Comparison between metamizole and triptans for migraine treatment: a systematic review and network meta-analysis

Objective The aim of this systematic review was to evaluate the efficacy of metamizole and triptans for the treatment of migraine. Methods Randomized controlled trials including people who received metamizole or triptan by multiple routes of administration and at all doses as treatment compared to subjects who received another treatment or placebo were included in the systematic review. The primary outcomes were freedom from pain at 2 hours; pain relief at 2 hours; sustained headache response at 24 hours; sustained freedom from pain at 24 hours. The statistical analysis of all interventions of interest were based on random effect models compared through a network meta-analysis. Results 209 studies meeting the inclusion and exclusion criteria were analyzed. Of these, 130 had data that could be analyzed statistically. Only 3.0% provided enough information and were judged to have a low overall risk of bias for all categories evaluated; approximately 50% of the studies presented a low risk of selection bias. More than 75% of the studies presented a low risk of performance bias, and around 75% showed a low risk of detection and attrition bias. Conclusion There is no evidence of a difference between dipyrone and any triptan for pain freedom after 2 hours of medication. Our study suggests that metamizole may be equally effective as triptans in acute migraine treatment. Mario Fernando Prieto Peres marioperes@yahoo.com


Introduction
M igraine is a highly prevalent condition manifesting as moderate or severe intermittent headache attacks with associated symptoms, lasting 4 to 72 hours if not properly treated. [1][2][3] Migraine is not only a headache but also a syndrome of various phases, each with its own distinct mechanisms and treatment approaches. Briefly, the migraine prodrome, or premonitory phase, can occur several hours to days before a headache and may be hypothalamically modulated, although other brainstem and limbic structures may play a causal role as well. 4 The relationship between migraine and cognition is complex. Cognitive symp-toms are part of the subjective experience of migraine attacks and contribute to attackrelated disability, interfering with work performance, family and social life, besides self-management of the attacks. This transient impairment may occur along all phases of a migraine attack. While pain is the main determinant of disability, cognitive dysfunction also contributes to attack-related impairment, and should be considered as a migraine therapeutic target, together with pain, to evaluate the efficacy of an acute attack treatment. 5 While it is clear that migraine attacks include some degree of cognitive impair-ment, in the long run, migraine is not associated with any significant impact on cognitive performance or age-associated cognitive decline in the general popu-lation. So, acute cognitive dysfunction during a migraine attack is reversible. However, individuals with more severe and frequent migraine attacks and sub-jects with chronic migraine tend to maintain cognitive difficulties between at-tacks. 5 The acute management of migraines includes the use of non-steroidal anti-inflammatory drugs (NSAIDs), acetaminophen, metamizole, ergots, and triptans. Gepants and ditans have been recently added to the list. 6 Metamizole is a well-established and highly used drug to treat acute attacks in emergency settings in Brazil, as well as the most common analgesic medication used for migraine treatment in the population. 7,8 Triptans represent a large therapeutic group with a good therapeutic profile, but their vasoconstriction adverse events warrant caution in patients with cardio-vascular risk. Other side effects, such as nausea, dizziness and chest symptoms, preclude some patients from using triptans, while a few patients do not respond well to triptans. Compliance and tolerability of triptans are certainly different for each medicine. Triptans are considered to be safe, with a very low potential risk of clinically significant serious adverse events. Contraindications to triptan use include uncontrolled hypertension, ischemic heart disease, coronary vasospasm, cerebrovascular disease, peripheral vascular disease, and basilar or hemiplegic migraine. 9,10 Metamizole and triptans are both major medications in the acute therapy arse-nal, however, they have never been directly compared.
To evaluate the efficacy of metamizole and triptans for the treatment of migraine, we conducted this systematic review and network meta-analysis to address the following focused questions: (1) "what is the evidence for the efficacy and safety of metamizole for the treatment of migraines compared with triptans?" and (2) "how effective are those treatments in improving cognitive dysfunction in patients with migraine?"

Methods
TThe systematic review has been developed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Protocols (PRISMA) statement [11][12][13] , using methodology described in the Cochrane Handbook for Systematic Review of Interventions. 14 This protocol was registered within the PROSPERO database (CRD42020216360).

Study eligibility criteria
Only randomized controlled trials reporting study-specific data for migraine outcomes in people who received metamizole or triptan as treatment were included in the systematic review. The population of interest was participants with migraine, of any age, gender and severity of migraine. We have used investigator-reported definitions (according to accepted diagnostic criteria, such as the International Classification of Diseases, or according to the criteria established by the International Headache Society). 15 We examined papers from all countries, subjects who have used metamizole or triptan treatment (test group), by multiple routes of administration (tablets, oral disintegrating tablets, injection, transdermal, nasal spray, rectal suppositories) and at all doses (any frequency or strength), compared to subjects who have received another treatment or placebo. Metamizole and Triptans were not allowed to be used in combination with other drugs. The primary outcomes were freedom from pain at 2 hours; pain relief at 2 hours; sustained headache response at 24 hours; sustained freedom from pain at 24 hours. Secondary outcomes were relief of other symptoms associated with migraine, specifically nausea, vomiting, photophobia and phonophobia, fatigue, dizziness, cognitive impairment, any adverse effects (AEs), withdrawals due to adverse events, use of rescue medication, patient satisfaction, absenteeism, functional disability and quality of life.
We excluded studies in which metamizole or triptan was not the intervention of interest, studies comparing combined metamizole preparations with another treatment, studies comparing combined triptan preparations with another treatment, studies where metamizole or triptan have not been studied in only one separate intervention group, studies in which migraine is not reported as the outcome of interest, studies that do not have adequate information regarding whether metamizole or triptan and its derivatives are not related to migraine improvement, studies involving secondary headache disorders (post-puncture headache, post-traumatic headache, cancer-related headache etc.), studies that do not have adequate information on the classification of primary headache or animal studies. There was no restriction of study setting.

Information sources
We searched the literature in the following databases: MEDLINE via PubMed, EMBASE, LILACS, EbscoHost and all references of the included studies, with no language restrictions from inception to November 2020. MesH terms and keywords were combined with Boolean operators and used as search strategies: #1 -migraine OR headache OR "tension-type headache"; #2 -dipyrone OR metamizole; #3 -triptan OR sumatriptan OR zolmitriptan OR rizatriptan OR naratriptan OR frovatriptan OR almotriptan OR eletriptan; #4 -#1 AND #2; #5 -#1 AND #3; #6 -#4 OR #5. Two reviewers screened all articles identified from the search independently. Any disagreements between reviewers were solved by discussion with a third reviewer to meet a consensus. Studies meeting the inclusion criteria underwent a validity assessment and data extraction. Reasons for rejecting studies were recorded for each study.

Data extraction (study characteristics and results) / Data management
Two reviewers extracted data independently. Disagreements were solved with discussion with a third reviewer. Data were transferred to Excel sheets for analysis. Measures of central tendency (mean or median) and dispersion (standard deviations and percentiles) for different biometric parameters were extracted. For continuous outcomes, the following was extracted: means, SD and sample sizes at baseline and follow-up. If these were unavailable, change scores or mean differences were extracted. For dichotomous outcomes, the number of cases and total sample size were extracted. Safety outcomes included the number of participants reporting any or serious AEs or withdrawn from the study because of AEs.
All interventions of interest were compared through a network meta-analysis. A graph summarized the results of interest, allowing us to easily assess the structure of existing evidence.

