Extensively drug-resistant TB (XDR-TB) present throughout the world

Extensively drug-resistant tuberculosis (XDR-TB) is present in every region of the world, participants at the 37^th Union World Conference on Lung Health in Paris were told this week.

“We do not know how common it is. We just know that it exists and it exists in different parts of the world” said Dr Ken Castro, Director of the Division of Tuberculosis Elimination, at the US Centers for Disease Control (CDC).

At a symposium on Tuesday, Dr Castro gave a brief report the the first global XDR-TB task force convened in October in Geneva and the preliminary plans to combat the illness. Then on Thursday, Dr Sarita Shah of the Albert Einstein College of Medicine in New York and the CDC gave a more detailed report on the outcome of that task force, which included a revised definition of XDR-TB (see below) as well as plans for future surveillance and reporting to better characterise the spread of the illness.

The evolution of XDR-TB MDR-TB

Although the recent outbreak of XDR-TB in South Africa, which appears to be very rapidly fatal in people with AIDS, has perhaps received most of the press, the XDR-TB story began somewhat earlier.

The story should actually begin with the evolution of multidrug-resistant (MDR-TB), defined as TB which is resistant to at least the two cornerstone TB drugs, isoniazid and rifampicin, that first emerged as a public health issue in the 1990s. MDR-TB may be caused by inadequate treatment of drug susceptible TB but once that has occurred, MDR-TB can also be transmitted to other individuals. The problem of MDR-TB has long been known to be a serious one, especially in Eastern Europe, the countries of the former Soviet Union, Peru and Asia.

The treatment of MDR-TB requires the prolonged use of expensive and often hard-to-tolerate second-line drugs which are divided into six classes including:

• an aminoglycoside, such as the injectable drugs amikacin and kanamycin

• a polypeptide such as the injectable drug capreomycin

• a fluoroquinolone such as ofloxacin or ciprofloxacin (as well as the new drugs moxifloxacin and gatifloxacin)

• a thioamide such as ethionamide and prothionamide

• a serine analogue such as cycloserine and terizidone; and

• para-aminosalicylic acid (PAS)

Generally, a minimum of four to six drugs are used, starting with any of the remaining first-line drugs to which the patient’s TB is still susceptible (including ethambutol and pyrazinamide), with one of the injectable drugs, a fluoroquinolone, and as many as needed or available from the last three classes of drugs. Unfortunately, the last three classes are less potent and drug susceptibility testing (DST) for these drugs has yet to be standardised — which makes it difficult to predict whether they will help or not.

“Inadequate treatment of MDR-TB will predictably result in the development of resistance to second-line drugs,” said Dr Shah. And once resistance to these second-line drugs develop, she stressed “it may be difficult, if not impossible in some settings to treat the disease successfully.”

Changes since the reports in Kwa-Zulu Natal, South Africa

The low frequency in Africa is probably more of a reflection of the scarcity of samples submitted from this part of the world — though according to a presentation on Thursday made by Dr Neel R Gandhi, one of the investigators of the cases in Tugela Ferry, it may have only evolved in the region relatively recently. These cases have previously been described on aidsmap (see here, here and here), and presentations at the conference in Paris on the South African XDR-TB epidemic will be discussed at greater length in a subsequent article).

But one aspect that really stood out about the cases in South Africa was the high mortality rate in people with HIV (44 out of 53 had been tested for HIV, though based upon clinical presentation, all of the cases may have been HIV infected). 52 out of 53 of the cases (98%) were dead within a median of 25 days from diagnosis.

This association of HIV, XDR-TB and mortality may extend to other regions of the world as well. For example, Dr Vaira Leimane of the national DOTS program in Latvia noted that while Latvia does not have a high burden of HIV, there was a higher percentage of people with HIV among the XDR-TB (10%) than among the MDR-TB cases (3%). 55% of the cases of XDR-TB ended in treatment failure, despite access to a large number of TB drugs in the country. “HIV may be a factor in worse treatment outcomes among the drug resistant patients,” said Dr. Leimane.

