Imperative that strategies to control tuberculosis include measures to treat individuals with MDR disease

Sep 20 2004

Challenging the notion that multidrug-resistant tuberculosis (MDR-TB) poses a low threat to populations, epidemiologists at the Harvard School of Public Health (HSPH) have constructed a mathematical model of the disease variant's potential course, projecting the likelihood of an eventual epidemic if no concerted action is taken soon to corral drug-resistant strains.

The model description appears in the October 2004 issue of Nature Medicine.

Most mutations that lead to drug resistance in TB strains make them less transmissible than drug-sensitive organisms. But the study authors argue that even if only a small fraction of mutated strains are transmissible -- as has been found in some recent laboratory and medical studies of TB-afflicted patients -- these mutated strains will eventually overwhelm and dominate drug-sensitive strains in an epidemic. Current TB containment strategies in many countries focus mainly on identifying and treating those patients with drug-sensitive disease.

TB is a major cause of infectious disease mortality throughout the world. The World Health Organization (WHO) has estimated that there are 8 million new cases of the disease and 2 million deaths per year. The airborne disease is spread through the respiratory route, and all population groups are at risk. After almost a century of decline in Europe and America, TB is increasing in Eastern Europe, Asia and among population groups whose immunity is already weakened by HIV/AIDS. TB strains resistant to a multiplicity of antibiotic drugs (MDR-TB) have been seen in every region evaluated recently by WHO, with "hot spots" in Russia, Eastern Europe, South Africa, China and Israel.

Current strategies to control TB include the implementation of what is called "directly observed treatment, short course" or DOTS, which focuses on case detection and treatment of drug-sensitive TB using a standardized regimen of antibiotics. Until recently, WHO and many national TB programs did not endorse programs designed to treat people with MDR-TB; these programs and their treatment regimens were considered to be expensive (costing more than 100 times standard care) and to have inadequate cure rates. In addition, many argued that MDR strains of TB were unlikely to be highly transmissible since the mutations that led to drug resistance may have damaged the organism, making it no longer capable of being transmitted from one person to another. Since multi-drug resistance originally arises through poor adherence by individuals to standard therapy, it was argued, limited resources should be directed toward strengthening DOTS rather than treating MDR-TB.

These arguments derive from laboratory and epidemiological data that suggested that some MDR strains were not as easily transmitted as drug-sensitive strains. Recent work, however, has shown that drug resistance arises from a wide range of different mutations in different genes in the M. tuberculosis organism. This work suggests that while some mutations do impair the ability of the organism to be transmitted from person to person, others have little impact on the organism. The authors incorporated this recent work into their mathematical model.

First author Ted Cohen, a postdoctoral fellow at HSPH, and co-author Megan Murray, assistant professor of epidemiology at HSPH, examined the long term effect of focusing TB programs on drug-sensitive strains under the assumptions that MDR strains develop from incomplete or interrupted therapy; that most MDR strains will have reduced transmissibility and that -- most importantly --- a small fraction of MDR strains will be only slightly less transmissible than drug-sensitive strains. They found that under these circumstances, MDR-TB will increase and dominate versus drug-sensitive TB in an epidemic, but that this may take several decades to occur.

"It is imperative that strategies to control tuberculosis include measures to treat individuals with MDR disease," said Cohen. "Others, including our colleagues at Harvard Medical School led by Paul Farmer and Partners in Health, have shown that appropriate treatment can cure these individuals in epidemic settings. It can be done. Our model shows why making this treatment available is absolutely necessary to avoid otherwise inevitable MDR-TB epidemics."

"We should not be reassured by studies that suggest mutated TB is less transmissible," said Murray. "Our model implies that even when the vast majority of mutated strains aren't easily transmitted from person to person, it only takes a few mutations that are transmissible to create MDR epidemics. This is a potential problem that can be avoided if concerted action is taken now to contain MDR-TB."

This study was supported by grants from the National Institute of Allergy and Infectious Diseases (NIAID) and The Bill and Melinda Gates Foundation.

http://www.hsph.harvard.edu/