U.S. flag

An official website of the United States government

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

OnabotulinumtoxinA for Injection (Botox): For the Prophylaxis of Headaches in Adults With Chronic Migraine (≥ 15 Days per Month With Headache Lasting 4 Hours a Day or Longer) [Internet]. Ottawa (ON): Canadian Agency for Drugs and Technologies in Health; 2015 Jul.

Cover of OnabotulinumtoxinA for Injection (Botox)

OnabotulinumtoxinA for Injection (Botox): For the Prophylaxis of Headaches in Adults With Chronic Migraine (≥ 15 Days per Month With Headache Lasting 4 Hours a Day or Longer) [Internet].

Show details

3RESULTS

3.1. Findings From the Literature

A total of two studies were identified from the literature for inclusion in the systematic review (Figure 1). The included studies are summarized in Table 4 and described in Section 3.2. A list of excluded studies is presented in APPENDIX 3: EXCLUDED STUDIES.

Figure 1. QUOROM Flow Diagram for Inclusion and Exclusion of Studies.

Figure 1

QUOROM Flow Diagram for Inclusion and Exclusion of Studies.

Table 4. Details of Included Studies.

Table 4

Details of Included Studies.

3.2. Included Studies

3.2.1. Description of studies

Two manufacturer-sponsored, multi-centre, randomized, DB, parallel-group, placebo-controlled, phase 3 superiority trials were included in this systematic review. Study 191622-079 (PREEMPT-1) will be referred to as Study 079 (N = 679),15,17 and Study 191622-080 (PREEMPT-2) will be referred to as Study 080 (N = 705).16,18 Both trials investigated the efficacy and safety of onabotulinumtoxinA as a headache prophylaxis in patients between 18 and 65 years of age. Patients had to have experienced 15 or more headache days per four-week period. A total dose of 155 U of onabotulinumtoxinA or placebo was administered IM as 31 fixed-site, fixed-dose injections across seven specifically-defined head/neck muscle areas. At the investigator’s discretion, the dose could be increased by an additional 40 U, using a “follow-the-pain” method, in up to three specific head/neck muscle areas that took into consideration the patient-reported usual location of predominant pain. These additional injections did not need to be consistent across treatment visits with respect to dose or number of injection sites. Both studies were identical in design (Figure 2).

Figure 2. Study Design (Studies 191622-079 and 191622-080).

Figure 2

Study Design (Studies 191622-079 and 191622-080). Source: Study 191622-079 Clinical Study Report, Study 191622-080 Clinical Study Report.

Study duration was 60 weeks, which included a four-week pre-randomization (baseline) phase, a 24-week, DB treatment phase and a 32-week OLE phase. Following the baseline phase (week −4 to day 0), patients who continued to meet the inclusion/exclusion criteria at day 0 were stratified according to medication overuse (yes/no), where medication overuse was determined by the frequency of use of acute headache pain medications during the baseline phase. Overuse of acute headache pain medications was defined as an intake of medication at least two days per week and at least 10 to 15 days per 28-day period (varying with medication category). After stratification according to medication overuse, within each stratum within each site, patients were randomly allocated in a blinded fashion to receive onabotulinumtoxinA or placebo in a 1:1 ratio.

Data presented in the systematic review are for the 24-week DB treatment phase from each trial. Data from the OLE phase are summarized in APPENDIX 6: SUMMARY OF OPEN-LABEL EXTENSION PHASE OF STUDY 079 AND STUDY 080.

3.2.2. Populations

a. Inclusion and exclusion criteria

Both trials were designed to enrol male or female patients between the ages of 18 to 65 years with a history of migraine headache disorder meeting any of the diagnostic criteria listed in the ICHD-2, Section 1, Migraine, with the exception of complicated migraine (i.e., basilar migraine, hemiplegic migraine, migrainous infarction, or ophthalmoplegic migraine). Patients were required to have, during the four-week baseline phase, 15 or more headache days with each day consisting of four or more hours of continuous headaches, and at least 50% of baseline headache days being migraine or probable migraine days, and at least four distinct headache episodes each lasting at least four hours.

Key exclusion criteria were the presence of a medical condition that might have put the patient at increased risk with exposure to onabotulinumtoxinA (e.g., a neuromuscular disease), the presence of any uncontrolled clinically important medical condition other than CM, a diagnosis of another headache disorder (e.g., complicated migraine, chronic tension-type headache, hypnic headache, hemicranias continua, new daily persistent headache), use of prophylactic headache medication within 28 days prior to start of baseline, previous use of botulinum toxin, a temporomandibular disorder, fibromyalgia, a psychiatric disorder, and/or a Beck Depression Inventory score of ≥ 24 at baseline.

b. Baseline characteristics

In both trials, there were no between-group differences with respect to baseline demographic characteristics. Median age of the included patients was approximately 42 years of age, and approximately 58% of patients were ≥ 40 years of age. The majority of patients (84.6% to 89.1%) were female. Patients were predominantly Caucasian (89.4% to 91.4%). The mean time since the onset of CM was 17.6 years to 20.6 years across treatment arms in the two trials, with 47.2% and 38.8% of patients having a time since onset of CM greater than 20 years in Studies 079 and 080, respectively. The mean age of onset of CM was 20.3 to 22.8 years of age across treatment arms in the two trials, with age of onset ranging between 1 year and 57 years of age. Table 5 below presents the baseline characteristics.

Table 5. Summary of Baseline Characteristics.

Table 5

Summary of Baseline Characteristics.

In Study 079, during the 28-day baseline period, when compared with the placebo group, the onabotulinumtoxinA group had fewer headache episodes (12.3 versus 13.4), and migraine/probable migraine episodes (11.5 versus 12.7). The onabotulinumtoxinA group had more cumulative hours of headache occurring on headache days (295.66 versus 274.88). No between-group differences were seen in other disease characteristics. The mean Headache Impact Test (HIT)-6 score was close to 65, and 94.5% of patients had a severe (≥ 60) HIT-6 category score. The mean number of headache days (as per patient diaries) was approximately 20.0 days, and the mean number of moderate to severe headache days (defined as four or more continuous hours of headache and a maximum severity of moderate or severe as per patient diary) was approximately 18 days in both treatment groups. The mean number of migraine/probable migraine days was approximately 19 days in both groups.

