Why is the therapeutic effect of acute antimigraine drugs delayed? A review of controlled trials and hypotheses about the delay of effect

Abstract

In randomised controlled trials (RCTs) of oral drug treatment of migraine attacks, efficacy is evaluated after 2 hours. The effect of oral naratriptan 2.5 mg with a maximum blood concentration (T_max) at 2 hours increases from 2 to 4 hours in RCTs. To check whether such a delayed effect is also present for other oral antimigraine drugs, we hand‐searched the literature for publications on RCTs reporting efficacy. Two triptans, 3 nonsteroidal anti‐inflammatory drugs (NSAIDs), a triptan combined with an NSAID and a calcitonin gene–related peptide receptor antagonist were evaluated for their therapeutic gain with determination of time to maximum effect (E_max). E_max was compared with known T_max from pharmacokinetic studies to estimate the delay to pain‐free. The delay in therapeutic gain varied from 1–2 hours for zolmitriptan 5 mg to 7 hours for naproxen 500 mg. An increase in effect from 2 to 4 hours was observed after eletriptan 40 mg, frovatriptan 2.5 mg and lasmiditan 200 mg, and after rizatriptan 10 mg (T_max = 1 h) from 1 to 2 hours. This strongly indicates a general delay of effect in oral antimigraine drugs. A review of 5 possible effects of triptans on the trigemino‐vascular system did not yield a simple explanation for the delay. In addition, E_max for triptans probably depends partly on the rise in plasma levels and not only on its maximum. The most likely explanation for the delay in effect is that a complex antimigraine system with more than 1 site of action is involved.

1. INTRODUCTION

Currently, the primary efficacy parameter in randomised controlled trials (RCTs) of orally administered drugs for treatment of migraine attacks is the percentage of pain‐free patients 2 hours after administration.1, 2, 3 Previously, the primary parameter was headache relief (a decrease from moderate or severe headache to none or mild) after 2 hours.4, 5 The 2‐hour criterion seems supportable for 2 reasons: First, the triptans are absorbed relatively quickly, and the time to maximum plasma concentration (T_max) ranges from 1 hour for https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=51 10 mg5 to 3 hours for https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=7191 2.5 mg.5 Second, rescue medication after 2 hours is permitted to avoid causing undue discomfort to study participants not responding to treatment.1, 2, 3

The 2‐hour responses for becoming pain‐free and achieving headache relief have subsequently been regarded as gold standards for efficacy and used in reviews, meta‐analyses, and Cochrane reviews of drugs for the treatment of migraine attacks.6, 7 https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=45 2.5 mg, however, was considered a slow‐acting and less effective triptan,8 so that headache relief also was evaluated after 4 hours, with some increase in efficacy observed from 2 to 4 hours.9 In a few RCTs comparing naratriptan, sumatriptan, and https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=40, the effect of all 3 oral triptans increased from 2 to 4 hours.10, 11, 12 This observation inspired the current review of a possible delay between T_max and time to maximum effect (E_max) of oral and parenteral antimigraine drug treatment.

2. METHODS

We searched PubMed with the terms “migraine attacks AND drug treatment AND controlled randomized trials”. Among the articles found by this search, only articles on RCTs of acute migraine treatment were searched by personal evaluation of abstract and methods section for results beyond 2 hours. In addition, articles with results beyond 2 hours were retrieved by hand search of reference lists of meta‐analyses4, 5, 13, 14 and Cochrane reviews6, 7, 15, 16, 17, 18, 19 on drugs for migraine attack treatment. For the main analysis, which was the evaluation of time–effect curves of oral drugs (Table 1), we included papers based on 2 conditions: if the T_max of the drug was available from pharmacokinetic studies, and if the E_max could be defined from the results of the RCTs.

Table 1.

Freedom from pain after oral administration in randomised controlled trial (RCTs) of sumatriptan 85 mg, zolmitriptan 2.5 mg, sumatriptan 85 mg combined in 1 tablet with naproxen 500 mg, naproxen 500 mg, ibuprofen solution 600 mg, diclofenac solution 50 mg, and the calcitonin gene‐related peptide antagonists rimegepant 75 mg and ubrogepant 25 mg. Rescue medication was permitted starting 2 hours after dosing*. Only statistically significant differences are presented as therapeutic gain (TG: percentage response after active drug minus percentage response after placebo)

