Rescue Therapy for Acute Migraine, Part 3: Opioids, NSAIDs, Steroids, and Post-Discharge Medications

Authors


D.E. Tepper, MD, Center for Headache and Pain, Neurological Institute, Cleveland Clinic C-21, 9500 Euclid Avenue, Cleveland, OH 44195, USA, email: debtepper@gmail.com

Abstract

Objective.— The final section of this 3-part review analyzes published reports involving the acute treatment of migraine with opioids, non-steroidal anti-inflammatory drugs (NSAIDs), and steroids in the emergency department (ED), urgent care, and headache clinic settings, as well as post-discharge medications. In the Conclusion, there is a general discussion of all the therapies presented in the 3 sections.

Method.— Using the terms (“migraine” AND “emergency”) AND (“therapy” OR “treatment”), the author searched MEDLINE for reports from ED and urgent care settings that involved all routes of medication delivery. Reports from headache clinic settings were included only if medications were delivered by a parenteral route.

Results.— Seventy-five reports were identified that compared the efficacy and safety of multiple acute migraine medications for rescue. Of the medications reviewed in Part 3, opioids, NSAIDs, and steroids all demonstrated some effectiveness. When used alone, nalbuphine and metamizole were superior to placebo. NSAIDs were inferior to the combination of metoclopramide and diphenhydramine. Meperidine was arguably equivalent when compared with ketorolac and dihydroergotamine (DHE) but was inferior to chlorpromazine and equivalent to the other dopamine antagonists. Steroids afford some protection against headache recurrence after the patient leaves the treatment center.

Conclusions.— All 3 opioids most frequently studied – meperidine, tramadol, and nalbuphine – were superior to placebo in relieving migraine pain, although meperidine combined with promethazine was not. Opioid side effects included dizziness, sedation, and nausea. With ketorolac being the most frequently studied drug in the class, NSAIDs were generally well tolerated, and they may provide benefit even when given late in the migraine attack. The rate of headache recurrence within 24-72 hours after discharge from the ED can be greater than 50%. Corticosteroids can be useful in reducing headache recurrence after discharge. As discussed in Parts 1, 2, and 3, there are effective medications for provider-administered “rescue” in all the classes discussed. Prochlorperazine and metoclopramide are the most frequently studied of the anti-migraine medications in the emergent setting, and their effectiveness is superior to placebo. Prochlorperazine is superior or equivalent to all other classes of medications in migraine pain relief. Although there are fewer studies involving sumatriptan and DHE, relatively “migraine-specific” medications, they appear to be equivalent to the dopamine antagonists for migraine pain relief. Lack of comparisons with placebo and the frequent use of combinations of medications in treatment arms complicate the comparison of single agents to one another. When used alone, prochlorperazine, promethazine, metoclopramide, nalbuphine, and metamizole were superior to placebo. Droperidol and prochlorperazine were superior or equal in efficacy to all other treatments, although they also are more likely to produce side effects that are difficult for a patient to tolerate (especially akathisia). Metoclopramide was equivalent to prochlorperazine, and, when combined with diphenhydramine, was superior in efficacy to triptans and NSAIDs. Meperidine was arguably equivalent when compared with ketorolac and DHE but was inferior to chlorpromazine and equivalent to the other neuroleptics. Sumatriptan was inferior or equivalent to the neuroleptics and equivalent to DHE when only paired comparisons were considered. The overall percentage of patients with pain relief after taking sumatriptan was equivalent to that observed with droperidol or prochlorperazine.

(Headache 2012;52:467-482)

In Part 1 of this review on physician-administered rescue therapy for acute migraine in the emergency department (ED), urgent care, and headache clinic infusion center settings, results of trials involving triptans, dihydroergotamine (DHE), and magnesium sulfate were presented. Information concerning migraine pathophysiology and the commonly used methodology for studies of migraine therapy in the ED also was included. Part 2 focused on studies involving dopamine antagonists, antihistamines, 5HT3 antagonists, valproate, and others (octreotide, lidocaine, nitrous oxide, propofol, and bupivacaine). The present paper, Part 3, includes studies involving opioids, non-steroidal anti-inflammatory drugs (NSAIDs), steroids, and post-discharge medications, as well as a general discussion of all therapies presented in the 3 sections.

