μ encoded by the GSTM1 gene and GST-θ encoded by GSTT1. Homozygous deletion of part of these genes (null genotype) results in enzyme deficiency and might therefore lead to hampered detoxification (8). Several studies (9, 10), among which our own Sudanese case-control study (3), showed that of populations exposed to aflatoxins, only subjects carrying the GSTM1-null genotype are at increased risk of HCC. After GST-μ, GST-θ showed highest efficiency for conjugation of glutathione to AFB1-8,9-exo-epoxide (Refs. 11 and 12; Fig. 1). Homozygous deletion of the GSTT1 gene was recorded in 24 to 38% of people from African origin (13). One study showed a positive association between GSTT1-null genotype and aflatoxin-albumin adduct levels among chronic HBV antigen carriers (10), but this was not confirmed by another study (14).

The role of mEH in AFB1-8,9-exo-epoxide hydrolysis is still unclear because of inconsistent results from biochemical as well as epidemiological studies (11). The encoding EPHX gene has two polymorphic sites occurring with allele frequencies of ∼30% (15, 16). In exon 3, the 113Y allele encodes tyrosine incorporation at position 113 of the resulting protein, whereas 113H codes for histidine. The allelic variants in exon 4, 139H and 139R, result in proteins with histidine and arginine at position 139, respectively (15). Both the 113H and the 139H allele are related to relatively decreased enzyme expression or stability (15, 17, 18) and may therefore be associated with an increased risk of HCC. Indeed, EPHX 113HH genotype increased the risk of HCC in a small case-control study (9) but was not associated to aflatoxin-albumin adduct levels in another study (14).

GSTT1 and EPHX polymorphism may especially be significant among subjects carrying the GSTM1-null genotype. Few studies (3, 9, 10, 14) addressed the role of GSTT1 and EPHX polymorphism in HCC etiology, and only one (14) simultaneously studied GSTM1 polymorphism. Moreover, there is some discrepancy between biochemical and epidemiological studies concerning possible roles for GST-μ, GST-θ, and mEH in AFB1 detoxification in ultimate aflatoxin-related HCC etiology. Therefore, we evaluated the possible roles of GSTT1 and EPHX polymorphism and their interaction with aflatoxin exposure (as estimated by peanut butter consumption) in HCC in a Sudanese case-control study, in all subjects and in the subgroup of GSTM1-null genotype carriers.

Population.

A case-control study investigating risk factors for HCC was conducted in Sudan between September 1996 and September 1998. The design and conduct of this study were described in detail by Omer et al.(3). In short, subjects residing in West and Central Sudan were eligible. HCC cases were diagnosed clinically by liver function test and ultrasound in five of the six hospitals in the capital Khartoum, situated between West and Central Sudan. Additional diagnosis by liver biopsy and histological examination was available for 95% of cases. All 150 contacted HCC patients provided oral informed consent. ∼5% of HCC patients died before they could be invited to participate in the study. Community-based controls were selected randomly in a 1:3 woman:men ratio, because HCC is more prevalent among men and in proportion to the respective population sizes of the two regions. Recruitment was done through sugar shops which hold a complete registration of inhabitants of their serving area. Registration at the shops is required for all inhabitants in the locality, irrespective of their income. In four randomly chosen localities in West Sudan and six localities in Central Sudan, 1 of the 10–15 sugar shops was selected randomly, and 20 households were selected from the sugar shop’s registration list. In each household, one control was recruited. All 205 invited subjects agreed to participate.

Data Collection.

On the basis of the experience obtained from an explorative study (19), a questionnaire was designed to assess peanut butter intake in a standardized manner. This questionnaire was administered orally by one of the authors (R. E. O.). All participants were interviewed personally, although some cases needed help of an accompanying family member because of serious illness. Cases were interviewed in the hospital, and controls were interviewed at home. Frequency of peanut butter consumption was assessed on a daily or weekly basis and was inquired both “in season” and “off-season,” the former corresponding to the period of the year with active agricultural practices, usually June–September. To assess quantity of consumption, we asked if peanut butter is eaten as a part of the meal or as a full meal, and we inquired the amount of peanut butter eaten per meal and the number of persons such a meal is shared with. If a person had stopped to consume peanut butter, the reason for stopping, the time since stopping, and usual peanut butter consumption before stopping were inquired.

Monthly frequency of peanut butter consumption was calculated as a weighted average of the frequencies of peanut butter consumption in season and off season (see Table 1, footnote c). The quantity of peanut butter consumed per month was subsequently calculated by multiplication of the average frequency of peanut butter consumption by the amount of peanut butter consumed per meal per person. The questionnaire also identified other HCC risk factors, such as hepatitis infection, smoking, and alcohol consumption.

For determination of genotypes and chronic hepatitis infection, blood samples were collected in 10-ml Venoject tubes. Blood samples of cases were drawn at the hospital, centrifuged immediately, and subsequently stored at −20°C. Because of severe illness, it was not possible to obtain a blood sample of 37 cases (25% of cases). Controls were sampled at home, and blood samples were first transferred in coolers at 4°C to regional hospital laboratories where they were centrifuged and then transferred to Khartoum National Health Laboratory. Here, all samples were kept at −20°C until transport to Wageningen, the Netherlands for further analysis. Of six controls (3%), no blood was available because of logistic difficulties. Thus, blood was collected from 113 cases and 199 controls.

Laboratory Analysis.

DNA was isolated from 200-μl whole blood using the QIAamp blood kit (Qiagen, Inc., Chatsworth, CA), stored at 4°C, and directly used as a template in PCR analyses. Because of small amounts and low quality of some blood samples, DNA isolation failed for samples of 1 case and 5 controls. In total, DNA of 112 cases and 194 controls was available for genotyping.

