Progressive Multifocal Leukoencephalopathy

by Mark Bowers

HIV is a member of a family of retroviruses that frequently find their way into brain tissue. The spectrum of HIV-related diseases in the brain is wide, caused either by the direct or indirect effects of the growth of HIV in the brain, or by various viruses, bacteria and fungi that cause opportunistic infections and complications. AIDS dementia complex (ADC) is the most widely recognized of these complications (see "AIDS Dementia Complex" in the December 1996 BETA). Another complication is progressive multifocal leukoencephalopathy (PML).

PML is rarely found in early HIV disease; when it does occur, it is almost always found in people with late-stage HIV disease. More PML infections are found in people with AIDS than in people with any other immunosuppressive disorder. In March 1996, Joseph Berger, MD, reported on a retrospective study of the incidence of PML in Americans with AIDS. For the period 1981-1995, as many as 5% of people with AIDS had PML either as their first AIDS-defining infection or one that developed after an AIDS diagnosis. Fewer than 89% survived more than 1 year after diagnosis, but 11% lived longer than a year with either partial or complete remission.

PML is progressive in that there is continuous expansion of brain lesions caused by the JC virus. Neurological abnormalities increase over time, frequently caused by multifocal (in several areas) lesions. However, it is common for a person with PML to have only one brain lesion instead of multifocal lesions. PML may cause dementia, but it is almost always seen with other neurologic problems. A characteristic finding is leukoencephalopathy, or degeneration of the white matter of the brain.

Until recently, neurologists were discouraged by the rapid course of PML disease; the average survival time from date of diagnosis was only 4-6 months. Dawn McGuire, MD, at the Western NeuroAIDS Program at Davies Medical Center in San Francisco, now paints a very different picture. According to McGuire, the immune restoration that frequently accompanies highly active antiretroviral therapy (HAART) may prove to have a significant impact on the incidence of PML. A retrospective study from London found that patients who took AZT monotherapy also had a reduced risk of PML. McGuire tells her patients to "ignore the literature that says you'll be dead in 4 months." The majority of her patients survive much longer.

PML has been particularly difficult to study. As many as two-thirds of American adults harbor the JC virus, but very few develop PML disease; these individuals usually become symptomatic only when they become severely immune suppressed. Clinical evaluation of these individuals is difficult. No drugs to combat JC virus infection have been approved by the Food and Drug Administration (FDA) and there is no accepted standard of care. In addition, diagnostic tools do not always help distinguish PML from other possible brain deficits or infections. By themselves, the results of routine cerebrospinal fluid (CSF) tests do not conclusively diagnose PML, nor do they easily exclude it.

Until recently, treatment for the disease was usually referred to euphemistically as "unsatisfactory." However, effective antiretroviral therapy may prevent PML, and new treatment drugs may hold promise. This article discusses the diagnosis, clinical evaluation and treatment of PML.

Symptoms

The symptoms of PML are not unique to PML. The neurological symptoms often associated with PML include disturbances in movement and coordination, including walking and posture. A limp or muscular weakness may be noticed on one side of the body, suggesting a lesion on the opposite side of the brain. (The neural wiring that connects one side of the body to the brain almost completely crosses to the opposite side.) The visual field of some individuals with PML is smaller and contains "visual field cuts," or missing parts of the visual field on the left or right of the environment. Other senses may also be affected.

PML lesions may affect higher cortical functions, causing aphasia (speech disorders) and apraxia (the inability to coordinate movements despite having the strength to do so). Memory loss, headache, confusion, dementia and seizures may also result from PML lesions.

The onset of PML disease varies. Mental and physical decline can be sudden and rapid, and people may require significant help with the tasks of daily living. Personality changes may create new challenges for caregivers. Occasionally, PML spontaneously goes into remission. Encouraging anecdotal reports suggest that highly active antiretroviral therapy including a protease inhibitor or the experimental use of cidofovir (Vistide) may reverse the decline usually seen with the disease. In some reported cases, the response to therapy has been so impressive that patients return to normal. McGuire reports that one patient's lesions seemed to disappear from radiologic studies. However, it is difficult to evaluate individual case reports when up to 10% of people with PML get better without any specific treatment.

