Peripheral Neuropathies
Neuropathies associated with cytomegalovirus infection
Nov. 16, 2024
MedLink®, LLC
3525 Del Mar Heights Rd, Ste 304
San Diego, CA 92130-2122
Toll Free (U.S. + Canada): 800-452-2400
US Number: +1-619-640-4660
Support: service@medlink.com
Editor: editor@medlink.com
ISSN: 2831-9125
Toll Free (U.S. + Canada): 800-452-2400
US Number: +1-619-640-4660
Support: service@medlink.com
Editor: editor@medlink.com
ISSN: 2831-9125
Worddefinition
At vero eos et accusamus et iusto odio dignissimos ducimus qui blanditiis praesentium voluptatum deleniti atque corrupti quos dolores et quas.
Progressive multifocal leukoencephalopathy (PML) is an opportunistic demyelinating infection of the central nervous system caused by JC virus (JCV, JCPyV), a polyomavirus that is widely distributed in human populations. PML is characterized clinically by the development of multifocal neurologic signs referable to cerebrum, or, less frequently, cerebellum or brainstem; spinal cord involvement is rare. Pathologically, PML is characterized grossly by multifocal areas of myelin loss. Microscopically, the disease is characterized by lytic infection of oligodendrocytes, and in many cases, the presence of atypical astrocytes. The disorder has been rare outside the setting of HIV infection, but 4% of untreated AIDS patients may succumb to the disease. PML has become of increasing concern in patients receiving aggressive immunosuppression for organ or stem cell transplantation or in patients treated with newer immunomodulatory agents, in particular natalizumab. In this article, the author reviews the pathogenesis, clinical features, diagnosis, and treatment of this disorder.
• Progressive multifocal leukoencephalopathy (PML) is an opportunistic demyelinating infection of the central nervous system caused by the human polyomavirus, JC virus (JCV). The disorder almost invariably affects immunosuppressed patients, in particular, those with impaired T-cell response. | |
• The major risk factor for development of PML is untreated AIDS. PML may affect up to 4% of such patients. | |
• PML also occurs in individuals receiving newer, more aggressive immunosuppressive regimens for organ or stem cell transplantation, as well as in individuals receiving immunomodulatory agents for treatment of multiple sclerosis or other disorders. The agent most frequently associated with PML is natalizumab. Cases of PML may occur in patients receiving drugs, such as dimethyl fumarate, fingolimod, brentuximab vedotin, or other agents. | |
• The diagnosis of PML should be considered in immunocompromised individuals presenting with multifocal neurologic signs or evidence of multiple white matter lesions on MRI. Specific detection of JCV is typically made by polymerase chain reaction analysis of cerebrospinal fluid. | |
• In occasional patients, almost always in the setting of HIV, JCV may also cause infection of cerebellar granule cells or cortical or hippocampal neurons. | |
• There is no proven antiviral therapy for PML. In patients with AIDS-associated PML, antiretroviral therapy may allow sufficient recovery of the immune system to allow stabilization or improvement (Historical note: ART: combined antiretroviral therapy was initially termed “highly active antiretroviral therapy” or “HAART” and subsequently “combined antiretroviral therapy” or “cART”, replaced by “ART”). Remission has also been reported after withdrawal of immunosuppressive drugs and, in patients with PML in the setting of natalizumab treatment, after cessation of therapy. Plasma exchange (PLEX) or immunoadsorption therapy have been used in natalizumab-associated PML, but their efficacy is unproven. | |
• Stabilization or improvement in PML has been reported in limited numbers of cases following treatment with the recombinant human methionyl granulocyte colony-stimulating actor, filgrastim, immune checkpoint inhibitors with or without concomitant treatment with interleukin 2, or T cells sensitized to a second human polyomavirus, BK virus (BKV, BKPyV). Definitive treatment for PML, however, remains unestablished. | |
• In both HIV-infected and iatrogenically immunosuppressed patients, restoration of immune function may result in immune reconstitution inflammatory syndrome (IRIS). |
Hallervorden, in 1930, reported two cases of a previously undescribed, apparently degenerative condition accompanied by central nervous system demyelination (80). Additional, similar cases were described by Winkelman and Moore, Bateman, Squires and Thannhauser, and Christensen and Fog (208; 10; 34). It was not until 1958, however, that Richardson, Astrom, and Mancall published the first case series of this disorder, describing its clinical features and its neuropathological findings of multifocal demyelination, nuclear enlargement or inclusions in oligodendrocytes, and bizarre alteration of individual astrocytes (07). These authors termed the condition “progressive multifocal leukoencephalopathy” (PML) and noted the close association of this condition with hematological malignancies, subsequently describing the association of the disorder with other immunocompromised states (149). The presence of inclusions within oligodendrocytes led first Cavanagh and then Richardson to suggest that the disease might represent an unusual sort of infection (29; 149). Evidence that PML was caused by a virus came in 1965, when Zu Rhein and Chou and Silverman and Rubinstein independently identified crystalline arrays of virions in progressive multifocal leukoencephalopathy oligodendrocytes (167; 218). The arrays most closely resembled those seen in cells infected with the mouse agent, polyoma virus, an agent not known at that time to have any human counterpart. Despite intense initial skepticism, this observation was universally confirmed by other investigators.
Attempts to culture an agent from PML brains were unsuccessful until 1974 when Padgett and colleagues successfully isolated the agent in primary cultures of human fetal brain cells and named the agent “JC virus” (JCV), using the initials of the patient (John Cunningham) from whose brain the virus had been recovered (139). In that year, a second, closely related human polyomavirus, BK virus (BKV, BKPyV), was recovered from human urine (65). It is now known that polyomaviruses are widespread agents in both mammalian and avian species and that four different polyomaviruses have been associated with human disease (77). Of these, JCV has been associated with almost all cases of PML. PML apparently caused by BK virus has been reported in rare patients (25; 24; 47; 125; 135); and BK virus has also been associated with rare cases of encephalitis (193; 04). Several early reports described identification of a third polyomavirus, SV40, a simian agent that was a contaminant of early lots of Salk and Sabin polio vaccines (165; 77). Subsequent studies employing molecular methods, however, have demonstrated that in all of these cases in which tissue was still available for study, the causative agent was JCV, and that SV40 had been a laboratory contaminant (172).