A multiplex PCR was done to determine presence or absence of the GSTM1 and GSTT1 genes simultaneously according to Arand et al.(20). As a positive PCR control, however, we used primers derived from β-globin (21) instead of histidine. Primers derived from GSTM1, GSTT1, and β-globin were 5′-GAACTCCCTGAAAAGCTAAAGC-3′ and 5′-GTTGGGCTCAAATATACGGTGG-3′, 5′-TTCCTTACTGGTCCTCACATCTC-3′ and 5′-TCACCGGATCATGGCCAGCA-3′, and 5′-CAACTTCATCCACGTTCACC-3′ and 5′-GAAGAGCCAAGGACAGGTAC-3′, respectively. After DNA electrophoresis on an ethidium bromide-stained agarose gel, the amplified products were visualized under UV light. The GSTM1 fragment was 215 bp, the GSTT1 fragment was 480 bp, and the β-globin fragment was 350 bp in size. GSTM1/T1 genotyping was conducted in duplicate. For two subjects, not enough DNA was left to repeat genotyping. Samples with inconsistent genotyping results (n = 2) were genotyped a third time, and these consistent results are presented in this paper.

Genetic polymorphism in EPHX exon 3 (113Y and 113H alleles) and exon 4 (139H and 139R alleles) were determined by RFLP analysis after amplification of the exons. For amplification of exon 3, we used primers described by Smith and Harrison (Ref. 22; 5′-GATCGATAAGTTCCGTTTCACC-3′ and 5′-ATCTTAGTCTTGAAGTGAGGAT-3′). A mismatch in the reverse primer incorporated an EcoRV restriction site in the amplicon of the 113Y allele resulting in digestion products of 23 and 140 bp. The 113H allele remained undigested. Amplification of exon 4 was done using primers described by Hassett et al. (Ref. 15; 5′-GGGGTACCAGAGCCTGACCGT-3′ and 5′-AACACCGGGCCCACCCTTGGC-3′), followed by restriction analysis with RsaI. The 139H allele was digested into two fragments (295 and 62 bp), and the 139R allele was digested into fragments of 174, 121, and 62 bp in size. The amplified products were visualized under UV light after DNA electrophoresis on an ethidium bromide-stained agarose gel.

Determination of chronic hepatitis B and C virus infection was done using Hepanostika HBV surface antigen and ORTHO HCV 3.0 ELISA test systems with enhanced SAVe test kits, according to the instructions of the manufacturer.

Data Analysis.

Analyses were restricted to subjects who completed the questionnaire and of whom genotyping results were available for at least one polymorphism, i.e., 112 cases and 194 controls. For 108 cases and 188 controls, information on all genotypes was available. For GSTM1 and GSTT1, results were available for 109 cases and 188 controls; EPHX genotype at exons 3 and 4 could be determined for 110 cases and 193 controls and for 110 cases and 184 controls, respectively. The GSTM1- and GSTT1-null genotypes were a priori considered as the high-risk genotypes. For EPHX, we first conducted the analyses separately for both polymorphic sites. Thus, when analyzing the exon 3 polymorphism, we did not consider the exon 4 polymorphism and vice versa. On the basis of the available literature on the association between EPHX genotype and phenotype, we considered 113HH and 139HH genotypes and the combination of these to be the EPHX high-risk genotypes (23). On the basis of the median peanut butter consumption in controls, we classified all subjects as low or high peanut butter consumers. Analyses were done using the SAS statistical software package (release 6.12). After univariate analyses, we adjusted for age (in tertiles, according to age distribution in the control group). Multiple variables (e.g., education level, job type, region of origin and hepatitis infection) were considered to be included in the model if numbers in the various cells were sufficiently large (i.e., >5 after one-way stratification). Variables remained in the model if their inclusion caused a change of ≥10% in the β-estimates. Although adjustment for age did not change the β-estimates, this variable was forced into the model because cases were significantly older than controls, and age is known to be related both to several HCC risk factors (e.g., alcohol consumption, smoking) and to the disease itself.

Table 1 shows the characteristics of the study population. Cases were older than controls, less educated, and more often resided in West Sudan. They also consumed significantly more peanut butter, both in amount as in frequency per month. Besides, more cases were chronic carriers of the hepatitis B or C virus. Results on GSTM1 have been published previously but are now presented jointly with other polymorphisms. Frequencies of the GSTM1- and GSTT1-null genotypes were similar for cases and controls. There were no statistically significant differences in frequency of EPHX variants in exons 3 and 4. However, cases more often had the EPHX 113HH genotype in combination with the EPHX 139HH genotype than controls (Table 1). Results of the West Sudanese population were similar to results of the total population. However, in contrast to the West Sudanese population, Central Sudanese cases and controls did not differ with respect to peanut butter consumption (both the amount and frequency) and EPHX genotype (Table 1). Allelic variants of the studied polymorphic genes occurred at similar frequencies among controls from West and Central Sudan; the biggest difference between the regions was observed for EPHX exon 3: 113HH and 113YH genotypes occurred in 10 and 30% of Western controls and in 6 and 23% of Central Sudanese controls, respectively. The results presented in Table 1 did not change significantly after adjustment for age.

In Table 2, ORs of HCC are shown for genetic polymorphism of GSTM1, GSTT1, and EPHX. Univariate ORs indicate that only the combination of the EPHX 113HH and 139HH increased the risk of HCC. Adjustment for age and region in a multivariate model did not change the results on GST polymorphisms, although the association between the EPHX 113HH and 139HH genotypes and HCC lost statistical significance. Although inclusion of a variable for hepatitis infection changed ORs significantly (by increasing ORs and 95% CIs for the GSTM1-null and EPHX 139HH and 139HR genotypes to statistically significant values), it was not included in the multivariate model because this substantially inflated the SEs of the β-estimates. This was probably attributable to limited data: only 16 controls showed evidence of hepatitis infection, of whom 1 presented the EPHX 113HH and 139HH genotype, and 5 carried the GSTM1-null genotype.

To investigate if genetic polymorphism of GSTT1 or EPHX modified the association between peanut butter consumption and HCC, we stratified for peanut butter consumption. Results are shown in Table 3. We showed previously that peanut butter consumption was a risk factor for HCC, especially in GSTM1-null genotype carriers (3). Interestingly, effect modification by GSTT1 or EPHX genotype was not observed (ORs for the interaction terms were 1.3 and 1.0, respectively). In contrast to GSTM1, GSTT1 and EPHX polymorphisms did not affect the association between peanut butter consumption and HCC in subgroups with relatively high peanut butter consumption. These results did not change significantly after adjustment for age and region.

