Study Designs
SAMARAI and SPARTAN6,7 were randomized, double-blind, placebo-controlled Phase 3 trials for acute treatment
of migraine, conducted in the United States, and, in the case of SPARTAN, the United
Kingdom and Germany also. The studies were almost identical in design. Key inclusion
criteria were a Migraine Disability Assessment score ≥11 (moderate or severe disability),
a history of migraine ≥1 year, 3-8 migraine attacks/month, <15 headache days per month,
and migraine onset at <50 years of age.
Patients were randomized equally to receive the first dose of lasmiditan 50 mg (SPARTAN
only), 100 mg, 200 mg, or placebo. Patients were instructed to take study drug on
an outpatient basis within 4 hours of onset of a migraine attack with a headache of
at least moderate severity and not improving. If needed for rescue or recurrence,
a second dose was taken 2 to 24 hours after first dose. For the second dose, patients
originally assigned lasmiditan were randomized to the same active dose or placebo
(2:1 ratio), and patients originally assigned placebo received placebo; randomization
to second dose occurred at the same time as randomization to first dose. The rescue
population referred to patients who did not achieve headache pain-free status at 2
hours, completed the 2-hour assessments, and took a second dose of study drug between
2 and 24 hours post-first dose. The recurrence population referred to patients who
achieved headache pain-free status at 2 hours but then experienced recurrence of mild,
moderate, or severe migraine pain and took a second dose of study drug up to 24 hours
from the first dose.
The primary and key secondary endpoints in SAMURAI and SPARTAN were the proportion
of patients who were headache pain-free and the proportion who were MBS free at 2
hours post-first dose.
This paper adheres to CONSORT guidelines.
Outcome measures
At baseline and at 0.5, 1, 1.5, 2, 3, 4, 24, and 48 hours after dosing, patients were
asked to rate their headache pain as none, mild, moderate, or severe. Additionally,
they were asked whether they were experiencing nausea, phonophobia, photophobia, or
vomiting. If they had nausea, phonophobia, or photophobia, they were asked, at baseline
and prior to dosing, to identify which was their most bothersome migraine associated
symptom. Patients were asked the following questions: 1) at all time points, “How much is your migraine interfering with your usual activities?” (responses reported using a 4-point numeric rating scale); and 2) at 2 hours, the
Patient Global Impression of Change (PGIC)question,8 “How do you feel after taking study medication” (responses recorded using a seven-point
Likert scale from “very much better” to “very much worse”).
Adverse Event Collection
At 0.5, 1, 1.5, 2, 3, 4, 24, and 48 hours after the first dose and again at the same
intervals after a second dose of study drug, , if taken, an electronic diary asked
“Do you feel anything unusual since taking the study medication that you have not felt
with a migraine before?”If the patient answered “yes”, they were to receive a call from the site to determine
if an adverse event (AE) had been experienced. AEs that occurred or worsened within
48 hours of dosing were considered treatment-emergent.
Statistical Analysis
Efficacy hypotheses were that, after a first dose of lasmiditan, a second dose of
lasmiditan would be superior to a second dose of placebo when taken for rescue or
when taken for recurrence. Since the number of patients choosing to take a second
dose of study medication could not be determined prior to the conduct of the study,
no formal power analyses were conducted.
Efficacy analyses were performed in the intention-to-treat (ITT) second dose population,
which was defined as patients who used a second dose of study drug to treat a migraine
and had any post-second dose headache severity or symptom assessments. This population
was further classified as ITT second dose rescue population (not pain-free at 2 hours
and took a second dose of study medication) and ITT second dose recurrence population
(pain-free at 2 hours and took a second dose of study medication). Safety analyses
utilized the second dose safety population, which was defined as all patients who
used 2 doses of study drug, regardless of whether they completed any study assessments.
Mantel-Haenszel odds ratios (ORs) for efficacy outcomes were computed and compared
between lasmiditan and placebo, with study as a covariate. For treatment comparisons,
an estimate of the OR of achieving a response, as well as the corresponding confidence
interval (CI) and p-value using Wald’s test, was computed. Time to second dose graphics
were created using Kaplan-Meier analyses. Continuous variables were summarized using
descriptive statistics; categorical variables were summarized using counts and percentages.
All efficacy endpoints were tested at a 2-sided significance level of 0.05.
Analyses were performed using SAS software, version 9.4/EG 7.1 (SAS Institute, Cary,
NC).