A systematic review of adverse events in placebo groups of anti-migraine clinical trials

PAINÒ 146 (2009) 261–269

www.elsevier.com/locate/pain

A systematic review of adverse events in placebo groups of anti-migraine clinical trials Martina Amanzio a,b,*, Luca Latini Corazzini a,b, Lene Vase c, Fabrizio Benedetti d,e a

Department of Psychology, University of Turin, Via Verdi 10, 10123 Turin, Italy Neuroscience Institute of Turin (NIT), University of Turin, Italy c Department of Psychology, University of Aarhus, Jens Chr. Skous Vej 4, DK – 8000 Århus C, Denmark d Department of Neuroscience, University of Turin Medical School, Corso Raffaello 30, 10125 Turin, Italy e National Institute of Neuroscience (INN), Turin, Italy b

a r t i c l e

i n f o

Article history: Received 24 March 2009 Received in revised form 4 June 2009 Accepted 13 July 2009

Keywords: Placebo Nocebo Adverse events Migraine Randomized clinical trials

a b s t r a c t In analgesic clinical trials, adverse events are reported for the painkiller under evaluation and compared with adverse events in the placebo group. Interestingly, patients who receive the placebo often report a high frequency of adverse events, but little is understood about the nature of these negative effects. In the present study, we compared the rates of adverse events reported in the placebo arms of clinical trials for three classes of anti-migraine drugs: NSAIDs, triptans and anticonvulsants. We identified 73 clinical trials in 69 studies describing adverse events in placebo groups: 8 were clinical trials with NSAIDs, 56 were trials with triptans, and 9 were trials with anticonvulsants. Studies were selected of all Medline/PubMed or CENTRAL referenced trials published until 2007. Adverse event profiles of the three classes were compared using a systematic review approach. We found that the rate of adverse events in the placebo arms of trials with anti-migraine drugs was high. In addition, and most interestingly, the adverse events in the placebo arms corresponded to those of the anti-migraine medication against which the placebo was compared. For example, anorexia and memory difficulties, which are typical adverse events of anticonvulsants, were present only in the placebo arm of these trials. These results suggest that the adverse events in placebo arms of clinical trials of anti-migraine medications depend on the adverse events of the active medication against which the placebo is compared. These findings are in accordance with the expectation theory of placebo and nocebo effects. Ó 2009 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.

1. Introduction Adverse events (AEs) in pharmacological treatments are important aspects to be considered in clinical decisions in order to both minimize the risk and maximize the benefit of drugs. It is widely known that different analgesics induce different side effects. For example, in the treatment of migraine, different pharmacological agents can be used, both in the prophylaxis and in the management of acute attacks, ranging from non-steroid anti-inflammatory drugs (NSAIDs) to triptans and anticonvulsants. These different classes of drugs induce different rates of adverse events. NSAIDs are associated with high rates of gastrointestinal symptoms, whereas anticonvulsants are known to induce more adverse events compared with the other two classes of drugs, as well as specific and unique side effects, such as anorexia [6,15,16]. In order to test both the efficacy of these compounds and their toxicity and tolerability, in clinical trials these medications are * Corresponding author. Address: Department of Psychology, University of Turin, Via Verdi 10, 10123 Turin, Italy. Tel.: +39 11 6702468; fax: +39 11 6702061. E-mail address: [email protected] (M. Amanzio).

compared with placebos. Primary–secondary therapeutic outcomes and the occurrence of adverse events are assessed in both the active medication arm and the placebo arm [18]. In a typical clinical trial, the subjects know they can receive either the active medication or the placebo and, accordingly, they are informed about the possible adverse events they may experience during the trial. Different studies on the placebo and nocebo effect have shown that expectations about the therapeutic outcome play a crucial role in the response to treatment [1,3,7,11,17]. This is true for both positive suggestions leading to positive outcomes (placebo responses) and negative suggestions leading to negative outcomes (nocebo responses). Therefore, informing subjects about the possible adverse events they may experience, may have a significant impact on their expectations and experiences of negative effects. On the basis of these considerations, we conducted an analysis of published randomized clinical trials on anti-migraine drugs that reported rates of adverse events in placebo groups. We hypothesized that the side effects reported in the placebo groups of clinical trials of the three classes of anti-migraine drugs (NSAIDs, triptans and anticonvulsants) may differ across the trials for specific adverse events. In particular, since anti-migraine drugs are expected

0304-3959/$36.00 Ó 2009 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.pain.2009.07.010

M. Amanzio et al. / PAINÒ 146 (2009) 261–269

262

to produce specific side effects, we postulated that this effect should also be seen in the corresponding placebo groups. We also hypothesized that this effect specifically depends on suggestions about adverse events and thereby on the patient’s and/or investigator’s expectations. The analysis also considered the role of other contributing factors such as sample size, race, gender, age, clinical characteristics of migraine, weight, withdrawal, ascertainment strategy (i.e., structured versus spontaneous adverse events report), Jadad score [10] and year of publication.

Potentially relevant articles identified and screened for retrieval (n =218)

Articles excluded as there were no description of the AEs in the active or in the placebo group (n=41)

Articles retrieved for more detailed evaluation (n=177)

2. Methods Articles excluded as they were conducted in relation to paediatric migraine (n=8)

2.1. Search procedure We conducted a search of randomized clinical trials published between 1988 and 2007 across the Medline database (PubMed) and the Cochrane Central Register of Controlled Trials (CENTRAL). All studies that included one of the following terms in the title were considered:

Articles considered for inclusion in the analysis (n= 169)

Articles excluded as they did not fulfil additional selection criteria * (n=20)

For triptans: ‘‘sumatriptan”; ‘‘almotriptan”; ‘‘elitriptan”; ‘‘zolmitriptan”; ‘‘rizatriptan”; ‘‘frovatriptan”; ‘‘naratriptan”. For NSAIDs: ‘‘rofecoxib”; ‘‘valdecoxib”; ‘‘diclofenac”; ‘‘ibuprofen”; ‘‘ketoprofen”; ‘‘indobufen”. For anticonvulsants: ‘‘topiramate”; ‘‘gabapentin”; ‘‘valproate”. We also searched published meta-analyses on anti-migraine trials to identify additional studies that were not included in the database. Finally, the references included in the published articles identified in these searches were used as an additional source to identify other clinical trials.

Total trials on AEs in placebo groups of antimigraine clinical trials included in the analysis (n=73) reported in 69 publications

* See appendix I (supplementary on-line material)

Fig. 1. Trial flow: selection of study reports.

2.2. Inclusion and exclusion criteria 2.3. Data abstraction and validity assessment Studies were selected for inclusion according to the following criteria: (1) Only randomized placebo-controlled clinical trials (RCT) of NSAIDs or anticonvulsants or triptans, which reported AEs, were included; (2) We excluded studies with insufficient reporting of side effects, i.e., either no reports of side effects at all, or combined reports of side effects for active and placebo groups together; (3) Crossover studies were excluded from the analysis. In fact, if placebo is given as the first treatment, one is measuring the effects of suggestion only, whereas if placebo is given as a second treatment one is measuring the effects of both suggestion and conditioning; (4) Studies comparing two classes of anti-migraine drugs in the same trial were excluded; (5) We also excluded children and studies in which migraine was associated with other neurological disorders; (6) We excluded studies using nasal spray medications and suppositories for migraine treatment because these routes of administration are not used for all the three classes of drugs; (7) Furthermore, we excluded some studies that suffered from methodological limitations (see the Appendix as supplementary on-line material); (8) Studies not published in English were excluded; (9) Double publications were also excluded. A total of 69 studies met the inclusion criteria and thus were included in our analysis. The final sample consisted of 56 trials for triptans reported in 52 publications, 9 trials for anticonvulsants reported in 9 publications and 8 trials for NSAIDs reported in 8 publications. The studies that were excluded are listed in Appendix 1 along with the reason for exclusion (see supplementary on-line material). In particular, we excluded 85 studies for triptans, 32 for anticonvulsants, and 32 for NSAIDs (see the trial flow represented in Fig. 1).