Pain‐free, %
			(TG in %; 95% CIs in %)
Ref. drug and dose	Number of patients	Tmax	1 h	2 h	Emax	Other time points
Oral sumatriptan 85 mg20, 21	723	1.5 h5	5% (2%;0.5–4%)	24% (15%;11–19%)	4 h, 42% (27%; 22–31%)	8 h, 48% (26%; 21–31%)
Placebo	742		3%	9%	15%	22%
Oral zolmitriptan 5 mg22	345	1.5 h5	4% (3%; 1–6%)	26% (16%; 10–22%)	3–4 h, 45% (27%; 20–33%)	8 h, 55% (21%; 14–28%)
Placebo	348		1%	10%	18%	34%
Oral sumatriptan 85 mg plus naproxen 500 mg20, 21	712	Sumatriptan 1 h23 naproxen 6 ha 23	5% (3%; 1–5%)	32% (23%; 19–27%)	4–6 h, 54% (40%; 36–45%)	10 h, 64% (39%; 34–44%)
Placebo	726		2%	9%	15%	25%
Oral naproxen 500 mg20, 21	706	2 h23		15% (6%; 3–10%)	8 h, 37% (15%; 10–20%)	10 h, 39% (14%; 10–19%)
Placebo	726			9%	22%	25%
Ibuprofen suspension 600 mg24	198	0.5 h25	9% (7%; 2–13%)	29% (17%; 8–25%)	4 h, 51% (23%; 12–32%)	8 h, 61% (20%; 10–30%)
Placebo	142		2%	12%	28%	41%
Diclofenac solution 50 mg26	343	15 min27	12% (8%; 4–12%)	25% (15%; 9–21%)	4 h, 43% (22%; 15–28%)	8 h, 46% (20%; 13–27%)
Placebo	347		4%	10%	22%	26%
	Number of patients	Tmax	2 h	4 h	Emax	Other time point
Rimegepant 75 mg28	537	2 h29	20% (7%; 2–11%)	43% (19%; 13–25%)	6 h, 54% (22%; 16–28%)	8 h, 66% (19%; 13–25%)
Placebo	535		13%	24%	32%
Ubrogepant 25 mg30	435	1 h (0.7–1.5)31	22% (7%; 2–15%)	48% (15%; 9–22%)	6 h, 67% (17%,b 11–24%)	8 h, 70% (14%; 8–20%)
Placebo	456		15%	33%	45%	56%

*Rescue medication used in oral randomised controlled trials presented in this and other tables and in the text is more frequent in the placebo group. Its use would cause a bias towards an effect in the placebo arm compared to the active arm, thus reducing TG; ^a the T_max of 6 h for naproxen in the combination with sumatriptan is considerably prolonged compared with the usual T_max of naproxen which is 2 h;23 ^bfor ubrogepant 50 mg no E_max could be estimated because both after 6 and 8 h the TG was 17% (8 h after drug administration was the latest time point for evaluation of the acute effect of ubrogepant).30

For RCTs of nonoral formulations (subcutaneous or by inhalation), only the T_max needed to be presented. In addition, in RCTs involving oral triptans (other than https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=54 and https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=60) and https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=3928 that registered effects at 2 to 4 hours, we evaluated any increase in effect between the 2 time points.

When both pain‐free and headache relief were reported, for evaluation of the time–effect curve, we used pain‐free relief as the most clinically relevant criterion.1, 2, 32 In RCTs, for each time point in the main analysis when a drug was superior to placebo, we calculated the therapeutic gain (TG; percentage response to active drug minus percentage response to placebo) with 95% confidence intervals (CIs).

3. RESULTS

3.1. Retrieval of articles from the literature

The PubMed search, see Methods, resulted in 585 hits but of these articles only 150 concerned.

RCTs with acute migraine drugs. Further evaluation of the articles resulted in only 7 articles reporting results beyond 2 hours. The additional literature search of meta‐analyses and Cochrane reviews, see Methods, resulted in retrieval of 24 relevant articles.

3.2. Oral therapies

Table 1 summarises the findings regarding oral administration of sumatriptan,20, 21 zolmitriptan,22 sumatriptan plus https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=5230,20, 21 naproxen,20, 21 https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2713,24 https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=2714,26 rimegepant28 and ubrogepant30 (Table 1).

Figure 1 illustrates the time–effect curves for becoming pain‐free after sumatriptan 85 mg and placebo up to 12 hours. Both curves show an increase with time, the TG had an estimated maximum (E_max) of 27% after 4 hours. The E_max for TG is estimated with some uncertainty.

Figure 1.