Until 1980, PML remained an extraordinarily rare condition such that an individual hospital might not see a case for many years. In addition, the diagnosis might be suspected clinically but could only be confirmed by brain biopsy or autopsy. The incidence of PML changed dramatically as AIDS became epidemic, and PML became a prominent opportunistic central nervous system infection in HIV-infected patients. Antemortem diagnosis of PML became possible by MRI and polymerase chain (PCR) analysis of CSF. Development of effective ART has led to a decrease in numbers of PML cases, and regression or stabilization of PML may occur in some, but not all, patients treated with antiretroviral therapy (94). In recent years, PML has been associated with monoclonal and other newer immunosuppressive agents, including natalizumab, efalizumab, rituximab, alemtuzumab, mycophenolate mofetil, etanercept, leflunomide, and brentuximab vedotin (98; 109; 186; 131; 146; 73; 162; 27; 77; 90).
To date, no successful antiviral treatment has been developed for PML (150). In recent years, however, three alternative approaches to therapy have been tried. The first of these, based on the recognition that JCV attachment and entry to host cells are mediated by α2,6-linked lactoseries tetrasaccharide c (LSTc) and 5-hydroxytryptamine receptors (5-HT2Rs), has led to attempts to treat PML using 5-HT2Rs antagonists such as mirtazapine (121); however, these have not proven definitively successful. A second and more promising approach stems from recognition that PML can stabilize if there is restoration of effective T lymphocyte response. This has led to efforts to treat PML using checkpoint inhibitors (pembrolizumab or nivolumab) or using methods, which modify host T cells (128; 43; 195; 71; 106; 105; 41). A third, potentially promising approach has been to treat PML using infusions of BK virus-specific T lymphocytes (128; 44; 88). Institution of ART or reduction in immunosuppressive regimens (eg, natalizumab) in patients with PML may result in a paradoxical inflammatory response (immune reconstitution inflammatory syndrome, or IRIS), which may lead to severe cerebral edema and death (11; 36; 37; 62; 77).
• PML is overwhelmingly associated with conditions producing impaired host T cell immune response. Chief among these is AIDS, followed by aggressive immunosuppressive therapy for autoimmune disease or organ or stem cell transplantation. Patients receiving natalizumab or other immunomodulatory agents are also at increased risk. | |
• PML is characterized clinically by development of focal neurologic signs referable to cerebrum, brainstem, or cerebellum. Retina is not affected, and spinal cord involvement is rare. | |
• The course of PML without intervention is one of progressive development of neurologic signs, culminating in a vegetative state. |
PML is almost invariably a disorder of immunocompromised patients. In the great majority of patients, PML begins insidiously, although, occasionally, disease onset may be so rapid as to suggest stroke or brain abscess. Initial symptoms and signs are commonly indicative of focal cerebral involvement and may include alterations in personality, changes in intellect, focal weakness, difficulty with motor skills, or sensory loss (149; 75; 42). Involvement of the dominant cerebral hemisphere may result in expressive or receptive dysphasia. Visual abnormalities occur in 50% of patients and may include Balint syndrome and actual cortical blindness (09; 202; 174). Occasionally, PML begins with signs of brainstem or cerebellar involvement; these may include ataxia, abnormalities of eye movements, difficulties with phonation or swallowing, or ataxia (07; 05; 157). Spinal cord involvement, although reported, is unusual (83; 194; 19; 130).
Prior to the use of ART in patients with AIDS, the course of PML was almost always one of progression to death. As discussed below, the use of ART has resulted in prolonged survival in many patients with PML. Similarly, remission of PML with clinical stabilization has been observed following withdrawal of immunosuppressive treatment (38; 36); treatment with filgrastim, an agent usually used for neutropenia (171); use of immune checkpoint inhibitors; and use of BKV-sensitized T lymphocytes (128; 88; 205; 43; 195; 71; 115; 44; 135). Without therapeutic intervention (eg, institution of ART in AIDS-PML or withdrawal of immunosuppressive agents, with or without plasma exchange), the course of progressive multifocal leukoencephalopathy is, with rare exception, remorselessly progressive. Initial symptoms are followed by the appearance of multifocal neurologic signs, increasing dementia, and eventual progression to a vegetative state. Most patients with non-AIDS progressive multifocal leukoencephalopathy die within 1 year, but death may occur within as little as 2 months, and cases have been reported with survivals of 8 to 10 years or longer (82; 170; 95; 113). In the absence of treatment with ART, survival in AIDS-progressive multifocal leukoencephalopathy averages 4 months (16; 70). Rarely, PML may spontaneously stabilize or remit.
A 68-year-old man with a history of intravenous drug abuse was diagnosed as having acquired immune deficiency syndrome 2 years prior to admission. He was noncompliant with therapy but remained in apparent good health until 2 months prior to admission when he developed diplopia, followed by slurred speech and right-sided weakness. The patient became increasingly demented, with worsening hemiparesis and incontinence of bowel and bladder. He was brought to the hospital after his family found him severely confused.
On examination the patient was an emaciated man who was drowsy but arousable. The patient exhibited a right-gaze preference and a flaccid right hemiparesis most marked in the upper extremity. Spontaneous movements were noted against gravity and to deep pain in the left upper and lower extremities. The patient responded to pain in all four extremities. Reflexes were hyperactive but symmetrical. Plantar response was flexor on the right and produced withdrawal on the left. Head CT scan was normal except for atrophy. MRI, however, showed multiple areas of altered white matter signal, consistent with PML. PCR of the patient’s CSF was positive for JCV but negative for Toxoplasma gondii, herpes simplex virus, herpes zoster virus, and cytomegalovirus. CSF cultures for bacteria, Mycobacterium tuberculosis, and fungi were negative, as was cryptococcal antigen. The patient declined treatment, and at the family’s request, the patient was discharged to a nursing home.
