Edinburgh Research Explorer
Trial of Cannabidiol for Drug-Resistant Seizures in the Dravet
Syndrome
Citation for published version:
Devinsky, O, Cross, JH, Laux, L, Marsh, E, Miller, I, Nabbout, R, Scheffer, IE, Thiele, EA, Wright, S & the
Cannabidiol in Dravet Syndrome Study Group 2017, 'Trial of Cannabidiol for Drug-Resistant Seizures in the
Dravet Syndrome', New England Journal of Medicine, vol. 376, no. 21, pp. 2011-2020.
https://doi.org/10.1056/NEJMoa1611618
Digital Object Identifier (DOI):
10.1056/NEJMoa1611618
Link:
Link to publication record in Edinburgh Research Explorer
Document Version:
Publisher's PDF, also known as Version of record
Published In:
New England Journal of Medicine
Publisher Rights Statement:
Copyright © 2017 Massachusetts Medical Society. Reprinted with permission.
General rights
Copyright for the publications made accessible via the Edinburgh Research Explorer is retained by the author(s)
and / or other copyright owners and it is a condition of accessing these publications that users recognise and
abide by the legal requirements associated with these rights.
Take down policy
The University of Edinburgh has made every reasonable effort to ensure that Edinburgh Research Explorer
content complies with UK legislation. If you believe that the public display of this file breaches copyright please
contact openaccess@ed.ac.uk providing details, and we will remove access to the work immediately and
investigate your claim.
Download date: 10. Aug. 2022
The new england
journal of medicine
n engl j med 376;21 nejm.org May 25, 2017
2011
established in 1812
May 25, 2017
vol. 376 no. 21
From the New York University Langone
Comprehensive Epilepsy Center, New York
(O.D.); the University College London
Great Ormond Street Institute of Child
Health (J.H.C.) and GW Pharmaceuticals
(S.W.) — both in London; Lurie Children’s
Epilepsy Center, Ann and Robert H. Lurie
Children’s Hospital of Chicago, Chicago
(L.L.); the Children’s Hospital of Philadel-
phia, Philadelphia (E.M.); Miami Children’s
Hospital, Miami (I.M.); Hôpital Necker–
Enfants Malades, Paris (R.N.); Florey Insti-
tute, Austin Health and Royal Children’s
Hospital, University of Melbourne, Mel-
bourne, VIC, Australia (I.E.S.); and Massa-
chusetts General Hospital, Boston (E.A.T.).
Address reprint requests to Dr. Devinsky
at the Department of Neurology, NYU
Langone School of Medicine, 223 E. 34th
St., New York, NY 10016, or at od4@ nyu
. edu.
* A complete list of investigators in the
Cannabidiol in Dravet Syndrome Study
Group is provided in the Supplementary
Appendix, available at NEJM.org.
Drs. Devinsky and Cross contributed equal-
ly to this article.
N Engl J Med 2017;376:2011-20.
DOI: 10.1056/NEJMoa1611618
Copyright © 2017 Massachusetts Medical Society.
BACKGROUND
The Dravet syndrome is a complex childhood epilepsy disorder that is associated with
drug-resistant seizures and a high mortality rate. We studied cannabidiol for the treat
-
ment of drug-resistant seizures in the Dravet syndrome.
METHODS
In this double-blind, placebo-controlled trial, we randomly assigned 120 children and
young adults with the Dravet syndrome and drug-resistant seizures to receive either
cannabidiol oral solution at a dose of 20 mg per kilogram of body weight per day or
placebo, in addition to standard antiepileptic treatment. The primary end point was the
change in convulsive-seizure frequency over a 14-week treatment period, as compared
with a 4-week baseline period.
