“What's interesting about Helicobacter is that everybody is focusing on its pathogenic role and not enough on its role as an indigenous organism,” Blaser said. He thinks that the microbes that live inside humans are not there accidentally—over the course of human history, they have adapted to humans and have made trade-offs to become obligate parasites. As a result, their evolution is inextricably linked to human evolution, as well as to human disease. “Our microbes are a part of human physiology… as much a part… as our liver, or our kidneys, or our heart,” Blaser commented. “They are a metabolic compartment or a series of metabolic compartments in the human body and how they behave and… their characteristics are relevant to health and disease.”

The link between chronic diseases and infectious agents dates as far back as the early 1900s, when F. Peyton Rous showed that viruses could cause cancer. Since then, many more researchers have tried to confirm similar links, despite difficulties using Koch's postulates to prove causation (see sidebar) and opposition from some members of the scientific community. Now, “the scientific arguments tend to be around ‘which' infectious agent and ‘how', rather than ‘whether',” said Ian Frazer, Director of the Centre for Immunology and Cancer Research at the University of Queensland, Australia. Efforts to determine ‘how' are now focusing on the distinction between acute infection and persistent colonization, and the long-term balance between man and microbe.

Evidence points to H. pylori being an ancient and persistent guest in the human gastrointestinal system, rather than an acute infection. But in the past few decades, the prevalence of H. pylori has declined, offering an unprecedented opportunity to track the demise of a member of the normal human flora. In developed countries, the bacteria now affects 20–50% of adults, but studies show that there has been a 50% decline in the prevalence of H. pylori in the USA since 1968 (www.helico.com). At least three factors have a role in its disappearance: cleaner water, less crowding and better hygiene, and the widespread use of antibiotics over the past 60 years.

The true nature of the bacteria's relationship with humans is only beginning to emerge, and reveals a complex mixture of pathogenicity and protection. “I don't use the term infection for Helicobacter anymore, I only use the term colonization,” Blaser said. “I think it's normal flora just like Escherichia coli or Bacteroides.” In addition to H. pylori, the human body houses hundreds of strains of commensal bacteria and viruses, which may persist for months, years or even a lifetime. For the most part, these microbes cohabit happily with their host, and rarely cause problems—indeed, their presence is often seen as beneficial rather than detrimental.

If it is not the mere presence of a microorganism that triggers disease, then other factors must be at play. Indeed, the biology of a short-term infection differs markedly from that of a persistent colonization, even for the same infectious agent. “Clearly, persistent infection is a dynamic state and not one of quiescent bacteria held in check by the force of the host immune system,” wrote Stanley Falkow, Professor of Microbiology and Immunology at Stanford University School of Medicine (CA, USA; Falkow, 2006). Thus, “[t]he outcome of confrontation between a human body and a pathogenic microorganism is determined by what each of the players brings to the fight,” according to a recent report on ‘Microbial Triggers of Chronic Human Illness' from the American Association of Microbiology (AAM; Carbone et al, 2005).

Although H. pylori may be the most well known, many other bacteria and viruses have been linked—either definitively or loosely—to chronic conditions (Table 1). Human papillomavirus (HPV) has an important role in cervical cancer, as do hepatitis B and C in hepatic cancer, Epstein–Barr virus in lymphoma and schistosomes in bladder cancer. Infectious agents have also been suggested to contribute to leukaemia, Crohn's disease, arthritis and rheumatic fever. And for decades, scientists have attempted to link chronic infection with cardiovascular disease, with mixed success.

However, despite the growing evidence, the role of infectious agents in a wide range of chronic diseases is still contested. Ruth Itzhaki, a professor in the Molecular Neurobiology Laboratory at the University of Manchester, UK, faced an uphill battle against some members of the Alzheimer's disease research community with her work on the link between herpes simplex virus type 1 (HSV1) and the disease. “All they are interested in is amyloid,” she said. “When we first showed the presence of viral DNA in elderly brains, we met huge incredulity. Since then about seven other groups have broadly substantiated this, and the results have generally been quietly accepted.”

More tenuous links have been suggested between microbial agents and conditions, such as obesity and diabetes. Human adenoviruses increase adiposity in humans (Atkinson et al, 2005) and other animal models (Whigham et al, 2006; So et al, 2005). Although Blaser noted that a link between H. pylori and obesity and diabetes in humans is “preliminary and not supported”, he pointed out that the regulation of hormones involved in energy homeostasis changes in the presence or absence of the bacteria. “Helicobacter status affects the status of hormone production in the stomach.”