• The causative agent of PML is the human polyomavirus, JC virus (JCV). In a handful of cases, a second human polyomavirus, BK virus, has been isolated without concomitant detection of JCV. | |
• JCV is a ubiquitous human agent, with a seroprevalence of over 50% to 70% by late adult life. In immunologically normal patients, initial JCV infection is mild or clinically inapparent and is followed by lifelong infection with persistence of virus in kidneys and possibly other organs including brains. Patients with persistent JCV infection may periodically shed virus in urine. | |
• JCV exists in two forms: an archetypal form that is thought to be the virus transmitted in nature, and variants with alterations or duplications in the virus regulatory region. JCV isolates from PML CSF or brains are typically variants. | |
• JCV induces PML under conditions of impaired host T cell response. It is not certain whether PML is the consequence of systemic viremia with CNS invasion or whether it arises following reactivation of viral infection within brain. | |
• In PML, JCV infects two cell populations in the brain: oligodendrocytes, which undergo lytic infection with resultant demyelination, and nonproductive infection of astrocytes: these cells do not undergo lytic infection but assume multinucleate or bizarre forms suggestive of astrocytes found in glial tumors. |
Human JCV infection. The causative agent in almost all cases of PML is the human polyomavirus, JCV, with rare cases of PML having been associated with infection by a second human polyomavirus, BK virus (25; 24; 47; 125; 135; 150). JCV is an unenveloped 45 nm virus whose genomic material is composed of supercoiled, circular, double-stranded DNA of 4963 base pairs (77; 42). The viral genome can be separated into an early region encoding nonstructural genes that have enzymatic function in viral replication and a late region that encodes the viral capsid proteins. Early and late regions are separated at their 5’ termini by regulatory region sequences of approximately 500 base pairs that control early and late gene expression and the initiation of viral DNA replication.
JCV is a ubiquitous human agent. The prevalence of antibody to JCV is 10% in children 5 years of age and has been reported to rise to approximately 70% by late adult life (136; 77; 150). Overall seroprevalence of anti-JCV antibodies is approximately 56%, with a higher seroprevalence rate in males than in females (23; 136). Initial infection is not known to cause symptomatic clinical disease. Work indicates that JCV uses the 5HT2A serotonin receptor to infect cells (77; 121). Other receptors may also exist; Chapagain and colleagues have reported that JCV is able to infect cultured human brain microvascular endothelial cells without using the serotonin receptor (31).
Persistent JCV infection and host immune response. Following initial infection, JCV has been shown to persist in kidneys and possibly in other tissues, including the brain (204; 59; 76; 49; 141; 104; 176; 117). JCV infection in immunologically normal patients is characterized by lifelong persistence with episodic reactivation. In immunologically normal patients, JCV persistence can result in recurrent episodes of asymptomatic viruria that may increase with age, and viruria is common during pregnancy, and under conditions of immunosuppression (77). In a study of urine samples from 20 normal donors, JCV was detected in urine from three (15%) of the patients (89). In this study, JCV viruria and viremia were detected in 25% of samples from immunosuppressed renal transplant patients. JCV has been detected in 48% of patients with PML (60). In older studies, viremia was reported in 13.5% or more of HIV-infected individuals without PML (53; 101), and in the era of more effective ART showing plasma JCV DNA (60), in 35% of individuals treated with natalizumab (116). Mechanisms of persistent JCV infection are poorly understood. Work with an analogous murine polyomavirus, murine pneumotropic virus (K virus), however, has shown that the virus does not persist in truly latent form, as do herpes simplex and herpes zoster viruses, but rather as a low-level productive infection, which is contained by T cell-mediated response and can reactivate following immunosuppression (79).
Pathogenesis of PML. The factors that give rise to PML have not been fully defined. Clearly, immunosuppression plays in important role. Du Pasquier and colleagues have shown that most normal individuals have circulating cytotoxic T lymphocytes (CTLs) specific for JCV and that an intact CTL response specific for JCV is an important factor in containment of PML once the infection begins (56; 55). Similarly, the greatly increased incidence of PML in AIDS strongly suggests that HIV infection predisposes to PML to a greater extent than any other condition. Studies by several investigators suggest that HIV tat protein may facilitate JCV replication (33; 103; 12; 209), and HIV tat protein has been detected in JCV-infected oligodendrocytes in PML brains (160; 185). However, PML affects only a minority of AIDS patients, despite that fact that the majority of these patients are persistently infected with JCV.
Changes in the virus itself are thought to contribute to the development of PML. JCV exists in two forms. The parent strain of JCV, termed “archetypal JCV” is the form in which the virus is believed to be transmitted from person to person and is the form that persists in renal tubular epithelial cells. Archetypal JCV is characterized by a single regulatory region. In contrast, 90% of strains of JCV recovered from PML brains are distinguished from the JCV archetype by rearrangements or deletions in the viral regulatory and coding regions (204; 08; 01; 132; 77). These strains have also been detected in tissues of normal as well as immunosuppressed individuals and may increase the ability of the agent to infect oligodendrocytes and astrocytes under conditions of immunosuppression (176). JCV has also been shown to infect B lymphocytes (64). It is not yet known, however, whether PML is the result of viremic spread of JCV to the central nervous system, possibly within B lymphocytes, or whether PML results from reactivation of JCV infection within brain.
PML associated with treatment with natalizumab. Natalizumab is a humanized monoclonal antibody against alpha4 integrin, a molecule involved in adhesion and migration of lymphocytes across endothelial cells. Subclinical reactivation of JCV with JCV viremia and viruria has been reported in patients receiving natalizumab (32), and it has been thought that the drug, by altering T cell-mediated immune surveillance in these individuals, causes reactivation of JCV infection and PML. Later data also indicate that natalizumab may have two other effects relevant to JCV replication and the development of PML: first, the virus causes migration of CD34+ cells and B cell precursor cells from bone marrow, increasing the likelihood of release in cells that can carry the virus into the peripheral circulation; second, natalizumab upregulates a family of transcription factors (POU2AF1/SpiB) in T cells from natalizumab-treated multiple sclerosis patients (124). SpiB possesses multiple binding sites on the JCV regulatory region and could enhance productive JCV infection in CD34+ hematopoietic precursor cells and CD19# B cells. Interestingly, this upregulation occurs over a period of roughly 2 years, consistent with the duration of natalizumab therapy in many patients prior to the development of PML (124). Studies involving natalizumab have also suggested that use of the agent may increase the rate of seroconversion to JCV (190). Increase in antibody titers to JCV has been reported in some but not all studies following natalizumab-treated patients (145; 190).