Early reports of XDR-TB

And indeed, over the last few years, anecdotal reports of virtually untreatable MDR-TB — or XDR-TB — had begun to crop up in TB programmes across the globe.

To investigate how extensive this emerging phenomenon was, investigators from the CDC and WHO and International Union Against Tuberculosis and Lung Disease (IUATLD) Global Project conducted a retrospective survey of all the TB isolates that had been submitted to the network of Supranational Reference Laboratories (SRLs) for second-line DST for any reason between 2000 and 2004. The SRLs are a network of laboratories which perform epidemiological TB surveillance and DST across six continents.

For the purposes of the study XDR-TB was defined as TB that was resistant to the two cornerstone TB drugs, rifampicin and isoniazid (which constitutes MDR-TB) plus at least three out of six of the second-line drugs used in TB treatment. However, Dr Shah pointed out that the lack of global standardisation of DST for these second-line drugs is a major limitation to being able to reliably identify how widespread the problem is.

49 countries had sent in samples to the 14 participating SRLs, but the data received were very sparse from African countries, and were not available from India or China. Also, since the isolates in the survey were from a referral population of often high risk patients, the researchers also included population level data from South Korea, the United States and Latvia.

The researchers published their findings World TB Day, March 24 of this year in the Morbidity and Mortality Weekly Report.

Out of 17,690 isolates that were submitted by the SRLs, 3,520 (20%) were MDR-TB, and out of these 347 (10% of the MDR-TB or 2% of the total isolates) were XDR-TB.

XDR-TB was present in each region of the world, although it was more commonly seen in former Soviet Union or Eastern European countries (14% of MDR isolates) and Asia, especially in the Republic of Korea (15% of MDR isolates), where they have been systematically performing DST for some time. It was relatively infrequent (around 6% of MDR) in the industrialised nations and even less frequent reported in Africa.

Changes in the definition of XDR-TB and increasing surveillance

She also noted that discrepancies in DST testing for second-line drugs frequently makes treatment management more difficult.

This was one of the issues considered by the aspect of the global XDR task force with laboratory specialists, clinicians and epidemiologists that convened in early October in Geneva. One of the problems with the initial definition of XDR-TB was the poor ability to get reproducible DST results with some of the second-line drugs. It also failed to stress the greater clinical importance of the resistance to the injectable and fluoroquinolone classes of TB drugs. Resistance to these drugs is consistently associated with a poorer clinical outcome.

So the task force has issued a new definition for XDR-TB which it believes will be more consistent, reproducible and clinically relevant. To be XDR-TB an isolate should include resistance to isoniazid and rifampicin (MDR-TB) plus resistance to any fluoroquinolone and resistance to at least one of the injectable drugs. Under the revised definition, 234 or 7% of the MDR-TB isolates from the survey in the MMWR were XDR-TB.

Both the task force and a Global Action plan to combat XDR-TB are calling for increased surveillance.

Dr Castro highlighted some of the challenges inherent in this aspect of this plan.

“If you were to ask me how common [XDR-TB] is or where it is happening, we wouldn’t know and we need to know. The next question is well, how are we going to find out? Well, what labs are there available? It really highlights a deficiency that has existed and prevailed out there that, I think, for too long we have been too complacent about. We need to improve rapid drug susceptibility testing and bring that closer to the patient,” he said.

Furthermore, “in order to determine estimates of burden and prevalence of XDR-TB, standardised surveillance and reporting of drug resistance is needed,” said Dr Shah. “Because so few countries are offering first-line DST, second-line testing is even less accessible in most countries.”

She said that currently the WHO/IUATLD Global Project recommends that all MDR-TB isolates detected should be sent into SRLs for second-line DST. A quality assurance system is under development.

Given the lack of laboratory capacity and staff in countries most at risk, she said that in the meantime, programmes should focus on the high-risk groups, with DST on known MDR-TB and failures. Finally, she recommended that HIV testing should also be integrated into routine drug resistance surveys.