Research

Original Investigation

Varicella-Zoster Virus Infections in Patients Treated

With Fingolimod
Risk Assessment and Consensus Recommendations
for Management
Ann M. Arvin, MD; Jerry S. Wolinsky, MD; Ludwig Kappos, MD, PhD; Michele I. Morris, MD; Anthony T. Reder, MD; Carlo Tornatore, MD, PhD;
Anne Gershon, MD; Michael Gershon, MD; Myron J. Levin, MD; Mauritz Bezuidenhoudt, MSc; Norman Putzki, MD

Editorial page 10
IMPORTANCE Varicella-zoster virus (VZV) infections increasingly are reported in patients with Author Audio Interview at
multiple sclerosis (MS) and constitute an area of significant concern, especially with the jamaneurology.com
advent of more disease-modifying treatments in MS that affect T-cell–mediated immunity.
Supplemental content at
jamaneurology.com
OBJECTIVE To assess the incidence, risk factors, and clinical characteristics of VZV infections
in fingolimod-treated patients and provide recommendations for prevention and
management.

DESIGN, SETTING, AND PARTICIPANTS Rates of VZV infections in fingolimod clinical trials are
based on pooled data from the completed controlled phases 2 and 3 studies (3916
participants) and ongoing uncontrolled extension phases (3553 participants). Male and
female patients aged 18 through 55 years (18-60 years for the phase 2 studies) and diagnosed
as having relapsing-remitting MS were eligible to participate in these studies. In the
postmarketing setting, reporting rates since 2010 were evaluated.

INTERVENTIONS In clinical trials, patients received fingolimod at a dosage of 0.5 or 1.25 mg/d,
interferon beta-1a, or placebo. In the postmarketing setting, all patients received fingolimod,
0.5 mg/d (total exposure of 54 000 patient-years at the time of analysis).

MAIN OUTCOMES AND MEASURES Calculation of the incidence rate of VZV infection per 1000
patient-years was based on the reporting of adverse events in the trials and the
postmarketing setting.

RESULTS Overall, in clinical trials, VZV rates of infection were low but higher with fingolimod
compared with placebo (11 vs 6 per 1000 patient-years). A similar rate was confirmed in the
ongoing extension studies. Rates reported in the postmarketing settings were comparable (7
per 1000 patient-years) and remained stable over time. Disproportionality in reporting
herpes zoster infection was higher for patients receiving fingolimod compared with those
receiving other disease-modifying treatments (empirical Bayes geometric mean, 2.57 [90%
CI, 2.26-2.91]); the proportion of serious herpes zoster infections was not higher than the
proportion for other treatments (empirical Bayes geometric mean, 1.88 [90% CI, 0.87-3.70]).
Corticosteroid treatment for relapses might be a risk factor for VZV reactivation.

CONCLUSIONS AND RELEVANCE Rates of VZV infections in clinical trials were low with
fingolimod, 0.5 mg/d, but higher than in placebo recipients. Rates reported in the
postmarketing setting are comparable. We found no sign of risk accumulation with longer
exposure. Serious or complicated cases of herpes zoster were uncommon. We recommend
establishing the patient’s VZV immune status before initiating fingolimod therapy and
Author Affiliations: Author
immunization for patients susceptible to primary VZV infection. Routine antiviral prophylaxis
affiliations are listed at the end of this
is not needed, but using concomitant pulsed corticosteroid therapy beyond 3 to 5 days article.
requires an individual risk-benefit assessment. Vigilance to identify early VZV symptoms is Corresponding Author: Norman
important to allow timely antiviral treatment. Putzki, MD, Novartis Pharma AG,
Fabrikstrasse 12-2.03.38, CH-4002
JAMA Neurol. 2015;72(1):31-39. doi:10.1001/jamaneurol.2014.3065 Basel, Switzerland (norman.putzki
Published online November 24, 2014. @novartis.com).