What PML Does

The nerve fibers that convey messages to and from the brain are insulated from one another for many of the same reasons that telephone wires are insulated. Myelin, the fat and protein coating that forms a living sheath for neural fibers, helps to speed the transmission of information to and from the brain and prevents "cross talk," or garbled messages. Like multiple sclerosis, PML is a disease that breaks down myelin (demyelination) by killing the cells that make this insulating material. PML lesions typically begin at the junction between the grey matter and the underlying white matter of the brain. Demyelination results in lesions. The brain lesions that define PML can affect motor and sensory neural networks in some cases and intellectual functions in others, depending on where the lesions are.

A commonly held theory about JC virus infection is that it is harbored in the kidneys until the host becomes immune suppressed, then travels through the blood to the brain. Once in the brain, it infects and kills oligodendrocytes, specialized neural cells that form myelin in the brain. The movement of JC virus to the brain may be mediated by B-cells, and may require some structural changes in the virus. Further study is required to unravel the pathogenesis of PML.

Diagnosis

One difficulty in establishing a diagnosis of PML is the similarity of symptoms among many neurologic diseases. The symptoms described above do not unmistakably point to just one neurologic disease. A person with PML may look and act like one with toxoplasmosis or central nervous system (CNS) lymphoma. Each disease has a different set of characteristic neuroimaging and diagnostic findings.

Brain Biopsy

Brain biopsy, which involves the physical removal and study of a small sample of brain tissue, has been the "gold standard" for diagnosing PML. A stereotactic biopsy allows a cutting tool to be precisely guided to the area of a lesion by use of 3-dimensional coordinates. McGuire recommends that physicians and patients inquire about rates of complications that accompany a brain biopsy at the hospital or institution where a biopsy is planned. Armed with this information, the physician and patient can together decide whether the risks of the procedure outweigh the benefits of knowing for sure if PML is the problem. Considerable research has been devoted to finding a safe and reliable alternative to biopsy.

Cerebrospinal Fluid Analysis

The cerebrospinal fluid (CSF) that bathes the brain and spinal cord contains increased proteins in general, myelin basic protein in specific, and more immunoglobulin content in a person with PML. However, none of these increases proves that a person has PML. McGuire has developed a polymerase chain reaction (PCR) test that is 92% sensitive and specific for JC virus DNA. This level of sensitivity allows McGuire to detect PML without resorting to radiologic studies in 92% of all cases. Several other widely available commercial assays detect JC virus DNA with similar reliability. Some clinicians consider radiologic studies combined with a positive result from a CSF PCR test sufficient to establish a diagnosis of PML, although a negative CSF PCR test result does not mean that a person does not have PML.

There is a concern about the way in which lumbar punctures or spinal taps are done to obtain CSF. Neurologist Christina Marra, MD, Assistant Professor of Neurology and Medicine at the University of Washington, is the primary investigator of an AIDS Clinical Trials Group study (ACTG 363) of cidofovir for the treatment of PML. Marra prefers to use "pencil-point" needles that do not leave holes in the dura mater that surrounds the CSF. When regular needles are used, holes in the dura mater allow some CSF leakage after a lumbar puncture, leading to headaches in about 1 in 10 patients. The rate of headache following a puncture with a "pencil-point" needle is quite small.

Blood Studies

PCR tests for JC virus in the blood are not very helpful in diagnosing PML. As many as one-third of all people with HIV have antibodies to JC virus in their blood, while two-thirds of the public have been exposed to JC virus at some point in their life. It is unlikely that simple blood tests will replace the combination of radiologic tests and analysis of the CSF in the near future.

Radiologic Studies

In a retrospective study by Berger and colleagues, computerized tomography (CT scans, or serial x-ray images) revealed low density white matter lesions that did not distort the shape of the brain, usually in the back part of the brain hemispheres. Not all patients later known to have PML showed abnormalities on CT. Enhancement of CT with contrast agents also failed to locate lesions in most of Berger's subjects. Many neurologists feel strongly that CT scans provide insufficient information to diagnose PML and are very costly, and therefore should not be done.

Magnetic resonance imaging (MRI, a technique to create detailed pictures of the brain without using x-rays) appeared superior to CT in Berger's study because it located more lesions in more subjects. MRI is a better way to see PML lesions below the surface of the brain, in the white matter. MRI is more expensive and takes longer to perform, requires a large space to house the apparatus and can be trying for individuals with claustrophobia. "Open" MRI scanners are available in larger cities for those with claustrophobia.

