|Home | About | Journals | Submit | Contact Us | Français|
Progressive multifocal leukoencephalopathy (PML) has become much more common with monoclonal antibody treatment for multiple sclerosis and other immune-mediated disorders.
We report 2 patients with severe psoriasis and fatal PML treated for ≥3 years with efalizumab, a neutralizing antibody to αLβ2-leukointegrin (LFA-1). In one patient, we conducted serial studies of peripheral blood and CSF including analyses of leukocyte phenotypes, migration ex vivo, and CDR3 spectratypes with controls coming from HIV-infected patients with PML. Extensive pathologic and histologic analysis was done on autopsy CNS tissue of both patients.
Both patients developed progressive cognitive and motor deficits, and JC virus was identified in CSF. Despite treatment including plasma exchange (PE) and signs of immune reconstitution, both died of PML 2 and 6 months after disease onset. Neuropathologic examination confirmed PML. Efalizumab treatment was associated with reduced transendothelial migration by peripheral T cells in vitro. As expression levels of LFA-1 on peripheral T cells gradually rose after PE, in vitro migration increased. Peripheral and CSF T-cell spectratyping showed CD8+ T-cell clonal expansion but blunted activation, which was restored after PE.
From these data we propose that inhibition of peripheral and intrathecal T-cell activation and suppression of CNS effector-phase migration both characterize efalizumab-associated PML. LFA-1 may be a crucial factor in homeostatic JC virus control.
Progressive multifocal leukoencephalopathy (PML) is a serious demyelinating disease of the CNS occurring predominantly in immunocompromised patients (reviewed in Koralnik1) and is caused by JC virus (JCV), a 40-nm, nonenveloped, double-stranded DNA, icosahedral polyomavirus.2 Serologic studies suggest prior JCV infection of 50%–80% of the general population.3 A neurotropic spread of replicating virus with development of PML occurs in diverse settings such as HIV infection, lymphoproliferative disorders, and organ transplantation.4 The risk of PML is increased by some immunomodulating therapies, including mycophenolate mofetil and monoclonal antibodies such as the anti-CD20 monoclonal antibody therapy rituximab.5 Particular interest has recently focused on the association of PML with natalizumab,6 a humanized monoclonal antibody directed against α4-integrin approved for treatment of relapsing forms of multiple sclerosis7,8 and Crohn disease.9
Efalizumab (Raptiva; Genentech, San Francisco, CA, and Merck-Serono, Geneva, Switzerland) is a humanized monoclonal antibody to the α-subunit of the integrin LFA-1 (CD11a) and was approved for the treatment of severe and therapy-refractory psoriasis until it was taken off the market in 2009 after the observation of PML. Efalizumab efficiently blocks CD11a on cells such as lymphocytes,10 thereby inhibiting their migration into the target tissue. In addition, efalizumab downregulates CD11a on target cells.11 In October and November 2008, 3 fatal cases of PML in patients treated with efalizumab were reported by the manufacturers.12,13 Here, we further describe 2 of those cases, including a detailed immunologic study of one case.
In both cases of PML, immunohistochemical analysis of brain tissue was performed to illustrate pathology, JCV infection, and immune cell infiltrates. For patient 1, we combined several methods for an immunologic workup: flow cytometric stainings of surface molecules and of bound efalizumab, transmigration assays over primary human endothelial cells, and T-cell receptor repertoire analysis via multiplex PCRs. A detailed description of the performed protocols can be found in e-Methods on the Neurology® Web site at www.neurology.org.
The local ethics committees in Würzburg and Cleveland approved the use of human subjects for this study. All patients gave informed written consent to participate in the study at a time when they were still competent to understand the usefulness of the study.
An HLA-A2+, 47-year-old man presented with progressive left hemiparesis, dysarthria, and memory difficulty. Past medical history was notable for psoriasis, for which he was treated sequentially with retinoic acid, PUVA therapy, fumaric acid, and methotrexate (November 2002–May 2003). He was then treated with efalizumab (1 mg/kg by weekly subcutaneous application) from 2006 on for 3 years before the diagnosis of PML was made. Efalizumab was discontinued 3 months after symptom onset. Brain MRI at that time showed multiple foci of increased T2 signal in the frontal and parietal lobes bilaterally. CSF showed 5 cells/μL, all mononuclear, with normal protein, negative oligoclonal bands, and detection of JCV by PCR (1,860 copies/mL), all confirming PML. Complete CSF data are found in figure 1.