Each study was coded using a structured coding scheme (the full version can be requested from the corresponding author), including information on report identification, methodology, subjects and treatment. Adverse events were categorized into 28 groups based on symptoms reported in all the selected studies (Table 1). Symptom frequency was coded for each patient. We considered studies that reported adverse events in the placebo groups according to the number of patients and/or frequency per patient. Target symptoms of interventions (e.g. headache, pain symptoms) were excluded from adverse event coding. All the adverse effects were recorded from each study by two different researchers who were blind to the aim of the study. The two researchers collected the quality assessment of reports of randomized clinical trials in terms of Jadad scores, considering the description and sequence of randomization, the double blind procedure, its appropriateness and the description of withdrawals and dropouts (range 0–5) [10]. Also ascertainment strategy (structured versus spontaneous) was collected. In particular, the assessment of side effects was categorized as ‘‘structured recording” versus ‘‘spontaneous reports” (whether or not they used a checklist or diary cards to collect symptoms). Cohen’s kappa was calculated from each measurement. We also analyzed mean differences with Chi-square and confidence intervals. 2.4. Data analysis Statistical analysis was performed by a researcher (L.L.C.) who was blind to the aim of the study. Mean inter-rater reliability was kappa = 0.798 for assigning a subject to a specific symptom

M. Amanzio et al. / PAINÒ 146 (2009) 261–269 Table 1 Symptoms observed in the placebo groups listed in alphabetic order. Symptoms Abdominal pain Anorexia OR/AND weight loss Attention difficulties Burning OR/AND flushing Chest discomfort Chills Diarrhea Dizziness Dry mouth Dyspepsia Fatigue Heaviness Injection side reaction Insomnia Language difficulties Memory difficulties Nasal signs and symptoms Nausea Numbness Paresthesia OR/AND tingling Pharyngitis Somnolence OR/AND drowsiness Stinging OR/AND pressure sensation Taste disturbance Tinnitus Upper respiratory tract infection Vomiting Weakness

category, reliability was kappa = 0.889 for quality assessment in terms of Jadad scores and kappa = 0.763 for structured versus spontaneous recording. Adverse events were categorized into 28 symptom categories, and symptom frequency was coded. For each study, we computed the number of subjects presenting each symptom in the placebo group. We then summed the counts of the separate studies in order to obtain the cumulative number of subjects reporting each symptom in order to combine data from individual studies. In addition, we computed percentage scores and 95% confidence intervals (CI) representing the number of patients with each recorded symptom. Non-overlapping CIs of the rates of adverse effects indicate significant differences among the different drug categories and a CI including 0 corresponds to the absence of a significant adverse effect rate. In order to compare the different classes of drugs for each symptom category, we entered the observed total number of subjects into 2  2 contingency tables and computed Chi-square tests. The aim of this analysis was to investigate whether there was an association between the variable defining the drug categories (NSAIDs versus triptans versus anticonvulsants) and the variable defining the presence–absence of the symptom in the aggregate groups. To control for multiple tests, we adopted a Bonferroni correction of alpha value. For instance, given that we compared triptans and anticonvulsants for 16 different symptoms, the adjusted significance level was set at 0.05/16 (approximated to 0.003125). Moreover, we used the Yates’ correction for continuity, required in one-degree-of-freedom situations [20]. We performed two different analyses, to achieve a precise measure of AEs; the first, described above and a second more inclusive. It is important to note that in the first analysis we excluded studies characterized by no report of that adverse effect from the computation of the cumulative number of subjects presenting an adverse effect. For structured studies, no report clearly means that this adverse effect was not present in the check-list adopted. Indeed, these studies report a value of 0 to indicate that an adverse effect (included in the check-list) was not present in their placebo group.

263

If we consider spontaneous reports studies, no report is a more problematic point to be considered. Indeed, it might mean that the adverse effect was not present in the placebo group. However, it might also indicate that participants who really did present this symptom did not report it for several reasons. Therefore, we decided to re-run our analysis considering no report of a symptom in a spontaneous report trial as absence of the symptom in the placebo group of this trial. In other words, this trial was equated here to a trial explicitly reporting absence of the symptom (0 value). This does not modify the cumulative number of subjects presenting each symptom, but it does modify the total number of participants from which the percentage is computed. We then entered the observed total number of subjects into 2  2 contingency tables and computed Chi-square tests. In this case we again adopted a Bonferroni correction of alpha value and the Yates’ correction for continuity. 2.5. Analysis of other contributing factors We also investigated the influence of two categorical factors on the frequency of the adverse events. The two factors were ‘‘assessment strategy” for side effects (structured versus spontaneous reports) and quality of studies (evaluated by means of the Jadad score). Their influence was tested first by dividing the trials into two categories and then by comparing them. As to the quality factor, we compared trials with low Jadad scores (below 2.5) to trials with high Jadad scores (above 2.5). We computed a 2  2 contingency table for each drug and each symptom, and ran a Chi-square test to investigate the association between the quality level and the presence–absence of that specific symptom. As far as the assessment strategy factor is concerned, we compared trials rated as structured to trials rated as spontaneous. We computed a 2  2 contingency table for each drug and each symptom, and ran a Chi-square test to investigate the association between the assessment strategy and the presence–absence of that specific symptom. In this case, we again adopted a Bonferroni correction of alpha value. Others variables such as sample size, sex, age, weight, diagnosis of migraine (with and without aura), migraine frequency, race, withdrawal and publication year were also considered in the analysis. A series of Kruskal–Wallis tests were run to compare the trials in the three different placebo arms of anti-migraine drugs. In the presence of a significant effect, a Mann–Whitney test was used to define the nature and direction of the global effect. 3. Results The characteristics of the three placebo arms of anti-migraine drugs are shown in Table 2. Analyses revealed significant differences between the three groups regarding the number of patients, with a larger number of subjects included in trials of triptan-placebos than those of anticonvulsant-placebos [Kruskal–Wallis H (2, 73) = 9.1313, p = .0104 and Mann–Whitney U Test = 98.00, p = .0034]. They also revealed a significant difference regarding sex, with a larger proportion of females in the triptan-placebo and NSAID-placebo classes than those in the anticonvulsant-placebo class [Kruskal–Wallis H (2, 68) = 8.1802, p = .0167]. In addition, analyses revealed a larger number of patients with no aura recruited in trials of triptan-placebos than those of anticonvulsant-placebos (Kruskal–Wallis H (2, 41) = 6.2196, p = .0446 and Mann–Whitney U Test = 10.00; p = .0194). Analyses also revealed significant differences between the three groups regarding migraine frequency, with a higher frequency for subjects included in trials of anticonvulsant-placebos than subjects recruited in NSAID-placebo group [Kruskal–Wallis H (2, 18) = 11.6701, p = .0029], and significant differences between the three

M. Amanzio et al. / PAINÒ 146 (2009) 261–269

264

Table 2 Characteristics of the placebo groups, divided into the three anti-migraine pharmacological classes against which the placebo groups were compared.