PML associated with other immunomodulatory agents. Individual cases of PML have also been reported with other monoclonal and newer immunomodulatory agents, including natalizumab, efalizumab, rituximab, ocrelizumab, ofatumumab, alemtuzumab, mycophenolate mofetil, etanercept, leflunomide, and brentuximab vedotin (21; 77; 129; 140; 169; 166). As of 2022, 12 cases of dimethyl fumarate–associated PML were reported (of 560,000 patients treated), all with some degree of lymphopenia, and nine with severe, prolonged lymphopenia (114). PML can develop in dimethyl fumarate-treated patients without prior exposure to natalizumab (93). Cases of PML have also been reported in patients receiving fingolimod. Although many of these cases occurred after transition from natalizumab, cases have also been described in patients who were natalizumab naïve (168; 14).
The mechanisms by which treatment with these agents leads to PML are incompletely understood. That natalizumab might have direct effects on B cell release from bone marrow and on JCV replication has been discussed above. Exposure over time to multiple immunosuppressive agents appears to increase the likelihood of PML (117). Studies involving natalizumab suggest that use of the agent may increase the rate of seroconversion to JCV (190). Stuve and colleagues reported that CSF of patients treated with natalizumab contains not only reduced numbers of CD4+ and CD8+ T cells, but also reduced numbers of CD10+ B lymphocytes and CD138+ plasma cells as compared to controls (173). These investigators also found that numbers of CSF mononuclear cells remained below expected normal values even after 6 months (173). Niino and colleagues, using a fibronectin layer as a surrogate for capillary endothelium, demonstrated that natalizumab reduced VLA-4 expression and migratory capacity of circulating immune cells and that inhibition of VLA-4 expression and migration varied not only among individual immune cell subsets but also from patient to patient (134). Although neither study specifically addressed the effect of natalizumab on host immune response to JCV, the studies by Stuve and Niino, taken together, suggest that the effects of natalizumab on human immune function – and on lymphocyte trafficking into the CNS – may be complex and also variable among individual patients. The spectrum of effects of the other immunosuppressive reagents associated with PML on host immune response—in particular within the central nervous system—have been less well studied, and the actual effect of any of these agents on JCV persistence and reactivation is not known, in particular when these agents are administered over long periods of time or in combination with other agents. Rituximab, which has been associated with over 50 cases of PML, is of particular interest: the monoclonal reacts with CD20 B cells, but not antibody-producing plasma cells or T cells, so that the immune status of patients treated with the agent might differ significantly from the state of T-cell deficiency found in most PML patients. However, most cases of PML associated with rituximab have occurred in cancer patients treated with multiple agents or in patients with other autoimmune disease. PML appears to be a relatively rare event in patients with multiple sclerosis receiving rituximab) (137). A study of patients with multiple sclerosis without clinical or MRI evidence of PML found that JCV DNA was detected in peripheral blood mononuclear cells (but not blood plasma) in 9 of 33 patients treated with natalizumab and in the CSF of two of these patients (30). JCV DNA was also found in peripheral blood mononuclear cells of three of six patients treated with beta interferons, none of whom had detectable CSF virus.
Pathology. The pathology of PML is distinctive. Although PML usually involves the cerebrum most extensively, pathological changes may also be present in the cerebellum or brainstem (149; 217; 75; 77). Spinal cord involvement is unusual. Grossly, cerebral cortex and deep gray matter appear normal, but areas of retraction within subcortical or deep white matter may indicate myelin loss. Histopathological examination of PML lesions demonstrates loss of oligodendrocytes in demyelinated areas. Surviving oligodendrocytes may have enlarged nuclei or contain actual intranuclear inclusions (149; 217; 75).
Astrocytes in and around PML lesions frequently develop hyperchromatic or multiple nuclei and mitotic figures (149; 217; 75). JCV nucleic acids and early and late viral proteins are present within nuclei of infected oligodendrocytes (75). Atypical astrocytes in PML contain viral nucleic acids but only occasionally express early or late viral proteins (03; 78; 160). Electron microscopic examination of PML-affected brains demonstrate crystalline arrays of viral particles within infected oligodendrocytes (217). Small numbers of viral particles can be found in occasional morphologically normal astrocytes but not within atypical astrocytes (217). Myelin breakdown and lipid-laden macrophages become evident as the disease progresses. Extensive inflammation is unusual in non-AIDS-PML, although small numbers of lymphocytes may be seen around vessels and in demyelinated areas. Demyelination in AIDS-PML is often more extensive than in non-AIDS cases, however, and brains may contain areas of actual necrosis (160; 188). Lymphocytic perivascular infiltrates typical for CNS infection by HIV are often evident and may, at times, be accompanied by parenchymal infiltrates, which may include macrophages and multinucleated giant cells (160; 188). Inflammation may be intense in cases of AIDS-PML developing an immune reconstitution inflammatory syndrome (IRIS) during treatment with ART (189; 119). Although atypical astrocytes have been considered a pathological hallmark for PML, these cells are often infrequent or absent in AIDS-PML brains and may also be rare or absent in non-AIDS cases. It has been recognized that JCV may infect cerebellar granule cells (54; 102; 45; 84; 74). The majority of these cases have been described in individuals with HIV infection, with or without accompanying PML and may be the presenting feature of AIDS (74), but cases have also been described in a child with C40 ligand deficiency (81) and patients with lymphoma both with and without treatment (45; 50). Analysis of JCV DNA from the granule cells of a patient with coexisting PML revealed a unique deletion in the C terminus of the VP1 gene not present in JCV DNA present in the patient’s white matter lesions, suggesting that the deletion, with its accompanying frame shift, may have enabled the virus to infect granule cell neurons (46). Retrospective immunohistological review of paraffin-embedded PML tissues has demonstrated that JCV infection of cerebellar granule cells may be relatively common and has been reported to occur in the absence of demyelinating lesion (211; 74). Wuthrich and Koralnik have also described involvement of cortical neurons in a significant number of patients with PML (212), as well as in hippocampal neurons and glia (210).