Mean percentage who were pain‐free after oral sumatriptan 85 mg (n = 723) and placebo (n = 742) up to 12 hours in 2 randomised controlled trials. Also shown is the mean therapeutic gain (difference between sumatriptan and placebo)

As shown in Table 1, the delay between T_max and E_max varies considerably among the 8 drugs or the drug combination: 1–2 h (zolmitriptan), 2½ h (sumatriptan), 3½ h (ibuprofen), 4 h (diclofenac, rimegepant, the combination of naproxen and sumatriptan), 5 h (ubrogepant) and 7 h (naproxen).

The therapeutic gain for pain‐free from 2 hours (the usual time point for evaluating efficacy in oral RCTs3) to the later maximum effect (E_max) increased for all drug as follows: ibuprofen 600 mg 6% (17–23%); diclofenac 50 mg 7% (15–225); naproxen 500 mg 9% (6–15%); ubrogepant 25 mg 10% (7–17%); zolmitriptan 5 mg 11% (16–27%); sumatriptan 85 mg 12% (15–27%); rimegepant 75 mg 15% (7–22%); and sumatriptan/naproxen combination 85 mg/500 mg 17% (23–40%).

3.3. Subcutaneous sumatriptan

For headache relief (a decrease in headache from moderate or severe to none or mild), sumatriptan had an effect after 10 minutes, TG = 12% (95% CI: 8–15%),33 whereas an effect for being pain‐free was detected first after 20 minutes, see Table 2.33

Table 2.

The effect of subcutaneous sumatriptan on pain‐free or headache relief response in 1 placebo‐controlled randomised controlled trial33

Subcutaneous sumatriptan 6 mg
Ref. drug and dose	Tmax	Number of patients	10 min	20 min	30 min	40 min	50 min	60 min
Pain‐free, % (TG in % + 95% CIs in %)
Sumatriptan 6 mg	10 min	734	2%	9% (8%; 5–10%)	17% (15%; 12–18%)	26% (22%; 18–26%)	36% (30%; 26–34%)	49% (39%; 34–44%)
Placebo		370	0%	1%	2%	4%	6%	10%
Median plasma concentrations after subcutaneous sumatriptan 3 mg (ng/mL; n = 18)34			42	35	22			9

The RCT33 showed a slow increase in TG for pain‐free response to 39% (95% CIs: 34–44%) at 60 minutes, the last time of evaluation. In a Cochrane Review of subcutaneous sumatriptan 6 mg the mean TGs for pain‐free were 34% (95% CIs: 32–37%) after 1 hour and 44% (95% CIs: 41–48%) after 2 hours.6 In contrast, there was no significant increase of TG for headache relief from 1 to 2 hours.6, 35

3.4. Inhaled dihydroergotamine

The T_max for oral inhalation of dihydroergotamine (https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=121) occurred after 10 minutes, followed by a rapid decrease in plasma concentrations36 (Table 3). However, the effect developed slowly up to 4 hours,36, 37 the last time point of measurements (Table 3).

Table 3.

Pain‐free after orally inhaled dihydroergotamine (DHE) and placebo36, 37

Pain‐free (TG in % + 95% CIs in %)
Ref. drug and dose	Number of patients	Tmax	10 min	30 min	1 h	2 h	4 h
36, 37 DHE 1 mg	369	10 min	1%	5% (4%; 1–7%)	15% (10%; 5–15%)	28% (18%; 12–23%)	39% (22%; 16–28%)
Placebo	370		1%	1%	5%	10%	17%
Mean DHE concentrations (pg/mL)36			3.400	1.800	1.100	490

3.5. Intravenous sumatriptan, olcegepant and acetylsalicylic acid lysinate

Table 4 shows the findings from a small (n = 30), placebo‐controlled, double‐blind RCT of intravenous sumatriptan 64 μg/kg.38 The results indicated an immediate effect on headache relief (14/15) within 20 minutes.

Table 4.

Headache relief and pain‐free in a small randomised controlled trial comparing intravenous sumatriptan (n = 15) with placebo (n = 15)38

Drug and dose	Headache relief within 20 min	Pain‐free within 20 min
Sumatriptan 64 μg/kg as a slow intravenous bolus over 3–4 min	93% (14/15)a TG, 80% (95% CI: 46–91%)	8/15b
Placebo	13% (2/15)	0/15

^a P < .001; ^b P < .01.

Some pain‐free results from a trial with the calcitonin gene‐related peptide (CGRP) receptor antagonist https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=702 (0.25 mg to 10 mg infused for 10 min) are shown in Table 5.39 Intravenous olcegepant 2.5 mg, the optimum dose39 was superior to placebo after 2 hours with a TG of 41% (95% CI: 23–58%). After 4 hours, the TG was 46% (95% CI: 25–63%).

