HIV-associated dementia has been described using a variety of terms including AIDS dementia complex (ADC), HIV encephalopathy, HIV-1-associated cognitive / motor complex, encephalitis and brain disease. It is thought to be due to the direct effects of HIV upon the central nervous system (CNS): the brain and spinal cord.

HIV and the brain

It is clear that the HIV usually enters the CNS very soon after infection. Around the time of seroconversion, some people with HIV develop various short-lived neurological problems such as encephalitis (inflammation of the brain), meningitis (inflammation of the meninges, the membranes which line the brain) or neuropathy (damage to the nerves). These syndromes are thought to be due either to HIV itself infecting cells, or to the immune system's response to HIV infection. They usually get better on their own. Although HIV-related damage is reversible, encephalopathy may cause irreversible damage and lasting disability.

Among asymptomatic people, there may be some mild abnormalities in the cerebrospinal fluid (CSF), the fluid that surrounds the brain and spinal cord. These include increased levels of white blood cells, protein or antibodies, as well as the presence of HIV itself. These do not usually result in any major problems or symptoms (McArthur 1989; Miller 1990; Karlsen 1993; Samuelsson 2006), although they may cause a slight impairment in speed of information processing (Baldewicz 2004). It appears that in the early stages of infection, the immune system is able to control HIV replication in the brain.

In the absence of anti-HIV treatments, dementia occurs most commonly among people with advanced disease, with a CD4 cell count below 50 cells/mm³, and often progresses rapidly. Among people taking highly active antiretroviral therapy (HAART), dementia or the early stages of neurological impairment may emerge when CD4 cell levels fall below 300 cells/mm³ and progress slowly. Alternatively, neurological functioning may improve or stabilise during treatment with HAART.

HIV-related neurological problems are common in children with AIDS. If untreated, HIV infection in children can lead to major developmental delays and problems in understanding, co-ordination and movement.

How HIV causes dementia

There remains some uncertainty about precisely how HIV in the CNS causes dementia. The leading theory suggests that HIV probably enters the brain within infected monocytes or lymphocytes, further infecting macrophages and brain cells such as astrocytes, microglia and oligodendrocytes.

In the presence of HIV, immune cells within the brain become over-active. They produce immune system messengers called cytokines or chemokines which can inhibit the growth and survival of brain cells, or trigger cell death. A cytokine called oncostatin M may be the most damaging of the cytokines, although it acts in concert with other cytokines. Low levels of transforming growth factor β 1 may enhance brain cell damage in people with HIV infection. Elevated levels of the chemokines MIP-1B, and low but detectable levels of MIP-1a, have been associated with dementia, although high levels of MIP-1a may protect against dementia (Ensoli, Letendre 1999). It has also been suggested that platelet-activating factor (PAF) may also contribute to neuronal damage in HIV-infection.

Various HIV genes such as tat and nef, and the gp120 protein in HIV's envelope, may have direct toxic effects upon cells. However, some researchers now believe that the HIV Tat protein has an indirect effect on the brain. Test tube studies have found that when a particular type of macrophage is exposed to Tat, it produced a soluble factor which produced profound brain degeneration including loss of neurones (Brana 1999).

Other research suggests that infected brain cells such as astrocytes appear not to function properly during HIV-associated dementia. Postmortems on three people with HIV have found that HIV evolves separately in different parts of the brain and that some strains of HIV are particularly suited to certain areas of the brain (Shapshak 1999).

Oxidative stress is one explanation for a number of neurological diseases including HIV-related dementia. There is evidence that infected cells express superoxide and nitric oxide which interact to produce a chemical called peroxynitrite that is toxic to the brain. One study found higher levels of these chemicals in demented versus non-demented HIV-infected people.

There is no evidence that a diagnosis of Kaposis sarcoma reduces an individuals future risk of developing HIV-associated dementia or other central nervous system disorders.

Other causes of neurological problems

Neurological problems in people with HIV can have a variety of causes other than HIV itself. Apart from the ways in which HIV itself can directly affect the nervous system, infections and tumours can occur in the nervous system when HIV damages the immune system. These include Cryptococcus, cytomegalovirus, Hodgkin's disease, neuropathy, non-Hodgkin lymphoma (NHL), progressive multifocal leukoencephalopathy (PML) and toxoplasmosis.