31

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: on 03/21/2018

 Research Original Investigation Varicella-Zoster Virus Infections in MS

V
aricella-zoster virus (VZV) is a neurotropic, epidermo- tion and management of VZV infections, predominantly HZ
tropic, and lymphotropic α-herpesvirus that infects more in adults with MS receiving DMTs, is a topic of clinical
than 90% of people worldwide.1 Primary infection with relevance.
VZV (varicella) is usually acquired in childhood or early adoles-
cence, and infection in adults is rare and often more severe than
in children. Fatal cases with multiple-organ diseases, such as
pneumonia, hepatitis, and coagulopathy, are significantly more
Methods
common in healthy adults than in children. Respiratory muco- Patient Population
sal epithelial cells are presumed to be the first site of infection. We based our analyses on the pooled data from the double-
The T cells become infected with VZV in the tonsils and re- blind phases of 6 completed phase 2 and 3 studies (3916 par-
gional lymph nodes and then transport virions to the skin, where ticipants) of fingolimod in MS and from any ongoing exten-
VZV replication results in the typical vesicular lesions of vari- sion phases of these studies as of August 31, 2012 (3553
cella. Varicella-zoster virus gains access to cranial and dorsal root participants) (clinicaltrials.gov identifiers: NCT00333138,
ganglia and likely to autonomic ganglia by T-cell viremia and N C T 0 0 2 3 5 43 0 , N C T 0 0 2 8 9 9 7 8 , N C T 0 03 4 0 83 4 , a n d
by retrograde transport from skin lesions through the afferent NCT00355134).8-10,16 Institutional review board approval was
fibers of the sensory nervous system. Similar to herpes sim- obtained at all sites, and participants provided written in-
plex virus types 1 and 2, VZV then establishes life-long latency formed consent, as detailed previously.8-10,16 Male and fe-
in the sensory ganglia.1 Antibodies and T cells specific to VZV male patients aged 18 through 55 years (18-60 years for the
are induced during primary infection and typically protect phase 2 studies) and diagnosed with relapsing-remitting MS
against symptomatic reinfections after new exposure in immu- were eligible to participate in these studies. Details of the study
nocompetent and most immunocompromised individuals. Vari- population have been presented elsewhere.8-10,16 Serologic test-
cella-zoster virus antibodies are likely to provide a first line of ing using an enzyme-linked immunoassay for VZV IgG anti-
defense against a new respiratory mucosal inoculation of the bodies was introduced as a protocol amendment during the
virus, whereas VZV-specific T-cell responses are the major host pivotal phase 3 studies. All patients with positive test results
defense against symptomatic reactivation of latent VZV, which after randomization continued the study. For patients with
results in herpes zoster (HZ), commonly termed shingles.1 In in- negative test results after randomization, increased vigilance
dividuals unable to produce VZV IgG antibodies, T-cell immu- for infections was implemented. Patients who were found to
nity is sufficient to protect against a second episode.2 Varicella- be seronegative for antibodies before randomization were not
zoster virus antibodies do not protect against HZ, regardless of included in these studies.
the levels detected in the serum.3 After mid-adulthood, T-cell–
mediated immunity declines with increasing age, and a signifi- Assessments and Analysis
cant decrease in VZV-specific early effectors and cytokine- Calculation of the incidence rate of VZV infections was based
producing CD4 + and CD8 + T cells increases the risk for on the reporting of adverse events in the respective trials. The
reactivation of latent virus and the incidence of HZ.4 incidence is reported per 1000 patient-years, which was de-
The available literature on the incidence of clinically symp- fined as the sum of days all patients in each treatment group
tomatic VZV infections or HZ in patients with multiple sclero- received the study drug divided by 365.25 days. The number
sis (MS) is limited. Herpes zoster has been reported in pa- and percentage of patients with VZV infection was summa-
tients who receive disease-modifying treatments (DMTs) for rized for patients who received concomitant systemic corti-
MS, such as natalizumab, alemtuzumab, and fingolimod.5-10 costeroid dosages of no more than 1000 mg/d for the treat-
Fingolimod, 0.5 mg/d (Gilenya [FTY720]; Novartis Pharma ment of relapses. For the postmarketing incidence of VZV
AG), has shown superior efficacy to the approved first-line treat- infection, we conducted a search of the US Food and Drug Ad-
ment, interferon beta-1a, and placebo in three phase 3 studies8-10 ministration Adverse Event Reporting System database to iden-
and is currently approved as a once-daily oral therapy for re- tify spontaneously reported adverse events consisting of
lapsing forms of MS. The therapeutic effects of fingolimod are HZ.17,18 The search included reported frequency and propor-
mediated via the modulation of sphingosine 1-phosphate (S1P) tion of HZ infection for all drugs, including fingolimod and
receptors by fingolimod phosphate, a phosphorylated active other DMTs for MS, from January 1, 1994, through June 30,
moiety of fingolimod. In vivo, fingolimod-phosphate binds as 2012. Point estimates of the disproportionality in reporting fin-
an analogue of S1P to S1P receptor 1 (S1P1) receptors on lympho- golimod vs other DMTs for MS were assessed using the ad-
cytes to trigger receptor internalization and degradation. This justed value of a ratio of observed to expected reports (em-
process leads to the inhibition of S1P signaling and the selec- pirical Bayes geometric mean [EBGM]). The lower limit of a 90%
tive retention of CCR7+ lymphocytes in the lymphoid organs. CI of the EBGM above the threshold of 2.0 is considered to in-
As a consequence, predominant levels of CD4-naive T cells and dicate a potential signal that deserves further investigation.
central memory T cells (TCM), but not effector memory T cells All these analyses were undertaken using the following pre-
(TEM), are reduced in the circulation.11,12 ferred terms from the Medical Dictionary for Regulatory Ac-
The recently approved MS DMTs all lead to distinct altera- tivities, version 15.119: herpes zoster, herpes zoster dissemi-
tions of immune surveillance.13-15 Infections with VZV may be nated, herpes zoster infection neurological, herpes zoster
seen more frequently with the more efficacious DMTs in clini- multidermatomal, herpes zoster ophthalmic, herpes zoster oti-
cal practice, and increased vigilance is warranted. Preven- cus, and postherpetic neuralgia.

32 JAMA Neurology January 2015 Volume 72, Number 1 jamaneurology.com

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: on 03/21/2018

 Varicella-Zoster Virus Infections in MS Original Investigation Research

Table 1. Incidence Rate of HZ Events per 1000 Patient-years by Severity in Clinical Trials and Extensionsa

No. of Cases Reported (Patient-year Rateb)

Double-blind Controlled Studies
Double-blind Controlled Studiesc and Long-term Extensionsd
Placebo Fingolimod, 1.25 mg/d
Fingolimod, 1.25 mg/d Fingolimod, 0.5 mg/d (1305 Switched to 0.5 mg/d Fingolimod, 0.5 mg/d
HZ Event (1661 Patient-years) (1716 Patient-years) Patient-years) (4402 Patient-years) (4136 Patient-years)
Skin involvement
≤2 Contiguous dermatomes 17 (10.2) 18 (10.4) 7 (5.3) 41 (9.3) 49 (11.8)
Bilateral or >2 contiguous 3 (1.8) 0 0 6 (1.3) 2 (0.4)
dermatomes
Disseminatede 1 (0.6) 0 0 1 (0.2) 0
Ophthalmic 0 1 (0.5) 0 5 (1.1) 4 (0.9)
Oticus 0 0 0 1 (0.2) 0
Encephalitis 0 0 0 0 0

Abbreviation: HZ, herpes zoster. study), 12 months (TRANSFORMS [Trial Assessing Injectable Interferon versus
a
Diagnosis of HZ is based on clinical symptoms and investigating physicians’ FTY720 Oral in Relapsing-Remitting Multiple Sclerosis]9; only fingolimod arms
judgment. Data cutoff was August 31, 2012. included), and 24 months (FREEDOMS8 and FREEDOMS II10) for fingolimod.
d
b
Calculated by the following formula: No. of cases/(patient-year × 1000). Treatment time in the double-blind and extension phases of the clinical trials
Patient-years were defined as the sum of the number of days receiving the was as long as 4 years for fingolimod, 0.5 mg/d (FREEDOMS,8 TRANSFORMS,9
study drug for all patients in each treatment group divided 365.25 days. and FREEDOMS II10 extension studies), and as long as 7 years for fingolimod,
c
1.25 mg/d, switched to 0.5 mg/d (phase 2 extension study).
Treatment time in the double-blind phase of the completed clinical trials is 3
e
months (vaccination study), 6 months (phase 2 study), and 24 months Indicates presence of bilateral lesions or lesions affecting more than 2
(FREEDOMS [FTY720 Research Evaluating Effects of Daily Oral therapy in contiguous dermatomes (cutaneous dissemination) or in other organs than
Multiple Sclerosis]8 and FREEDOMS II10) for placebo; 3 months (vaccination the skin.

