Less Frequent Natalizumab Dosing Dramatically Reduces PML Risk

Caroline Helwick

February 03, 2018

SAN DIEGO — Less frequent dosing of natalizumab was associated with a statistically significant, 94% lower risk for progressive multifocal leukoencephalopathy (PML) compared with standard dosing in patients with multiple sclerosis (MS) who test positive for the anti–JC virus antibody, an analysis of a large database has shown.

"Our work is hypothesis-generating, and we plan further analyses, but I think the numbers are very powerful and clinically meaningful," Lana Zhovtis Ryerson, MD, from NYU Langone Health at New York University, told Medscape Medical News. While noting that the analysis did not examine efficacy of the novel schedule, she suggested, "Perhaps less frequent infusions — every 35 to 43 days — is the safer schedule. I think a lot of clinicians will see these data and start dosing natalizumab differently."

PML is a known risk associated with natalizumab treatment, at least at the approved dose of 300 mg intravenously every 4 weeks. The late-breaking research, presented here at the third annual forum of the Americas Committee for Treatment and Research in Multiple Sclerosis (ACTRIMS), aimed to determine whether longer intervals between dosing might ameliorate this risk.

"Extended-interval dosing (EID) is sometimes practiced, especially in patients at high risk of PML, with the aim of reducing PML risk while maintaining efficacy, but prior studies have been inconclusive on EID's impact on PML risk," Dr Ryerson said.

The mandatory US risk evaluation and mitigation strategies program for natalizumab, called TOUCH, offers the largest data source available on PML risk. As of June 1, 2017, the registry contained information on 90,038 patients with MS.

From that registry, this study drew 35,132 patients who had tested positive for the anti–JC virus antibody and who had received natalizumab at dosing intervals of 3 weeks or more and 12 weeks or less. Standard dosing was based on average dosing intervals of 3 or more to less than 5 weeks, while EID was based on average dosing intervals of 5 weeks or more to 12 weeks or less.

"TOUCH involves real-world data where clinicians are not necessarily infusing every 28 days, routinely. Out of this dataset, we are parsing out those patients who have consecutive, prolonged EID," she explained.

Study Methods

Baseline demographics were well balanced across dosing groups, but the EID group had more exposure to natalizumab (most had been receiving the standard dose for 2 years before switching to EID). PML hazards in EID and standard dose cohorts were compared, with adjustment for the covariates of age, sex, prior immunosuppressant use, initiation calendar year, and number of infusions.

The investigators examined the effect of EID in three ways:

  • Average dosing intervals in the last 18 months of recorded infusion history, to test whether dosing history in the last 18 months affects PML risk (primary analysis). EID was defined as 15 or fewer infusions in the last 18 months (540 days).

  • Any consecutive 6-month period of EID dosing in the infusion history, which tested whether any EID period across the dosing history affected PML risk (secondary definition). An EID infusion was an infusion with 10 or fewer doses occurring in the prior 365 days.

  • Average dosing interval over the full infusion history, which tested whether a primary EID dosing history affects PML risk (tertiary definition). EID was defined as 10 or fewer infusions per year based on the total number of infusions divided by total follow-up time.

"Because this is real-world data, and we were initially blind to the PML findings, we didn't know the best way to approach the analysis. We don't know whether it's the entire infusion history that matters, or infusions in the last 18 months, so we created multiple definitions," she explained.

PML Extremely Rare With Extended Interval

In the primary analysis, average dosing interval was 29 days for the standard dose and 36 for EID.  Median exposure was 44 months for the standard dose and 59 for EID. Most patients had more than 2 years of the standard dose before switching to EID.

The development of PML was almost eliminated by dosing natalizumab less frequently (hazard ratio [HR], 0.06; 95% confidence interval [CI], 0.01 - 0.22; P < .0001).

Table. Cumulative Probability of PML in Primary Analysis

Time Point

PML Cases/Patients at Risk (n/n)

Standard Dosing


5 y



7 y




After 7 years, Dr Ryerson added, "the Kaplan-Meier curves keep separating, but the number of patients at risk becomes much lower. Still, PML risk continues to go up," and the benefit of EID continues to be seen, Dr Ryerson said.

By the secondary definition (testing EID any time in the dosing history), a significant difference between the EID and standard dose groups was also observed (HR, 0.012; P < .0001).

Similarly, while the numbers were small for the tertiary definition, a significant effect was also seen there (P = .024).  

"By the tertiary definition, there were no cases at all of PML in the EID group," she noted. Asked whether this suggests EID should be initiated from the start of natalizumab, she told Medscape Medical News the finding is "intriguing" but is based on a small cohort of 812 patients.

As the TOUCH program does not collect effectiveness data, additional studies are needed to establish whether the effectiveness of natalizumab is maintained with EID. The current study also does not capture associations between outcomes and the JC antibody index. The investigators hope to pursue these questions in further analyses.

She also acknowledged there could be selection bias favoring EID in that most of these patients had already been treated for about 2 years with the standard dose and had remained negative for JC virus antibody.  It is possible, therefore, that they were self-selected to be antibody negative regardless of dosing interval.

"Interesting, Persuasive Data"

Jeffrey I. Greenstein, MD, president of the Multiple Sclerosis Research Institute in Philadelphia, Pennsylvania, called the findings "very, very interesting" and "persuasive."

Dr Greenstein's laboratory is exploring the connection between JC virus antibody and risk for PML, primarily with natalizumab but also with other monoclonal antibodies, which he pointed out also pose this risk. He said the current study fits with much of what he has been finding.

"The data are very important," he commented. "I think this is the first time someone has shown that you can reduce the risk of being on natalizumab without really sacrificing therapeutic efficacy, and I think that's a paradigm change in terms of how we should view the drug."

This study was supported by Biogen. Dr Ryerson received personal compensation from Biogen and Teva for speaker and advisory board activities and research support from Biogen. Dr Greenstein has a research grant from Biogen.

Americas Committee for Treatment and Research in Multiple Sclerosis (ACTRIMS). LB250. Presented February 2, 2018.

For more Medscape Neurology news, join us on Facebook and Twitter


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.