TB and HIV coinfection

Training Session:

Dr Anton Pozniak

This is a report of the training session on TB and HIV coinfection that was given to the UK-CAB by Dr Anton Pozniak. Reading material for this meeting and the slide set are on the i-Base website. Report by Simon Collins.

The main subject in this talk will be:

  • chemoprophylaxis – how to prevent TB by giving drugs to protect against it
  • treatment and side effects
  • when to treat, how, and with what
  • IRIS – Immune Reconstitution Inflammatory Syndrome
  • MDR – multi-drug resistance

The slide set accompanying this talk are available to download from:


Note the following abbreviations are commonly used in slides and studies for the following drugs:

Izoniazid IHZ H
Pirazimamide PZM Z
Rifampicin RFP R
Rifabutin RBT Rb
Rifapentine RPT Rp
Ethanbutol ETH E

The relationship between TB and HIV

The relationship between TB and HIV prevalence in adults is very clear – as HIV incidence rises so does TB – especially in the developing world.

We collected data back in 1990 showing that in Zimbabwe over 80% of people with TB were coinfected with HIV – but this was seen as unimportant back then and it was rejected from World AIDS conference. Now it is recognised as a crucial fact.

It makes treating TB a major problem – and putting programes together is difficult. TB doctors are concerned that HIV doctor will dominate. The WHO in Geneva is concerned about TB, but I personally believe that even if the WHO closed, that TB programmes would still continue.

Any country with HIV will have TB – and with increased travel to and from Africa and Eastern European raises this is as a health concern in the UK. Russia has huge TB problem.

London figures from TB increases from African and Asian visitors – and the Home Office dispersal policy makes this worse. Some units outside London have little or expertise of refugees or TB and this is an important issue for advocates.


TB is rising in the UK, with 7,000 formal notifications in England and Wales a year, up from the 1987 low point of just over 5,000 (if nothing like the 1913 figure of 117,000). Two-thirds of confirmed cases are in foreign-born patients, up from 45% in 1988, but disease surveillance reports, as with HIV/Aids and hepatitis, give no indication of immigration status. Doctors believe a substantial proportion of TB cases might be explained by poor housing, poor diet and limited access to medical services.

A recent pilot study by the Home Office screened 5000 asylum seekers in Kent but failed to find one case of potential infectious TB.

Ref: The Guardian, 7 Feb 2003:

Use of prophylaxis treatment

Given that treatment for TB are very effective, many people thought that the solution would be to identify patients at high risk of TB and treat them to prevent infection. Only a few doctors though this was not as straight forward as it appeared – and finally studies have born this out. It is worth explaining why use of both primary and secondary prophylaxis is not generally recommended.

One third of the worlds population has been exposed to TB but this is not the same as having active TB. There is often confusion between infection and active disease. This is similar to being exposed to chicken pox – most people are exposed to the virus as children. They then fight the infection and recover but they still carry the virus throughout their lives – though they no longer have active disease. Once you are infected with TB however, then there is the potential for the disease to reactivate in the future.

TB is diagnosed in various ways – sputum, xrays, skin test etc – but it is not always easy to differentiate between active and dormant? However the primary cause of reactivation is HIV – so for a prophylaxis programme to work you would need to be able to identify these people at highest risk.

Multiple studies using the tuberculin skin test found that if you are HIV-positive and had a positive skin test, then giving isoniazid prophylaxis provides 50-80% protection from getting TB in the future.

This sounds fantastic – use a skin test – perhaps the new blood test – BUT, unfortunately this only works for HIV-positive people who have a high CD4 count. The skin test does not work if your CD4 count is under 400 – and only 8-10% people have a count >400.

When isoniazid was given to the general population (HIV-positive and HIV-negative) – the protective value was lost.

In spite of this, data from many different countries has been published showing a protective benefit from izoniazid prophylaxis compared to placebo [Bucher, AIDS 1999;13:501].

But, this may be protective for only a short period. Taking liver toxic drugs for 1 year you may only get protection for 2-3 years. Could we give izoniazid forever? – This is certainly not recommended – and if TB infection occurs later, there will be resistance to isoniazid, which is a critical drug required for treatment.

There has been only one unpublished study about duration of protection [Quigley, 2001] – and this showed that protection wore off after 2 years and adherence was not assessed in the study – and izoniazid is a difficult drug to take.

Could other drugs be used for prophylaxis?

Although studies have shown that 2 month rifampicin + pirazimamide (R+Z) may be as good at one year of izoniazid, rifampicin resistance is terrible if you later get TB.

