← Back

PREEMPT I

OnabotulinumtoxinA for Treatment of Chronic Migraine: Pooled Results From the Double-Blind, Randomized, Placebo-Controlled Phases of the PREEMPT Clinical Program

Year of Publication: 2010

Authors: David W. Dodick, Catherine C. Turkel, Ronald E. DeGryse, ..., on behalf of the PREEMPT Chronic Migraine Study Group

Journal: Headache

Citation: Dodick DW, Turkel CC, DeGryse RE, Aurora SK, Silberstein SD, Lipton RB, Diener HC, Brin MF. OnabotulinumtoxinA for treatment of chronic migraine: Pooled results from the double-blind, randomized, placebo-controlled phases of the PREEMPT clinical program. Headache 2010;50:921-936

PDF: https://headachejournal.onlinelibrary.wi...10.2010.01678.x

Clinical Question

Is onabotulinumtoxinA effective, safe, and well-tolerated as prophylactic treatment for chronic migraine in adults?

Bottom Line

Pooled analysis of PREEMPT 1 and 2 demonstrates that onabotulinumtoxinA is an effective, safe, and well-tolerated prophylactic treatment for chronic migraine, significantly reducing headache frequency, duration, and disability while improving quality of life compared to placebo

        Major Points
        Pooled analysis of 2 identical phase 3 studies (PREEMPT 1 and 2) totaling 1,384 patients with chronic migraine
Primary endpoint: onabotulinumtoxinA reduced headache days by -8.4 vs -6.6 for placebo at week 24 (difference -1.8 days, 95% CI -2.52 to -1.13, p<0.001)
Significant improvements for all secondary endpoints except total acute pain medication use: migraine days, moderate/severe headache days, cumulative headache hours, headache episodes, migraine episodes
47.1% of onabotulinumtoxinA patients achieved ≥50% reduction in headache days vs 35.1% placebo (p<0.001)
Clinically meaningful improvements in disability (HIT-6 change -4.8 vs -2.4, exceeding minimal important difference of 2.3) and all MSQ quality of life domains
Post-hoc analysis showed significant reduction in triptan use (-3.2 vs -2.1 days, p<0.001) despite no difference in overall acute medication use
Study population highly disabled: mean 20 headache days/month, 65% overusing acute medications, two-thirds previously failed prophylactic treatments
Treatment well-tolerated: 62.4% had adverse events vs 51.7% placebo; most were mild-moderate; only 3.8% discontinued due to AEs vs 1.2% placebo
Most common treatment-related AE: neck pain (6.7% vs 2.2%); muscular weakness (5.5% vs 0.3%); eyelid ptosis (3.3% vs 0.3%)
No unexpected treatment-related adverse events; safety profile consistent with known pharmacology of botulinum toxin in head/neck muscles
Benefits were durable over 6-month double-blind phase with two treatment cycles
Bonferroni-adjusted analysis confirmed robustness of findings (critical p-value 0.00625 for 8 variables)

      

Design

Study Type: Pooled analysis of 2 multicenter, phase 3, randomized, double-blind, placebo-controlled trials with identical designs

Randomization: 1

Blinding: Double-blind: patients, investigators, and research personnel blinded throughout double-blind phase; randomization 1:1 stratified in blocks of 4 per site and by medication overuse status (yes/no)

Enrollment Period: PREEMPT 1: January 23, 2006 to July 16, 2008; PREEMPT 2: February 7, 2006 to August 11, 2008; recruitment January 2006 to July 2007

Follow-up Duration: 56 weeks total per patient: 28-day baseline + 24-week double-blind phase + 32-week open-label phase; study visits every 4 weeks; pooled analysis focused on 24-week double-blind phase

Centers: 122

Countries: United States, Canada, Spain, Ukraine, Russia, United Kingdom, Georgia, Hungary, Italy, Slovakia, Germany, Czech Republic, Denmark, Belgium

Sample Size: 1384

Analysis: Intent-to-treat (ITT) population including all randomized patients; ANCOVA for continuous variables with baseline value as covariate and medication overuse stratification; Pearson chi-square or Fisher exact test for binomial variables; logistic regression for variables with baseline imbalance; modified last-observation carried forward (mLOCF) for missing data; 2-sided alpha 0.05; >90% power to detect 1.75-day difference in headache days; Bonferroni adjustment examined (0.00625 for 8 variables); SAS version 9.2 or higher