Risk of bias within individual studies
The risk of bias of the included studies was evaluated according to the Cochrane Collaboration's Tool for Assessing Risk of Bias. 14 Briefly, randomization and allocation methods (selection bias), completeness of follow-up period/incomplete outcome data (attrition bias), masking of patients (performance bias) and examiners (detection bias), selective reporting (reporting bias), and other forms of bias were classified as adequate (+), inadequate (-), or unclear (?). Based on these domains, the overall risk of bias was categorized as follows: 1) low risk of bias; 2) unclear risk of bias; or 3) high risk of bias.

Summary measures
To inform on comparative efficacy, effectiveness, and safety between all interventions, we conducted a network meta-analysis. We modeled log odds ratios using the conventional logistic regression network meta-analysis setup. 16 The network meta-analysis was based on logistic model with random study effects. 17

Assessment of inconsistency
Consistency was assessed by comparison of the conventional network meta-analysis model for which consistency was assumed with a model that does not assume consistency (a series of pairwise meta-analyses analyzed jointly). If the trade-off between model fit and complexity favors the model with assumed consistency, this model was preferred. Moreover, we calculated the difference between direct and indirect evidence in all closed loops in the network; inconsistent loops were All of the included studies were published between 1991 and 2019. Most of them were large, multicenter, doubleblind, placebo-controlled trials conducted in a variety of countries in the five continents.
The number of randomized patients who received some treatment totaled 94,570 subjects, diagnosed with migraine headaches according to the International Headache Society criteria for migraine. Excluding four studies that did not mention participants' sexes30-33, female participants were the majority, with approximately 84.2%.
The types of treatments varied widely both in dosage and route of administration. The seven types of triptans (almotriptan, eletriptan, frovatriptan, naratriptan, rizatriptan, sumatriptan, and zolmitriptan) appeared among the selected studies. However, sumatriptan was the most common and was found in 18 different variations: it appeared with oral, subcutaneous, nasal spray, iontophoretic transdermal patch (TDS) and suppository administration, combining dosages from 1 mg to 200 mg.
Some treatment arms used triptan in combination therapies with other drugs: frovatriptan 2.5 mg + dexketoprofen 25 mg or 37.5 mg 34 ; naratriptan 2.5 mg orally + rectal suppository of prochlorperazine 25 mg 35 ; rizatriptan 10 mg + acetaminophen 1,000 mg orally 36 , and rizatriptan 10 mg + dexamethasone 4 mg orally 37 . All studies with combination therapy included in this review had an exclusive triptan arm and a placebo comparative arm.
Only 6 studies with metamizole to treat migraine (dosage of 500 mg and 1,000 mg orally and IV) met all the inclusion criteria to be considered in this systematic review. Four studies utilizing the intravenous metamizole route were performed in Brazil 8, [38][39][40] , one in Spain 41 and the last one in Turkey 42,43 , with the last two studies including oral metamizole.
Most studies included in this review were conducted in the adult population and only 13 were carried out with adolescents. [43][44][45][46][47][48][49][50][51][52][53][54][55] Most of the selected studies evaluated improvement or complete relief of headache after 1 hour, 2 hours and in the first 24 hours, although some studies have evaluated different times of symptomatic relief after 30 minutes, in addition to the use of rescue medication in the period. Some studies have evaluated relief of migraine-associated symptoms, such as photophobia, phonophobia, nausea and vomiting. The characteristics of the included studies is in supplementary table (Table 1).
The results of the risk of bias assessment are shown in figure 2. Of all 209 studies, only six (3.0%) provided enough information and were judged to have a low overall risk of bias for all categories evaluated; 146 studies had insufficient information, mainly in the selective reporting domain, so the overall risk of bias was unclear, and 46 presented a high overall risk of bias. Approximately 50% of the studies presented a low risk of selection bias. More than 75% of the studies presented a low risk of performance bias, and around 75% have shown a low risk of detection and attrition bias.
The statistical analysis below was based on random effect models from network meta-analyses. The calculations were made with the netmeta package of the R library, developed by Rücker et al. 56 and based on the methodology described in Schwarzer et al. 57 Since there are no direct comparisons of dipyrone versus any of the triptans, consistency tests were not performed. Figure 3 illustrates the connections between the active substances. The thickness of the edges indicates the weights of the direct comparisons. Studies with more than two treatments were excluded in this analysis. 20,36 There is no evidence of a difference between dipyrone and any triptan.   Figure 5 illustrates the connections between the active substances.

Pain relief after 2 hours of medication
The confidence intervals for differences in pain relief ratios after 2 hours of medication between triptan and placebo versus dipyrone are shown in the forest plot. There is no evidence of a difference between dipyrone and any triptan.   The primary efficacy endpoint was headache relief by two grades on a 5-grade face scale at 2 hours. Sumatriptan (n 53/83; 64%); placebo (n 32/83; 39%) Ahonem et al. 43 Adolescents -12.0 years (SD 2.4; range 6.1 to 16.1 years) n=116 (63 girls and 53 boys) Oral rizatriptan 5 mg (20 to 39 kg); Oral rizatriptan 10 mg (40 kg or more); Placebo Two doses of rizatriptan and a matching placebo were administered at home during three attacks.
Ninety-six patients used all three treatments, 10 used two, and 10 only the first. At 2 hours, the primary endpoint (headache relief by two grades on a five-grade face scale at 2 hours) was reached twice as often after both treatments of rizatriptan [first n=71/96 (74%); second n=70/96 (73%)] as after placebo [n = 35/96 (36%)] (p<0.001). Already at 1 hour, rizatriptan was clearly more effective as headache relief was reported by 50% (n = 48/96) and 55% (n = 53/96) of children after the first and the second dose of rizatriptan, compared to 29% (n= 28/96) after placebo (p=0.004). Rizatriptan was superior at 3 and 4 hours, and the other endpoints also favored rizatriptan. Efficacy of rizatriptan was constant over the two treated attacks, and the findings were similar in children using the dose of 5 and 10 mg. The use of the higher 10 mg adult dose in adolescents caused adverse effects with a frequency comparable to what has been observed in adults. But no serious adverse effects were observed. Adults -Placebo 37.6 years; dipyrone: 44.2 years; n=60 (31 women and 29 men) Dipyrone 1 g in 10 ml saline. Placebo (intravenous injection of 10 ml saline) Patients receiving dipyrone showed a statistically significant improvement (p<0.05) of pain compared to placebo up to 30 min after drug administration. The therapeutic gain was 30% in 30 min and 40% in 60 min. The number of patients needed to be treated for at least one to have benefit was 3.3 in 30 min and 2.2 in 60 min. There were statistically significant reductions in the recurrence (dipyrone = 25%, placebo = 50%) and use of rescue medication (dipyrone = 20%, placebo = 47.6%) for the dipyrone group. Bigal et al. 37 Adults -18 to 55 years n= 35 (all female) Rizatriptan 10 mg + dexamethasone 4 mg Rizatriptan 10 mg Dexamethasone 4 mg Tablets for 6-attack crossover study.