Other regimens can have TB-protective benefit – ie twice weekly rifampicin + pirazimamide – had high rates of hepatitis related deaths in HIV-negative – therefore why would you want to give this to HIV-positive people? So there is the potential of 2 or 6 months R+Z – but these are expensive drugs, can cause resistance, and there can be uncertainties of continuous drug supply etc

There is good intent behind the idea of prophylaxis – but because of the many complications this is not the answer for most settings.

Proportion of people in programmes shown in the slide below indicate not only the practical difficulty of treatment but also the unreliability of the data that is published from these studies.

Powerpoint Slide

Even in the studies from Thailand, Uganda and Brazil, where 100% of people in the study were HIV-positive, even in Thailand and Brazil where 100% entered the study – only 89% and 75% started prophylaxis treatment. And adherence in those taking treatment is less than 70%.

In Uganda – all were HIV-positive – but only 50% entered the study, and only 35% of those took drug and only 70% of those were adherent etc. Only half people returned for their HIV test result and only half took treatment. This is all WHO generated data and they still concluded that ‘this is feasible’ – but the data contradicts their conclusions.

You have to be very careful with all data – especially where the conclusions contradict the data – which is not that uncommon. It results in giving people unrealistic hope.

It is also important to remember that the skin test itself is not straight-forward:

  • it requires specialist skills
  • you need to return after 48 hours (with Heaf test this can be one week)
  • it requires supply of tuberculin
  • need to avoid hepatitis contamination to people, etc

It is sometimes possible to treat all in a contained population – ie in prisons, the army or mineworkers, or household contacts, but the length of treatment is still a real problem – and without continued protection there is risk of infection.

Secondary prophylaxis

Secondary prophylaxis is where you treat people who have already had an infection in order to prevent the infection or disease from occurring again.

With TB, the prevalence of TB in the region you are concerned with is crucial.

  • In industrialised countries, risk of relapse and reinfection after TB treatment low (approx 2%), so secondary prophylaxis is not required
  • In regions with high TB prevalence, risk of reinfection may be significant

With TB (and TB studies) you need to know whether people who later get active TB are getting a relapse of their previous infection, or being reinfected with a new virus. Again TB prevalence in the area studies is crucial. One study of mineworkers in South Africa showed that recurrent TB in HIV-positive workers was largely due to reinfection with a new TB virus [Sonnenburg et al, Lancet 2001;358:1687 & Lancet 2002;359:1619-1620].

Data showing benefits again has problems.

The Haiti study gave izoniazid or nothing (ie unblinded without a placebo), so when anyone appeared with possible symptoms in the no-treatment group, they erred on the side of caution and always diagnosed TB in every case – but this data still gets published in the Lancet.

The cynical approach to getting an establish medical career is to cause health care disaster, then write about it – then get published in the Lancet or New England Journal of Medicine…

click on either slide for a larger view

But even when there may be a benefit of secondary prophylaxis it is unclear which treatment should be used and for how long – and how to avoid general resistance etc?

Rates of TB by CD4 count show a direct relationship in all setting (ie Italy, US and South Africa) with CD4 count – both >200 and >350, so the best prevention for TB is to boost CD4. TB can occur at any CD4 count but the risk is greatly reduced and effect on TB incidence in countries using HAART is huge (up to 99% in the US).

Also works for other infections – therefore all programmes need to identify HIV infections and low CD4 – for all countries.

Therefore a summary of when to use secondary TB prophylaxis is in:

  • countries with significant TB
  • where there are HIV programmes
  • for all when CD4 <200 – as PCP
  • stop when CD4 has increased >200

Issues when treating TB in HIV-positive people

This is largely medical – for TB social factors like poverty, overcrowding, governments and political will etc are also important. With overcrowding for example, you need to know that if somebody sneezes into the open air, then TB droplets stay in the air for 1-2 days (aerosolised). If they cover their mouth then the droplets fall to the ground and are not infectious.

If TB is in the larynx then it comes out with each breath – so someone singing can be highly infectious, especially as you can be infected without symptoms for 1-2 years.

One of the major problems in treating HIV-positive people for TB, and which make TB different form many other OIs (PCP, CMV, lymphoma etc) where you would treat both HIV and the OI at the same time, is the interactions between TB drugs and HIV drugs.

Drugs are absorbed and processed in liver, where enzymes break drugs down (especially CYP3A4) – and the major enzymes breaking down nearly all drugs. One common example is that the enzyme for processing alcohol in reduced in Asian people – and that alcohol tolerance is therefore much reduced). Drugs are then eliminated in urine and faeces.