McGuire and Marra agree that radiologic results that are consistent with PML combined with PCR studies of the CSF are sufficient to accurately diagnose PML most of the time. McGuire is reluctant to base a PML diagnosis solely on radiologic studies. She observes that radiologic studies that misdiagnose PML are rarely published, while successes are almost invariably reported. This provides a distorted picture of the value of radiologic studies by themselves.

Autopsy

Researchers who study the incidence of PML have historically relied on autopsy information to provide a real picture of the number of people affected by PML. Because the number of people who have brain lesions caused by JC virus at autopsy is higher than the number of people definitively diagnosed with PML, pathologists speculate that a significant number of PML cases are missed during life. Efforts are underway to accurately identify individuals with JC virus infection and PML early.

Treatment

Recent interest has focused on treating PML indirectly with highly active antiretroviral therapy, usually consisting of 2 nucleoside analog drugs in combination with a protease inhibitor drug. David Clifford, MD, at Washington University in St. Louis has begun to collect a database of physicians' anecdotal experiences with HAART in people with newly diagnosed or progressing PML. By August 7, Clifford had natural history data from 16 individuals. The goal of Clifford's observational database is to document HAART use in people with PML, and to chart the course of HIV viral load and any complications that may arise once individuals are on his registry. Clifford recommends "pulling out the stops" with regard to aggressive antiretroviral therapy, as well as closely monitoring viral load. He informally observes that the people who are now getting PML generally do not seem to have undetectable HIV viral loads. He encourages physicians with patients who have PML to work diligently to decrease HIV viral load and to share their experiences with him and the NeuroAIDS Research Consortium (NARC). A website has been established for the PML registry at www.neuro.wustl.edu/narc.

Three case reports of people who stabilized or dramatically improved while receiving HAART have recently appeared in BETA (September 1996, page 40; March 1997, page 29; June 1997, page 52).

For several years, people with PML were treated with cytosine arabinoside (Ara-C), either by intravenous administration or by introduction of the drug directly into the CSF. Most work with the drug was done by Caroline Britton, MD, at Columbia-Presbyterian Hospital in New York City. Her positive results created a general impression that Ara-C might be an effective treatment.

A recent AIDS Clinical Trials Group study (ACTG 243) compared the 2 routes of administration of Ara-C combined with the best antiretroviral therapy for the individual to the best antiretroviral therapy alone. (Because this study began before any protease inhibitor drugs were approved, many participants did not receive combination therapy including a protease inhibitor.) The study was halted after 60 participants were treated. No statistical difference in survival could be found among the study arms, suggesting that either too few people were studied to detect a small benefit to the use of Ara-C, or that there was no effect at all. This was an important study, because many physicians elected to treat their patients with Ara-C based on scant clinical evidence that the drug worked. There have also been isolated reports that combinations or alternations of Ara-C and interferon-alpha produced some clinical improvement, but these combination strategies have not been subjected to rigorous clinical evaluation.

Test-tube studies have suggested that cidofovir kills several viruses related to the JC virus. Belgian researchers Andrei, Snoeck, Vandeputte and DeClercq have published encouraging findings that show that cidofovir kills viruses related to JC virus, and Gary Blick, MD, of Blick Medical Associates in Greenwich, CT, has submitted a case report for publication detailing successful treatment with cidofovir of an AIDS patient diagnosed with PML. The only way to be sure that cidofovir is a useful treatment for PML is through clinical study.

Other drugs that have been used as treatment for PML include prednisone, acyclovir, vidarabine and adenine arabinoside. Interferon-alpha and interferon-beta have been combined with other drugs or given alone. At the Fourth Conference on Retroviruses and Opportunistic Infections in Washingtion, DC, in January, 1997, Justin McArthur, MD, and colleagues from Johns Hopkins University reported the effect of interferon-alpha on survival by examining the records of 104 people with HIV and PML. They found that untreated patients lived an average of 121 days compared to 325 for those treated with interferon-alpha. They concluded that interferon-alpha use may delay progression, improve survival and be associated with remission in HIV-associated PML.