PML is almost invariably a disease of immunocompromised patients (75; 77). Prior to the advent of AIDS, the disorder was extremely rare (13). The disorder was most common in older individuals of both sexes and was essentially confined to patients immunocompromised because of hematological malignancies and lymphomas and patients immunosuppressed because of cancer chemotherapy or organ transplantation (13; 75). Occasional additional cases were reported in patients with tuberculosis, sarcoidosis, or nontropical sprue (75). Only a handful of cases were observed in children, most of whom had severe combined immune deficiency syndromes (75). With the advent of the AIDS epidemic, PML became much more common. Prior to the advent of ART, 4% to 6% of AIDS patients could be expected to die of PML (87; 13). The disorder tended to affect younger individuals, and cases were also seen in children with AIDS. The overall incidence and mortality of PML in HIV-infected individuals has decreased with use of ART, with many cases occurring in individuals not attending HIV clinics (57; 28). As discussed above, the occurrence of PML has become of increasing concern in patients treated with natalizumab and other newer immunosuppressive agents (147; 90). As of January 2018, there had been 756 reported cases of PML in patients treated with natalizumab, with a global incidence of 4.19 per 1000 people. Risk of PML in natalizumab patients treated for less than 18 months is 0.96 cases per 1000 patients treated, rising to 1.71 per 1000 patients thereafter. Small numbers of cases have been reported in patients treated with alemtuzumab, dimethyl, or other fumarates or with fingolimod, with or without prior treatment with natalizumab (52; 168; 69; 14; 129; 166; 114). A small number of cases have also been reported in patients with hematological malignancies treated with the anti-CD30 monoclonal antibody-drug conjugate Brentuximab vedotin (27; 147; 77). In contrast to patients treated with other biological agents, onset of PML may occur within weeks of initiation of therapy in brentuximab-treated patients (27; 77).
At this point in time, there are no vaccines or specific antiviral therapies to prevent or control JCV infection. Thus, prevention of PML relies on prevention or treatment of its major associated disorder, HIV infection; avoidance, insofar as possible, of natalizumab treatment in JCV-positive patients; and careful monitoring of patients under aggressive treatment with other immunosuppressive or immunomodulatory agents.
Prevention of PML in HIV infection and AIDS. The use of increasingly sophisticated regimens of ART in this patient group has resulted in a fall in numbers of PML cases from 14.8 of 1000 patients per year in 1996 to 2.6 in 2005 and 0.8 in 2011 (94). Increasingly, PML in AIDS has become associated with patients untreated or poorly compliant with ART.
Prevention of PML in patients treated with natalizumab. Three factors, alone or in combination, appear to increase risk of PML: presence of serum antibody to JCV, duration of natalizumab therapy, and prior use of other immunosuppressive agents (72; 20; 117).
Earlier approaches to prevent PML in patients receiving biological and possibly other immunosuppressive agents have included studies of patient plasma and CSF for the appearance of JCV DNA and analysis of patient sera for the presence of anti-JCV antibodies (155). The utility of screening blood or CSF for JCV DNA sequences has been questioned, however (153), and patients may develop JCV viremia, or even detectable CSF virus, but not progress to PML (215; 30). Dominguez-Mozo and colleagues have reported detection of JCV DNA in peripheral blood mononuclear cells in 23% of multiple sclerosis patients treated with natalizumab over a 3- to 39-month period (51). In patients without PML, viral DNA was archetypal and, in most patients, was present intermittently. In contrast, in two natalizumab-treated multiple sclerosis patients developing PML, both the archetype strain and the potentially neurotropic variants were detected in peripheral blood mononuclear cells and cerebrospinal fluid (51). These data suggest that monitoring of JCV genotype might be of value in assessing PML risk; however, the genomic analysis employed is intrinsically more complex than is antibody testing, and the utility of the method needs confirmation by other investigators.
The presence of antibodies to JCV has been intensely studied as a tool for risk stratification in individuals receiving natalizumab, based on the concept that individuals who are seropositive for JCV will be at a higher risk for developing PML (181; 85; 161; 117). In a study of 2253 patients being tested for anti-natalizumab antibodies, Trampe and colleagues found that 58.8% of patients tested positive for anti-JCV antibodies (181). In that study, all 10 patients developing PML had anti-JCV antibodies (13/15 samples prior to PML diagnosis and 2/15 at the time of diagnosis); antibody titers tended to be higher in these patients than in anti-JCV antibody-positive patients who did not develop PML (181). In evaluating 99,571 patients treated with natalizumab, Bloomgren and colleagues found that all 54 natalizumab-treated patients developing PML had detectable anti-JCV antibodies prior to PML diagnosis (20). In that study, risk of PML was lowest among patients negative for JCV antibodies (≤0.09 cases per 1000 patients treated). In contrast, patients who were positive for anti-JCV antibodies, had taken immunosuppressants before the initiation of natalizumab therapy, and had received 25 to 48 months of natalizumab treatment had an estimated incidence of 11.1 cases per 1000 patients (20). One study has suggested that serum anti-JCV titers may rise prior to the onset of PML (196). In one case report, however, PML was diagnosed in an elderly patient whose JCV serology had remained negative up to 2 weeks before the diagnosis of PML was made (63). It should also be noted that patients may have persistent JCV infection in the absence of antibody response and that antibody titers may vary over time. Careful studies combining detection of anti-JCV antibodies and presence of JCV in urine or plasma as measured by PCR have demonstrated that 37% of seronegative patients may nonetheless have JCV viruria, indicating ongoing persistent infection (15). In a study of 49 patients receiving monthly infusions of natalizumab, Major and colleagues found that 17 of 49 patients (35%) had JCV viremia at some point in time and that six the 17 patients exhibiting viremia were seronegative (116). In the study by Trampe and colleagues, 9.8% of patients converted from seronegative status to seropositive status; however, 4.7% of patients reverted from seropositive status to seronegative, indicating that some individuals may fluctuate in antibody response (181). Taken together, these data indicate that lack of antibody to JCV conveys a significantly lower risk for developing PML, but that seronegativity may not exclude persistent JCV infection.
