© 2008 American Thoracic Society doi: 10.1164/rccm.200809-1380ED
Of Blind Men and Elephants: Making Sense of Extensively Drug-resistant Tuberculosis
WHO Collaborating Centre for TB and Lung Diseases
PATH
Division of Clinical Infectious Diseases, Medical Clinic When extensively drug-resistant tuberculosis (XDR-TB)* was first reported in March 2006 (1), tuberculosis was propelled overnight from the ranks of neglected diseases to sensational peril. An outbreak in KwaZulu Natal, South Africa with a very high fatality rate among the 53 cases reported (all with HIV co-infection) kindled an urgent interest in understanding the risks of XDR-TB for both individual patients and public health (2). As of today, XDR strains of M. tuberculosis have been found in almost 50 countries in all regions of the world (3). This outbreak generates a number of questions requiring urgent answers: Are the clinical characteristics of XDR-TB different from those of multidrug-resistant tuberculosis (MDR-TB)? Is XDR-TB a death sentence? What approach to treatment might yield the best results? And the most troubling question of all: When it comes to XDR-TB, are we going back to the time before there were anti-TB drugs at all (4)?
The scientific community is still puzzling over answers. We find ourselves in the same conundrum that faced the blind men in describing the elephant: is everyone partly right, and are we all mostly wrong at the same time? In this issue of the Journal (pp. 1075–1082), Kim and coworkers provide additional evidence for debate as they report findings on XDR-TB treatment outcomes and long-term survival in South Korea (5). Kim and colleagues' findings confirm some of the evidence previously presented, offer interesting hypotheses to test, and highlight contemporary challenges. On the issue of clinical differences, Kim and coworkers, like others, found that individuals with XDR-TB more frequently had a history of previous treatment with second-line agents, had more advanced disease, and were more often hospitalized at treatment initiation than their counterparts with MDR-TB. In comparing outcomes for MDR-TB versus XDR-TB, Kim and colleagues found that XDR-TB was the strongest predictor of poor outcome and decreased long-term survival in their cohort. This corroborates earlier findings by TBNET in Europe, the first group to demonstrate that XDR-TB carries higher (5.45 times) risk of death than MDR-TB and requires a longer time to achieve sputum smear and culture conversion (6). In the European cohort followed for 42 months (7), treatment success was only 39.3% among those with final outcome recorded. These findings were confirmed by another Korean study, in which 53.5% success was achieved (8). In the study by Kim and coworkers, which has the added strength of following patients for 3 to 7 years after initial treatment outcome, almost half of those with XDR-TB died, with 84% of the deaths attributable to tuberculosis. This survival rate is similar to that for the natural course of tuberculosis before the discovery of streptomycin in 1943. From current work, it appears that while XDR-TB is not always a death sentence, particularly in HIV-negative patients, the prognosis is poor when compared with outcomes of patients afflicted by less resistant strains. How well could patients with XDR-TB do under ideal conditions, and what would those conditions be? In the study by Kim and colleagues, few of the cases with XDR-TB were treated with more than two to three active drugs, which begs the question of what might have been if more drugs or other interventions were available for use in these patients, and may explain the high mortality associated with XDR-TB in their cohort. In contrast, a recent prospective study by Mitnick and coworkers (9) from Peru reported 60.4% treatment success, and a report from Russia showed a favorable outcome in 48% of patients with XDR-TB, with no increase in mortality (10). The Peruvian and Russian patients had a combination of interventions: aggressive treatment regimens and management of side effects, as well as an excellent package of social support interventions to improve chances for success. Still, results are sobering. It is clear from the work done to date that rapid diagnostics to shorten the time to effective treatment initiation for drug-resistant disease and new drugs to meet the challenge of XDR-TB are essential to improve outcomes. Regrettably, a new drug for treatment of tuberculosis has not been licensed in decades. Moxifloxacin and linezolid show promise in some circumstances (11). Other studies suggest that aggressive treatment and case management may be the best hope for patients with XDR-TB (12). Kim and colleagues note that surgical resection was the only positive predictor of treatment success in their cohort, for both MDR-TB and XDR-TB. Although the numbers are small, this finding warrants serious further study. Their work is substantiated by similar effects noted elsewhere (9, 12). While we are not quite back at square one in treating XDR-TB, it appears that some of the interventions from the early days of tuberculosis treatment, including surgery and nutritional support, may again prove useful. At this early stage in interpreting the impact of XDR-TB, all findings must be taken with healthy skepticism because of the difficulties inherent in studying this entity. The majority of research to date relies on retrospective cohorts with relatively small numbers of cases to evaluate. In addition, the technical difficulties in testing for resistance to all the XDR-TB–defining drugs, even in high-income countries, hamper our ability to accurately categorize patients (13). The study by Kim and coworkers, like many others, did not have access to data for all XDR-TB–defining drugs, and thus XDR-TB prevalence in the cohort may have been underestimated. The messages that we should take from attempts made thus far to describe the elephant that is XDR-TB are several, and related. First, while the benefits of aggressive treatment are yet to be fully evaluated, the costs of not treating XDR-TB aggressively have been universally high for individual patients and public health. Second, while medical intervention is crucial, it alone cannot produce optimal results—adjunct therapies and psychosocial support are keys to yielding positive outcomes. Third, coordinated, prospective research that can collect representative data using standard definitions and outcome measures (14, 15) on large patient populations is critical to our success in conquering XDR-TB. What we must not forget is that preventing XDR-TB through vigorous tuberculosis control programs residing in strong health systems is still the best answer to our unanswered questions (16). Respiratory medicine specialists have a critical role to play in winning this battle. As we wait for new diagnostics and drugs that can meet the challenge of XDR-TB, we must work with what we presently have to create the optimal conditions for their success, and thus seize the opportunity we have to eliminate tuberculosis. FOOTNOTES
* XDR-TB is now defined as disease caused by organisms resistant to isoniazid, rifampin, any fluoroquinolone, and at least one injectable drug: capreomycin, kanamycin, or amikacin.
Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. REFERENCES
Related articles in AJRCCM:
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
From a famous Indian legend and the poem by John Godfrey Saxe, "The Blindmen and the Elephant," in which six blind men disagree over the nature of an elephant based on the body parts they are feeling (tusk, trunk, tail, etc.) because they cannot see the whole. 
