Many infectious agents have been posited to cause cancer. The IARC biannually reviews epidemiologic and basic science information on putative cancer-causing agents. It then classifies these agents as definite, probable, possible, or unlikely causes of cancer. To date, they have carried out extensive evaluations of liver flukes, schistosomes, H. pylori, hepatitis viruses, human herpesvirus 8 (HHV-8), HPV, HTLV, human immunodeficiency virus (HIV), and Epstein-Barr virus (EBV; Table 12-1).
Table 12-1 Infectious Agents Purported to Cause Malignancy
IARC Classification* Attributable Fraction t
Viruses Hepatitis B Hepatitis C
Papillomaviruses 16 and 18 Papillomaviruses 31 and 33 Papillomaviruses 2 and other papillomaviruses HTLV-1
HHV-8 HIV-1 HIV-2 Bacteria Helicobacter pylori
Mycobacterium ulcerans Salmonella typhi Chronic or recurrent UTIs Chronic osteomyelitis or wound infections Parasites Opisthorchis viverrini Clonorchis sinensis Schistosoma haematobium Schistosoma japonicum
Hepatocellular carcinoma Group 1 Hepatocellular carcinoma and MALT lymphoma Group 1
Cervical, anogenital cancers Group 1
Cervical, anogenital cancers Group 2A
Cervical, anogenital cancers Group 2B
Adult T-cell leukemia Group 1
Burkitt's lymphoma Group 1
Non-Hodgkin's lymphoma Group 1
Hodgkin's disease Group 1
Nasopharyngeal carcinoma Group 1
Gastric adenocarcinoma NS
Kaposi's sarcoma Group 2A
ID related cancers Group 1
ID related cancers Group 2A
Gastric adenocarcinoma Group 1 Gastric lymphoma
Immunoproliferative small intestinal NS
Squamous cell carcinoma (skin) NS
Bladder cancer NS
Squamous cell carcinoma NS
Cholangiocarcinoma Group 1
Cholangiocarcinoma Group 2A
Bladder cancer Group 1
Hepatocellular and rectal cancers Group 2B
90%-98% for all viruses
*From "Overall Evaluations of Carcinogenicity to Humans" as evaluated in IARC Monographs Volumes 1-82. Web site: 188.8.131.52/monoeval/
crthall.html. Last updated January 9, 2004, consulted April 7, 2004. ^From references 14 and 15.
Other infectious agents that have not yet been evaluated by the IARC, but are suspected of causing cancer, are also listed in Table 12-1. Because the great majority of these pathogens are reviewed elsewhere in this book, we will not discuss their life cycles, epidemiology, or clinical manifestations. Instead, we focus here on the general features and control of the principal infection-induced malignancies.
A simplified scheme of infection-related oncogenesis can be used to categorize the malignancies into one of two types: either tumors caused by the integration of oncogenic DNA into the host cell or those induced by chronic inflammation. Basically, all types of oncogenic agents enhance cell growth. The mechanisms by which this occurs can be quite variable, however, ranging from expression of viral oncogenes for cell growth factors to stimulation of growth by inflammatory intermediaries. Agents with oncogenes, however, directly immortalize cells, while inflammatory carcinogenic agents do not. Moreover, it is thought that immunosuppression may be important for development of malignancies related to oncogene integration. This is not thought to be the case for inflammation-related malignancy; in fact, immunosuppression, by lessening the inflammatory response, could potentially protect against inflammation-related cancers. This was highlighted in a recent review by Smukler and Ratner who state "the observed lack of a significant increase in the incidence of hepatocellular carcinoma [in HIV infected patients] may be caused, in part, by a possible reduction in HBV-induced hepatic damage in HIV coinfected people. Immune impairment in HIV-positive patients leads to higher HBV replication levels with less severe liver damage because of a blunted HBV-specific immune response and less consequent cirrhosis."16 A disproportionate lack of necrotizing inflammation in HIV-HBV coinfected subjects supports this possibility.17 Finally, cancers related to inflammation appear to occur only after a long incubation period. There is no such temporal relationship between oncogenic viruses and cancer. This latter point is most readily demonstrated by the short incubation period between EBV infection and the development of Burkitt's lymphoma in young children in sub-Saharan Africa.
