Mycobacterium avium and M. intracellulare are very similar bacteria which are usually grouped together. In the United States, they are usually described as Mycobacterium avium complex, or MAC. They belong to the same genus as the organism which causes tuberculosis (TB; M. tuberculosis).

By the early 1990s, MAI was the most commonly reported bacterial infection among people with AIDS in the United States. One study found that 43% of people who survived two years after being diagnosed with AIDS had detectable MAI organisms in their blood (Nightingale 1992). Since the introduction of highly active antiretroviral therapy (HAART), the incidence of MAI has fallen significantly among HIV-infected people, and people who have been diagnosed with MAI live longer.

MAI organisms are found in the soil, including potted plant soil, and tap water throughout the world. They are generally thought to enter the body in food, water or inhaled dust. The organisms are probably not transmitted from one person to another.

In people with healthy immune systems, MAI rarely causes illness. However, in people with HIV who have CD4 counts less than 100 cells/mm³, and often below 20 cells/mm³, MAI can spread widely or 'disseminate' throughout the body, affecting almost any organ, but especially the liver, spleen and bone marrow. They mainly infect cells of the reticuloendothelial system, a group of related cells in the body that includes macrophages.

Symptoms

Disseminated MAI organisms spread widely throughout the body, infecting the blood, spleen, liver, bone marrow and lymph nodes. This may cause high fevers, severe anaemia, night sweats, chills, weight loss, loss of appetite and weakness. If the gut is involved, symptoms can include chronic diarrhoea with malabsorption of nutrients, and abdominal pain due to ulcers. If the organisms have spread to the lymphatic system, symptoms can include enlarged lymph nodes, liver and spleen. MAI can occasionally cause mouth ulcers.

Laboratory tests often show that people with MAI are anaemic, have low levels of platelets in their blood (neutropenia) and high levels of protein, including the liver enzyme alkaline phosphatase.

Diagnosis

The symptoms of MAI can have many possible causes, so MAI is usually diagnosed by testing whether MAI organisms can be detected in the blood. If the bacteria can be found in the blood, the individual is described as having MAI bacteraemia. MAI can also be diagnosed by trying to detect the organisms from samples of the liver, bone marrow, lymph nodes, cerebrospinal fluid, spleen or rectum.

Growing, or 'culturing' the bacteria in the laboratory can take several weeks. However, newer tests, called radiometric assays, can give results in several days. A test called an acid-fast smear can detect the presence of mycobacteria within minutes, but cannot tell whether the bacteria are MAI or M. tuberculosis. New polymerase chain reaction (PCR) tests that look for MAI's DNA are currently being evaluated.

Risk factors

People in whom MAI organisms can be detected in the faeces or saliva seem to be at increased risk of developing MAI. However, this is not a reliable way of diagnosing systemic MAI since two-thirds of people with MAI never have detectable organisms in their faeces or saliva.

A study of risk factors for MAI has found that people who drank bottled water more than three times per week had a greater chance of developing MAI than those who drank tap water, although tap water was still a risk factor (Anwar 1998). Another study found that tap water was not a risk factor: instead, consumption of mineral water and raw seafood were associated with risk of MAI (von Reyn 2002). However, given that M. avium is common in the environment, avoiding certain foods or mineral water is probably not protective against MAI infection.

Prophylaxis

Current United States guidelines recommend that HIV-positive people whose CD4 cell count is below 50 cells/mm³ should start to take clarithromycin (Klaricid / Klaricid XL) or azithromycin (Zithromax) as MAI prophylaxis, or rifabutin (Mycobutin) if neither of these drugs can be tolerated. This recommendation is made on the basis of evidence that people who take MAI prophylaxis live longer than people who do not take prophylaxis. If MAI does develop while someone is on prophylaxis, they should switch to a combination drug treatment regimen.

In the United Kingdom, many doctors do not routinely prescribe MAI prophylaxis. They express concern that people receiving clarithromycin or azithromycin as prophylaxis may develop resistance, resulting in attacks of MAI disease that may be hard to treat. They also question whether any of the drugs' efficacy in preventing or delaying the actual symptoms of MAI (rather than MAI bacteraemia) or prolonging survival has been clearly established. There have also been reports that people who are exposed to TB while receiving rifabutin may develop TB strains that are resistant to rifampicin, a drug that is a mainstay of TB treatment regimens.

Guidelines published by the United States Public Health Service recommend that people can stop MAI prophylaxis with little risk of relapse if they have sustained immune recovery due to highly active antiretroviral therapy (HAART). Specifically, people who have never had MAI may cease prophylaxis when they have a CD4 cell count above 100 cells/mm³ for over three months. Maintenance therapy or 'secondary prophylaxis' may be stopped when the CD4 cell count has been above 100 cells/mm³ for six months, and the patient has completed twelve months of MAI therapy and is free of symptoms of MAI. The guidelines recommend people who have previously had MAI should restart prophylaxis if their CD4 count falls below 200 cells/mm³.

A large American study found that stopping azithromycin was safe, although two cases of MAI occurred among people not on active MAI prophylaxis. Participants had recorded previous CD4 cell count low-points below 50 cells/mm³, but on entry to the study their average CD4 cell count was over 200 and 63% had viral loads below 500 copies/ml (Currier 2000).