In GSTM1-null genotype carriers, presence of the GSTT1-null genotype or the EPHX 113HH and 139HH genotypes did not increase the risk of HCC; ORs and 95% CIs were 0.7, 0.3–1.9 and 2.0, and 0.4–11.2, respectively. Strikingly, only those with GSTM1 non-null genotypes in combination with the EPHX 113HH and 139HH genotypes were at an increased risk of HCC (OR, 5.7; 95% CI, 1.2–28.2). Similarly, subjects with GSTT1 non-null genotypes and the EPHX 113HH and 139HH genotypes had an increased risk of HCC (OR, 22.2; 95% CI, 2.4–205.8). Because none of the cases and only 3 controls carried a combination of all three high-risk genotypes, i.e., GSTM1 null, GSTT1 null, and EPHX 113HH and 139HH, we could not evaluate if the presence of this combination specifically increased the risk of HCC.

GSTT1-null genotype alone or in combination with peanut butter consumption was not a risk factor for HCC in this Sudanese case-control study. EPHX polymorphism might play a modest role in HCC. However, no interaction with peanut butter consumption was observed, indicating that the mEH enzyme may not be significant in aflatoxin detoxification.

As in all case-control studies, several types of bias might have occurred in this study. Selection bias probably did not occur, because almost all suspected HCC cases were referred to one of the participating hospitals. Therefore, we estimate that only few cases from the two selected regions were missed. Only cases being able to travel to Khartoum were included in this study, and of these, we analyzed results of those providing blood and for whom genotyping results were available. These might be the relatively mildly diseased ones, because some cases might have died before reaching the hospital and blood sampling failed for severely ill cases. It is unlikely that this introduced selection bias, because HCC is fatal in almost all cases, and survival time is very short (1). Moreover, we do not expect GST and EPHX genotypes to influence HCC survival. We think that the control population correctly reflects the Sudanese population in the two regions with respect to habitual peanut butter consumption and exposure to other HCC risk factors, because both regions are inhabited by populations with culturally determined, relatively constant food habits.

Controls were frequency matched to cases for sex. No age matching was done, because information on age was not available from sugar shop registries. Because cases were older than controls, we adjusted for age in all analyses, although this did not change the results significantly. Controls were enrolled from the same two regions as the cases, proportional to the population size of each region. These regions were chosen to investigate if differences in HCC prevalence may be explained by differences in aflatoxin exposure. Because Central Sudan has more inhabitants than West Sudan, whereas HCC is more prevalent in West Sudan, controls were more likely to be from Central Sudan, and cases were more likely to be West Sudanese. Homogeneity with respect to ethnicity should be considered because like many populations, both the Western and Central study populations may be composed of subjects from different ethnic groups. For cultural reasons, we chose not to record the ethnicity of the participants. Although region (i.e., Central and West) is not a definite predictor of ethnicity, the Central Sudanese population is predominantly Arab whereas people of West Sudan are more related to Chadians. We controlled for possible differences in ethnic composition between the two regions by the inclusion of region as a cofactor in our multivariate model. Stratification for region showed that results of the West Sudanese population were similar to results of the total population. Thus, in West Sudan, peanut butter consumption and EPHX 113HH/139HH genotypes were risk factors for HCC. In Central Sudan, however, those factors were not related to HCC. Because aflatoxin contamination of peanuts is lower in Central than in West Sudan (19), aflatoxin is probably a less significant determinant of HCC in Central Sudan. The finding for EPHX genotype may be attributable to differences in the ethnic composition of the two regions, as we explained above.

Information bias is not a major consideration in this study, because the population has no knowledge on aflatoxin contamination and its potential hazards. Moreover, we used a standardized questionnaire, and all interviews were conducted by the same person.

Because aflatoxin exposure could not be measured directly, consumption of peanut butter was used to estimate the exposure of the population. Peanut butter is considered to be the main source of aflatoxins in this population: (a) it is a frequently consumed staple food highly contaminated with aflatoxins, especially with AFB1 (2, 19, 24); (b) other staple foods that may be contaminated with aflatoxins (i.e., corn, sorghum, and millet) were reported to contain aflatoxins at concentrations below the reported hazard level (25); and (c) peanut butter consumption was indeed positively associated with the increased risk of HCC in our study, whereas the consumption of corn, sorghum, and millet was not.

In Sudan, life-style patterns are relatively simple because of limited food choice, limited amount and types of jobs, and religious uniformity. We therefore think our data are quite complete with respect to assessment of potential confounders, such as socioeconomic status (estimated by marital status, job type, income, and education level) and potential HCC risk factors (i.e., consumption of other main food types and alcohol consumption, smoking, and hepatitis). All variables assessing socioeconomic status were associated with HCC but did not confound the association between peanut butter consumption and HCC. Nevertheless, residual confounding could still have occurred. Because adjustment of the studied associations for potential confounders weakened our estimates only marginally, we expect that residual confounding did not distort the studied associations significantly.

The frequency of the GSTM1-null genotype in our control group was in-between frequencies reported for Europeans and sub-Saharan Africans (13, 16, 26), and the frequency of the GSTT1-null genotype was similar to prevalences found in other African studies (13, 14). Because EPHX polymorphism is known to vary greatly across people of different ethnicity (14, 16), our results on EPHX genotype frequencies cannot be easily extrapolated to other populations. Adjustment of our results for region neutralized the slight difference in EPHX genotype frequencies between the two regions.

Like GSTM1, the GSTT1-null genotype itself was not associated with HCC in this study. In contrast to GSTM1, GSTT1 did not modify the relation between peanut butter consumption and HCC. This may suggest that GSTT1 polymorphism is not important in aflatoxin metabolism or in HCC etiology. Possibly, conjugation of glutathione to AFB1-8,9-exo-epoxide is efficiently catalyzed by GST-μ, and in the case of GST-μ deficiency, GST-θ does not take over the function of GST-μ, but the metabolite is detoxified via hydrolysis instead. Our results correspond with those of a study in Gambia including 357 healthy subjects, showing that GSTT1 polymorphism was not related to aflatoxin-albumin adduct levels (14). Chen et al.(10), however, reported that HBV-positive, GSTT1-null genotype carriers had increased albumin adduct levels and were at an increased risk of HCC in a small study of 32 cases and 73 controls.