NSAIDs trials Jadad score (range 0–5) Sample size (valid N) Females (valid N) Age (years) Weight (kg) Patients with aura (valid N) Patients without aura (valid N) Withdrew in general (valid N) Withdrew due to AE (valid N) Race (valid N) White Black Asian Caucasian Migraine frequency (months) Rescue medicine (valid N) Pain severity: mild (valid N) Pain severity: severe (valid N) Year of publication (valid N) Triptans trials Jadad score (range 0–5) Sample size (valid N) Females (valid N) Age (years) Weight (kg) Patients with aura (valid N) Patients without aura (valid N) Withdrew in general (valid N) Withdrew due to AE (valid N) Race (valid N) White Black Asian Caucasian Migraine frequency (months) Rescue medicine (valid N) Pain severity: mild (valid N) Pain severity: severe (valid N) Year of publication (valid N) Anticonvulsants trials Jadad score (range 0–5) Sample size (valid N) Females (valid N) Age (years) Weight (kg) Patients with aura (valid N) Patients without aura (valid N) Withdrew in general (valid N) Withdrew due to AE (valid N) Race (valid N) White Black Asian Caucasian Migraine frequency (months) Rescue medicine (valid N) Pain severity: mild (valid N) Pain severity: severe (valid N) Year of publication (valid N)

Mean

SD

Range

8 8 7 7 0 7 1 5 3

4.12 114 99.14 39.08 – 32 143 31.40 1

0.83 79.90 77.65 1.86 – 20.54 – 44.62 1.73

3–5 17–221 10–193 36.4–41 – 16–61 143–143 0–108 0–3

1 2 0 1 1 3 3 3 8

98 29.50 – 174 5.2 27.37 108 71.33 1999

– 19.09 – – – 36.14 36.75 3.05 6.47

98–98 16–43 – 174–174 5.2–5.2 6–69.1 74–147 68–74 1988–2006

56 56 52 52 5 39 37 36 38

3.61 167.27 138.19 38.06 70.32 19.56 118.40 16.36 0.39

1.00 108 86.73 7.54 12.92 13.35 72.26 23.37 0.68

2–5 26–558 21–401 14–55.5 54–86 4–64 25–325 0–95 0–2

21 11 9 7 9 34 25 25 56

140.76 10.18 2.22 139.14 3.50 88.73 89.08 48.36 2003

111.56 9.13 3.03 95.67 0.63 64.95 59.58 30.74 3.06

21–428 0–23 0–9 41–302 2.6–4.3 4–272 2–264 1–100 1992–2007

9 9 9 9 5 3 3 8 7

4.44 75.22 62.11 40.61 76.34 26.67 38.33 29 7.71

.53 53.04 47.03 2.42 2.65 19.60 28.71 24.66 4.64

4–5 14–161 9–140 38.1–44.4 74.1–80.7 6–45 14–70 2–73 2–14

4 4 3 0 8 0 0 0 9

97.25 12.50 1.00 – 8.22 – – – 2004

25.45 9.68 1.00 – 5.91 – – – 2.11

64–126 7–27 0–2 – 4.2–20 –– – – 2001–2007

groups regarding withdrawal due to AEs, with a larger number of withdrawals in trials of anticonvulsant-placebos than those of triptan-placebos and NSAID-placebos [Kruskal–Wallis H (2, 48) = 21.7713, p < .0001]. There were also significant differences between the three groups regarding Jadad scores, with higher scores for trials of anticonvulsant-placebos than those of triptan-placebos [Kruskal–Wallis H (2, 73) = 6.9021, p = .0317 and Mann–Whitney U Test = 131.00, p = .0215]. Conversely, no significant differences in race, age, weight, migraine without aura and migraine duration were observed

between placebo groups. Analysis of year of publication revealed no significant differences between the three classes of anti-migraine drugs [Kruskal–Wallis H (2, 73) = 4.036, p = .1329]. Tables 3 and 4 show the number and percentage, respectively, of AEs across the different placebo groups. The rates of specific AEs in the placebo arms varied widely across trials, as shown by the statistical comparisons in Table 5 and in Fig. 2. In particular, the placebo arm of NSAIDs trials produced more AEs than triptan-placebos in terms of dry mouth, nausea and vomiting. On the other hand anticonvulsant-placebos produced more AEs than NSAIDs in terms of fatigue, somnolence and dizziness. Anticonvulsant-placebos also produced more AEs than triptans in terms of fatigue, nausea, paresthesia, somnolence, dizziness and burning. These results were also confirmed by the more conservative analysis figured in Table 6, except for the symptoms dizziness and burning. In this table we reported the comparison of AEs in placebo groups of the three classes of anti-migraine drugs considering no report of a symptom as the absence of that symptom in the placebo group of the spontaneous trials. Interestingly this analysis revealed specific AEs not found in the analysis of Table 5; in particular, NSAID-placebos produced more AEs than triptans in terms of abdominal pain, dyspepsia and weakness. Anticonvulsant-placebos produced more AEs than both NSAIDs and triptans in terms of anorexia, fatigue, memory difficulties, paresthesia, somnolence and upper respiratory tract infection; but also more AEs in terms of abdominal pain, attention difficulties, diarrhea, nasal sign, and taste disturbance than triptans. In considering the role of the quality score as a contributing factor, we observed significant differences in terms of low and high Jadad scores in placebo groups of triptans trials for two symptoms, dizziness and somnolence. No other difference was observed (Table S2, see supplementary on-line material). In order to control this aspect we re-ran the statistical tests comparing triptans with the other two classes of trials for the two mentioned symptoms. As far as dizziness is concerned, the 2  2 Chi-square analysis revealed a significant difference between triptans and anticonvulsants, v2(1, N = 6147) = 26.14, p < 0.0001, but not between triptans and NSAIDs, v2(1, N = 6405) = 0.13, p = 0.721. As far as somnolence is concerned, the 2  2 Chi-square analysis revealed a significant difference between triptans and anticonvulsants, v2(1, N = 5368) = 16.72, p < 0.0001, but not between triptans and NSAIDs, v2(1, N = 5410) = 3.98, p = 0.046. These results mirror those obtained considering all studies regardless of the Jadad score (see Table 5), suggesting that the quality factor as measured through the Jadad score does not affect the differences found in comparing the different classes of drugs. All the studies with anticonvulsants and NSAIDs had Jadad scores above 2.5, thus we did not analyze these classes further. Supplementary Table S1 shows the number of adverse events in placebo groups of NSAID, triptan and anticonvulsant trials considering no report of a symptom as the absence of that symptom. Supplementary Table S3 shows the comparisons between structured versus spontaneous assessment in placebo groups of NSAID and triptan trials. Whereas all the studies with anticonvulsants were spontaneous reports, we observed significant differences between structured and spontaneous assessment in studies of triptans for one symptom, nausea, and in studies of NSAIDs for two symptoms, fatigue and nausea. No other difference was found. We provided supplementary materials figured in tables S4 and S5 representing comparison of adverse events in placebo groups considering structured and spontaneous report trials, respectively. These analyses confirmed the data obtained in Tables 5 and 6. 4. Discussion There are a number of studies showing that patients’ expectations can modify the rates of reported side effects, for example in

