Susan Jeffrey

September 25, 2014

BOSTON — A new report describes cases of progressive multifocal leukoencephalopathy (PML) that have been seen to date among patients with multiple sclerosis (MS) treated with fingolimod (Gilenya, Novartis Pharma AG), all but 1 of whom had previously been treated with natalizumab (Tysabri, Biogen Idec), known to carry a risk for PML.

The lingering question is whether there is any risk for PML with the new oral agent, but the authors suggest that the 1 patient not previously treated with natalizumab may have had neuromyelitis optica (NMO) rather than MS, and had several atypical features.

"With more than 100,000 patients exposed, we cannot actually put a risk of PML to fingolimod patients in MS," Norman Putzki, MD, global program medical director for Novartis, Basel, Switzerland, concluded.

Examination of these patients, some of whom were found to have had PML on MRI before beginning fingolimod therapy, suggests that baseline neurologic examination and MRI before initiation of fingolimod may assist, at least retrospectively, in attributing causality to natalizumab vs fingolimod, he pointed out.

"Lastly, vigilance for PML is warranted following natalizumab discontinuation, irrespective of consecutive treatment approaches," Dr. Putzki said.

The findings were presented here in a well-attended talk at MS Boston 2014, the 2014 Joint Americas and European Committees for Treatment and Research in Multiple Sclerosis (ACTRIMS/ECTRIMS) meeting.

Uncertain Risk

Fingolimod, the first of now several oral medications to treat relapsing MS, is an S1PR modulator that reversibly retains naive T cells and central memory T cells in peripheral lymphoid tissues, the authors note. In the clinical trials supporting approval of this drug, no cases of PML were reported during more than 21,000 patient-years of exposure.

The combined clinical trial and postmarketing exposure to fingolimod now exceeds 135,800 patient-years, and of these, 15% to 20% of patients switched to the oral drug from previous treatment with natalizumab, Dr. Putzki noted.

On the other hand, PML is a well-known risk of natalizumab therapy, and risk can persist beyond actual treatment, the authors say. Natalizumab recipients going off treatment should be monitored for at least 6 months for any sign of emerging PML symptoms.

PML risk with natalizumab is increased with 1 or more risk factors, including JC virus (JCV) seropositivity, treatment with natalizumab for longer than 24 months, and previous immunosuppressive therapy. The PML risk for patients who are JCV positive and taking natalizumab beyond 24 months and up to 48 months is estimated at 3 per 1000 for those not treated with prior immunosuppressive therapy and 13 per 1000 for those having had such treatment, he said.

In April 2012, Novartis reported a case of PML in a patient receiving fingolimod, but the patient had previously been treated with natalizumab for 2.5 years and was positive for antibodies to JCV.

However, in August 2013, the US Food and Drug Administration announced it was investigating a case, reported by Novartis the month before, of a patient with PML who was receiving fingolimod and had not previously been treated with natalizumab.

In this latest report, Dr. Putzki and colleagues from Washington University School of Medicine in St. Louis, Missouri, Mayo Clinic in Rochester, Minnesota, and the University of Ottawa and Ottawa Hospital Research Institute, Ontario, Canada, aimed to investigate and describe the characteristics of PML cases using the fingolimod safety database. They also sought to estimate the hypothetical risk of developing PML among fingolimod-treated patients previously exposed to natalizumab.

In all, 11 cases of PML have been reported to date in patients with MS receiving fingolimod who were previously treated with natalizumab. Exposure to natalizumab in these cases ranged from 2.5 to 6.0 years, with a median of 4 years.

Onset of PML in 5 cases, seen on retrospective review of MRI scans, occurred before the first dose of fingolimod. In 4 cases, onset after the start of fingolimod ranged from 3 weeks to 6 months. In the other 2 cases, the time of onset was unknown.

In the 5 patients in whom PML onset was suspected before receipt of fingolimod, the washout period between treatments ranged from 5 weeks to 6 months; however, data were not complete for several of the patients. For example, while outcomes for 4 of the patients were complete recovery in 2, condition improving in 1, and condition unchanged in 1, the outcome is unknown in 1 case, and JCV PCR data are not available for 3 of the 5.

In the 4 patients in whom PML was reported after the fingolimod switch, the washout period was longer, between 1 month and 5 months, and time to onset of PML after the first dose of fingolimod ranged from 3 weeks to 6 months. Outcomes for these 4 patients were called condition improving in 2, condition unchanged in 1, and death from pulmonary embolism in 1 patient.

However, again, JCV polymerase chain reaction results were not available in 2 patients, and no information at all was available on 2 of the patients who were completely lost to follow-up, which Dr. Putzki pointed out is a function of postmarketing reporting. "This is not a perfect system," he noted.