We found that the combination of EPHX 113HH and 139HH genotypes increased the risk of HCC. In a study in China including 52 cases and 116 controls, McGlynn et al.(9) found that carriers of the 113HH genotype were at an increased risk of HCC. Wild et al.(14), however, reported that this genotype was not associated with aflatoxin-albumin adducts. Both research groups did not study the EPHX exon 4 polymorphism. We found no indications for the EPHX polymorphism to modify the relationship between peanut butter consumption and HCC. Biochemical studies showed that mEH-catalyzed hydrolysis is probably not a rate-limiting step in aflatoxin detoxification, because hydrolysis of the AFB1-8,9-exo-epoxide can occur spontaneously at a rate comparable with that of mEH-catalyzed hydrolysis (27). The role of mEH in detoxification of polycyclic aromatic hydrocarbons might be more significant (28). This could explain the increased risk of liver (9), ovarian (29), and colon cancer (30) associated with allelic variants of EPHX. However, in vitro studies did not reveal a clear genotype-phenotype correlation for the polymorphic sites of EPHX(15, 18, 31, 32). This might imply that the studied polymorphic sites are linked to other genetically polymorphic sites, e.g., sites in noncoding regions of the gene, thereby modifying regulation of gene transcription (33).

Wild et al.(14) found that the presence of GSTM1- and GSTT1-null genotypes in combination with the EPHX 113HH genotype was related to elevated aflatoxin-albumin adduct levels, although not with statistical significance. We could not evaluate the effect of such a combination, because none of the cases and only 3 controls in our study had this combination of genotypes. However, the combinations of GSTM1 or GSTT1 non-null genotypes with the EPHX 113HH and 139HH genotypes increased the risk of HCC, which is not in line with the proposed metabolism (see Fig. 1). These unexpected findings might be attributable to chance because of the small number of observations per category, but they might also indicate that mEH is involved in other metabolic pathways rather than AFB1 detoxification.

We have estimated that as much as/up to 50% of HCC cases in Sudan might be attributed to HBV infections (34). Hepatitis infection has been reported to modify the relation between genetic polymorphism and HCC (9, 35). However, we could not study such effect modification in our study population because the number of controls with evidence of hepatitis infection was too small.

Although being the largest study on aflatoxin-associated HCC and genetic polymorphism to date, the population was still small to study relatively weak gene-environment interactions. Especially our results on EPHX should be interpreted with care, because 113 HH and 139 RR genotypes occurred at low frequency, and after stratification, some cells had <5 observations.

In conclusion, our results do not indicate that GSTT1 plays a role in HCC. EPHX polymorphism might be related to HCC, although the encoded enzyme may not be important in AFB1 detoxification in which GST-μ plays a major role. This seems to be in line with in vitro studies (11). Because EPHX polymorphism appears to be associated with cancer risk, the specific effect of this polymorphism on gene expression and enzyme function and the role of the mEH enzyme in various metabolic pathways need to be addressed in future studies.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

        
1

Supported partially by the Sudanese Standard and Meteorology Organization, SSMO. Wageningen University provided scientific support and facilitated personnel exchange.

                
3

The abbreviations used are: HCC, hepatocellular carcinoma; AFB1, aflatoxin B1; mEH, microsomal epoxide hydrolase enzyme; GST, glutathione-S-transferase; GST-μ, glutathione-S-transferase μ enzyme; GSTM1, glutathione-S-transferase μ 1 gene; GST-θ, glutathione-S-transferase θ enzyme; GSTT1, glutathione-S-transferase θ 1 gene; EPHX, epoxide hydrolase gene; HBV, hepatitis B virus; OR, odds ratio; CI, confidence interval; SE, standard error.

Fig. 1.

Proposed metabolism of AFB1. Adapted from Guengerich et al.(7) and Eaton et al.(6). AFM1, aflatoxin M1; AFQ1, aflatoxin Q1; CYP, cytochrome P450.

Fig. 1.

Proposed metabolism of AFB1. Adapted from Guengerich et al.(7) and Eaton et al.(6). AFM1, aflatoxin M1; AFQ1, aflatoxin Q1; CYP, cytochrome P450.

