Analysis: Not all good news in new WHO MDR-TB recommendations

In a move that could help save the hearing of thousands of people with TB, the World Health Organisation (WHO) is advising wider use of newer and repurposed drugs, bedaquiline and linezolid in place of injections that cause deafness. Rather perplexingly, the WHO is however also maintaining its recommendation for a shorter and cheaper, but much less impressive treatment regimen and advised that the new South African regimen should only be used under conditions of operational research.

In their announcement on Friday the WHO recommends two different treatment regimens for the treatment of MDR-TB – a long and a short regimen. They do not indicate that either of the two is preferable, thus leaving the door open for governments to decide. This is problematic, since the long regimen is almost certainly superior to the other in terms of patient outcomes, if not in terms of price.

The WHO long regimen

The long regimen recommended for MDR-TB (and RR-TB – TB resistant only to rifampicin) is composed of bedaquiline, linezolid, levofloxacin (or moxifloxacin), and cycloserine or clofazimine taken for 18–20 months – with some drugs taken for a shorter period. For more details on which drugs are to be used for which duration we will have to wait for the full WHO guidelines expected later this year.

One of the most notable things about this regimen is that it does not contain any injections, and that patients will thus not be exposed to the high rates of hearing loss and other severe side effects associated with injectable drugs such as kanamycin, capreomycin, and amikacin. There will of course still be side effects – linezolid in particular is a tough drug to tolerate and can cause anaemia and nerve damage amongst others.

The long duration and the inclusion of powerful drugs like bedaquiline and linezolid makes this a potent regimen that should result in many more people being cured than is currently the case in most countries. Almost as important, is that this regimen will cure people without causing any hearing loss in the process – something that constitutes a major advance in how we treat MDR-TB.

The WHO shorter regimen

The WHO’s recommended short regimen consists of amikacin, moxifloxacin, prothionamide, clofazimine, pyrazinamide, high-dose isoniazid, and ethambutol taken for nine to 12 months. This is slightly different from the WHO’s previously recommended short regimen that contained kanamycin instead of amikacin. Either way, it still contains injections and comes with a risk of hearing loss.

That the WHO should recommend this short regimen in the same breath as the longer regimen is perplexing. In their explanation of how to construct the longer regimen, the WHO classifies all the key MDR-TB drugs into three groups where the drugs are “ranked based on the latest evidence about the balance of effectiveness to safety”. Group A contains “medicines to be prioritised”, group B contains “medicines to be added next” and group C “Medicines to be included to complete the regimens and when agents from Groups A and B cannot be used”. It is made very clear that drugs from groups A and B are the preferred choices and drugs from group C are only backups. The perplexing thing is that four of the seven drugs in the WHO’s recommended short regimen are from group C – in the WHO’s own words the lowest rank group in terms of the balance of effectiveness to safety. The shorter regimen also contains one drug from group A and one from group B. The seventh drug, isoniazid, is mostly used for drug-sensitive TB – MDR-TB is by definition TB that is resistant to isoniazid and rifampicin. In some cases however high-dose isoniazid might be able to overcome resistance- which is why it appears in this regimen.

Either way, this shorter regimen disappointed last year in a major trial called STREAM stage I where it failed to show non-inferiority to (that it was as good as) the old WHO-recommended 18-24-months of treatment. Since the shorter regimen is cheaper, and the WHO has left the door open to its use, some less progressive countries might opt for this inferior regimen and end up dragging their feet when it comes to phasing out the injectables and implementing the longer regimen discussed above.

To its credit, the WHO does recommend that “decisions to start newly diagnosed patients on the standardized shorter MDR-TB regimen should be made according to patient preference and clinical judgement…”. It is hard to imagine well-informed patients and clinicians preferring this regimen over the longer regimen. In reality of course patients and clinicians often do not in fact have such choices.