OPIOIDS

Opioids can modulate nociceptive input to the trigeminocervical complex and frequently are used to treat pain in the ED. Opioids do not, however, affect the inflammatory processes or the neurovascular changes that occur in migraine. Meperidine is the opioid most studied for ED headache treatment.1 Meperidine has poor oral bioavailability, and so, is usually given parenterally. Side effects include sedation, respiratory and cardiac depression, and increased risk of dependency. Nalbuphine is a semisynthetic opioid with mixed agonist–antagonist analgesic properties. Its most frequent side effect is sedation. Nalbuphine can produce respiratory depression but is considered to have low abuse potential.2

Tramadol hydrogen chloride is an “atypical” opioid used for pain management.3 Tramadol weakly binds the mu opioid receptor and also inhibits serotonin and norepinephrine re-uptake. Tramadol is better tolerated than other opioids because of its low impact on the respiratory, cardiac, and gastrointestinal (GI) systems at therapeutic doses.

There is evidence that all opioids can sensitize the central nervous system to pain (especially in migraineurs) and increase the risk of medication-overuse headache.4

Klapper and Stanton compared meperidine 75 mg plus hydroxyzine 75 mg intramuscular (IM) with DHE 1 mg plus metoclopramide 10 mg to treat patients whose abortives had failed.5 Pain reduction measured on a 4-point pain scale (PPS) (4-PPS) was greater with DHE/metoclopramide (−2.14 vs −0.86, P = .006), as was the percentage with headache relief (93% vs 21%, P < .001). Carleton et al compared meperidine 1.5 mg/kg plus hydroxyzine 0.7 mg/kg IM with DHE 1 mg plus hydroxyzine 0.7 mg/kg IM.6 There was similar efficacy in pain reduction measured on a visual analog scale (VAS) (55.7% vs 53.5%, P = .81). Dizziness (DHE 2% vs meperidine 15%, P < .05) and drowsiness (DHE 21.5% vs meperidine 22.6%) were the most common side effects. Davis et al compared meperidine 75 mg plus promethazine 25 mg IM with ketorolac 60 mg IM, and rates of headache relief at 1 hour were not significantly different (63.3% vs 50%, respectively; P = .37).7 Duarte et al found that headache relief was similar between meperidine 100 mg plus hydroxyzine 50 mg IM and ketorolac 60 mg IM.8 Harden et al also found pain reduction (VAS) to be similar between meperidine 50 mg plus promethazine 25 mg IM, ketorolac 60 mg IM, and placebo/ normal saline (NS) IM (60% vs 44.4% vs 54.5%).9 Lane et al compared meperidine 0.4 mg/kg plus dimenhydrinate 25 mg intravenous (IV) with chlorpromazine 0.1 mg/kg IV (up to 3 doses); pain reduction (VAS) was significantly greater with prochlorperazine (−70.6 vs −44.5, P < .05).10

Larkin and Prescott compared meperidine 75 mg IM with ketorolac 30 mg IM, and headache relief at 1 hour was greater with meperidine (44% vs 13%, P < .02).11

Richman et al compared meperidine 1.5 mg/kg IM with droperidol 2.5 mg; there was no difference in pain reduction (VAS) (−37 vs −47, P = .33).12

Belgrade et al compared meperidine 75 mg IM plus hydroxyzine 50 mg IM with DHE 1 mg plus metoclopramide 10 mg IV and to butorphanol 2 mg IM.13 Pain reduction (VAS) was greater with DHE plus metoclopramide (−59) and butorphanol (−54) vs meperidine/hydroxyzine (−37, P < .01). Patients experienced greater than 90% pain reduction more often in the DHE/metoclopramide group (38%) vs butorphanol (16%) or meperidine/hydroxyzine (0%, P < .01). The most common side effects with each were: meperidine/hydroxyzine = sedation (18%); DHE/metoclopramide = nausea (33%), dysphoria (43%), restlessness (19%), and flushing (29%); and butorphanol = dizziness (21%) and nausea (26%). Scherl and Wilson found that 1 hour headache relief was similar when comparing meperidine 75 mg plus promethazine 25 mg IM with DHE 0.5 mg plus metoclopramide 10 mg IV (77.2% vs 86.2%, P = .37).14 Stiell et al compared meperidine 75 mg plus dimenhydrinate 50 mg IM with methotrimeprazine (not available in the USA) 37.5 mg IM and found similar pain reduction (VAS) (−46.6 vs −40.3, P = .27).15

Alemdar et al compared tramadol 100 mg IV with placebo/NS IV.3 The percent pain free at 1 hour was not greater with tramadol (29.4% vs 11.8%, P = .40). Pain reduction (VAS) at 1 hour was greater in the tramadol group (70.6% vs 35.3%, P = .04). No side effects were observed at 1 hour. Engindeniz et al compared tramadol 100 mg IM with diclofenac sodium 75 mg IM, and headache relief was the same in both groups (80%).16

Tek and Mellon compared nalbuphine 10 mg IM with hydroxyzine 50 mg IM, with hydroxyzine plus nalbuphine IM, and with placebo/NS IM.2 For migraineurs without aura only, headache relief at 1 hour was greatest with nalbuphine alone compared with the other treatments (nalbuphine −2.16 vs nalbuphine/hydroxyzine −1.42 vs hydroxyzine −1.00 vs placebo −0.89, P < .01).

Table 1 summarizes results from the studies involving meperidine, tramadol, and nalbuphine.

Table 1.—. Studies Comparing Meperidine (MEP), Tramadol (TMD), and Nalbuphine (NLB) Individually to Other Agents for Migraine Therapy. Superior Performance is Indicated by Boldfaced Characters; if Treatment 1 Superior, Then Also Underlined; if Treatment 2 Superior, Then Also Italicized
Study First Author (year)Group 1 (G1) Treatmentmg or mg/kgRouteGroup 2 (G2) Treatmentmg or mg/kgRouteVenue – # If Multiple SitesResearch Design% WomenMean AgeGroup 1 # PtsGroup 2 # Pts
Cicek (2004)17MEP50IMPCBNAIMEDR/DB/P88394948
Larkin (1992)11MEP75IMKET30IMEDR/DB77331615
Richman (2002)12MEP1.5IMDPD2.5IMEDR/DB73321415
Harden (1996)9MEP50IMPCBNAIMEDR/DB/P80331010
PlusPMZ25IMNA 
Scherl (1995)14MEP75IMDHE0.5IVEDR70311314
PlusPMZ25IMMTC10IV 
Davis (1995)7MEP75IMKET60IMEDR/DB81352220
PlusPMZ25IMNA 
Stiell (1991)15MEP75IMMTM37.5IMEDR/DB76323737
PlusDMH50IMNA 
Lane (1989)10MEP0.4IMCPZ0.1IMEDR/DB85312224
PlusDMH50IMNA 
Duarte (1992)8MEP100IMKET60IMEDR/DB80352525
PlusHDX50IMNA 
Carleton (1998)6MEP1.5IVDHE0.5IVED11R/DB82327878
PlusHDX0.7IVHDX0.7IV 
Belgrade (1989)13MEP75IMDHE1IVEDR63302122
PlusHDX50IMMTC10IV 
Klapper (1993)5MEP75IMDHE1IVCLR/DB1414
PlusHDX75IMMTC10IV 
Engindeniz (2005)16TMD100IMDCF75IMEDR/SB2020
Alemdar (2007)3TMD100IVPCBNAIVEDR/SB/P82401717
Tek (1987)2NLB1.5IMPCBNAIMEDR/DB/P75301818
Study First Author (year)% Pain Free% Pain Relief% Sustained Pain Relief% Disability Free% Requiring Rescue% Patient Satisfaction
G1G2G1G2G1G2G1G2G1G2G1G2
  1. — = no data obtained. Research design – DB = double blind; P = placebo controlled; R = randomized; SB = single blind.

  2. Route – IM intramuscular; IV = intravenous. Treatments – CPZ = chlorpromazine; DCF = diclofenac; DHE = dihydroergotamine; DMH = dimenhydrinate; DPD = droperidol; HDX = hydroxyzine; KET = ketorolac; MTC = metoclopramide; MTM = methotrimeprazine; PCB placebo; PMZ = promethazine. Venue – CL = clinic; ED = emergency department; NA = not applicable.

Cicek (2004)1765564157
Larkin (1992)1130644133873
Richman (2002)1249534333
Harden (1996)96055
Plus
Scherl (1995)14  71863354
Plus
Davis (1995)7212255506450
Plus
Stiell (1991)15  595177822730
Plus
Lane (1989)10  568550 8
Plus
Duarte (1992)8  2024
Plus
Carleton (1998)6  5654293614321821
Plus
Belgrade (1989)13  038>90 pain reduction
Plus
Klapper (1993)5  2193  
Plus
Engindeniz (2005)1635458080656070652020
Alemdar (2007)3291271355929
Tek (1987)2  5523

NSAIDS

Non-steroidal anti-inflammatory drugs (NSAIDs) can inhibit the neuroinflammatory cascade, prostaglandin synthesis, and platelet aggregation associated with the release of vasoactive substances, all processes that are involved in the initiation and prolongation of migraine. The cyclooxygenase COX1/COX2 inhibitors such as ketorolac and indomethacin can inhibit the release of prostaglandins that activate nociceptive neurons in the spinal trigeminal nucleus, a process that leads to central sensitization in migraine.18

Seim et al found ketorolac 30 mg IV reduced headache less at 1 hour than prochlorperazine 10 mg IV (−40 vs −71, P = .04).19 Meredith et al found ketorolac 30 mg IV to be more effective than sumatriptan 20 mg nasal spray.20 Jakubowski et al21 compared ketorolac 15 mg IV bolused twice successively (30 mg total) with sumatriptan 6 mg subcutaneous (SQ) for the delayed treatment (4 hours) of migraine with cutaneous allodynia. No patient in the sumatriptan group was pain free after 2 hours, compared with 64% in the ketorolac group at 1 hour (P < .05). Patients in the sumatriptan group were then given ketorolac after the 2-hour assessment, and an hour later, 71% of these patients were pain free (P < .05). All patients who were pain free were also allodynia free (and vice versa). There were 9 patients who did not respond to ketorolac. All of these nonresponders had a history of opioid use, as compared with only one of the 19 patients who did respond to ketorolac (P < .05).

Three studies comparing ketorolac 60 mg IM with a combination of meperidine and promethazine or hydroxyzine yielded similar results. Davis et al compared ketorolac 60 mg IM with promethazine 25 mg IM plus meperidine 75 mg IM and found that there was no difference between the 2 groups in percentage pain free at 30 minutes (promethazine/meperidine 20.8% vs ketorolac 22.3%).7 Harden et al found no difference in headache relief between ketorolac 60 mg IM (44.4%), meperidine 50 mg plus promethazine 25 mg IM (60%), or placebo/NS IM (54.5%).9 Duarte et al found no difference in pain reduction (VAS) for ketorolac 60 mg IM compared with meperidine 100 mg IM plus hydroxyzine 50 mg IM (−33.5 vs −33.7, P = .76); the percentage complaining of nausea and drowsiness was not significantly less for ketorolac (28% vs 48%, P = .15).8

Larkin and Prescott compared ketorolac 30 mg IM with meperidine 75 mg IM; the percentage pain free at 1 hour was greater for meperidine (30% vs 6%, P < .05), as was sustained pain freedom at 24 hours (44% vs 13%, P < .02), and no adverse events were reported in either group.11 Klapper and Stanton found that ketorolac 60 mg IM was less effective in treating migraine than DHE 1 mg IV plus metoclopramide 5 mg IV; headache relief (4-PPS) at 1 hour was greater for DHE/metoclopramide (78% vs 33%, P = .031).22

Bigal et al compared metamizole (MMZ, also called dipyrone and not available in the USA) 1000 mg IV with placebo/NS IV.23 For migraineurs without aura, pain reduction (11-PPS) with MMZ was more effective than placebo at 30 and 60 minutes (−4.0 vs −1.2, P < .01 and −6.0 vs −3.0, P < .01). For patients with aura, the reduction was also significantly greater with MMZ at 30 and 60 minutes (−4.9 vs −0.9, P < 0.01 and −6.4 vs −1.9, P < .01). Krymchantowski et al then compared MMZ 1000 mg IV with lysine clonixinate (LC, an NSAID not available in the USA) 200 mg IV, delivered as a 25 mL infusion over 5 minutes.24 The percentage pain free at 60 and 90 minutes was greater with LC (60 minutes: 33% vs 13%, P < .001; 90 minutes: 86.7% vs 73%, P < .001). There was no follow-up post-discharge. Patients receiving LC had more injection-site pain than those receiving MMZ (13/15 vs 3/15, P < .0001). Engindeniz et al compared diclofenac sodium 75 mg IM with tramadol 100 mg IM. Headache relief at 2 hours was the same for both groups (80%).16 Ellis et al compared ibuprofen 600 mg oral (PO) with metoclopramide 10 mg IV and placebo; pain reduction (VAS) was similar for metoclopramide and metoclopramide/ibuprofen (−75 vs −50) but was greater in both these groups (P < .01) than in the ibuprofen or placebo groups, which were the same (−25).25

Table 2 summarizes the studies involving ketorolac, diclofenac, ibuprofen, LC, and MMZ (the last not an NSAID).

Table 2.—. Studies Comparing Ketorolac (KET), Diclofenac (DCF), Ibuprofen (IBU), Metamizole (MMZ), and Lysine Clonixinate (LYC) Individually to Other Agents for Migraine Therapy. Superior Performance is Indicated by Boldfaced Characters; if Treatment 1 Superior, Then Also Underlined; if Treatment 2 Superior, Then Also Italicized
Study First Author (year)Group 1 (G1) Treatmentmg or mg/kgRouteGroup 2 (G2) Treatmentmg or mg/kgRouteVenue – # If Multiple SitesResearch Design% WomenMean AgeGroup 1 # PtsGroup 2 # Pts
Seim (1998)19KET30IVPCZ10IVEDR/DB92333529
Jakubowski (2005)21KET30IVSTP6SQEDR/DB1414
Meredith (2003)20KET30IVSTP20NSEDR/DB86341316
Larkin (1992)11KET30IMMEP75IMEDR/DB77331516
Harden (1996)9KET60IMMEP50IMEDR/DB/P80331010
PlusPMZ25IM 
also vsPCBNAIM 10
Duarte (1992)8KET60IMMEP100IMEDR/DB80352525
PlusHDX50IM 
Klapper (1991b)22KET60IMDHE1IVCLR/DB99
PlusMTC5IV 
Davis (1995)7KET60IMMEP75IMEDR/DB81352022
PlusPMZ25IM 
Engindeniz (2005)16DCF75IMTMD100IMEDR/DB2020
Ellis (1993)25IBU600POPCBNAPOEDR/DB/P1010
Bigal (2001)23MMZ1000IVPCBNAIVEDR/DB/P73329530
 NAwith aura69355430
Krymchantowski (2008)24MMZ1000IVLYC200IVEDR/SB93311515
Study First Author (year)% Pain Free% Pain Relief% Sustained Pain Relief% Disability Free% Requiring Rescue% Patient Satisfaction
G1G2G1G2G1G2G1G2G1G2G1G2
  1. — = no data obtained. Research design – DB = double blind; P = placebo controlled; R = randomized; SB = single blind. Route – IM = intramuscular; IV = intravenous; NS = nasal spray; PO = oral; SQ = subcutaneous. Treatments = DHE = dihydroergotamine; HDX = hydroxyzine; MEP = meperidine; MTC = metoclopramide; PCB = placebo; PCZ = prochlorperazine; PMZ = promethazine; STP = sumatriptan; TMD = tramadol. Venue – CL = clinic; ED = emergency department; NA = not applicable.

Seim (1998)1952681714
Jakubowski (2005)21640
Meredith (2003)207727
Larkin (1992)1163013447338
Harden (1996)94460
Plus
also vs55
Duarte (1992)82420
Plus
Klapper (1991b)22  3378891006711
Plus
Davis (1995)7222150645055
Plus
Engindeniz (2005)16453580806065657020208075
Ellis (1993)252931
Bigal (2001)237538
 7623
Krymchantowski (2008)247387

STEROIDS

The rate of headache recurrence within 24-72 hours following discharge from the ED can exceed 50%, and corticosteroids typically are used in the ED in an attempt to reduce the frequency of such recurrence.26,27 The steroids can act to suppress the sterile inflammation underlying migraine and may reduce the likelihood of headache recurrence and ED “recidivism.”

Klapper and Stanton compared dexamethasone 6 mg IV plus metoclopramide 5-10 mg IV with DHE 0.75-1 mg IV plus metoclopramide 5-10 mg IV and to placebo/NS IV; both DHE/metoclopramide (82%) and dexamethasone/metoclopramide (78%) provided a greater percentage of patients with headache relief at 30 minutes than placebo (20%, P < .002) but were not significantly different from one.28 Baden and Hunter compared dexamethasone 10 mg IV with placebo/NS IV as adjuvant treatment to prevent recurrence 48-72 hours post-ED discharge; headache recurrence was lower with dexamethasone (12.9% vs 58.3%, P < .001), and there was an equal incidence of side effects (19.4% vs 20.8%, P = 1.0), none serious.29 Friedman et al compared dexamethasone 10 mg IV with placebo/NS IV for prevention of headache recurrence within 24 hours.30 Both groups received metoclopramide 20 mg plus diphenhydramine 25 mg IV (which could be repeated twice) for initial treatment. The percentage headache free at discharge who remained so at 24 hours was similar (dexamethasone 25% vs placebo 19%, P = .34). When the subgroup of patients whose headache duration was more than 72 hours at ED presentation was analyzed separately, the difference in sustained pain freedom almost met the criterion for statistical significance (dexamethasone 38% vs placebo 13%, P = −.06). In the dexamethasone group, 6% reported a “burning sensation” at the injection site. Rowe et al compared dexamethasone 15 mg IV with placebo/NS IV in preventing migraine recurrence 48-72 hours and 7 days post-discharge.31 The percentage reporting severe headache recurrence was similar for both dexamethasone and placebo at 48-72 hours (22% vs 32%) and at 7 days (28% vs 40%). Of note, headache recurrence was more likely to occur if the pain rating on the VAS at discharge was >20 mm (P < .05). Innes et al compared dexamethasone 24 mg IV with placebo/NS IV in preventing recurrence of severe headache.27 Although the percentage with severe headache at 48 hours was greater for placebo (18% vs 45%, P = .005), there was no difference in the frequency of experiencing any degree of headache recurrence (65% vs 67%). Thirty-seven adverse events were reported for dexamethasone and 47 for placebo, the most common being drowsiness (34%), restlessness (24%), and nausea (21%). Donaldson et al compared dexamethasone 24 mg IV with placebo/NS IV, and the rate of headache recurrence was similar in the 2 groups at both the 3-day (dexamethasone 35% vs placebo 45%, P = .38) and 30-day follow-up (43% vs 47%, P = .68).32 Feisseler et al compared either dexamethasone 10 mg IV or prednisone 40 mg PO daily ×2 days vs placebo (either NS IV or lactulose PO).33 Only patients with IV access were given IV steroid or placebo. Headache recurrence at 24-72 hours of follow-up was not significantly different for steroid vs placebo (22% vs 32%, respectively; P = .21).

Table 3 summarizes results from the studies involving steroids.

Table 3.—. Studies Comparing Magnesium and Dexamethasone (DXM) Individually to Other Agents for Migraine Therapy. Superior Performance is Indicated by Bolding; if Treatment 1 Superior, Then Also Underlined; if Treatment 2 Superior, Then Also Italicized
Study First Author (year)Group 1 (G1) Treatmentmg or mg/kgRouteGroup 2 (G2) Treatmentmg or mg/kgRouteVenue – # If Multiple SitesResearch Design% WomenMean AgeGroup 1 # ptsGroup 2 # pts
Klapper (1991a)28DXM6IVPCBNAIVCLR/DB/P1110
 MTC5-10IV         
Friedman (2007)30DXM6IVPCBNAIVED4R/DB/P853710699
 MTC5-10IVMTC5-10IV      
Innes (1999)27DXM24IVPCBNAIVCLR/DB/P80354949
Donaldson (2008)32DXM24IVPCBNAIVED2R/DB/P81366253
Rowe (2008)31DXM15IVPCBNAIVED2R/DB/P81356462
Fiesseler (2011)33DXM10IVPCBNAIVEDR/DB/P86379182
ORPRD40 × 2POPCBX2PO      
Baden (2006)29DXM10IVPCBNAIVEDR/DB/P64343124
Study First Author (year)% Pain Free% Pain Relief% Sustained Pain Relief% Disability Free% Requiring Rescue% Patient Satisfaction
G1G2G1G2G1G2G1G2G1G2G1G2
  1. — = no data obtained. Research design – DB = double blind; P = placebo controlled; R = randomized; Route – IV = intravenous; PO = oral. Treatments – MTC = metoclopramide; PCB = placebo; PRD = prednisone. Venue – CL = clinic; ED = emergency department; NA = not applicable.

Klapper (1991a)28  8220  450    
Friedman (2007)3055472519675913137678
Innes (1999)27    8255      
Donaldson (2008)32    6555    7271
Rowe (2008)31    7868      
Fiesseler (2011)33    7868  3644  
OR            
Baden (2006)29  70668742      

POST-DISCHARGE MEDICATIONS

Many headache patients leave the ED without complete headache relief, and up to 60% still are experiencing significant pain 24 hours post-discharge.34-36 In addition, and as stated in the previous section on steroids, even many of those who respond to initial treatment have recurrence of their headaches within 24-72 hours.

In an open-label study, Akpunonu et al provided all patients with sumatriptan 100 mg PO to use if their headaches recurred after discharge; of the 92 patients (88% women, mean age 40) who had mild or no pain on discharge, 57 (62%) had recurrence of their headache, and 37/57 (65%) reported meaningful relief 1 hour after taking sumatriptan.37 Using a randomized, double-blind design, Friedman et al compared sumatriptan 100 mg PO with naproxen 500 mg PO for use as a rescue medication for post-discharge headaches.38 The majority of the patients were treated with a parenterally administered dopamine antagonist prior to discharge. By 48 hours post-discharge, 280/410 patients (68%) had experienced recurrent headache, 196 took medications, and response to rescue this therapy could be analyzed in 188 (naproxen: 96 patients and oral sumatriptan: 92 patients; 85% were women and mean age was 36). Pain reduction (11-PPS) was equivalent in the 2 groups (naproxen – 4.3 vs sumatriptan −4.1). The percentage of patients who reported they would utilize their assigned rescue medication again was similar (naproxen 71% vs sumatriptan 75%), as was the percent reporting side effects (naproxen 19% vs sumatriptan 26%), with the most common complaints in both groups being GI upset, dizziness, and drowsiness.

Secobarbital is a relatively long-acting barbiturate with hypnotic action. In a randomized, double-blind, placebo-controlled trial, Gerhardt et al compared secobarbital 100 mg PO with placebo for prevention of migraine recurrence.39 Two tablets were provided to each patient upon discharge. One tablet was to be taken upon arriving at home. Patients were instructed to sleep, and if not asleep in 1 hour, to take a second dose. The percentage of patients with sustained headache relief at 24 hours was greater for secobarbital than placebo (94% vs 50%, P < .02). No adverse events were reported.

DISCUSSION AND CONCLUSIONS: OPIOIDS, NSAIDS, STEROIDS, AND POST-DISCHARGE MEDICATIONS

The opioids assessed, meperidine, tramadol, and nalbuphine, were superior to placebo in relieving migraine pain, although meperidine combined with promethazine was not. Meperidine 75 mg was superior to ketorolac 30 mg IM in providing headache relief but was similar to ketorolac 60 mg IM even when combined with an antihistamine. Meperidine 75 mg IM or 1.5 mg/kg IV was similar in pain relief to DHE 0.5 mg IV but inferior to DHE 1 mg IV; a factor complicating the interpretation of these comparisons, however, is that both drugs were combined with different anti-emetics/antihistamines in 3 out of the 4 studies. Meperidine was also similar to methotrimeprazine in pain relief but was inferior to chlorperazine. Meperidine was not compared with prochlorperazine, droperidol, or sumatriptan. Lastly, tramadol 100 mg IM yielded pain relief similar to diclofenac 75 mg IM.

Headaches treated with opioids have a high rate of recurrence after the analgesic effect of the opioid wears off. Opioids may, however, make the patient drowsy enough to sleep, which often terminates headaches. Side effects of opioids included sedation, dizziness, GI discomfort, nausea and vomiting, and akathisia (although the last occurs much less frequently than with the dopamine antagonists).

Opiates/opioids often are used in treating patients with status migrainosus. In one study, however, patients in status became pain free after treatment with ketorolac and sumatriptan only if they had used no opiates/opioids whatsoever in the previous 6 months.22 This result reinforces the notion that opioids may exert a persistent pro-nociceptive effect – a “pain-memory state”– that prevents the reversal of central sensitization.4 Along the same line, Ho et al found that rizatriptan was less effective as a migraine abortive if patients had recent exposure to opioids.40

As such, concern for creating opiate/opioid dependency in migraine patients may not be the most important reason for not recommending this class of drugs as first-line therapy. Even so, the habituation rate associated with the use of opioids for migraine treatment is estimated at 13 out of 1 million patients, and ED staff rightfully are concerned at the prospect of contributing to an already established habituation or even true addiction, especially in patients with the latter who misrepresent themselves as having headaches.41,42 When one considers that the opioids studied were superior only to placebo and ketorolac 30 mg IM, the issue of their potential pro-nociceptive effect, abuse/addiction issues, and the evidence that the use of opioids appears to render relatively migraine-specific abortive medications less effective, it is recommended that opiates/opioids not be used as first-line therapy for migraine pain in the ED or clinic.

