Home Search Abstracts View Session E-mail Abstract Author


Session 135 Poster Abstracts
Treatment Issues in Tuberculosis and HIV Co-Infection
Session Day and Time: Wednesday, 1:30 - 3:30 pm
Poster Hall


795
Identification of a Multi-drug-resistant Tuberculosis Cluster as a Cause of Death among HIV-co-infected Patients in Rural South Africa
A Moll1, Neel Gandhi*2,3, R Pawinski4, J Andrews2, K Zeller2,5, U Lalloo4, W Sturm4, and G Friedland2
1Church of Scotland Hosp & Philanjalo, Tugela Ferry, South Africa; 2Yale Univ, New Haven, CT, US; 3Emory Univ, Atlanta, GA, US; 4Nelson R Mandela Sch of Med, Durban, South Africa; and 5Brown Univ, Providence, RI, US

Background:  Roughly two-thirds of all active tuberculosis (TB) patients in rural KwaZulu Natal, South Africa, are co-infected with HIV. Case fatality rates are nearly 40%, despite the availability of TB therapy by the World Health Organization’s directly observed therapy short-course (DOTS) strategy. The causes of death have not been well evaluated, but are often attributed to “end-stage AIDS.” With the recent introduction of ART medications, several patients have died despite having good virologic response to ART. Further investigation revealed that patients were infected with a highly resistant strain of multi-drug-resistant (MDR) TB. We sought to determine the extent of this highly resistant strain of MDR TB among patients in this district.

Methods:  Increased surveillance with sputum culture and drug susceptibility testing was initiated in January 2005 for patients admitted to the TB wards of a provincial hospital in rural South Africa, and for outpatients with cough. Spoligotyping was performed on sputum samples found to have resistance to all tested TB drugs (isoniazid, rifampin, ethambutol, streptomycin, ciprofloxacin, and kanamycin).

Results:  Among 93 positive sputum cultures collected from January through June 2005, 40 (43%) were found to have MDR TB; 26 (28%) had identical susceptibility patterns with resistance to all first- and second-line drugs tested. Chart review of previous MDR TB cases revealed 5 additional cases with high level of resistance, bringing the total to 31 in the past year. Spoligotyping available for 11 of these 31 highly resistant patients revealed 9 (82%) with the same strain. Of the 31 highly resistant MDR TB patients, 27 (87%) have died. Median survival after sputum was collected was 13 days (range, 1 to 200 days). HIV status is known for 29 of 31 patients, all of whom were HIV+, with CD4 counts ranging from 6 to 283 cells/mm3. Of the total, 55% of patients had been previously treated for TB, of which 33% successfully completed therapy. All patients with MDR TB had been hospitalized in the past year.

Conclusions:  Increased surveillance for MDR TB in a rural South African community has uncovered a strain with high levels of drug resistance as a previously unrecognized cause of death among HIV/TB co-infected patients. This cluster highlights the potential dangers of MDR TB transmission in high prevalence HIV and TB populations. As more communities roll out HIV care and ART, the need for improved TB diagnostic, therapeutic, drug sensitivity, and infection control procedures is critical for further reducing mortality in co-infected patients.