comitant administration of corticosteroids as a risk factor for VZV

Results infection or severity in patients receiving fingolimod. No appar-
ent relationship between HZ and systemic corticosteroid use for
Fingolimod Clinical Studies Data fingolimod (all dosages) or placebo was noted. Of 41 cases in the
Although the overall incidence of VZV infection was low, the in- fingolimod arm, only 5 occurred during or within 30 days of con-
cidence was higher in fingolimod recipients (11 per 1000 patient- comitant corticosteroid use (Table 2). However, corticosteroid
years) compared with placebo recipients (6 per 1000 patient- use was limited to a maximum of 5 days per the protocol.
years) in the MS randomized clinical trials. More cases occurred
during the first 6 months of fingolimod treatment. Uncompli- Postmarketing Data
cated cutaneous HZ involving a single dermatome or 2 adja- Reporting Rate of VZV Infections
cent dermatomes was the most common presentation of HZ As of February 28, 2013, a total of 60 873 patients had received
across all treatment groups (Table 1). No increase in incidence fingolimod therapy in the postmarketing setting, which con-
was observed for up to 7 years in the extension studies, and the stituted approximately 54 000 patient-years of drug expo-
profile of patients who developed HZ did not differ from that sure. The reporting rate of HZ infection is 7 per 1000 patient-
of patients without HZ with respect to age, duration of fingoli- years and has remained stable since the launch of fingolimod
mod treatment, other treatments, or previous DMT use. in 2010 (Figure 1). These incidence rates are slightly lower than
Two fatal cases of VZV infection were reported in pa- those observed in the clinical trials.8-10,16 In a separate analysis
tients treated with fingolimod. In a clinical trial, a patient who performed using databases that report spontaneous adverse
had received fingolimod, 1.25 mg/d, for 10 months died of dis- events to identify any disproportionate reporting of an ad-
seminated VZV primary infection (reported in May 2008).9 The verse event,17,18 higher rates of HZ were observed for fingoli-
second fatal case was VZV reactivation (reported in April 2013) mod (EBGM, 2.57 [90% CI, 2.26-2.91]) compared with all other
that occurred in a patient who was seropositive for VZV anti- DMTs for MS (Novartis Pharma AG; annual Periodic Safety Up-
bodies before initiating fingolimod therapy, 0.5 mg/d, for 6 date Report to US Food and Drug Administration; submitted on
months in a postmarketing observational study (the prior DMT October 18, 2013 [data not available in the public domain]).
was natalizumab, and the washout period was approxi-
mately 3 months) (Novartis Pharma AG; annual Periodic Safety Severity
Update Report to US Food and Drug Administration; submit- Serious or complicated cases of HZ were uncommon. The in-
ted on October 18, 2013 [data not available in the public do- volvement of more than 2 dermatomes was reported in 27 of
main]). Both cases received approximately 10 days of con- 339 HZ cases (8.0%). The overall incidence rate of compli-
comitant corticosteroid therapy. Details of the cases are given cated HZ is estimated at 1 per 1000 patient-years.17,18 The dis-
in the eAppendix in the Supplement. proportionate reporting assessment showed that the propor-
After this second case was reported, integrated analysis of tion of serious HZ infections during fingolimod treatment is
the data was conducted to assess the potential impact of con- not higher (lower confidence limit, <2) compared with those

jamaneurology.com JAMA Neurology January 2015 Volume 72, Number 1 33

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: on 03/21/2018

 Research Original Investigation Varicella-Zoster Virus Infections in MS

Table 2. Incidence Rate of HZ Events per 1000 Patient-years by Corticosteroid Use in Controlled Clinical Trialsa

No. of Patients With Event/No. of Patients Treated With Corticosteroids (%)

Fingolimod, 1.25 mg/d Fingolimod, 0.5 mg/d Placebo
VZV-Related Adverse Event (n = 1313) (n = 1212) (n = 866)
Yes
Within 30 d of the last dose of corticosteroidb 3/445 (0.7) 2/387 (0.5) 0
Outside 30-d corticosteroid window 10/445 (2.2) 9/387 (2.3) 4/460 (0.9)
No 432/445 (97.1) 376/387 (97.2) 456/460 (99.1)
b
Abbreviations: HZ, herpes zoster; VZV, varicella-zoster virus. Includes 500 to 1000 mg of intravenous methylprednisolone daily for 3 to 5
a
Treatment time in the double-blind phase of clinical trials was 6 months days with or without a subsequent tapering dosage of oral corticosteroids
(phase 2 study) and 24 months (FREEDOMS8 and FREEDOMS II10) for placebo (most often prednisone) for 1 to 3 weeks. Specific decisions as to dosage,
and 12 months (TRANSFORMS9; only fingolimod arms included) and 24 duration of treatment, and whether a taper is used are based on the type of
months (FREEDOMS8 and FREEDOMS II10) for fingolimod. relapse and the clinicians’ judgment.