People with HIV can also develop a number of other neurological disorders, including brain damage due to lack of oxygen resulting from lung infections, liver dysfunction or clotting within the veins. Strokes may also complicate the course of AIDS.

Likewise, many treatments may cause mental disturbances or lack of alertness. People with AIDS may be particularly sensitive to the side-effects of anti-psychotic drugs and others which are active within the central nervous system. The following common treatment drugs can sometimes cause various neuropsychiatric effects, including depression, irritability, delirium, confusion and hallucinations: aciclovir (Zovirax), interferon alfa (IntronA / Roferon-A / Viraferon); amphotericin B (Fungilin / Fungizone / Abelcet / AmBisome / Amphocil); AZT (zidovudine, Retrovir), corticosteroids, co-trimoxazole (Septrin), efavirenz (Sustiva), ganciclovir (Cymevene), isoniazid, methotrexate, pentamidine (Pentacarinat), procarbazine, vinblastine (Velba) and vincristine (Oncovin).

Furthermore, symptoms of depression or severe anxiety, such as confusion and forgetfulness, may be similar to symptoms of dementia.

Symptoms of HIV-associated dementia

HIV-associated dementia begins with slight changes in behaviour, intellect and co-ordination. Friends and family notice forgetfulness, changes in personality and symptoms normally characteristic of depression, such as loss of appetite and motivation. Tasks that require concentration or complex thought become difficult. These symptoms are interspersed with sharp mood swings or mania. Motor skills gradually deteriorate: hands become clumsy, movement is slow and unsteady and eye movements become jerky.

Cognitive impairment is the term used to described difficulties with thinking, problem-solving and concentration while neurological impairment is a general term which refers to the symptoms of dementia affecting thought, movement and personality.

Post mortem examinations have shown that HIV-associated dementia involves the loss of the outer coating (myelin) from nerve cells and the accumulation of various cells in the brain matter. HIV itself can be detected in the brain in various blood cells (macrophages and monocytes), brain cells (microglia, oligodendrocytes and astrocytes) and the cell walls of capillaries. HIV is not uniformly distributed throughout the brain. Areas of the brain with high levels of HIV include the basal ganglia and the hippocampus.

Diagnosis of HIV-associated dementia

In the early stages of HIV-associated dementia, standard neuropsychological tests can detect subtle changes in memory, concentration and information processing. In the later stages, HIV-associated dementia can be diagnosed from its symptoms. Abnormalities in the CSF may include an excess of white blood cells, protein and immunoglobulin G (IgG) and the presence of HIV and the HIV protein p24. Brain scans using computerised axial tomography (CAT) scans and magnetic resonance imaging (MRI) can reveal distinctive signs but often results are inconclusive.

HIV viral load in the CSF roughly correlates with the severity of HIV-associated dementia among people with CD4 counts below 200 cells/mm³, but at higher CD4 counts there is no clear correlation between HIV viral load in the CSF and cognitive function. Viral load in the CSF may also be increased by infections such as cryptococcal meningitis.

Incidence of HIV-associated dementia

Estimates of the incidence of HIV-associated dementia vary, depending on the definition of dementia being used. Researchers from John Hopkins University have estimated that 15% of HIV-infected people may develop dementia and that 30% may show some signs of neurological impairment (McArthur 1999). In Uganda, around 11% of HIV-positive patients have dementia, although this rises to 50% in patients with CD4 cell counts below 200 cells/mm³.

HIV dementia became less common after the introduction of AZT. One British study found that only 2% of people with AIDS taking AZT developed HIV-associated dementia between 1982 and 1988, compared with 20% of people not taking AZT.

However, whether the widespread uptake of HAART has reduced the number of new cases of HIV-associated dementia is less clear. Some studies have shown reduced incidence and prevalence of dementia since HAART became available (Brodt 1997; Maschke 2000; Sacktor 2001) but other studies have failed to confirm this (Dore 1999, 2002; Ives 2001). Despite this uneven evidence, it is clear that HAART can slow or reverse neurological damage due to HIV-associated dementia.

Higher viral load in the blood and the CSF, and lower CD4 cell count have been associated with a greater risk of HIV dementia. People who have used cocaine or amphetamines, have injected drugs, or with a history of alcohol abuse or dependence, may be more likely to develop dementia (Maschke 2000; Zhang 1998; Nath 2001; Bouwman 1998; Dougherty 2002; Becker 2004). A number of studies have indicated that older age is associated with an increased incidence of cognitive impairment and dementia in HIV-positive patients (Becker 2004; Cherner 2004; Valcour 2004a,b).

