What Drugs Cause PML (Progressive Multifocal Leukoencephalopathy)?

Several classes of immunosuppressive and immunomodulatory drugs have been linked to progressive multifocal leukoencephalopathy, a rare but often devastating brain infection caused by the JC virus. The best-known culprit is natalizumab (Tysabri), used in multiple sclerosis and Crohn’s disease, but PML cases have also been tied to rituximab, fingolimod, dimethyl fumarate, certain chemotherapy agents, transplant drugs, and newer targeted therapies like ibrutinib. The common thread is not a single mechanism but a shared outcome: deep suppression of the immune surveillance that normally keeps JC virus in check.

How JC Virus and Immune Suppression Set the Stage

Most people carry JC virus (JCV) without knowing it. The virus typically sits dormant in the kidneys, lymphoid tissue, and even the brain, causing no symptoms as long as the immune system keeps it under control.¹ When that immune surveillance breaks down, JCV can reactivate and infect oligodendrocytes, the brain cells responsible for producing myelin. The virus replicates inside these cells until they burst, releasing more virus to neighboring cells and progressively destroying the protective myelin coating around nerve fibers.² This spreading destruction of white matter is what causes PML’s symptoms: weakness, vision loss, cognitive decline, and difficulty with speech or coordination.

PML is not simply a drug side effect. It occurs in any context where the immune system is deeply impaired, including untreated HIV/AIDS, certain blood cancers, and organ transplantation. What drugs do is create a window of vulnerability by depleting or trapping the very immune cells that would normally patrol for JCV reactivation. The risk differs sharply between drug classes and even between individual patients on the same drug, which makes understanding the specific agents and their risk profiles genuinely useful.

Natalizumab and the Highest Drug-Associated PML Risk

Natalizumab, sold as Tysabri, carries the strongest and most extensively documented association with PML of any single drug. It works by blocking a protein on the surface of immune cells that helps them cross the blood-brain barrier. This is effective for calming the brain inflammation of multiple sclerosis and Crohn’s disease, but it also prevents immune cells from reaching the brain to fight JCV. The U.S. Food and Drug Administration imposed a Risk Evaluation and Mitigation Strategy (REMS) program specifically because of this danger, estimating the incidence of PML at roughly 11 per 1,000 among patients with the highest-risk profile: those who carry JCV antibodies, have previously used other immunosuppressants, and have been on natalizumab for more than two years.³

Three factors define the risk stratification doctors use before and during natalizumab therapy. The first is JCV antibody status: patients who test negative for JCV antibodies have an extremely low risk of PML, because the virus is probably not present in their body to reactivate. The second is treatment duration, with risk climbing after roughly two years of continuous use. The third is whether the patient has previously taken another immunosuppressant.⁴ Beyond simple positive-or-negative antibody testing, a quantitative measure called the JCV antibody index helps further sort risk. Patients with anti-JCV antibody levels at or below an index of 0.45 appear to remain at low cumulative risk regardless of whether they previously used other immunosuppressants.⁵ Among patients without prior immunosuppressant use, antibody index levels are significantly higher in those who go on to develop PML than in those who do not.⁶

Monitoring under the natalizumab REMS program includes periodic MRI surveillance and JCV antibody testing. The current recommendation involves monitoring for PML at three and six months after initiation and every six months thereafter.³ If a patient’s antibody index rises or other risk factors accumulate, the decision often shifts to discontinuing natalizumab and switching to a different therapy.

Rituximab and Other Anti-CD20 Antibodies

Rituximab targets CD20, a protein found on B cells, and has been approved for non-Hodgkin lymphoma, chronic lymphocytic leukemia, rheumatoid arthritis, and certain types of vasculitis. By depleting B cells, it creates a prolonged gap in one arm of the immune system, which can allow JCV to gain a foothold.⁷ A large adverse-event analysis identified 57 PML cases in HIV-negative patients who had received rituximab. The overwhelming majority, 52 of 57, had underlying lymphoproliferative disorders, while a handful had autoimmune conditions like lupus or rheumatoid arthritis.⁸

For patients using rituximab for rheumatoid arthritis or vasculitis, the absolute risk is very small. A post-marketing analysis estimated the reporting rate at about 2.56 per 100,000 patients for rheumatoid arthritis and fewer than 1 per 10,000 for the vasculitis indications. Every patient who developed PML had at least one additional risk factor beyond rituximab itself, such as prior immunosuppressant use or the underlying autoimmune disease.⁹ This makes it genuinely difficult to separate the drug’s contribution from the underlying disease, a challenge that recurs across nearly every drug linked to PML.

