Alemtuzumab and PML
“Gavin, you didn’t tell me that alemtuzumab, or cladribine for that matter, could cause progressive multifocal leukoencephalopathy (PML). In fact, you said the opposite, that immune reconstitution therapies (IRTs) would be associated with a low risk of PML. Can you please explain yourself?”
What triggered this question was a recent report of PML in a person with MS (pwMS) treated with alemtuzumab. However, the explanation is in the small print and is not captured in the abstract.
Gerevini et al. Immune profiling of a patient with alemtuzumab-associated progressive multifocal leukoencephalopathy. Mult Scler. 2019 Apr 9:1352458519832259.
A 31-year-old woman had been diagnosed with MS at the age of 23 years when she had manifested mild sensory symptoms and typical MRI lesions. She had been treated with several different treatments over the years:
- Interferon β (IFNβ) 1a, 22mg (2009–2011)
- IFNβ 1a, 44mg (2011–2012)
- Fingolimod (2012– 2013)
- Glatiramer acetate (2013–2014)
- Dimethyl fumarate (2014–2016).
- Alemtuzumab (Course 1, December 2016; Course 2, December 2017)
In February 2018, 14 months after the first cycle and 2 months after the second cycle of alemtuzumab, the patient presented with seizures and was found to have PML. After developing immune reconstitution inflammatory syndrome (IRIS) JCV-DNA in the CSF became negative and her symptoms from PML gradually improved.
John Cunningham, or JC, virus (JCV)
PML is caused by a mutant strain of JCV. JCV belongs to the family of viruses called human polyomaviruses. To get PML you first need to be infected with JCV. Please note we have not identified a specific syndrome due to primary JCV infection. In other words, when you become infected with JCV you won’t be aware of it. As we detect JCV in urine and saliva we think the virus is transmitted via contact with urine and/or saliva from a person who is shedding the virus. Please note not all people who are JCV seropositive (have antibodies in their blood to JCV) shed the virus. Whether JCV non-shedders are still infected with the virus is unknown at present. However, at present, we assume that all MSers who are JCV seropositive are still infected with the virus and are at risk of PML.
To cause PML, the JCV has to mutate and be able to infect cells so-called glial cells in the brain. On average the virus acquires 4–5 mutations in its surface, or coat protein, and several mutations in its regulatory region. These mutations allow the virus to become neurotrophic, i.e. acquire the ability to infect glial cells and cause PML, and importantly to escape recognition by the immune system. The antibodies against JCV that are found in the blood of MSers with PML do not necessarily recognise the JCV mutant in their brains, this tells us that the PML mutant strain of JCV is an escape variant and is not been recognised by the immune system. I suspect that in MSers on natalizumab who develop PML the virus is able to hide from the immune system in the brain; hidden behind a blood-brain-barrier that is sealed off from immune cells that would normally find the virus and destroy it. This is almost certainly why the risk of getting PML on natalizumab is orders of magnitude higher than on other MS drugs in particular IRTs.
It is clear that JCV is more likely to mutate and cause PML if you are immunocompromised. Being immunocompromised increases the chances of the JCV mutating, escaping immune detection and elimination by the immune system. It is clear that the virus takes months to years to become pathogenic, i.e. being able to cause PML, which is why there is carryover risk of PML when you switch from an immunosuppressive drug onto natalizumab or for that matter alemtuzumab. Similarly, there is a similar carryover risk when you switch from natalizumab to other immunosuppressive drugs. It is highly unlikely that the pre-PML mutant strains disappear when you switch from natalizumab to another DMT. This is why it is particularly risky when you switch from natalizumab to an immunosuppressive DMT that has an irreversible, or slowly reversible, mode of action such as that associated with the IRTs. In this alemtuzumab-associated case of PML, I am convinced the prior immunosuppressive treatments are relevant. It is likely that the virus had already started mutating whilst this woman was on fingolimod and then DMF and completed its evolution to a pathogenic PML-inducing strain during the short period of immunosuppression from the two courses of alemtuzumab. It is highly likely that if this woman had not been treated in the past with immunosuppressive therapies she would have developed PML on alemtuzumab.
Lymphopaenia is a risk factor for PML but is a relative risk factor. The lower you lymphocyte count the greater the chance of you getting PML. The risk of PML and other opportunistic infections increase when the total lymphocyte count drops below 800/mm3 (WHO grade 1/2 boundary) and in particular below 500 /mm3 (WHO grade 2/3 boundary). This risk is also associated with how long you remain lymphopaenic. In the case of DMF- and fingolimod-associated PML, most of the patients have persistent and prolonged lymphopaenia. In the case of the IRTs, the lymphopaenia is very short-lived and explains why PML-risk is low with alemtuzumab, cladribine, mitoxantrone and HSCT.
Although this patient was relatively young older age is a risk factor for developing PML. This is particularly noteworthy in the case of DMF-associated PML. In addition, all the patients with ‘de novo PML’, without an identifiable risk factor, I have seen have been old (>60 years). Therefore, age-related immune dysfunction, or immunosenescence, is a risk factor for PML.
Mode of action
The mode of action of DMTs clearly plays a role in causing PML. The more the CNS compartment is immunocompromised the greater the risk of PML. This explains why PML risk is so high with natalizumab a selective adhesion molecule blocker. Natalizumab prevents immune surveillance of the CNS. In terms of the other MS DMTs, it is the DMTs that blunt longterm cytotoxic CD8+ T-cell function that has the highest risk of PML; these include the S1P modulators (fingolimod, siponimod, etc.) and dimethyl fumarate. In contrast, IRTs may have a short-term impact on CD8+ T-cells but once the immune system is reconstituted these anti-viral cells recover and keep PML away. The latter observation explains why the patient above developed IRIS and recovered from her PML.
At present knowing your JCV serostatus at baseline is helpful in predicting PML risk. Being seropositive and having a high anti-JCV antibody index are risk factors for developing PML. On natalizumab a rising anti-JCV antibody index is also an important risk factor, i.e. it indicates that the virus must be replicating actively within the body and stimulating the antibody response (booster effect). Unlike natalizumab, fingolimod and DMF are generalised immunosuppressive drug and hence may affect the antibody response to JCV. Therefore, I would advise against using the JCV antibody index as part of any risk stratification strategy in patients on immunosuppressive therapies. You can use an online PML risk calculator to calculate your PML risk.
It is likely that genetic factors, in particular, HLA and variants in immune response genes, for example, expression of the adhesion molecule L-selectin, are also associated with PML risk.
In summary, (1) previous immunosuppressive treatment (including natalizumab, fingolimod and DMF treatment), (2) duration of immunosuppression, (3) level of lymphopaenia, (4) JCV serostatus, (5) level of JCV antibody index (natalizumab-treatment only), (6) genetic (HLA), (7) immunological factors (L-selectin, lymphocyte subset function), (8) viral factors, (9) mode of action of drug and (10) age play a role in the pathogenesis of PML.
So yes, I said PML risk associated with alemtuzumab and cladribine would be low. I have never said the risk would be zero. This case is complicated by previous immunosuppressive therapies and the PML is unlikely to have been caused by alemtuzumab alone. The fact that this patient recovered after immune system reconstitution supports the premise that PML risk will remain low for the IRTs.