Nathan Geffen, TAC

A poster by Shenoi and colleagues at the World Lung Conference examined survival characteristics of people with XDR TB in Tugela Ferry, Kwazulu-Natal. [1]

This was a case controlled study. It included patients registered at Tugela Ferry with new active TB. Patients also had had culture and drug susceptibility testing demonstrating resistance to isoniazid, rifampin, ciprofloxacin, and kanamycin. Survivors were defined as patients who were alive at least six months from sputum collection. Non-survivors were defined as patients who died within six months from sputum collection. Non-survivors were matched to survivors by year of XDR TB diagnosis.

The researchers found records of 460 patients with XDR TB from 2005 to 2008. Of these 122 were confirmed alive at time of receipt of DST results and referred for XDR TB treatment. Of these, 68 met the definition of survivors and 74 met the definition of non-survivors. Remarkably, only one patient in each arm was HIV-negative.

The XDR TB regimen used in Tugela Ferry was ethambutol, pyrazinamide, ethionamide cycloserine or Terizidone, Para-aminosalicylic acid (since 2007) and capreomycin for at least six months followed by an 18 month continuation phase that excluded capreomycin. Amoxicillin/clavulanate and clarithromycin were added to the regimen if one of the above was discontinued due to toxicity or if patients were still TB-culture positive after 12 months of treatment. Patients were also treated with antiretrovirals if there CD4 count was lower than 200 cells/mm3.

Survival was associated with several factors. The most significant and intuitively sensible of the treatment characteristics associated with survival were having started XDR TB treatment (97% v 28%, p<0.0001), having ever received ART (82% v 42%, p<0.0001) and having a higher baseline CD4 count  at XDR diagnosis (162[IQR:72-260] v 84 [IQR:40-157] cells/mm3, p=0.004). Being smear negative also predicted survival (37% v 79%, p<0.0001). The majority of patients on ART initiated therapy prior to diagnosis of XDR TB. The authors indicate that ART may have unmasked XDR TB IRIS in some patients.

The authors emphasise that laboratory diagnostic delay is a major impediment to detecting XDR TB. They also suggest that smear negativity and better multi-organ functioning suggested less disease severity at baseline. They state that concurrent therapy for both XDR TB and HIV coinfection is feasible and associated with successful XDR TB outcomes.

comment

Although this is a small case controlled study, this research is important for understanding how to improve survival in XDR TB patients. The new South African HIV treatment guidelines, in line with these findings, now provide for ART for all HIV-positive patients with drug-resistant TB, irrespective of CD4 count. The new guidelines are described in this issue of HTB South.

Notable is the apparent extremely high mortality rate. From the 460 patients identified, only 128 were confirmed alive. While this does not mean that all of the remainder died, it is very probable that most of them did. Furthermore, of the confirmed survivors, less than half lived six months. The likely availability of new TB drugs in the next few years will probably help reduce the drug-resistant mortality rate. But, even more important is finding ways to diagnose drug-resistant TB quicker, primarily by getting DST results back to clinics and hospitals faster. This will probably have the greatest short to medium-term impact on reducing drug-resistant TB mortality.

It would be useful if the authors clarified some of their findings in greater detail (some of the data on their poster is unclear or counterintuitive) and published this valuable study in a journal.

As the authors recommend, prospective studies to examine their findings in greater detail are needed.

Ref: Shenoi SV et al. 2009. Comparison of characteristics of XDR TB survivors with those of non-survivors in rural Kwa-Zulu Natal, South Africa. 40th World Lung Conference, Cancun 2009.