In Study 080, there were no between-group differences with respect to baseline disease characteristics. The mean HIT-6 score was approximately 65, and approximately 91% of patients had a severe HIT-6 category score. In both treatment groups, the mean number of headache days was approximately 20 days and the mean number of moderate to severe headache days was approximately 18 days. The mean number of migraine/probable migraine days was approximately 19 days in both treatment groups. Total mean cumulative hours of headache occurring on headache days was close to 290 hours. In both treatment groups, the mean number of headache episodes and migraine/probable migraine episodes was approximately 12 episodes and 11 episodes, respectively.

A total of 97.5% (662 out of 679) and 97.6% (688 out of 705) of patients used acute medications to treat headache pain, with a mean intake of medication(s) at baseline of approximately 30 intakes and 25 intakes in Studies 079 and 080, respectively. A total of 68.1% (462 out of 679) and 63.0% (444 out of 705) of patients overused acute headache pain medications at baseline in Studies 079 and 080, respectively. A total of 61.9% (420 out of 679) and 65.1% (459 out of 705) of patients had previously used other headache prophylactic medications prior to study enrolment in Studies 079 and 080, respectively. Table 6 presents baseline medication use.

Table 6. Baseline Medication Use.

Table 6

Baseline Medication Use.

3.2.3. Interventions

a. Double-blind phase (day 0 to week 24)

In both trials, onabotulinumtoxinA or placebo was administered at day 0 and at week 12 during the DB phase.

A minimum dose of 155 U onabotulinumtoxinA (purified neurotoxin complex) or placebo (saline) was administered IM as 31 fixed-site, fixed-dose injections across seven specific head/neck muscle areas. In addition, at the investigator’s discretion, an additional 40 U of onabotulinumtoxinA or placebo could be administered using a “follow-the-pain” paradigm in up to three specific head/neck muscle areas. The active treatment and placebo were supplied in identical glass vials.

The onabotulinumtoxinA dose, the number of injection sites per muscle, the areas to be injected, and the total dose were based on the results of two exploratory phase 2 studies (191622-038 and 191622-039).

b. Open-label extension phase (week 24 to week 56)

All patients who continued into the OLE phase received onabotulinumtoxinA injections at week 24, week 36, and week 48. The same doses and regimens were used as described in the DB phase.

c. Concomitant medications

Patients were allowed to take acute headache pain medications, which were recorded using a daily diary. Any headache prophylactic medication was prohibited during the trial and within 28 days prior to baseline of Studies 079 and 080 (week −4 to day 0).

3.2.4. Outcomes

a. Efficacy

Study 079

Primary efficacy end point: In Study 079, the primary efficacy end point was the frequency of headache episodes per 28-day period ending with week 24 compared with baseline. The 28-day period ending with week 24 was defined as day 57 to day 84 following the second injection. The number of headache episodes during the 28-day run-in period prior to randomization served as the “baseline.”

Secondary efficacy end points: The secondary efficacy outcomes were frequency of headache days per 28-day period, the frequency of migraine/probable migraine headache episodes per 28-day period, the frequency of migraine/probable migraine days per 28-day period, and the frequency of acute headache pain medication intakes per 28-day period.

Other: The following post hoc efficacy analyses were undertaken: frequency of moderate/severe headache days, total cumulative hours of headache occurring on headache days, and proportion of patients with severe HIT-6 category scores.

Study 080

Primary efficacy end point: In the first study protocol of Study 080, the efficacy end points were the same as Study 079; however, two weeks prior to the time at which the database was locked and treatment unblinded, the protocol was amended. The primary efficacy end point was changed from the frequency of headache episodes per 28-day period to the frequency of headache days per 28-day period, ending with week 24 compared with baseline.

Secondary efficacy end points: With the protocol change, the secondary outcomes became frequency of migraine/probable migraine days per 28-day period, frequency of moderate/severe headache days per 28-day period, total cumulative hours of headache occurring on headache days per 28-day period, proportion of patients with severe HIT-6 impact category score per 28-day period, and frequency of headache episodes per 28-day period.

In both studies, data required for the evaluation of all the headache characteristics and the use of acute headache pain medications were derived from self-reported diaries, in which efficacy measures were recorded daily by patients for the duration of the study using a validated electronic telephone diary. The start/stop times of each headache, headache-specific characteristics, symptoms associated with headache, the effect of physical activity on each headache, and use of acute headache pain medication were reported by patients on a daily basis using the electronic telephone diary.

Definitions of efficacy outcomes used in Studies 079 and 080:

  • Headache episode: patient-reported headache pain that lasted at least four continuous hours per patient diary.
  • Headache day: a day (00:00 to 23:59) with four or more continuous hours of headache.
  • Migraine episode: pain that lasted at least four continuous hours and that met the ICHD-2 criteria of migraine without aura or migraine with aura.
  • Probable migraine episode: pain that lasted at least four continuous hours and that met the ICHD-2 criteria of probable migraine.
  • Migraine/probable migraine days: a day (00:00 to 23:59) with four or more continuous hours of migraine or probable migraine headache.
  • Acute headache pain medication use: a patient-reported intake of medication(s) to treat headache pain where a patient reported that they took medication, regardless of the dose or number of types of medication taken at the same time. It was possible to have multiple intakes within a given day for each patient.
  • Acute headache pain medication overuse: ≥ 15 days per four-week period and at least two days per week for any simple analgesic intake, or ≥ 10 days per four-week period and at least two days per week for intake within a category for at least one category among ergotamines, triptans, opioids, and combination analgesic medications, or for such intake combined across at least two categories among ergotamines, triptans, analgesics (including simple and combination analgesic medication), and opioids.
  • Moderate/severe headache day: a day (00:00 to 23:59) with four or more continuous hours of headache and a maximum severity of moderate or severe as per patient diary.

Of note, some of the efficacy analyses mentioned above were not defined a priori as parameters to be measured in the protocol. Specifically, in Studies 079 and 080, the efficacy analyses assessed per 28-day period, but not defined a priori, included: 25%, 50%, 75%, and 100% decrease from baseline in the frequency of headache days, headache episodes, migraine/probable migraine headache days, migraine/probable migraine headache episodes, and daily headache impact score. Daily headache impact scores were collected from electronic patient diaries and derived from daily patient-reported assessments of headache impact. The scores ranged from 1 to 5 with 1 = no impact, 2 = little impact, 3 = moderate impact, 4 = severe impact, and 5 = unable to do anything. However, this instrument has not been validated in patients with CM, and no MCID is available. Furthermore, in Study 080, the frequency of migraine/probable migraine episodes and the frequency of acute headache pain medication intakes per 28-day period were not defined a priori.