He then focused on the 1 patient with a purported diagnosis of PML without previous exposure to natalizumab. This 46-year-old woman from eastern Europe had a rapidly progressive disease course that was initially assumed to be MS, he noted. She was treated with high cumulative doses of steroids, interferon, and a few weeks of azathioprine, after which she received fingolimod for 7 months. "The patient had quite a serious disability very early in the disease," he explained.

They were able to review her records and found that her clinical and imaging features, as well as the presence of AQP4 antibodies, pointed instead to a diagnosis of NMO spectrum disorder.

The PML diagnosis was based on the presence of JCV in the cerebrospinal fluid, although the number of copies in this case was low, 371 copies/mL found in a test done in July 2013; previous test results had been negative.

"I am not suggesting that this case did not have PML, but if there was PML our experts could not define the onset of PML because, based on the available MRI scan that we have received, there were no typical PML lesions," he said.

The patient died in March 2014; the reported cause of death was progressive neurologic disability, complicated by nosocomial bronchopulmonary infection and cardiorespiratory arrest. No autopsy was performed.

Strengths of the authors' report include the large population exposure, longer treatment duration, and "real world" medical practice, useful for generation of safety signals, he said. Missing, inaccurate, or duplicate data; uncertainty about causality; and issues about under-reporting or stimulated reporting mean that the result is only a reporting rate, not an incidence rate.

Still, given that more than 89,000 patients were being prescribed fingolimod globally at the time of this analysis and 15% of patients taking fingolimod were estimated to have previous natalizumab exposure, Dr. Putzki said that an excess risk during overlapping pharmacodynamics effect when switching from natalizumab to fingolimod "cannot be established."

They estimate a reporting rate of 0.82 (95% confidence interval, 0.41 - 1.4) cases per 1000 patients who discontinued natalizumab, based on the 11 cases.

No "Compelling" Evidence of Risk

Asked for comment on these findings, Robert J. Fox, MD, Mellen Center for Multiple Sclerosis at Cleveland Clinic, Ohio, said 2 major questions must be addressed. One is whether MS therapies that follow the use of natalizumab increase the risk for PML, he said, "and the corollary to that is does there need to be a washout of Tysabri before starting that follow-on therapy, whatever that therapy is."

He pointed out that many of the cases had PML already evidenced on MRI before starting fingolimod, highlighting the importance of ensuring there is no PML at the time of transition off of natalizumab.

"The reason that time of transition is so important is because some of those patients will have a relapse when they transition from Tysabri to another therapy, and it can be tricky to figure out whether one is treating a relapse or treating PML," Dr. Fox told Medscape Medical News. "So it's vitally important to make sure that one gets a transition MRI so you have a stake in the ground to say this is where you are now, at the time of transition."

The other observation is that it did not appear to him, he said, that the washout period helped prevent PML at the time of the switch. "Or to put it another way, it doesn't look like washout had anything to do with developing PML and that these PML cases look like they developed PML while they were still on Tysabri," he said.

He seemed satisfied that the 1 case without previous natalizumab therapy was not likely PML. "When I viewed the MRIs, the brain lesions looked more likely to have arisen from NMO instead of PML," he said. "Aside from the positive JCV test, there was nothing to suggest PML in that patient — neither the clinical history nor the MRI findings were typical of PML. So to me, it doesn't provide any evidence that fingolimod is associated with PML."

However, he did express concern about the 1 patient who ultimately died because there was a long washout of 5 months and 6 months on treatment passed before symptoms developed. "That's the only case that suggests that fingolimod may increase the risk of PML, and we would certainly want to know more details about that case. What were the MRI findings? Was there an MRI at the time of transition? Was there an MRI during treatment with fingolimod?"

JCV polymerase chain reaction was not available in that case, typically considered a requirement for a confirming a case of PML, he noted. "Now the patient has since passed from a pulmonary embolism so we're not going to gain additional information, but understanding that the previously collected data will be very, very important for us to understand this case."

Overall, though, he says he has not seen any "compelling" data of a possible PML risk associated with fingolimod treatment. The number of patient-years of follow-up cited by the company — 135,800 patient-years — is "impressive," he said.

"That's a lot of patient-years, so if there was an association between Gilenya and PML, I think we should have seen more compelling evidence by now," Dr. Fox noted. "That's not to say that we won't want to continue looking and being vigilant, but if there is a signal it's likely going to be very small if it's there at all."

The work was funded by Novartis. Dr. Putzki is an employee of Novartis, maker of fingolimod. Dr. Fox reports he has received personal consulting fees from Novartis, Biogen Idec, GlaxoSmithKline, MedDay, Questcor, Teva, and XenoPort; has served on advisory committees for Biogen Idec and Novartis; and has received research grant funding from Novartis.

MS Boston 2014: The 2014 Joint Americas and European Committees for Treatment and Research in Multiple Sclerosis (ACTRIMS/ECTRIMS) meeting. Abstract FC3.1. Presented September 12, 2014.


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.