Table 5.

Pain‐free (PF) after intravenous infusion of olcegepant or placebo.39 In addition to the optimum 2.5 mg dose (based on headache response), the combined results for the 3 highest doses (2.5 mg, 5 mg, and 10 mg) are presented. time points are calculated after start of the 10‐minute infusion

PF (TG in % + 95% CIs in %)
Olcegepant dose	Number of patients	30 min	1 h	2 h	4 h
2.5 mg	32	3% (1/32)	16% (5/32)	44% (14/32; 41%; 23–58%)	56% (18/32; 46%; 25–63%)
Combined PF after 2.5 mg + 5 mg + 10 mg.	60	5% (3/60)	15% (9/60)* (13%; 0.03–24%)	35% (21/60; 33%; 18–45%)	48% (29/60; 39%, 21–52%)
Placebo	41	2% (1/41)	2% (1/41)	2% (1/41)	10% (4/41)

^* P = .045, Fisher's exact test.

Headache relief after intravenous acetylsalicylic acid lysinate 1800 mg in 1 RCT40 showed a maximum TG after 90 minutes (Table 6).

Table 6.

Headache relief after intravenous acetylsalicylic acid lysinate (AAL), and placebo in 1 randomised controlled trial40

Headache relief (TG in % + 95% CIs in %).
Drug and dose	Number of patients	30 min	60 min	90 min	120 min
AAL 1800 mga	119	33% (26%; 11–36%)	61% (42%; 25–55%)	73% (52%; 35–64%)	74% (48%; 30–61%)
Placebo	42	7%	20%	22%	25%

^acorresponds to aspirin 1000 mg

3.6. Three additional oral triptans at 2 and 4 hours postadministration

Results of headache relief up to 4 hours were published for 13 RCTs of naratriptan, eletriptan and frovatriptan. The mean TGs for headache relief increased from 2 to 4 hours as follows: from 35% (95% CI: 29–40%) to 47% (95% CI: 41–52%) in 3 RCTs12, 41, 42 for eletriptan 40 mg; from 20% (95% CI: 16–20%) to 28% (95% CI: 23–31%) in 5 RCTs10, 11, 12, 43, 44 for naratriptan 2.5 mg; and from 13% (95% CI: 10–16%) to 29% (95% CI: 27–32%) in 6 RCTs4, 45, 46, 47, 48 for frovatriptan 2.5 mg. We identified no results for efficacy of oral almotriptan 12.5 mg after 2 hours.

3.7. Effect of the oral 5‐HT_1F receptor agonist lasmiditan

https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=3928 (200 mg), an oral 5‐HT_1F receptor agonist, has a T_max of 1.5 hours.49 In a large phase 3 RCT, lasmiditan 200 mg (n = 518) was superior to placebo (n = 524), with 32% pain‐free after 2 hours on treatment vs 15% pain‐free on placebo.50 The TG for being pain‐free after 2 hours was thus 17% (95% CI: 12–22%); after 3 hours, it was 19% (18–37%); and after 4 hours, it was 20% (22–42%; results for 3 and 4 hours were provided by Eli Lilly & Co.).

3.8. Additional early results with rizatriptan, a quickly absorbed oral triptan, and results with intranasal butorphanol

In 5 RCTs,51, 52, 53, 54, 55 the mean TG for becoming pain‐free with oral rizatriptan 10 mg was 9% (95% CI: 7–11%) after 1 hour at T_max,5 but that increased to 34% (95% CI: 31–37%) after 2 hours. At the same time points, plasma concentrations decreased from 20 ng/mL (1 hour) to 13 ng/mL (2 hours).56

http://bcp14090.docx (T_max = 0.33 h)57 1 mg (plus an optional 1 mg after 1 h; n = 107) was compared with placebo (n = 50) in 1 RCT.58 After 1 hour, the TG for pain‐free was 12% (95% CI: −0.01% to 19%; not significant). The TG was 19% (95% CI: 18–38%) after 2 hours, 23% (95% CI: 9–34%) after 4 hours, and 19% (95% CI: 4–31%) after 6 hours.58 At 2 hours, 8% in the butorphanol group and 24% in the placebo group had used rescue medication.58

4. DISCUSSION

4.1. Time–effect curves after oral and nonoral administration of antimigraine drugs

A delayed effect of a drug vs plasma levels can involve several factors.59 These factors include slow equilibration of the drug at the receptor site, delay in the distribution of the drug to the effect site, and initiation of physiological responses by the drug at the receptor mediated by secondary messengers or the turnover of biological intermediates.59 A combination of these 3 mechanisms probably is common; generally, if more than 1 mechanism acts in the clinical effect, it will be the rate‐limiting process with the slowest time–effect curve, in turn determining the time course of effect.59 A likely example is the slow equilibration of dihydroergotamine at the receptor site,60 resulting in a considerable delay of effect,36, 61, 62 see e.g. Table 3.

Based on the results for the time to become pain‐free in 8 RCTs (see Table 1), we constructed time–effect curves for some quite different, acute oral migraine drugs (2 triptans, 3 NSAIDs, the combination of triptan and NSAID, and 2 oral CGRP receptor antagonists (Table 1); and evaluated the delay from T_max to E_max.

The delays from T_max to E_max varied from 1–2 h to 7 h, demonstrating the lack of correlation between drug concentrations and drug effects in migraine treatment. Further support comes from the increase in TG from 2 to 4 hours after treatment with naratriptan,10, 11, 12, 43, 44 eletriptan,12, 41, 42 frovatriptan,10, 46, 47, 48 and the 5‐HT_1F receptor agonist lasmiditan.50 The similar delay in effect from T_max for 4 quite different groups of oral acute migraine drugs (triptans, NSAIDs, a 5HT_1F receptor agonist and CGRP receptor antagonists) indicates that this delay is a general phenomenon after oral administration of acute migraine drugs.

In contrast, time–effect curves after nonoral administration are not so uniform. For subcutaneous sumatriptan with a T_max of 10 minutes,34 the pain‐free response increased slowly from 20 to 60 minutes in 1 RCT,33 and increased further from 1 to 2 hours based on a Cochrane Review,6 Thus, the E_max for pain‐free response after subcutaneous sumatriptan cannot been defined. In contrast, for headache relief after subcutaneous sumatriptan the effect is the same after 1 and 2 hours.6 Traditionally, the effect of subcutaneous sumatriptan has been evaluated as headache relief after 1 hour.5, 35 Since the response (pain‐free) after 2 hours is the usual measure with most data available, it is also the most interesting time point for subcutaneous drugs when they are compared with other application forms. Notwithstanding, earlier time points are very useful in addition.

The effect of orally inhaled DHE (T_max 10 min) develops slowly over 4 hours (Table 3).36, 37 This pace probably traces to a tight binding of DHE to the 5‐HT_1B/1D receptor, resulting in slow onset and long duration of action.60, 61, 62

For intranasal butorphanol (T_max 0.33 h57), the E_max for becoming pain‐free was 3.5 hours.58 In contrast, for post‐operative pain, intranasal butorphanol works much faster, with an E_max up to 60 minutes.63

We identified a delay to becoming pain‐free with 2 intravenous administrations: acetylsalicylate lysine39 had an E_max of 90 to 120 minutes, and with olcegepant, the effect time increased from 2 to 4 hours.40 In contrast, intravenous sumatriptan resulted in a TG for headache relief of 93% within 20 minutes.38 The lack of delay for intravenous sumatriptan, with maximum plasma levels in the same range (70–80 ng/mL) as subcutaneous sumatriptan,64, 65 could theoretically be the result of a quicker distribution to target sites by intravenous administration. This in turn could lead to, for example, constriction of extracerebral arteries35 combined with a quick blockade of neurotransmission in the trigeminal nucleus66 (see below).

Thus, intravenous administration of 3 different drugs—a prostaglandin synthesis inhibitor, a CGRP receptor antagonist, and a 5‐HT_1B/1D receptor agonist— was most likely to result in different time–effect curves.

4.2. Probable dependence of effect size on the rise in plasma concentrations, indicating a complex modulation of trigemino‐vascular pain

The size of E_max with similar plasma levels of a triptan drug can vary depending on the administration route.67 For example, subcutaneous naratriptan 1 mg with a maximum concentration in plasma (C_max) of 16 ng/mL resulted in a TG of 27% for becoming pain‐free after 2 hours.67 In contrast, oral naratriptan 2.5 mg with a C_max of 13 ng/mL resulted in a TG for being pain‐free of 16% after 2 hours.13, 68 Similar results were reported for the subcutaneous sumatriptan vs oral sumatriptan.67 When dosed to similar C_max values, the 2 administration modes result in different speeds of absorption, with an increasing rate of rise in plasma levels for subcutaneous vs oral administration.

“The time course of the pharmacological response is often influenced by in vivo homeostatic feedback mechanisms that may be operative”.69 Such mechanisms may explain observations such as e.g. complex pharmacological effect vs time profiles, and that an effect can depend on the rate of administration of the drug.69 The latter is probably observed with the larger effect of subcutaneous compared to oral administration of sumatriptan. These differences may also indicate that these drugs work on a complex migraine pain system. Such a system with multiple sites of actions and interactions could result in a delay of effect, as observed in RCTs of acute migraine treatment.

4.3. Summary of several likely sites of actions of triptans that could contribute to the interplay in a complex system for treatment of migraine pain

Triptans are the only drugs that have been extensively evaluated in both oral and parenteral RCTs35 and have undergone extensive pharmacological testing in isolated tissues, in vivo in animals and in humans. A review of the pharmacology of these drugs is thus likely to be instructive for understanding antimigraine mechanisms.

These possible sites of action are based largely on pharmacological studies in animals70, 71, 72 (and a few in humans73, 74). Only a few studies investigated the time to onset of effect (e.g., neuronal) with an intact blood–brain barrier.66, 75 Apart from unknown sites, there are 5 possible mechanisms for the effect of triptans in the trigemino‐vascular pain system:

1. Sumatriptan constricts arteries quickly: Maximum constriction of isolated human coronary arteries is elicited in vitro within 4 minutes.76 In 1 study using angiography of the middle meningeal artery (MMA), local intravenous (2 mg, n = 7) and subcutaneous (6 mg, n = 2) sumatriptan caused constriction of the artery in 7 of 9 patients, often within 3–5 minutes.77 With magnetic resonance (MR) angiography, however, dilatation of extracerebral arteries was not observed during spontaneous migraine attacks.78 This finding called into question the idea that arterial dilatation was involved. So far, only the extracranial part of the MMA has been investigated with MR angiography.78, 79 Improved MR angiography techniques have helped reveal that the intracranial part of the MMA is dilated after sildenafil and CGRP treatment in healthy volunteers77, 80 and investigations during migraine attacks with this imaging method of MMA are awaited.

2. Neuropeptide release from perivascular meningeal afferents and subsequent neurogenic inflammation have been hypothesised to be an important mechanism of migraine pain.81, 82 Experimentally, triptans inhibit neuropeptide release from trigeminal afferents, a mechanism that has been speculated to contribute to the antimigraine effect of triptans83, 84 Neuropeptide release causes arterial dilatation (through CGRP), protein plasma extravasation (through substance P), and degranulation of mast cells,84 3 elements of neurogenic inflammation. However, neurogenic inflammation is not the cause but the result of afferent activation. Furthermore, neuropeptides do not effectively activate primary afferents of the dura mater.85 Clinically, treatment with subcutaneous sumatriptan during migraine normalises increased levels of CGRP in the external jugular vein,86 suggesting inhibition of neuropeptide release. The source of the released CGRP has not been clarified.

3. The trigeminal ganglion as a target for antimigraine drugs is an attractive option because the ganglion lacks a blood–brain barrier.87 Different 5‐HT₁ receptor subtypes, including 5‐HT_1F, have been localised on trigeminal ganglion neurons, so that the trigeminal ganglion appears to be a target for triptans.88, 89 Triptans binding to Gi‐protein–coupled 5‐HT₁ receptor subtypes may act on trigeminal ganglion neurons via different mechanisms.90, 91 Sumatriptan inhibits CGRP release from the trigeminal ganglion, which can be reversed by a 5‐HT_1B/D antagonist.92 Sumatriptan inhibits voltage‐gated calcium currents in rat trigeminal ganglion neurons and prevents sensitisation of neurons innervating the dura mater against inflammatory mediators, which can be blocked by a 5‐HT_1D antagonist.93 In addition, activation of the 5‐HT_1D receptor subtype decreases the membrane excitability of mouse trigeminal ganglion neurons.91

4. Triptans probably inhibit nociceptive transmission within the trigeminal nuclear brainstem complex (TBNC). The central terminals of trigeminal afferents projecting there may be equipped with 5‐HT₁ receptors as targets for triptans.94 Because these sites are located inside the blood–brain barrier, triptans may have limited access, and an action may depend mainly on the degree (or possible on the speed) of their penetrance.95 In cats, superior sagittal sinus stimulation showed zolmitriptan but not sumatriptan inhibiting neuronal activity in the TBNC.96, 97, 98 In contrast, sumatriptan blunts the c‐fos increase induced by blood in the cisterna magna in rats,99 and intravenous sumatriptan blunts the blood flow in the TBNC induced by stimulation of the trigeminal ganglion, beginning at 5 minutes after administration, with a maximum after 60 minutes.66 If zolmitriptan and sumatriptan cross the blood–brain barrier with a relevant difference in speed or amount and this effect is important, an effect difference in migraine might be expected, but there is none. The TGs for being pain‐free after 2 hours are sumatriptan 100 mg 21% (95% CI: 20–23%)6 and zolmitriptan 2.5 mg 20% (95% CI: 18–22%).7

5. The thalamus has been described as being most likely to be involved in dysfunctional pain modulation and processing in migraine.100 In addition, during migraine attacks, the network connectivity of the right thalamus with pain‐modulating and pain‐encoding cortical areas is affected.101 In 1 study in rats, intravenous or microiontophoretically naratriptan potently and reversibly modulated trigemino‐vascular thalamic neuronal firing triggered by electrical stimulation of the superior sagittal sinus.102 The peak inhibition was observed between 5 and 10 minutes after intravenous administration;102 however, the blood–brain barrier was probably not intact because of insertion of the recording electrode.

4.4. The low efficacy of acute migraine drugs and possible contributions of the concept of a delay in migraine models for developing better drugs

Most migraine patients consider complete relief of head pain, rapid onset of action and no recurrence as important attributes of acute migraine therapy.32 In addition, patients consider relief without adverse events to be important.32

Those who use triptans often find the drugs satisfactory.103, 104 However, only a minority of migraine patients (35% in France104 and 19% in the USA103, 105) use triptans.

Headache experts often judge current medical treatment of migraine attacks as unsatisfactory. According to a 2015 review in Lancet Neurology, “the management of patients with migraine is often unsatisfactory because available acute and preventive therapies are either ineffective or poorly tolerated”.106

Despite the emphasis that migraine patients place on rapid complete relief of head pain,32 for most current, oral drugs, a TG toward freedom from pain after 2 hours is far from these wishes, see Table 7. The lack of optimum acute migraine treatment has led to 2 types of approaches for improvements: development of new drug formulations for current drugs, and novel drug targets based on a better understanding of headache pathophysiology.107 As shown in Table 7 (section 2, New administration forms, and section 3, Recently developed drugs), neither approach has yielded increased efficacy. For the gepants (ubrogepant and rimegepant), tolerability is improved compared with most triptans,28, 31 and these drugs and lasmiditan can be administered to patients with cardiovascular problems. In addition, nonresponders to triptans may possibly respond to the new drugs.

Table 7.

Therapeutic gain (TG) with 95% confidence intervals (CIs) for being pain‐free (PF) 2 hours after drug administration for migraine attacks. Examples of commonly used standard drugs (for other antimigraine drugs, see13, 14) are shown, together with new administration routes for current drugs and recently developed drugs with different pharmacological attributes

Section 1. Standard drugs for the treatment of migraine attacks
Oral aspirin 900/1000 mg TG, 13% (95% CI: 9%–16%)108	Oral sumatriptan 50 mg TG, 16% (95% CI: 14%–18%)6	Oral sumatriptan 100 mg TG, 21% (95% CI: 20%–23%)6	SC sumatriptan 6 mg TG, 44% (95% CI: 41%–48%)6	Oral zolmitriptan 2.5 mg TG, 20% (95% CI: 18%–21%)7
Oral eletriptan 40 mg TG, 26% (95% CI: 22%–9%)109	Oral rizatriptan 10 mg TG, 34% (95% CI: 31%–37%)47, 48, 49, 50, 51	Combination of sumatriptan 85 mg and naproxen 500 mg TG, 23% (95% CI: 19%–26%)20, 21	Oral ibuprofen 400 mg TG, 14% (95% CI: 11%–17%)18	Oral naproxen 500 mg TG, 9% (95% CI: 6%–12%)16

Section 2. New administration routes for current drugs
Intranasal zolmitriptan 5 mg TG, 23% (95% CI: 20%–26%)110, 111	Soluble diclofenac 50 mg TG, 14% (95% CI: 9%–18%)17	Nasal‐delivered sumatriptan powder 22 mg TG, 17% (95% CI: 5%–28%)112	Orally inhaled dihydroergotamine 1 mg TG, 18% (95% CI: 12%–23%)31, 33	Transdermal patch with sumatriptan TG, 9% (95% CI: 3%–15%)113

