INFECTION INTO DISEASE RENÉ J. DUBOS, Ph.D., ScD. {hon.), M.D. {hon.)* In 1886, E. L. Trudeau carried out an experiment to test the effect of living conditions on susceptibility to tuberculosis. Here is the account as given in his autobiography (1): ... I naturally began to wonder, iftuberculosis was a germ disease and the germs had already gained access to the body, how a change ofclimate, rest, fresh air and food could influence the disease. In seeking an experimental answer to this question I decided on the following experiment: Lot i, offive rabbits, were inoculated . . . andput under the best surroundings oflight, food and air attainable. Lot 2, offive rabbits, inoculated at the same time and in the same way, were put under the worst conditions of environment I could devise: and Lot 3, offive rabbits, were put under similar bad conditions without being inoculated. Lot i, I turned loose on a little island in front ofmycamp at Paul Smith's, where they ran wild all summer in thefresh air and sunshine, and were provided with abundant food. Lot 2 and Lot 3 were put in a dark, damp place where the air was bad, confined in a small box and fed insufficiently. The results showed that of the rabbits allowed to run wild under good conditions, all, with one exception, recovered. OfLot 2, the same as Lot I, but put in unfavorable surroundings, four rabbits died within three months and the organs showed extensive tuberculosis. Lot 3, uninoculated animals, were then killed and, though emaciated, they showed no tuberculous disease. This showed me conclusively that bad surroundings ofthemselves could not produce tuberculosis, and that when once the germs had gained access to the body the course of the disease was greatly influenced by a favorable or an unfavorable environment. To my knowledge, no one has attempted to repeat Trudeau's experiment . This is regrettable because the findings, if valid, would provide a useful model for the study of environmental and host factors in tuberculosis —-and even more because they highlight one ofthe most neglected aspects ofthe germ theory ofdisease, namely, the fact that infection rarely produces fatal disease under natural circumstances. * The Rockefeller Institute for Medical Research, Sixty-sixth Street and York Avenue, New York 21, New York. 425 Problems ofinfection have been studied largely through laboratory experiments so designed that the pathogen causes death or at least severe disturbance in most ofthe infected individuals—plants or animals. These experimental models do correspond to real situations in nature. The infection of man by Bacillus pestis, of the rabbit (Oryctolagus) in Australia and in Europe by the myxoma virus, and ofthe American chestnut tree by the fungus Endothia parasitica are examples of naturally occurring infections with a tremendous mortality. But, while destructive epidemics ofthis sort have caused some ofthe cataclysms ofhistory and continue to be important , they certainly do not represent the most usual or the most significant aspect of microbial disease in our communities. Much infection in the Western world at the present time is from parasites ubiquitous in the environment and occurring normally in a very large percentage ofhealthy individuals, which cause pathological disorders only when natural resistance to them has been undermined by the stresses and strains of life. Trudeau's rabbits were all infected with tubercle bacilli (as are most human adults), but the tuberculous infection evolved into fatal disease only in the animals placed in unfavorable living conditions (as is so often the case with man). One mechanism through which the physiological state ofthe host can influence the course ofthe infectious process is by affecting the susceptibility oftissue cells to the constituents or products ofthe infective agent. It is certain that the presence of pathogens in the tissues—even in very large numbers—does not necessarily result in gross pathological changes. Trypanosoma lewisi in rats and the virus of aster yellows in mosquitoes are examples ofhost-parasite relationships in which the infective agents multiply extensively in vivo without damaging appreciably, ifat all, the health of the infected hosts. In tissue cultures likewise, extensive multiplication of tubercle bacilli or ofcertain viruses has been observed intracellularly without concomitant cytopathogenic effects. Experiments have revealed, on the other hand...