Close modal
Table 1

General characteristics of the study population

Total populationWest Sudanese populationCentral Sudanese population
Cases (n = 112)Controls (n = 194)Cases (n = 71)Controls (n = 80)Cases (n = 41)Controls (n = 114)
Demographics       
 Region, n from West Sudan (%) 71 (63.4) 80 (41.2)a NAb NA NA NA 
 Age, mean (SD) 57.0 (12.2) 44.9 (10.9)a 56.9 (11.7) 48.2 (11.4)a 56.9 (13.1) 42.5 (10.0)a 
 Gender, n males (%) 86 (76.8) 146 (75.3) 56 (78.9) 60 (75.0) 29 (72.5) 86 (75.4) 
 Education, n illiterate (%) 78 (69.6) 67 (34.5)a 50 (70.4) 38 (47.5)a 27 (67.5) 29 (25.4)a 
Peanut butter consumption, mean (SD)       
 Average frequency (times/month)c 11.3 (11.8) 7.9 (9.1)a 12.7 (13.6) 7.2 (8.5)a 9.1 (7.4) 8.4 (9.5) 
 Average consumption (kg/month)c 0.68 (1.1) 0.29 (0.6)a 0.83 (1.22) 0.28 (0.47)a 0.44 (0.77) 0.30 (0.63) 
Other risk factors, n (%)       
 Hepatitis B infectiond 45 (41.3) 14 (7.3)a 31 (44.9) 7 (8.8)a 14 (35.0) 7 (6.2)a 
 Hepatitis C infectiond 13 (11.9) 3 (1.6)a 6 (8.7) 1 (1.3)a 7 (17.5) 2 (1.8)a 
 Positive history of alcohol consumption 44 (39.3) 60 (30.9) 29 (40.9) 32 (40.0) 15 (37.5) 28 (24.6) 
 Positive history of smoking 45 (40.2) 76 (39.2) 27 (38.0) 33 (41.3) 18 (45.0) 43 (37.7) 
Genotype, n (%)       
GSTe       
  GSTM1 null 47 (42.7) 73 (38.8) 29 (40.9) 29 (38.2) 17 (44.7) 44 (39.3) 
  GSTT1 null 39 (35.8) 71 (37.8) 27 (38.0) 29 (38.2) 12 (31.6) 42 (37.5) 
EPHX exon 3e       
  113YY 68 (61.8) 128 (66.3) 43 (61.4) 47 (59.5) 25 (62.5) 81 (71.1) 
  113YH 28 (25.5) 50 (25.9) 16 (22.9) 24 (30.4) 12 (30.0) 26 (22.8) 
  113HH 14 (12.7) 15 (7.8) 11 (15.7) 8 (10.1) 3 (7.5) 7 (6.1) 
EPHX exon 4e       
  139HH 63 (57.3) 102 (55.4) 40 (57.1) 37 (52.1) 23 (57.5) 65 (57.5) 
  139HR 44 (40.0) 69 (37.5) 28 (40.0) 28 (39.4) 16 (40.0) 41 (36.3) 
  139RR 3 (2.7) 13 (7.1) 2 (2.9) 6 (8.5) 1 (2.5) 7 (4.6) 
EPHX exon 3 and 4 combinede       
  113HH and 139HH 12 (10.9) 7 (3.8)a 11 (15.7) 3 (4.2)a 1 (2.5) 4 (3.5) 
Total populationWest Sudanese populationCentral Sudanese population
Cases (n = 112)Controls (n = 194)Cases (n = 71)Controls (n = 80)Cases (n = 41)Controls (n = 114)
Demographics       
 Region, n from West Sudan (%) 71 (63.4) 80 (41.2)a NAb NA NA NA 
 Age, mean (SD) 57.0 (12.2) 44.9 (10.9)a 56.9 (11.7) 48.2 (11.4)a 56.9 (13.1) 42.5 (10.0)a 
 Gender, n males (%) 86 (76.8) 146 (75.3) 56 (78.9) 60 (75.0) 29 (72.5) 86 (75.4) 
 Education, n illiterate (%) 78 (69.6) 67 (34.5)a 50 (70.4) 38 (47.5)a 27 (67.5) 29 (25.4)a 
Peanut butter consumption, mean (SD)       
 Average frequency (times/month)c 11.3 (11.8) 7.9 (9.1)a 12.7 (13.6) 7.2 (8.5)a 9.1 (7.4) 8.4 (9.5) 
 Average consumption (kg/month)c 0.68 (1.1) 0.29 (0.6)a 0.83 (1.22) 0.28 (0.47)a 0.44 (0.77) 0.30 (0.63) 
Other risk factors, n (%)       
 Hepatitis B infectiond 45 (41.3) 14 (7.3)a 31 (44.9) 7 (8.8)a 14 (35.0) 7 (6.2)a 
 Hepatitis C infectiond 13 (11.9) 3 (1.6)a 6 (8.7) 1 (1.3)a 7 (17.5) 2 (1.8)a 
 Positive history of alcohol consumption 44 (39.3) 60 (30.9) 29 (40.9) 32 (40.0) 15 (37.5) 28 (24.6) 
 Positive history of smoking 45 (40.2) 76 (39.2) 27 (38.0) 33 (41.3) 18 (45.0) 43 (37.7) 
Genotype, n (%)       
GSTe       
  GSTM1 null 47 (42.7) 73 (38.8) 29 (40.9) 29 (38.2) 17 (44.7) 44 (39.3) 
  GSTT1 null 39 (35.8) 71 (37.8) 27 (38.0) 29 (38.2) 12 (31.6) 42 (37.5) 
EPHX exon 3e       
  113YY 68 (61.8) 128 (66.3) 43 (61.4) 47 (59.5) 25 (62.5) 81 (71.1) 
  113YH 28 (25.5) 50 (25.9) 16 (22.9) 24 (30.4) 12 (30.0) 26 (22.8) 
  113HH 14 (12.7) 15 (7.8) 11 (15.7) 8 (10.1) 3 (7.5) 7 (6.1) 
EPHX exon 4e       
  139HH 63 (57.3) 102 (55.4) 40 (57.1) 37 (52.1) 23 (57.5) 65 (57.5) 
  139HR 44 (40.0) 69 (37.5) 28 (40.0) 28 (39.4) 16 (40.0) 41 (36.3) 
  139RR 3 (2.7) 13 (7.1) 2 (2.9) 6 (8.5) 1 (2.5) 7 (4.6) 
EPHX exon 3 and 4 combinede       
  113HH and 139HH 12 (10.9) 7 (3.8)a 11 (15.7) 3 (4.2)a 1 (2.5) 4 (3.5) 
a

Significantly different from cases (P < 0.05).

b

NA, not applicable.

c

Average frequency calculated as: [3 × freq “in season” + 9 × freq “off-season”] ×

\(\frac{1}{12}\)
⁠; average consumption calculated as: [3 × freq “in season” + 9 × freq “off-season”] × portion size ×
\(\frac{1}{12}\)
.

d

Information on hepatitis B and hepatitis C virus infections was available for 109 cases and 193 controls; among West Sudanese participants, 1 case and 1 control had evidence of both hepatitis B and C virus infection.

e

Information on GSTM1 and GSTT1 genetic polymorphism was available for 109 cases and 188 controls, on EPHX exon 3 for 110 cases and 193 controls, and on EPHX exon 4 for 110 cases and 184 controls.