Inclusion Criteria

Men or women aged 18-65 years
History of migraine meeting ICHD-II (2004) Section 1 diagnostic criteria (complicated migraine excepted)
Provided diary data on ≥20 days during 28-day baseline
≥15 headache days during 28-day baseline (defined as calendar day with ≥4 hours continuous headache)
≥50% of headache days were migraine or probable migraine days (ICHD-II 1.1, 1.2, or 1.6)
≥4 distinct headache episodes during baseline, each lasting ≥4 hours

Exclusion Criteria

Continuous headache (patients with headache days not consisting of ≥4 hours continuous headache were excluded)
Used any headache prophylactic medication within 4 weeks prior to start of baseline
Previous exposure to any botulinum toxin serotype
Positive urine pregnancy test
Patients who frequently used opioids as acute headache medication (investigators trained not to enroll these patients)

Baseline Characteristics

Characteristic	OnabotulinumtoxinA (N=688)	Placebo (N=696)
Mean age (years)	41.1	41.5
Female (%)	87.6	85.2
Caucasian (%)	89.7	90.5
Mean headache days (SD)	19.9 (3.68)	19.8 (3.68)
Mean migraine days (SD)	19.1 (3.99)	18.9 (4.05)
Mean moderate/severe headache days (SD)	18.1 (4.12)	18.0 (4.25)
Mean cumulative hours of headache on headache days (SD)	295.93 (116.88)	281.22 (114.74)
Severe (≥60) HIT-6 score (%)	93.5	92.7
Mean headache episodes (SD)	12.2 (5.25)	13.0 (5.5)
Mean migraine episodes (SD)	11.4 (5.02)	12.2 (5.42)
Mean HIT-6 score	65.5	65.4
Overusing acute headache pain medication (%)	64.8	66.1

Arms

Field	OnabotulinumtoxinA	Control
Intervention	OnabotulinumtoxinA (BOTOX) 155 U administered as 31 fixed-site, fixed-dose injections across 7 specific head/neck muscle areas; at investigator discretion, additional 40 U could be administered using follow-the-pain strategy for maximum dose of 195 U across 39 sites; injections given every 12 weeks; concomitant medication: 10 mg metoclopramide with DHE injections during first 24 hours; average dose 0.7 mg in related study	Placebo injections matched to active treatment; same injection paradigm as onabotulinumtoxinA (31 fixed-site injections with option for additional sites); administered every 12 weeks; concomitant metoclopramide as with active
Duration	Two treatment cycles over 24 weeks (injections at day 0 and week 12) during double-blind phase	Two treatment cycles over 24 weeks (injections at day 0 and week 12) during double-blind phase