MSQ is a disease-specific instrument which assesses the impact of migraine on a patient’s HRQoL. MSQ version 2.1 is a 14-item instrument measuring the following three domains: role function–restrictive (RR, seven items assessing how migraines limit one’s daily social and work-related activities), role function–preventive (RP, four items assessing how migraines prevent these activities), and emotional function (EF, three items assessing the emotions associated with migraines). For each domain, scores range from 0 (high function) to 100 (low function). MCIDs for the within-group differences for MSQ-RR, MSQ-RP, and MSQ-EF were estimated to be 10.9, 8.3, and 12.2, respectively. The MSQ was completed by patients at day 0, at week 12, at week 24, and at week 56. Please see APPENDIX 5: VALIDITY OF OUTCOME MEASURES for additional information regarding the description and validation of the MSQ.

HIT-6 comprises six items that measure pain, social functioning, role functioning, vitality, cognitive functioning, and psychological distress.22 Total HIT-6 scores range from 36 to 78. A total score of ≤ 49 indicates little or no impact, 50 to 55 indicates some impact, 56 to 59 indicates substantial impact, and ≥ 60 indicates severe impact of the disease on the daily life of the respondent. A between-group difference of HIT change scores of 2.3 units suggests a clinical improvement in a patient’s headache condition. HIT-6 was completed by patients at all scheduled DB visits beginning on day 0, as well as at all office visits in the OLE phase. Please see APPENDIX 5: VALIDITY OF OUTCOME MEASURES for additional information regarding the description and validation of the HIT-6 instrument.

In addition the generic questionnaire Euro-QoL Visual Analogue Scale (EQ-VAS) was completed by patients at day 0 and week 24. This instrument is not disease specific and no MCID for CM is available.

Harms: Safety data were presented for each included study through week 24 (DB treatment phase). Long-term safety and tolerability data are presented from the OLE phase in APPENDIX 6: SUMMARY OF OPEN-LABEL EXTENSION PHASE OF STUDY 079 AND STUDY 080.

3.2.5. Statistical analysis

a. Study 079

The primary efficacy analysis was change from baseline in the frequency of headache episodes per 28-day period ending with week 24 in the intention-to-treat (ITT) analysis set, as defined in the Analysis Populations section below. An analysis of covariance (ANCOVA) of the change from baseline was conducted, with treatment and medication-overuse strata as the main effects and the baseline frequency of headache episodes as covariate, using modified last observation carried forward (mLOCF [the mean across all patients in the previous 28-day period]). The analyses of secondary outcomes (frequency of headache days per 28-day period, frequency of migraine/probable migraine days per 28-day period, frequency of migraine/probable migraine episodes per 28-day period, and frequency of acute headache pain medication intakes per 28-day period) were conducted in a similar manner as the primary outcome using ANCOVA of the change from baseline in the ITT analysis set, with treatment and medication-overuse strata as the main effects and the baseline count for each of the outcomes as covariate, using mLOCF.

For the primary outcome, a two-sided test with P value less than or equal to 0.05 was considered as statistically significant; treatment-by-subgroup interactions were examined at the 0.10 level. In order to adjust the pre-specified type 1 error rate of 0.05 for all five pre-specified primary or secondary outcomes, a Bonferroni adjustment was applied post hoc to compare P values at a critical level of 0.01.

Study 080

The primary efficacy analysis was change from baseline in the frequency of headache days per 28-day period ending with week 24 in the ITT analysis set, as defined in the Analysis Populations section below. An ANCOVA of the change from baseline was conducted, with treatment and medication-overuse strata as the main effects and the baseline frequency of headache days as covariate, using mLOCF. The analyses of secondary outcomes (frequency of migraine/probable migraine days, frequency of moderate/severe headache days, total cumulative hours of headache occurring on headache days, and frequency of headache episodes) were conducted in a similar manner as the primary outcome using ANCOVA of the change from baseline in the ITT analysis set, with treatment and medication-overuse strata as the main effects and the baseline count for each of the outcomes as covariate, using mLOCF. Proportions of patients with severe HIT-6 total scores were compared between-treatment groups by Pearson’s chi-square tests.

For the primary outcome, a two-sided test with P value less than or equal to 0.05 was considered as statistically significant; treatment-by-subgroup interactions were examined at the 0.10 level. In order to control the type 1 error rate for multiple secondary end points, a gatekeeping approach for five ranked secondary outcomes was used, in which the test of any lower-ranked secondary end point was not considered statistically significant if the P value of a higher ranked secondary end point was > 0.05. The hierarchical order of the secondary end points was as follows: frequency of migraine/probable migraine days, frequency of moderate/severe headache days, total cumulative hours of headache occurring on headache days, proportion of patients with severe HIT-6 category scores, and frequency of headache episodes.

In both Studies 079 and 080, in order to analyze other efficacy outcomes, the Wilcoxon rank-sum test was used in the comparisons of the change from baseline between-treatment groups for outcomes with ordered response categories and for continuous outcomes. ANCOVA with a baseline covariate included in the analysis was used for outcomes with a continuous response range, and Pearson’s chi-square tests or Fisher’s exact test were used for the comparisons between-treatment groups for outcomes where the data were essentially binomial.

In both Studies 079 and 080, primary and secondary efficacy outcomes were summarized using the following subgroup factors:

  • investigator centre
  • age (< 40/> 40 years)
  • gender (male/female)
  • race (Caucasian/non-Caucasian)
  • acute headache pain medication overuse (yes/no)
  • patients who have failed three or more prior oral prophylactic medications (post hoc analysis)

The CDR protocol included a subgroup by duration of illness; however, such analysis was not undertaken by the manufacturer.

For primary and secondary efficacy outcomes, in order to have data for all patients in the ITT population for the ITT analysis, missing values were estimated according to the following method:

  • If at least 20 but less than 28 days of data were reported, data were prorated to a 28-day period equivalence.
  • If less than 20 days of a 28-day period data were reported, then the scores were imputed using a mLOCF imputation analysis, which involved the mean across all patients in the previous 28-day period.

a. Analysis populations

The primary analysis population used for the efficacy analyses in both studies was the ITT analysis set. The various analyses populations are defined below:

  • Safety Analysis Set: safety analyses were performed using the safety population, consisting of all patients who received the study treatment at day 0. For the DB phase, the patients were analyzed for safety as treated at the first treatment.
  • ITT Analysis Set: Patients were analyzed according to the randomization assignment, regardless of actual treatment received. In order to have data for all patients in the ITT population for the ITT analysis, missing values were imputed for the ITT analysis, as described above.