Table 2

Genotypes of GSTM1, GSTT1, and EPHX and risk of HCC

OR (95% CI)
UnivariateAdjusted for ageMultivariatea
GSTM1    
 Present 1.00 (Referent) 1.00 (Referent) 1.00 (Referent) 
 Null 1.18 (0.73–1.90) 1.26 (0.73–2.16) 1.26 (0.73–2.19) 
GSTT1    
 Present 1.00 (Referent) 1.00 (Referent) 1.00 (Referent) 
 Null 0.92 (0.56–1.50) 0.94 (0.54–1.64) 0.94 (0.54–1.65) 
EPHX exon 3    
113YY 1.00 (Referent) 1.00 (Referent) 1.00 (Referent) 
113YH 1.05 (0.61–1.82) 0.94 (0.51–1.73) 0.93 (0.50–1.71) 
113HH 1.76 (0.80–3.85) 1.68 (0.69–4.09) 1.51 (0.61–3.75) 
EPHX exon 4    
139RR 1.00 (Referent) 1.00 (Referent) 1.00 (Referent) 
139HR 2.75 (0.75–10.25) 3.67 (0.89–15.06) 3.63 (0.88–15.02) 
139HH 2.68 (0.73–9.76) 3.19 (0.80–12.79) 3.22 (0.80–13.01) 
EPHX exon 3 and 4 combined    
 All other combinations 1.00 (Referent) 1.00 (Referent) 1.00 (Referent) 
113HH and 139HH 3.10 (1.18–8.12) 2.97 (0.99–8.85) 2.56 (0.83–7.95) 
OR (95% CI)
UnivariateAdjusted for ageMultivariatea
GSTM1    
 Present 1.00 (Referent) 1.00 (Referent) 1.00 (Referent) 
 Null 1.18 (0.73–1.90) 1.26 (0.73–2.16) 1.26 (0.73–2.19) 
GSTT1    
 Present 1.00 (Referent) 1.00 (Referent) 1.00 (Referent) 
 Null 0.92 (0.56–1.50) 0.94 (0.54–1.64) 0.94 (0.54–1.65) 
EPHX exon 3    
113YY 1.00 (Referent) 1.00 (Referent) 1.00 (Referent) 
113YH 1.05 (0.61–1.82) 0.94 (0.51–1.73) 0.93 (0.50–1.71) 
113HH 1.76 (0.80–3.85) 1.68 (0.69–4.09) 1.51 (0.61–3.75) 
EPHX exon 4    
139RR 1.00 (Referent) 1.00 (Referent) 1.00 (Referent) 
139HR 2.75 (0.75–10.25) 3.67 (0.89–15.06) 3.63 (0.88–15.02) 
139HH 2.68 (0.73–9.76) 3.19 (0.80–12.79) 3.22 (0.80–13.01) 
EPHX exon 3 and 4 combined    
 All other combinations 1.00 (Referent) 1.00 (Referent) 1.00 (Referent) 
113HH and 139HH 3.10 (1.18–8.12) 2.97 (0.99–8.85) 2.56 (0.83–7.95) 
a

Multivariate model adjusted for age (tertiles) and region of origin.

Table 3

Genotypes of GSTM1, GSTT1, and EPHX as risk factors for HCC, stratified for peanut butter consumptiona

GenotypeORs and 95% CIs, age adjustedbORs and 95% CIs, multivariatec
LowaHighaLowaHigha
No. of cases/controlsOR (95% CI)No. of cases/controlsOR (95% CI)OR (95% CI)OR (95% CI)
All genotypes 33/98 1.00 (Referent) 78/96 2.62 (1.50–4.57) 1.00 (Referent) 2.50 (1.43–4.38) 
GSTM1       
 Present 25/55 1.00 (Referent) 37/60 1.57 (0.77–3.17) 1.00 (Referent) 1.43 (0.70–2.93) 
 Null 8/38 0.55 (0.21–1.47) 39/35 2.75 (1.30–5.78) 0.55 (0.21–1.47) 2.60 (1.22–5.54) 
GSTT1       
 Present 18/58 1.00 (Referent) 51/59 2.71 (1.30–5.63) 1.00 (Referent) 2.71 (1.30–5.63) 
 Null 14/35 1.25 (0.51–3.11) 25/36 2.28 (0.99–5.23) 1.26 (0.51–3.11) 2.28 (0.99–5.23) 
EPHX exon 3       
113YY 22/65 1.00 (Referent) 45/63 2.47 (1.24–4.94) 1.00 (Referent) 2.26 (1.12–4.56) 
113YH 6/24 0.74 (0.25–2.23) 22/26 2.34 (1.02–5.38) 0.66 (0.22–2.01) 2.28 (0.99–5.25) 
113HH 4/8 1.47 (0.35–6.20) 10/7 4.54 (1.33–15.49) 1.35 (0.31–5.78) 3.75 (1.07–13.12) 
EPHX exon 4       
139RR 21/54 1.00 (Referent) 41/48 1.90 (0.12–31.48) 1.00 (Referent) 1.70 (0.10–28.24) 
139HR 10/30 2.48 (0.24–25.30) 34/39 7.85 (0.83–74.67) 2.37 (0.23–24.07) 7.18 (0.76–67.98) 
139HH 1/9 2.88 (0.30–27.44) 2/4 6.54 (0.69–61.37) 2.82 (0.30–26.80) 6.02 (0.64–56.22) 
EPHX exons 3 and 4 combined       
 All other combinations 29/89 1.00 (Referent) 68/88 2.57 (1.43–4.62) 1.00 (Referent) 2.43 (1.34–4.39) 
113HH and 139HH 3/4 3.02 (0.51–18.01) 9/3 7.68 (1.65–35.69) 2.60 (0.41–16.69) 6.31 (1.32–30.05) 
GenotypeORs and 95% CIs, age adjustedbORs and 95% CIs, multivariatec
LowaHighaLowaHigha
No. of cases/controlsOR (95% CI)No. of cases/controlsOR (95% CI)OR (95% CI)OR (95% CI)
All genotypes 33/98 1.00 (Referent) 78/96 2.62 (1.50–4.57) 1.00 (Referent) 2.50 (1.43–4.38) 
GSTM1       
 Present 25/55 1.00 (Referent) 37/60 1.57 (0.77–3.17) 1.00 (Referent) 1.43 (0.70–2.93) 
 Null 8/38 0.55 (0.21–1.47) 39/35 2.75 (1.30–5.78) 0.55 (0.21–1.47) 2.60 (1.22–5.54) 
GSTT1       
 Present 18/58 1.00 (Referent) 51/59 2.71 (1.30–5.63) 1.00 (Referent) 2.71 (1.30–5.63) 
 Null 14/35 1.25 (0.51–3.11) 25/36 2.28 (0.99–5.23) 1.26 (0.51–3.11) 2.28 (0.99–5.23) 
EPHX exon 3       
113YY 22/65 1.00 (Referent) 45/63 2.47 (1.24–4.94) 1.00 (Referent) 2.26 (1.12–4.56) 
113YH 6/24 0.74 (0.25–2.23) 22/26 2.34 (1.02–5.38) 0.66 (0.22–2.01) 2.28 (0.99–5.25) 
113HH 4/8 1.47 (0.35–6.20) 10/7 4.54 (1.33–15.49) 1.35 (0.31–5.78) 3.75 (1.07–13.12) 
EPHX exon 4       
139RR 21/54 1.00 (Referent) 41/48 1.90 (0.12–31.48) 1.00 (Referent) 1.70 (0.10–28.24) 
139HR 10/30 2.48 (0.24–25.30) 34/39 7.85 (0.83–74.67) 2.37 (0.23–24.07) 7.18 (0.76–67.98) 
139HH 1/9 2.88 (0.30–27.44) 2/4 6.54 (0.69–61.37) 2.82 (0.30–26.80) 6.02 (0.64–56.22) 
EPHX exons 3 and 4 combined       
 All other combinations 29/89 1.00 (Referent) 68/88 2.57 (1.43–4.62) 1.00 (Referent) 2.43 (1.34–4.39) 
113HH and 139HH 3/4 3.02 (0.51–18.01) 9/3 7.68 (1.65–35.69) 2.60 (0.41–16.69) 6.31 (1.32–30.05) 
a