Rifampicin is a crucial TB drug – and you need to take for it 6 months – without rifampicin you need > 18 months treatment.

Rifampicin increases 3A4 to very high levels – (turning the liver from a Scoda to Lambourgini!) and as these are the enzymes used to metabolise PIs and NNRTIs, they are cleared really quickly from your body, and you only get a small fraction of the levels that you need.

Effects of rifampicin on HIV drugs:

saquinavir 80 % decrease
ritonavir 35 % decrease
indinavir 92 % decrease
nelfinavir 82 % decrease
amprenavir 81 % decrease
nevirapine 37 % decrease
efavirenz 26 % decrease

Reverse transcriptase drugs (nucleosides) are not affected.

However, given the benefits of reduced risk of TB, once CD4 is increased (as with other OIs) and that some people are diagnosed with very low CD4 counts, several HIV drugs may be possible:

  • use 3-4 RTIs (triple nukes are now not recommended, but 4 could be ok
  • sqv/rtv – some data 1000/100 (with 600mg rifampicin 3 days a week) and 1600/100 (with 600mg rifampicin daily) – but there is no long term data and in very small numbers, but is a possibility
  • nvp (including increasing nevirapin to 300BID) was studied in only 6 pts and would have serious concerns about stepping up dosing and risk of severe side effects
  • using EFV at higher dose of 800mg daily if body weight > 60kg may be possible

Other studies have suggested using rifabutin at reduced doses:

PowerPoint Slide

Three reports at the 2003 Retrovirus Conference also reported some success:

  • 109 HIV +ve patients with TB
    Only risk factor for TB relapse was low CD4 count
    Nettles R et al. 10th CROI, Boston MA, February 2003. Abs 137
  • 98 HIV +ve patients on 2 NRTIs + EFV + rifampicin
    Co-administration of EFV + rifampicin was well-tolerated and immunologically effective
    Patel A et al. 10th CROI, Boston MA, February 2003. Abs 138;
  • 98 HIV +ve patients on 2 NRTIs + EFV 600mg + rifampicin
    80% had TB resolution
    EFV 600mg was sufficient to treat HIV/TB patients on rifampicin
    Pedral-Samapio D et al. 10th CROI, Boston MA, February 2003. Abs 784

You can also make use of TDM to confirm drug levels of HIV drugs if you are able to get results back within few days or a week, especially for gender etc and NVP

The US suggest using rifabutin for HIV-negative patients with pulmonary TB may be as good as rifapentin but studies are still underway. Rifapentin is not broken down by CYP3A4 so you can use other ARVs, but there are other complications.

There are also additional ongoing concerns that the absorption of TB drugs may be different between HIV-positive and negative individuals, sometimes potentially higher. and sometimes potentially lower, though will little data for confidence.

Data on duration of treatment is similarly sparse. US say treat for one year but there is no data behind this – they say they ‘err .on side of caution’ but then need to analyse data. 4/6 studies showed an acceptable (<5%) relapse rate after 6 months. 2 showed relapse >9% but whether this was relapse of reinfection was not studied in one and was found to be a new infection in another.

Generally 6 months treatment is thought sufficient in uncomplicated and drug-sensitive cases, with longer courses for CNS disease, MDR disease or non-rifampicin regimens

Q- What if someone has HIV/CMV/TB etc and a low CD4 count

A-Priority for treatment is TB. It makes people feel dreadful and response to treatment is quick – they feel much better in 2-3 weeks.

Side effects of HIV treatment are increased with TB treatment:

22% stopped or modified due to peripheral neuropathy – most due to d4T – izoniazid causes PN and so does HIV – so with everything combined rates are much higher;

19% rash;

11% hepatitic toxicity (but not HCV related)

These figures are significant enough to stop drugs – and are very bad for future treatment,

Q-What advice to the global find to fight TB when there is no access to HIV treatment?

A-That if you just treat TB – and many coinfected people will not survive – this is only likely to spread MDR – the answer is to treat HIV.

Examples of TB regimens used are shown below:

Powerpoint Slide

When to start HIV treatment?

Even fairly recently, the arguments against starting both TB treatment and HAART both together at the same time (and therefore delaying HAART) were fairly convincing:

  • HIV is a chronic disease
  • Adherence may be compromised
  • Toxicity management is more complex
  • Immune restoration may produce “paradoxical reactions”

We used to wait for 2-3 months for TB drugs to work and then start HAART – but an important study [Dean et al, AIDS 2001] showed that 27/188 coinfected patients developed further AIDS complications – most of these people were not on HAART (24/27). Median CD4 count in this group was 70 cell/mm3 and 90% still have CD4 <100 after 4 months TB treatment. Of the 16/188 people died only 4 were on HAART. On the basis of this data we moved to treating all on HAART.