Section 3. Recently developed drugs for the treatment of migraine attacks
Oral lasmiditana 200 mg TG, 17% (95% CI: 12%–20%)46	Oral ubrogepantb 100 mg TG, 9% (95% CI: 5%–14%)109	Oral rimegepantb ODTc 75 mg TG, 10% (95% CI: 6%–14%)114	Oral telcagepantb,c 280/300 mg TG, 17% (95% CI: 14%–20%)109

^alasmiditan is a 5‐HT_1F receptor agonist; ^bubrogepant, rimegepant and telcagepant are calcitonin gene‐related peptide receptor antagonists; ^can oral dissolving tablet; ^dtherapeutic gain (TG) for telcagepant is shown in order to compare it to the TGs for ubrogepant and rimegepant. Development of telcagepant was stopped due to hepatic toxicity.

SC, subcutaneous.

Optimum successful treatment of migraine attacks has been obtained with some high doses and parenteral administration forms of triptans. In the small placebo‐controlled RCT of intravenous sumatriptan 64 μg/kg, the TG for headache relief within 20 minutes was 80% (95% CI: 46–91%; see above).38 In another RCT, subcutaneous naratriptan 10 mg (n = 34) resulted in 88% of patients achieving freedom from pain after 2 hours. The TG for naratriptan was 71% (95% CI: 52–81%), and naratriptan was superior to sumatriptan 6 mg (n = 47; 55% pain‐free for a difference of 33%; 95% CI: 13–49%).68 Both naratriptan (71%), and sumatriptan (53%) caused more adverse events than placebo (22%; P < .05). Maker Glaxo did not develop parenteral naratriptan any further because the main emphasis at that time was on tolerability of triptans.

Animal and human models (often drug‐induced migraine pain) have been developed (see70, 71, 72, 73, 74, 115, 116, 117, for review) and should be feasible for investigating 2 main factors for satisfactory treatment of migraine attacks: E_max (maximum proportion pain‐free) and rapid onset of action (speed of absorption of drug and delay between pharmacokinetics and pharmacodynamics of the drug). As noted above, E_max probably also depends on speed of absorption of the drug.

4.5. Clinical implication of the delay between pharmacokinetics and pharmacodynamics

In this review we have estimated the delay in effect from T_max to E_max, but clinically it is also important to look at the early part of the time–effect curve of an important delay beyond the plasma concentrations. Early results with the quickest orally absorbed triptan illustrate this fact: rizatriptan has a T_max of 1 hour,5 but in 2 RCTs with rizatriptan 10 mg (n = 705) vs placebo (n = 241),53, 54 the TG for freedom from pain was 8% (95% CI: 5–11%) after 1 hour and 35% (95% CI: 29–40%) after 2 hours. Meanwhile, plasma concentrations of rizatriptan were 25 ng/mL after 1 hour and 14 ng/mL after 2 hours. 57 Some delay is also obvious for subcutaneous sumatriptan 6 mg (T_max = 10 min; E_max = 1 h;33 see Table 1).

4.6. An additional note on the possible delay in acute treatment of tension‐type headache

In a meta‐analysis of 4 double‐blind randomised, controlled trials in episodic tension‐type headache pain‐free after a combination of aspirin, paracetamol and caffeine (Excedrin^R) was superior to paracetamol, and placebo was evaluated up to 4 hours.118 The mean T_max for oral paracetamol is approximately 40 minutes;119, 120 but the pain‐free (PF) response on headache increases from 2 hours (paracetamol 21% PF, n = 2748 attacks, vs placebo 18% PF, n = 1376 attacks; TG = 3%; 95% CI: 0.4% to 5%) to 4 hours (paracetamol 57% PF vs placebo 48% PF, TG = 9%; 95% CI: 5% to 12%), thus demonstrating a delay between pharmacokinetics and effect of paracetamol on episodic tension‐type headache. We have not found results for efficacy beyond 2 hours for aspirin and ibuprofen, often used in the treatment of tension‐type headache, but in future studies with tension‐type headache these drugs should be evaluated also after 4 hours, in order elucidate whether a pharmacodynamic delay is present generally in tension‐type headache treatment similar to the results in migraine. Also, such studies would be interesting in other pain disorders.

4.7. Nomenclature of targets and ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to HARMACOLOGY.121

COMPETING INTERESTS

There are no competing interests to declare.

Tfelt‐Hansen P, Messlinger K. Why is the therapeutic effect of acute antimigraine drugs delayed? A review of controlled trials and hypotheses about the delay of effect. Br J Clin Pharmacol. 2019;85:2487–2498. 10.1111/bcp.14090