Rizatriptan (RI) and Dexamethasone (DE)
The primary endpoint of this study was 24-hour sustained relief and the secondary was 24-hour sustained pain-free. A total of 1,173 patients were treated with almotriptan or sumatriptan. There were no significant differences between the 2 treatment groups in terms of satisfaction with pain relief; however, patients in the almotriptan group were significantly more satisfied (less bothered) with side effects than those receiving sumatriptan (p=0.016). Connor  Across attacks, headache relief in the 20, 10, and 5 mg drug and placebo groups was experienced 120 minutes postdose by 60%, 54%, 44%, and 32% of patients, respectively (p<0.05 for each sumatriptan nasal spray group vs placebo, for the 10-mg vs 5-mg drug group, and for the 20-mg vs 5-mg drug group). Two thirds of the 20 mg patients treating 3 attacks experienced relief at 2 hours postdose for at least 2 of 3 attacks. Clinical disability scores at 120 minutes in the 20, 10, and 5mg drug and placebo groups reflected no or mild impairment in 70%, 67%, 57%, and 50% of patients, respectively (p<0.05 for the 10 or 20 mg drug group vs placebo group, and for the 20-mg vs 5-mg drug group). Similar efficacy rates were observed for nausea, photophobia, and phonophobia. The incidence of adverse events was not dose related. The most frequently reported adverse event in the active treatment groups was taste disturbance (bad, bitter, or unpleasant). The main result of this study was the significant difference (p=0.001) in efficacy, expressed as headache relief from grade 3 or 2 to grade 1 or 0, within 2 hours after administration of L-ASA and sumatriptan compared to placebo. Placebo was significantly inferior to both verum drugs (p<0.0001). Sumatriptan achieves a higher rate of headache-free patients after 2h, however was associated with a significantly higher incidence of adverse events. The primary efficacy endpoint was headache response (improvement from severe or moderate to mild or no pain) at 2 h -Eletriptan 80 mg 68%, Eletriptan 40 mg 54%, Cafergot 33 % a n d Pl a c e b o 2 1 % ( p < 0 .0 1 fo r a l l c o mp a ri so n s ) . Secondary efficacy measures: pain-free rates at 2 h -Eletriptan 80mg 38%, eletriptan 40mg 28%, Cafergot -10%, Placebo -5%. In the first attack, 221 of 302 participants (73%) did not achieve 2-hour pain relief with sumatriptan and were randomized to treatment of their second attack with almotriptan 12.5 mg or placebo. Of the 198 sumatriptan nonresponders who treated their second attack (99 almotriptan; 99 placebo), 70% had severe headache pain at baseline. Two-hour pain-relief rates were significantly higher with almotriptan compared to placebo (47.5% vs 23.2%; p<0.001). A significant treatment effect for almotriptan was also seen in pain-free rates at 2 hours (33.3% vs 14.1%; P < .005) and sustained freedom from pain (20.9% vs 9.0%; p<0.05). In the second attack, 7.1% of patients in the almotriptan group experienced adverse events compared to 5.1% in the placebo group (P = .77). Díez Zolmitriptan 5mg nasal spray Placebo Treat up to two migraine attacks, each with a single dose of study medication The headache response rate at 2 hours post-dose was 66.2% for the zolmitriptan group, compared with 35.0% for the placebo group (p<0.001). Zolmitriptan nasal spray also produced significantly higher headache response rates than placebo at all earlier timepoints assessed, starting as early as 15 minutes postdose (p<0.001). Similar results were obtained for the analysis of the first attack. Significantly higher pain-free rates were obtained with zolmitriptan nasal spray, compared with placebo, from 15 minutes post-dose onward (p<0.005). Zolmitriptan nasal spray was also significantly superior to placebo for headache response at 4 hours, sustained headache response at 24 hours and sustained pain-free rate at 24 hours. Zolmitriptan nasal spray was well tolerated, with most adverse events being of short duration and mild or moderate intensity. Dowson  The primary efficacy assessment was migraine pain relief, improvement from severe or moderate pain to mild or no pain, at 2 h after treatment. Response rates, stratified for variation in baseline pain levels, for both almotriptan doses were equivalent to sumatriptan and significantly better than placebo. Other efficacy assessments confirmed the equivalence of the almotriptan groups with the sumatriptan group. Almotriptan 12.5 mg was as well tolerated as placebo (P=0.493) and significantly better tolerated than sumatriptan (P<0.001), in terms of the overall incidence of adverse events. There was no statistically significant difference in the incidence of adverse events between almotriptan 25 mg and sumatriptan 100 mg (P=0.376). Dowson  The efficacy results were consistent for attacks one and two: 2 h after treatment in attacks one and two, 53 (73%) and 51 (81%;) of the sumatriptan-treated subjects, respectively, reported headache relief (reduction of a severe or moderately severe headache to a mild or no headache), compared with 27 (31%) and 18 (29%) of the placebo-treated subjects (p<0.001). Within 24 hours of treatment in attack one, 28 (53%) and 14 (52%) of the initial responders to sumatriptan and placebo, respectively, experienced headache recurrence. Consistently response -The percentage of subjects who reported a response in 2 out of 3 headaches was significantly higher on E40 (66% -74/112) and E80 (72% -72/100) compared with placebo (15% -5/34) (p<0.001).

Fernandes et al. 39
Adults -Dipyrone -male 32.2 (DP 12.6) female 34.3 (DP 9.6); Metoclopramide -male 29 (DP 8.6) female 35.8 (DP 6.4) years. n=27 (14 male and 13 female) Dipyrone IV Metoclopramide IV One migraine attack Among male patients, it was observed that the percentage variation in pain scores before and after treatment was, in general, greater in patients treated with metoclopramide than in patients treated with dipyrone (p=0.052). No difference was seen between female patients (p=0.748).

Ferrari et al. 100
Adults -Sum100 mg+Sum100 mg -40.5 ±10.5 Sum100 mg+Placebo -40.5 ± 10.5 n=1246 (1021 female and 225 male) Sumatriptan 100 mg Placebo Tablets to treat up to three migraine attacks The primary objective of the study was to compare headache improvement from moderate or severe at time 0 to none or mild at 4 h on sumatriptan 100 mg + sumatriptan 100mg and 100 mg + placebo.
Headache improvement -2 h after the first dose of 100 mg sumatriptan -56% Sum+Sum vs 55% Sum+Plac. Headache improvement at 4 h -77% Sum+Sum vs 80% Sum+Plac. Headache recurrence -25% Sum+Plac vs 22% Sum+Sum Freitag et al. 101 Adults -Isometheptene Combination-40.9 ± 9.6 Sumatriptan succinate -43.3 ± 9.6 n=126 (112 female and 14 male) Isometheptene Combination (65 mg isometheptene, 100 mg dichloralphenazone, and 325 mg acetaminophen) -5 capsules taken over several hours Sumatriptan Succinate 25 mg (2 doses) capsules Placebo Single migraine attack Primary outcome: % of patients with mild or no headache at 2 hours and/or 4 hours after the first dose of study medication, and % of patients who did not suffer a recurrence of headache within 24 hours of the initial dose. No or mild headache at 4 hours -Isometheptene combination 76% vs Sumatriptan 80% (X2 = 0.22). Patients without headache or had only a mild migraine at 24 hours -82% in both groups.

