A review of five different routes to the synthesis of the grilled or fried food carcinogen 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline (8-MeIQx) and of closely related compounds developed in different laboratories is presented. Interest in the synthesis of these amines began in the late 1970s when the research group led by Professor T. Sugimura (National Cancer Center, Tokyo) detected extremely high mutagenicity in the charred parts of fish and meat that could not be explained only by the presence of polyaromatic hydrocarbons. A number of new mutagenic heterocyclic amines have been detected, isolated and identified since then (de Meester, 1989; Overvik and Gustafsson, 1990; Felton and Knize, 1991; Jägerstad et al., 1991). It is still not entirely clear how these compounds are formed during cooking. For the "IQ-group" of the amines (2-amino-3-methylimidazo-quinoline and -quinoxaline congeners), a formation pathway from Maillard reaction products and creatinine was conceived by Professor K. Olsson (this laboratory) and presented at the 183rd meeting of the American Chemical Society, Las Vegas in 1982 (Jägerstad et al., 1983). However, the amounts of the amines formed during cooking or in model reaction systems are very small. Therefore, efficient and unambiguous synthetic methods yielding the compounds in isomerically pure form are required for reference purposes in analytical work and structure-biological activity studies. For instance, compare the mutagenicity of 4,8- and 5,8-DiMeIQx, and PhIP and its 3-methyl isomer shown on the following page. The pure compounds are also required in large quantities for long-term animal feeding studies. The length of this article does not allow a presentation of the published synthetic methods for all the heterocyclic amines. Therefore, the syntheses of only one of the food mutagens, 8-MeIQx, and some related compounds are presented. This will hopefully demonstrate the sort of problems the organic chemist encounters and some of their possible solutions. For a relatively recent reference list covering the synthesis, isolation, detection, formation and biological activity of these food mutagens and carcinogens the reader is referred to Hatch et al. (1988).
A significant proportion of patients with hepatocellular carcinoma (HCC) are infected with hepatitis C virus (HCV). This finding suggests that HCV infection is a major risk factor for the development of HCC. It is presently unclear whether HCV has a direct oncogenic effect on infected hepatocytes or whether continuous cell regeneration due to the chronic necroinflammatory process predisposes hepatocytes to mutations and malignant transformation. Except for rare cases, HCC in chronic HCV infection is always associated with cirrhosis. We examined a series of 138 consecutive cirrhotic livers with chronic hepatitis C that had been removed during transplantation for evidence of macroregenerative nodules (MRNs), liver cell dysplasia of large and small cell types, and HCC. MRNs and liver cell dysplasia are currently considered to be precursors of HCC. HCCs were present in 38 livers (28%) and were multifocal in half of the cases. MRNs were identified in 34 livers (25%). The number of MRNs ranged from 1 to 5 in 28 patients and was greater than 5 in 6 patients. In 14 of 34 livers with MRNs, there were associated HCCs (41%). Eight MRNs contained microscopic HCC. No microscopic HCC was found outside of MRNs; however, grossly apparent HCCs might have arisen from MRNs. Large liver cell dysplasia (LLCD) was frequently observed. It was present in 97 livers (70%) with or without MRNs and/or HCCs. Small liver cell dysplasia (SLCD) was seen in 8 livers with MRNs and/or HCCs and in 1 liver without MRN and HCC. These findings suggest that in chronic HCV infection, multifocal HCC is often found. MRN may represent one pathway in hepatocarcinogenesis. LLCD, similar to that found in chronic hepatitis B virus (HBV) infection, is a common finding in HCV-infected livers with cirrhosis, and appears not to be directly related to the development of HCC. SLCD is rarely seen, but may represent an important step in malignant transformation.