3.3. Patient Disposition

Of the 1,713 patients screened for Study 079, a total of 679 patients were randomized and 674 received at least one dose of the study drug. Five randomized patients did not receive any study drug (four in the placebo treatment group and one in the onabotulinumtoxinA treatment group). In Study 080, 1,621 patients were screened, of which 705 were randomized; all of them received at least one dose of the study drug. The primary reason for screening failure in both studies was the failure to meet all of the inclusion/exclusion criteria, especially the criterion regarding the minimum number headache days during the pre-randomization phase.

Across both studies, discontinuations ranged from 6.7% to 13.2% in individual treatment arms as detailed in Table 7. Overall, the primary reasons for discontinuation were: AEs (highest proportions in the onabotulinumtoxinA treatment groups of Study 079 [3.2%] and Study 080 [2.3%]); lost to follow-up (highest proportions in the placebo treatment groups of Study 079 [4.4%] and Study 080 [2.2%]); personal reasons (highest proportions in both treatment groups of Study 079); and “other” (it was unclear how “other” was defined).

Table 7. Patient Disposition.

Table 7

Patient Disposition.

3.4. Exposure to Study Treatments

The mean (median) total dose of onabotulinumtoxinA received by patients in the active group at day 0 and at week 12 ranged from 165.1 U (155.0 U) to 165.8 U (155.0 U) in Study 079, and from 163.0 U (155.0 U) to 164.3 U (155.0 U) in Study 080. In both treatment groups, the mean number of injection sites was approximately 33 in Study 079 and 32.5 in Study 080. Table 8 summarizes the exposure by randomized treatment for both included trials.

Table 8. Exposure to Study Treatments (Safety Population).

Table 8

Exposure to Study Treatments (Safety Population).

3.5. Critical Appraisal

3.5.1. Internal validity

a. Study Design

Statistical significance should be viewed with caution due to inflated type 1 error rates and different alpha levels required to claim statistical significance, where a P value less than 0.05 would not indicate statistical significance. In Study 079, in order to adjust the pre-specified type 1 error rate of 0.05 for all pre-specified primary or secondary outcomes, a Bonferroni adjustment was applied post hoc to compare P values at a critical level of 0.01. This adjustment did not take into consideration all the outcomes measured, including HRQoL measures. In Study 080, in order to control the type 1 error rate for multiple secondary end points, a gatekeeping approach for five ranked secondary outcomes was used, in which the test of any lower-ranked secondary end point was not considered statistically significant if the P value of a higher ranked secondary end point was > 0.05. The type 1 error was not adjusted for HRQoL outcomes and subgroup analyses. The problem with this approach is that only certain outcomes were selected; hence, the gatekeeping approach did not take into consideration all outcomes measured, including HRQoL measures. In addition, no criteria were stated on how the outcomes were ranked, and the ranking used was different than the ranking stated in the protocol of this review. This indicates selective reporting.

In Study 080, two weeks before the primary database lock and treatment unblinding, the initial pre-specified primary end point was changed from being “frequency of headache episodes” to “frequency of headache days.” These changes were driven by the results from Study 079; however, it is unlikely these changes affected the analysis of the study.

Even though the study was blinded, blinding could have been broken. Because of AEs (such as paralysis of the forehead muscles), patients may have guessed that they were receiving onabotulinumtoxinA. The guidelines for controlled trials of prophylactic treatment of CM in adults mention that “unblinding may be a significant factor in acute and prophylactic placebo-controlled migraine trials. Subjects and investigators should be questioned at the end of the trial regarding their opinion as to what treatment group (active or placebo) the subject was assigned to during the study.”5 In Studies 079 and 080, patients were not asked at any time if they could determine whether they were on onabotulinumtoxinA or placebo treatment. In addition, in two phase 2 studies (191622-038 and 191622-039), approximately 86% of patients correctly determined that they were receiving onabotulinumtoxinA and 58% correctly guessed they were receiving placebo.23 In Studies 079 and 080, it may be that the efficacy of onabotulinumtoxinA was overestimated; this would affect the results of patient-reported outcomes and the more subjective outcomes such as days of headache/migraine and episodes of headache/migraine. Furthermore, investigators may also have been able to determine the allocated treatment group.

To compare treatment effects in subgroups in a randomized controlled trial (RCT), a test of interaction should be used. If the result of the interaction test is not significant, then there is no observable subgroup effect; however, such analysis was not reported by the manufacturer.

b. Population

In Study 079, there was a baseline imbalance between the two treatment groups in terms of headache and migraine episodes. These were higher in the placebo group than the onabotulinumtoxinA group. Furthermore, the total cumulative hours of headache was higher in the onabotulinumtoxinA group than in the placebo group. This imbalance between the two groups may have explained why the effect sizes obtained in Study 079 were not as great as those obtained in Study 080.

It was reported in both studies that the minimum age of onset of CM was one year old, putting into question the internal validity of the data or the diagnosis. If there was an error in data entry, then one would question why it was not fixed. Furthermore, a diagnosis of CM at one year of age would be highly improbable.

c. Interventions

No standardized approach was used to determine which patients would require a dose > 155 U. At the investigator’s discretion, the dose could be increased by up to an additional 40 U using a “follow-the-pain” method. In addition, the study was not designed to determine the clinical benefit of treatment at a dose higher than 155 U.

Both trials compared onabotulinumtoxinA versus placebo. Despite the lack of other approved prophylactic medications for CM in Canada, there are medically accepted therapies such as propranolol, amitriptyline, and topiramate, which are used off-label. A comparative trial against one of these agents would have been clinically relevant. In fact, several onabotulinumtoxinA trials have been conducted against medically accepted prophylactic medications. Head-to-head trials have compared onabotulinumtoxinA to topiramate,24,25 to divalproex sodium,26 and to amitriptyline for the prevention of CM.27 All trials used a CM dosing regimen of onabotulinumtoxinA that is not approved by HC.

d. Outcomes

Patients were withdrawn from prophylaxis medications four weeks prior to the trial run-in phase. Improvement in outcomes from baseline may have been exaggerated (i.e., less well controlled at baseline).

A definition of what constituted a pain-free interval between two headache or migraine episodes was not provided, and hence, the reliability of frequency of episodes as an outcome may be questioned.

Headache/migraine episodes and days were derived from patient diaries; however, self-reporting is subject to individual variability in reporting accuracy and completion. Furthermore, no descriptions of the electronic telephone diary and its validation were provided.