Low consumption defined as consumption of ≤135 g of peanut butter/month, high consumption defined as consumption of >135 g of peanut butter/month, based on the median consumption among controls.

b

Adjusted for age (tertiles).

c

Adjusted for age (tertiles) and region of origin.

We thank staff of hospitals, outpatient clinics, and health registries in Khartoum, North Kordofan State, and Gazira State for cooperation in the conduct of this study.

1
London W. T., McGlynn K. A. Liver cancer Schottenfeld D. Fraumeni J. F. eds. .
Cancer Epidemiology and Prevention
,
:
772
-793, Oxford University Press New York  
1996
.
2
Habish H. A., Abdulla M. H. The incidence of aflatoxin in Sudanese groundnuts.
Trop. Sci.
,
13
:
279
-287,  
1971
.
3
Omer R. E., Verhoef L., van’t Veer P., Idris M. O., Kadaru A. M. Y., Kampman E., Bunschoten A., Kok F. J. Peanut butter intake, GSTM1 genotype and hepatocellular carcinoma: a case-control study in Sudan.
Cancer Causes Control
,
12
:
23
-32,  
2001
.
4
Parkin, D. M., Whelan, S. L., Ferlay, J., Raymond, L., and Young, J. Cancer Incidence in Five Continents, Vol. 7. IARC Scientific Publ., Lyon, France: IARC, 1997.
5
Wang J. S., Groopman J. D. DNA damage by mycotoxins.
Mutat. Res.
,
424
:
167
-181,  
1999
.
6
Eaton D. L., Ramsdell H. S., Neal G. E. Biotransformation of aflatoxins Eaton D. L. Groopman J. D. eds. .
The Toxicology of Aflatoxin: Human Health, Veterinary, and Agricultural Significance
,
:
45
-71, Academic Press, Inc. San Diego, CA  
1994
.
7
Guengerich F. P., Johnson W. W., Ueng Y. F., Yamazaki H., Shimada T. Involvement of cytochrome P450, glutathione S-transferase, and epoxide hydrolase in the metabolism of aflatoxin B1 and relevance to risk of human liver cancer.
Environ. Health Perspect.
,
104 (Suppl. 3)
:
557
-562,  
1996
.
8
Liu Y. H., Taylor J., Linko P., Lucier G. W., Thompson C. L. Glutathione S-transferase μ in human lymphocyte and liver: role in modulating formation of carcinogen-derived DNA adducts.
Carcinogenesis
,
12
:
2269
-2275,  
1991
.
9
McGlynn K. A., Rosvold E. A., Lustbader E. D., Hu Y., Clapper M. L., Zhou T., Wild C. P., Xia X. L., Baffoe-Bonnie A., Ofori-Adjei D., Chen G. C., London W. T., Shen F. M., Buetow K. H. Susceptibility to hepatocellular carcinoma is associated with genetic variation in the enzymatic detoxification of aflatoxin B1.
Proc. Natl. Acad. Sci. USA
,
92
:
2384
-2387,  
1995
.
10
Chen C. J., Yu M. W., Liaw Y. F., Wang L. W., Chiamprasert S., Matin F., Hirvonen A., Bell D. A., Santella R. M. Chronic hepatitis B carriers with null genotypes of glutathione S-transferase M1 and T1 polymorphisms who are exposed to aflatoxin are at increased risk of hepatocellular carcinoma.
Am. J. Hum. Genet.
,
59
:
128
-134,  
1996
.
11
Guengerich F. P., Johnson W. W., Shimada T., Ueng Y. F., Yamazaki H., Langouet S. Activation and detoxication of aflatoxin B1.
Mutat. Res.
,
402
:
121
-128,  
1998
.
12
Johnson W. W., Ueng Y. F., Mannervik B., Widersten M., Hayes J. D., Sherratt P., Ketterer B., Guengerich F. P. Conjugation of highly reactive aflatoxin B1–8,9-exo-epoxide catalyzed by rat and human glutathione transferases: estimation of kinetic parameters.
Biochemistry
,
36
:
3056
-3060,  
1997
.
13
Rebbeck T. R. Molecular epidemiology of the human glutathione S-transferase genotypes GSTM1 and GSTT1 in cancer susceptibility.
Cancer Epidemiol. Biomarkers Prev.
,
6
:
733
-743,  
1997
.
14
Wild C. P., Yin F., Turner P. C., Chemin I., Chapot B., Mendy M., Whittle H., Kirk G. D., Hall A. J. Environmental and genetic determinants of aflatoxin-albumin adducts in the Gambia.
Int. J. Cancer
,
86
:
1
-7,  
2000
.
15
Hassett C., Aicher L., Sidhu J. S., Omiecinski C. J. Human microsomal epoxide hydrolase: genetic polymorphism and functional expression in vitro of amino acid variants.
Hum. Mol. Genet.
,
3
:
421
-428,  
1994
.
16
Masimirembwa C. M., Dandara C., Sommers D. K., Snyman J. R., Hasler J. A. Genetic polymorphism of cytochrome P4501A1, microsomal epoxide hydrolase, and glutathione S-transferases M1 and T1 in Zimbabweans and Venda of southern Africa.
Pharmacogenetics
,
8
:
83
-85,  
1998
.
17
