2009 : WHAT WILL CHANGE EVERYTHING?

paul_w_ewald's picture
Professor of Biology, Amherst College; Author, Plague Time
SCANNING FOR THE NONHUMAN IN HUMANS

Thirteen decades ago, Louis Pasteur and Robert Koch led an intellectual revolution referred to as the germ theory of disease, which proposes that many common ailments are caused by microbes. Since then the accepted spectrum of infectious causation has been increasingly steadily and dramatically.  The diseases that are most obviously caused by infection were accepted as such by the end of the 19th century; almost all of them were acute diseases.  Acute diseases with a transmission twist, mosquito-borne malaria for example, were accepted a bit later at the beginning of the 20th century.  Since the early 20th century, the spectrum has been broadened mostly by recognition of infectious causation of chronic diseases.  The first of these had distinctly infectious acute phases, which made infectious causation of the chronic disease more obvious. Infectious causation of shingles, for example, was made more apparent by its association with chicken pox.  Over the past thirty years, the spectrum of infectious causation has been broadened mostly through inclusion of chronic diseases without obvious acute phases.  With years or even decades between the onset of infection and the onset of such diseases, demonstration of infectious causation is difficult. 

Technological advances have been critical to resolving the ambiguities associated with infectious causation of such cryptic infectious causation.  In the early 1990s Kaposi's Sarcoma Associated Herpes Virus was discovered using a molecular technique that stripped away the human genetic material from Kaposi's sarcoma cells and found what remained.  A similar approach revealed Hepatitis C virus in blood transfusions.  In these cases there were strong epidemiological signs that an infectious agent was present.  When the cause was discovered, acceptance did not have to confront the barrier of entrenched opinions favoring other non-infectious causes.  If such special interests are present the evidence has to be proportionately more compelling.  Such is the case for schizophrenia, atherosclerosis, Alzheimer's disease, breast cancer, and many other chronic diseases, which are now the focus of vehement disagreements.

Advances in molecular/bioinformatic technology are poised to help resolve these controversies.  This potential is illustrated by two discoveries, which seem cutting edge now, but will soon be considered primitive first steps.  About a decade ago, one member of Stanford team scraped spots on two teeth of another team member, and amplified the DNA from the scrapings.  They found sequences that were sufficiently unique to represent more than 30 new species.  This finding hinted at the magnitude of the challenge--tens or perhaps even hundreds of thousands of viruses and bacteria may need to be considered to evaluate hypotheses of infectious causation. 

The second discovery provides a glimpse of how this challenge may be addressed.  Samples from prostate tumors were tested on a micro-array that contained 20,000 DNA snippets from all known viruses.  The results documented a significant association with an obscure retrovirus related to one that normally infects mice. If this virus is a cause of prostate cancer, it causes only a small portion that occurs in men with a particular genetic background. Other viruses have been associated with prostate cancer in patients without this genetic background. So, not only may thousands of viruses need to be tested to find one correlated with a chronic disease, but even then it may be one of perhaps many different infectious causes.

The problems of multiple pathogens and ingrained predispositions are now coming to a head in research on breast cancer.   Presently, three viruses have been associated with breast cancer: mouse mammary tumor virus, Epstein Barr virus, and human papillomavirus.  Researchers are still arguing about whether these correlations reflect causation.  If they do, these viruses account for somewhere between half and about 95% of breast cancer, depending on the extent to which they act synergistically. Undoubtedly array technology will soon be used to assess this possibility and to identify other viruses that may be associated with breast cancers.

There is a caveat. These technological advancements provide sophisticated approaches to identifying correlations between pathogens and disease.  They do not bridge the gulf between correlation and causation.  One might hope that with enough research all aspects of the pathological process could be understood, from the molecular level up to the whole patient.  But as one moves from molecular to the macro levels, the precision of interpretation becomes confounded by the complex web of interactions that intervene, especially in chronic diseases.   Animal models are generally inadequate for chronic human diseases because the disease in animals is almost never quite the same as the human disease. The only way out of this conundrum, I think, will be to complement the technological advancements in identifying candidate pathogens with clever clinical trials.  These clinical trials will need to use special states, such as temporary immune suppression, to identify those infections that are exacerbated concurrently with exacerbations of the chronic disease in question.  Such correlations will then need to be tested for causation by treatment of the exacerbated infection to determine whether the suppression of the infection is associated with amelioration of the disease.  

Why will this process change things?   For those of us who live in prosperous countries, infectious causes are implicated but not accepted in most common lethal diseases: cancers, heart attacks, stroke, Alzheimer's disease.  Infectious causes are also implicated in the vast majority of nonlethal, incapacitating illness of uncertain cause, such as arthritis, fibromyalgia, and Crohn's disease.  If infectious causes of these diseases are identified medical history tells us that they will tend to be resolved. 

A reasonable estimate of the net effect would be a rise in healthy life expectancy by two or three decades, pushing lifespan up against the ultimate boundary of longevity molded by natural selection, probably an age range between 95 and 105 years.  Being pushed up against this barrier people could be expected to live healthy lives into their 90s and then go downhill quickly.  This demographic transition toward healthy survival will improve productivity, lower medial costs, and enhance quality of life.  In short it will be one of medicine's greatest contributions.