M. Amanzio et al. / PAINÒ 146 (2009) 261–269

265

Table 3 Number of adverse events in placebo groups of NSAID, triptan and anticonvulsant trials. Symptoms

NSAIDs

Abdominal pain Anorexia Attention difficulties Burning Chest discomfort Chills Diarrhea Dizziness Dry mouth Dyspepsia Fatigue Heaviness Injection side reaction Insomnia Language difficulties Memory difficulties Nasal signs Nausea Numbness Paresthesia Pharyngitis Somnolence Stinging Taste disturbance Tinnitus Upper respiratory tract infection Vomiting Weakness

Triptans n

Anticonvulsants

n

N

10

337

3

N 289

1 0 2

17 52 143

14 16 7 7

519 376 622 246

43 34 3 1 179 48 13 100 7 41

2356 4258 641 80 6850 2747 890 6803 1234 451

64

2 284 7 51 1 152 3 3 0

193 7660 633 5830 1150 5501 989 284 273

337 52

92 3

2099 193

90

856

2

182

6

571

1 30 4

n

N

6 24 6 10

142 548 341 115

1 8 20 5 5 56

27 228 261 161 156 642

0 1 20 17 48

14 174 548 276 605

37

562

30

529

6

448

27 3

234 142

Note. n, number of patients reporting symptoms; N, total sample size.

motion-induced nausea [13], allergic drug reactions [14] and tension [8]. The present study was aimed at testing the hypothesis that, before entering the double blind arm of a treatment trial, the examiner’s instructions about the possible adverse events asso-

ciated with a specific anti-migraine drug may affect the outcomes in the placebo group. In particular, if the active treatment is represented as an anticonvulsant drug, the examiner may suggest that the patients will experience those AEs that are specific to this

Table 4 Percentage of adverse events in placebo groups of NSAID, triptan and anticonvulsant trials. Symptoms

NSAIDs Percentage

Abdominal pain Anorexia Attention difficulties Burning Chest discomfort Chills Diarrhea Dizziness Dry mouth Dyspepsia Fatigue Heaviness Injection side reaction Insomnia Language difficulties Memory difficulties Nasal signs Nausea Numbness Paresthesia Pharyngitis Somnolence Stinging Taste disturbance Tinnitus Upper respiratory tract infection Vomiting Weakness

Triptans Inferior C.I.

Superior C.I.

Percentage

Anticonvulsants Inferior C.I.

Superior C.I.

2.967

1.144

4.791

1.038

0.139

2.215

5.882 0.000 1.399

6.588 0.000 0.553

18.353 0.000 3.350

2.697 4.255 1.125 2.846

1.302 2.202 0.296 0.753

4.093 6.309 1.955 4.938

1.825 0.798 0.468 1.250 2.613 1.747 1.461 1.470 0.567 9.091

1.284 0.531 0.061 1.238 2.235 1.257 0.672 1.184 0.148 6.421

2.366 1.066 0.997 3.738 2.991 2.237 2.249 1.756 0.986 11.761

4.688

1.036 3.708 1.106 0.875 0.087 2.763 0.303 1.056 0.000

0.403 3.284 0.291 0.636 0.083 2.330 0.040 0.141 0.000

2.476 4.131 1.921 1.114 0.257 3.196 0.646 2.254 0.000

11.963 15.192

4.383 1.554

3.507 0.204

5.259 3.313

10.514

8.458

12.570

1.099

0.428

2.625

1.051

0.214

1.888

1.563

1.563

8.902 7.692

5.842 0.192

Percentage

Inferior C.I.

Superior C.I.

4.225 4.380 1.760 8.696

0.871 2.665 0.355 3.470

7.580 6.095 3.164 13.921

3.704 3.509 7.663 3.106 3.205 8.723

3.926 1.103 4.413 0.404 0.404 6.538

11.333 5.915 10.913 5.807 6.006 10.907

0.000 0.575 3.650 6.159 7.934

0.000 0.557 2.078 3.303 5.778

0.000 1.707 5.221 9.015 10.089

6.584

4.531

8.636

5.671

3.698

7.644

1.339

0.268

2.410

11.538 2.113

7.415 0.285

15.662 4.511

Note. Values are given as percentages and 95% confidence intervals are indicated (Inferior = inferior value of the interval, Superior = superior value of the interval).

M. Amanzio et al. / PAINÒ 146 (2009) 261–269

266

Table 5 Comparison of adverse events in placebo groups of NSAID, triptan and anticonvulsant trials. Symptoms

NSAIDs versus triptans P value

NSAIDs versus anticonvulsants P value

Triptans versus anticonvulsants P value

Abdominal pain Anorexia Attention difficulties Burning Chest discomfort Chills Diarrhea Dizziness Dry mouth Dyspepsia Fatigue Heaviness Injection side reaction Insomnia Language difficulties Memory difficulties Nasal signs Nausea Numbness Paresthesia Pharyngitis Somnolence Stinging Taste disturbance Tinnitus Upper respiratory tract infection Vomiting Weakness

0.160

0.674

0.069

0.739 0.887 0.495

0.938

A > T 0.000

0.970

0.979 NSAIDs > T 0.003 0.739 0.142

A > NSAIDs 0.003 0.699 0.128 A > NSAIDs 0.004

0.389 0.518 A > T 0.000 0.343 0.226 A > T 0.000

NSAIDs > T 0.000

0.116

0.011 A > T 0.000

0.933

0.007

A > T 0.000

0.021

A > NSAIDs 0.000

A > T 0.000 0.996

0.429

NSAIDs > T 0.002

0.020

0.298 0.059

Note. T, Triptans; A, Anticonvulsants; the symbol > specifies which class has a higher percentage of adverse events in placebo groups. p = Chi-square level of significance, p in bold = significant difference with alpha < 0.05 after Bonferroni adjustment.

pharmacological class, such as fatigue, anorexia, somnolence and paresthesia [15]. On the other hand, if the treatment consists of NSAIDs versus placebo, the suggested AEs may be gastrointestinal and dry mouth [6]. We found specific side effects in the placebo arms of anti-migraine trials when analyzing the three groups of drugs. We observed that the side effects that are expected for the active drug against which the placebo is compared, are also more frequent in the placebo group. In particular, anticonvulsant-placebos appear to have a higher rate of AEs than the other two classes of anti-migraine drugs, as demonstrated by the two different analyses of Tables 5 and 6. The more inclusive analysis enable us to describe the presence of specific side effects for the anticonvulsant-placebo class, that is, anorexia, memory difficulties, paresthesia and upper respiratory tract infection: none of the other two placebo groups presented these symptoms in the spontaneous reports data collection studies. Moreover, it is also important to note that a larger number of patients in the anticonvulsant-placebo group discontinued the study (withdrawals due to AEs) than those in the triptanplacebo and NSAID-placebo groups. It is worth noting that the side effect profiles reported in the clinical trials analyzed in this study are not influenced by variables such as race, age, weight, migraine without aura and migraine duration. We also excluded any influence in terms of year of publication of the studies in the three different classes considered. Since we specifically wanted to analyze possible effects of suggestion, our analysis did not include a trial design comparing two different classes of anti-migraine drugs or crossover studies. In fact, the possible contribution of pharmacological conditioning in crossover trials could represent a confounding factor.

