a Darwinian selective process of "survival of the fittest." Without competition for the hosts' tissues, these bacilli then became the dominant subspecies.

Fortunately, two other medications were discovered shortly thereafter— p-aminosalicylic acid and isonicotinic acid hydrazide (isoniazid). Clinicians soon recognized that if all these drugs were given simultaneously, drug resistance did not emerge and lifetime cures of tuberculosis finally were achievable. Subsequent research showed that the explanation for this was as follows: (i) Random bacterial mutations that conferred resistance to individual drugs occurred infrequently during microbial replication, approximately once in 105–108 (4). (ii) These mutations were unlinked; therefore, the probability of a microbe spontaneously developing resistance to two drugs was the product of the individual risks or 1 in 105 × 1 in 106 = 1 in 1011 (3). Because the number of bacilli in a patient, even with extensive disease, rarely exceeds 109, it was highly improbable that multiresistant mutants would occur spontaneously (4). Thus, when isoniazid and streptomycin were given together, the isoniazid killed the mutants resistant to streptomycin and vice versa, ultimately eliminating the bacteria from the body.

In the 1950s and 1960s tuberculosis specialty hospitals (sanatoria) and clinics were widely available throughout the industrialized nations. Based on public fear of the disease and aggressive professional programs, successful treatment—despite the need for 24 months of drug therapy—was accomplished in the great majority of cases. However, as these two elements lost intensity and social disruptions became more pervasive in our society, adherence to treatment plans was eroded. Clinicians and public health authorities were hopeful that with newer, more powerful drugs the duration of treatment could be reduced sufficiently to combat noncompliance. But, despite reducing the required time from 24 to 6 months, irregular or incomplete adherence rose steadily over the past two decades (5).

As a consequence, the prevalence of drug-resistant strains of M. tuberculosis has risen dramatically in certain regions or populations. At the dawn of the treatment era, roughly 1–2% of strains of M. tuberculosis were seen to have significant drug resistance, almost universally to only one drug (6); in the 1960s and 1970s, that rate in the U.S. hovered around 3–5% (7, 8). However, over the past decade the national rate has risen steadily (9). In New York City, where a variety of elements, including poverty, substance abuse, and deteriorating public health programs, combined to confound tuberculosis control, 33% of tuberculosis strains recovered in April 1992 were resistant to at least one drug, and 19% were resistant to two or more agents (10). Tragically, in some developing nations where resources



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement