The recent surge of publicity denying HIV's causal role in AIDS begs an examination of the relationship between a specific microorganism and a specific disease. The question is this: What scientific proof is required to establish cause and effect?
HIV and AIDS-Related Diseases
AIDS is defined by the Centers for Disease Control and Prevention (CDC) as the presence of a positive HIV antibody test and one or more of the illnesses known as opportunistic infections (OIs) or a CD4 cell count below 200 cells/mm3. AIDS wasting syndrome, which technically is not considered an OI, and three cancers -- non-Hodgkin's lymphoma, Kaposi's sarcoma (KS), and cervical carcinoma -- are AIDS-defining in the presence of a positive HIV antibody test. HIV and AIDS have been continuously linked in time, place, and certain population groups (e.g., children born to HIV-infected mothers) since 1984, the year HIV was discovered. Besides millions of cases of AIDS, some people receiving cancer chemotherapy develop AIDS-like illnesses. And in people receiving organ transplants, the immune system is deliberately suppressed and AIDS-like illnesses can result. Neither of these two examples, however, results in an AIDS diagnosis. Moreover, those illnesses resolve when the causative medications are stopped.
Recent improvements in laboratory testing, particularly polymerase chain reaction (PCR) testing, allow detection of HIV genetic material in people with AIDS or a positive HIV antibody test result. Interestingly, a 1995 study of 230,000 people with AIDS found that 168 were HIV-seronegative. Today it is known that the immune system depletion characteristic of advanced HIV disease can suppress the production of antibodies to HIV, which may explain this earlier finding.
The vast majority of people with AIDS in resource-poor countries have not had an HIV antibody test; even in developed countries, not all people with AIDS have had viral load tests. Those who have had viral load tests showing nondetectable virus generally have no additional tests to confirm the presence of HIV genetic material.
However, the above facts, while strongly suggesting a relationship between HIV and AIDS, do not by themselves establish a causal relationship.
The German scientist Robert Koch (1843-1910) established four requirements, or postulates, that must be demonstrated to prove that a specific microorganism causes a particular disease:
The microorganism must be found in all cases of the disease.
It must be possible to isolate the microorganism from the host and grow it in pure culture (in the laboratory).
The microorganism must reproduce the original disease when introduced into an experimental animal.
The microorganism must be recoverable from that animal.
The scientific research community widely accepts that fulfillment of these postulates demonstrates causality. Even scientists such as Peter Duesberg, Ph.D., who says he believes that HIV does not cause AIDS, accept that the postulates put forth by Robert Koch are sufficient to establish causality. While in 1987 Dr. Duesberg noted that HIV did not fit all of Koch's criteria, developments in HIV/AIDS research in the 13 years since that statement have shown that HIV does indeed fulfill Koch's postulates as the cause of AIDS.
Conclusive Data from Reputable Scientific Research
Regarding postulate one, PCR testing allows researchers to document cell-associated proviral HIV in persons with AIDS who have been tested (proviral DNA detection is a research test, not one of the common FDA-approved viral load tests). Prior to this technology, HIV was often difficult to find. In addition, combining PCR testing with the common viral load tests has documented the presence of HIV genes as RNA freely floating in the blood plasma, outside of cells, in persons with a positive antibody test not taking anti-HIV medication. (Viral load testing looks for virus; the ELISA and Western blot tests look for antibodies to HIV.)
Regarding postulate two, improvements in laboratory culture techniques have allowed the growth of HIV in vitro (in laboratory models) from blood samples obtained from persons with AIDS who have undergone such testing and from almost all persons with a positive antibody test without AIDS who have undergone such testing.
The last two postulates stipulate that inoculating the organism into an animal model (i.e., exposing or infecting the animal) leads to the same disease and that the organism is recoverable from that animal. The evidence satisfying these postulates was established in 1997, when Francis J. Novembre, Ph.D., and colleagues from Emory University in Atlanta, GA, published in the Journal of Virology that a chimpanzee inoculated with HIV ten years earlier had developed an AIDS-defining OI. Prior to the OI, the HIV RNA viral load had increased (partially documenting recovery of the organism from the animal model) and the CD4 cell count had decreased in the chimpanzee. Cultures of blood from the animal also were positive for HIV, establishing recoverability of the organism. Subsequently, blood from that chimp was transfused into a second, healthy chimpanzee. This second chimpanzee later had an increase in the HIV viral load and a decrease in the CD4 cell count.
Prior to this 1997 report, fulfillment of Koch's third and fourth postulates was lacking. Interestingly, the incubation period for clinical AIDS in this chimpanzee, with whom humans share 98% gene homology (structural similarity), was essentially equivalent to the average incubation period in humans -- ten years. This finding and publication were reported in the September 1997 issue of BETA. While evidence from one chimpanzee may not seem compelling to the lay person, in the scientific arena and in conjunction with other, cumulative data, it is considered persuasive. A good source of information on this topic is www.niaid.nih.gov.
Ancillary and Epidemiologic Supporting Evidence
Several reports document the transmission of the organism to a human host that reproduces the original disease, and the subsequent recovery of the microorganism in that second person. For example, at least three laboratory workers developed AIDS after accidental exposure to concentrated HIV in the laboratory. All three developed immunosuppression and related opportunistic diseases, including Pneumocystis carinii pneumonia (PCP), following infection. In all three cases, HIV was isolated, sequenced, and shown to be the infecting strain of the virus.
In addition, the development of AIDS following known HIV seroconversion has been repeatedly observed in widely diverse populations, including each of the following:
pediatric and adult blood transfusion cases
hemophiliacs who received infected blood clotting factor protein infusions
monogamous sexual partners of those transfusion recipients
health-care workers with accidental needlestick or other occupational exposure, similar to the lab technicians described above
male-to-male and male-to-female sexual transmission
injection drug users with secondary sexual transmission
extremely rare outbreaks including transmission to dental patients from an infected dentist (e.g., the Kimberly Bergalis case in Florida)
In conclusion, although the specific molecular mechanisms of HIV's causative role in AIDS are not yet completely understood, Koch's postulates have been fulfilled, thus establishing causality.
Tim Teeter is Associate Director of Treatment Support and Publications at the San Francisco AIDS Foundation.
Related BETA Articles: "Special Report on HIV & AIDS"; "HIV is the Only Cause of AIDS: The Potential for Journalism to Impact the Public Health"
Burnett, M. and White, D.O. Natural History of Infectious Disease. Cambridge University Press, Cambridge. 1972.
Cohen, J. Fulfilling Koch's postulates. Science 2666(5191): 1647. December 1994.
National Institute of Allergy and Infectious Diseases (NIAID) Fact Sheet. The evidence that HIV causes AIDS. July 1995.
Novembre, F.J. and others. Development of AIDS in a chimpanzee infected with human immunodeficiency virus type 1. Journal of Virology 71(5): 4086-4102. May 1997.
Weiss, R.A. and Jaffee, H.W. Duesberg, HIV, and AIDS. Nature 345(6277): 659. June 1990.
Back to the SFAF BETA Spring, 2000 contents page.