RESULTS
The median frequency of convulsive seizures per month decreased from 12.4 to 5.9 with
cannabidiol, as compared with a decrease from 14.9 to 14.1 with placebo (adjusted
median difference between the cannabidiol group and the placebo group in change in
seizure frequency, −22.8 percentage points; 95% confidence interval [CI], −41.1 to −5.4;
P = 0.01). The percentage of patients who had at least a 50% reduction in convulsive-
seizure frequency was 43% with cannabidiol and 27% with placebo (odds ratio, 2.00;
95% CI, 0.93 to 4.30; P = 0.08). The patient’s overall condition improved by at least one
category on the seven-category Caregiver Global Impression of Change scale in 62% of
the cannabidiol group as compared with 34% of the placebo group (P = 0.02). The fre
-
quency of total seizures of all types was significantly reduced with cannabidiol
(P = 0.03), but there was no significant reduction in nonconvulsive seizures. The per
-
centage of patients who became seizure-free was 5% with cannabidiol and 0% with
placebo (P = 0.08). Adverse events that occurred more frequently in the cannabidiol
group than in the placebo group included diarrhea, vomiting, fatigue, pyrexia, somno
-
lence, and abnormal results on liver-function tests. There were more withdrawals from
the trial in the cannabidiol group.
CONCLUSIONS
Among patients with the Dravet syndrome, cannabidiol resulted in a greater reduction in
convulsive-seizure frequency than placebo and was associated with higher rates of adverse
events. (Funded by GW Pharmaceuticals; ClinicalTrials.gov number, NCT02091375.)
abstract
Trial of Cannabidiol for Drug-Resistant Seizures
in the Dravet Syndrome
Orrin Devinsky, M.D., J. Helen Cross, Ph.D., F.R.C.P.C.H., Linda Laux, M.D., Eric Marsh, M.D., Ian Miller, M.D.,
Rima Nabbout, M.D., Ingrid E. Scheffer, M.B., B.S., Ph.D., Elizabeth A. Thiele, M.D., Ph.D.,
and Stephen Wright, M.D., for the Cannabidiol in Dravet Syndrome Study Group*
The New England Journal of Medicine
Downloaded from nejm.org at UNIVERSITY OF EDINBURGH LIB on June 9, 2017. For personal use only. No other uses without permission.
Copyright © 2017 Massachusetts Medical Society. All rights reserved.
n engl j med 376;21 nejm.org May 25, 2017
2012
The
new england journal
of
medicine
S
eizures are difficult to control in
the Dravet syndrome, a rare genetic form
of epileptic encephalopathy primarily due
to loss-of-function mutations in the SCN1A gene.
Interest in cannabidiol for the treatment of epi-
lepsy was generated by media reports of efficacy
in children with the Dravet syndrome.
1
Four small
trials of cannabidiol had yielded mixed results.
2-5
A series of in vitro and in vivo preclinical models
of seizure showed that cannabidiol had activity
against convulsive seizures.
6
Subsequently, the
safety and effectiveness of a standardized oral
solution of cannabidiol was tested in an open-
label trial involving 214 children and young
adults with drug-resistant epilepsy.
7
We conducted
a randomized, double-blind, placebo-controlled
trial of cannabidiol to treat drug-resistant epi-
lepsy in the Dravet syndrome.
Methods
Trial Design and Oversight
This was a multinational, randomized, double-
blind trial of adjunctive cannabidiol versus pla-
cebo in children and young adults 2 to 18 years
of age with the Dravet syndrome whose seizures
were not controlled by their current antiepileptic-
drug regimen. The trial comprised a 4-week base-
line period, a 14-week treatment period (2 weeks
of dose escalation and 12 weeks of dose main-
tenance), a 10-day taper period, and a 4-week
safety follow-up period. The trial was approved
by the review board or ethics committee at each
participating institution and was conducted in
accordance with the principles of the Declara-
tion of Helsinki and the International Confer-
ence on Harmonisation Good Clinical Practice
guidelines. All the patients or their parents or
legal representatives provided written informed
consent, and children mature enough to under-
stand the trial provided assent. Patients could
withdraw at any point without prejudice.