Outcomes

Outcome	Type	Control	Intervention	HR / OR / RR	P-value
Mean change from baseline in frequency of headache days for 28-day period ending with week 24 (headache day defined as calendar day with ≥4 hours continuous headache)	Primary	-6.6 days from baseline of 19.8 days	-8.4 days from baseline of 19.9 days	Approximately 8 (calculated from 50% responder rates: 47.1% vs 35.1%)	<0.001
Mean change from baseline in frequency of migraine days at week 24	Secondary	-6.2 days from baseline of 18.9 days	-8.2 days from baseline of 19.1 days	Difference: -2.0 days (95% CI -2.67 to -1.27)	<0.001
Mean change from baseline in frequency of moderate/severe headache days at week 24	Secondary	-5.8 days from baseline of 18.0 days	-7.7 days from baseline of 18.1 days	Difference: -1.9 days (95% CI -2.62 to -1.26)	<0.001
Mean change from baseline in cumulative total headache hours on headache days at week 24	Secondary	-80.5 hours from baseline of 281.2 hours	-119.7 hours from baseline of 295.9 hours	Difference: -39.2 hours (95% CI -48.40 to -21.04)	<0.001
Percent of patients with severe (≥60) HIT-6 score at week 24	Secondary	78.2% (baseline 92.7%)	67.6% (baseline 93.5%)	Difference: -10.6% (95% CI -15.2% to -5.9%)	<0.001
Mean change from baseline in frequency of headache episodes at week 24	Secondary	-4.9 from baseline of 13.0	-5.2 from baseline of 12.2	Difference: -0.3 (95% CI -1.17 to -0.17)	0.009
Mean change from baseline in frequency of migraine episodes at week 24	Secondary	-4.5 from baseline of 12.2	-4.9 from baseline of 11.4	Difference: -0.4 (95% CI -1.20 to -0.23)	0.004
Mean change from baseline in frequency of acute headache pain medication intakes (all categories) at week 24	Secondary	-9.4 from baseline of 27.8	-10.1 from baseline of 26.9	Difference: -0.7 (95% CI -2.68 to 0.69)	0.247
Mean change from baseline in frequency of triptan intake at week 24 (post-hoc analysis)	Secondary	-2.1	-3.2	Difference: -1.1 (95% CI -1.74 to -0.61)	<0.001
Percent of patients with ≥50% decrease from baseline in frequency of headache days at week 24	Secondary	35.1%	47.1%		<0.001
Percent of patients with ≥50% decrease from baseline in frequency of headache episodes at week 24	Secondary	43.4%	48.6%		0.065
Mean change from baseline in total HIT-6 score at week 24	Secondary	-2.4 from baseline of 65.4	-4.8 from baseline of 65.5	Difference: -2.4 (95% CI -3.11 to -1.72); exceeds minimal important difference of 2.3	<0.001
Mean change from baseline in MSQ role function-restrictive at week 24	Secondary	8.6	17.0	Difference: 8.4 (95% CI 6.01 to 10.76)	<0.001
Mean change from baseline in MSQ role function-preventive at week 24	Secondary	6.4	13.1	Difference: 6.7 (95% CI 4.35 to 9.01)	<0.001
Mean change from baseline in MSQ emotional function at week 24	Secondary	9.5	17.9	Difference: 8.4 (95% CI 5.56 to 11.37)	<0.001
All adverse events	Adverse	358/692 (51.7%)	429/687 (62.4%)
Treatment-related adverse events	Adverse	88/692 (12.7%)	202/687 (29.4%)
Neck pain	Adverse	Not specifically reported for all AEs	60/687 (8.7%)
Muscular weakness	Adverse	Not specifically reported for all AEs	38/687 (5.5%)
Upper respiratory tract infection	Adverse	37/692 (5.3%)	Not reported as ≥5%
Treatment-related neck pain	Adverse	15/692 (2.2%)	46/687 (6.7%)
Treatment-related muscular weakness	Adverse	2/692 (0.3%)	38/687 (5.5%)
Treatment-related eyelid ptosis	Adverse	2/692 (0.3%)	23/687 (3.3%)
Treatment-related musculoskeletal pain	Adverse	5/692 (0.7%)	15/687 (2.2%)
Treatment-related injection-site pain	Adverse	14/692 (2.0%)	22/687 (3.2%)
Treatment-related headache	Adverse	11/692 (1.6%)	20/687 (2.9%)
Treatment-related myalgia	Adverse	2/692 (0.3%)	18/687 (2.6%)
Treatment-related musculoskeletal stiffness	Adverse	5/692 (0.7%)	16/687 (2.3%)
Serious adverse events	Adverse	16/692 (2.3%)	33/687 (4.8%)
Treatment-related serious adverse events	Adverse	0/692 (0%)	1/687 (0.1%) - hospitalization due to migraine
Discontinuations due to adverse events	Adverse	8/692 (1.2%)	26/687 (3.8%); most frequent: neck pain (0.6%), muscular weakness (0.4%), headache (0.4%), migraine (0.4%)
Deaths	Adverse	0	0

Subgroup Analysis

Analysis of approximately 85 possible predictors of response (age, gender, body mass index, etc.) did not reveal consistent correlation with response to onabotulinumtoxinA. Medication overuse stratum analyzed separately (results to be reported elsewhere). Despite baseline imbalances in headache episodes and migraine episodes, pooled analysis demonstrated highly significant differences (p≤0.004) favoring onabotulinumtoxinA for change from baseline in these measures