Treatment for PML included 5 courses of plasma exchange (PE) (30 mL/kg body weight per exchange) over 11 days, to accelerate depletion of efalizumab14 and mefloquine (1,000 mg loading dose, then 250 mg PO weekly), based on preliminary reports of antiviral effects in vitro.15 Over the following several weeks, left hemiparesis progressed and left-sided neglect was observed. Repeat CSF studies showed normal cell count, but increased JCV load (20,500 copies/mL). In view of the relentlessly progressive neurologic disorder, a prolonged course of IV immunoglobulin (240 g total) was given with the rationale that it might downregulate a potentially fulminant immune reconstitution inflammatory syndrome (IRIS), although there is currently no direct evidence supporting this treatment. The clinical signs and course of IRIS observed in this patient appeared rather unusual and did not present as an acute or subacute neurologic deterioration. Still, the tentative diagnosis of IRIS was made in view of all clinical, imaging, and CSF analyses. The patient developed seizures and further declined mentally. MRI showed further progression of CNS lesions, but no contrast enhancement. CSF studies now demonstrated pleocytosis (138 cells/μL) and further increased JCV load (106,000 copies/mL), supporting the diagnosis of IRIS. Seizures were successfully treated with valproic acid. Over the next weeks, the patient became stuporous and died in coma 6 months after symptom onset. A complete autopsy was performed.
A 70-year-old man presented with a 1-week history of depression, headache, social withdrawal, language difficulties, and peculiar behavior. Past medical history was notable for psoriasis for 22 years, during the last 4 of which he received efalizumab (1 mg/kg subcutaneous weekly). Before treatment with efalizumab, the patient did not take any systemic medications for psoriasis and received only topical creams, including those containing coal tar solution, salicylic acid, and sulfasalazine. Brain MRI showed increased T2 signal in the bilateral cerebellar, left parietal, occipital, and right medial temporal lobes, and, in addition, a left posterior temporal venous angioma (figure e-1). Mental status declined over the next several weeks. Repeat brain MRI showed progression of brain lesions, none of which enhanced. Results of routine CSF studies (which did not include JCV PCR) were unremarkable. Efalizumab was discontinued 1 month after symptom onset.
Mental status continued to decline over the next several weeks. He developed marked ataxia, visual tracking difficulty, and suicidal ideation. Repeat CSF testing showed JCV by PCR, and PML was diagnosed. He was treated with 3 courses of PE (1 plasma volume per exchange) and mefloquine (750 mg PO weekly).
Despite therapy, he developed right hemiparesis and became increasingly confused and agitated. Repeat brain MRI 10 days after PE showed extensive confluent lesions with laminar necrosis in the lateral left parietal lobe and enhancement involving the medial left occipital-parietal lobe. He became obtunded and died 2 months after onset of symptoms. A brain-only autopsy was performed.
The macroscopic and microscopic features of both brains were similar and diagnostic of PML. No other significant neuropathology was observed. Macroscopically, multiple regions of softening and cavitation were seen in the white matter of the cerebrum (figure 2A) and cerebellum. Microscopically, these lesions corresponded to myelin loss (figure 2, B and C) with relative sparing of axons at a variable degree (not shown). The demyelinated areas ranged from large and confluent lesions in regions of gross abnormality (figure 2B) to multiple small foci of demyelination in macroscopically normal–appearing areas (figure 2C). Demyelinated areas showed macrophages with Luxol fast blue inclusions, reactive astrocytes, and oligodendrocytes with intranuclear viral inclusions (figure 2D). Affected oligodendrocytes were immunoreactive with antibodies to JCV antigens (figure 2E),16 and electron microscopy demonstrated viral particles (figure 2F). Perivascular lymphocytes were conspicuous in many lesions (figure 2G). CD3+ T lymphocytes were both perivascular and intraparenchymal (figure 2H). The majority of CD3+ T cells were CD8+ (figure 2I). CD4+ T cells and CD20+ B cells were also present (figure 2, J and K). In some lesions, plasma cells were identified by morphology and CD138 immunostaining (figure 2, L and M).