A patient with an incomplete response to acute headache pain medication taken for a headache, or a patient getting temporary relief from acute medication, could count one headache episode as two headache episodes.

The subgroup analysis for patients who failed three or more prior prophylaxis medications was post hoc. No interaction test for this subgroup was undertaken. The result of this analysis must be interpreted with caution. In addition, a subgroup analysis for patients who did not fail three or more prior prophylaxis medications was not provided. Hence, a conclusion about these two subgroups cannot be made.

The reduction in acute medication use could potentially confound the interpretation of the results, as it may lead to a reduction in headache frequency/severity if the headaches are due to medication overuse. Approximately 65% of patients included in the study were overusing acute headache pain medication; a change in acute medication use might be reflected as a change in headache status, especially if patients had a diagnosis of MOH instead of CM.

When patients reported that they used acute headache pain medication, that use was measured as “the intake of medication(s) to treat headache pain”, regardless of the dose or the number of types of medication taken at the same time. This method did not accurately measure the use of acute headache pain medication.

The use of mLOCF imputation might have biased the results. Specifically, the mLOCF method of imputation involved calculating the mean result for the outcome of interest across all patients in the previous 28-day period. Since pain could change from day to day, imputation using the previous 28-day period might not be appropriate. In addition, imputations at week 24 were performed in about 20% of the patients in both Study 079 and 080.19 But when sensitivity analysis using “observed data” without imputing for missing values was performed, there was no change in the direction of results, indicating that the mLOCF was appropriate.

The proportion of patients who withdrew from the trials was approximately 13% for Study 079 and 9% for Study 080. One reason given was classified as “other,” which was undefined. The Food and Drug Administration (FDA) clinical reviewer report identified that, in Study 079, five patients in the onabotulinumtoxinA treatment arm and four patients in the placebo treatment arm discontinued treatment due to a lack of efficacy, yet eight of these patients were reported in the “other” category in the CSR.21 If the “other” category included withdrawals due to lack of efficacy, then the reasons for withdrawals have not been reported adequately in the studies. Furthermore, the number of patients included in the observed data analyses was approximately 10% lower than the number of patients who completed the DB phase. This indicates that approximately 10% of patients did not complete their diaries. This may have introduced bias, which may have affected the internal validity of both trials.

3.5.2. External validity

Studies 079 and 080 were designed when ICHD-2 criteria were still in use and prior to the publication of ICHD-3. ICHD-2 used a strict definition of migraine (migraine occurring on 15 or more days per month for more than three months), whereas the ICHD-3 describes CM as a headache (tension-type-like or migraine-like) occurring on 15 or more days per month for more than three months, and which has the features of migraine headaches on at least eight days per month. This latter definition of CM was meant to better reflect the population of patients seen in clinical practice. Although patients with tension-type headaches were specifically excluded from Studies 079 and 08, the inclusion criteria used in both studies were more in line with ICHD-3.

The ICHD-3 excludes medication-overuse headache from the diagnosis of CM. It is believed that approximately 50% of patients are wrongly diagnosed with CM while they should be diagnosed with MOH.1 However, neither study excluded patients with MOH. Both studies stratified patients according to history of acute medication overuse to account for this, but some patients could still have been misclassified as CM patients when in fact they should have been diagnosed with MOH. As per the clinical expert involved in the review, in clinical practice there is considerable overlap between MOH and CM, making it very difficult to distinguish between them. Thus, the trials examined likely reflect the real-world situation.

The exclusion of patients who were on therapy that is also used as prophylactic headache treatment (for example beta blockers used for hypertension and antidepressants used for depression) may have excluded CM patients with comorbid illnesses; hence, the efficacy of onabotulinumtoxinA has not been explored in such a subgroup of patients.

Patients continued to receive onabotulinumtoxinA even after the number of headache days decreased to < 15 days, which would change the diagnosis of CM to EM. It is not clear if the reduction in headache days would be sustained if patients were to stop receiving onabotulinumtoxinA. Previous onabotulinumtoxinA trials conducted in patients with EM showed limited benefits.10,28

3.6. Efficacy

Only those efficacy outcomes identified in the review protocol are reported below (Section 2.2, Table 3). See APPENDIX 4: DETAILED OUTCOME DATA for detailed efficacy data.

3.6.1. Health-related quality of life

a. MSQ

Change in HRQoL from baseline to week 12 and to the primary end point (week 24) was measured using the MSQ (see APPENDIX 5: VALIDITY OF OUTCOME MEASURES). This end point was not listed within the pre-specified primary and secondary outcomes, and it is unclear if an adjustment for type 1 error for this outcome was made.

In Study 079, at week 24, patients treated with onabotulinumtoxinA had a greater decrease from baseline in mean scores for the three MSQ domains than patients treated with placebo. Changes from baseline for RR scores were −16.8 for onabotulinumtoxinA versus −8.8 for placebo (P < 0.001 for between-group difference). Changes from baseline for RP scores were −12.6 for onabotulinumtoxinA versus −7.6 for placebo (P = 0.005 for between-group difference). Changes from baseline for EF scores were −16.9 for onabotulinumtoxinA versus −10.0 for placebo (P = 0.001 for between-group difference).

Similar results were found in Study 080, where changes from baseline were −17.2 for onabotulinumtoxinA versus −8.4 for placebo in RR scores (P < 0.001 for between-group difference), −13.5 for onabotulinumtoxinA versus −5.4 for placebo in RP scores (P < 0.001 for between-group difference), and −19.0 for onabotulinumtoxinA versus −9.1 for placebo in EF scores (P < 0.001 for between-group difference).

In both studies, patients who received onabotulinumtoxinA treatment had a mean change from baseline scores that exceeded the established MCID within-group differences of −10.9 (RR), −8.3 (RP) and −12.2 (EF), while patients receiving placebo did not exceed the respective MCIDs of 10.9, 8.3, and 12.2 points. Results were presented for observed data without imputation for missing values (Table 9 and Table 14).

Table 9. Baseline and Mean Change From Baseline at Week 24 in MSQ Scores (Observed Data Without Imputation for Missing Values).

Table 9

Baseline and Mean Change From Baseline at Week 24 in MSQ Scores (Observed Data Without Imputation for Missing Values).

b. EQ-VAS

There were no statistically significant between-group differences in EQ-VAS at week 24.