subjects.20-23 In an age-adjusted retrospective analysis of claims

Figure 1. Cumulative Reporting of Varicella-Zoster Virus (VZV) Infections
in the Postmarketing Setting databases, HZ incidence was approximately twice as high in
patients with MS as in the general population (6 per 1000 pa-
10 tient-years) (Novartis Pharma AG; annual Periodic Safety Up-
date Report to US Food and Drug Administration; submitted
on October 18, 2013 [data not available in the public do-
8
main]). The observed rate of HZ was higher in patients receiv-
1000 Patient-years of Exposure
Observed Reporting Rate per

ing fingolimod compared with those receiving placebo in the

6
clinical trials. The rates were similar and remained stable for
patients continuing to receive fingolimod during the long-
term extensions for up to 7 years. In the postmarketing set-
4 ting, reporting rates were lower than those of clinical trials, and
no increase in the cumulative reporting rate since 2010 was ob-
served (exposure, 54 000 patient-years as of February 28, 2013).
2
However, the postmarketing reporting of spontaneous ad-
VZV only verse events has limitations, such as underreporting, lack of
VZV + herpes unspecified
0 exposure data, validation of clinical details, and potential bi-
September 1, 2011, to March 1, 2012, to September 1, 2012, to
ases (eg, stimulated reporting).24 Effective risk mitigation, such
February 29, 2012 August 31, 2012 February 28, 2013 as VZV immunization recommended by the fingolimod label,
may affect the actual risk. Comparison of the fingolimod ex-
Herpes unspecified indicates VZV infection with further information regarding perience with that of other DMTs is difficult owing to study
the infection type and/or severity not specified in the spontaneous report. Error
differences in patient characteristics, reporting of HZ occur-
bars indicate 90% CIs.
rence, and concomitant medications (Figure 2).6,7,25-29 Based
on reports of adverse events, HZ was documented more of-
for other MS DMTs (EBGM, 1.88 [90% CI, 0.87-3.70]) (Novar- ten with fingolimod than with other DMTs, such as interferon
tis Pharma AG; annual Periodic Safety Update Report to US Food beta, but the proportion of complicated HZ cases was not higher
and Drug Administration; submitted on October 18, 2013 [data (reporting rate, 1 per 1000 patient-years in the postmarketing
not available in the public domain]). Noncutaneous manifes- setting). We acknowledge that the analysis of HZ severity is
tations were uncommon. The reporting rate of visceral dis- based on limited case numbers and the reported 90% CI is wide
semination was less than 0.1 per 1000 patient-years. Only a few (0.80-3.70). Disproportionality analysis methods other than
cases of recurrent nonserious dermatomal HZ cases have been EBGM are available but were not applied for this analysis.30
reported to date (Novartis Pharma AG; annual Periodic Safety These reporting rates may be influenced by increasing aware-
Update Report to US Food and Drug Administration; submit- ness about VZV infections over time and by recommenda-
ted on October 18, 2013 [data not available in the public do- tions in the product information and educational materials (for
main]). Gilenya) and may be confounded by differences in MS dis-
ease characteristics and previous DMT exposure.
To control VZV reactivation, T-cell immunity is critical.
Central memory T cells are intermediate memory T cells that
Discussion do not yet display effector function. However, TCMs are the
Although it is unclear whether patients with MS have in- basis for long-term immunity and form the reactive memory,
creased susceptibility to HZ, several population-based epide- rapidly proliferating and giving rise to TEMs on strong anti-
miological and case-control studies suggest that they might gen challenge. In the case of VZV infections, fingolimod may
be at greater risk for HZ compared with healthy control reduce the recirculation of specific CD4+ TCMs by retaining

34 JAMA Neurology January 2015 Volume 72, Number 1 jamaneurology.com

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: on 03/21/2018

 Varicella-Zoster Virus Infections in MS Original Investigation Research

Figure 2. Incidence of Varicella-Zoster Virus (VZV) Infections in Fingolimod Trials and Comparison With Other
Treatments and Conditions

Fingolimod Clinical Trials and Extensions

Interferon-beta
Placebo
Controlled Trials Fingolimod, 0.5 mg/d
Fingolimod, 1.25 mg/d
Controlled Trials +
Extensions

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75
Incidence Rate per 1000 Patient-years
Natalizumab Placebo-Controlled Studies25 of MS

Natalizumab

Placebo

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75
Incidence Rate per 1000 Patient-years

DEFINE,26 CONFIRM,27 and Phase 228

Dimethyl Fumarate,
240 mg TID

Dimethyl Fumarate, Vertical dashed lines at 3.7 and 5.6

240 mg BID per 1000 patient-years are
background rates in the general
Placebo
population and patients with multiple
sclerosis (MS) that one would expect
in a population of the same age
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 distribution as the patients treated in
Incidence Rate per 1000 Patient-years fingolimod clinical trials. Because the
incidence of VZV varies with age, the
REFLEX29
background rates were adjusted to
Rituximab the age of the patient population
enrolled in fingolimod clinical trials.
The shaded area throughout the
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75
graph represents the incidence rate in
controlled clinical trials for
Incidence Rate per 1000 Patient-years
fingolimod, 0.5 mg/d.6-8,10,16,25-29
Error bars indicate 90% CIs.
CARE-MS I7 and CARE-MS II6 BID indicates twice daily; CARE,
Interferon-beta CARE-MS I Comparison of Alemtuzumab and
CARE-MS II Rebif Efficacy in Multiple Sclerosis;
CARE-MS I CONFIRM, Comparator and an Oral
Alemtuzumab,
12 mg/d CARE-MS II Fumarate in RRMS; DEFINE,
Determination of the Efficacy and
Alemtuzumab, CARE-MS II
Safety of Oral Fumarate in
24 mg/d
Relapsing-Remitting Multiple
Sclerosis (RRMS); RA, rheumatic
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 arthritis; REFLEX, Randomized
Incidence Rate per 1000 Patient-years Evaluation of Long-term Efficacy of
Rituximab in RA; TID, thrice daily.

them in the lymph nodes, thereby preventing their wider dis- Studies showing intact T-cell immune responses to non-
tribution to the sites of VZV replication, such as the skin (and specific and specific stimuli during fingolimod treatment sug-
possibly the ganglia). As a consequence, the generation of gest that the pharmacodynamic effect of fingolimod leading
TEMs from TCMs would be restricted to tissues that harbor to the ubiquitous reduction of circulating TCMs may not be re-
sufficient numbers of TCMs, which may reduce VZV-specific sponsible for increased HZ incidence.12 However, informa-
immune control and may be especially problematic as a tion about the effects on VZV-specific T-cell responses is lim-
result of confounding factors, such as immune suppression ited at present.31 Varicella-zoster virus–specific T cells are
by corticosteroid use. In contrast, antiviral CD8 T-cell immu- normally present at low frequencies, and the occurrence of HZ
nity would be largely preserved during treatment with fingo- infection in individuals without underlying disease is associ-
limod because CD8 T-cell immunity predominantly results ated with a rapid and substantial increase in these cells.32 Vari-
from CCR7-linked TEMs that are not retained in the lymph cella-zoster virus–specific T cells have multiple functions, in-
nodes by the drug.11,12 cluding production of key cytokines that orchestrate local

