HIV-TB co-infection: They’re already sitting in our waiting rooms

Dr Eric Goemaere is medical co-ordinator for MSF in South Africa. He is based in Khayelitsha, the largest township in the Western Cape province, also the location of the comprehension HIV and TB care project, which MSF manages in conjunction with the Western Cape government. Eric, a former General Director of MSF Belgium, has been with MSF since 1984. He has been in Khayelitsha since 1999.

Increasingly we hear of HIV-TB co-infection, especially in southern Africa; and then we hear of latent TB. What is latent TB? Is it TB?

Typically people co-infected with HIV will potentially present with a very different clinical picture, with regard to TB: mild symptoms, sometimes merely slow wasting, as there is hardly any inflammatory reaction, and worse, the common TB test – sputum smear – repetitively reads negative. This is mainly due to low bacterial concentration in the lungs.

With a 70% co-infection rate in Khayelitsha and 80% in Lesotho, we are regularly confronted with so-called latent TB presenting symptoms which have obliged us to develop elaborated diagnostic algorithms, sometimes difficult to follow by nurses. In such environments we have up to 30% of extra-pulmonary forms, which make diagnosis even more difficult.

How do you test for the TB disease?

A standardised test – the sputum test – invented more than a century ago, is insufficiently sensitive and even more so in patients that are co-infected. Far too many nurses are still sending patients back home as their sputum smear results come back negative while symptoms are present. A study in 2003, done in Khayelitsha, showed that amongst all TB cases notified during the year, only 18% were diagnosed with a positive smear. A major progress in recent years is access to rapid culture tests and the inclusion of culture in the first line of the diagnostic algorithm: This doubles our sensitivity of detection.

Culture is unfortunately far from reality in most African environments where MSF works, but thanks to a joined effort between PIH and FIND, it is now becoming a reality, even in Lesotho. We need to further explore the possibilities of rapid culture in a very simple lab setting, as it will save a lot of smear-negative patients from being diagnosed either too late, or not at all. Despite all of these progresses, TB remains our number one killer in Khayelitsha, indicating how far we are from a decent diagnostic for co-infected patients.

In the past 15 years new TB cases have tripled in countries with high HIV prevalence, but in 2006 less than one percent of people living with HIV/AIDS were screened for TB. Less than one percent?

I would attribute this official figure mainly to the diagnostic difficulties as explained before. We are training our staff to weigh patients at each visit and to systematically question them for classical TB symptoms. Here again, knowledge of HIV status is extremely important as symptoms will differ radically from those who are HIV positive.

We have for too long kept the illusion that applying DOTS, together with smear results central to the diagnosis, would control the epidemic. This might be true in countries with low HIV prevalence, but figures exist proving the contrary in high HIV-prevalent settings. The problem now is to validate smear-negative algorithms, which are nurse friendly: TB programmes in these settings are nurse-based and needs to remain so or we risk losing many more patients as access to doctors in rural areas are often limited. We validated such an algorithm in Khayelitsha two years ago, a relatively privileged setting as we have access to culture, CRP and chest X-ray facilities, and potentially advice from a doctor. We are now testing another algorithm in rural Lesotho, in a much more resource-constrained environment

What is combined care for HIV-TB co-infection?

Mortality in co-infected patients is much higher than in patients only suffering TB; their TB evolves much faster as their cellular immune system no longer has control whilst TB infection increases the viral load. No time to waste here: it is a matter of a few weeks before patients die.

MSF now promotes routine testing with an "opt out" option in TB services, with some success as testing rates reaches more than 90%. It is vital for TB patients to know if they are HIV infected, and if so – given cotrimoxazole prophylaxis and CD count done – to see if patients are eligible for ARVs. If eligible, they will then be referred to an ART service, the prognostic being radically different.

I believe this is more and more happening across southern Africa. We could call it the "minimal package".

So what exactly is the one-stop service offered in Khayelitsha?

In several clinics in Khayelitsha, jointly with the Cape Town City health department, we have pushed for much more integration. Both TB and HIV services, including ART, are fully integrated. This necessitated decentralising ART initiation to most peripheral clinics with extensive nurse training in order to treat both diseases.

It is easy to see the advantages for the patient as this becomes a one-stop service: precise synchronisation of both treatments; necessary adjustments for drug-drug interactions and side effects; and, close monitoring.

Of course, what immediately comes to mind is the risk of nosocomial infection: as a result of having HIV positive and TB patients waiting together in the same waiting room. We have some preliminary evidence showing that, in fact, the relative risk has hardly increased as many were already sitting in these waiting rooms, for weeks, and undiagnosed. This has required extensive infection-control work, mainly based on the use of natural ventilation, controlled patient flow, and UV.

HIV-TB co-infection can lead to immune reconstitution syndrome. This almost sounds positive, yet we are very concerned about this syndrome; why is that?

Immune reconstitution syndrome is a paradoxical reaction: it happens in previously diagnosed TB patients, usually in the initial two months of treatment – we kick start ART and, after a couple of weeks, TB symptoms resurface and sometimes even more severely, while the patient was clearly improving on TB treatment before ART was started.

An alternative form, called "unmasking TB”, is the appearance of TB symptoms, nonexistent before initiating ART, a few weeks after initiation. What happens here is that as soon as ART reduces the VL, a pool of T-lymphocytes, which were in hiding, protecting themselves from HIV in the blood, reappears massively in the blood flow. They immediately begin patrolling to find any existing foreign antigen, then they find TB antigens and unleash a massive inflammatory syndrome with very florid TB-like symptoms. This can be deadly, mainly in case of TB meningitis, hence we need to always screen for TB before starting ART.

This then another reason why it is important to have the one-stop for HIV-TB care.