This page has moved. Click here to view page.

 

AIDS Dementia Complex

HIV/CNS infections: it appears that infections of the nervous system - we see this particularly in the spinal fluid but probably the same thing is true in the brain - that it’s an intrinsic part of the ecology of HIV. One of the things that is puzzling about that is that early infection for a protracted period of time, despite this infection or exposure to the nervous system, it is entirely asymptomatic, even though we might find abnormalities in the spinal fluid. Only under certain conditions late in AIDS dementia complex

There are two variants that we encompass within the definition of AIDS dementia complex, and sometimes they are separated out. One is the dementia or the brain problem, the cerebral problem, which is a combination really of cognitive and

Myelopathy. This syndrome is a coherent syndrome that has a hierarchy of symptoms and signs, although with some variability. Early on patients, the major problem is slowing of both cognitive and motor spheres. So patients complain of some difficulty with concentration, focusing on things and so forth, and they may also have mild motor slowing. Later on, that is those that progress - and they don’t all progress - other domains are involved cognitively and

Then there’s a sub-group within this pathology of patients who have true HIV infection or HIV encephalitis. In this case marked by macrophage infiltration and you probably can’t see but a number of these macrophages are multinucleated and that’s the so-called multinucleated-cell encephalitis, which is true HIV encephalitis. These are HIV-infected macrophages. The major cell that is productively infected in the brain, which has been known for some time, is the macrophage or microglial cell. Subsequently more recently it has been shown that other cells can be infected, although they are usually not

So we always have the problem of: how do you explain dysfunction if the cells infected are just macrophages? They are not the functional elements of the brain. At times the neural imaging will be normal. But the characteristic findings are brain atrophy, so clearly this process affects the brain. The brain shrinks and sometimes we see white matter abnormalities of the diffuse variety. At times these will be misread as PML, but the appearance really isn’t the same as that of PML. It tends to be a diffuse and fluffy white matter involvement and if you had PML this extensive, the patient would be virtually vegetative and these patients don’t have 

Now the vacuolar myelopathy: let me just dwell a moment on that. A variety of myelopathies can occur in HIV infection but the vacuolar myelopathy is what we call a non-segmental or non-focal myelopathy. That is, there is not a level on the body where it changes from abnormal to normal, but rather it’s a myelopathy that seems to shade toward the chordal end so that the legs are more affected. They are hyperreflexic, spastic. The arms would be less affected although they may be a little

This is the vacuole myelopathy. This is a severe case. This bubbly appearance is edema within the myelin sheath and we still don’t know the pathogenesis. It’s not straightforward HIV infection of the spinal cord, so it’s pathogenesis seems to

What’s happened to dementia in the era of HAART? The disorder, the more severe dementia, is a disorder that occurs in individuals with low CD4 count, with a high systemic viral burden and systemic immune activation. So it’s an any-stage HIV complication. Very much following the profile of the major opportunistic infections. Although we don’t have clear data, I think anecdotal data and emerging information - and even extrapolation from information earlier in therapy with AZT monotherapy - all suggest together that HAART has a protective effect against the dementia just like it does with opportunistic infections.

Pathogenesis: the way we think, or the general thought about the pathogenesis of the disease is that the virus drives the process. It’s the major prime mover of the process but that there are these secondary events which we can encompass under what we call neuropathogenic pathways that then lead to the disease phenotype. In terms of the infection, just to reiterate a bit, the important part of the infection, or the aspects that have been identified, is that this occurs in a setting usually of poor control of systemic infection. There is CNS invasion of the virus. There is probably autonomous infection of

Very briefly, the general idea just to represent another way is that the infection initiates the process. Various viral products probably trigger other events involving whole expression of toxic host genes that then target… that lead to abnormality. That

Let me talk a little about CSF studies that relate to the nature of the infection. And this simple cartoon is just a reminder that the brain and the CSF are separate. They are not the same thing, so when we look at CSF there’s always the caveat that we are looking at something that is a little bit different than looking at the brain itself. Of course we look at CSF because we are able to relatively non-invasively sample it, whereas obviously we can’t sample the brain and particularly sample the brain over time. Both the CSF and the brain have barriers that separate them from the blood so there’s a selection of what goes into the brain and what goes into the CSF. These barriers are somewhat different but many of the properties are sufficiently similar that we think that this can be looked at as a parallel compartment. There also is an element of interchange between the brain and CSF so certain things can leak in different directions. Particularly leak out of the brain into the CSF. So the importance is that when we look at the CSF we can look at virus, we can look at virus sub-types, we can look at infection and we can look at immune activation molecules and so forth, and, at least hypothetically, see that they may reflect the same kind of things that are going on in the brain. But we always have the reservation that they may not be identical.