Studies on the natural course of virus-associated hepatocellular carcinoma (HCC) in high risk areas, particularly hepatitis B virus (HBV), have shown a stage of persistent liver cell hyperplasia characterized by a low level elevation of serum alpha-fetoprotein (AFP). We have recently identified a population of epithelial cells with distinct structure and expression of cytokeratin and AFP in non-neoplastic liver tissues from humans with HBV-associated HCC. These cells were characterized by oval nuclei, scant pale cytoplasm, small cell size, and cross-reactivity to a monoclonal antibody against rat oval cells. These human epithelial cells, putative human oval-type cells, stained strongly positive for cytokeratin 19 and displayed considerable heterogeneity in AFP and albumin expression. These findings suggest that a cell population structurally and phenotypically similar to oval cells seen in the early stages of chemical hepatocarcinogenesis in the rat is also present in humans in regenerating liver lesions observed in HBV-associated HCC. Hepatitis B surface antigen (HBsAg) was detected in 69, 81, and 85% of oval-type cells, transitional cells, and hepatocytes, respectively, but not in bile ducts or ductular cells. Also, high levels of expression of transforming growth factor alpha (TGF-alpha) were frequently seen in oval-type and transitional cells expressing HBsAg. These data suggest the possibility that oval-type cells are a target cell population for HBV infection; in the presence of elevated TGF-alpha expression, these cells may constitute a progenitor population for human HCC.
The 5' end of the hepatitis C virus (HCV) genome encodes structural proteins of the virion. The first gene encodes a highly basic core protein. Immediately downstream of the core gene are regions which encode the envelope proteins (E1 and E2) of the virus. Artificial expression and secretion of immunologically active envelope proteins have proven to be a substantial challenge due to the high degree of glycosylation and the existence of certain hydrophobic domains contained within these sequences. Bacterial cell expression of recombinant HCV envelope proteins results in products that are not glycosylated and are poorly immunogenic. Emphasis has shifted to the use of mammalian cell lines (human embryonic kidney [HEK] and Chinese hamster ovary [CHO] cells) for the expression of glycosylated, immunologically active envelope proteins. Using HEK cells, E1 is expressed intracellularly but is not secreted from the cells. When E1 is cloned in fusion with a C-terminal truncated E2 protein, both proteins are detected intracellularly; however, only E2 is secreted. When the E1/E2 processing site is interrupted by constructing deletion mutants, the unprocessed E1/E2 fusion protein can be secreted from the cells. Quantifiable expression and secretion of a truncated E2 protein is now possible using CHO cells and SV40-based vectors. The HCV E2 glycoprotein expressed from CHO cells is highly antigenic; a strong humoral response to this antigen develops in persons infected with HCV. Antibodies to E2 are found in 95% of patients with detectable HCV RNA in their sera. The presence of antibodies to E2 is not indicative of viral clearance and therefore the role these antibodies play in protective immunity, if any, is unclear.
The heterocyclic aromatic amine 2-amino-3-methylimidazo [4,5-f] quinoline (IQ) is one of three heterocyclic amine mutagens currently being evaluated for carcinogenic activity in nonhuman primates, primarily cynomolgus monkeys. IQ was administered by gavage five times a week at doses of 10 or 20 mg/kg. Thus far IQ induced tumors in 50 percent of the monkeys at the 10 mg/kg dose and in 85 percent of the monkeys at the 20 mg/kg dose. Because H and E sections of the myocardium from IQ-treated animals demonstrated inflammatory infiltrate and studies of IQ-DNA adducts in monkeys, by the 32P-postlabeling method, showed high levels of adducts in the heart a systematic study of cardiac pathologic changes associated with chronic administration of IQ was undertaken. Both light and electron microscopic abnormalities were seen. The possibility exists that heterocyclic aromatic amines may be etiologic factors for human cancer and myocardial disease.
2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) is the most abundant compound of the amino-imidazoazaarens (AIA) group of muta-/carcinogens isolated from the crust of fried meat. PhIP is principally activated via P450IA2 dependent N2-hydroxylation. A major metabolic pathway is N2-glucuronidation of the proximate 2-hydroxyamino-PhIP metabolite and excretion via bile to the intestine. After bacterial hydrolysis the proximate metabolite may be esterified by the intestinal cells and cause genetic damage. 2-Hydroxyamino-PhIP formed in vivo may be further oxidized presumably to 2-nitro-PhIP which reacts directly with glutathione through substitution of the nitro group. Detoxification is principally via P450IA1 dependent ring-hydroxylation followed by sulfation or glucuronidation. Direct glucuronidation also occurs. PhIP metabolism was examined in freshly isolated hepatocytes from rat, mouse, hamster and guinea pig. Activation was evaluated by the total level of covalent binding of PhIP to macromolecules. Rat hepatocytes had the lowest rate of metabolism, both to reactive and detoxified metabolites. The major products were 4'PhIP-sulfate, PhIP-glucuronide and 2-hydroxyamino-PhIP glucuronide, whereas in the mouse hepatocytes mainly 4'PhIP-sulfate was found. The level of covalent binding in the mouse hepatocytes exceeded those of the rat. An extensive metabolism was seen in guinea pig hepatocytes, the major products being 4'PhIP-sulfate, 4'-O-PhIP glucuronide, PhIP-glucuronide and 2-hydroxyamino-PhIP-glucuronide. The relative amount of PhIP covalently bound to macromolecules in guinea pig hepatocytes was low. Hamster hepatocytes had the highest level of covalently bound PhIP. The main metabolites were 4'PhIP-sulfate, 4'O-PhIP-glucuronide and PhIP-glucuronide. Minor amounts of 2-hydroxyamino-PhIP-glucuronide was produced in the hamster. Several unknown PhIP metabolites were formed in the hamster and guinea pig. Direct detoxification of PhIP and further metabolism of 2-hydroxyamino-PhIP to reactive and/or detoxified metabolites are important for the resulting covalent binding.