Subgroup Analyses

Subpopulation: Patients with Three or More Prior Prophylactics: In Study 079 at week 24, patients treated with onabotulinumtoxinA had a greater decrease in mean MSQ scores for two domains (RR with P = 0.010 and EF with P = 0.024 for between-group differences) from baseline than patients treated with placebo. The RP domain score change was not statistically significantly different between onabotulinumtoxinA and placebo. Patients who received onabotulinumtoxinA treatment had a change from baseline scores that exceeded the established MCID for the RR and EF domains of MSQ.

In Study 080, patients treated with onabotulinumtoxinA had a greater reduction in mean scores for all three MSQ domains (RR with P < 0.001, RP with P < 0.001, and EF with P = 0.007 for between-group differences) from baseline than patients treated with placebo. Patients who received onabotulinumtoxinA treatment had a change from baseline MSQ scores that exceeded the established MCID for all three domains (RR, RP, and EF). Detailed results are presented in Table 15.

Other Subpopulations: In both studies for male patients, there was no statistically significant difference between onabotulinumtoxinA and placebo for all MSQ domains. Patients in both treatment groups had a change from baseline scores that exceeded the established MCID for some domains. Results for female patients were consistent with the results obtained overall (Table 16).

In both studies, the onabotulinumtoxinA subgroup of patients who overused acute headache medication had a change from baseline scores that exceeded the established MCID for all domains, while for those in the placebo group, the change did not exceeded the established MCID for any of the three domains. Similarly, in the subgroup of patients who did not overuse acute headache medication, patients in placebo group achieved the established MCID except for the RP domain in Study 079 and RR domain in Study 080 (Table 17).

3.6.2. Other Patient-Reported Outcomes

In Study 080, the proportion of patients with a “severe” HIT-6 category score (≥ 60) was listed as a secondary outcome, and statistical adjustments for type 1 error were made. Other HIT-6 outcomes were not listed within the pre-specified primary and secondary outcomes.

Mean changes from baseline in total HIT-6 score favoured onabotulinumtoxinA over placebo with P < 0.001 for between-group differences in both studies. The between-group difference at week 24 was 2.3 in Study 079 and 2.5 in Study 080. This between-group difference met the MCID of 2.3 (Table 10).

Table 10. Change From Baseline at Week 24 in HIT-6 (MLOCF).

Table 10

Change From Baseline at Week 24 in HIT-6 (MLOCF).

When the four groupings of HIT-6 (little or no impact, some impact, substantial impact, or severe impact) were compared at week 24, results also favoured onabotulinumtoxinA over placebo, with P < 0.001 in both studies. Detailed results are presented in Table 18. In addition, Figure 3 and Figure 4 present, respectively, the proportion of patients with a “severe” HIT-6 score (≥ 60) and the mean change from baseline in HIT-6 scores at different time points. These figures showed that benefit was obtained after the first treatment, with little additional benefit after the second treatment at week 12.

Patients treated with onabotulinumtoxinA had a greater mean decrease from baseline in daily headache impact score than patients treated with placebo at week 24, with P = 0.002 and P < 0.001 in Study 079 and Study 080, respectively. No MCID in CM is available for this instrument. Detailed results are presented in Table 18.

a. Subgroup Analyses

Subpopulations: Patients With Three or More Prior Prophylactics

In Study 079, the difference in the mean changes from baseline in total HIT-6 scores between the onabotulinumtoxinA group and the placebo group did not meet the established MCID of 2.3 at week 24, while in Study 080 the difference was 2.9, exceeding the established MCID. Detailed results are presented in Table 19.

Other Subpopulations

In both studies, the difference in the mean changes from baseline in total HIT-6 scores between the onabotulinumtoxinA group and the placebo group did not meet the established MCID of 2.3 for male patients at week 24, but the MCID was exceeded for female patients. Detailed results are presented in Table 20.

In both studies, among patients coded as medication overuse “Yes” or “No” the difference in the mean changes from baseline in total HIT-6 scores between the onabotulinumtoxinA group and the placebo group met or exceeded the MCID (Table 21).

Patients treated in the onabotulinumtoxinA group had a greater mean decrease from baseline in daily headache impact score than patients treated with placebo at week 24 for the subpopulations of patients with three or more prior prophylactic, females, patients who overused and did not overuse acute headache pain medication, while for the subpopulation of male patients the difference was not statistically significant in both studies (Table 22).

In both studies, the difference in the mean changes from baseline in total HIT-6 scores between the onabotulinumtoxinA group and the placebo group met the established MCID of 2.3 for the subgroup of patients who overused acute headache medication, while in the subgroup of patients who did not overuse acute headache medication, the difference exceeded the established MCID in Study 080 only (Table 23).

3.6.3. Acute headache pain medication intake

Acute headache pain medication intake was pre-specified as a secondary efficacy outcome in Study 079, while it was not pre-specified as a primary or a secondary outcome in Study 080. At week 24, in Study 079, in both the onabotulinumtoxinA and placebo groups, the mean decrease from baseline in the frequency of acute headache pain medication intakes per 28-day period was −10.1 intakes for the onabotulinumtoxinA group versus −9.8 intakes for the placebo group. The mean decrease from baseline in the frequency of acute headache pain medication days per 28-day period was −5.8 days for the onabotulinumtoxinA group versus −5.8 days for placebo group. In addition, there was an overall reduction in the use and overuse of acute headache pain medications. None of the aforementioned between group comparisons was statistically significant, however.

In Study 080, in both the onabotulinumtoxinA and placebo groups, the mean decrease from baseline in the frequency of acute headache pain medication intakes per 28-day period was −9.7 intakes for the onabotulinumtoxinA group versus −8.1 intakes for placebo group (P = not statistically significant). The mean decrease from baseline in acute headache pain medication days per 28-day period was −6.4 days for the onabotulinumtoxinA group versus −4.8 days for placebo group (P < 0.001). In addition, in both the onabotulinumtoxinA and placebo groups, there was an overall reduction in the use of acute headache pain medications, with a larger reduction in the overuse of acute headache pain medications in the onabotulinumtoxinA group than in the placebo group by approximately 9%, with P = 0.017.

Detailed results are presented in Table 11, Figure 5, Figure 6, Figure 7, and Figure 8. Figure 5 to 8 present results of the mean change from baseline in acute headache pain medication intake, mean change from baseline in frequency of acute headache pain medications days, proportion of patients who used acute headache pain medications, and proportion of patients who overused acute headache pain medications per 28-day period at different time points, respectively. These figures indicate that improvements were observed after the first treatment, with smaller improvements noticed after the second treatment.

Table 11. Acute Headache Pain Medication Intake.

Table 11

Acute Headache Pain Medication Intake.