Let me just go on to describe some general things about the CSF in HIV. First of all, is a CSF examination worthwhile in demented patients? I think that the bottom line is generally no. In terms of routine studies, both the protein and the cell count are variable. These are old data but actually they are good data since there was a larger series that could be collected. And the general thing that Harry said is true, that later in infection, and in the case of demented patients the more severe the dementia, there is a tendency for these things to actually be more normal. That is, the protein to be more in the normal range rather than elevated, which it characteristically is in early HIV infection, and the cell count to actually go down. This is true in demented and non-demented patients. Late HIV infection, the cell count tends to normalize if anything. Here’s one of these funny things. This dollar sign is supposed to be a beta.

People have attempted to look at both markers of infection and immune activation. I’ll just dismiss this in a minute with the newest information that’s been reported. Has to do with the chemokines in the spinal fluid. And there’s a recent report that MCP1 and ranties are elevated in the spinal fluid of demented patients and appear to correlate with the degree of dementia, just as earlier reported for beta to microglobulin, neopterin and the endogenous neurotoxin quinaldinic acid. Whether these just reflect and are secondary and just markers, or whether they tell us something about the pathogenic processes I think awaits further study. But this is an interesting and recent report showing that. Other beta chemokines were not elevated, interestingly. Early studies with P24 were not very helpful and we will talk a little bit about the viral load in spinal fluid. For patients who have a CD4 count above 200 it appeared that the viral load in the CSF correlated principally with the viral load in the plasma. And they seem to have a kind of constant relationship, and that’s what you see in this regression line.

On the other hand, people with lower CD4 count, those didn’t correlate and actually the viral load of the spinal fluid appeared to correlate with neurological abnormality or neuropsychological testing abnormality. The kind of inference they made from this, which I think is an appropriate one, or hypothetically, is that as HIV infection advances the CSF infection and perhaps brain infection, begins to become more autonomous and not just a reflection of traffic of infection into the nervous system, but rather endogenous or local infection of the nervous system and obviously that might be important in the genesis of dementia if the same thing occurs in the brain. In the same issue of the annals, the Baltimore group of Justin MacArthur and colleagues showed that there was a general rise in the CSF virus load in relation to the increasing severity of the dementia. There’s a tremendous amount of overlap here, but from a population standpoint there seems to be a correlation. You can see also here, indirectly, that when you get up here to the severely demented patients the ratio of CSF to plasma appears to rise, which also was interesting. So there’s a general correlation that again reinforces the idea that in general the dementia appears when the viral load is higher in the plasma. So, more viremic, the more likely the dementia is to occur.

There are our own data which really reinforce the same thing. In terms of diagnosis they indicate that there is a lot of overlap in the level of CSF in the red squares viral load with the degree of dementia, and in individual cases it is probably not helpful. There’s a tendency here in this very early data that the ratio may be upset in the ones with more severe dementia. But the bottom line is that it is not useful diagnostically. You may have seen these kinds of diagrams earlier in the week, and without going into any details, these are some early studies. We’ve extended these studies going along with the idea of Ellis that there is a disconnection with more severe infection. What Bob Grant and Natalie Inken and others have found in the CSF is that in later infection with low CD4 count, the viruses, the phylogeny of the viruses in the CSF and plasma when they are compared are more divergent. Whereas early in infection it seems to be the same virus. Again, going along with the idea that early infection virus traffics through the nervous system, transiently. Late in infection, it tends to diverge and may be more autonomous within the brain, or at least selective.

We’ve done some studies looking at viral load responses in the spinal fluid. I’m probably getting behind, so I’ll go through these very quickly. We studied patients with a range of CD4 counts, a range of dementia, and looked and compared the kinetics in spinal fluid of response to HAART with those in plasma. These axes also didn’t print appropriately but this is HIV RNA and a log scale on this side. This is one of the patients that we saw with a relatively high, or preserved CD4 count 337. He’s one of these surprising patients that comes in perfectly normal and you do an LP and he has 28 cells. Actually as we’ve done these studies prospectively, it sort of reinforces what Harry has described and others have described in the past, that there can be an asymptomatic pleocytosis with entirely unexpected. This is a person that you have no idea about this. And I go back and question him. He had no headache, no nothing. At any event he was naive. He was started on a simple regimen of Combivir and nelfinavir and he had a rapid - this is a very short scale, this is 15 days - a rapid decrement in his plasma, and the CSF, if anything, came down more rapidly. This is what we’ve generally found in those with a high CD4 count who haven’t been treated or are not resistant, but there’s a parallel decrement. If anything the CSF looks like it responds better. What we found in the larger series, the half life - these half lives are long but it’s kind of an artifact of the limited number of data points we have, but in general we found that the CSF half life of virus is a little bit less than that of plasma. When we have better data the half lives are somewhere in the range, in the plasma, of 1-1.6 days, and in the CSF just about the same. Maybe a minor fraction slower. So the general thing is that the responses in the CSF are similar to those in plasma in most patients with AIDS dementia complex.

But we do find exceptions to this and this is a marked exception. A patient treated just with two nucleosides earlier, but what you can see is he had a rapid decrement in his plasma and his CSF hardly moved in this short time frame. That’s one of the patients with the dissociation between the two viruses. They were fairly remotely related to each other. What we AIDS dementia complex