jamaneurology.com JAMA Neurology January 2015 Volume 72, Number 1 35

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: on 03/21/2018

 Research Original Investigation Varicella-Zoster Virus Infections in MS

fluenza vaccine and the induction of circulating influenza-

Box 1. Recommendations to Mitigate Risk for VZV Infection specific T cells were comparable. Despite CD4+ and CD8+ T-cell
During Fingolimod Treatment lymphopenia, the recently activated T cells may override the
dependency on S1P1 signaling for egress from the lymph nodes
Varicella (Chickenpox) Vaccination: Prevention of Primary Varicella by down-modulation of CCR7.34 Thus, the functionality of T
Evidence of immunity to varicella should be sought before initiating
cells appears to be intact during fingolimod therapy. A recent
MS therapy.
study31 assessed the immune responses to VZV in patients
Patients with no varicella immunity should be vaccinated (if they have
treated with fingolimod compared with healthy controls, un-
no contraindication) with 2 doses of Varivax/Varilrix separated by 4
treated patients with MS, and patients treated with other DMTs
weeks.a
(glatiramer acetate or interferons). The study reported that fin-
In case of treatment with high-dose corticosteroids, vaccination
golimod treatment of patients with MS lowered VZV-specific
should be administered at least 30 days after the last corticosteroid
dose. immunity and suggested that subclinical VZV reactivation,
demonstrated by polymerase chain reaction detection of VZV
Live-attenuated vaccines are contraindicated during fingolimod
therapy. Varicella vaccination should not be given during fingolimod DNA in the saliva, was higher among patients treated with fin-
treatment; initiation of treatment with fingolimod should be golimod compared with untreated patients with MS and
postponed for 1 month after the second dose of vaccine. healthy controls. However, patient groups were combined in
Corticosteroid Treatment
the statistical analysis, and the numbers were small. Varicella-
Use of corticosteroids may increase the risk for VZV reactivation. zoster virus DNA was not detected in blood, and none of the
fingolimod-treated patients developed clinically apparent HZ.31
When possible, corticosteroid treatment for relapses should be
limited (3-5 days of intravenous methylprednisolone without Compared with CD4+ TEMs, CD8+ TEM cells are less af-
tapering) in patients receiving fingolimod. fected by fingolimod treatment and might help to control VZV
Corticosteroid treatment with a higher dosage or for a longer reactivation despite a drop in CD4 cell levels. Other immune
duration should be evaluated on an individual basis. responses, such as production of interferon-γ by CD3+ T cells
Antiviral prophylaxis is not recommended routinely but can and by tissue-resident T cells, may provide antiviral control that
be considered in patients requiring prolonged or repeated is unaffected by fingolimod.11 In some cases, a subclinical re-
high-dose corticosteroid treatment. activation may trigger the recruitment of lymphocytes from
Vigilance Toward Early Recognition and Treatment
naive and TCM pools in lymph nodes to support the TEM re-
of VZV Infections sponse and prevent clinically apparent VZV infection.
In patients who have evidence of dissemination, including
symptoms of severe abdominal pain, MS therapy should be Recommendations for Management and Risk Mitigation
discontinued and intravenous antiviral therapy as per the current Three areas of risk minimization for VZV infections in pa-
CDC-ACIP guidelines35 should be initiated. tients with MS have been identified. These may also apply to
Antiviral treatment in case of suspected VZV infection should be patients treated with DMTs other than fingolimod (Box 1).
implemented as per CDC-ACIP guidelines.
Patients receiving immunomodulatory therapies should be Prevention of Primary VZV Infection
educated to identify signs and symptoms of primary varicella and Any adult who has not had varicella should be vaccinated (in
recurrent HZ infections with VZV and report them promptly.
the absence of contraindications) with one of the licensed live
Prolonged antiviral treatment may be required for attenuated varicella vaccines because primary VZV infection
immunocompromised patients.
is significantly more severe and occasionally fatal in adults. The
Abbreviations: CDC-ACIP, Centers for Disease Control and Prevention– VZV immune status of adults should be determined at the time
Advisory Committee on Immunization Practices; HZ, herpes zoster; MS,
of MS diagnosis 35(Box 2). However, commercially available as-
multiple sclerosis; VZV, varicella-zoster virus.
says have limited sensitivity; 10% to 15% false-negative and 8%
a
Zostavax (Merck & Co, Inc) is a live-attenuated vaccine approved for the to 11% false-positive results often occur.36,37 Previous VZV in-
prevention of HZ in otherwise healthy elderly people. It contains a 14-fold
increased viral concentration compared with Varivax (Merck & Co, Inc).
fection can also be confirmed by the assessment of the pa-
Safety in the MS population is not yet established. tient’s history of varicella diagnosis by a health care profes-
sional (Box 2). 35 If a patient with a history of household
exposure did not develop varicella, preexisting immunity may
cell-mediated control of infection and cytotoxic effects against be the reason. Few individuals with MS will have been given
infected cells.32 In a 4-week study in healthy volunteers, fin- varicella vaccine because it was only licensed in 1995 or later.
golimod at the 0.5 mg/d dosage had a mild to moderate inhibi- However, if the patient’s medical record shows that he or she
tory effect on T-cell–dependent and T-cell–independent im- received 2 doses of varicella vaccine, serologic testing is not
mune responses. 33 However, the immunogenicity of the indicated and additional doses of vaccine are not needed.
neoantigen keyhole limpet hemocyanin was similar to the ef- If the clinical history and serologic testing suggest that the
fect observed with placebo. Recall antigen responses, as as- patient is susceptible to primary VZV infection, varicella vac-
sessed by anti–tetanus toxoid immunogenicity and delayed- cination should be given as listed in Box 1. Vaccination of in-
type hypersensitivity, were not affected.33 In a smaller study dividuals who may be in the false-negative serologic group is
that included fingolimod-treated patients with MS and un- not detrimental. The vaccine should not be given while the pa-
treated healthy controls, humoral immune responses to in- tient is receiving antiviral drugs active against VZV to avoid