The funding source, GW Pharmaceuticals, was
responsible for the trial design (with input from
investigators and other experts), trial manage-
ment, site monitoring, trial pharmacovigilance,
data analysis, and statistical analysis. GW Phar-
maceuticals prepared and provided the active
treatment and placebo. Trial procedures were
reviewed at multisite investigator meetings. Ser-
vices were used for clinical laboratory testing;
bioanalytical laboratory testing; design of the
case-report form; data management; trial-agent
distribution, returns, and destruction; the inter-
active voice-response system; diagnosis of the
Dravet syndrome and seizure classification; and
translation of documents. The authors vouch for
the accuracy and completeness of the reported
data and analyses and for the adherence of the
trial to the protocol (available with the full text
of this article at NEJM.org). The authors affirm
that they approved the final draft of the manu-
script.
Patients were eligible if they had an estab-
lished diagnosis of the Dravet syndrome, were
taking one or more antiepileptic drugs, and had
had four or more convulsive seizures during the
28-day baseline period. An independent review
of the previously documented diagnosis of the
Dravet syndrome and the classification of seizure
type was conducted for each patient by an inde-
pendent panel appointed by the Epilepsy Study
Consortium, under a standard protocol (see the
protocol). All medications or interventions for
epilepsy, including a ketogenic diet and vagus-
nerve stimulation, were stable for 4 weeks before
screening and were to remain unchanged
throughout the trial. The dose of cannabidiol
used in the trial was recommended by an inde-
pendent data and safety monitoring committee
(see the protocol), whose members reviewed
data from a dose-ranging pharmacokinetic and
safety evaluation of three doses of cannabidiol
(5, 10, and 20 mg per kilogram of body weight
per day) and identified the maximum dose that
was safe and was not associated with unaccept-
able side effects.
Procedures
After informed consent was obtained, patients
entered a 4-week baseline period. The investiga-
tor trained the caregiver to record daily seizure
information. Patients who satisfied all eligibility
criteria were randomly assigned in a 1:1 ratio to
receive cannabidiol or matching placebo, in ad-
dition to their stable antiepileptic-drug regimens.
Cannabidiol oral solution contained 100 mg of
cannabidiol per milliliter. The placebo solution
was identical to the cannabidiol solution except
for the absence of cannabidiol. The dose was
escalated up to 20 mg per kilogram per day (or
the equivalent volume of placebo) with the use
of a 14-day dosing regimen that was approved
by the data and safety monitoring committee.
The New England Journal of Medicine
Downloaded from nejm.org at UNIVERSITY OF EDINBURGH LIB on June 9, 2017. For personal use only. No other uses without permission.
Copyright © 2017 Massachusetts Medical Society. All rights reserved.
n engl j med 376;21 nejm.org May 25, 2017
2013
Cannabidiol for Seizures in the Dravet Syndrome
All doses were administered twice daily. At the
end of the treatment period, the cannabidiol and
placebo solutions were tapered (10% each day)
over a period of 10 days. After trial completion,
all patients could enter a long-term open-label
study.
Each day, patients or their caregivers recorded
the number and type of convulsive seizures
(tonic, clonic, tonic–clonic, or atonic) for the
primary end-point measure of convulsive-seizure
frequency, using an interactive voice-response
system. Clinical laboratory assessments were
performed at baseline and after 2, 4, 8, and 14
weeks of the trial regimen, as well as at the end
of the taper period for those patients who did
not enter the open-label extension study or who
withdrew early and tapered the trial agent.
End Points
The primary end point was the percentage change
per 28 days from the 4-week baseline period in
convulsive-seizure frequency during the 14-week
treatment period among patients who received
cannabidiol as compared with placebo. The
treatment period extended from randomization
to the end of the 14-week trial or the date of the
last dose. The maintenance period extended from
the end of the 2-week dose-escalation period to the
end of the 14-week trial or the date of the last
dose. The intention-to-treat analysis set included
all patients in the safety analysis set who had
postbaseline efficacy data.
