What is PML?

PML stands for progressive multifocal leukoencephalopathy and is a relatively rare, potentially fatal disease of the brain caused by a viral infection. PML is characterised by progressive damage to the brain’s white matter, typically in multiple different areas – hence the descriptor ‘multifocal’. It is triggered by the JC virus (JCV), which causes a persistent infection that is normally kept under control by your immune system. (JC refers to John Cunningham, the patient in whom PML was first described; the tradition of naming viruses after patients has now stopped.) The JC virus is harmless and only causes problems in people who are immunocompromised or have a weakened immune system.

You can get infected with the JC virus at any time. Infections start in childhood and increase gradually; by adulthood, about 50 ̶ 60% of the general population are infected with the virus. JCV causes an asymptomatic primary, or initial, infection; in other words, you have no known ill effects. The virus then resides in the kidney and possibly the bone marrow and lymphoid tissue. Infected people intermittently shed the virus in their urine and saliva, spreading the infection to others; we estimate that about 0.5% of the population gets infected every year.

Your immune system responds to the virus by making antibodies, i.e. you become JCV– seropositive. We detect these antibodies with a simple blood test. We assume that all JCV-seropositive people are still infected with the virus.

How does the JC virus cause PML?

If you are immunocompetent, your immune system keeps the virus in check. If your immune system is compromised, however, JCV replicates and mutates. Some of the mutants acquire the ability to infect glial cells in the brain. Once the mutant JCV infects the glial cells, it hijacks the cells’ machinery and reproduces itself, causing the cells to burst and release thousands of new viruses to infect adjacent cells. When a critical number of glial cells are infected and destroyed, focal neurological symptoms develop. The type of symptoms associated with PML depends on the areas of the brain infected with the virus. PML symptoms are non-specific and can mimic an MS relapse, leading to misdiagnosis initially. Some common PML presenting symptoms include a change in cognition, personality and performance of complex motor tasks (apraxia), accompanied by seizures.

How do you diagnose PML?

PML is diagnosed clinically with the aid of a magnetic resonance imaging (MRI) scan and cerebrospinal fluid to detect the virus using a lab test called JCV-DNA PCR. Occasionally a brain biopsy is required to make the diagnosis, but this is rare nowadays.

How common is PML in people with MS?

PML is quite rare in people with MS and was not reported until the era of immunosuppressive DMTs. (This lack of earlier reporting may be because of misdiagnosis.)

PML is particularly a problem with natalizumab. Natalizumab blocks the trafficking of lymphocytes into the central nervous system (CNS) and thereby prevents the immune system surveying the brain and spinal cord for viruses. PML also occurs with other immunosuppressive therapies, unrelated to previous natalizumab treatment, including fingolimod and dimethyl fumarate.

Carry-over PML

Some people with MS have developed PML while on alemtuzumab, teriflunomide, fingolimod, rituximab or ocrelizumab as a carry-over effect from previous natalizumab therapy. We assume here that the PML was asymptomatic when they stopped natalizumab and only presented after they had switched to another DMT. PML is classified as ‘carry-over’ if it occurs within 12 months of stopping natalizumab. However, I don’t believe the risk from natalizumab ever goes away because PML is a complex disease and takes time to develop. The mutant strain that evolves to potentially cause PML may persist in the body long term and hence increase your risk, or at least lower your threshold, for developing PML in the future if you remain on immunosuppression.

How do I know if I’m at risk of PML?

JCV-seropositive test result

Everyone with MS on natalizumab is at risk of PML. If you are JCV-seronegative your risk of PML is very low, unless you become infected with the virus; each year, about 0.5 ̶ 2% of people with MS who are JCV-seronegative become seropositive. If you are JCV-seropositive your risk increases with duration of natalizumab treatment; it is particularly low if you have been on natalizumab less than 12 ̶ 24 months.