Use of the JCV antibody index to predict natalizumab-treated patients at risk. The major tool developed for risk assessment in JCV-positive patients treated with natalizumab is the commercially available JCV index. Use of this assay, over time, demonstrated that natalizumab-treated patients developing PML maintained significantly higher anti- JCV antibody index values over time than did patients not developing PML (143). In a study of 285 patients on natalizumab, Sgarlata and colleagues detected statistically significant increase of JCV index during natalizumab treatment period; the authors also detected an increase in antibody index in individuals moved from natalizumab to fingolimod (164). In 2017, Ho and colleagues reviewed data from four different studies (STRATA, STRATIFY-2, TOP, and TYGRIS) and observed that the cumulative risk of PML with a positive JCV serostatus following 72 natalizumab infusions is roughly 27 per 1000 with prior IS exposure, and 17 per 1000 without (85). An index value below 0.20 was regarded as seronegative, that between 0.20 and 0.40 as indeterminate and requiring further testing, and that above 0.40 as evidence for JCV seropositivity. Cut-off points of 0.9 and 1.5 were used to define three categories of JCV index value: low (0.9), medium (> 0.9 to 1.5) and high (> 1.5) (85). An index of greater than 0.9 is associated with increased risk of PML (85; 148). Use of the JCV index has been reviewed by Tugemann and Berger, who noted that index values and their correlation with PML risk differed between Europe and the United States, and that further refinement of the index is needed (183).
Reduction of PML risk by use of extended natalizumab dosing intervals. Natalizumab is conventionally administered at 4-week intervals. Work over the last 4 years, however, has shown that administration of the agent at 6-week intervals preserves effectiveness (213; 61; 154; 142). In a large study of patients enrolled in the natalizumab TOUCH program, Ryerson and colleagues demonstrated that administration of natalizumab at intervals of 35 to 43 days over time preserved drug effectiveness but significantly decreased PML risk (154).
Several other infectious agents that may mimic PML clinically are common in immunosuppressed patients, especially those with AIDS. These include central nervous system invasion by Toxoplasma gondii, Cryptococcus neoformans, Mycobacterium tuberculosis, Candida albicans, or Listeria monocytogenes (22; 110). Central nervous system lymphoma may also mimic PML, particularly in AIDS patients (35; 127; 58; 187). In addition, a relapsing and remitting demyelinating leukoencephalomyelopathy has been described in AIDS patients (18). It must be kept in mind that AIDS and other profoundly immunocompromised patients may develop multiple CNS disorders concurrently; thus, a diagnosis of PML does not exclude the presence of other, more readily treatable conditions. By the same token, the presence of other CNS conditions in such patients does not exclude the possibility that PML may also be present. Brain biopsy to rule out treatable conditions may, thus, be crucial when the diagnosis of PML is at all in question or when there is concern about potentially treatable coexisting conditions.
• PML should be suspected in any immunocompromised patient who develops multifocal neurologic signs, particularly in the setting of HIV infection, aggressive immunosuppression for organ or stem cell transplantation, or treatment with immunomodulatory drugs, especially natalizumab. | |
• MRI is the most important study for detection of white matter injury suggesting PML. It is important, in reviewing MRI images, to remember that PML may begin with a single lesion. | |
• Specific diagnosis of PML is usually made by PCR studies of CSF. Not all patients will have positive CSF, however, and brain biopsy may be necessary in patients in whom the diagnosis is strongly suspected but CSF PCR is negative. |
PML should be considered in any immunocompromised patient, particularly in a patient with HIV infection who develops progressive neurologic deficits and especially if the neurologic examination or MRI suggests involvement of multiple areas of brain. Similarly, PML should be considered in patients with multiple sclerosis treated with natalizumab or other biological agents, in particular brentuximab vedotin, who develop progressive demyelinating disease or in patients on aggressive immunosuppressive medications for organ or stem cell transplant. Symptoms suggesting spinal cord but not cerebral involvement make the diagnosis extremely unlikely. Hematological studies and blood chemistries are unhelpful in the diagnosis of PML. Cerebrospinal fluid is usually normal but may occasionally contain increased protein or, rarely, a lymphocytic pleocytosis (07; 149; 75). CSF oligoclonal bands may be present in a minority of patients (122; 96).
The most useful screening study for PML is magnetic resonance imaging. MRI findings in PML consist of altered signal in subcortical and deep white matter, most clearly evident on FLAIR and T2-weighted scans.
Classically, only a minority of PML patients, usually with AIDS, develop PML lesions exhibiting gadolinium enhancement on MRI (133). Gadolinium enhancement may occur more frequently in patients developing PML in the setting of natalizumab therapy (06; 197). MRI diagnosis of PML may be particularly challenging in patients with multiple sclerosis, in whom there is already multifocal demyelination and in whom gadolinium enhancement may signal active multiple sclerosis lesions (156; 197). In a retrospective review of the first 40 natalizumab-treated patients diagnosed as having PML, Yousry and colleagues identified a number of features suggestive of PML (216). These included lesion size (greater than 3 cm); subcortical localization; presence of lesions in the basal ganglia; hyperintensity on T2, FLAIR, and DWI images with hypointensity on T1 images; lesions that are sharp towards gray matter and ill-defined towards white matter; and absence of atrophy. Additional features suggestive of PML include the presence of small, punctate T2-hyperintense lesions in the immediate vicinity of main lesions and a tendency for lesions to increase in size and for new lesions to appear. Detection of a punctate pattern on T2- weighted imaging of hyperintense or enhancing brain punctate lesions, thought to represent the presence of CD8-positive T-cells within the perivascular spaces, has also been suggested as a biomarker separating PML lesions from those of multiple sclerosis (86). Hyperintensity of lesions on T1-weighted images has been thought to suggest PML-IRIS (PML-IRIS) (198; 62).
For many years, definitive diagnosis of PML during life required brain biopsy with identification of characteristic histological findings or, more reliably, identification of JCV nucleic acids or viral antigens by in situ nucleic acid hybridization or immunohistochemistry (75; 77). Isolation by tissue culture methods was slow and required use of human fetal brain cultures, which were often difficult to obtain. However, PCR methods have come to provide an accurate, highly sensitive method of identifying JCV in CSF in PML patients. In AIDS patients not treated with ART, specific identification of JCV in CSF could be achieved in 80% to 90% of cases (68; 199; 200; 123; 151). PCR has also been used to increase diagnostic accuracy in studies of brain biopsy or autopsy material (160; 180; 184).