Fig. 2. Bar chart plots of the percentage of AEs in the placebo arms of NSAID (white bars), triptan (grey bars) and anticonvulsant trials (black bars). The error bars (thin black lines) refer to the confidence intervals of each percentage. Note: Diff = difficulties; ISR = injection side reaction; URT = upper respiratory tract.

Both patients’ and experimenters’ expectations may have affected the AEs occurrence in the placebo groups. For example, on the one hand, it has long been known that the investigator’s expectations may influence the outcome in behavioural research [9,19], thus they may have influenced the AEs in these anti-migraine trials as well. On the other hand, the patient’s expectation effects have been widely documented in a variety of conditions [1–3,7,11,17]. In fact, the occurrence of nocebo effects in both the clinical and the experimental setting has been investigated in detail and some of its neurobiological underpinnings have been unraveled [4,5,12]. Therefore, in the clinical trials we review here, both investigator’s and patient’s expectations may have played a role. In addition patients may also have strong desires to avoid the AEs. Previous studies have shown that a strong desire to avoid a therapeutic outcome, such as an AE, along with a low expectation of being able to avoid the outcome, may lead to negative emotions and to a decrease in placebo effects [21,22]. Therefore, expectations of AEs are likely to combine with desires to avoid AEs in the facilitation of nocebo effects. We also considered the influence of the ascertainment strategy used in the collection of symptoms (structured versus spontaneous data collection) and the quality assessment in terms of Jadad scores for each study included in the analysis. This choice was grounded on the fact that some authors have found that the use of specific assessment methods, i.e., structured check-lists (struc-

M. Amanzio et al. / PAINÒ 146 (2009) 261–269 Table 6 Comparison of adverse events in placebo groups of NSAID, triptan and anticonvulsant trials considering no report of a symptom as the absence of that symptom in the placebo group (see Section 2). Symptoms

NSAIDs versus triptans P value

NSAIDs versus anticonvulsants P value

Triptans versus anticonvulsants P value

Abdominal pain

NSAIDs > T 0.0000

0.2965

A > T 0.0000

A > NSAIDs 0.0006 0.1684 0.1339

A > T 0.0000

Anorexia Attention difficulties Burning Chest discomfort Chills Diarrhea Dizziness Dry mouth Dyspepsia Fatigue

0.6421 0.3868 0.0646 0.0644 0.9836 NSAIDs > T 0.0000 NSAIDs > T 0.0006 0.8540

Heaviness Injection side reaction Insomnia Language difficulties Memory difficulties

0.8242 0.1949

Nasal signs Nausea

0.2589 NSAIDs > T 0.0000 0.8741 0.6250

Numbness Paresthesia Pharyngitis Somnolence Stinging Taste disturbance Tinnitus Upper respiratory tract infection Vomiting Weakness

0.0450 0.1118 0.4165 0.4165 0.0644

NSAIDs > T 0.0000 NSAIDs > T 0.0000

0.8040 0.0862 0.3530 0.0069

A > T 0.0000 0.0174 0.1549 0.8757 A > T 0.0000 0.1030 0.9966

0.7659

0.0288

A > NSAIDs 0.0000

A > T 0.0000 0.8424 0.0505

0.7546 A > NSAIDs 0.0020 0.0051 0.0637

A > NSAIDs 0.0000 A > NSAID 0.0001 0.1684 0.7546 A > NSAIDs 0.0002 NSAIDs > A 0.0000 0.0278

0.2002 A > T 0.0000 A > T 0.0000 A > T 0.0000 0.7927 A > T 0.0000 0.1615 A > T 0.0000 0.7079 A > T 0.0000 A > T 0.0000 0.0108 0.7322

Note. T = triptans, A = anticonvulsants; the symbol > specifies which class has the highest percentage of adverse events in placebo groups, an empty cell means that the computation was impossible (for instance, with a row of the 2  2 matrix only containing zeros). p = Chi-square level of significance, p in bold = significant difference with alpha < 0.05 after Bonferroni adjustment.

tured data collection) yields higher rates of side effects than spontaneous patient reports, in placebo-controlled trials of statin drugs [18]. Moreover, considering the variability of the data collection process, given the possible unreliability of non-structured or non-standardized AEs assessment [18], we compared studies with high versus low Jadad scores and structured versus spontaneous symptom data collection. In triptan-placebo trials we observed significant differences between low and high Jadad scores for dizziness and somnolence. However, the differences for these two AEs remained significant, and persisted when we restricted the analysis to studies with high Jadad scores, demonstrating that the differences we observed were not dependent on Jadad scores. Although it was not possible to perform a comparison between structured and spontaneous assessments for trials with anticonvulsants (as these were all spontaneous), we observed significant differences for nausea only in triptan-placebos and NSAID-placebos and for fatigue only in NSAID-placebos. In our analysis structured assessment approaches did not reveal much higher rates for symptoms than unstructured (spontaneous) approaches, except for nausea in NSAID-placebos.

267

Some possible limitations of our study also need to be mentioned. Meta-analyses and systematic reviews are prone to selection biases. In addition, a limited number of factors can usually be analyzed, although we tried to control as many factors as possible, such as race, age, weight, migraine without aura and migraine duration. As discussed above, a further problem is represented by missing information on side effects in clinical trials. This was the reason why we performed two analyses, as shown in Tables 4–6, which show a different distribution of side effect profiles across different placebo groups. In conclusion, we believe that this study offers important considerations for the evaluation of anti-migraine trials, the need for caution in interpreting side effect profiles of different anti-migraine drugs (e.g. NSAIDs, triptans and anticonvulsants) and for greater consideration of the respective placebo groups, and the importance of achieving an optimum compromise between ethical concerns due to the need to inform patients about expected AEs and the need to reduce expectancy effects. For example, if the investigator’s expectations influence the frequency of reported AEs, it would be important for the examiner to be blind about expected side effects of the drug being evaluated. Likewise, informed consents should be written carefully in order to avoid the occurrence of negative expectations in the patient. Although this is not easy to do and it is often difficult to separate pharmacodynamic AEs from psychological AEs, clinical trialists should consider the possibility of nocebo effects whenever an AE is reported. We believe that the present study emphasizes the need to develop new designs and new procedures of AE assessment in clinical trials, in order to both minimize the risk and maximize the benefit of drugs to the patient’s advantage. Acknowledgments All authors report no competing interests. This work was supported by grants from Fondazione San Paolo and Regione Piemonte. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the on-line version, at doi:10.1016/j.pain.2009.07.010. References [1] Amanzio M, Pollo A, Maggi G, Benedetti F. Response variability to analgesics: a role for non-specific activation of endogenous opioids. Pain 2001;90:205–15. [2] Barsky AJ. Amplification, somatization, and the somatoform disorders. Psychosomatics 1992;33:28–34. [3] Benedetti F. Mechanisms of placebo and placebo-related effects across diseases and treatments. Annu Rev Pharmacol Toxicol 2008;48:33–60. [4] Benedetti F, Amanzio M, Vighetti S, Asteggiano G. The biochemical and neuroendocrine bases of the hyperalgesic nocebo effect. J Neurosci 2006;26:12014–22. [5] Benedetti F, Lanotte M, Lopiano L, Colloca L. When words are painful: unraveling the mechanisms of the nocebo effect. Neuroscience 2007;147:260–71. [6] Burke A, Smyth EM, Fitzgerald GA. Analgesic-antipyretic and antiinflammatory agents, pharmacotherapy of gout. In: Brunton LL, Lazo JS, Parker K, editors. Goodman & Gilman’s. The pharmacological basis of therapeutics. New York: Mc Graw-Hill; 2006. p. 671–716. [7] Enck P, Benedetti F, Schedlowski M. New insights into the placebo and nocebo responses. Neuron 2008;59:195–206. [8] Flaten MA, Simonsen T, Olsen H. Drug-related information generates placebo and nocebo responses that modify the drug response. Psychosom Med 1999;61:250–5. [9] Gracely RH, Dubner R, Wolskee PJ, Deeter WR. Placebo and naloxone can alter postsurgical pain by separate mechanisms. Nature 1983;306:264–5. [10] Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds JM, Gavaghan DJ, McQuay HJ. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996;17:1–12. [11] Kirsch I. Response expectancy as a determinant of experience and behavior. Am Psychol 1985;40:1189–202.