This study describes histopathological findings in 26 nonhuman primates (cynomolgus and rhesus monkeys) with liver tumors after dosing with IQ for time periods of 27 to 72 months. In most of the cases 2-5 well-defined tumor nodules were present in the liver at the time of autopsy. In 5 cases there were numerous nodules occupying most of the liver. Intrahepatic vascular invasion was observed in 14 cases, and lung metastases in 6 cases. The histology of 72 nodules from 21 cases with well-defined tumors and the 5 cases with massive involvement was studied. The majority of the tumors represented well to moderately well-differentiated hepatocellular carcinoma (HCC). Trabecular pattern was most common, followed by solid and pseudoglandular patterns. Five nodules displayed poorly differentiated pleomorphic variant of HCC with large, often multinucleated, tumor cells. Spindle cell and clear cell variants of HCC were seen in one nodule each. Two nodules were classified as cholangiocarcinomas. Immunostaining for cytokeratins 7 and 19 did not prove to be reliable markers for tumors of bile duct origin, since they were also expressed in some of the poorly differentiated HCC. The most common microscopic finding of the noncancerous portions of the liver were clear cell foci, composed of large, glycogen rich hepatocytes with small eccentric nuclei. Other lesions commonly found were dysplastic hepatocyte foci around the central veins and proliferating bile ductular-like (oval) cells along the portal areas and portal tracts.
Hepatitis C virus (HCV) is an important etiological agent in the development of chronic liver diseases such as chronic hepatitis, cirrhosis, and hepatocellular carcinoma (HCC). The virus, identified only recently, contains a single-stranded RNA genome of positive polarity, is distantly related to pestiviruses and flaviviruses, and has been classified as the first member of a third genus within the family Flaviviridae. Extensive analysis of HCV genomic sequences demonstrated that this virus possesses significant genetic heterogeneity. Different regions of the viral genome demonstrate a varying degree of heterogeneity; the regions coding for the putative envelope proteins are the most variable sites between different isolates. Furthermore, HCV circulates as a quasispecies in the host. During the course of acute and chronic infection, the sequence composition of the HCV population in one patient has been found to change sequentially with an extremely high rate of nonconserved nucleotide changes in the hypervariable region I (HVR1) of HCV. Such sequence changes alter the antigenicity of the epitopes coded within HVR1 so that these are not always recognized by preexisting antibodies. It has been suggested that this could represent one mechanism by which HCV evades host immune surveillance and may account for the high rate of chronicity observed in such infections. Continuous viral replication may, in turn, lead to the development of chronic liver disease, including HCC, in infected individuals. To date, at least nine major genetic groups (genotypes 1-9) and more than 30 subgroups of HCV have been recognized based on genetic differences. A distinct difference has been observed in the genotype distribution in Africa compared with other continents. Recent data have suggested a difference in pathogenesis and in the outcome of interferon therapy in individuals infected with HCV of certain genotypes. For example, genotype 1b (II) seems to be associated with more severe liver disease, including HCC, and with a poorer response to interferon therapy. The extensive genetic heterogeneity of HCV may have serious implications for the diagnosis, treatment and prevention of hepatitis C as well as in understanding the biology of infection by this important human pathogen.