Raised JCV antibodies

The level of antibody against JCV also predicts risk. In people with a raised JCV antibody index, the high or rising level of antibodies indicates ongoing active infection, which boosts the antibody response. This means the virus is active, possibly mutating, so there is a higher risk of developing PML. People with MS who were previously on an immunosuppressive therapy such as mitoxantrone or azathioprine are also at high risk of developing PML. Immunosuppressive therapies presumably allow the virus to escape immune surveillance, to acquire the necessary PML-associated mutations and thus to put you at higher risk.

Not everyone with MS who is JCV-seropositive sheds virus; a subset of individuals may have cleared the virus from the body and hence be at low risk of PML. This may explain why a persistently low index of antibodies to JCV indicates a low risk of PML, i.e. it suggests past infection and no active infection at present.

Immunosuppressive therapy

Immunosuppressive therapies also blunt the immune response to the virus and affect the JCV antibody index. People with MS previously exposed to immunosuppression are still at high risk of developing PML even if they have a low JCV antibody index. This makes the index unreliable in people with MS previously exposed to an immunosuppressive therapy. The following table and graph summarise these risk factors.

Key risk factors for developing PML; risk increases with increasing time on Tysabri. Based on manufacturer’s February 2023 data and modified from Barts-MS PML Risk Guide. Extrapolated extended interval dosing values show ~94% risk reduction compared with standard dosing.
IS, immunosuppressive therapy; JCV, JC virus; PML, progressive multifocal leukoencephalopathy.

Extended interval dosing

Reducing the risk of PML

At present, several factors can help us assess the risk of PML:  JC virus testing (negative and positive), level of JCV antibodies (antibody index), previous exposure to immunosuppression, and treatment duration. We also have frequent MRI monitoring (3 ̶ 4-monthly) to detect PML early and plasma exchange to remove natalizumab as a backup option if a patient develops PML. Another option to reduce the risk of PML may be extended interval dosing (EID).

I have several patients who, despite being JCV-seropositive, insist on staying on natalizumab rather than trying an alternative treatment. A few patients, even after switching to another DMT to prevent getting PML, have opted to go back onto natalizumab. The reason typically relates to the return of MS fatigue, or brain fog, after stopping natalizumab. After restarting natalizumab, patients come back and say, ‘I feel well, my fatigue has gone, and my thinking is clear’. This is why anything that decreases the risk of PML for patients on natalizumab is good.

Rationale for extended interval dosing

The theory behind EID is that some cells are less sensitive than others to the effects of natalizumab; if you delay the next natalizumab infusion by 1 or 2 weeks, the saturation of the surface receptors drops below a threshold and allows those cells to traffic into the CNS. If these cells with lower sensitivity to natalizumab are the antiviral T cells and/or the natural killer cells that fight viruses, this could allow immune surveillance of the CNS to occur and prevent PML from developing. By achieving the correct EID, the saturation of the immune cells that cause MS (possibly the memory B cells) is sufficient not to allow MS to reactivate. Clearly, not all cells are equal when it comes to the effect of natalizumab. Importantly, several other adhesion molecules impact the adhesion (stickiness) of immune cells to the blood vessels in the CNS. A delicate balance between the availability of different accessory adhesion molecules could also make the difference.

When these principles were adopted by several neurologists in the USA, the data emerging from their centres suggested they were correct in hypothesising that the risk of developing PML was reduced when JCV-seropositive people with MS received EID natalizumab.

Comparison of extended and standard interval dosing

Biogen, the manufacturer of Tysabri, sponsored some large studies to explore this theory.^1,2 Using the so-called TOUCH program (Tysabri Outreach: Unified Commitment to Health), which is a mandatory database of all people with MS receiving Tysabri in the USA, statisticians identified more than 35,000 anti-JC virus antibody-positive patients on Tysabri; they compared those on EID with those on standard interval dosing (SID) for PML risk.¹ The TOUCH programme is real-life data, not a clinical trial database, so the periods of EID are variable. To deal with this, the statisticians defined three types of EID with increasing stringency. The remarkable finding was that EID was seen to reduce the risk of PML significantly compared with SID in two of the analyses; in the most stringently defined cohort of EID there were rare cases of PML.