Serial peripheral blood and CSF analyses were performed in patient 1 (peripheral blood at 8 time points and CSF at 5 time points). Assessments included efalizumab binding, migration capacity, CD4+ and CD8+ T-cell spectratypes (which indicate clonal expansions), naive, memory, and effector phenotype, and JCV specificity. Data comprised effects on CD4+ and CD8+ T cells of PML during efalizumab treatment and the response to PE-mediated removal of efalizumab.
Efalizumab binding to T cells diminished after PE and reached the level found in untreated subjects about 3 weeks after the final PE (figure 3A).
LFA-1, the molecular target of efalizumab, is involved in T-cell migration. To mimic the influence of efalizumab on the passage of immune cells across the blood-brain barrier (BBB),17 we assessed the migration of peripheral blood mononuclear cells (PBMCs) in an in vitro model of the human BBB.18 On admission (e.g., before PE and in the presence of efalizumab), PBMCs migrated poorly. The migratory capacity began to increase 2 weeks after the initial PE and increased steadily thereafter (figure 3B). In line with the notion that efalizumab not only blocks LFA-1 but also induces downregulation of LFA-1 expression,19 T cells showed steadily increasing levels of surface LFA-1 after PE (figure 3C). Moreover, migrated CD8+ T cells showed a higher expression of LFA-1 than their nonmigrated counterparts (figure 3C), consistent with a crucial role of LFA-1 in this migration assay. As disease controls, we evaluated peripheral blood cells from patients with HIV-associated PML (n = 2) in this assay. The migratory rate of cells from patients with HIV-associated PML was similar to that of cells from healthy donors (n = 3), excluding the possibility that PML itself is associated with diminished migration. After PE, migration of cells obtained from the patients with efalizumab-associated PML was still impaired compared with that of controls (figure 3D).
Using CDR3 spectratyping, we assessed the T-cell receptor (TCR) repertoire at several time points in peripheral blood and CSF during life and in different CNS regions at autopsy in patient 1. Spectratyping allows detection of clonal T-cell expansions, indicating a T-cell–mediated immune response.20 Before PE, cells in the CSF showed few clonal expansions, indicating that the extent of CNS antigen-specific responses was limited. After PE, the numbers of clonally expanded cells rose, although the absolute cell numbers remained low. The presence of clonally expanded T cells in the CSF thus hints toward an early reconstitution of T-cell function (figure 3E). Subsequent pleocytosis was temporally associated with the rise of Gaussian-distributed TCRs (figure 1), putatively indicating restored T-cell migration into the CSF and intrathecal antigen-mediated restimulation.21 This pleocytosis coincided with the onset of IRIS (figure 3F, blue arrow). All CSF samples analyzed had a red blood cell count of <1/μL.
After PE, we used flow cytometry to assess naive, central memory (Tcm), and effector memory (Tem) bulk CD4+ and CD8+ T cells and other immune cell populations, both in the periphery and in the CNS22,23 (figure 3G, table e-1). Immediately after completing PE, there was a predominance of naive (CCR7+CD45RA+) CD4+ and CD8+ T cells in peripheral blood (47% and 54%, respectively), similarly to those observed before PE (table e-1) and in HIV-positive patients with PML, as well as in healthy donors. Within 3 weeks after the last PE, however, peripheral CD4+ T cells in the blood had shifted to be mostly Tcm (68%), whereas CD8+ T cells were mainly Tem (53%). At this time point, CSF showed mild pleocytosis (14 cells/mL) with 78% of CD8+ cells and 43% of CD4+ T cells being Tem, strikingly different from the virtual absence of Tem from CSF of healthy subjects or patients with multiple sclerosis. CD4+ Tem in the periphery did not reach high numbers until the last available time point (5 weeks after PE). Of note, we did not observe an increase in the percentage of CD4+CD8+ double-positive T cells during or after PE in the periphery or in the CSF (data not shown).