Dementia in the age of highly active antiretroviral therapy

The argument that the course of HIV dementia has been significantly changed by HAART rests on the premise that damage to the brain is only slowed, but not prevented. Although antiretroviral therapy can improve cognitive and motor function in people with dementia, it does not eradicate HIV dementia nor does it prevent HIV dementia from developing. Thus some researchers have argued that HIV dementia has become a protracted disorder rather than an acute dementing disease (Kusdra 2002). Nevertheless, patients with cognitive impairment are still 2.5 times more likely to die than those without, even while taking HAART (Tozzi 2004).

Evidence that HIV dementia now occurs earlier in HIV disease provides further support for the argument that HAART has changed the nature of HIV dementia. People may now be at greater risk of dementia when their CD4 cell counts are above 100 or 200 cells/mm³. An Australian study found that the average CD4 cell count of people developing dementia as their first AIDS-defining condition in 1997 was 170 cells/mm³ compared to 70 cells/mm³ between 1992 and 1995 (Dore 1999, 2002). American researchers have also suggested a larger proportion of new cases of HIV dementia may be occurring among people with CD4 cell counts between 200 and 350 cells/mm³ (Sacktor 2001).

However, not all research supports this reformulation of HIV-associated dementia. For instance, a comparison of an HIV cohort from 1994 / 1995 and another from 1998 / 1999 found no difference in the occurrence of HIV dementia or specific neurological abnormalities when controlling for factors such as CD4 cell count and age (Sacktor 2002). There is also evidence from a retrospective study that progression of HIV dementia was actually highly variable before the advent of HAART (Bouwman 1998).

In contrast to the view that dementia may develop slowly despite antiretroviral therapy, it is possible that maintaining a strong immune system will protect against HIV dementia. If the immune system plays a key role in controlling HIV replication in the brain and thus preventing dementia, it is plausible, but unproven, that maintaining a healthy CD4 cell count will reduce the risk of dementia, regardless of which drugs are used.

The effect of antiretroviral therapy on the relationship between age and dementia remains unclear. Although age was associated with dementia in a cross-sectional analysis of HIV-positive people in the United States regardless of antiretroviral treatment, an Italian study has found that age is only associated with an increased risk of dementia in patients not taking antiretroviral therapy (Lorenzini 2004; Valcour 2004a,b). In contrast, one study of Italian patients taking antiretroviral therapy did find a significant influence of age on the incidence of dementia (Antinori 2004).

Treating dementia with highly active antiretroviral therapy

It was initially feared that HAART would not be effective against HIV in the brain, because only very low levels of some of the most potent drugs cross the blood-brain barrier. However, recent research has shown that anti-HIV combination therapy can reverse the signs of HIV-associated dementia in up to 60% of patients. This may partly be a consequence of improvements in the strength of the immune system throughout the body, and partly the result of anti-HIV effects in the brain and CSF.

When a person with HIV-associated dementia starts taking HAART, viral load in the fluid around the brain as well as in the blood may decline (Lanier 2001). Immune activation in the brain may normalise and levels of neurotoxins fall producing improved neurological function (Gendelman 1998; Chang 1999; Kusdra 2002). Normalised blood flow in the brain may also play a role in improved neurological function (Herning 2001).

Several studies have shown clinical improvements in many people with HIV-associated dementia after starting HAART or switching to a new regimen. An Italian study found that 81% of 26 people with advance HIV disease had neurocognitive impairment before treatment, but this fell to 50% and 22% after six and 15 months of treatment with an antiretroviral combination (Tozzi 1999). Functions such as memory, concentration and mental processing improved following treatment. An American team reported that 60% of 30 people treated with HAART showed neurological improvement, although the other 40% had neurological progression (Dougherty 2002).

One case study found that a treatment-naive woman with severe HIV-associated dementia had most of her cognitive and motor deficits resolve after taking AZT, 3TC (lamivudine, Epivir), indinavir (Crixivan) and the anti-inflammatory painkiller ibuprofen for six weeks. Tests found that inflammatory neurotoxins were suppressed by the treatment.