Clinical implications

I have acted on this finding and have offered EID to my patients taking natalizumab who are at risk of PML. It is advisable to transition to 6-weekly EID over several months so as not to precipitate pre-infusion worsening of symptoms. I am now recommending three infusions at 5-weekly intervals before moving to 6-weekly infusions.

The question to consider is whether EID will be associated with some loss of natalizumab effectiveness. A recent study showed no loss of efficacy with EID.² In addition, the study was not clear on finding a personalised dose. Therefore, all patients on EID received it every 6 weeks.

I personally am thrilled by these results. Why? Because anything that derisks PML for people with MS on natalizumab is good, particularly for people with more advanced MS. This is important because natalizumab is effective in more advanced MS, particularly in slowing down or preventing worsening of hand and arm function.

Can you treat PML?

The short answer is no. Some potential treatments for PML have been proposed, but none has been shown to work. In the MS context, you need immune reconstitution to clear the virus from the brain, and herein lies the problem. When you remove natalizumab with either plasma exchange or by waiting for it to wash out spontaneously, your immune cells start re-trafficking into the brain and you develop encephalitis. This is called IRIS (immune reconstitution inflammatory syndrome) and it can be potentially very dangerous. Therefore, in patients with a large PML burden or PML in strategic brain areas such as the brainstem, we try to reduce the damage associated with IRIS by giving steroids. Anecdotal experience suggests steroids work.

Is there another strategy that we can try? The anti-HIV drug, maviroc, blocks a particular chemokine receptor (CCR5) on lymphocytes and may help prevent or dampen down IRIS. T cells, including cytotoxic CD8+ T cells, use the CCR5 receptor to cross the blood ̶ brain barrier. Blocking CCR5 seems to dampen down IRIS, and in two reported cases it appeared to prevent IRIS-related damage. Clearly, maviroc as a monotherapy is not enough to stop the immune system clearing the JC virus from the CNS. To determine whether maraviroc does this more effectively than steroids will require a clinical trial.

The mainstay of treating natalizumab-associated PML is reversal of the natalizumab effect. Plasma exchange can speed this up, i.e. removal of the plasma and hence the circulating natalizumab. When natalizumab levels in the peripheral blood fall sufficiently low, the receptors become active again and immune system re-trafficking occurs, allowing your own T cells to fight the infection. However, recent data suggest that benefits of plasma exchange are marginal and for this reason most centres don’t perform plasma exchange to treat PML.  A problem arises when we can’t reconstitute CNS immunosurveillance. This can happen after immune reconstitution therapy, particularly with alemtuzumab or possibly cladribine, or in people with persistent lymphopaenia. One strategy here is an immune transplant, i.e. giving donor anti-JCV lymphocytes to someone with PML, matched to their own HLA (human leukocyte antigens), to fight the JCV infection. This strategy has helped several people recover from PML who might otherwise have died.

I hope that cases of natalizumab-associated PML become increasingly rare. Now that we have derisking strategies and other safer, highly effective DMTs, should we continue to put people with MS at such a high risk of PML? However, until we get a drug that clears JCV from the body we will never remove the PML risk completely. It is a complication of immunosuppression and therefore it will remain a rare complication of our MS treatments. Further information about natalizumab can be found under DMT: Details

References

1. Ryerson LZ, et al. Risk of natalizumab-associated PML in patients with MS is reduced with extended interval dosing. Neurology 2019;93:e1452 ̶ 62.
2. Foley JF, et al. NOVA study investigators. Comparison of switching to 6-week dosing of natalizumab versus continuing with 4-week dosing in patients with relapsing-remitting multiple sclerosis (NOVA): a randomised, controlled, open-label, phase 3b trial. Lancet Neurol 2022;21:608 ̶ 19.