Other anti-CD20 monoclonal antibodies, including ocrelizumab (widely used in MS), ofatumumab, and obinutuzumab, share the same B-cell depleting mechanism and have also appeared in PML case reports and FDA adverse-event databases. Ofatumumab and obinutuzumab were among the newer drugs with notable proportional reporting ratios for PML in an FDA adverse-event review.¹⁰

Fingolimod, Dimethyl Fumarate, and Other MS Therapies

Fingolimod (Gilenya) was the first oral disease-modifying therapy approved for MS and works by trapping lymphocytes inside lymph nodes, reducing the number circulating in the blood. This peripheral lymphopenia appears to raise the risk of opportunistic infections, including PML.¹¹ PML cases linked to fingolimod have been reported mostly in patients who had previously been on natalizumab or another immunosuppressant, which complicates the picture. Still, cases in patients without prior immunosuppressive therapy have also surfaced, and the drug’s label now warns about PML.

Dimethyl fumarate (Tecfidera), another oral MS drug, works through a different mechanism but also reduces lymphocyte counts in some patients. An analysis of PML cases among dimethyl fumarate-treated MS patients found an incidence of roughly 0.02 per 1,000 patients. In addition to persistent severe lymphopenia, older age appeared to be a risk factor.¹² The pattern here, as with fingolimod, centers on sustained low lymphocyte counts. Patients whose blood counts drop and stay low for extended periods seem to be the ones at risk, which is why routine blood monitoring is recommended during treatment with either drug.

These MS therapies collectively illustrate a theme: PML risk from immunomodulatory drugs often mirrors changes that look like premature aging of the immune system. One review found that the lymphocyte shifts induced by high-risk MS therapies resemble the patterns seen in immunosenescence, the natural age-related weakening of immune function.¹³ That observation fits with the broader finding that older patients on these drugs tend to be at higher risk for PML.

Efalizumab, a Cautionary Withdrawal

Efalizumab (Raptiva) was a biologic therapy approved for moderate-to-severe psoriasis. It blocked a surface molecule on T cells needed for their activation and migration. After several confirmed PML cases surfaced, the drug was voluntarily withdrawn from the U.S. market in July 2009.¹⁴ Efalizumab’s withdrawal is sometimes cited as the clearest example of a drug being pulled specifically because of PML risk, and it helped shape the vigilance now applied to newer biologics before and after approval.

Chemotherapy Agents and Purine Analogues

Before the era of monoclonal antibodies, PML in non-HIV patients was most frequently seen in people undergoing intensive chemotherapy for blood cancers. Purine analogues like fludarabine are among the older drugs most associated with PML. Two separate case reports documented PML after fludarabine treatment in patients with chronic lymphocytic leukemia and other low-grade lymphoproliferative diseases, with the authors concluding that fludarabine may increase PML risk in this population.¹⁵¹⁶

A larger study of PML in patients with lymphoproliferative disorders found that purine analogues were the most commonly received therapy among recent PML cases, with age over 55, male sex, and severely low CD4 cell counts identified as risk factors.¹⁷ High-dose chemotherapy followed by stem cell transplantation was the second most frequent treatment in that cohort. The median time from purine analogue or transplant treatment to PML diagnosis was about 11 months.

Brentuximab vedotin, an antibody-drug conjugate approved in 2011 for Hodgkin lymphoma and anaplastic large cell lymphoma, also carries a PML warning. An analysis identified five patients who developed PML after brentuximab treatment, including two who were immunocompetent at the time, challenging the assumption that only patients with already-compromised immune systems are vulnerable.¹⁸

Ibrutinib and Other Targeted Small Molecules

Ibrutinib, a Bruton’s tyrosine kinase inhibitor used for chronic lymphocytic leukemia and certain lymphomas, has been linked to PML in a growing number of reports. An FDA adverse-event review found five PML cases associated with ibrutinib alone and five more with ibrutinib combined with rituximab, with a 90% fatality rate among those cases.¹⁰ Several case reports have reinforced this association.¹⁹²⁰ The complicating factor is that most patients receiving ibrutinib have already undergone multiple lines of therapy, often including rituximab, making it hard to know how much of the risk belongs to ibrutinib versus the accumulated immune damage from prior treatments.

Idelalisib, another kinase inhibitor used in blood cancers, was also flagged in the same FDA review as having elevated PML reporting ratios. Both drugs suppress signaling pathways important for B-cell survival and function, so their association with PML fits the broader pattern of B-cell-targeting therapies and JCV vulnerability.