268

M. Amanzio et al. / PAINÒ 146 (2009) 261–269

[12] Kong J, Gallub RL, Polich G, Kirsch I, Laviolette P, Vangel M, Rosen B, Kaptchuk TJ. A functional magnetic resonance imaging study on the neural mechanisms of hyperalgesic nocebo effect. J Neurosci 2009;28:13354–62. [13] Levine ME, Stern RM, Koch KL. The effects of manipulating expectations through placebo and nocebo administration on gastric tachyarrhythmia and motion-induced nausea. Psychosom Med 2006;68:478–86. [14] Liccardi G, Senna G, Russo M, Bonadonna P, Crivellaro M, Dama A, D’Amato M, D’Amato G, Canonica GW, Passalacqua G. Evaluation of the nocebo effect during oral challenge in patients with adverse drug reactions. J Investig Allergol Clin Immunol 2004;14:104–7. [15] McNamara JO. Pharmacotherapy of the epilepsies. In: Brunton LL, Lazo JS, Parker K, editors. Goodman & Gilman’s. The pharmacological basis of therapeutics. New York: Mc Graw-Hill; 2006. p. 501–26. [16] Peroutka SJ. Drugs effective in the therapy of migraine. In: Hardman JG, Limbird LE, editors. Goodman & Gilman’s. The pharmacological basis of therapeutics. New York: Mc Graw-Hill; 1996. p. 487–502. [17] Price DD, Finniss DG, Benedetti F. A comprehensive review of the placebo effect: recent advances and current thought. Annu Rev Psychol 2008;59:565–90. [18] Rief W, Avorn J, Barsky AJ. Medication-attributed adverse effects in placebo groups. Implications for assessment of adverse effects. Arch Int Med 2006;166:155–60. [19] Rosenthal R. Experimenter effects in behavioral research. New York: Appleton Century Crofts; 1966. [20] Runyon RP, Haber A. Fundamentals of behavioral statistics. AddisonWesley: Publishing Company; 1971. [21] Vase L, Robinsen ME, Verne GN, Price DD. Increased placebo analgesia over time in irritable bowel syndrome (IBS) patients is associated with desire and expectation but not endogenous opioid mechanisms. Pain 2005;115:338–47. [22] Vase L, Price DD, Verne GN, Robinson ME. The contribution of changes in expected pain levels and desire for pain relief to placebo analgesia. In: Price DD, Bushnell MC, editors. Psychological methods of pain control: basic science and clinical perspectives, progress in pain research and management. Seattle: IAPS Press; 2004. p. 207–32.

Further reading Following studies are included in the analysis. [23] Ahrens SP, Farmer MV, Williams DL, Willoughby E, Jiang K, Block GA, Visser WH. Efficacy and safety of rizatriptan wafer for the acute treatment of migraine; Rizatriptan Wafer Protocol 049 Study Group. Cephalalgia 1999;19:525–30. [24] Bomhof M, Paz J, Legg N, Allen C, Vandormael K, Patel K, the RizatriptanNaratriptan Study Group. Comparison of rizatriptan 10 mg vs. naratriptan 2.5 mg in migraine. Eur Neurol 1999;42:173–9. [25] Brandes JL, Saper JR, Diamond M, Couch JR, Lewis DW, Schmitt J, Neto W, Schwabe S, Jacobs D. MIGR-002 Study Group. Topiramate for migraine prevention: a randomized controlled trial. JAMA 2004;291:965–73. [26] Brandes JL, Kudrow D, Cady R, Tiseo PJ, Sun W, Sikes CR. Eletriptan in the early treatment of acute migraine: influence of pain intensity and time of dosing. Cephalalgia 2005;25:735–42. [27] Cady RC, Ryan R, Jhingran P, O’Quinn S, Pait G. Sumatriptan injection reduces productivity loss during a migraine attack. Arch Intern Med 1998;158:1013–8. [28] Cady R, Martin V, Mauskop A, Rodgers A, Hustad CM, Ramsey KE, Skobieranda F. Efficacy of rizatriptan 10 mg administered early in a migraine attack. Headache 2006;46:914–24. [29] Carpay J, Schoenen J, Ahmad F, Kinrade F, Boswell D. Efficacy and tolerability of sumatriptan tablets in a fast-disintegrating, rapid-release formulation for the acute treatment of migraine: results of a multicenter, randomized, placebo-controlled study. Clin Ther 2004;26:214–23. [30] Carrieri PB, Orefice G, Sorge FA. Double-blind placebo-controlled trial of indobufen in the prophylaxis of migraine. Acta Neurol Scand 1988;77:433–6. [31] Codispoti JR, Prior MJ, Fu M, Harte CM, Nelson EB. Efficacy of non prescription doses of ibuprofen for treating migraine headache: a randomized controlled trial. Headache 2001;41:665–79. [32] Dahlof C, Björkman R. Diclofenac-K (50 and 100 mg) and placebo in the acute treatment of migraine. Cephalalgia 1993;13:117–23. [33] Dahlöf C, Tfelt-Hansen P, Massiou H, Fazekas A, Almotriptan Study Group. Dose finding, placebo-controlled study of oral almotriptan in the acute treatment of migraine. Neurology 2001;57:1811–7. [34] Diener HC, Tfelt-Hansen P, de Beukelaar F, Ferrari MD, Olesen J, Dahlöf C, Mathew N, on behalf of the Study Group. The efficacy and safety of sc alniditan vs. sc sumatriptan in the acute treatment of migraine: a randomized, doubleblind, placebo-controlled trial. Cephalalgia 2001;21:672–9. [35] Diener HC, Bussone G, Van Oene JC, Lahaye M, Schwalen S, Goadsby PJ. TOPMAT-MIG-201(TOP-CHROME) Study Group. Topiramate reduces headache days in chronic migraine: a randomized, double-blind, placebo-controlled study. Cephalalgia 2007;27:814–23. [36] Dowson AJ, MacGregor EA, Purdy RA, Becker WJ, Green J, Levy SL. Zolmitriptan orally disintegrating tablet is effective in the acute treatment of migraine. Cephalalgia 2002;22:101–6.