Hassett C., Lin J., Carty C. L., Laurenzana E. M., Omiecinski C. J. Human hepatic microsomal epoxide hydrolase: comparative analysis of polymorphic expression.
Arch. Biochem. Biophys.
,
337
:
275
-283,  
1997
.
18
Laurenzana E. M., Hassett C., Omiecinski C. J. Post-transcriptional regulation of human microsomal epoxide hydrolase.
Pharmacogenetics
,
8
:
157
-167,  
1998
.
19
Omer R. E., Bakker M. I., van’t Veer P., Hoogenboom R. L., Polman T. H., Alink G. M., Idris M. O., Kadaru A. M., Kok F. J. Aflatoxin and liver cancer in Sudan.
Nutr. Cancer
,
32
:
174
-180,  
1998
.
20
Arand M., Muhlbauer R., Hengstler J., Jager E., Fuchs J., Winkler L., Oesch F. A multiplex polymerase chain reaction protocol for the simultaneous analysis of the glutathione S-transferase GSTM1 and GSTT1 polymorphisms.
Anal. Biochem.
,
236
:
184
-186,  
1996
.
21
Bell D. A., Taylor J. A., Paulson D. F., Robertson C. N., Mohler J. L., Lucier G. W. Genetic risk and carcinogen exposure: a common inherited defect of the carcinogen-metabolism gene glutathione S-transferase M1 (GSTM1) that increases susceptibility to bladder cancer.
J. Natl. Cancer Inst.
,
85
:
1159
-1164,  
1993
.
22
Smith C. A., Harrison D. J. Association between polymorphism in gene for microsomal epoxide hydrolase and susceptibility to emphysema.
Lancet
,
350
:
630
-633,  
1997
.
23
Pastorelli R., Guanci M., Cerri A., Negri E., LaVecchia C., Fumagalli F., Mezzetti M., Cappelli R., Panigalli T., Fanelli R., Airoldi L. Impact of inherited polymorphisms in glutathione S-transferase M1, microsomal epoxide hydrolase, cytochrome P450 enzymes on DNA, and blood protein adducts of benzo(a)pyrene-diolepoxide.
Cancer Epidemiol. Biomarkers Prev.
,
7
:
703
-709,  
1998
.
24
IARC. Some Naturally Occurring Substances: Food Items and Constituents, Heterocyclic Aromatic Amines and Mycotoxins. Monographs on the Evaluation of Carcinogenic Risks to Humans. IARC Sci. Publ., 56: 245–395, 1993.
25
Abdel-Rahman A. M., Osman N. A., Idris M. O. Survey of some cereal grains and legume seeds for aflatoxin contamination in the Sudan.
Zentbl. Mikrobiol.
,
144
:
115
-121,  
1989
.
26
Zhao L., Alldersea J., Fryer A., Tighe A., Ollier B., Thomson W., Jones P., Strange R. Polymorphism at the glutathione S-transferase GSTM1 locus: a study of the frequencies of the GSTM1 A, B, A/B and null phenotypes in Nigerians.
Clin. Chim. Acta
,
225
:
85
-88,  
1994
.
27
Johnson W. W., Yamazaki H., Shimada T., Ueng Y. F., Guengerich F. P. Aflatoxin B1 8,9-epoxide hydrolysis in the presence of rat and human epoxide hydrolase.
Chem. Res. Toxicol.
,
10
:
672
-676,  
1997
.
28
Seidegard J., DePierre J. W. Microsomal epoxide hydrolase: properties, regulation and function.
Biochim. Biophys. Acta
,
695
:
251
-270,  
1983
.
29
Lancaster J. M., Brownlee H. A., Bell D. A., Futreal P. A., Marks J. R., Berchuck A., Wiseman R. W., Taylor J. A. Microsomal epoxide hydrolase polymorphism as a risk factor for ovarian cancer.
Mol. Carcinog.
,
17
:
160
-162,  
1996
.
30
Harrison D. J., Hubbard A. L., MacMillan J., Wyllie A. H., Smith C. A. Microsomal epoxide hydrolase gene polymorphism and susceptibility to colon cancer.
Br. J. Cancer
,
79
:
168
-171,  
1999
.
31
Omiecinski C. J., Hassett C., Hosagrahara V. Epoxide hydrolase-polymorphism and role in toxicology.
Toxicol. Lett.
,
112–113
:
365
-370,  
2000
.
32
Hassett C., Robinson K. B., Beck N. B., Omiecinski C. J. The human microsomal epoxide hydrolase gene (EPHX1): complete nucleotide sequence and structural characterization.
Genomics
,
23
:
433
-442,  
1994
.
33
Raaka S., Hassett C., Omiencinski C. J. Human microsomal epoxide hydrolase: 5′-flanking region genetic polymorphisms.
Carcinogenesis
,
19
:
387
-393,  
1998
.
34
Omer R. E., van’t Veer P., Kadaru A. M. Y., Kampman E., El Khidir I. M., Fedail S. S., Kok F. J. The role of hepatitis B and hepatitis C viral infections in the incidence of hepatocellular carcinoma in Sudan.
Trans. R. Soc. Trop. Med. Hyg.
,
95
:
1
-5,  
2001
.
35
London W. T., Evans A. A., Buetow K., Litwin S., McGlynn K., Zhou T., Clapper M., Ross E., Wild C., Shen F. M., Chen G. C., M’Boup S. Molecular and genetic epidemiology of hepatocellular carcinoma: studies in China and Senegal.
Princess Takamatsu Symp.
,
25
:
51
-60,  
1995
.