Transplant Drugs and Mycophenolate

Organ transplant recipients take long-term immunosuppressive regimens to prevent rejection, and PML has been reported in this population. A retrospective study of more than 32,000 kidney transplant recipients found that PML is rare overall, with the incidence density in mycophenolate mofetil users estimated at about 14.4 cases per 100,000 person-years, compared with zero cases among non-users. The difference was not statistically significant, partly because mycophenolate use was so widespread in the cohort that it was hard to find an adequate comparison group.²¹ Transplant-associated PML likely reflects the cumulative immunosuppressive burden rather than any single drug, since recipients typically take combinations of calcineurin inhibitors, antimetabolites, and corticosteroids together.

CAR T-Cell Therapy

Among newer immunotherapies, chimeric antigen receptor (CAR) T-cell therapy has emerged as another setting in which PML can develop. A reported case involved a patient treated with CAR T cells for non-Hodgkin lymphoma who had multiple overlapping risk factors, including a history of rituximab treatment and chronic lymphopenia. The authors argued that CAR T-cell therapy likely contributed significantly, given that the patient’s last rituximab dose was two years prior and the lymphoma had been in remission.²² As CAR T-cell use expands, more data will be needed, but the same fundamental dynamic applies: a therapy that profoundly reshapes the immune system can open a window for JCV.

Why It Is Hard to Pin the Blame on a Single Drug

A persistent challenge in studying drug-associated PML is that many of the patients who develop it were already at elevated risk from their underlying disease, prior treatments, or both. One classification framework proposed three criteria for evaluating a drug’s PML risk: whether the underlying condition itself predisposes to PML even without the drug, how long after starting the drug PML appears, and how frequently PML is observed.²³ By those standards, natalizumab stands out because PML occurs in MS patients who would not normally be at risk for it, and the incidence is measurable enough to study in detail. For drugs like rituximab and ibrutinib, the picture is muddier because the blood cancers they treat already suppress immunity and carry their own PML risk.²⁴

This does not mean those drugs are off the hook. It means that quantifying their individual contribution is harder than it looks in a case report, and clinicians typically weigh PML risk as part of the cumulative immune burden a patient has already accumulated.

When PML Develops and What Comes Next

Diagnosing PML starts with recognizing a set of characteristic MRI findings and confirming JCV DNA in the cerebrospinal fluid. A recent diagnostic study found that quantitative PCR for JCV in cerebrospinal fluid had a positive predictive value of 100% when combined with compatible imaging findings, though nearly half of confirmed cases had viral loads below the lowest quantifiable standard, meaning a negative or equivocal PCR does not rule PML out.²⁵ Characteristic imaging patterns include the “Milky Way sign,” a punctate pattern of white matter involvement, and the “rim-and-core” pattern, among others.

For natalizumab-associated PML, the standard management is to stop the drug and perform plasma exchange (plasmapheresis) to rapidly clear it from the bloodstream. This restores immune cell trafficking to the brain, but it also triggers a dangerous complication: immune reconstitution inflammatory syndrome, or IRIS. In one study of natalizumab-associated PML, every patient developed IRIS.²⁶ IRIS occurs because the returning immune cells mount an aggressive inflammatory response against the JCV-infected tissue, and this inflammation can itself cause significant brain damage.²⁷ Managing IRIS often requires corticosteroids to tamp down the inflammation without completely suppressing the immune response needed to control JCV.

Long-term outcomes for natalizumab-associated PML are sobering. A nationwide Austrian study with follow-up extending to 11 years found that patients’ disability scores roughly doubled on average, from a median of 3.5 before PML to 6.5 at last assessment, and several patients converted to progressive MS within three years of PML.²⁸ Among survivors, functional disability generally stabilized about six months after diagnosis, though recovery was partial at best.²⁹

Checkpoint Inhibitors as a Potential Treatment

One of the more unexpected recent developments involves using immune checkpoint inhibitors, drugs typically used to fight cancer, as a treatment for PML itself. The rationale is that PML patients’ T cells often express high levels of PD-1, a molecule that acts as an “off switch” for immune cells. Blocking PD-1 can reinvigorate these exhausted T cells and help them fight JCV. In a small study of eight patients treated with pembrolizumab, five showed clinical improvement or stabilization along with reduced JCV levels in their cerebrospinal fluid and stronger anti-JCV immune responses. The other three showed no meaningful benefit.³⁰ This approach is still early and unproven at scale. Larger studies are needed to determine who responds and why some patients do not.³¹ There is also an inherent tension in using checkpoint inhibitors for patients whose PML was caused by immunosuppression for an autoimmune disease: reigniting the immune system could simultaneously worsen the autoimmune condition or trigger severe IRIS.