Rizatriptan 10 mg ODT Placebo Single migraine attack
There was a greater percentage of patients with elimination of nausea at 2 hours (primary efficacy endpoint) in the rizatriptan ODT group compared with the placebo group (70.3% vs 62.0%), P = 0.165, odds ratio (95% CI) = 1.45 (0.86, 2.46) -not statistically significant. There was a significantly greater percentage of patients who achieved 2-hour pain relief (secondary efficacy endpoint) with rizatriptan ODT compared with placebo (69.7% vs 54.3%), P = .012, odds ratio (95% CI) = 1. 94 16.1%, and 0.6% of placebotreated patients. The differences in level of functional disability between the 2 treatment groups were statistically significant at 2 hours (P =0.007; Cochran-Mantel-Haenszel, stratified by center) and at 4 hours (P <0.001). Resolution of pain was associated with a normal level of function; at 2 hours posttreatment, 91.7% of patients in the total population who achieved pain-free reported normal function compared with 44.8%, 8.0%, and 0% of patients with mild, moderate, and severe pain, respectively. The absence compared with the presence of photophobia, phonophobia, and nausea at 2 hours also was associated with less disability (P < 0.0001 for each symptom). Treatment with almotriptan compared with placebo resulted in consistently better 24-hour MQoL scores with significant results for all 3 migraine headache attacks in the social function and feelings/concern domains. A logistic regression model determined that pretreatment functional level (P = 0.0117), pretreatment pain intensity (P = 0.0089), and pretreatment MIDAS score (P = 0.0152) were significant covariates of the proportion of patients who achieved normal function at 2 hours posttreatment. Pain relief 2h -RA 90%, R 77%, A 70%, P 46%. RA was statistically superior to A and P. Pain-freedom 2h -RA 54%, R 40%, A 26%, and p 15%. RA was statistically superior to A and P. Pain relief sustained 24h -RA 62%, R 53%, A 42% and P 15%. RA was statistically superior to P only. RA was statistically superior to A for absence of phonophobia (85% vs 60%, P = .009) and statistically superior to P for absence of phonophobia (85% vs 67%, P = .039), absence of nausea (92% vs 72%, P = .021), and absence of functional disability (65% vs 41%, P = .024). At 2 hours -59% of metoclopramide subjects and 35% of sumatriptan subjects were pain-free.

Friedman et al. 105
Adults -TMB/DPH 34 (9.7) and Sumatriptan 32 (8.9) years n=40 (37 female and 3 male) Trimethobenzamide 200 g + diphenhydramine 25 mg (TMB/ DPH) as a single intramuscular injection Sumatriptan 6mg SC By 2 hours sumatriptan subjects had improved by a mean of 6.1 and the TMB/DPH subjects had improved by a mean of 4.4 (95% CI for difference of 1.7: −0.1 to 3.4). By 24 hours after medication administration, sumatriptan subjects had a mean improvement from baseline of 4.9 as compared to 5.3 for TMB (95% CI for difference of −0.4: −2.4 to 1.6). The need for rescue medication was comparable between the groups. No serious or frequent adverse effects were noted in either group Friedman  A reduction in headache pain from moderate/severe at baseline to mild or no pain 2 h after taking study medication with no moderate or severe recurrence within 24 h (Primary Endpoint) was reported by 39%, 38% and 32% of patients treated with zolmitriptan, sumatriptan and placebo, respectively, with no significant difference between treatment groups. In patients with moderate headache at baseline, complete response was significantly greater following zolmitriptan than after placebo (48% vs. 27%; P50.01); there was no significant difference between sumatriptan and placebo groups (40% vs. 27%). In patients with severe baseline headache (where a greater reduction in headache intensity is required for a headache response), there was no significant difference between any groups in complete headache response rates. Secondary objectives were to compare headache and pain-free response rates at 1, 2 and 4 h post-dose. In addition, other secondary objectives were to compare the proportion of patients whose migraine-associated symptoms were effectively treated, use of escape medication after 2 h, incidence of recurrence, meaningful migraine relief, time to meaningful migraine relief, and degree of activity impairment at 1, 2, 4 and 24 h. For these secondary endpoints, active treatment groups were significantly superior to placebo for: 1-, 2-and 4-h headache response (e.g. 2-h headache response rates: zolmitriptan 59%; sumatriptan 61%; placebo 44%; P=0.01 vs. placebo); pain-free response rates at 2 and 4 h; alleviation of nausea and vomiting; use of escape medication and restoration of normal activity. The incidence of adverse events was similar between zolmitriptan and sumatriptan groups but was slightly lower in the placebo group.

ASAA
Peres MFP, Scala WAR, Salazar R Comparison between metamizole and triptans for migraine treatment: a systematic review and network meta-analysis AAC group experienced significantly greater pain intensity reduction or pain relief than those taking S50 or placebo. Pain intensity reduction and pain relief score for Sumatriptan 50mg group were higher than placebo group, but not to a statistically significant degree. Reduction of headache pain intensity from moderate/severe to mild/none -Sumatriptan 50mg group was significantly superior to AAC (30 minutes postdose); AAC group was significantly greater than in the Sumatriptan 50mg group (2, 3 and 4 hours postdose). The response rate of AAC versus placebo was significant from 90 minutes. The rate of response in the Sumatriptan 50mg group was greater than that in the placebo group at all time points, but not to a statistically significant degree. Utilization of the rescue medication showed statistically significant difference between Sumatriptan 50mg group (11.9% subjects) versus the AAC group (1.5% of subjects), at 4 hours postdose. Goldstein  Two-hour pain free rates were 16%, 40%, and 50% in the placebo group, sumatriptan 50 mg group, and the sumatriptan 100 mg group respectively (p < 0.001, active treatment groups vs placebo). The percentage of subjects who sustained a pain-free response for both 50 mg and 100 mg sumatriptan groups (24% and 27%) was significantly higher than in the placebo group (6%). After 4 hours, 25% of the 50 mg sumatriptan group and 13% of the 100 mg sumatriptan group experienced worsening of their migraine pain, compared to 46% of placebo patients (both p<0.001).