Rifabutin was licensed for prescription as prophylaxis against MAI in 1993. Two clinical trials showed that a dose of 300mg per day was able to delay the development of MAI bacteraemia in people with CD4 counts below 200 cells/mm³. Rifabutin recipients were also less likely to develop fever or fatigue than people given placebo, although there was no difference in other possible MAI symptoms such as night sweats, abdominal pain or diarrhoea. A subsequent analysis of the pooled data from these trials also suggested that rifabutin prophylaxis prolonged survival: after one year in the trials, 28% of the placebo recipients had died, compared with only 20% of those taking rifabutin (Moore 1995).

Clarithromycin was licensed for MAI prophylaxis in the United States in December 1995 but has not been approved as a preventive treatment for MAI in Europe. An international trial found that it reduced the risk of developing MAI by over two thirds compared with placebo and produced a survival advantage (Pierce 1996). Two studies which have found that clarithromycin is more effective than rifabutin at delaying MAI bacteraemia or MAI disease among people with CD4 cell counts below 100 cells/mm³, but they did not report a survival advantage (Hewitt 1999; Benson 2000). Combining both clarithromycin and rifabutin was not significantly better than clarithromycin alone, but caused more side-effects.

However, of the people who did develop MAI despite taking clarithromycin, over half had developed clarithromycin-resistant strains of MAI. Other trials have suggested that combination regimens such as clarithromycin plus clofazimine or ethambutol may have increased efficacy.

Azithromycin was licensed for MAI prophylaxis in the United States in June 1996 but has not been approved as a preventive treatment for MAI in Europe. One study showed that azithromycin was more effective on its own or in combination with rifabutin than rifabutin monotherapy at preventing MAI bacteraemia (Havlir 1996). It can be taken in a convenient once-weekly dose. However, 11% of people who developed MAI despite taking an azithromycin-containing regimen had resistant MAI strains. MAI strains that are resistant to azithromycin are also resistant to clarithromycin, and vice versa.

A trial of clofazimine monotherapy was stopped as it appeared to have no effect in preventing MAI.

Treatment

Prior to the introduction of antiretroviral treatment for HIV, anti-MAI drugs seemed to be unable to eradicate the organisms from the body completely although they did reduce the amount of bacteria in the blood, alleviate symptoms and improve survival. Since the introduction of HAART, the prognosis of people diagnosed with MAI has improved and complete eradication of MAI from the body has been documented (Aberg 1998). A blood culture test can be used to determine whether MAI bacteria have been eradicated from the body. Treatment for MAI is lifelong unless sustained immune recovery due to HAART occurs.

MAI treatment should generally consist of a combination of two or more drugs, one of which should be clarithromycin or azithromycin. Ethambutol is commonly the second drug prescribed in addition to azithromycin or clarithromycin. Side effects of clarithromycin and azithromycin are mainly gastrointestinal such as nausea and diarrhoea, and usually mild and short-lived.

MAI organisms can become resistant to clarithromycin when the drug is used on its own, and people with resistant organisms are more likely to experience a resurgence of MAI. Doses higher than 500mg twice daily are not recommended as they have been associated with poorer survival in several clinical trials.

In the United Kingdom, treatment combinations also often include rifabutin. It causes relatively few side-effects, such as reductions in white blood cells and platelets, rashes, gastrointestinal upsets and darkening of the skin. There have also been reports of eye inflammation (uveitis) among people receiving high doses. Drug interactions between rifabutin and antiretroviral drugs may make dose adjustments necessary.

Other drugs used in combinations include rifampicin (Rifadin / Rimactane), gentamicin (Cidomycin / Genticin), amikacin (Amikin), ethambutol, ciprofloxacin (Ciproxin) and sparfloxacin. Ethambutol and clarithromycin are commonly prescribed together because they each enhance the strength of the other.

Another drug, clofazimine, has been used in combinations, but in July 1996 American researchers warned against this drug after a study suggested that it offered no added benefits, and was associated with an increased risk of death.

Mycobacterium avium intracellulare in the age of highly active antiretroviral therapy

MAI has become less common since the widespread use of HAART. However, people who have recently started HAART may be at risk of an unusual form of MAI, characterised by inflamed, swollen lymph nodes and fever (Hassell 2001; Aberg 2002). These symptoms are thought to be caused by the recovering immune system over-reacting to low levels of MAI infection that were present before treatment began. One study found that localised MAI during the first few weeks of HAART occurred among people with CD4 counts up to 210 cells/mm³ (Phillips 2002). Other manifestations of MAI during immune reconstitution include a cavitary lung lesion and, rarely, a painful nodule under the skin.

Several studies have reported no recurrences of MAI among people with immune recovery who stopped prophylaxis (Gill 1998; El Sadr 2000; Furrer 2000; Rossi 2001). At the population level, HAART itself has been associated with a 73% reduced risk of MAI but MAI prophylaxis in addition to HAART does not further reduce the risk of MAI (Mary-Krause 2000). Nevertheless, there is some evidence that prophylaxis may continue to provide some benefits. A small proportion of people do have a recurrence of MAI after they stop prophylaxis, even in the presence of a CD4 cell count above 100 cells/mm³ (Currier 2000; Kirk 2002; Zeller 2002). A Canadian study has found that azithromycin prophylaxis still has value for people on HAART with CD4 counts below 75 cells/mm³, but that as CD4 cell counts rise on HAART, prophylaxis may be ceased (Phillips 2002).