[37] Dowson AJ, Massiou H, Laínez JM, Cabarrocas X. Almotriptan is an effective and well-tolerated treatment for migraine pain: results of a randomized, double-blind, placebo controlled clinical trial. Cephalalgia 2002;22:453–61. [38] Drescher MJ, Elstein Y. Prophylactic COX 2 inhibitor: an end to the Yom Kippur Headache. Headache 2006;46:1487–91. [39] Eletriptan Steering Committee in Japan. Efficacy and safety of eletriptan 20 mg, 40 mg and 80 mg in Japanese migraineurs. Cephalalgia 2002;22:416–23. [40] Elkind AH, Satin LZ, Nila A, Keywood C. Frovatriptan use in migraneurs with or at high risk of coronary artery disease. Headache 2004;44:403–10. [41] Färkkilä M, Olesen J, Dahlöf C, Stovner LJ, ter Bruggen JP, Rasmussen S, Muirhead N, Sikes C. Elitriptan for the treatment of migraine in patients with previous poor response or tolerance to oral sumatriptan. Cephalalgia 2003;23:463–71. [42] Freitag FG, Collins SD, Carlson HA, Goldstein J, Saper J, Silberstein S, Mathew N, Winner PK, Deaton R, Sommerville K, Depakote ER, Migraine Study Group. A randomized trial of divalproex sodium extended-release tablets in migraine prophylaxis. Neurology 2002;58:1652–9. [43] Garcia-Ramos G, MacGregor EA, Hilliard B, Bordini CA, Leston J, Hettiarachchi J. Comparative efficacy of eletriptan vs naratriptan in the acute treatment of migraine. Cephalalgia 2003;23:869–76. [44] Goadsby PJ, Ferrari MD, Olesen J, Stovner LJ, Senard JM, Jackson NC, Poole PH, for the Eletriptan Steering Committee. Eletriptan in acute migraine: a doubleblind, placebo-controlled comparison to sumatriptan. Neurology 2000;54:156–63. [45] Goldstein J, Keywood C, on behalf of the 251/96/14 Study Group. Frovatriptan for the acute treatment of migraine: a dose-finding study. Headache 2002;42:41–8. [46] Ishkanian G, Blumenthal H, Webster CJ, Richardson MS, Ames M. Efficacy of sumatriptan tablets in migraineurs self-described or physician-diagnosed as having sinus headache: a randomized, double-blind, placebo-controlled study. Clin Ther 2007;29:99–109. [47] Jelinski SE, Becker WJ, Christie SN, Ahmad FE, Pryse-Phillips W, Simpson SD. Pain free efficacy of sumatriptan in the early treatment of migraine. Can J Neurol Sci 2006;33:73–9. [48] Kellstein DE, Lipton RB, Geetha R, Koronkiewicz K, Evans FT, Stewart WF, Wilkes K, Furey SA, Subramanian T, Cooper SA. Evaluation of a novel solubilized formulation of ibuprofen in the treatment of migraine headache: a randomized, double-blind, placebo-controlled, dose-ranging study. Cephalalgia 2000;20:233–43. [49] Klapper J, Lucas C, Røsjø Ø, Charlesworth B, on behalf of the ZODIAC study group. Benefits and treating highly disabled migraine patients with zolmitriptan while pain is mild. Cephalalgia 2004;24:918–24. [50] Landy S, Savani N, Shackelford S, Loftus J, Jones M. Efficacy and tolerability of sumatriptan tablets administered during the mild-pain phase of menstrually associated migraine. Int J Clin Pract 2004;58:913–9. [51] Loder E, Silberstein SD, Abu-Shakra S, Mueller L, Smith T. Efficacy and tolerability of oral zolmitriptan in menstrually associated migraine: a randomized, prospective, parallel-group, double-blind, placebo-controlled study. Headache 2004;44:120–30. [52] Loder E, Freitag FG, Adelman J, Pearlman S, Abu-Shakra S, on behalf of the zolmitriptan study 014 study group. Pain-free rates with zolmitriptan 2.5 mg ODT in the acute treatment of migraine: results of a large double-blind placebo- controlled trial. Curr Med Res Opin 2005;21:381–9. [53] Mannix LK, Loder E, Nett R, Mueller L, Rodgers A, Hustad CM, Ramsey KE, Skobieranda F. Rizatriptan for the acute treatment of ICHD-II proposed menstrual migraine: two prospective, randomized, placebo-controlled, double-blind studies. Cephalalgia 2007;27:414–21. [54] Mathew NT, Schoenen J, Winner P, Muirhead N, Sikes CR. Comparative efficacy of elitriptan 40 mg versus sumatriptan 100 mg. Headache 2003;43:214–22. [55] Mathew NT, Finlayson G, Smith TR, Cady RK, Adelman J, Mao L, Wright P, Greenberg SJ, AEGIS Investigator Study Group. Early intervention with almotriptan: results of the AEGIS trail (AXERTÒ Early Migraine Intervention Study). Headache 2007;47:189–98. [56] Mei D, Capuano A, Vollono C, Evangelista M, Ferraro D, Tonali P, Di Trapani G. Topiramate in migraine prophylaxis: a randomized double-blind versus placebo study. Neurol Sci 2004;25:245–50. [57] Nett R, Landy S, Shackelford S, Richardson MS, Ames M, Lener M. Pain-free efficacy after treatment with sumatriptan in the mild pain phase of menstrually associated migraine. Obstet Gynecol 2003;102:835–42. [58] Newman L, Mannix LK, Landy S, Silberstein S, Lipton RB, Putnam DG, Watson C, Jöbsis M, Batenhorst A, O’Quinn S. Naratriptan as short-term prophylaxis of menstrually associated migraine: a randomized, double-blind, placebocontrolled study. Headache 2001;41:248–56. [59] Olesen J, Diener HC, Schoenen J, Hettiarachchi J. No effect of eletriptan administration during the aura phase of migraine. Eur J Neurol 2004;11:671–7. [60] Pascual J, Vega P, Diener HC, Allen C, Vrijens F, Patel K, RizatriptanZolmitriptan Study group. Comparison of rizatriptan 10 mg vs. zolmitriptan 2.5 mg in the acute treatment of migraine. Cephalalgia 2000;20:455–61. [61] Pascual J, Falk RM, Piessens F, Prusinski A, Docekal P, Robert M, Ferrer P, Luria X, Segarra R, Zayas JM. Consistent efficacy and tolerability of almotriptan in the acute treatment of multiple migraine attacks: results of a large, randomized, double-blind, placebo-controlled study. Cephalalgia 2000;20:588–96. [62] Pfaffenrath V, Cunin G, Sjonell G, Prendergast S. Efficacy and safety of sumatriptan tablets (25 mg, 50 mg, and 100 mg) in the acute treatment of

M. Amanzio et al. / PAINÒ 146 (2009) 261–269

[63]

[64] [65]

[66]

[67]

[68]

[69]

[70]

[71]

[72]

[73]

[74]

[75]