Jensen et al. 127
Adults -43 years (range 20-65) n=138 (125 female and 13 male) Sumatriptan 6 mg subcutaneous Placebo Sumatriptan 6 mg sc was significantly better than placebo at 30, 60, 90 and 120 min after injection in relieving moderate or severe headache to mild or none as well as relieving any headache to none. At 60 min after injection, the treatment response rate was 61% for sumatriptan and 6% for placebo.
During the following open-phase trial of four attacks treated with sumatriptan, treatment response rates were 68-74%. During the total of 538 attacks treated, 12 attempts at using the self-injector failed. In the double-blind and open phases, 81% and 90% of patients respectively found the device easy or very easy to use. Adverse effects were benign and short lasting, but led 7 patients to discontinue the study. The onset of significant pain relief was apparent 15 minutes after treatment with zolmitriptan nasal spray. At 1 hour after the dose, zolmitriptan nasal spray produced a higher headache response rate than did placebo (58.1% vs 43.3%). Zolmitriptan nasal spray was also significantly superior to placebo in improvement in pain intensity, pain-free rates, sustained resolution of headache, and resolution of associated migraine symptoms. Return to normal activities was also consistently faster with zolmitriptan nasal spray than with placebo, with less use of any escape medication. Treatment with zolmitriptan nasal spray was well tolerated. Linder   Metamizole (0.5 and 1 g) oral Acetylsalicylic acid (1 g) oral Placebo The pain intensity reduced steadily for all three active treatments in comparison with placebo up to 4h after administration. The analgesic efficacy of 0.5 and 1 g metamizol vs placebo was highly statistically significant for sum of pain intensity differences, maximum pain intensity difference, number of patients with at least 50% pain reduction, time to 50% pain reduction, maximum pain relief and total pain relief. A trend towards an earlier onset of a more profound pain relief of 0.5 and 1 g metamizol over 1 g Acetylsalicylic acid was noticed. Adverse events were experienced during the treatment phase of the study in all groups, but differences statistics were not observed. Global assessment of tolerability by the patients was good or satisfactory in more than 90% of all patients.

Massiou et al. 143
Adults -aged 18 to 65 years n=257 (Female only) Naratriptan 2.5 mg Placebo single migraine attack A higher percentage of subjects in the naratriptan group (58%) reported complete pain relief 4 h after medication than in the placebo group (30%) (P < 0.001). Significant differences between the naratriptan and placebo groups and in favor of naratriptan were also found for: total pain relief at 2 h (P = 0.004), sustained pain-free response within 4-24 h (P < 0.001), absence of all associated symptoms at 2 and 4 h (P =0.004), ability to work and carry out daily activities at 2 h (P = 0.036), and patient overall satisfaction (P <0.001). Mathew  Pain-free response at 2 hours after early treatment was noted in 70% of attacks in the rizatriptan group and in 22% in the placebo group (P < .01). Pain-free response at 1 hour occurred in 45% and 8% attacks, respectively (P < .01). When the attacks were categorized by headache severity at the time of treatment, the pain-free response at 2 hours was higher for mild attacks than for moderate or severe attacks (P < .01). Sustained pain-free response after treatment was significantly higher for attacks treated with rizatriptan (60%) than for those treated with placebo (17%) (P < .001). Adverse events were reported in 62 attacks (29%) in the rizatriptan group and 15 attacks (14%) in the placebo group. The primary outcomes for these analyses were migraine pain intensity and migraine related disability. Average pain intensity for Sumatriptan 22 mg nasal treated attacks was significantly lower than sumatriptan treated ones at all time points from 10 to 90 minutes (P < .05 for all). The mean portion of the models showed that Sumatriptan 22 mg nasal had significantly lower disability from 10 to 90 minutes. Patients scored the severity of their headache on a 100-mm visual analog scale (VAS) of pain prior to medication and again 1 hour after medication. Differences between initial and 1-hour scores were analyzed. Before treatment, no difference existed between the groups in the intensity of headache. One hour after medication, the sumatriptan group had a decrease in pain score of 22.937 mm and the ketorolac group a decrease of 71.462 mm on the VAS. The decrease in pain score with ketorolac was significantly greater than that with sumatriptan (P < 0.001).

Miljkovic et al. 150
Adults -18 to 64 years n=201 (168 female and 33 male) Sumatriptan tablet Ergotamine combination tablet (propyphenazone, caffeine, camylofin chloride, mecloxamine citrate) Placebo Higher percentage of patients was completely free of the headache 2 hours after dose administration in the ergotaminebased medication group compared to the sumatriptan group, regardless whether all (51.12 % vs 33.70 %) or only repeated attacks were taken into account (50.91 % vs 23.73 %). The salvage therapy (diclofenac) utilization rate was also lower in the ergotamine-based medication group (relative risk 0.61).
The time to a pain-free response was significantly (P < .05) higher with Indomethacin 25 mg + prochlorperazine 4mg and caffeine 75 mg suppository than with sumatriptan in the first, second, and total attacks. Headache relief rates in the total attacks at 2 hours postdose were 71% with Indomethacin 25 mg + prochlorperazine 4mg and caffeine 75 mg suppository and 65% with sumatriptan, without any statistically significant difference between the drugs.

ASAA
Peres MFP, Scala WAR, Salazar R Comparison between metamizole and triptans for migraine treatment: a systematic review and network meta-analysis Pain intensity was significantly lower in the propofol group 30 minutes after treatment (P = 0.001); however, after 1 and 2 hours, there were no significant differences between the groups. The need for antiemetic therapy and the recurrence of symptoms were significantly lower in the propofol group (P = 0.045 and P = 0.001, respectively). Headache-free Naratriptan 1mg 50% versus Placebo 25%, (P=.003). More patients treated with naratriptan 1 mg were headache free compared with placebo (23% versus 8%), although statistical tests were not performed. Significantly more patients treated with naratriptan 1 mg reported menstrual associated migraine 50% or less compared with placebo-treated patients. Patients treated with naratriptan 1 mg, also had significantly fewer menstrual associated migraine days compared with placebo-treated patients. Sumatriptan (20 mg plus 20 mg) was more effective than placebo at relieving headache, defined as reduction in severity from moderate or severe to mild or none, at 60 and 120 min. At 120 min, 75% of patients in the sumatriptan group reported headache relief, compared with 32% of patients in the placebo group (p<0.001); 53% of patients in the sumatriptan group were completely pain-free compared with 11% in the placebo group. Nausea incidence was significantly lower in sumatriptan group compared with placebo at both 60 min (17 vs. 43%; p=0.014) and 120 min (14 vs. 38%; p=0.021). Photophobia was significantly lower in sumatriptan group, compared with placebo at 60 min (28 vs. 57%; p=0.013) and 120 min (19vs. 51%; p=0.005). Sumatriptan was significantly more effective at reducing functional disability of patients at 30 min (p=0.024) and at 60 and 120 min (p<0.001). However, similar number of patients reported migraine recurrence, within 24 h in both treatment groups. Sumatriptan was also significantly more effective at reducing the incidence of nausea (p < 0.001), vomiting (p < 0.01) and photophobia/phonophobia (p < 0.001) 2h after treatment, and fewer patients on sumatriptan (24%) than on Cafergot (44%, p < 0.001) required other medication after 2h. The overall incidence of patients reporting adverse events was 45% after sumatriptan and 39% after Cafergot; the difference was not significant. The total number of attacks relieved (severe or moderate pain reduced to mild or no pain) at 2 h post-dose was significantly higher (P<0.001) after treatment with almotriptan 6.25 or 12.5 mg compared with placebo (60% and 70% vs. 38%, respectively). Moreover, a consistent response was achieved across and within patients for almotriptan 6.25 or 12.5 mg compared with placebo (pain relief in at least 2 out of 3 attacks within 2h for 64% and 75% vs. 36%, respectively) and less than one-third of the patients relapsed within 24h.
Almotriptan was well tolerated with no significant differences between the almotriptan and placebo treatment groups in the percentage of patients reporting adverse events. The headache response rates with zolmitriptan doses ≥ 2.5 mg were 44 to 51% at 1 hour, 65 to 67% at 2 hours, and 75 to 78% at 4 hours (all significantly superior to placebo). Also, zolmitriptan effectively relieved migraine-associated symptoms such as nausea, photophobia and phonophobia, and reduced activity impairment.