The importance of inflammation in inducing cancer was first exemplified by two diseases not uncommon to the tropics: chronic, draining osteomyelitis and Buruli ulcers (caused by Mycobacterium ulcerans). These chronic, inflammatory conditions, if left untreated, cause aggressive squamous-cell carci-nomas.18 In each instance, carcinogenesis appears to depend not on specific oncogenic genes or gene products produced by the infectious agent, but rather on the organism's ability to persist despite the host's efforts to combat its presence. Inflammation is thought to induce cancer by stimulating production of reactive oxygen and nitrogen species that damage DNA, proteins, and membranes. Inflammation may also increase the prolifer-ative rate of infected tissues, promoting development of tumors. Proliferation is a promutagenic effect, permitting more cells to be susceptible to DNA damage and mutation. It also fosters selective growth of mutant clones.19-23 We now recognize that many infections, including H. pylori, S. haematobium, O. viverrini, and HBV, are at least in part linked to cancer by their ability to foster an inflammatory response. In support of this, phenotypes of H. pylori that cause more inflammation are more closely related to malignancy.24 Treatment of S. haematobium and O. viverrini leads to marked decreases in inflammatory cells within the bladder and liver, respectively, and associated declines in chromosomal damage.
As stated previously, the risk of developing inflammation-related cancer also appears to be intimately tied to the duration of the inflammatory process. This has been particularly well established for HBV infection but appears also to be the case for H. pylori and probably other agents as well.25,26 The chronicity of infection implies that the carcinogenic effects of infection are likely to be promotional rather than resulting from a single initiating event.
Oncogenic viruses—such as EBV, HTLV-1, HHV-8, and human papillomaviruses—may lead more directly to cancer by integration of viral oncogenes or oncogene promoters into nuclear DNA. Resultant viral oncoproteins may for instance interact with and inactivate host tumor suppressor proteins such as p53, uncoupling normal growth control processes and leading to cellular transformation.27,28 Oncogenic viruses may also induce production of cell growth factors, stimulating the host cell's reproduction, and accordingly their own. Other viruses, such as HIV, are not direct carcinogens but, by limiting the host's immune response, foster carcinogenic effects of normally more innocuous agents. For example, HIV-related cellular immunodeficiency fosters carcinogenesis by chronic colonizers such as EBV and permits infection with HHV-8, an unusual agent in immunocompetent hosts, with resultant Kaposi's sarcoma (KS). KS in persons with HIV infection occurs at a rate thousands of times higher than in the non-HIV population.29 In HIV-infected patients, lymphomas also occur at much higher rates than in the general population. Although the specific cause remains to be determined, much evidence points to the activity of lym-photropic viruses, particularly EBV. Whereas both EBV and HTLV-1 have long been implicated in the development of lymphoid malignancies, additional viral and bacterial agents are now thought to have lymphoproliferative activity. These include HHV-8, hepatitis C virus (HCV), H. pylori, Borrelia burgdorferi, and possibly Campylobacter jejuni.30-33
The etiology of infection-induced cancer is made more complex by multifactorial causality of certain tumors on the one hand, and multiple outcomes following certain infections on the other. The importance of a specific agent in causing malignancy (its "attributable risk") depends on its prevalence. For example, in high-incidence areas of hepatitis, primary hepatocellular carcinoma (HCC) is commonly associated with chronic infection with HBV or HCV Shin and associates34 found strong evidence that HBV and HCV were independent risk factors for HCC in Korea, where liver cancer mortality is reported to be the highest in the world. In areas where hepatitis viruses are less prevalent, they will be a less common factor in HCC. In Sweden, for example, Kaczynski and coworkers tested 113 HCC patients for serologic or immuno-cytochemical evidence of such infection. They found no signs of chronic HBV infection, and HCV reactivity occurred in only 7 of 64 patients.35 Thus, chronic viral hepatitis appeared to play a minor role in induction of HCC in Sweden, where alcohol-related cirrhosis is a much more common risk factor.
Certain infections may be important, but not sufficient, for the induction of cancer. Just as most smokers do not get lung cancer, most infections with potentially carcinogenic pathogens do not proceed to a malignancy. For example, the large and diverse group of HPVs is implicated in causation of a variety of proliferative conditions, some of which (warts, epithelial cysts, intraepithelial neoplasias, anogenital, orola-ryngeal and oropharyngeal papillomas, keratoacanthomas, and other types of hyperkeratoses) are relatively benign. At least 90% of cancers of the cervix and more than 50% of other anogenital cancers are attributed to certain high-risk HPV types that may act as solitary carcinogens or in concert with cofactors. The progression of HPV-associated squamous epithelial lesions to cervical cancer may be enhanced by certain concurrent sexually transmitted agents, such as herpes simplex virus or chlamydia, and is clearly associated with HIV-induced immunosuppression.15 Aflatoxin B1 is widely considered to be a cocarcinogen with HBV; although either is capable of inducing HCC, the combination appears to behave synergistically in regions where both agents are endemic.36 Although 90% of the world's population is thought to be infected with EBV, the two tumors classically linked to EBV occur in limited geographic areas. Burkitt's lymphoma is found in hyperendemic malarial areas of Africa and New Guinea, and nasopharyngeal carcinoma in parts of Asia, suggesting the presence of specific, geographically limited, cofactors.37,38
Was this article helpful?