Abstract
We compared the incidence of multidrug resistance in 150 consecutive Mycobacterium tuberculosis isolates obtained from a rural center (in Sakawar, India) and an urban tertiary care center (in Mumbai, India). The study highlights an alarmingly high percentage of multidrug-resistant M. tuberculosis isolates in Mumbai (51%) as compared with that at the rural center (2%).
The worldwide emergence of multidrug-resistant tuberculosis (MDR-TB) is a major threat to tuberculosis (TB) control. The recrudescence of TB in India has been exacerbated by the emergence and spread of MDR-TB as well as the expanding HIV epidemic. With an estimated 2 million new cases of TB and 500,000 TB-related deaths in India annually, India accounts for one-quarter of the global TB burden [1–3]. MDR-TB is now encountered in India with increasing frequency, with reports from many parts of the country [4–6]. However, the absence of a surveillance network and the lack of reliable prevalence studies make it difficult to evaluate the true extent of MDR-TB in India. We therefore sought to measure the extent of initial and acquired resistance at our tertiary care center in Mumbai, India's largest city, and at a nearby rural center (Sakawar clinic, in the Thane district of India).
For urban cases, samples were collected consecutively from patients suspected of having TB who had cough and fever for >15 days accompanied by weight loss and radiological findings consistent with TB. Ours is a multispecialty tertiary care hospital located in central Mumbai and receiving referral patients from all over the city. Of the total of 1169 patients who were tested for TB, 363 were culture positive. Testing for susceptibility to antitubercular drugs was performed for 150 patients for whom follow-up was possible. Sixty-four (43%) of these 150 patients were referred to our hospital by general practitioners, 50 (33%) were referred by district hospitals, and 36 (24%) were referred by our institution. For rural cases, samples were collected from Ramakrishna rural health and welfare center, Sakawar, which is located ∼100 km from Mumbai. Samples were collected only from patients suspected of having TB.
Samples from a total of 184 patients were collected, and drug susceptibility test results were available for 150 patients. Drug susceptibility testing was performed using Lowenstein-Jensen media by the absolute concentration method [7]. The drugs tested and their concentrations were as follows: isoniazid (INH), 0.1, 0.2, and 0.4 μg/mL; pyrazinamide (Pza), 50, 100, and 150 μg/mL; ethambutol (Emb), 1, 2, and 4 μg/mL; rifampicin (Rmp), 20, 40, and 80 μg/mL; streptomycin (SM), 2, 4, and 8 μg/mL; para-aminosalicylic acid (PAS), 0.25, 0.5, and 1 μg/mL; thioacetazone (Th), 1, 2, and 4 μg/mL; ethionamide (Eta), 10, 20, and 40 μg/mL; ciprofloxacin (Cp), 4, 8, and 16 μg/mL; and sparfloxacin (Sp), 1, 2, and 4 μg/mL. For each drug listed, the middle concentration was considered the breakpoint MIC. Standard strain M. tuberculosis H37Rv was used as quality control.
Patients were interviewed with the standard questionnaire to determine their history of anti-TB drug therapy and were classified as having either initial or acquired resistance, according to the standard definitions [2]. Resistance to a particular drug with or without resistance to other anti-TB drugs was denoted as “any resistance.” Resistance to isoniazid and rifampicin with or without resistance to other anti-TB drugs was defined as “multidrug resistance.”
Detailed results are as shown in table 1. Eighty percent of isolates obtained from the Mumbai patients were resistant to at least 1 drug, and 51% were multidrug resistant. Sixty-seven percent of those patients with no history of previous treatment were resistant (i.e., were classified as having initial resistance) to at least 1 drug, and 30% were multidrug resistant. A very high proportion of isolates were resistant to 2 particular combinations of drugs (SM, INH, Emb, and Rmp, 25% of isolates; and SM, INH, and Rmp, 14% of isolates) which suggests that a few multidrug-resistant strains may be circulating widely, thus accounting for a significant proportion of overall MDR-TB. This scenario is consistent with the fact that, in Mumbai, most patients with MDR-TB continue to have positive sputum smear results and are managed in the community.