We assessed JCV-specific CD8+ T cells at 2 time points during the course of the disease: before PE (day 10 in the hospital) and after PE (day 22 in the hospital) (figure 3, H and I). Using the HLA-A*0201/JCV VP1p36 peptide to stimulate total PBMCs for 14 days in vitro, we did not find any evidence of JCV-specific memory CD8+ T cells either before or after the PE. There was no expansion of JCV tetramer–specific CD8+ T cells or secretion of inflammatory cytokines (interferon-γ and tumor necrosis factor-α) by these cells. In line with these data, spectratyping of T cells in CNS, in CSF, and in the periphery did show various clonal expansions, none of which could be attributed to JCV specificity (figure e-2A). However, there were clonal expansions detected in more than one tissue compartment. One notable Vβ6+ CD8+ clone (Vβ6 CASS LGAT NEKLF Jβ1.4) was prominently expanded in CSF, CNS, and peripheral blood at more than one time point. We assume that this clone is myelin-specific as suggested by another study.24 This notion is in line with our data showing massive, putatively non-JCV–related T-cell reactions in periphery and CSF.
TCR repertoire and viral load were evaluated in various regions. The MRI scans taken at 2 time points are shown to illustrate the sites of progressive CNS pathology over a long time period (figure 4A). CNS tissues taken from the regions indicated in figure 4B showed various degrees of clonal expansions (figure e-2B), with white matter containing larger numbers of expansions than gray matter (figure 4B).
We present 2 fatal cases of efalizumab-associated PML, both after a similar course of disease. In patient 1, we performed serial T-cell phenotyping and functional assessments of T cells from blood, CSF, and brain tissue.
Three months after onset of clinical PML, the majority of peripheral T cells in patient 1 showed a naive phenotype consistent with lack of differentiation toward effector functions, which was in line with the lack of clonal expansions in CSF and periphery. As expected, blockade of LFA-1 also dramatically impaired T-cell migration.
Removal of efalizumab by PE induced a series of immunologic changes. Immediately after PE, clonal expansions in the CSF indicated the reconstitution of the proliferative function of primed central T cells. The subsequent appearance of effector memory lymphocytes (CD8+ T cells 3 weeks after PE and CD4+ T cells 5 weeks after PE) indicates a strong peripheral immune reaction with the potential for enhanced T-cell migration across the blood-CSF barrier.25 The absence of a virus-specific immune response before and after PE was documented by lack of JCV-specific memory CD8+ T cells and was observed despite several months of clinically evident PML.26 We propose that the increase in clonal CD8+ T-cell numbers seen after PE in patient 1 represented a poly-specific expansion of CD8+ T cells rather than of JCV-specific CD8+ T cells. This hypothesis is consistent with the observed increase in CSF viral load after PE, plausibly related both to insufficient virus control and BBB malfunction along with incipient IRIS. This time point of the disease course (hospital day 22) also corresponded to the restoration of memory CD8+ T-cell activation and accumulation of CD8+ T cells in the CSF, a site of immune surveillance where T cells are restimulated by local antigen presentation.17,27–30 The restimulation of lymphocytes in the subarachnoid space requires physical interactions between T-cell LFA-1 and antigen-presenting cell (for CD4+ T cells) or target cell (for CD8+ T cells) intercellular adhesion molecule-1.31 Lymphocyte restimulation by antigen elicits production of cytokines, thereby influencing the cerebrovascular endothelium and promoting transmigration of leukocytes across the BBB.25 Thus, efalizumab seems to impair 3 aspects of adaptive antiviral immunity: 1) activation and proliferation of naive T cells; 2) restimulation of memory T cells with antigen; and 3) T-cell migration to sites of infection.17 In agreement with the hypothesis that T-cell surveillance is crucial to hold JCV in check, the migration of T cells was indeed impaired: in vitro BBB assays showed that cells from patient 1 migrated poorly before PE and slowly regained migratory function after efalizumab removal, but never reaching healthy control levels. Improvement of migration was observed in vitro after about 50% of cell-associated efalizumab was removed, corresponding to recovery of LFA-1 expression on lymphocytes.
The central phase of immune reconstitution was followed by recovery of peripheral immune cell function, including T-cell migration and maturation. Subsequent parenchymal inflammation associated with IRIS was reflected by increasing lesions on MRI without contrast enhancement and clinically by newly occurring seizures along with deterioration of vigilance.