Hepatocellular carcinoma is one of the most important cancers worldwide and at least two hepatitis viruses, hepatitis B virus (HBV) and hepatitis C virus (HCV) appear to be etiologically associated with its development. Regardless of the nature of this association, control of these two viruses would likely be effective in the prevention of hepatocellular carcinoma. The control of HBV is achievable through proper utilization of highly effective vaccines against HBV. However, the development of useful vaccines against HCV will require more knowledge about the genetic and serologic heterogeneity of this virus and the nature of the host's immune response to it.
Hepatitis B virus (HBV) is a co-factor in some hepatocellular carcinomas (HCC). Chronic infection with HBV is a risk factor for tumor development, suggesting the accumulation of cellular genetic changes. HBV DNA is frequently found integrated at random sites in HCC, with chromosomal deletions and rearrangements being common at the sites of viral integration. Tumor suppressor gene p53 is frequently altered in HCC. Environmental carcinogens are factors in HCC development in certain geographic locations. HBV encodes a protein (X) known to transactivate viral and cellular genes; the X gene is often retained in HCC. To learn more about X gene function. We employed the yeast two-hybrid genetic system to seek X-interactive proteins. A cellular protein, designated XAP-1, was recovered that interacts specifically with the X protein. XAP-1 is the human homologue of the monkey UV-damaged DNA-binding protein (UV-DDB); the UV-DDB protein functions in DNA repair and is defective in some xeroderma pigmentosum group E patients. The interaction between XAP-1 and HBV X protein was confirmed by several independent methods. This suggests that cellular DNA repair processes may be affected by HBV and that the resulting genetic instability may contribute to hepatocellular carcinogenesis. A unifying model of the molecular basis of HBV involvement in HCC development is presented. Fundamental components of the model are chronic infection by HBV and viral effects on cellular DNA repair. This model has implications for the possible role of HCV infection in the induction of HCV-associated HCC.
The metabolic activation of the heterocyclic amine carcinogen, 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP), was examined in dogs and rats as models for urinary bladder and colon carcinogenesis, respectively. The results indicate that unconjugated N-OH-PhIP is not excreted in the urine after oral dosing with PhIP and that the two isomeric N-glucuronides of N-OH-PhIP, which are formed as major metabolites, are stable under acidic conditions. These data suggest that PhIP is unlikely to serve as a urinary bladder carcinogen in either species. Using metabolic inhibitors, bile duct ligation, and intravenous dosing studies, a new hypothesis for colorectal carcinogenesis is proposed involving N-oxidation of PhIP by hepatic cytochrome P-4501A2 (CYP1A2) and O-acetylation by the polymorphic acetyltransferase (NAT2). The resulting N-hydroxy and N-acetoxy metabolites both appear to be transported through the circulation to the colon mucosa, forming covalent DNA adducts. Glucuronidation and reaction with glutathione appear to serve as detoxification pathways. In humans, individuals who are phenotypically rapid metabolizers for both CYP1A2 and NAT2 are significantly higher (p = 0.0015) in colorectal cancer/poly cases vs. controls; and PhIP-DNA adducts can be detected in human colon samples. These studies provide strong evidence that PhIP and other heterocyclic amines play an important role in the etiology of human colorectal cancer.
Estimating the cancer risk posed by heterocyclic amines depends on measuring how chemical dose influences measurable indicators of cancer progression. This data ideally should encompass the range of actual human exposure, at the low dose end, and laboratory animal studies, at the high dose end. Accelerator mass spectrometry (AMS) has been used to measure the absorption, fate, and DNA adduct dosimetry of the heterocyclic amines PhIP and MeIQx at doses equivalent to human consumption following single-dose administration and chronic daily dosing. AMS is a nuclear physics technique which specifically counts nuclei of cosmogenic isotopes, rather than relying on decay. For tracing 14C, sensitivity is increased 10(6)-fold relative to decay counting. We have found that tissue clearance rates for [2-(14)C]-PhIP are rapid (t1/2 = 1 h) at low dose (41 ng/kg), with most of the radiocarbon distributed to the liver and G.I. tract. MeIQx-DNA adduct levels decrease linearly with dose (5 mg/kg-500 ng/kg) in single dose exposures. Likewise, the biologically available dose of [2-(14)C]-MeIQx decreases linearly with decreasing dose (5 mg/kg-1 ng/kg). On chronic daily dosing, it takes 40 days for adducts to reach steady-state in tissues and adduct levels appear to decrease linearly with decreasing dose, except possibly at very low doses. DNA binding of PhIP involves both sulfation or acetylation of the N-hydroxylated PhIP. Quantitatively, sulfation appears to be an important pathway for PhIp activation in rodent tissue cytosols while acetylation appears quantitatively more important in human tissue cytosols. The greatest activity is in liver and intestinal tissues for both pathways. The specific DNA adducts formed in vivo and in vitro from exposure to PhIP and MeIQx are likely guanine adducts. These data suggest that DNA adduct dosimetry responds linearly with dose but may become sub-linear at very low doses for chronic exposure and that factors other than DNA adduction may be critical to explain these heterocyclic amines' tumorigenicity.