Resistance to first-line drugs and combinations of first-line drugs among isolates of Mycobacterium tuberculosis from an urban and a rural region in India.
Resistance to first-line drugs and combinations of first-line drugs among isolates of Mycobacterium tuberculosis from an urban and a rural region in India.
As the Hinduja Hospital is a tertiary referral center, patients attending this hospital are more likely to have been unresponsive to therapy or to have had relapse, resulting in a bias toward drug-resistant isolates. However, three-quarters of the isolates in this study were from sputum samples obtained by district level hospitals and general practitioners. The rate of multidrug resistance among these isolates was similar (P < .05, not significant) to that from our own institution, indicating a high rate of MDR-TB in the community. A number of recent studies have reported increasing rates of MDR-TB in urban settings [4–6]. This suggests that lack of response to first-line therapy (i.e., multidrug resistance) is a significant problem in the community. The rural samples were most probably free of ascertainment bias, because virtually all patients with TB in that region report exclusively to the rural health and welfare center at Sakawar. The rate of MDR-TB in this rural area was comparable with rates in other areas in India [3]
These rates of drug resistance among clinical M. tuberculosis isolates in Mumbai are among the highest reported worldwide; this identifies Mumbai as a hot spot for MDR-TB and has major public health implications. One major reason for the high rate of MDR-TB may be the erratic and highly variable treatment prescribed by doctors in Mumbai's private sector. This, together with overcrowding, HIV infection, inadequate public health TB control measures, and a lack of suitable treatment or containment of MDR-TB cases, could all contribute to a major MDR-TB epidemic in Mumbai unless effective interventions are rapidly instituted. Susceptibility testing is not performed as a routine practice in India, and, sadly, only a few susceptibility tests are being done in a handful of laboratories. The high incidence of MDR-TB in our urban setting stresses the need for drug susceptibility testing to be done for every patient whose culture is positive for M. tuberculosis. The therapy given to the patients in the city needs to be strictly monitored, since many doctors frequently prescribe inadequate therapy [8].
The marked contrast between the high rates of drug resistance and multidrug resistance among isolates obtained from urban patients and the low rates among isolates obtained from rural patients is probably related to the following factors. First, therapy at the rural center was free and supervised on a weekly basis (i.e., a semi–directly observed therapy [DOT] protocol). Second, the rural patients had less access to multiple doctors and multiple drugs, both of which are associated with poor and variable prescribing habits in cities [8]. Third, for epidemiological reasons, there may be less onward transmission of multidrug-resistant strains in rural areas with low population densities. The method of weekly supervised dispensing of drugs at Sakawar could be a good model for treatment elsewhere. This approach could help to control the impending epidemic in Mumbai, though there could be several operational obstacles in implementing this approach in a city such as Mumbai. Moreover, a high preexisting prevalence of MDR-TB would raise doubts about the efficacy and suitability of DOT in this context and may require development of a new, modified control strategy. Use of directly observed therapy short-course-plus (DOTS-plus) is an unrealistic option in India. The prolonged supervision it would entail is beyond the capacity of India's overburdened public health system, which is already struggling to cope with the workload imposed by DOT. Besides, the DOTS-plus approach demands that reliable laboratories for susceptibility testing be available.
To better understand the epidemiology and transmission dynamics of MDR-TB in Mumbai, molecular epidemiology studies are needed to establish what proportion of MDR-TB cases are due to particular strains and what proportion of MDR-TB cases are due to ongoing transmission. Moreover, there are an increasing number of reports that the Beijing genotype of M. tuberculosis is associated with drug resistance, especially in Southeast Asian countries [9, 10]. It would be interesting, from this point of view, to determine the clonality of MDR-TB strains in Mumbai. Improved understanding of the epidemiology of TB through strain typing, therefore, will be important for designing public health interventions at the regional and national level, and we have recently initiated a molecular epidemiology study of MDR-TB in Mumbai to serve the cause.
Acknowledgments
We thank all the doctors and volunteers of the Ramakrishna Mission Rural Health and Welfare Center, Sakawar, District Thane, India, for all their cooperation and help.
Comments