It is important to keep in mind, however, that a minority of individuals treated with ART and with PML associated with other conditions may have PML without JCV detectable by PCR in CSF (100; 111). In this regard, a study by Marzocchetti and colleagues provides important information. These workers compared the diagnostic sensitivity of CSF PCR in patients with suspected PML treated with ART to that obtained in CSF of patients studied in the pre-ART era. They found that diagnostic sensitivity of PCR had dropped from 89.5% in the pre-ART era to 57.5% (120). In their study, failure to detect JCV DNA in CSF was associated with exposure to ART at disease onset and with higher CD4 counts. Similar lower diagnostic yield could also occur in non-AIDS patients with less extensive disease. In a review of 28 patients with natalizumab-associated PML diagnosed by MRI and follow-up, Wijburg and colleagues found that 29% had negative CSF PCR for JCV (206). One case has also been reported in which PCR analysis of CSF failed to detect JCV in a case of rapidly PML whereas JCV DNA was readily detected in PCR studies targeting a different region of the genome (108). In an additional case, occurring in a patient with untreated discoid lupus erythematosus, JCV DNA was detected in serum but not in CSF (191). Work by Ferretti and colleagues suggests that PCR analysis of plasma for JCV DNA may complement CSF PCR, in particular where CSF PCR is negative (60). In their study, JCV DNA was detected in plasma of 49 of 103 (48%) patients with PML versus 4 of 144 (3%) of controls without PML. Brain biopsy may be necessary in cases of PML in which CSF PCR is negative (91; 150).
As noted above, approximately 33% of immunosuppressed patients excrete JCV in their urine, and viral excretion detectable by PCR is common in immunologically normal persons. Thus, detection of polyomavirus virions or nucleic acids in urine has usually been considered of little value in the diagnosis of PML (77). Detection of anti-JCV antibody in serum or CSF is of limited value in the diagnosis of PML. In general, PML patients, with or without HIV, may exhibit higher titers of antibody against JCV than do healthy controls; anti-JCV antibodies have been detected in CSF in 76% of cases (201; 181). However, because over half of individuals have serum antibodies to JCV, detection of anti-JCV antibodies, without detailed controls, has little value in the diagnosis of PML. Furthermore, many patients with PML fail to develop a significant rise in antiviral antibody titers within serum or CSF (138; 99). Rare cases of AIDS-PML may be asymptomatic during life and only detected at autopsy (144). As mentioned above, AIDS patients in particular may have multiple ongoing CNS infections. The role of serum antibody detection in risk stratification prior to beginning natalizumab–and possibly other agents–is discussed above.
• Specific antiviral therapy for PML has not been developed. | |
• The most important therapeutic approach to PML is restoration of host-immune response. This may be by using ART in AIDS or reducing or stopping immunosuppressive regiments. PLEX and immunoadsorption have been used in natalizumab-associated PML, but the efficacy of these approaches has been called into question. | |
• Improvement in PML in some patients has been reported in patients treated with filgrastim or with immune checkpoint inhibitors, and in patients receiving infusions of T cells specific for a second, closely related polyomavirus, BK virus. | |
•Patients surviving PML may be left with significant neurologic defects. |
PML, untreated, is a fatal disease. Attempts to control the disease, over time, have involved three approaches: efforts to develop effective antiviral therapy; efforts to alter cellular uptake of JCV; and attempts to halt or reverse the disease by restoring host immune response.
Early studies of therapy for PML were hindered by the extraordinary rarity of the condition, precluding any sort of organized therapeutic trial; the lack of any imaging modality to allow presumptive diagnosis and early institution of treatment; the inability to make a specific virus diagnosis by any procedure short of brain biopsy; and the absence of any means of assessing response to therapy, other than clinical examination. In early individual case reports, treatment of PML with adenine arabinoside or acyclovir was unsuccessful; cytosine arabinoside therapy was associated with clinical improvement in some PML patients but, overall, was largely without benefit (75; 77).
The increase in the number of PML cases associated with AIDS provided both impetus and patient numbers for organized therapeutic trials. These trials were facilitated by the ability of MRI to provide provisional radiological diagnosis of PML and also by the ability of PCR to identify JCV within CSF or stereotactic brain biopsy and to quantitate viral burden of both HIV and JCV in CSF. To date, although there have been reports of improvement in individual case reports, no approach to treatment using antivirals has proven successful. This has included systemic and intrathecal cytosine arabinoside when used in AIDS patients, camptothecin, topotecan, and cidofovir used independently or in combination with ART (75; 118; 48; 77). Although Aksamit and colleagues treated 19 non-AIDS PML patients with intravenous cytosine arabinoside at 2 mg/kg for 5 days and observed neurologic stabilization or improvement in seven patients (02), further studies of cytosine arabinoside have not been carried out in non-AIDS patients. There are several reports of clinical improvement or stabilization in PML patients treated with mefloquine. However, a trial of the agent in PML was terminated early because of failure to document efficacy (39).
An alternative approach used both with and without antiviral treatment has employed agents such as mirtazapine, which binds to the serotonin receptor presumably used by JCV. Like mefloquine, there have been multiple case reports in which these agents have been thought to modify disease course. A review of the literature concerning this agent, however, failed to produce clear evidence that it alters disease course (92).
The most promising approach to the treatment of PML has involved measures to restore host immune response. This was first noted following restoration of CD4 counts in patients with AIDS following ART. Stabilization of PML in patients treated with natalizumab has been reported following discontinuation of the agents, with or without plasma exchange, or immunoadsorption to remove the agent from the circulation (112; 203; 38; 37; 207). The efficacy of these measures has not been clearly proven; however, literature review of 219 cases of natalizumab-associated PML treated with plasma exchange did not find evidence of therapeutic benefit (107), and a second study suggests that plasma exchange may be associated with a longer duration of PML-IRIS (158).
Three novel approaches have been employed to enhance host immune response in PML. In a retrospective series with natalizumab-associated PML treated with subcutaneous filgrastim (granulocyte-colony stimulating factor; G-CSF) with or without plasma exchange, Stefoski and colleagues reported 100% survival (171). A second approach, initially used by Muftuoglu and colleagues, has been to treat patients with BK virus-specific T lymphocytes. These investigators reported treating three immunosuppressed PML patients with ex vivo-expanded, partially HLA-matched, third-party-produced, cryopreserved BK virus-specific T cells (128). Donor-derived T cells could be detected in the CSF of all patients. In two patients, treatment was followed by alleviation of clinical signs and imaging changes, with clearing of JCV from the CSF. The other patient had reduction in viral load and stabilization of symptoms. Rubinstein and colleagues described the course of four younger patients who developed PML in the course of hematopoietic stem cell treatment and received third-party BKV-specific T cells (152). Of the four, one patient treated almost immediately after diagnosis was able to clear JCV from blood and cerebrospinal fluid, with resultant stabilization of neurologic decline. The other three patients, in whom treatment was initiated later in the course of PML, exhibited continued neurologic decline despite reduction in viral load (152).