Another complication with both the WHO short regimen and the South African short regimen (discussed below), is that a significant number of people will have TB that is resistant to one or more of the seven drugs in these regimens. In the absence of widespread drug-susceptibility testing, this means many people will be given drugs that are ineffective against their TB, but that nevertheless contributes to side effects.

The South African short regimen

Two months ago, South Africa also announced a new treatment regimen for MDR-TB, essentially taking the shorter regimen discussed above and swapping out the injectable in that regimen for bedaquiline. While we have not seen final treatment guidelines from South Africa’s Department of Health, responding to a Spotlight query the Department previously indicated that the regimen would contain bedaquiline, Moxifloxacin, Ethionamide, Clofazimine, High dose Isoniazid, Ethambutol and Pyrazinamide.

As with the WHO-recommended shorter regimen, this modified version of the shorter regimen contains a number of group C drugs – or drugs with a relatively poor balance between safety and effectiveness. Replacing the injectable with bedaquiline is however a positive move and should result in both increased cure rates and a dramatic reduction in hearing loss when compared to the WHO’s short regimen.

Thus, while the South African regimen is preferable to the WHO’s shorter regimen, it is probably not preferable to the WHO’s longer regimen – although this last judgement is based on relatively weak evidence and may turn out to be wrong as evidence accumulates in the coming years.

In its rapid communication on Friday the WHO effectively advises against the wide rollout of the South African regimen, stating that “evidence is currently lacking on the effect of replacing any of the agents with alternatives in the shorter regimen (e.g replacing the injectable with bedaquiline or other oral agents…)”. They then go on to advise that any variations to the standardized shorter MDR-TB regimen should only be considered under operational research conditions – i.e. not rolled out widely as is the case in South Africa.

While it is true that there is relatively limited evidence for the use of bedaquiline in a shorter regimen, this relative lack of evidence should be seen in the context of the minefield of other evidence gaps in MDR-TB, the relatively poor performance of the WHO shorter regimen, and the well-established link between the injectables and hearing loss. On balance, it is hard to see why any well-informed person with MDR-TB would choose to take the WHO-recommended shorter regimen rather than the South African version of the shorter regimen, and yet that is what the WHO recommends.

Not on firm ground

Finally, in a line that might well go unnoticed by someone skim-reading The new WHO communication, the WHO states that two of the injectables, kanamycin and capreomycin, are no longer recommended “given increased risk of treatment failure and relapse associated with their use in longer MDR-TB regimens.”

It is a line worth re-reading.

It suggests that over decades thousands and thousands of people have been subjected to months of painful, hearing-loss-causing  injections, not for a better chance at getting healthy again, but for an “increased risk of treatment failure and relapse”. It is a staggering thought that suggests a deep and serious problem in the way we collectively conducted TB research in recent decades. If the WHO is correct that these drugs increase these risks, how is it that we are only finding this out now?

Chronic underinvestment in TB research, and particularly in large randomised controlled trials, has arguably made it possible for ineffective, or even harmful drugs to stay part of standard DR-TB treatment regimens for decades. Even for an apparent breakthrough drug like bedaquiline, we have not yet seen phase III data – although we have seen enough good data from other sources to justify forging ahead.

With several important randomised controlled trials into various MDR-TB treatment regimens underway, we are now slowly moving away from this chronic problem of insufficient evidence in TB – although dramatically increased funding for TB research is needed if we are to make a clean break with this shameful past of underinvestment.

As MDR-TB trials like NEXT-TB and END-TB are reported in the coming years, we will no doubt see many more changes to WHO and national guidelines as the ground beneath our feet slowly firms up. Until then the reality is that we remain on relatively unsteady ground. We cannot quite discount the possibility that even an increasingly safe and effective -looking drug like bedaquiline might turn out to have an achilles heal. For now though, there is more than enough reason for patients and treatment programmes to embrace both bedaquiline and the rest of the WHO’s new longer regimen as the best way forward given imperfect evidence.