Treatment outcomes in HIV positive and negative people with drug resistant TB in Khayelitsha, Cape Town

An analysis of treatment outcomes in HIV positive and negative people with drug resistant (DR) tuberculosis (TB) in Khayelitsha showed similar treatment success and long-term mortality in the presence of ART.

This study – presented at the 2014 Southern African Clinicians Society Conference – showed greater mortality in HIV positive participants before and during DR-TB treatment. But, loss to follow up was greater in HIV negative participants, leading to higher mortality after this occurred.

Khayelitsha is a township in Cape Town with a population of about 500,000. About half the population live in informal dwellings. In 2012 antenatal prevalence of HIV was 34% and the vertical transmission rate 1.7%.

Three quarters of about 6,000 TB cases that are registered in Khayelitsha each year are in HIV positive people. Approximately 200 TB cases are DR-TB. The success rate for treating DR-TB is about 46% and the mortality rate about 20%. In March 2014, 11 health facilities in Khayelitsha were providing TB and HIV care and 28, 738 people were receiving ART.

Vivian Cox from Médecins Sans Frontières (MSF) presented a retrospective analysis of treatment outcomes and mortality from routine Khayelitsha DR-TB data.

Of 839 diagnosed DR-TB cases, 607 were in HIV positive and 232 in HIV negative people. Respectively 1% and 11% were presumed to have MDR-TB, 21% and 12% were rifampicin mono-resistant, 61% and 62% had MDR-TB and 16% and 15% MDR-TB plus resistance to 2nd line TB drugs.

Of those diagnosed 736 started DR-TB treatment. Of the 116 (100 HIV positive) who did not start treatment, 62 died before treatment started, 55 (89%) HIV positive and 7 (11%) negative, p<0.001. A total of 470/507 HIV positive people who started DR-TB treatment were on ART and 440/470 had final treatment outcomes available. Of 216 HIV negative people starting DR-TB treatment, 189 had final treatment outcomes available.

Similar proportions, 48% vs 47%, of the HIV positive and negative groups respectively achieved treatment success (defined as cure and treatment completion). Loss to follow up occurred in 27% vs 37%, p=0.01. There was no difference in the proportion of the two groups with treatment failure: 6% vs 7%. But a greater proportion of HIV positive people died: 18% vs 9%, p=0.004.

Similar proportions died after loss to follow up: 26% vs 21% at a median of 6.7 (IQR 2.5-16.1) vs 5.5 (IQR 3.5-13.3) months. A greater proportion of the HIV positive group died after treatment failure: 77% vs 50% at a median of 0 (0-0.2) vs 5.7 (1.5-17.5) months.

Overall mortality was 132 in the HIV positive group, IR/100py 19 (95% CI 16-22) vs 39 in the HIV negative group, IR/100py 17 (95% CI 12-22); IRR 1.14 (95% CI 0.80-1.67).

Dr Cox noted that loss to follow up masks mortality in HIV positive and negative people. Overall treatment success and mortality in HIV positive people on ART is similar to that in HIV negative people.