To identify environmental, viral, and genetic factors that may influence the risk of developing hepatocellular carcinoma (HCC), large prospective studies are being conducted in Haimen City, China and Senegal, and a case-control study of genetic variation in the detoxification of aflatoxin-B1 was carried out in Shanghai, China. Analysis of 78 HCCs that have occurred among 51,020 men enrolled in a large prospective study in Haimen City, China showed a strong association of HCC with chronic hepatitis B virus (HBV) infection. There were also significant associations of HCC risk with occupation (farming), history of a clinical episode of hepatitis in adulthood, and a family history of HCC. Study of 52 HCC cases and 116 controls for genetic polymorphisms and HCC risk showed a significant association with epoxide hydrolase (EPHX) mutant alleles (1/2, 2/2) and a borderline association with homozygous deletion of the glutathione-S-transferase mu (GSTM1) gene. There was a multiplicative interaction of these polymorphisms with chronic HBV infection such that HBsAg-positive persons who were GSTM1 null and were EPHX 1/2 or 2/2 had 135 times the risk of HCC as HBsAg-negative persons with the wild type genotypes for GSTM1 and EPHX. The risk of HCC is not uniform among persons with chronic HBV or HCV infections. Studies of genetic, viral, and environmental interactions may permit identification of those individuals at highest risk within groups at increased risk of HCC. Prevention strategies could then be targeted at those individuals.
2-amino-1-methyl-6-(4-hydroxyphenyl)imidazo[4,5-b]pyridine (4'-OH-PhIP) was mutagenic, inducing 180 revertants of Salmonella typhimurium TA98 per 100 micrograms with S9 mix and was formed by heating a mixture of creatine, tyrosine and glucose. It was detected in broiled beef at a level of 21.0 ng per g of broiled beef, which is comparable to the level of 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP). Two new mutagens were isolated from bacteriological-grade beef extract using a new Salmonella tester strain, YG1024, which has a much higher O-acetyltransferase level than TA98. These mutagens were identified as 2-amino-4-hydroxymethyl-3,8-dimethylimidazo[4,5-g]quinoxaline (4-CH2OH-8-MeIQx) and 2-amino-1,7,9-trimethylimidazo[4,5-g]quinoxaline(7,9-DiMeIgQx++ +). The amounts of these mutagenic heterocyclic amines (HCAs) in beef extract were 6.0 ng and 53 ng per g of beef extract, respectively. 4-CH2OH-8-MeIQx induced 326,000 revertants of YG1024 and 99,000 revertants of TA98 per microgram with S9 mix, while 7,9-DiMeIgQx induced 13,800 and 670 revertants of YG1024 and TA98, respectively, per microgram in the presence of S9 mix. The levels of nine previously reported HCAs in cooked meats and fish and in beef extract were determined quantitatively. The level of PhIP was highest (0.56 approximately 69.2 ng/g), followed by that of 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline (MeIQx) (0.64 approximately 6.44 ng/g), and those of other HCAs were 0.03 approximately 2.50 ng/g. Mainstream smoke condensates of five Japanese brands of cigarettes contained four HCAs, 3-amino-1,4-dimethyl-5H-pyrido[4,3-b]indole (Trp-P-1), 3-amino-1-methyl-5H-pyrido[4,3-b]indole (Trp-P-2), 2-amino-9H-pyrido[2,3-b]indole (A alpha C) and 2-amino-3-methyl-9H-pyrido[2,3-b]indole (MeA alpha C), at levels of 0.02 approximately 13.5 ng per cigarette and sidestream smoke condensates of two brands of cigarettes contained these HCAs at levels of 0.14 approximately 2.72 ng per cigarette. PhIP was not detected in any sample of mainstream or sidestream smoke condensate.