36 JAMA Neurology January 2015 Volume 72, Number 1 jamaneurology.com

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: on 03/21/2018

 Varicella-Zoster Virus Infections in MS Original Investigation Research

blocking efficacy. If a vaccine recipient develops a rash pre-

sumed to be caused by the vaccine virus, appropriate antivi- Box 2. Evidence of Immunity to Varicella (CDC-ACIP
ral therapy with acyclovir or related drugs should be initiated Recommendations)35
promptly.38
Assessing the level of VZV T-cell–mediated immunity in Criteria for Determining Persons Who Can Be Considered Immune
to Varicella
patients with MS is not possible in the clinical setting. Repeti-
Documentation of age-appropriate vaccination with varicella
tive attempts to quantitate circulating VZV IgG antibodies have
vaccine
no predictive value for assessing the risk for HZ. Consistent with
Laboratory evidence of immunitya or laboratory confirmation of the
the epidemiological data from the general population, more
disease
than 95% of the adult MS population undergoing testing in the
Birth in the United States before 1980b
fingolimod clinical trials was found to be seropositive for VZV
Diagnosis or verification of history of varicella disease by a health care
antibodies. Based on limited testing results, fingolimod had
provider
no apparent effect on the VZV antibody status of participants
Diagnosis or verification of history of HZ by a health care provider
in the clinical trials.
Abbreviations: CDC-ACIP, Centers for Disease Control and
Prevention–Advisory Committee on Immunization Practices; HZ, herpes
Prevention of HZ
zoster.
The risk for HZ infection is reduced in individuals without un-
a
derlying diseases when they are given a single-dose HZ vac- Commercial assays may lack sensitivity for detecting vaccine-induced
immunity.
cine (Zostavax; Merck & Co, Inc).39 However, the efficacy and
b
safety of this live-attenuated vaccine have not been estab- For health care personnel, pregnant women, and immunocompromised
persons, birth before 1980 should not be considered evidence of immunity.
lished in patients with MS.
Concomitant treatment with corticosteroids may in-
crease the risk for VZV reactivation. Duration of corticoste-
roid treatment for relapses in fingolimod clinical trials was life of 9 days,40 continuing its use in patients presenting with
limited (3-5 days of treatment with intravenous methyl- uncomplicated HZ is reasonable because treatment interrup-
prednisolone without tapering). Decisions regarding dosage tion is unlikely to affect the recovery from HZ in response to
and duration of corticosteroid treatment in individual pa- antiviral therapy.
tients require the physician’s judgment. Current data do not
support the general use of antiviral prophylaxis in patients re-
ceiving fingolimod given the low incidence of HZ infection. Conclusions
However, antiviral prophylaxis may be considered when
prolonged or repeated high-dose corticosteroid use is deemed Analysis of all available data has not identified obvious risk fac-
necessary. tors for VZV infections, such as duration of treatment, age, pre-
vious treatment, or concomitant corticosteroid use, in pa-
Vigilance Toward Early Symptom Recognition and tients treated with fingolimod. That corticosteroids suppress
Treatment of HZ the immune system and extended corticosteroid treatment is
Patients receiving fingolimod should be educated about early expected to be associated with additional risks are well
symptoms and signs of varicella or HZ and the need to report established. Therefore, the longer-term use of concomitant cor-
them promptly, as well as the possible increased risk when con- ticosteroids (beyond 3-5 days of high-dose intravenous
comitant corticosteroids are prescribed. Antiviral treatment for methylprednisolone for MS relapses) requires a careful risk-
suspected VZV infection should be implemented as per the benefit assessment. Proper patient education along with imple-
guidelines for such treatment in immunocompromised mentation of the recommendations and risk mitigation strat-
patients.38 The route of administration and the dose of acy- egies (Boxes 1 and 2) provide a systematic approach
clovir are determined by the extent of the immunosuppres- for management of VZV infections in patients receiving
sion and severity of infection. Because fingolimod has a half- fingolimod.

ARTICLE INFORMATION Medical Center, Chicago, Illinois (Reder); take responsibility for the integrity of the data and
Accepted for Publication: September 1, 2014. Department of Neurology, MedStar Georgetown the accuracy of the data analysis.
University Hospital, Washington, DC (Tornatore); Study concept and design: Arvin, Wolinsky, Kappos,
Published Online: November 24, 2014. Department of Pediatrics, Columbia University, Tornatore, M. Gershon, Levin, Putzki.
doi:10.1001/jamaneurol.2014.3065. New York, New York (A. Gershon); Department of Acquisition, analysis, or interpretation of data:
Author Affiliations: Department of Pediatrics, Pathology and Cell Biology, Columbia University, Arvin, Wolinsky, Kappos, Morris, Reder, Tornatore,
Stanford University School of Medicine, Stanford, New York, New York (Michael Gershon); A. Gershon, Levin, Bezuidenhoudt, Putzki.
California (Arvin); Department of Neurology, The Department of Pediatrics, University of Colorado Drafting of the manuscript: Arvin, Wolinsky,
University of Texas Health Science Center at Anschutz Medical Campus, Aurora (Levin); Novartis Kappos, Reder, Tornatore, Levin, Bezuidenhoudt,
Houston (Wolinsky); Department of Neurology, Pharma AG, Basel, Switzerland (Bezuidenhoudt, Putzki.
University Hospital Basel, Basel, Switzerland Putzki). Critical revision of the manuscript for important
(Kappos); Department of Infectious Diseases, Author Contributions: Mr Bezuidenhoudt and Dr intellectual content: Arvin, Wolinsky, Kappos,
University of Miami, Miami, Florida (Morris); Putzki had full access to all the data in the study and Morris, Reder, Tornatore, A. Gershon, M. Gershon,
Department of Neurology, University of Chicago Levin, Putzki.

jamaneurology.com JAMA Neurology January 2015 Volume 72, Number 1 37

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: on 03/21/2018

 Research Original Investigation Varicella-Zoster Virus Infections in MS

Statistical analysis: Reder, Bezuidenhoudt, Putzki. and editing the paper for submission under author 15. Bielekova B, Becker BL. Monoclonal antibodies
Study supervision: Kappos, Putzki. guidance. No financial compensation was given for in MS: mechanisms of action. Neurology. 2010;74
Conflict of Interest Disclosures: Dr Arvin received this support. (suppl 1):S31-S40.
a consulting fee from Novartis for participation in a 16. Kappos L, Antel J, Comi G, et al; FTY720 D2201
workshop on herpes zoster and fingolimod and for REFERENCES Study Group. Oral fingolimod (FTY720) for
her work in preparing this manuscript. She 1. Arvin AM, Gilden D. Varicella-zoster virus. In: relapsing multiple sclerosis. N Engl J Med. 2006;
conducted this research as part of a personal Howley PM, Knipe DM, eds. Fields Virology. Vol 2. 7th 355(11):1124-1140.
outside consulting arrangement with Novartis; the ed. Philadelphia, PA: Lippincott Williams & Wilkins 17. US Food and Drug Administration. FDA Adverse
research and research results are not in any way Press; 2013:2015-2057. Event Reporting System (FAERS): latest quarterly
associated with Stanford University. Dr Wolinsky 2. Good RA. Recent studies on the data files. http://www.fda.gov/Drugs
has received fees from Novartis as a consultant and immunodeficiencies of man. Am J Pathol. 1972;69 /GuidanceComplianceRegulatoryInformation
steering committee member for their drug (3):484-490. /Surveillance/AdverseDrugEffects/ucm082193
development programs in multiple sclerosis (MS). .htm. Accessed February 28, 2013.
Dr Kappos participated during the last 24 months as 3. Weinberg A, Zhang JH, Oxman MN, et al; US
principal investigator, member, or chair of the Department of Veterans Affairs (VA) Cooperative 18. WHO Collaborating Centre for International
planning and steering committees or the advisory Studies Program Shingles Prevention Study Drug Monitoring. Uppsala Monitoring Centre: UMC
boards in corporate-sponsored clinical trials in MS Investigators. Varicella-zoster virus–specific WHO databases. http://www.who-umc.org. Accessed
and other neurological diseases sponsored by immune responses to herpes zoster in elderly February 28, 2013.
Actelion, Addex, Allozyne, Bayer HealthCare participants in a trial of a clinically effective zoster 19. Harmonisation for Better Health. Introductory
Pharmaceuticals, Bayer Schering Pharma, Biogen vaccine. J Infect Dis. 2009;200(7):1068-1077. Guide MedDRA, version 15.1. September 2012.
Idec, CLC Behring, GeNeuro SA, Genzyme, 4. Berger R, Florent G, Just M. Decrease of the http://www.meddra.org/sites/default/files
GlaxoSmithKline, Lilly, Merck Serono, Mitsubishi lymphoproliferative response to varicella-zoster /guidance/file/intguide_15_1_English_0.pdf.
Pharma, Novartis, Octopharma, Ono Pharma, virus antigen in the aged. Infect Immun. 1981;32(1): Accessed February 28, 2013
Praxicon, Roche, sanofi-aventis, Santhera, Siemens, 24-27. 20. Lenman JA, Peters TJ. Herpes zoster and
Teva, and XenoPort and has lectured at medical 5. Fine AJ, Sorbello A, Kortepeter C, Scarazzini L. multiple sclerosis. BMJ. 1969;2(5651):218-220.
conferences or in public on various aspects of the Central nervous system herpes simplex and
diagnosis and management of MS, often sponsored 21. Ragozzino MW, Kurland LT. Epidemiologic
varicella zoster virus infections in investigation of the association between herpes
by nonrestricted educational grants to the natalizumab-treated patients. Clin Infect Dis. 2013;
institution from one or another of the above listed zoster and multiple sclerosis. Neurology. 1983;33
57(6):849-852. (5):648-649.
companies. Honoraria and other payments for all
these activities have been used exclusively for 6. Coles AJ, Twyman CL, Arnold DL, et al; CARE-MS 22. Ross RT. The varicella-zoster virus and multiple
funding of research in his department. Research II Investigators. Alemtuzumab for patients with sclerosis. J Clin Epidemiol. 1998;51(7):533-535.
and the clinical operations (nursing and patient care relapsing multiple sclerosis after disease-modifying
therapy: a randomised controlled phase 3 trial. 23. Ross RT, Cheang M, Landry G, Klassen L,
services) of the MS Center in Basel have been Doerksen K. Herpes zoster and multiple sclerosis.
supported by nonrestricted grants from one or Lancet. 2012;380(9856):1829-1839.
Can J Neurol Sci. 1999;26(1):29-32.
more of these companies and by grants from the 7. Cohen JA, Coles AJ, Arnold DL, et al; CARE-MS I
European Union, the Swiss MS Society, the Swiss Investigators. Alemtuzumab versus interferon beta 24. Berlin JA, Glasser SC, Ellenberg SS. Adverse
National Research Foundation, and the Gianni 1a as first-line treatment for patients with event detection in drug development:
Rubatto, Novartis, and Roche Research relapsing-remitting multiple sclerosis: a randomised recommendations and obligations beyond phase 3.
Foundations. Dr Morris received a consulting fee controlled phase 3 trial. Lancet. 2012;380(9856): Am J Public Health. 2008;98(8):1366-1371.
from Novartis for the workshop. Dr Reder received 1819-1828. 25. Peripheral and Central Nervous System Drugs
a grant for translational research consultation and is 8. Kappos L, Radue EW, O’Connor P, et al; Advisory Committee. Natalizumab briefing book
a trialist for Novartis and most other companies and FREEDOMS Study Group. A placebo-controlled trial from Biogen to the FDA for the 2006 advisory
agencies trying to cure MS. Dr Tornatore received a of oral fingolimod in relapsing multiple sclerosis. committee. 2006. http://www.fda.gov/ohrms
consulting fee from Novartis. Dr A. Gershon has N Engl J Med. 2010;362(5):387-401. /dockets/ac/06/briefing/2006-4208B1_02
served as a paid Novartis consultant. Dr M. Gershon _01Biogenbriefingmaterial.pdf. Accessed February
has received research grants from the National 9. Cohen JA, Barkhof F, Comi G, et al; 28, 2014.
Institutes of Health. Dr Levin received a consulting TRANSFORMS Study Group. Oral fingolimod or
intramuscular interferon for relapsing multiple 26. Gold R, Kappos L, Arnold DL, et al; DEFINE
fee from Novartis for issues that preceded Study Investigators. Placebo-controlled phase 3
developing this manuscript, receives funds as a sclerosis. N Engl J Med. 2010;362(5):402-415.
study of oral BG-12 for relapsing multiple sclerosis.
consultant for Merck Research Laboratories, shares 10. Calabresi P, Radue E-W, Goodin D, et al. Safety N Engl J Med. 2012;367(12):1098-1107.
in a patent for Zostavax, and serves on an and Efficacy of Fingolimod in Patients With
adjudication committee for GlaxoSmithKline for an Relapsing-Remitting Multiple Sclerosis (FREEDOMS 27. Fox RJ, Miller DH, Phillips JT, et al; CONFIRM
investigational herpes zoster vaccine. Mr II): a double-blind, randomised, placebo-controlled Study Investigators. Placebo-controlled phase 3
Bezuidenhoudt and Dr Putzki are employees of phase 3 study. Lancet Neurol. 2014;13(6):545-556. study of oral BG-12 or glatiramer in multiple
Novartis Pharma AG. sclerosis. N Engl J Med. 2012;367(12):1087-1097.
11. Chun J, Hartung H-P. Mechanism of action of
Funding/Support: This study was supported by oral fingolimod (FTY720) in multiple sclerosis. Clin 28. Kappos L, Gold R, Miller DH, et al; BG-12 Phase
Novartis Pharma AG, who convened the Advisory Neuropharmacol. 2010;33(2):91-101. IIb Study Investigators. Efficacy and safety of oral
Board and provided manuscript writing support. fumarate in patients with relapsing-remitting
12. Sallusto F, Geginat J, Lanzavecchia A. Central multiple sclerosis: a multicentre, randomised,
Role of the Funder/Sponsor: Novartis Pharma AG memory and effector memory T cell subsets: double-blind, placebo-controlled phase IIb study.
contributed to the design and conduct of these function, generation, and maintenance. Annu Rev Lancet. 2008;372(9648):1463-1472.
studies and to the collection, management, Immunol. 2004;22:745-763.
analysis, and interpretation of the data. Employees 29. Cohen SB, Emery P, Greenwald MW, et al;
13. di Nuzzo L, Orlando R, Nasca C, Nicoletti F. REFLEX Trial Group. Rituximab for rheumatoid
had the opportunity to review the manuscript. The Molecular pharmacodynamics of new oral drugs
funding source had no role in the preparation or arthritis refractory to anti-tumor necrosis factor
used in the treatment of multiple sclerosis. Drug therapy: results of a multicenter, randomized,
approval of the manuscript and decision to submit Des Devel Ther. 2014;8:555-568.
the manuscript for publication. double-blind, placebo-controlled, phase III trial
14. Stüve O, Marra CM, Jerome KR, et al. Immune evaluating primary efficacy and safety at
Additional Contributions: Uma Kundu, MPharm, surveillance in multiple sclerosis patients treated twenty-four weeks. Arthritis Rheum. 2006;54(9):
and Hashem Salloukh, PhD, Medical with natalizumab. Ann Neurol. 2006;59(5):743-747. 2793-2806.
Communications, Novartis Pharma AG, provided
medical writing support, including literature search, 30. US Department of Health and Human Services.
preparation of manuscript draft, revising the draft, Guidance for industry: good pharmacovigilance