Subgroup Analyses

In both studies, in both the onabotulinumtoxinA and placebo groups, there was a decrease from baseline in the frequency of acute headache pain medication intakes for all subgroups of interest: patients who overused and did not overuse headache pain medication, males, females, and patients who had used three or more prior prophylactics. The effect size was greater for the subgroup of patients who overused headache pain medication compared with those who did not, and in females compared with males.

The between-group difference for the mean change from baseline in acute headache pain medication intake was statistically significant in Study 080 in the subgroup of patients who had used three or more prior prophylactics. All other between-group comparisons were not statistically significant. Detailed results are presented in Table 23.

3.6.4. Other efficacy outcomes

Reduction from baseline in headache days and migraine/probable migraine days per 28-day period, frequency of headache days and migraine/probable migraine days per 28-day period, number of moderate/severe headache days per 28-day period, total cumulative hours of headache occurring on headache days per 28-day period, reduction from baseline in headache episodes and migraine/probable migraine episodes per 28-day period, frequency of headache episodes and migraine/probable migraine episodes per 28-day period, emergency room (ER) visits and hospitalizations for migraine symptoms, and time lost due to migraine were reported in both studies.

a. Per cent reductions in headache days per 28-day period

In both studies, the reduction in the number of headache days per 28-day period was assessed using observed data (without imputation for missing values). In both studies, a greater proportion of patients in the onabotulinumtoxinA group had a 25% reduction, a 50% reduction, and a 75% reduction in headache days per 28-day period compared with the placebo group. The proportion of patients with a 100% reduction in headache days was similar in both treatment groups. Results are presented in Table 24. Figure 9 and Figure 10 present the proportion of patients with 25%, 50%, 75%, and 100% reduction in headache days per 28-day period at different time points in Studies 079 and 080, respectively.

Frequency of headache days per 28-day period

This was the primary end point for Study 080 (after the protocol change prior to unblinding) and a secondary end point for Study 079. The change in number of headache days per 28-day period was assessed using the ITT population, with missing values imputed using mLOCF. At week 24, patients treated with onabotulinumtoxinA had a greater decrease from baseline in frequency of headache days per 28-day period (LSMean = −7.8 in Study 079 and −9.2 in Study 080) than patients treated with placebo (LSMean = −6.4 in Study 079 and −6.9 in Study 080). P values for between-group differences were 0.006 and < 0.001 for Study 079 and Study 080, respectively. Detailed results are presented in Table 24 and Figure 13.

Results for subgroups of interest are presented in Table 25. At week 24, the between-group difference in the mean change from baseline in the frequency of headache days per 28-day period favoured onabotulinumtoxinA for all subgroups, including patients who overused headache medications and those who did not, males (except Study 079 where placebo was better), females, and patients who had used three or more prior prophylactics.

Per cent reductions in migraine/probable migraine days per 28-day period

In both studies, the reduction in the number of migraine/probable migraine days per 28-day period was assessed using observed data (without imputation for missing values). In both studies, a greater proportion of patients in onabotulinumtoxinA group had a 25% reduction, a 50% reduction, and a 75% reduction in migraine/probable migraine days per 28-day period compared with the placebo group. The proportion of patients with a 100% reduction in migraine/probable migraine days was approximately similar between both treatment groups at week 24. Results are presented in Table 24. Figure 11 and Figure 12 present the proportion of patients with 25%, 50, 75%, and 100% reduction in migraine/probable migraine days per 28-day period at different time points in Studies 079 and 080, respectively.

Frequency of migraine/probable migraine days per 28-day period

This was a secondary end point in both studies. The change in the number of migraine/probable migraine days per 28-day period was assessed using the ITT population with missing values imputed using mLOCF. At week 24, patients treated with onabotulinumtoxinA had a greater decrease from baseline in frequency of migraine/probable migraine days per 28-day period (LSMean = −7.6 in Study 079 and −8.8 in Study 080) than patients treated with placebo (LSMean = −6.0 in Study 079 and −6.5 in Study 080). P values for between-treatment group differences were 0.006 and < 0.001 for Study 079 and Study 080 respectively. Detailed results are presented in Table 24 and Figure 16.

Results for the subgroups of interest are presented in Table 26. At week 24, the between-group difference in the mean change from baseline in the frequency of migraine/probable migraine days per 28-day period favoured onabotulinumtoxinA for all subgroups, including patients who overused headache medications and those who did not, males (except Study 079 where both treatment groups were similar), females, and patients who had used three or more prior prophylactics.

Frequency of moderate/severe headache days per 28-day period

This was a secondary end point in Study 080 and a post hoc analysis in Study 079. The change in number of moderate/severe headache days per 28-day period was assessed using the ITT population, with missing values imputed using mLOCF. At week 24, the between-group difference in the mean change from baseline in frequency moderate/severe headache days per 28-day period was −1.5 days in Study 079 and −2.4 days in Study 080 (P < 0.001), with fewer moderate/severe headache days with onabotulinumtoxinA. Detailed results are presented in Table 24 and Figure 14.

Results for the subgroups of patients with three or more prior prophylactics are presented in Table 25. At week 24, the between-group difference in the mean change from baseline in frequency moderate/severe headache days per 28-day period was −1.9 days in Study 079 and −3.8 days in Study 080, with fewer moderate/severe headache days with onabotulinumtoxinA.

Total Cumulative Hours of Headache Occurring on Headache Days per 28-day Period

This was a secondary end point in Study 080 and a post hoc analysis in Study 079. The change in total cumulative hours of headache occurring on headache days per 28-day period was assessed using the ITT population with missing values imputed using mLOCF. At week 24, the between-group difference in the mean change from baseline in the total cumulative hours of headache was approximately −30 hours in Study 079 and approximately −40 hours in Study 080 (P < 0.001), with fewer total cumulative hours of headache with onabotulinumtoxinA. Detailed results are presented in Table 24 and Figure 15.

Results for the subgroups of patients who had used three or more prior prophylactics are presented in Table 25. At week 24, the between-group difference in the mean change in total cumulative hours of headache occurring on headache days per 28-day period was approximately −52 hours in Study 079 and approximately −58 hours in Study 080, with fewer total cumulative hours of headache with onabotulinumtoxinA.

Frequency of headache episodes per 28-day period

This was the primary end point for Study 079 and a secondary end point in Study 080. The change in number of headache episodes per 28-day period was assessed using the ITT population with missing values imputed using mLOCF. The between-group difference in the mean change from baseline in frequency of headache episodes per 28-day period at week 24 was −0.4 episodes in Study 079 (P = NS) and −1.0 episodes in Study 080 (P = 0.003), with fewer headache episodes with onabotulinumtoxinA. Detailed results are presented in Table 27.

