HIV dementia: are viral or host factors driving differences between subtypes?

Although researchers have identified differences between HIV subtypes in their potential for causing disease in the brain, findings presented at two recent international meetings on HIV-related neurological problems suggest that differences between individuals’ immune responses – the `host environment` for the virus – might also determine why some populations appear to suffer more frequently from HIV dementia. Some of these differences could be inherited, while some may be the result of HIV infection or other comorbidities (such as HCV).

For example, some of the viral variants that evolve over the course of infection and that may be involved in the pathogenesis of HIV dementia may lie in wait, and only start to cause neurological problems when there are changes in the host environment.

For example, viruses containing an HIV envelope variant identified in research presented by Dr Dana Gabuzda of the Dana Farber Cancer Institute appear to be highly susceptible to neutralising antibodies. As long as there are high levels of these antibodies, these viral variants appear to be kept in check. Thus, the variant that is associated with HIV dementia might only become more prevalent once the body’s immune surveillance (and antibody production) begins to fail.

Brian Wigdahl, of the Drexel University College of Medicine, presented research focusing on other changes that could occur to the virus over time in a person, which could also grow more important in the context of immune dysregulation.

In particular, he identified variations in the long terminal repeat (LTR) of HIV in a number of clades, that, when stimulated by high levels of inflammatory cytokines common in later stage disease, preferentially stimulate viral transcription in monocytic cells. These variants were found in roughly a quarter of the LTRs derived from demented individuals but seldom if ever found in LTR’s derived from a non-demented population.

One thing that became apparent during Wigdahl’s talk, however, was that different subtypes of viruses could make different changes at different positions on the LTR, which could functionally serve the same purpose.

“During the course of disease, there are different selective pressures that affect the spectrum of genotypes in people,” said Wigdahl. These viruses may “be waiting for an opportunity or a change in the physiology of the host to perhaps play a role in the pathogenesis of disease. The LTR [variant] that I talked about, we think is an LTR that was selected for to evade the immune response that now can take advantage of end stage disease, by using cytokine dyregulation, i.e, higher levels of IL-6 to promote replication. That’s just one example from the past but I think there are going to be many other examples like that.”

Ultimately, given time, the virus, whatever subtype, could evolve to exploit the particularly weaknesses of its host.

While it is very important to keep an eye on the virus, any explanation of how the virus causes disease is probably incomplete without a better understanding of its host. One possibility that wasn’t really fully explored at this meeting was that there could be significant differences in host susceptibility to HAD or MCMD that differ from one population to another.

Dr Sunil Ahuja of the University of Texas described how HIV pathogenesis could be affected by genetic differences affecting the expression of CCR5 and its ligands (such as MIP-1-alpha-P). High levels of MIP-1-alpha have been associated with lower risks of dementia in some studies, however, as yet there are few data to show whether population-based differences in the expression of this ligand specifically affect the risk of neurological disorders.

Dr. David Haas presented data suggesting that certain inherited polymorphisms in mitochondrial DNA (which are maternally inherited), appear to be associated with a much greater risk of developing neuropathy during nucleoside analogue treatment. One polymorphism in particular, which is associated with mitochondrial haplogroup T, was found among patients with neuropathy though there could well be others.

Going back to the clinic

However, given how little is known about the existence or importance of such host factors, is it possible to say that there are any clinical difference between clades?

According to Dr Ellis, it may be difficult to compare any of these studies side by side, because, in addition to the possible viral determinants of neuropathogenesis, there could be a number of significant confounders in these studies, including the differences in host susceptibility, and co-infections and other morbidities.

He described one hypothetical case of how confounding factors could make it difficult to compare clades. For example, Clade B and C viruses could coexist within the same geographical region, but perhaps Clade C was introduced into the injecting drug-using (IDU) population. This population might be somewhat older (e.g., age 45 + years), be less educated, and have a higher prevalence of hepatitis C coinfection (HCV has recently been shown to enter the CNS and may increase neurological problems).

Clade B on the other hand may be more common among men who have sex with men, who tend to be somewhat younger (~28 years old), have more education and a lower HCV coinfection rate. Such major differences between the hosts could make it extremely difficult to ferret out whether there are major differences between clades. And when the comparisons are across studies and different geographical regions, these confounders would only be compounded.