ASAA
Rates of headache recurrence, headache persistence, and use of escape medication were lower with zolmitriptan doses ≥ 2.5 mg than with placebo. In patients with persistent or recurrent headache, a second zolmitriptan dose effectively treated both headache and nonheadache symptoms. There was no statistically significant improvement between zolmitriptan 10 mg (2 x 5 mg tablet) and placebo for the primary efficacy variable headache response at 2 hours, nor any of the secondary variables tested. Two-hour headache response rates were 54%, 53%, and 57% for zolmitriptan 10, 5, and 2.5 mg, respectively, and 58% for placebo. Two-hour pain-free rates were 25%, 19%, and 23% for zolmitriptan 10, 5, and 2.5 mg, respectively, and 20% for placebo. Zolmitriptan was well tolerated, with a tolerability profile similar to the pattern seen in adults. Russell  The primary efficacy endpoint was headache relief 120 minutes after the first administration of study drug. Headache relief -62 to 63% patients in the sumatriptan 20-mg, 43 to 54% sumatriptan 10-mg, 29 to 35% of placebo (p <0.05). Pain-free 2h -31 to 32% sumatriptan 20-mg groups, 20 to 23% sumatriptan 10-mg, 4 to 20% placebo (p < 0.05). Incidence of nausea, photophobia, and phonophobia reduced after sumatriptan 20mg (p < 0.05), sumatriptan nasal spray 10 mg compared with placebo reduced the incidence of nausea (p < 0.05).

Ryan et al. 26
Adults Response at 2 hours range from 27% to 46% for frovatriptan compared with 21% to 27% for placebo. Likewise, at 4 hours, frovatriptan was consistently significantly more effective than placebo to provide headache relief. Response forfrovatriptan ranged from 56% to 65% compared with 31% to 38% for placebo (p<0,001). Frovatriptan was also significantly superior to placebo at rendering patients pain-free. At 2 hours, the proportion of patients pain-free was 9% to 14% for frovatriptan compared with 2% to 3% for placebo (p<0,001). At 4 hours post dose, 27% to 32% of patients takins frovatriptan were pain-free compared with 9% to 14% in the placebo group (p<0,001). Ryan  Response at 2 hours ranged from 27% to 46% for frovatriptan compared with 21% to 27% for placebo. Likewise, at 4 hours, frovatriptan was consistently significantly more effective than placebo to provide headache relief. Response for frovatriptan ranged from 56% to 65% compared with 31% to 38% for placebo (p<0,001). Frovatriptan was also significantly superior to placebo at rendering patients pain-free. At 2 hours, the proportion of patients pain-free was 9% to 14% for frovatriptan compared with 2% to 3% for placebo (p<0,001). At 4 hours post dose, 27% to 32% of patients taking frovatriptan were pain-free compared with 9% to 14% in the placebo group (p<0,001). Frovatriptan 2.5mg tablet Placebo Patients could take up to two doses per attack for headache recurrence within 24 hours of the first dose Response at 2 hours ranged from 27% to 46% for frovatriptan compared with 21% to 27% for placebo. Likewise, at 4 hours, frovatriptan was consistently significantly more effective than placebo ar providing headache relief. Response for frovatriptan ranged from 56% to 65% compared with 31% to 38% for placebo (p<0.001). Frovatriptan was also significantly superior to placebo at rendering patients pain-free. At 2 hours, the proportion of patients pain-free was 9% to 14% for frovatriptan compared with 2% to 3% for placebo (p<0.001). At 4 hours post dose, 27% to 32% of patients taking frovatriptan were pain-free compared with 9% to 14% in the placebo group (p<0.001). Headache response rates were 12% at 1 hour and 31% at 2 hours for placebo; 24% at 1 hour and 50% at 2 hours for sumatriptan 50 mg; 27% at 1 hour and 53% at 2 hours for sumatriptan 100 mg; 30% at 1 hour and 64% at 2 hours for eletriptan 40 mg; and 37% at 1 hour and 67% at 2 hours for eletriptan 80 mg. More patients receiving eletriptan 80 mg achieved a 1-hour headache response than did patients receiving sumatriptan 50 mg (p < 0.05). All doses of eletriptan were superior to sumatriptan at 2 hours for headache response and complete pain relief (p < 0.05). Significantly more patients on eletriptan 80 mg achieved headache response in all attacks than did patients receiving sumatriptan. Eletriptan 40 mg was superior to both sumatriptan doses in functional improvement (p < 0.005). The 40-and 80-mg doses of eletriptan were significantly more effective than placebo or sumatriptan in reducing the associated migraine symptoms of nausea, photophobia, and phonophobia after 2 hours. The 40-and 80-mg doses of eletriptan were significantly superior to oral sumatriptan or placebo in achieving and sustaining both headache response and pain-free response at 24 hours. The superior efficacy of both eletriptan doses was associated with higher rates of patient acceptability than sumatriptan 50 mg (p < 0.05). Eletriptan and sumatriptan were well tolerated.

Sang et al. 176
Adults -mean age 40 ± 9 years. n=44 (20 male and 24 female) Ly293558 (nonselective AMPA/ KA (GluR5) receptor antagonist with 1.2 mg/kg IV Sumatriptan 6 mg SC Placebo Simgle migraine attack The primary efficacy variable was the headache response rate, i.e. headache score improvement from moderate/severe at baseline to mild/none at 2 h. Response rates were 69% for LY293558 (P = 0.017 vs. placebo), 86% for sumatriptan (P < 0.01 vs. placebo) and 25% for placebo. LY293558 and sumatriptan were superior to placebo (P < 0.01 for all comparisons) on all other measures of improvement in pain and migraine associated symptoms. Fifteen percent of patients who took LY293558 reported adverse events, 53% patients who took sumatriptan and 31% of those who received placebo reported adverse events.