The secondary end-point measures were the
Caregiver Global Impression of Change (CGIC),
assessed on a 7-point Likert-like scale that used
three categories of improvement (slightly im-
proved, much improved, or very much improved),
three categories of worsening (slightly worse,
much worse, or very much worse), and an option
of “no change”; the number of patients with a
reduction in convulsive-seizure frequency of at
least 25%, at least 50%, at least 75%, and 100%;
reduction in total seizure frequency and reduc-
tion of seizure subtypes; the duration of seizure
subtypes, as assessed by the Caregiver Global Im-
pression of Change in Seizure Duration (CGICSD)
on a 3-point scale (decrease, no change, or in-
crease in average duration); sleep disruption,
assessed on a numerical rating scale from 0 to
10, with higher scores indicating greater disrup-
tion; the change in the score on the Epworth
Sleepiness Scale (range, 0 to 24, with higher
scores indicating greater daytime sleepiness);
the score on the Quality of Life in Childhood
Epilepsy questionnaire (range, 0 to 100, with
higher scores indicating better function); the
age-standardized score on the Vineland Adaptive
Behavior Scales, second edition (Vineland-II;
range, 20 to 160, with higher scores indicating
better behavioral adaptation); the number of hos-
pitalizations due to epilepsy; the number of pa-
tients with the emergence of seizure types that
had not occurred during the baseline period;
and the use of rescue medication.
The safety profile of cannabidiol was assessed
on the basis of the number, type, and severity of
adverse events as well as the Columbia Suicide
Severity Rating Scale (for patients ≥6 years of
age, when appropriate), vital signs, electrocardio-
graphic variables, laboratory safety variables, and
physical examination variables; safety end points
were monitored at each visit. The palatability of
the trial agents was assessed by caregivers on a
5-point scale, ranging from “liked it a lot” to
“did not like it at all.”
Statistical Analysis
A total of 100 randomly assigned patients were
planned. We calculated that this sample size
would provide 80% power to detect an absolute
difference of 32 percentage points between groups
in the primary end point in an intention-to-treat
analysis, with a standard deviation of 56% and a
two-sided significance level of 5%. Randomiza-
tion was performed and assigned independently,
held centrally, and not divulged to any other per-
son involved in the trial until after database lock.
Analysis of the primary end point was per-
formed with the use of a Wilcoxon rank-sum
test. An estimate of the median difference be-
tween cannabidiol and placebo, together with
the 95% confidence interval, was calculated with
the use of the Hodges–Lehmann approach. Sen-
sitivity analyses of this primary end point were
prespecified in the trial protocol and statistical
analysis plan.
The percentage of patients with a reduction in
convulsive-seizure frequency from baseline of at
least 25%, at least 50%, at least 75%, or 100%
was analyzed with the use of a Cochran–Mantel–
Haenszel test and presented with odds ratios.
The changes from baseline in the CGIC and the
CGICSD were analyzed with the use of an ordi-
nal logistic-regression model. For the secondary
The New England Journal of Medicine
Downloaded from nejm.org at UNIVERSITY OF EDINBURGH LIB on June 9, 2017. For personal use only. No other uses without permission.
Copyright © 2017 Massachusetts Medical Society. All rights reserved.
n engl j med 376;21 nejm.org May 25, 2017
2014
The
new england journal
of
medicine
end points, there were no adjustments of P val-
ues for multiple comparisons.
Results
Trial Population
At 23 centers in the United States and Europe,
177 patients were screened and 120 underwent
randomization (Fig. 1). The characteristics of the
trial groups were similar (Table 1). The mean
age of the patients was 9.8 years (range, 2.3 to
18.4), and 52% were male. The baseline convul-
sive-seizure frequency was a median of 13.0
seizures per month (range, 3.7 to 1717). A total
of 108 patients (90%) completed the treatment
period (52 of 61 patients [85%] in the cannabi-
diol group and 56 of 59 patients [95%] in the
placebo group). A total of 12 patients (10%)
withdrew from the trial before completion (9 in
the cannabidiol group and 3 in the placebo
group). Of the 108 patients who completed
the trial, 105 entered the open-label extension
study.