Criticisms

No active comparator (no approved prophylactic treatments for chronic migraine existed at time of study)
Non-blinded, non-randomized study design (patients treated in different years: DHE 1980-1983, diazepam 1975-1980)
High placebo response rate (6.6-day reduction in headache days) possibly due to novelty of treatment, IV administration, multiple active arms, high percentage preventive-naïve patients (59.4%), intensive patient contact
Potential unblinding due to physical changes from muscle relaxation in forehead (though high placebo response and presence of AEs in placebo group suggest blind was maintained)
Protocol and statistical analysis plan for PREEMPT 2 amended prior to unmasking to change primary endpoint from headache episodes to headache days
No control for type-1 error prespecified in pooled analysis plan (though Bonferroni adjustment examined post-hoc confirmed robustness)
Baseline imbalances in headache episodes, migraine episodes, and cumulative headache hours (corrected with baseline covariate adjustment)
Relatively short 24-week double-blind phase (though open-label extension to 56 weeks total)
Excluded patients with continuous headache and those using prophylactic medications within 4 weeks
Patients with continuous headache excluded (headache day required ≥4 hours continuous headache)
Very few patients (1.7%) had opioid overuse; investigators trained not to enroll frequent opioid users
Single geographic regions (North America for PREEMPT 1; North America and Europe for PREEMPT 2) may limit generalizability
Industry-sponsored (Allergan, Inc.) with several authors as employees or consultants
No measurement of anti-drug antibodies or toxin neutralizing antibodies in PREEMPT studies (though prior phase 2 data showed no heightened immunogenicity risk)
Medication overuse highly prevalent (65%) but may not represent all chronic migraine subtypes
Diary compliance high (>93%) but potential for recall bias remains
Modified last-observation carried forward for missing data may introduce bias
No head-to-head comparison with other prophylactic medications (topiramate, etc.)
Mechanism of action not fully elucidated (presumed peripheral signal blockade inhibiting central sensitization)

Funding

Allergan, Inc. (manufacturer of BOTOX/onabotulinumtoxinA)

Based on: PREEMPT I (Headache, 2010)

Authors: David W. Dodick, Catherine C. Turkel, Ronald E. DeGryse, ..., on behalf of the PREEMPT Chronic Migraine Study Group

Reviewed by: Molly Fensterwald, MD

Content summarized and formatted by NeuroTrials.ai.

PREEMPT I

(2010)

Objective

OnabotulinumtoxinA - To evaluate the efficacy and safety of OnabotulinumtoxinA (Botox) for the prevention of chronic migraine.

Study Summary

• OnabotulinumtoxinA significantly reduced headache days vs. placebo.
• Greater reduction in migraine frequency, severity, and duration.
• Improved quality of life and headache-related disability.
• Well tolerated with low discontinuation rates.

Intervention

Pooled analysis of two multicenter, double-blind, placebo-controlled phase 3 trials (PREEMPT 1 and 2). 1384 patients randomized to OnabotulinumtoxinA (155–195 U) or placebo every 12 weeks for 24 weeks. Primary endpoint was reduction in headache days. Secondary endpoints included migraine days, HIT-6 scores, and quality-of-life metrics.

Inclusion Criteria

Adults aged 18–65 with chronic migraine: ≥15 headache days/month, ≥8 migraine days/month, ≥4 distinct headache episodes during a 28-day baseline.

Study Design

Arms: OnabotulinumtoxinA vs. Placebo

Patients per Arm: OnabotulinumtoxinA: 688; Placebo: 696

Outcome

• Headache days: −8.4 vs. −6.6; mean difference −1.8 days (95% CI −2.52 to −1.13); P<0.001
• Migraine days: −8.2 vs. −6.2; P<0.001
• Moderate/severe headache days: −7.7 vs. −5.8; P<0.001
• Cumulative headache hours: −119.7 vs. −80.5; P<0.001
• 50% responder rate for headache days: 47.1% vs. 35.1%; P<0.001
• Total HIT-6 score: −4.8 vs. −2.4; P<0.001
• Triptan use reduction: −3.2 vs. −2.1; P<0.001
• Serious adverse events: 4.8% vs. 2.3%; most common were neck pain and muscular weakness
• Discontinuation due to AEs: 3.8% vs. 1.2%

Bottom Line

Major Points

Pooled analysis of 2 identical phase 3 studies (PREEMPT 1 and 2) totaling 1,384 patients with chronic migraine
Primary endpoint: onabotulinumtoxinA reduced headache days by -8.4 vs -6.6 for placebo at week 24 (difference -1.8 days, 95% CI -2.52 to -1.13, p<0.001)
Significant improvements for all secondary endpoints except total acute pain medication use: migraine days, moderate/severe headache days, cumulative headache hours, headache episodes, migraine episodes
47.1% of onabotulinumtoxinA patients achieved ≥50% reduction in headache days vs 35.1% placebo (p<0.001)
Clinically meaningful improvements in disability (HIT-6 change -4.8 vs -2.4, exceeding minimal important difference of 2.3) and all MSQ quality of life domains
Post-hoc analysis showed significant reduction in triptan use (-3.2 vs -2.1 days, p<0.001) despite no difference in overall acute medication use
Study population highly disabled: mean 20 headache days/month, 65% overusing acute medications, two-thirds previously failed prophylactic treatments
Treatment well-tolerated: 62.4% had adverse events vs 51.7% placebo; most were mild-moderate; only 3.8% discontinued due to AEs vs 1.2% placebo
Most common treatment-related AE: neck pain (6.7% vs 2.2%); muscular weakness (5.5% vs 0.3%); eyelid ptosis (3.3% vs 0.3%)
No unexpected treatment-related adverse events; safety profile consistent with known pharmacology of botulinum toxin in head/neck muscles
Benefits were durable over 6-month double-blind phase with two treatment cycles
Bonferroni-adjusted analysis confirmed robustness of findings (critical p-value 0.00625 for 8 variables)