Also, the studies that have been used to date have rarely used the exact same methods to assess neurological function. One common problem, pointed out by Dr Clifford as well, is that when introducing a test into a different setting, not only does it have to be translated into the local language and adapted to the culture, the test has to be performed in well-matched controls in order to establish what the local norms are.

Dr Ellis shared some more recent data from an HRNC-sponsored study in Pune, India demonstrating the importance of establishing local norms. When using neuropsychiatric battery of tests that had not yet been adapted for the setting, the Indian subjects naturally scored lower on some indicators than US populations would: 91% of the HIV-infected patients appeared to be impaired!

But once the test was standardised to the local population, 47% of the HIV-positive subjects vs 14% of the HIV-negative subjects were neurologically impaired, which is quite similar to rates seen in the US.

Notably these rates are higher than those reported from India previously — although earlier at the conference Dr Manisha Ghate, Assistant Director of the National AIDS Research Institute in Pune also reported the results of a pilot study using a battery of neurocognitive tests in 30 HIV-1C infected patients (along with age and education matched HIV-negative controls).

“The results suggest that cognitive difficulties are present among individuals with the clade C virus in India, with as many as 56% of the patients with advanced HIV meeting the criterion for impairment in two cognitive domains,” she said.

The jury is still out on the subject, however, and future plans include a collaboration with the US National Institutes of Mental Health and researchers in Pune to conduct a 600 patient study to determine the prevalence and nature of HIV-associated neurocognitive impairment (HCNI) in individuals who are infected with HIV-1C and to determine what the impact of ART on HNCI in this population might be (by treating those with less than 200 CD4 cells, and randomising those with CD4 cells between 200-500 to ART or no ART). The study will also assess the viral genetics associated with HNCI in people with HIV-1C in India.

Dr. Ellis noted that the HNRC is also developing a common protocol to be adapted for use at multiple sites across the world, including India, Brazil, China and Romania — places where multiple clades co-exist in similar populations.

In addition, the NIH is putting out requests for proposals to fund research into brain disorders in the developing world, for collaborative research between high income and low to middle-income country scientists. Two types of grants will be available, R21 planning grants, which should help researchers develop capacity and establish local norms, and major R01 grants open only to those who have previously been R21 grantees.

Kanmogne in Cameroon is one of the first recipients of one of the grants. Her initial goals are to normalise and adapt neuropsychiatric tests currently used by HNRC to diagnose HAD, MCMD, and neurological impairments or deficits in the local population. The project will also conduct molecular studies of isolates from the patients with neuroAIDS, looking for molecular determinants of susceptibility to neuroAIDS, and to see whether there is a relationship between viral genotype/CRFs/URFs and susceptibility to neuroAIDS.

She and other researchers from West Africa (namely Nigeria, Senegal, and The Gambia) also announced the launching of the West Africa NeuroAIDS Consortium, which hopes to foster collaboration to take better advantage of the opportunities for neurological research relevant to populations living with HIV/AIDS in West Africa. Some of the particular strengths of the region, according to Dr David Clifford, who led a lunch time discussion with numerous presentations from potential Consortium participants, include the existence of well characterised and studied cohorts of people living with HIV-2, as well as several very large and well-funded programmes where HAART is being introduced to thousands of people with HIV.

Prasad, who has a role as a mentor in the programme, said that although HIV-2 is less likely to lead to rapid progression, it is quite a mature epidemic, and there are several reports of neurological problems now developing in some patients with HIV-2. He noted that, although, for the most part, HIV-2’s tat protein is usually the same as most non-subtype C HIV-1, the presence of the cysteine31 to serine tat mutation is higher in HIV-2 than most of the HIV-1 subtypes. So one proposed collaborative project would be to identify whether this mutation is associated with less dementia, and to test isolates from these patients in the SCID mouse model.

Given the great diversity of HIV in the region, this could just be a launching pad, once the actual incidence of HAD and other HIV-related neurological problems are better described in the region.

“I think these studies are critically important,” said Dr Ellis, “and they represent an opportunity to increase our awareness of the importance of brain disease due to HIV infection. “

Related articles

Part one - Are some subtypes of HIV more likely to cause neuroAIDS than others?

Part two - HIV dementia: why might some viruses cause more problems than others?