Results for the subgroups of interest are presented in Table 28. The between-group difference in the mean change from baseline in frequency of headache episodes per 28-day period at week 24 was less than one point for all subgroups. There were statistically significant differences in favour of the onabotulinumtoxinA group in the subgroup of patients who overused headache medication, female patients, and patients who had used three or more prior prophylactics in Study 080 only. There was no statistically significant difference at week 24 for any subgroup of patients in Study 079.

Frequency of migraine/probable migraine episodes per 28-day period

This was a secondary end point in Study 079 and was not defined a priori as a parameter to be measured in the protocol for Study 080. The change in the number of migraine/probable migraine episodes per 28-day period was assessed using the ITT population, with missing values imputed using mLOCF. The between-group difference in the mean change from baseline in frequency migraine/probable migraine episodes per 28-day period at week 24 was −0.5 episodes in Study 079 (P = NS) and −0.9 episodes in Study 080, with fewer migraine/probable migraine episodes with onabotulinumtoxinA. Detailed results are presented in Table 27.

Results for the subgroups of interest are presented in Table 28. At week 24, the between-group difference in the mean change from baseline in frequency migraine/probable migraine episodes per 28-day period was less than 2.0 points for all subgroups. There were statistically significant differences in favour of the onabotulinumtoxinA group in the subgroup of patients who overused headache medication and in the subgroup of patients who had used three or more prior prophylactics. In Study 079, the P value was less than 0.05 at week 24 for the subgroup of patients who overused headache medication.

Emergency room visits and hospitalizations due to migraine symptoms

The mean change in number of ER visits and hospitalizations due to migraine symptoms was assessed using observed data (Table 29). The baseline numbers of ER visits and hospitalizations were less than one episode in the previous three-month period. At the end of the DB phase, the within-group differences and the between-group differences were less than 0.2 episodes. The frequency of visits and stays is too low to interpret these results.

Work Status and Productivity

The mean changes in number of hours worked, the days of work missed, the days of reduced work productivity, and the proportion of patients not working due to migraine were assessed using observed data. In some of the analyses, data from US patients were used. Results are presented in Table 30. Between-group differences were similar except for Study 080, in which the proportion of patients not working due to migraine at week 24 was 4.6% for the onabotulinumtoxinA group and 9.2% for the placebo group.

3.7. Harms

Only those harms identified in the review protocol are reported below (see 2.2.1, Protocol).

3.7.1. Adverse events

The proportion of patients who experienced at least one AE was higher in the onabotulinumtoxinA group (59.7% in Study 079 and 65.1% in Study 080) than in the placebo group (46.7% in Study 079 and 56.4% in Study 080) Table 12.

Table 12. Harms.

Table 12

Harms.

Overall, the most frequent AE associated with onabotulinumtoxinA was neck pain (8.2% in Study 079 and 9.8% in Study 080). In comparison, neck pain was reported in 3.3% and 2.2% of placebo-treated patients in Studies 079 and 080, respectively. Muscular weakness (5.9% and 5.2%) and headache (4.4% and 4.6%) were the next most frequently reported AEs in the onabotulinumtoxinA-treated patients (compared with 0% and 0.6% in Study 079, and with 3% and 3.4% in Study 080 in placebo-treated patients).

Other AEs that occurred with higher frequency in the onabotulinumtoxinA-treated patients when compared with placebo were eyelid ptosis, injection site pain, musculoskeletal pain, muscle spasms, musculoskeletal stiffness, myalgia, and migraine.

3.7.2. Serious adverse events

The proportion of patients with at least one SAE was higher in the onabotulinumtoxinA group (5.3% in Study 079 and 4.3% in Study 080) than in the placebo group (2.4% in Study 079 and 2.2% in Study 080) Table 12. In Study 080, SAEs reported by two or more patients in any treatment group were pneumonia (0.6% in the onabotulinumtoxinA group versus 0.3% in the placebo group), breast cancer (0.6% in the onabotulinumtoxinA group versus 0% in the placebo group), and migraine (0.9% in onabotulinumtoxinA group versus 0.3% in the placebo group); in Study 079, uterine leiomyoma was the only SAE reported by two or more patients (0.6% of patients in onabotulinumtoxinA group versus none in the placebo group).

3.7.3. Withdrawals due to adverse events

WDAEs occurred in 4.1% and in 3.5% of onabotulinumtoxinA-treated patients compared with 0.9% and 1.4% of placebo-treated patients in Study 079 and Study 080, respectively (Table 12). The most frequent reasons for withdrawals were headache in Study 079 and migraine in Study 080.

3.7.4. Mortality

No deaths were reported in either trial.

3.7.5. Adverse events of special interest

Drug hypersensitivity was reported by 0.3% and 0.6% of onabotulinumtoxinA-treated patients compared with 0% and 0.3% of placebo-treated patients in Study 079 and Study 080, respectively. Dysphagia was reported by 0.9% and 0.6% in the onabotulinumtoxinA-treated patients versus 0.3% and 0% in the placebo group for Study 079 and 080, respectively. Palpitations was only reported in Study 080 by 0.6% of patients in the onabotulinumtoxinA group. No other AEs of interest identified in the protocol were reported.

Image sr0345f3
Image sr0345f4
Image sr0345f5
Image sr0345f6
Image sr0345f7
Image sr0345f8
Image sr0345f9
Image sr0345f10
Image sr0345f13
Image sr0345f11
Image sr0345f12
Image sr0345f16
Image sr0345f14
Image sr0345f15
Copyright © CADTH 2015.

You are permitted to make copies of this document for non-commercial purposes, provided it is not modified when reproduced and appropriate credit is given to CADTH. You may not otherwise copy, modify, translate, post on a website, store electronically, republish, or redistribute any material from this document in any form or by any means without the prior written permission of CADTH.

Please contact CADTH’s Vice-President of Corporate Services at ac.htdac@secivresetaroproc with any inquiries about this notice or other legal matters relating to CADTH’s services.

Except where otherwise noted, this work is distributed under the terms of a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International licence (CC BY-NC-ND), a copy of which is available at http://creativecommons.org/licenses/by-nc-nd/4.0/

Bookshelf ID: NBK344280
Contents

Views

  • PubReader
  • Print View
  • Cite this Page

In this Page

Other titles in this collection

Recent Activity

ClearTurn OffTurn On

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...