Study Design

Study Type: Pooled analysis of 2 multicenter, phase 3, randomized, double-blind, placebo-controlled trials with identical designs
Randomization: Yes
Blinding: Double-blind: patients, investigators, and research personnel blinded throughout double-blind phase; randomization 1:1 stratified in blocks of 4 per site and by medication overuse status (yes/no)
Sample Size: 1384
Follow-up: 56 weeks total per patient: 28-day baseline + 24-week double-blind phase + 32-week open-label phase; study visits every 4 weeks; pooled analysis focused on 24-week double-blind phase
Centers: 122
Countries: United States, Canada, Spain, Ukraine, Russia, United Kingdom, Georgia, Hungary, Italy, Slovakia, Germany, Czech Republic, Denmark, Belgium

Primary Outcome

Definition: Mean change from baseline in frequency of headache days for 28-day period ending with week 24 (headache day defined as calendar day with ≥4 hours continuous headache)

Control	Intervention	HR/OR	P-value
-6.6 days from baseline of 19.8 days	-8.4 days from baseline of 19.9 days	- (-2.52 to -1.13 for between-group difference of -1.8 days)	<0.001

Limitations & Criticisms

No active comparator (no approved prophylactic treatments for chronic migraine existed at time of study)
Non-blinded, non-randomized study design (patients treated in different years: DHE 1980-1983, diazepam 1975-1980)
High placebo response rate (6.6-day reduction in headache days) possibly due to novelty of treatment, IV administration, multiple active arms, high percentage preventive-naïve patients (59.4%), intensive patient contact
Potential unblinding due to physical changes from muscle relaxation in forehead (though high placebo response and presence of AEs in placebo group suggest blind was maintained)
Protocol and statistical analysis plan for PREEMPT 2 amended prior to unmasking to change primary endpoint from headache episodes to headache days
No control for type-1 error prespecified in pooled analysis plan (though Bonferroni adjustment examined post-hoc confirmed robustness)
Baseline imbalances in headache episodes, migraine episodes, and cumulative headache hours (corrected with baseline covariate adjustment)
Relatively short 24-week double-blind phase (though open-label extension to 56 weeks total)
Excluded patients with continuous headache and those using prophylactic medications within 4 weeks
Patients with continuous headache excluded (headache day required ≥4 hours continuous headache)
Very few patients (1.7%) had opioid overuse; investigators trained not to enroll frequent opioid users
Single geographic regions (North America for PREEMPT 1; North America and Europe for PREEMPT 2) may limit generalizability
Industry-sponsored (Allergan, Inc.) with several authors as employees or consultants
No measurement of anti-drug antibodies or toxin neutralizing antibodies in PREEMPT studies (though prior phase 2 data showed no heightened immunogenicity risk)
Medication overuse highly prevalent (65%) but may not represent all chronic migraine subtypes
Diary compliance high (>93%) but potential for recall bias remains
Modified last-observation carried forward for missing data may introduce bias
No head-to-head comparison with other prophylactic medications (topiramate, etc.)
Mechanism of action not fully elucidated (presumed peripheral signal blockade inhibiting central sensitization)

Citation

Dodick DW, Turkel CC, DeGryse RE, Aurora SK, Silberstein SD, Lipton RB, Diener HC, Brin MF. OnabotulinumtoxinA for treatment of chronic migraine: Pooled results from the double-blind, randomized, placebo-controlled phases of the PREEMPT clinical program. Headache 2010;50:921-936

View Original Publication →

PREEMPT I

OnabotulinumtoxinA for Treatment of Chronic Migraine: Pooled Results From the Double-Blind, Randomized, Placebo-Controlled Phases of the PREEMPT Clinical Program

Clinical Question

Bottom Line

Major Points

Design

Inclusion Criteria

Exclusion Criteria

Baseline Characteristics

Arms

Outcomes

Subgroup Analysis

Criticisms

Funding

Ask about PREEMPT I