But to conclude that all disease-causing agents are bad is certainly too simple. For H. pylori colonization, for example, “there's a biological cost to carrying it—some people will get ulcers, and some people will get cancer,” Blaser said. “But there may be some biological benefits.” In parts of the world where Helicobacter incidence is decreasing, ulcer disease and gastric cancer are also declining in incidence, as is expected. However, the number of cases of esophageal cancer, specifically adenocarcinoma of the esophagus, is rising. “The data indicate that it's the most rapidly rising cancer in developed countries like US, UK, Norway, and Australia, where this has been studied,” Blaser said.

The question then is whether a link exists between these cancers and the decrease in H. pylori incidence. Paradoxically, over the course of a lifetime, H. pylori decreases gastric acidity. “You'd think there should be a lot of acid because of ulcer disease but ulcer disease is not well understood,” Blaser explained. Essentially, the inflammatory process caused by long-term H. pylori colonization reduces gastric function in terms of acid production. As people age, and as their stomachs age, they produce less acid—part of the pathogenesis of gastric cancer. For an individual without H. pylori, the stomach pumps out full levels of acid for decades, thus affecting their unprotected esophagus, “which, I would argue, has never in evolutionary history been in this situation before.”

Other factors—not yet fully understood—undoubtedly play a part. Not everyone who has Helicobacter will develop gastric cancer, and not everyone without the bacteria will develop esophageal cancer. “These are not all or none relationships,” Blaser explained. “It's about trends.” When the incidence of a disease rises so rapidly in such a short space of time, evolutionary factors cannot be the cause—environmental factors must be to blame. And “the place to begin looking is in our human environment, our indigenous microbes. That's the smoking gun,” Blaser said.

Part of the reason why H. pylori receives such attention is because Marshall and Warren's research led to a cure: antibiotics are now routinely used to cure gastric ulcers. Similar excitement surrounded the development of a prophylactic vaccine against HPV infection, based on research by Frazer and colleagues at the University of Queensland. Gardasil™ (Merck & Co; Whitehouse Station, NJ, USA) and Cervarix™ (GlaxoSmithKline; Uxbridge, UK) are 100% effective in preventing HPV16 and HPV18 infections. Frazer and his colleagues are now working on therapeutic vaccines to eliminate existing HPV infections. “We hope that the answers obtained from that work might be more generically applicable to the field of therapeutic vaccines,” he said.

But if other infectious agents share H. pylori's two-sided nature, will such eradication attempts cause more trouble than they are worth? The difficulties in balancing the risks and benefits “should not be a hindrance to developing vaccines and antimicrobials”, according to G. Singh Chhatwal, who studies the role of streptococci in rheumatic heart disease and other diseases, at the Department of Microbial Pathogenesis and Vaccine Research at the German Research Centre for Biotechnology in Braunschweig. Frazer, conversely, sees no danger in eradicating HPV to decrease the risk of cervical cancer. “High risk HPVs, such as HPV16, are all bad news, wherever they are in the body,” he said. “No epidemiological evidence suggests that their elimination would lead to any new problems.” In fact, most of the HPV types that humans carry are benign and successful ectoparasites—they cause no disease. HPV16 is what Frazer called a “successful failure”—successful because it is so widespread, but a failure “because it causes enough problems for the 1% of people who get cancer attributable to it that we will go out of our way to eliminate it by developing a vaccine!” There are other, less successful failures in the HPV family that are not as widespread but cause cancer, and they too will be targeted. “We'll then be left with the truly successful HPVs that cause no disease,” Frazer said.

The high-profile cases of gastric ulcers and cervical cancer may offer hope that more chronic diseases could be cured, but this assumption still largely ignores the human role in the interaction between host and microbe. “Unless we know that, then the story's not complete,” Chhatwal said. “It's important to understand these problems in terms of risk relationships,” Blaser said. “In my mind, Helicobacter doesn't cause ulcers [and] gastric cancer, it increases the risk for ulcers [and] gastric cancer—just as smoking doesn't cause lung cancer, it increases risk.” Frazer agreed: “Behavioral risks for cancer are clearly additive to the infectious risks.” Itzhaki's research highlights the importance of genetic factors: she and her colleagues found that HSV1 is a strong risk factor for Alzheimers' in carriers of the type 4 allele of the apolipoprotein E gene (Itzhaki et al, 1997; Itzhaki & Wozniak, 2006).

As eradication programmes are based on cost–benefit analyses—both financial and biological costs—their implementation will probably differ depending on the disease. Speaking in a personal capacity, Kathryn Carbone, one of the authors of the AAM report, said that, although each medical decision balances advantages and disadvantages, “as facts come to light, they contribute to changing the risk–benefit of any medical decision.” What makes these decisions even more difficult is that the disadvantages—such as the appearance of an unexpected disease—can take years or even decades to reveal themselves.