A third approach to enhancing host immune response has involved the use of immune checkpoint inhibitors (41). Cortese and associates treated eight PML patients, including patients both with and without HIV with pembrolizumab. The authors reported stabilization of disease and fall in viral load (43). In a single case study of a patient treated with pembrolizumab and interleukin 2 (IL-2), Goereci and colleagues reported clearance of JCV DNA from the CSF of a PML patient, accompanied by minimal clinical improvement (Goereci e al 2020). Mahler and colleagues have reported MRI stabilization in two PML patients treated with pembrolizumab and IL-2, one of whom exhibited marked clinical improvement (115). Walter and colleagues reported similar disease stabilization and fall in viral load following treatment with a second check point inhibitor, nivolumab (195). This patient, it should be noted, did not have an underlying disease and was not known to be immunocompromised. Volk and colleagues reported stabilization of PML in two of five patients treated with pembrolizumab (192). Although these early reports suggest possible therapeutic efficacy, it is clear that not all patients respond. Lambert and colleagues, in a literature review of 35 cases of PML treated with immune checkpoint inhibitors, found that a history of therapeutic immune suppression and the use of an immunosuppressive therapy at treatment initiation were significantly associated with a poor response (105).
PML and immune reconstitution inflammatory syndrome (IRIS). Patients with AIDS-PML may develop an immune reconstitution inflammatory syndrome with initiation of ART (189; 182; 178; Blanchardière 2017; 62). Immune reconstitution inflammatory syndrome may also be seen with reduction of immunosuppressive therapy in transplant recipients, other severely immunosuppressed patients, or patients treated with monoclonal agents in whom the agent was withdrawn and plasma exchange or other measures were used to reduce levels of circulating monoclonal (112; 203; 177; 126; 77). Development of IRIS in PML patients withdrawn from natalizumab appears to be almost universal and has also occurred following transition of patients from natalizumab to fingolimod (97; 26). The syndrome is characterized by severe CNS inflammation and cerebral edema, which may lead to brain herniation and death. In this setting, treatment with methylprednisolone or other corticosteroids may be lifesaving (178). Decompressive craniotomy has been used in one case of PML-IRIS that had caused severe cerebral edema (179). Immune reconstitution inflammatory syndrome appears to be more common in patients developing PML in association with natalizumab and then withdrawn from the agent than it is in AIDS-PML patients following initiation of ART (38). In a few individual reports, the chemokine receptor 5-positive (CCR5+) T antagonist, maraviroc, was reported to attenuate the severity of IRIS in a patient withdrawn from natalizumab after diagnosis of PML (67). However, a subsequent report of three patients with natalizumab-PML documented failure of maraviroc to alter the course of PML or to protect against PML-IRIS (159).
Remission of non-AIDS PML was reported in older literature, both spontaneously and following reduction of immunosuppressive medications (163; 138). Remission of AIDS-PML was first reported during zidovudine-induced stabilization of the patient's HIV infection (17; 40). Increasing sophistication of ART over time has significantly reduced the frequency of PML in patients with AIDS. In a study from Spain, Casado and colleagues reviewed a series of 72 AIDS patients under care from 1996 to 2011 (28). Mortality in 1996, prior to combined antiretroviral treatment was 14.8 of 1000 cases per year, falling to 0.8 in 2011. Sixty-four percent of cases in 2011 were patients not attending regular clinic visits (28). During this period, 1-year survival increased from 55% to 79% at 1 year for patients with CD4+ count above 100 cells/mm3 at diagnosis. Prolonged survival (up to 12 years) has also been reported (214). However, AIDS-PML remains a dangerous disease, and not all AIDS-PML patients experience significant remission during ART treatment. In a study of patients with PML in the setting of HIV, outcome was dismal: death or referral to hospice in patients with PML alone was 47% and 82% with PML-IRIS (175). In this study, median survival time was 266 days in PML patients and 109 patients in the PML-IRIS group (175).
The course of PML in patients receiving monoclonal agents or aggressive immunosuppressive regimens tends to be more favorable because these individuals off treatment have fundamentally normal immune systems and their immunocompromised state can be reversed by withdrawal of treatment or other enhancement of their immune response. As of February 2017, mortality for individuals with PML associated with receiving natalizumab for multiple sclerosis and Crohn disease was 23%. This is a much better survival rate than in PML associated with any other condition. However, the majority of survivors have been left with significant neurologic deficits. As noted above, many of these cases also underwent plasma exchange or immunoadsorption to remove residual circulating natalizumab, although the efficacy of these measures is questionable (107; 158). Development of immune reconstitution inflammatory syndrome (IRIS) is a major concern in these patients, and onset of PML-IRIS de novo frequently occurs after discontinuation of natalizumab (177; 66).
Pregnancy has been reported to increase urinary excretion of JCV but is not accompanied by an increase in the incidence of PML.
Anesthesia has no effect on the occurrence or course of PML.
All contributors' financial relationships have been reviewed and mitigated to ensure that this and every other article is free from commercial bias.
John E Greenlee MD
Dr. Greenlee of the University of Utah School of Medicine has no relevant financial relationships to disclose.
See ProfileNearly 3,000 illustrations, including video clips of neurologic disorders.
Every article is reviewed by our esteemed Editorial Board for accuracy and currency.
Full spectrum of neurology in 1,200 comprehensive articles.
Listen to MedLink on the go with Audio versions of each article.
MedLink®, LLC
3525 Del Mar Heights Rd, Ste 304
San Diego, CA 92130-2122
Toll Free (U.S. + Canada): 800-452-2400
US Number: +1-619-640-4660
Support: service@medlink.com
Editor: editor@medlink.com
ISSN: 2831-9125
Peripheral Neuropathies
Nov. 16, 2024
Infectious Disorders
Nov. 15, 2024
Infectious Disorders
Nov. 12, 2024
Neurogenetic Disorders
Oct. 23, 2024
Neuroimmunology
Oct. 10, 2024
Infectious Disorders
Oct. 08, 2024
Neuro-Oncology
Oct. 03, 2024
Neuroimmunology
Oct. 03, 2024