Clinical Studies

A pilot study of cidofovir (ACTG 363) has been designed by researchers in the ACTG and the NARC and at Gilead Sciences. Primary investigator Marra says that the dosing and schedule of administration of cidofovir will be the same in the upcoming PML study as they are for the treatment of cytomegalovirus (CMV) retinitis. Participants will be given intravenous cidofovir at 5 mg/kg once a week for 2 weeks, then once every 2 weeks for a total of 13 doses. The study lasts 24 weeks. This study will recruit participants in whom PML has been proven by a combination of positive CSF PCR and typical clinical and MRI findings. Marra intends to monitor changes in the quantity of JC virus DNA in the CSF to see if changes reflect changes in clinical response. She will also look for evidence of CMV in the CSF, since CMV is known to cause inflammation of the brain (encephalitis) and cidofovir may have a positive effect on encephalitis. People with positive PCR tests for CMV in the CSF will not be included in the PML study.

Marra is cautious about the use of cidofovir. She is wary of side effects such as iritis (inflammation of the eye) and hypotony (decreased pressure inside the eye), both of which are potentially serious, sight-threatening complications that occurred in 7.3% of participants in a clinical study of cidofovir. Participants in Marra's multicenter study will be closely monitored for the development of these conditions. She will also monitor participants for kidney toxicity (protein in the urine occurred in 48% and serum creatinine was elevated in 8% of participants in one clinical study) and low blood cell counts (neutrophil counts less than 750 cells/mm³ occurred in 31% of participants in one clinical study). For local study sites, call the AIDS Clinical Trials Information Service (ACTIS) toll-free at 800-874-2572.

Pharmacia and Upjohn, Gilead Science's marketing partners outside the United States, are conducting a study of the safety and efficacy of cidofovir in the treatment of PML in HIV seropositive individuals in Europe. This will be a randomized, double blind, placebo-controlled study with an open-label extension phase. The study will compare 6 months of cidofovir or placebo given to participants at the time of diagnosis or progression of PML. All participants will be offered an additional 6 months of access to cidofovir at the conclusion of the study.

McGuire is studying topotecan, a topoisomerase inhibitor, for the treatment of PML at the Western NeuroAIDS Program. Topotecan inhibits an enzyme that is involved with making usable DNA. Topotecan was approved for treating metastatic cancer of the ovary after another treatment has already failed. The drug is administered intravenously by a central catheter continuously for 21 days with 7-day breaks between infusions. Topotecan has been shown to inhibit both JC virus and HIV. Previous use of Ara-C is technically an exclusion for participation in this study, although it has been shown to be ineffective. Study sponsors have assured BETA that a protocol exception will allow people who have been previously treated with Ara-C to participate.

Topotecan has been shown to penetrate well into the brain, and the long continuous exposure to it permits the drug to work at the correct stage in the cell cycle. Side effects include bone marrow suppression (81%), nausea (74.8%), vomiting (55.5%), diarrhea (39.3%) and hair loss (59.3%). The drug is also being studied as a treatment for AIDS-related Kaposi's sarcoma and non-Hodgkin's lymphoma. For further information, contact the Western NeuroAIDS Program at 415-565-6368.

NARC has preliminary plans to study a related topoisomerase inhibitor called 9-amino-camptothecin. Neurooncologist John Henson, MD, at Massachusetts General Hospital is now developing a proof-of-concept study.

Conclusion and Future Directions

The current outlook for people diagnosed with PML is nowhere near as bleak as it was only 2 years ago. Perhaps the greatest gain has been a perceived decline in the number of individuals who are diagnosed with PML, widely attributed to increased use of HAART. Antiretroviral drugs may protect against PML. Important strides have recently been made in the ability to accurately and efficiently diagnose PML without resorting to stereotactic brain biopsy. McGuire and Marra both speculate that quantitative PCR tests for JCV DNA in the cerebrospinal fluid will become very useful in monitoring the effectiveness of treatments for PML. Although the unofficial standard treatment for PML, Ara-C, was found to be ineffective when carefully scrutinized, topoisomerase inhibitors such as topotecan may prove effective, and clinical studies are clearly needed. The anti-CMV drug cidofovir shows considerable promise and has enthusiastic supporters. Both studies will fill slowly, as the number of individuals being diagnosed each month may be declining. This alone is reason enough for celebration.