[76]

migraine: defining the optimum doses of oral sumatriptan. Headache 1998;38:184–90. Rapoport AM, Ramadan NM, Adelman JU, Mathew NT, Elkind AH, Kudrow DB, Earl NL. Optimizing the dose of zolmitriptan (Zomig, 311C90) for the acute treatment of migraine. A multicenter, double-blind, placebo-controlled, dose range-finding study. The 017 Clinical trial Study Group. Neurology 1997;49:1210–8. Rapoport A, Ryan RR, Goldstein J, Keywood C. Dose range-finding studies with frovatriptan in the acute treatment of migraine. Headache 2002;42:S74–83. Rothner AD, Wasiewski W, Winner P, Lewis D, Stankowski J. Zolmitriptan oral tablet in migraine treatment: high placebo responses in adolescents. Headache 2006;46:101–9. Sakai F, Iwata M, Tashiro K, Itoyama Y, Tsuji S, Fukuuchi Y, Sobue G, Nakashima K, Morimatsu M. Zolmitriptan is effective and well tolerated in Japanese patients with migraine: a dose-response study. Cephalalgia 2002;22:376–83. Sandrini G, Färkkilä M, Burgess G, Forster E, Haughie S, for the Elitriptan Steering Committee. A double-blind, placebo-controlled, multiple migraine attack study. Neurology 2002;59:1210–7. Saper J, Dahlof C, So Y, Tfelt-Hansen P, Malbecq W, Loeys T, Barraclough E, Klipfel M, Lines C, Visser H, Reines S, Yuen E, Rofecoxib Protocol 162 Study Group. Rofecoxib in the acute treatment of migraine: a randomized controlled clinical trial. Headache 2006;46:264–75. Savani N, Brautaset NJ, Reunanen M, Szirmai I, Ashford EA, Hassani H, Saiers J. A double-blind placebo-controlled study assessing the efficacy and tolerability of 50 mg sumatriptan tablets in the acute treatment of migraine. Sumatriptan Tablets S2CM07 Study Group. Int J Clin Pract Suppl 1999;105:7–15. Sheftell F, Ryan R, Pitman V, Eletriptan Steering Committee. Efficacy, safety and tolerability of oral eletriptan for treatment of acute migraine: a multicenter, double-blind, placebo-controlled study conducted in the United States. Headache 2003;43:202–13. Sheftell FD, Dahlöf CG, Brandes JL, Agosti R, Jones MW, Barrett PS. Two replicate randomized, double-blind, placebo-controlled trials of the time to onset of pain relief in the acute treatment of migraine with a fastdisintegrating/rapid-release formulation of sumatriptan tablets. Clin Ther 2005;27:407–17. Silberstein S, Tepper S, Brandes J, Diamond M, Goldstein J, Winner P, Venkatraman S, Vrijens F, Malbecq W, Lines C, Visser WH, Reines S, Yuen E. Randomized, placebo-controlled trial, of rofecoxib in the acute treatment of migraine. Neurology 2004;62:1552–7. Silberstein SD, Neto W, Schmitt J, Jacobs D, MIGR-001 Study Group. Topiramate in migraine prevention: results of a large controlled trial. Arch Neurol 2004;61:490–5. Silberstein SD, Hulihan J, Karim MR, Wu SC, Jordan D, Karvois D, Kamin M. Efficacy and tolerability of topiramate 200 mg/d in the prevention of migraine with/without aura in adults: a randomized, placebo-controlled, double-blind, 12-week pilot study. Clin Ther 2006;28:1002–11. Silberstein SD, Lipton RB, Dodick DW, Freitag FG, Ramadan N, Mathew N, Brandes JL, Bigal M, Saper J, Ascher S, Jordan DM, Greenberg SJ, Hulihan J, Topiramate Chronic Migraine Study Group. Efficacy and safety of topiramate for the treatment of chronic migraine: a randomized, double-blind, placebocontrolled trial. Headache 2007;47:170–80. Silvestrini M, Bartolini M, Coccia M, Baruffaldi R, Taffi R, Provinciali L. Topiramate in the treatment of chronic migraine. Cephalalgia 2003;23: 820–4.

269

[77] Spierings ELH, Rapoport AM, Dodick DW, Charlesworth B. Acute treatment of migraine with zolmitriptan 5 mg orally disintegrating tablet. CNS Drugs 2004;18:1133–41. [78] Stark R, Dahlöf C, Haughie S, Hettiarachchi on behalf of the Eletriptan Steering Committee. Efficacy, safety and tolerability of oral eletriptan in the acute treatment of migraine: results of a phase III, multicenter, placebo-controlled study across three attacks. Cephalalgia 2002;22:23–32. [79] Steiner TJ, Diener H-C, MacGregor EA, Schoenen J, Muirhead N, Silkes CR. Comparative efficacy of elitriptan and zolmitriptan in the acute treatment of migraine. Cephalalgia 2003;23:942–52. [80] Storey JR, Calder CS, Hart DE, Potter DL. Topiramate in migraine prevention: a double-blind, placebo-controlled study. Headache 2001;41:968–75. [81] Teall J, Tuchman M, Cutler N, Gross M, Willoughby E, Smith B, Jiang K, Reines S, Block G. Rizatriptan (MAXALT) for the acute treatment of migraine and migraine recurrence. A placebo-controlled, outpatient study. Rizatriptan 022 Study Group. Headache, 1998;38:281–87. [82] Tepper SJ, Cady R, Dodick D, Freitag FG, Hutchinson SL, Twomey C, Kuhn TA. Oral sumatriptan for the acute treatment of probable migraine: first randomized, controlled study. Headache 2006;46:115–24. [83] Tfelt-Hansen P, Teall J, Rodriguez F, Giacovazzo M, Paz J, Malbecq W, Block GA, Reines SA, Visser WH, on behalf of the Rizatriptan 030 Study Group. Oral rizatriptan versus oral sumatriptan: a direct comparative study in the acute treatment of migraine. Headache 1998;38:748–55. [84] Tuchman M, Hee A, Emeribe U, Silberstein S. Efficacy and tolerability of zolmitriptan oral tablet in the acute treatment of menstrual migraine. CNS Drugs 2006;20:1019–26. [85] Van Gerven JM, Schoemaker RC, Jacobs LD, Reints A, Ouwersloot-van der Meij MJ, Hoedemaker HG, Cohen AF. Self-medication of a single headache episode with ketoprofen, ibuprophen or placebo, home-monitored with an electronic patient diary. Br J Clin Pharmacol 1996;42:475–81. [86] Visser WH, Ferrari MD, Bayliss EM, Ludlow S, Pilgrim AJ. Treatment of migraine attacks with subcutaneous sumatriptan: first placebo-controlled study. The Subcutaneous Sumatriptan International Study Group. Cephalalgia 1992;12:308–13. [87] Visser WH, Winner P, Strohmaier K, Klipfel M, Peng Y, McCarroll K, Cady R, Lewis D, Nett R, Rizatriptan Protocol 059 and 061 Study Groups. Rizatriptan 5 mg for the acute treatment of migraine in adolescents: results from a double-blind, single-attack study and two open-label, multiple-attack studies. Headache 2004;44:891–9. [88] Wendt J, Cady R, Singer R, Peters K, Webster C, Kori S, Byrd S. A randomized, double-blind, placebo-controlled trial of the efficacy and tolerability of a 4-mg dose of subcutaneous sumatriptan for the treatment of acute migraine attacks in adults. Clin Ther 2006;28:517–26. [89] Winner P, Lewis D, Visser WH, Jiang K, Ahrens S, Evans JK, on behalf of the Rizatriptan Adolescent Study Group. Rizatriptan 5 mg for the acute treatment of migraine in adolescents: a randomized, double blind, placebo-controlled study. Headache 2002;42:49–55. [90] Winner P, Mannix LK, Putnam DG, McNeal S, Kwong J, O’Quinn S, Richardson MS. Pain-free results with sumatriptan taken at the first sign of migraine pain: two randomized, double-blind, placebo-controlled studies. Mayo Clin Proc 2003;78:1214–22. [91] Winner P, Linder SL, Lipton RB, Almas M, Parsons B, Pitman V. Eletriptan for the acute treatment of migraine in adolescents: results of a double-blind, placebocontrolled trial. Headache 2007;47:511–8.