Combination therapy improves thinking and concentration to a greater extent that single drug therapy in dementia patients (Sacktor 1999). However, it is the nucleoside analogue reverse transcriptase inhibitors (NRTIs) such as abacavir (Ziagen), d4T (stavudine,Zerit) and AZT which are able to cross the blood-brain barrier which are thought to provide much of the benefit to people with HIV dementia.

Much of the evidence about the effectiveness of anti-HIV drugs against dementia relates to AZT, not least because for many years it was the only available anti-HIV drug. AZT can prevent and improve HIV-associated dementia, especially when given at higher doses such as 1000mg per day, although lower doses can also be effective. Placebo-controlled and observational trials have found that very few people develop dementia while they are taking AZT.

For those who start AZT therapy after the onset of dementia the degree of improvement depends on how much damage has been done, and so people with very advanced CNS disease may not benefit.

Other than AZT, a number of anti-HIV drugs do cross the blood-brain barrier successfully. Abacavir crosses the blood-brain as well as AZT and may benefit people with dementia. One study of abacavir in people with HIV-associated dementia found that the addition of abacavir to HAART did not produce clinical or virological benefits. However, this non-significant result has been attributed to the study design which did not allow for the full benefits of HAART to stabilise before abacavir was added (Lanier 2001).

d4T has also been shown to reduce neurological symptoms in people who have previously taken AZT (Arendt 1998). The addition of d4T consistently reduces HIV viral load in the CSF to undetectable levels, whereas protease inhibitors have a variable effect (Gisolf 2000). ddI (didanosine, Videx / VidexEC) monotherapy also has some activity against dementia (Yarchoan 1990; Wolters 1991).

The non-nucleoside reverse transcriptase inhinitors (NNRTIs) nevirapine (Viramune), delavirdine (Rescriptor) and efavirenz (Sustiva) cross the blood-brain barrier reasonably well, but their effectiveness at preventing or treating dementia has not been established. One study found that combinations including efavirenz can suppress HIV to undetectable levels in the CSF, while another reported neurological improvement in dementia patients taking efavirenz- and nevirapine-based combinations (Tashima 1999; von Giesen 2002). The protease inhibitor indinavir also crosses the blood-brain barrier to some extent.

However, there is still uncertainty surrounding the importance of including brain-penetrant drugs in a HAART regimen. For example, in one study, improvement in cognitive impairment was shown to be associated with the degree of virological suppression in the CSF and the number of brain-penetrating antiretroviral drugs in a HAART combination (Letendre 2004), while a similar study failed to reveal the same association (Antinori 2004). Another recent study has shown that patients switching to a new HAART regimen experience an improvement in neurological symptoms regardless of whether it contains at least one brain-penetrant drug or not (Robertson 2004).

Other treatments

Some theories suggest that the HIV envelope protein gp120 may have toxic effects in the brain, so drugs which block the binding of gp120 to the CD4 molecule on human cells are being studied. One such drug is peptide T which was reported in a preliminary study to improve cognitive problems. However, larger controlled studies found no clear evidence of any beneficial effects of the drug, although an adjusted subgroup analysis left open the possibility of some benefits for people with higher CD4 cell counts or a greater degree of cognitive impairment before treatment began. See Peptide T in Drugs used by people with HIV: Entry and fusion inhibitors for research details.

An experimental drug called lexipafant has improved the memories of people with HIV-related cognitive dysfunction in a small, preliminary study. Larger, long-term studies are being planned (Schifitto 1999).

Similarly, selegiline patches may also improve neurological impairment in individuals with HIV-related dementia. Two pilot studies of selegiline patches found improved verbal memory and psychomotor speed test results, although other indicators did not improve (Sacktor 2000).

Trials have also studied nimodipine (Nimotop), a drug that changes the level of calcium in the brain. Calcium levels are thought to be disrupted by gp120. However, an American trial, ACTG 162, was stopped in 1995 after no benefits of the drug were seen.

Another theory is that excessive levels of the cytokine tumour necrosis factor α (TNFα) cause brain problems. Drugs that reduce the levels of TNFα, such as oxpentifylline (Trental), are also being tested.

It is also possible to treat some of the symptoms of HIV-associated dementia. People who have slowed reactions may benefit from psychostimulants such as methylphenidate (Ritalin). People who are agitated or manic can be treated with neuroleptics such as risperidone (Risperdal), starting with low doses.