Guidelines (with little evidence) for timing of treatment is now therefore:

  • TB treatment must be started urgently
  • the urgency of HIV treatment therefore depends on CD4 count
    • o if <100 – start HAART asap (1-2 weeks after TB)
    • o if 100-200 – risk of progression is low – so start HAART at 2 mo when settled on TB treatment
    • o if >200 – don’t use HAART but monitor

Most decisions should be based on data but not always there. In Spain and Italy they say HAART can wait but wait but would treat with a low CD4 but they keep these people out of studies… A new UK study – looking at immediate treatment vs delayed by 2 months will give an answer.

In summary, treatment of TB/HIV coinfection is very complicated – for timing, duration, dosing and choice of both HIV and TB drugs.

The solution that many doctors choose – and it is not a stupid one – is to contact a specialist, such as Anton Pozniak at the Chelsea and Westminster Hospital.

You are dealing with a situation with complicated interaction of (at least) two potentially fatal infections, both of which may require immediate treatment, and the recommended first-line treatments for each interact with each other in a complicated way that could jeopardise the effectiveness of each.

In the UK there are few cases of coinfection so this is still possible and the structure within the NHS is strengthening the use of support networks in this way.

This is also recommended for us as advocates when we hear of cases of coinfection, which through policies such as dispersal of asylum seekers may be increasingly being managed by clinics with little experience of HIV and TB coinfection.

BHIVA guidelines for TB/HIV are about to be published.

IRIS syndrome – Immune Reconstitution Inflammatory Syndrome

This presents as an angry infected site – often in neck, lymph glands, back – but we have seen this is most parts of the body.

It is seen in people with TB – and starts with a swollen lymph node. This should not be operated on – if removed by a surgeon it can lead to a long term seeping wound – always refer to specialist.

The wound can also come back – and any puss should only be removed by needle.

IRIS involves:

  • worsening of original symptoms
  • US say no evidence of TB – we think probably TB and low CD4 count
  • high fevers in most (but other causes)
  • related to when starting ARV but not TB treatment – and may be related to starting early in respect to TB treatment – possibly at any CD4 but may be only those with low CD4
  • may respond to steroids? – US say ok – may improve symptoms but reverse when stopped.
  • In UK at Chelsea and Westminster we use IL-2 and Gm-CSF
  • Use recurrent aspiration – DON’T biopsy – don’t use plastic surgeon to remove scars afterwards (as nerves can be brought up into scar during the healing process)

Possible causes include:

  • over activity of immune system – but NOT of CD4 – an increased proliferation of peripheral blood mononuclear cells (PBMCs) and an interferon response to tuberculin antigen
  • possibly genetic factors – but this has been seen in all people/races
  • gene produce inflammation – with reduced cytokine THFA-309 etc
  • increased levels of IL-6 are seen (but unclear whether cause or effect)

Multi-drug Resistant TB (MDR –TB)

Factors responsible for MDR outbreaks include:

  • when control programmes are not good (programmes disinvested in TB in 1982)
  • inadequate adherence
  • infection control procedure breakdown – with coinfection you breathe in same bugs but are more likely to get TB (10% risk a year with HIV rather than 10% over lifetime with HIV-negative)
  • immuno-suppressed convergence in many places

An example was given of an outbreak of MDR in Buenos Aires

162 patients in HIV unit and MDR outbreak lasted over 2-3 years

Minimum cost to manage an MDR patient in the UK is >£60,000

These patients were on an open ward (nightingale ward) and were infected by a non-adherence nurse with TB

  • 87 died without knowing they had TB
  • 49 died on standard TB drugs
  • 10 died on tailored TB drugs
  • only remained16 alive on tailored TB drugs

These are terrible results.

102 were resistance to 6 or more TB drugs

8 resistant to >10 TB drugs (and most doctors cannot name 10 TB drugs)

Using genetic markers on 92 cases – showed that 77 of these 92 were from index case

So how do we control TB (other than by running fast and holding your breath)?

In Buenos Aires contact tracing was through publishing names in newspaper!

In the UK this is more likely to be by visits from healthcare workers.

Link to top of page

This entry was posted in Presentations. Bookmark the permalink. Both comments and trackbacks are currently closed.