38 JAMA Neurology January 2015 Volume 72, Number 1 jamaneurology.com

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: on 03/21/2018

 Varicella-Zoster Virus Infections in MS Original Investigation Research

practices and pharmacoepidemiologic assessment. placebo-controlled, parallel-group, multiple-dose 37. Schmid DS, Jumaan AO. Impact of varicella
2005. http://www.fda.gov/downloads study. J Clin Pharmacol. 2012;52(12):1879-1890. vaccine on varicella-zoster virus dynamics. Clin
/regulatoryinformation/guidances/ucm126834.pdf. 34. Mehling M, Hilbert P, Fritz S, et al. Microbiol Rev. 2010;23(1):202-217.
Accessed August 14, 2014. Antigen-specific adaptive immune responses in 38. Arvin AM. Antiviral therapy for varicella and
31. Ricklin ME, Lorscheider J, Waschbisch A, et al. fingolimod-treated multiple sclerosis patients. Ann herpes zoster. Semin Pediatr Infect Dis. 2002;13(1):
T-cell response against varicella-zoster virus in Neurol. 2011;69(2):408-413. 12-21.
fingolimod-treated MS patients. Neurology. 2013;81 35. Marin M, Güris G, Chaves SS, et al. Prevention 39. Harpaz R, Ortega-Sanchez IR, Seward J.
(2):174-181. of varicella: recommendations of the Advisory Prevention of herpes zoster: recommendations of
32. Zerboni L, Nader S, Aoki K, Arvin AM. Analysis Committee on Immunization Practices (ACIP). 2007. the Advisory Committee on Immunization Practices
of the persistence of humoral and cellular immunity http://www.cdc.gov/mmwr/preview/mmwrhtml (ACIP). http://www.cdc.gov/mmwr/preview
in children and adults immunized with varicella /rr5604a1.htm. Accessed February 28, 2014. /mmwrhtml/rr57e0515a1.htm. Published May 25,
vaccine. J Infect Dis. 1998;177(6):1701-1704. 36. Sauerbrei A, Wutzler P. Serological detection of 2008. Accessed February 28, 2014.
33. Boulton C, Meiser K, David OJ, Schmouder R. varicella-zoster virus–specific immunoglobulin G by 40. David OJ, Kovarik JM, Schmouder RL. Clinical
Pharmacodynamic effects of steady-state an enzyme-linked immunosorbent assay using pharmacokinetics of fingolimod. Clin Pharmacokinet.
fingolimod on antibody response in healthy glycoprotein antigen. J Clin Microbiol. 2006;44(9): 2012;51(1):15-28.
volunteers: a 4-week, randomized, 3094-3097.

jamaneurology.com JAMA Neurology January 2015 Volume 72, Number 1 39

Copyright 2014 American Medical Association. All rights reserved.

Downloaded From: on 03/21/2018