PML, the brain disorder that will not go away in the antiretroviral era

Progressive multifocal leukoencephalopathy (PML) as a cause of serious central nervous system (CNS) complications and death among people with HIV has increased since the introduction of antiretroviral therapy (ART), according to several presentations made last month at the Second HIV Infection and Central Nervous System: Developed and Resource-Limited Settings meeting in Venice.

“This is certainly something which is not going away,” Dr Igor Koralnik of Beth Israel Deaconess Medical Center and Harvard Medical School, one of the world’s leading PML experts told the audience. And despite improvements in survival since ART (more than half of the people diagnosed with PML now survive a year or more), “it is still one of the deadliest opportunistic infections in HIV. [And] approximately 80% of survivors have significant neurological sequelae. So it’s really a terrible disease even if you survive PML.”

What’s worse, he said, is “that there is no cure, there is not even a race for a cure and that’s the really sad part of PML.” However, he and other researchers at the conference described some advances in the basic science understanding of PML, which could pave the way to the development of therapeutics, especially novel immune-based therapies — possibly in combination with drugs that block the serotonin receptor which the JC virus that causes PML uses to enter cells in the brain.

“It has been described early in the AIDS epidemic that approximately 5% of patients with AIDS will develop PML,” said Dr Koralnik. This was in industrialised settings; the incidence in resource-limited settings is either low or unknown.

But even though it is a relatively rare disease, the polyomavirus JC virus (JCV) that causes it is ubiquitous. In fact, most people are infected with the JC virus during childhood, and by adulthood approximately 85% are seropositive worldwide (though like HIV, different strains of JC virus are prevalent in different parts of the globe (Delbue)). The infection is asymptomatic as long as the immune system is functioning normally, and the virus remains quiescent in the kidney tubular epithelial cells or in the lymphoid organs. It can often be found by PCR secreted in the urine of healthy individuals.

It is not clear how commonly the virus exists latently in the brain in healthy people. Dr Koralnik rather believes that when the immune system is suppressed (due to HIV, leukaemia or organ transplant drugs or chronic steroid use), the virus begins to replicate elsewhere and then spreads via the blood to the CNS. In the case of HIV disease, this usually only occurs when the CD4 count falls below 200 cells/mm³, but there have been reports of patients with CD4 cell counts of 500 cells/mm³ and even higher when diagnosed with PML — as opposed to when the JCV infection actually becomes established in the brain. Some of the cases diagnosed at higher CD4 cell counts may actually be due to PML-IRIS (see related article).

Nevertheless, once the immune system is suppressed, JCV can enter the CNS and infect astrocytes (glial cells which protect and support neurons) and oligodendrocytes, the cells that produce myelin, the protective sheath covering nerve fibres in the CNS. This ultimately causes the death of the oligodendrocytes, the subsequent loss of myelin (demyelination), and the formation of multiple lesions (usually) in the white matter of the brain. However, there have also been occasional reports of grey matter involvement.

The initial clinical presentation is usually subacute or, rather, insidious. At first, the condition may seem like a stroke that gradually gets worse, often starting with weakness in the arms or legs (sometimes on only one side of the body), sometimes progressing to paralysis with altered vision (blurred or double vision) that may progress to blindness. Changes in mental status may be subtle at first, starting with confusion and worsening to dementia. Aphasia — trouble producing or understanding coherent speech — is common. If untreated, PML can lead to a coma or vegetative state, with death occurring in less than six months in most patients, while some people with PML have a much slower course of progression.

Diagnosis is usually based upon detection of JCV by PCR in cerebrospinal fluid (CSF), which, before the advent of ART at least, was highly sensitive and specific for PML. According to a presentation by Dr Serena Delbue of the University of Milan, high CSF levels of JCV are also predictive of poorer outcomes — and thus might be useful for clinical management of patients.

Since ART, however, some studies have reported that JCV PCR in the CSF is not quite as sensitive (detected in ~58%) — and so brain biopsies may still be necessary to confirm diagnosis in some patients. (There could be a couple of alternate explanation for these cases — see PML IRIS below). In suspected cases where the CSF is JCV-negative, it might also be easier to look for JCV in the blood — according to a poster presentation at the conference by Sala et al., around 40% of PML patients have detectable JCV in plasma. However in most cases, JCV-DNA levels were only detectable in the blood after onset of neurological symptoms, and thus could not be used either to predict or monitor outcomes.

Computerised tomography (CT scans) and magnetic resonance imaging (MRI) findings characteristic of classical PML typically show multiple (although there is occasionally just one) focal lesions in the white matter, without mass effect and usually without enhancement (there may be some differences in enhancement in people who develop PML-IRIS (see related article).

Limited access to these diagnostic methods could partly explain why the reported incidence of PML is so low outside the developing world. “It occurred to me during this conference that it is the only opportunistic infection which is more frequent in the resource developed setting,” said Dr Koralnik.

But the lower rate of PML in resource-limited settings could also be due to differences in host susceptibility or differences in the local strains of JC virus. Likewise, HIV has been theorised to specifically interact with JC virus, so the difference in PML incidence could possibly reflect the differences between HIV subtypes (and their respective neurotropisms) across the globe.

However, Dr Koralnik believes it might also simply be a case of people dying of other causes or prior to diagnosis in resource-poor settings. Another possibility, as Dr. Hakim from the University of Zimbabwe pointed out in an earlier session at the meeting, is that there's no incentive to spend scarce resources to investigate something that can’t be treated.

And there is no treatment for JC virus or PML. The literature is full of case reports or anecdotes of this or that drug (Cytosine arabinosine, Ara C), topotecan, cidofovir — all rather difficult to tolerate medications) having an effect in some patients but not others — but so far there is little convincing evidence that any of these drugs can be relied upon to alter the course of the disease.(see related article on future treatment research)

The best way to manage PML is to try to resolve the underlying immune suppression — in the case of HIV, with ART. As already noted, survival before ART was dismal, less than 10%, but ART can lead to some dramatic improvements — including complete clearance of JCV from the CNS. Still, much of the neurologic damage may be permanent.

But even though ART (plus the effect of cotrimoxazole prophylaxis against toxoplasmosis) have led to a dramatic decrease in most other opportunistic infections of the CNS, rates of PML have remained rather constant.

“In the era of [HA]ART, the incidence of PML seems rather stable over time,” said Dr Adriana Ammassari of the National Institute for Infectious Diseases "Lazzaro Spallanzani", in Rome. She described data drawn from the Italian Registry Investigative Neuro Aids (IRINA) that show the relative proportion of PML vis-à-vis other CNS opportunistic infections has increased. For instance, PML was the cause of 28% of the focal brain lesions detected in 2000 compared to 16% in 1994.

PML continues to occur in other settings as well.

“I used to see approximately one new case a month at the Beth Israel Deaconess Medical Centre,” said Dr Koralnik, “but for some reason, in the first three months of 2007 I have seen nine cases and I don’t know if there is a mini-epidemic going on [or what].”

The continued persistence of PML could be due to the fact that it can occur at relatively higher CD4 cell counts (~ 200 cells/mm³). Thus, in contexts where people are unaware of their HIV status until late in the course of HIV disease or where ART is delayed until the CD4 cell count falls near 200, the immune system may have already weakened enough to allow JCV to spread to the brain.