Santanello et al. 177
Adults -Rizatriptan 10mg -36.8 (SD 9); Rizatriptan 5mg -37.6 (SD 8.2); Rizatriptan 2.5mg -38.7 (SD 9.1); Placebo -39.7 (SD 9.7) years n=247 (222 female and 25 male) Rizatriptan 2.5, 5 and 10mg Placebo Oral One migraine attack Statiscally significant mean improvements were observed for those treated with rizatriptan 10mg compared with those treated with placebo on three of five domains: social functioning (p=0,007), migraine symptoms (p=0.005), and feeling/concerns (p=0.015). Patients who took the 5-mg and 10mg rizatriptan doses were significantly less disabled as 2h than those who took placebo (p=0.003); however, the patients who took 2,5mg rizatriptan remained about as functionally disabled as patients on placebo. Savi et al. 178 Adults -37±9 years n=125 (99 female 26 male) Frovatriptan 2.5 mg Rizatriptan 10 mg Capsules Treat 1-3 attacks Patient's preference for one drug or the other did not differ between the study treatments. Frovatriptan was chosen mainly because of the rapid speed of action (71% of patients), good tolerability (42% of patients), and reduction in pain severity (33%). A relevant result of study was that recurrence rate within 48 h were significantly lower under frovatriptan than under rizatriptan. These differences may be explained by the different pharmacokinetics of the two drugs. Frovatriptan has a time to maximum concentration typically of 2 to 3 h, but the longest half-life among triptans, greater 5-HT1B binding receptor potency, and multiple pathways metabolism. The headache recurrence was significantly less frequent with frovatriptan than under rizatriptan.
In the post-hoc analysis, transdermal sumatriptan was markedly superior to placebo for pain relief and freedom from pain, nausea, photo and phonophobia at 1 and 2 hours post-activation. Sumatriptan 100 mg tablet oral Placebo Three migraine attack Headache was relieved by the first sumatriptan dose in about 70% of patients, but the second dose did not produce significantly more relief than placebo, either in non-responders or in the group as a whole, nor did it reduce other symptoms (photophobia, nausea, vomiting, etc.) at 8 h, or influence the incidence of headache recurrence.
The drug was well-tolerated, and a further single dose was effective in treating recurrence after initial relief. Of those patients who treated at least one attack and expressed a view of the treatment, 80% (n = 1056) said that they would take the medication again. Use of frovatriptan reduced the occurrence of menstrually associated migraine. The incidence of menstrually associated migraine during 6-day was 67% for placebo, 52% for frovatriptan 2.5 mg QD, and 41% for frovatriptan 2.5 mg BID. Both frovatriptan regimens were superior to placebo (p < 0.0001), and the BID regimen was superior to the QD regimen (p < 0.001). Both frovatriptan regimens also reduced menstrually associated migraine severity (p < 0.0001), duration (p < 0.0001), and the use of rescue medication (p < 0.01 QD; p < 0.0001 BID) in a dose-dependent manner.
The incidences of the associated symptoms of migraine were significantly lower at 2 hours following sumatriptan 50 mg plus naproxen sodium 500 mg treatment versus placebo (P < .001).
The frequencies and types of adverse events reported did not differ between treatment groups, with dizziness and somnolence being the most common. During the first hours after intake of the study medication, the objective behavioral parameeters showed no significant effect time and no significant differences between naproxen and naratriptan, but naratriptan showed improve of symptoms and the interval between treatment and relief was significantly shorter after intake of naratriptan.

ASAA
Peres MFP, Scala WAR, Salazar R Comparison between metamizole and triptans for migraine treatment: a systematic review and network meta-analysis The two-hour pain-free response rate was higher in patients who treated persistent headache of any intensity with any dose of zolmitriptan compared with placebo. However, there were no statistical differences between the two-hour headache response rate for 5mg, 2,5 mg of zolmitriptan or placebo. Headache relief rates after rizatriptan 10 mg were consistently higher than sumatriptan at all time points up to 2 hours, with significance at 1 hour (37% versus 28%, P = 0.010). All active agents were significantly superior to placebo with regard to headache relief and pain freedom at 2 hours (P < or = 0.001).
The primary efficacy endpoint of time to pain relief through 2 hours demonstrated that, after adjustment for age imbalance, rizatriptan 10 mg had earlier onset than sumatriptan 100 mg (P = 0.032; hazard ratio 1.21). Rizatriptan 10 mg was also superior to sumatriptan on pain-free response (P = 0.032), reduction in functional disability (P = 0.015), and relief of nausea at 2 hours (P = 0.010).
Significantly fewer drug-related clinical adverse events were reported after rizatriptan 10 mg (33%, P = 0.014) compared with sumatriptan 100 mg (41%). Reduction in headache severity was observed at 2 hours (P < .001) and at 4 hours (P <.001) from headache onset, with no difference between the two treatment regimens (P = .34). A significant decrease in clinical disability at 2 hours (P < .001) and at 4 hours (P < .001) was observed, with no difference between the two treatment regimens (P = .28). The pain-free state at 4 hours was reported in a higher proportion with the naratriptan/ placebo regimen (50% vs 25%), but the trial size would need to be doubled to significantly prove the endpoints. Resolution of adverse effects was similar with both regimens at 2 hours and at 4 hours, although nausea resolved more often for those using the naratriptan/prochlorperazine regimen.

ASAA
Peres MFP, Scala WAR, Salazar R Comparison between metamizole and triptans for migraine treatment: a systematic review and network meta-analysis Visser et al. 208 Adults -Placebo: 39 (SD 10); Sumatriptan 1mg: 41 (SD 11); Sumatriptan 2mg: 40 (SD 11); Sumatriptan 3mg: 39 (SD 10). n=685 (165 male and 520 female) Sumatripan 1, 2 and 3 mg SC Placebo One migraine attack By 30 minutes post dose 17% (95% CI 8% to 27%) more patients had improved with 1 mg sumatriptan, 22% (95% CI 13% to 32%) with 2 mg sumatriptan and 34% (95% CI 24% to 44%) with 3 mg sumatriptan than with placebo (p < 0.001 for all three comparisons versus placebo). The number of patients who were improved increased significantly with increasing dose (p < 0.002; chi-square test for trend). Complete resolution of pain was obtained at 30 min by 5% of placebo-treated patients, 9% of patients treated with 1 mg sumatriptan and by 14% treated with 2 mg or 3 mg sumatriptan, respectively. Visser et al. 209 Adults -range: 18  The proportion of patients with headache relief was 18% for placebo; 46% for sumatriptan; and 52% for 10-mg, 56% for 20mg, and 67% for 40-mg rizatriptan. All differences with placebo were statistically significant (P<.001), and 40-mg rizatriptan was superior to sumatriptan (P=.01). The proportion of patients who became free of pain at 2 hours was 3% for the placebo-treated group; 22% for the sumatriptantreated group; and 26%, 35%, and 47% for the group of patients who took the 10-, 20-, and 40-mg doses of rizatriptan, respectively (all differences with placebo, P<.005; 40-mg rizatripan vs sumatriptan, P=.001). The recurrence of headache within 24 hours was found to be equal across all treatment groups-approximately 40%. Adverse events (most commonly short-lasting mild or moderate dizziness and drowsiness) occurred more frequently after a 40-mg dose of rizatriptan was given than after other treatments. Sumatriptan 5mg, 10mg, or 20mg nasal spray Placebo Headache relief 1-hour postdose was significantly greater for patients using 10 mg (56%) and 20 mg (56%) of sumatriptan nasal spray compared with placebo (41%). Headache relief 2 hours postdose was significantly greater for patients using 5 mg of sumatriptan nasal spray (66%) compared with placebo (53%) and approached statistical significance for 20 mg (63%) compared with placebo (53%). Complete relief 2 hours postdose was significantly greater for patients using 20 mg of sumatriptan nasal spray compared with placebo (36% vs 25%, respectively). Photophobia and phonophobia were significantly reduced 2 hours postdose for sumatriptan nasal spray (20 mg), compared with placebo (36% vs 48% and 25% vs 44%, respectively). Taste disturbance was the most commonly reported adverse event (2%, 19%, 30%, and 26% for placebo, 5 mg, 10 mg, and 20 mg, respectively). No drug-related serious adverse events or clinically relevant changes were reported. Winner et al. 55 Adolescents -mean age 14 years n=296 (161 female and 135 male) Rizatriptan 5mg oral Placebo The percentage of patients pain-free at 2 hours was 32% for rizatriptan 5 mg versus 28% for placebo (P.474). The percentage of patients with pain relief (reduction of predose pain intensity to mild or none) at 2 hours was 66% for rizatriptan versus 56% for placebo (P.079). Compared with placebo, rizatriptan significantly improved functional disability at 1.5 and 2 hours, and nausea at 1 and 1.5 hours. Rizatriptan 5 mg was well tolerated. The most commonly reported adverse events among patients receiving rizatriptan were dry mouth, dizziness, asthenia/fatigue, nausea, and somnolence.