Special thanks to Christina Marra, MD, Dawn McGuire, MD, Sally Nuessle, PharmD, and David Clifford, MD, for their comments or for review of this article. Any errors of fact or omission are solely the responsibility of the author.

Mark Bowers is Managing Editor of Treatment Publications at the San Francisco AIDS Foundation.

References

American Foundation for AIDS Research. AIDS/HIV Treatment Directory 8:1. February 1996.

Andrei G and others. Activity of various compounds against murine and human polyoma virus. Antimicrobial Agents and Chemotherapy 41:587-593. March 1997.

Berger J and others. Progressive multifocal leukoencephalopathy associated with human immunodeficiency virus infection. Annals of Internal Medicine 107:78-87. 1987.

Cutler R. Demyelinating disease. In Scientific American Medicine. Rubenstein and Federman, editors. Scientific American Inc, New York. 1997.

Domingo P and others. Remission of progressive multifocal leukoencephalopathy after antiretroviral therapy. The Lancet 349:1554. May 24, 1997.

Eliot B and others. 2-5 year remission of AIDS-associated progressive multifocal leukoencephalopathy with combined antiretroviral therapy. The Lancet 149:850. March 22, 1997.

Fong I and others. The natural history of progressive multifocal leukoencephalopathy in patients with AIDS. Clinical Infectious Diseases 20:1305-1310. 1995.

Giesen H and others. Diagnostic criteria and clinical procedures in HIV-1 associated progressive multifocal leukoencephalopathy. Journal of the Neurological Sciences 147:63-72. 1997.

Hammer S. Slow virus infections. In Scientific American Medicine. Rubenstein and Federman, editors. Scientific American Inc, New York. 1997.

Huang S and others. Survival prolongation and symptom palliation in HIV-seropositive patients with PML treated with alpha-interferon. 4th Conference on Retroviruses and Opportunistic Infections. January 22-26, 1997. Washington, DC. Abstract 341.

Levy R and others. Neurological manifestations of the acquired immunodeficiency syndrome (AIDS): Experience at UCSF and review of the literature. Journal of Neurosurgery 62:475-495. April 1985.

McGuire D and others. JC virus DNA in cerebrospinal fluid of human immunodeficiency virus-infected patients: predictive value for progressive multifocal leukoencephalopathy. Annals of Neurology 37:395-399. 1995.

Luttman S and others. High-signal lesions in the midbrain on T1-weighted MRI in an HIV-infected patient. Neuroradiology 39:136-138. 1997.

Quinlivan E and others. Subclinical central nervous system infection by the JC virus in patients with AIDS. Journal of Infectious Diseases 166:80-85. 1992.

Shah K. Polyomaviruses. In Fundamental Virology 2nd edition. Fields and Knipe, editors. Raven Press, New York. 1991.

SmithKline Beecham. An open, comparative Phase II study of immediate versus delayed treatment with topotecan HCl given as a continuous 21-day infusion every 28 days to patients with AIDS-related progressive multifocal leukoencephalopathy. SK&F 104864-A, Protocol 111. November 21, 1996.

Snoeck R and others. Treatment of progressive multifocal leukoencephalopathy (PML) with cidofovir in an AIDS patient. Ninth International Conference on Antiviral Research. Urabandai, Fukushima, Japan. May 19-24, 1996. Abstract 113.

Snoeck R and others. (S)-1-(3-hydroxy-2-phosphonylmethoxypropyl) cytosine (Cidofovir, HPMPC) is a potent inhibitor of murine polyoma virus replication both in vitro and in vivo. 35th Interscience Conference on Antimicrobial Agents and Chemotherapy. San Francisco, CA. September 17-20, 1995. Abstract H49.

Vago L and others. JCV-DNA and BKV-DNA in the CNS tissue and CSF of AIDS patients and normal subjects. Study of 41 cases and review of the literature. Journal of Acquired Immune Deficiency Syndromes and Human Retrovirology 12:139-146. 1996.
Vollmer-Haase and others. Efficacy of camptothecin in progressive multifocal leucoencephalopathy. The Lancet 349:1366. May 10, 1997.