The mutagenic activity and the mass amount of heterocyclic amines responsible for the mutagenic activity have been measured in some cooked foods. Cooked meats are the predominant source of mutagenic activity in the diet with values ranging from 0 to 10,000 revertants per gram reported in the Ames/Salmonella test with strain TA98. Several heterocyclic amines are present and have been quantified using solid-phase extraction followed by HPLC. Frying at higher temperatures and for longer times produces the greatest mutagenic response, and concomitantly, the largest amounts of heterocyclic amines. Most of the mutagenic activity in fried meat samples can be accounted for by MeIQx(2-amino-3,8-dimethylimidazo[4,5-b]quinoxaline), DiMeIQx (2-amino-3,4,8-dimethylimidazo [4,5-f]quinoxaline) and IQ (2-amino-3-methylimidazo [4,5-f]quinoline), although other heterocyclic amines are present and PhIP (2-amino-3-methyl-6-phenylimidazo[4,5-b]pyridine) mutagenic activity becomes significant at higher temperatures. Non-meat products such as baked breads can also form significant mutagenic activity, particularly when overcooked. Commercially prepared hamburgers made from meat substitutes such as tofu, wheat gluten or tempeh and fried at 210 degrees C have up to 10% of the mutagenic activity of a fried beef patty cooked under the same conditions. When detected, amounts of heterocyclic amines in fried beef patties range from a total of 0.35 ng/g for commercial beef hamburgers to 142 ng/g for a beef patty cooked over a barbecue. Dietary intake is expected to have a large range, from less than one microgram per day to over 50 micrograms per day based on current knowledge of known heterocyclic amine chemicals and heterocyclic amine-containing foods.
Cooking, heat processing, and pyrolysis of protein-rich foods induce the formation of structurally related heterocyclic aromatic amines that have been found to be mutagenic in bacteria, mammalian cells in culture and mice. All these compounds are potent mutagens and most are active below 1 ng/plate, in Ames/Salmonella tester strain TA1538 in the presence of S9 liver microsomal preparations from rat, mouse, or hamster. They are also potent in strains TA98, TA97, moderately active in TA1537, weakly active in TA100, and virtually inactive in TA1535 and TA102. Thus, they show powerful frameshift activity in reverting specific GC-rich sequences, but do not cause base substitution mutations or revert an AT-rich sequence. They are 100-fold less active in the uvrB+, repair-proficient strain TA1978, and in the case of 2-amino-3-methylimidazo [4,5-f] quinoline (IQ), cause insertions and large deletions not seen in TA1538.
Chronic hepatitis C virus (HCV) infection is implicated in the etiology and pathogenesis of a relatively small proportion of hepatocellular carcinomas (HCC) in black residents of southern Africa. The major risk factor for the tumor in this population is chronic hepatitis B virus (HBV) infection acquired early in life. Only 13.2% of black patients with HCC are currently infected with HCV (as judged by the presence of HCV RNA in serum) alone, and another 4.3% are infected with both HCV and HBV. Antibody to HCV can be detected with second generation assays in 19.5% of HCC patients. Fifty percent of the HCC patients have only current HBV infection and a further 35.9% show markers of past HBV infection (including 9.1% who also are currently infected with HCV). The prevalence of current HBV infection is significantly lower in HCV RNA-positive than in HCV RNA-negative patients, and this difference is independent of the sex, age, or geographical location of the patients. Only 5.4% of HCC patients are not infected with either HBV or HCV. Patients with HCV-associated HCC are appreciably older than those with tumors not associated with HCV and those with tumors associated with HBV, and the differences are independent of geographical location of the patients. Patients with HCV-associated HCC are more likely to be urban dwellers and less likely to be rural dwellers than are those with HBV-related tumors. Differences in sex distribution and alpha-fetoprotein levels just fail to reach statistical significance between patients with HCV-related and HBV-related HCC. The pathogenesis of HCV-related HCC in black Africans, as in all populations, is uncertain. A direct carcinogenic effect appears unlikely, and the available evidence favors an indirect effect mediated via chronic necroinflammatory hepatic disease. Because very few of our patients undergo surgery or have a necropsy performed, the relationship between HCV-associated HCC and cirrhosis in black Africans has not been analyzed.