ASAA
Peres MFP, Scala WAR, Salazar R Comparison between metamizole and triptans for migraine treatment: a systematic review and network meta-analysis Propacetamol (1 g) IV Rizatriptan 5 mg oral Propacetamol showed superior efficacy at 1 h and there was no significant difference at 30 min or at 2 h. This indicates that propacetamol is at least as effective as rizatriptan in the treatment of acute migraine attacks.

ASAA
Peres MFP, Scala WAR, Salazar R Comparison between metamizole and triptans for migraine treatment: a systematic review and network meta-analysis

Discussion
Our study suggests that overall response to triptans is as effective as that observed with dipyrone in acute migraine treatment.
Several studies involving triptans have evaluated the efficacy of different doses compared to placebo or included an evaluation of comparative efficacy between different triptans or doses. In general, all types of triptan were more effective than placebo in relieving migraine, with a good safety standard, although some drugs have achieved similar results as those for placebo. Attention to exceptional positive results for placebo in migraine treatment in some double-blind studies should be given, especially in the adolescent population. Symptoms related to migraine (nausea and vomiting, phonophobia and photophobia) also had a good response with triptans.
In general, triptans were effective in relieving associated symptoms and reducing clinical disability compared to placebo. A poor response to one triptan does not predict a poor response to other agents belonging to the class.
Regarding dipyrone (metamizole) results, Bigal showed that the number of patients required to be treated with dipyrone 1 g by intravenous injection compared to placebo for at least one to benefit was 3.3 in 30 min and 2.2 in 60 min. There were statistically significant reductions in recurrence (dipyrone = 25%, placebo = 50%) and use of rescue medication (dipyrone = 20%, placebo = 47.6%) for the dipyrone group. 38 A few studies have evaluated the restoration of functional ability after a migraine crisis and, to a lesser extent, lost time from work. A good number of studies evaluated the possibility of returning to normal functions or the number of patients who were able to return to normal activities after an average of 2 hours from initial treatment for a migraine episode. All studies involved triptans and no study was performed with dipyrone. 24 Barbanti et al evaluated equivalent work time loss after a migraine attack, and the results showed 1.9 ± 2.3 and 2.5 ± 4.7 hours lost from work for sumatriptan 100 mg and 50 mg, respectively, compared with 3.5 ± 4.3 for placebo. Sumatriptan 100 mg was also able to better restore functional ability. 64 Freitag et al. (REF) evaluated normal function disability, bed rest required, and ER/hospitalization resulting from a migraine attack in order to compare almotriptan and placebo responses at 2h-and 4h-posttreatment. The study showed that pain resolution was associated with a normal level of function, and the absence of photophobia, phonophobia, and nausea at 2 hours was also associated with less disability. In the study, treatment with almotriptan compared with placebo resulted in consistently better 24hour quality of life scores, with restored social function. A logistic regression model determined that pretreatment functional level and pretreatment pain intensity were significant covariates of the proportion of patients who achieved normal function at 2 hours posttreatment. 103 Dasbacj et al. 84 demonstrated that rizatriptan decreased the total number of lost work hours by 1.1h per treated migraine attack compared with placebo.
Most studies that evaluated migraine in the menstrual period involved triptans. 97 Some studies have associated hormonal drugs and mainly NSAIDs with the use of triptan in one of the tested arms, with good therapeutic results in general, especially when there was an association of a triptan with a NSAID, with superior results when compared to the drug alone. Naproxen, ketoprofen and ibuprofen were the most common NSAIDs evaluated in the studies. 20,30,33,34,48,86,106,187 Tullo et al. evaluated the factors that influenced the selection of a treatment for migraine, comparing frovatriptan and zolmitriptan in the selected study, and found the following order of priority: 1) speed of action; 2) reduction in pain intensity and 3) absence of side effects. 207 On the other hand, Savi et al. 178 demonstrated the following order of choice by patients: rapid speed of action, good tolerability and reduction in pain severity, being decisive for the selection of frovatriptan over rizatriptan. Although these studies have evaluated triptans, rapid pain relief appears to be the main attribute of drug selection for migraine relief. 92,171,207 Regarding the question presented in this study: "what is the evidence for the efficacy and safety of metamizole for the treatment of migraines compared with triptans?" The result is that overall response to metamizole is as effective as that observed with triptans in acute migraine treatment. The second point of evaluation in this systematic review was: "how effective are those treatments in improving cognitive dysfunction in patients with migraine?" Unfortunately, cognitive improvement is not a goal evaluated in most studies included in the review. A few triptan studies showed that pain resolution was associated with a normal level of function, and also a logistic regression model determined that pretreatment functional level and pretreatment pain intensity were significant covariates of the proportion of patients who achieved normal function at 2 hours posttreatment. There are no data regarding cognitive dysfunction improvement related to metamizole utilization, so it may just be an inference related to metamizole.
This systematic review involved different forms of administration and doses of metamizole and triptans, which allowed us to have a complete and comprehensive view of studies involving both studied medications in migraine treatment, but limits some more direct comparisons between doses and routes of administration. Most studies performed with triptans utilized oral administration and most studies with metamizole in this review utilized the intravenous route of administration.
No direct comparisons between metamizole and triptans have been performed in a controlled and randomized clinical study and most studies involving triptans have been conducted in European countries and the US.
The main weakness of this systematic review and metaanalysis is the small number of studies involving metamizole included. The literature on metamizole is scarce. In the setting of the present analyses, only 5 articles with metamizole had a placebo arm and the estimates obtained were all indirect. This fact is directly related to the absence of drug availability in expressive markets, such as the US and some European countries. Studies with metamizole included in this review were limited to Brazil, Spain and Turkey. 8,37,[39][40][41][42] Despite the adverse event of agranulocytosis being the main reason for metamizole withdrawal from the market in some countries, this health risk was not proven true in the pharmacovigilance data and other scientific evidence generated in countries that maintained product commercialization. [215][216][217] The data did not show a significant difference between metamizole and triptans in neither pain relief nor pain absence 2 hours after medication. In support of relief within 24 hours after medication, eletriptan, rizatriptan and zolmitriptan showed statistically different proportions from metamizole. There is no evidence of a difference between metamizole and triptans in absence of pain 24 hours after medication.
Considering the equivalence of therapeutic benefit and adverse events with triptans, especially cardiovascular ones, in addition to pharmacoeconomic aspects, as metamizole is far cheaper than triptans, metamizole could be a good medicine option for migraine treatment.

Conclusion
Metamizole may be equally effective as triptans in acute migraine treatment, with a good tolerability profile and a potentially better cost-benefit ratio with significant implications to healthcare policies. More studies are necessary to confirm our results.