Features of this patient's immunologic evaluation suggest that the mechanisms by which efalizumab and natalizumab lead to PML are different. First, efalizumab does not cause release of CD34+ bone marrow progenitor cells into the circulation (table e-1), a prominent effect of natalizumab, which could possibly lead to release of virus from a bone marrow site of nonpathogenic replication.32 Moreover, natalizumab, but not efalizumab, reduces CSF cell counts by more than 70%.33 Because T-cell immune surveillance of the CNS takes place in the CSF-containing subarachnoid space,34 natalizumab limits cells available to perform this function. In contrast, efalizumab appears to impair intrathecal antigen-presenting cell/antigen-mediated restimulation of CD4+ and for CD8+ T cells. Both natalizumab and efalizumab block and downregulate leukocyte-derived integrins, through which T cells cross the BBB to eliminate pathogens and virus-infected cells, and both monoclonal antibodies have similar pharmacokinetics.35,36 Importantly in patient 1, most of the clonally expanded T cells were found in the CNS. Only occasionally were individual expanded T-cell clones found in blood, CSF, and CNS tissue because the blockade of LFA-1 prevented migration of effector cells across the BBB. Clonal expansions in the CNS were numerous, indicating that the immune activation and T-cell expansion in patient 1 probably occurred within the parenchyma. Of note, in comparison with natalizumab, for which roughly 75% of patients with PML survive,37 all 3 patients with efalizumab-related PML died. From our data, it can be speculated that interference with LFA-1 might be associated with several steps of JCV defense including immune cell migration, T-cell priming and subsequent restimulation, and target cell lysis.
Our data allow formulation of a provisional hypothesis relating to JCV immune surveillance under physiologic conditions, its failure due to efalizumab, and the immune reconstitution after PE (figure e-3). We suggest that efalizumab inhibits antiviral T-cell activation and expansion as well as migration across the BBB, intrathecal lymphocyte restimulation with antigen, and target cell lysis, leading to development of PML. The cases described in this report may help to conceive strategies to interfere with the malfunction of immune cells and of the BBB that are relevant for the prevention and effective treatment of PML.
The authors thank Prof. Ortwin Adams, Dr. Linda Bonzel, Prof. Ralf Gold, Prof. Justus Müller, and Prof. Laszlo Solymosi for their support, Andrea Staudigel, and Theresa Moritz for excellent technical assistance, and Dr. James McMahon, Cleveland Clinic Anatomic Pathology, for performing and reading the electron microscopic preparations.
Supplemental data at www.neurology.org
Dr. Schwab performed experiments, analyzed the data, and wrote the first draft of the manuscript. Dr. Ulzheimer coordinated patient recruitment and participated in writing the first draft of the manuscript. Dr. Fox coordinated experiments and edited the manuscript. T. Schneider-Hohendorf performed experiments and analyzed the data. Dr. Kieseier coordinated patient recruitment and edited the manuscript. Dr. C. Monoranu coordinated sample processing and analyzed neuropathology sections. Dr. Staugaitis performed histologic analyzes, coordinated sample processing, and edited the manuscript. Dr. Welch coordinated patient recruitment and sample logistics. Dr. Jilek performed experiments and analyzed the JCV immunology data. Dr. Du Pasquier analyzed the JCV immunology data and edited the manuscript. Dr. Brück analyzed neuropathology sections and edited the manuscript. Dr. Toyka coordinated patient care and recruitment and edited the manuscript. Dr. Ransohoff coordinated experiments and data analyses, and participated in writing and editing the manuscript. Dr. Wiendl coordinated experiments, data processing, and participated in writing and editing the manuscript.
