Share this article on:

News from the AAN Annual Meeting: New Risk Data on PML Puts Hard Numbers on Antibody Status, Immunosuppressants, and Treatment Duration

Robinson, Richard

doi: 10.1097/


HONOLULU—The risk of developing progressive multifocal leukoencephalopathy (PML) for a multiple sclerosis (MS) patient on natalizumab (Tysabri) is almost 100 times greater if the patient has been taking the drug for more than two years, has a prior history of immunosuppressant use, and tests positive for antibodies to the JC virus, compared to a patient with none of these three risk factors. So conclude investigators of a study by Biogen Idec, the manufacturer of natalizumab, presented here at the 2011 AAN annual meeting.

But, said one expert not involved with the study, there are far more unknowns to be gleaned from future studies. “It is now almost universally accepted that prior exposure to immunosuppressive agents increases the risk of PML, especially if the patient is seropositive, but there are a lot of unknowns,” said Olaf Stuve, MD, associate professor of neurology at University of Texas Southwestern Medical School in Dallas. For example, still to be determined are the effects of dose, duration, and the risk of agents such as intravenous immunoglobulin or plasma exchange, he added. “Those are thing neurologists would really like to know before committing to treating patients long-term with natalizumab.”

“It has to be made clear to the community that the antibody test still lacks validation,” Dr. Stuve said. “A lot of neurologists think that a negative test means the patient is safe taking natalizumab…it should be stated more clearly that this isn‘t necessarily the case.”

Back to Top | Article Outline


As of March 2011, 102 patients worldwide who were taking natalizumab have developed PML, an opportunistic brain infection by the JC virus. The overall incidence of PML is about 1 case in 1000 treated patients. But it has become clear that patients vary in their risk of developing the disease, based on duration of treatment and prior immunosuppressant use, and, presumably, pre-existing infection with the virus.

Quantifying the relative risks required accumulating enough data to overcome the statistical fluctuations inherent in small numbers, said Al Sandrock, MD, PhD, senior vice president of development at Biogen Idec, and lead author of the study. “Initially, it was hard to make conclusions about risk factors when you only have a handful of cases. For instance, most of the early PML cases were men, whereas most MS patients are women. Only after we got more cases did we realize that [the preponderance of men] was a fluke of having too few patients.”

To develop a comprehensive risk profile, Dr. Sandrock drew data from clinical trials, post-marketing surveillance programs, and an independent MS registry in Sweden. Information on prior immunosuppressant use was not available for all patients taking natalizumab, and so data from a global surveillance program was used as a proxy. Average anti-JCV antibody prevalence among all MS patients was determined from samples collected in clinical studies, using a two-step ELISA test developed by Biogen Idec. Antibody status of patients who developed PML was determined in the same way, when samples were available.

The data on PML incidence versus treatment duration indicated that the number of cases per 1000 patients rose steadily through the third year of treatment. It appeared to decline somewhat after that, but, Dr. Sandrock pointed out, the number of patients may be too small to make meaningful conclusions about risk after the three-year time point.



Forty-five percent of natalizumab-treated patients who developed PML had received prior treatment with an immunosuppressive drug, compared to 20 percent of natalizumab-treated patients overall. The agents most commonly used were mitoxantrone, methotrexate, cyclophosphamide, azathioprine, and mycophenolate. Dr. Sandrock noted that virtually all patients had also been previously treated with one of the first-line disease-modifying treatments, either an interferon or glatimer-acetate, which are not considered immunosuppressants, and which are not currently implicated in PML risk.

Why prior use of an immunosuppressant increases PML risk is unknown, but Dr. Sandrock said it is consistent with what is known about the viral life cycle. Humans typically acquire the virus in adolescence, when it takes up residence in the periphery, including in the kidney. Movement to the brain “takes years, if not decades, but immunosuppressant drug use may affect that process,” increasing the likelihood the brain will become infected.

Antibody status is also a risk factor — perhaps the most obvious one, but somewhat complicated nonetheless. “I think it would be very difficult to imagine how one could get PML in the absence of a prior infection with JC virus,” Dr. Sandrock said. “But the antibody test, while good, nonetheless has a false negative rate of 2- to 4-percent.”

So far, all the PML cases have tested positive for the virus. “We haven‘t had any cases where someone was antibody negative and developed PML,” he said, but samples from only 25 of the 102 patients have been available and analyzed.

Because of the potential for a false negative, Dr. Sandrock calculated the risk of PML in a person with a negative test (regardless of other risk factors) as 0.11 per 1000 patients. “The calculation is a hypothetical, based on what would happen if the next case were negative.” Biogen Idec is currently seeking approval to include antibody status as a risk factor on the prescribing information for natalizumab.

For patients who are antibody positive, the risk of PML in natalizumab treatment was 0.35 in those without prior immunosuppressant use and less than two years of treatment, and rose to 2.8 after two years of treatment. For antibody-positive patients who had been previously treated with immunosuppressants, the risk was 1.2 with less than two years of treatment, and rose to 8.1 after two years.

Dr. Sandrock noted the risk assessment is ongoing, with monthly updates posted on a physician-only website — “We will continue to have regular communications with prescribers as long as they think it is useful to have, and regulatory authorities agree.”

Back to Top | Article Outline


“I think the manufacturers have to be applauded for their efforts,” said Dr. Stuve. “To my knowledge, this is the first time that a manufacturer has made this kind of effort to identify an at-risk group for any therapeutic agent in our patient population.”

But, he said, while studies show that patients are willing to take some risk in treatment, “with a range such as this, that should be looked at again. The question has to be asked, are there patients who should not be treated with natalizumab?”

Dr. Stuve said it would be appropriate for leaders in the field to come together to develop recommendations for treatment based on risk. It would have to be “a dynamic process,” he said, with re-evaluation as new data become available. “We are learning as we go along.”

©2011 American Academy of Neurology