Patients had previously tried a median of 4.0
antiepileptic drugs (range, 0 to 26) and were tak-
ing a median of 3.0 (range, 1 to 5). The most
common were clobazam (65%), valproates (all
forms, 59%), stiripentol (42%), levetiracetam
(28%), and topiramate (26%). The most common
type of convulsive seizure was generalized tonic–
clonic, in 94 patients (78%), with secondarily
generalized tonic–clonic seizures in 25 patients
(21%) and other convulsive-seizure types less
frequently. Nonconvulsive seizures were reported
in 37 patients in the cannabidiol group (61%)
and 41 patients in the placebo group (69%).
Developmental delay was observed in 114 of the
118 children with available data and was de-
scribed as severe or profound in 56 (48%) and
mild or moderate in 58 (50%).
Adherence to the data acquisition and voice-
response system was 97% for the cannabidiol
group and 98% for the placebo group during the
baseline period and 97% and 96%, respectively,
during the treatment period. The mean (±SD)
number of days on which a dose was missed was
0.6±2.1 in the cannabidiol group and 0.6±2.9 in
the placebo group.
Seizure Frequency
In the cannabidiol group, the primary end point
of convulsive-seizure frequency decreased from
a median of 12.4 seizures per month (range, 3.9
to 1717) at baseline to 5.9 (range, 0.0 to 2159)
over the entire treatment period (Table 2), repre-
senting a median change of −38.9% (interquar-
tile range, −69.5 to −4.8) from baseline. In the
placebo group, the median monthly convulsive-
seizure frequency decreased from 14.9 (range,
3.7 to 718) to 14.1 (range, 0.9 to 709), represent-
ing a median change of −13.3% (interquartile
range, −52.5 to 20.2). The adjusted median differ-
ence in convulsive seizures between the canna-
bidiol group and the placebo group was −22.8
percentage points (95% confidence interval [CI],
−41.1 to −5.4; P = 0.01). Prespecified sensitivity
analyses supported the primary analysis (Fig. S1
in the Supplementary Appendix, available at
NEJM.org). The difference in favor of cannabi-
diol was seen in the first month of the mainte-
nance period, during which the median number
of convulsive seizures per month declined from
12.4 to 5.0 in the cannabidiol group and from
14.9 to 13.0 in the placebo group (P = 0.002).
Secondary End Points
The results of the secondary end-point measures
are shown in Table 3. The end point of a reduc-
tion in convulsive-seizure frequency by 50% or
more during the treatment period occurred in
43% of the patients in the cannabidiol group
and in 27% of the patients in the placebo group
(odds ratio, 2.00; 95% CI, 0.93 to 4.30; P = 0.08).
During the treatment period, 3 patients in the
cannabidiol group and no patients in the placebo
group were free of seizures (P = 0.08). For total
seizures (all seizure types), the median frequency
of seizures per month decreased from 24.0 to
13.7 in the cannabidiol group (adjusted reduc-
tion, 28.6%), versus a decrease from 41.5 to 31.1
in the placebo group (adjusted reduction, 9.0%).
The adjusted median difference between groups
of −19.2 percentage points was significant (P = 0.03).
For reduction in nonconvulsive seizures, there
was no significant difference between groups
(P = 0.88). Rescue medication was used by 36
patients (59%) in the cannabidiol group and by
41 patients (69%) in the control group.
On the CGIC scale, 37 of 60 caregivers (62%)
judged their child’s overall condition improved
in the cannabidiol group, as compared with 20
of 58 caregivers (34%) in the placebo group
(P = 0.02). There was no significant difference
between groups in the sleep-disruption score
The New England Journal of Medicine
Downloaded from nejm.org at UNIVERSITY OF EDINBURGH LIB on June 9, 2017. For personal use only. No other uses without permission.
Copyright © 2017 Massachusetts Medical Society. All rights reserved.