Dr. Schwab reports no disclosures. Dr. Ulzheimer has received speaker honoraria from Novartis and sanofi-aventis; served as a consultant for Allergan, Inc. and Biogen Idec; and has received funding for travel from Merck Serono and Bayer Schering Pharma. Dr. Fox has received speaker honoraria from Biogen Idec and Teva Pharmaceutical Industries Ltd. has served as a consultant for Biogen Idec, Avanir Pharmaceuticals, Genentech, Inc., and Novartis; has served on clinical trial advisory committees for Biogen Idec and the US Department of Defense (Army); has received/receives research support from Biogen Idec and the National Multiple Sclerosis Society (RG 4091A3/1; RG 4103A4/2; RC 1004-A-5), which includes funding to study CCSVI; and serves on the editorial boards of Neurology® and Multiple Sclerosis. T. Schneider-Hohendorf reports no disclosures. Dr. Kieseier has served on scientific advisory boards for Biogen Idec, Novartis, sanofi-aventis, Roche, and Merck Serono; serves on speakers' bureaus for Biogen Idec, Novartis, sanofi-aventis, Bayer Schering Pharma, Merck Serono, and Teva Pharmaceutical Industries Ltd.; and has received speaker honoraria, funding for travel, and research support from Bayer Schering Pharma, Biogen Idec, Merck Serono, Novartis, Roche, sanofi-aventis, Talecris Biotherapeutics, and Teva Pharmaceutical Industries Ltd. Dr. Monoranu and Dr. Staugaitis report no disclosures. Dr. Welch serves on speakers' bureaus for and has received funding for travel and speaker honoraria from Forest Laboratories, Inc. and Novartis and serves/has served as a consultant for Teva Pharmaceutical Industries Ltd., Biogen Idec, GlaxoSmithKline, and Bayer Schering Pharma. Dr. Jilek reports no disclosures. Dr. Du Pasquier serves on scientific advisory boards for Biogen Idec, Merck Serono, and Novartis; has received funding for travel or speaker honoraria from Biogen Idec, sanofi-aventis, Merck Serono, and Bayer Schering Pharma; serves on the editorial boards of the European Journal of Neurology and the Journal of Neurovirology; and receives research support from the Swiss National Foundation, the Swiss Society for Multiple Sclerosis, and the Biaggi Foundation. Dr. Brück serves on scientific advisory boards for Teva Pharmaceutical Industries Ltd./sanofi-aventis and Biogen Idec; serves on speakers' bureaus for and has received speaker honoraria from Teva Pharmaceutical Industries Ltd., sanofi-aventis, Merck Serono, Biogen Idec, and Bayer Schering Pharma; and serves on the editorial boards of Acta Neuropathologica, Brain Pathology, and Multiple Sclerosis International. Dr. Toyka serves on scientific advisory boards for Medac GmbH, Bayer Schering Pharma, Teva Pharmaceutical Industries Ltd., sanofi-aventis, and Merck Serono, as Chairman of the Medical Advisory Board of the German MS Society, and as Chairman of the Multiple Sclerosis Treatment Consensus Group, Europe; has received funding for travel and speaker honoraria from Bayer Schering Pharma, Talecris Biotherapeutics, Lundbeck, Inc., and Merck Serono; serves on the editorial boards of Current Treatment Options in Neurology and Therapeutic Advances in Neurological Disorders; serves as a consultant for Bayer Schering Pharma, sanofi-aventis/Teva Pharmaceutical Industries Ltd., Merck Serono, EBEWE Pharma, and Rehab Clinic Bavaria, Germany; and receives research support from sanofi-aventis, Teva Pharmaceutical Industries Ltd., Bayer Schering Pharma, Merck Serono, DFG (German Research Foundation), BMBF (German Federal Ministry of Education and Science), Binational German-Polish Grant for MS Dendritic Cell Research, State of Bavaria Research Funds and University Research Funds, Gemneinnützige Hertie Foundation (GHS), and the German MS Society Research Fund. Dr. Ransohoff serves on scientific advisory boards for ChemoCentryx, Inc., Vertex Pharmaceuticals, Merck Serono, Wyeth, and GlaxoSmithKline; has received fees for preclinical consulting or honoraria for academic presentations from Novartis, Johnson & Johnson, Biogen Idec, Boehringer Ingelheim, Merck Serono, Pfizer Inc, and Teva Pharmaceutical Industries Ltd. serves as an Associate Editor for Neurology®; and receives research support from the NIH, the National MS Society, the Dana Foundation, and the Nancy Davis Center Without Walls. Dr. Wiendl has received funding for travel and speaker honoraria from Bayer Schering Pharma, Biogen Idec/Elan Corporation, sanofi-aventis, Merck Serono, and Teva Pharmaceutical Industries Ltd.; has served/serves as a consultant for Merck Serono, Medac, Inc., sanofi-aventis/Teva Pharmaceutical Industries Ltd., Biogen Idec, Bayer Schering Pharma, Novartis, and Novo Nordisk; and receives research support from Bayer Schering Pharma, Biogen Idec/Elan Corporation, sanofi-aventis, Merck Serono, and Novo Nordisk.