PML and Tysabri: the Risk
In my first posting introducing Tysabri I mentioned its most significant side effect: progressive multifocal leukoencephalopathy (PML.) This is the name of the potentially fatal brain infection caused by the JC virus in patients with compromised immune systems. Unfortunately, we can’t predict who will get PML and who won’t. However there are three clear risk factors that are known to increase the chances of getting PML:
- the length of time one has been on Tysabri
- prior treatment with immunosuppressants
- prior exposure to the JC virus
1. Length of Time on Tysabri
We know that the risk of PML is directly associated with the length of time someone has been on the medication. While the risk does not appear to go up indefinitely, the medicine is quite safe during the first 12-18 months of treatment. For patients who have been on the medicine for less than 2 years, the risk of contracting PML is 0.3:1000. One strategy that some doctors employ is to use Tysabri for 18-24 months, particularly in patients with active disease. After that time, patients can then decide if they feel comfortable continuing on Tysabri or switching to a safer, though likely less effective, agent. For people who have been on the medicine for over 2 years, the overall risk is about 1.5: 1,000. This means that if 2,000 people took the medicine for 2 years, one could expect 3 cases of PML. This is a substantial increase for patients who have been on the medicine for less than 2 years.
The risk does not appear to increase indefinitely, however, and the risk of PML actually seems to go down for people who have been on the medicine for more than 3 years. Based on the available data, the rate is about 1:1000 for people who have been on the medicine for this length of time. It is possible that most patients who are unfortunate enough to be infected with PML are going to contract the disease around the 2-year mark. It is possible that the risk of PML will peak at around 2 years and decline after that. At this time, it is premature based on the available data to say what the risk of PML will be in people who have been on the medicine for longer than 4 years. This information should be available in the upcoming years as more patients stay on Tysabri for this length of time.
2. Prior Treatment with Immunosuppressants
We also know that the risk of getting PML is more common in patients who have been treated previously with immunosuppressive agents. This may be one reason that PML is more common in patients in Europe than the United States, where these medications are more commonly used. Note that the standard medications used in MS (such as Avonex, Betaseron, Copaxone, and Rebif) are NOT immunosuppressants, though many patients mistakenly believe they are. Steroids, at least used in short courses as is the case with most MS patients, are also not considered immunosupressants. Prior use of these medications does not increase the risk of PML.
The immunsupressants that have been associated with PML in patients on Tysabri include:
Although some doctors use these medications, especially cyclophosphamide (cytoxan), I think this is a mistake for most patients as an initial treatment, as it makes Tysabri a much riskier medication if one ever wanted to use it.
3. Exposure to the JC Virus
Finally, patients are at greater risk if they have been exposed to the JC virus in the past. Prior infection by the JC virus is required for the development of PML. Biogen, the company that manufactures Tysabri, is conducting an ongoing study, known as the STRATIFY 2 study. The purpose is to screen patients with MS who are taking or might take Tysabri to determine if they have been previously infected with the JC virus and are at risk of contracting PML. On August 15th 2011, this test was made commercially available through Quest Diagnostics.
Unfortunately, about 60 percent of patients, at least in New York, have been exposed to this virus at some point in their lives, so most patients are in the “high risk” group. In some countries, such as the UK, the rate of exposure to the virus is substantially lower. Thus far, no patients who have tested negative for the JC virus antibody have contracted PML. In contrast, of the 40 patients tested to date who have had PML, all of them have tested positive for the antibody. So this test appears to be a powerful tool for detecting a subset of patients who are at very low risk of contracting PML. No test is perfect, however, and there is a small chance that a patient may test negative for the antibody, yet actually have been infected with the JC virus. Additionally, patients who are negative for the JC virus can always be infected with it, so it is important to repeat the test on a periodic basis (at least annually) in those patients who test negative.
The company is naturally very interested in further determining which of the patients who test positive for JC virus exposure are at risk of actually contracting PML. After all, even in patients who tested positive for the virus, the odds of getting PML are still much less than one percent, assuming they have not been on immunosupressants previously. Hopefully, it won’t be long before we can make more accurate predictions about which patients are at the highest of contracting PML and for whom Tysabri should not be considered.
For patients who test negative for the JC virus, however, the risk of contracting PML is likely exceedingly low (estimated to be less than one in 10,000), and for these patients Tysabri can be taken with minimal risk. I do not hesitate to suggest Tysabri for these patients if there is any indication that the older agents are not going to control their disease or these medicines are intolerable to the patient. In contrast, patients who test positive for the JC virus antibody need to have more active disease to justify the use of Tysabri. It is up to doctors and patients, however, to try to strike the right balance between the risk posed by the medicine and the risk posed by the disease. There is no formula.
New information about this subject is emerging frequently, and I will try to make updates if there are any significant changes.
I am available via e-mail at email@example.com and will try to answer all questions.