Ketorolac is a cyclooxygenase COX1/COX2 inhibitor that appears to be able to reverse both peripheral sensitization by inhibiting the neuroinflammatory cascade in the meninges and central sensitization associated with cutaneous allodynia. Ketorolac effectively treated acute migraine in patients with cutaneous allodynia who did not respond to sumatriptan SQ. Ketorolac 30 mg IV or 60 mg IM is more efficacious than sumatriptan nasal spray and less efficacious than prochlorperazine or DHE plus metoclopramide but similar to meperidine plus promethazine or hydroxyzine. In the only study which included a placebo arm, an unusually high rate of pain relief was reported by the both the ketorolac and placebo groups. Halving its IM dose to 30 mg resulted in ketorolac being less efficacious in reducing pain than meperidine.

Two studies investigated the pain-reducing efficacy of IV MMZ (not an NSAID) and LC (an NSAID), 2 parenteral medications commonly used for pain relief in Brazil but not available in the USA. Pain reduction was greater with MMZ than placebo for migraine both with and without aura. LC, in turn, outperformed MMZ in terms of the percentage pain free at 90 minutes, although the response rates for both were fairly impressive (LC 86.7% vs MMZ 73%). Diclofenac sodium was as effective as tramadol in reducing migraine pain, but ibuprofen PO was not more effective than placebo.

The NSAIDs are nonsedating and can be combined with most other medications. While side effects can occur even after a single dose, these side effects generally are fewer and milder than those associated with narcotics, opiates/opioids, and dopamine antagonists. Contraindications to NSAID use include a history of GI bleeding, other bleeding risks, and renal impairment. Due to a recent report demonstrating a significantly increased risk of miscarriage, pregnancy now may represent a contraindication to NSAID use, and NSAIDs should not be administered to nursing mothers.

There was a significant difference in the percentage of patients with sustained headache relief in 2 of the 6 studies that compared dexamethasone IV (6-24 mg single dose in ED) or oral prednisone (40 mg times 2 days) with placebo. In all 4 remaining studies, the number of patients with sustained pain relief was greater in the steroid groups, although not sufficient to be significant. Using a paired t-test on the percentage headache relief from these 6 studies, there is an overall significant difference between these scores for patients receiving dexamethasone compared with those receiving placebo (dexamethasone 69% vs placebo 51%, t = 2.9, d.f. = 5, P = .03).

Side effects of single-dose dexamethasone were relatively few and benign. Patients receiving dexamethasone were more likely to report dizziness and less likely to report nausea than patients receiving placebo. There were equally low frequencies of reported numbness/tingling, drowsiness, restlessness, and swelling in the both steroid and placebo arms. Repeated long-term use of steroids may increase the risk for osteoporosis and aseptic osteonecrosis of the femoral head and knees.43 Steroids should be used with caution in patients with diabetes.

When patients are treated in the outpatient setting for headache lasting more than 72 hours, they may receive a course of steroids (dexamethasone, prednisone) for 3-5 days or more until headache free for 24 hours.27 A single dose of dexamethasone IV (or 2 daily doses of prednisone used in 1 study) may not be sufficient to produce much of an effect on the rate of headache recurrence.

Patients in the 2 studies designed to treat headache recurrence experienced a recurrence rate exceeding 60%. The study designed to prevent headache recurrence by inducing sleep with secobarbital demonstrated a 6% headache recurrence rate in the secobarbital group vs 50% in the placebo group. Both oral sumatriptan 100 mg and naproxen sodium 500 mg appear effective more often than not in relieving headache that recurs after ED discharge. The use of secobarbital, with its sedative and sleep-inducing properties, appears to increase the percentage pain relief after ED discharge compared with those patients in the placebo group.

GENERAL DISCUSSION AND CONCLUSIONS

This section is intended as a general discussion of all the studies presented in this 3-part review of physician-administered rescue therapy for acute migraine in the ED, urgent care, and headache clinic infusion center settings. The conclusions are based on the current paper and those published previously in this journal.44,45

Analysis of the large number of studies presented in this review confirms that a definitive and optimally effective ED migraine rescue regimen cannot be determined on the basis of current published data. In an attempt to compare migraine treatments without relying solely on the pair-wise comparisons typical of the methodology used in the studies reviewed, the author determined and compared weighted averages of the percentages of pain relief. These weighted averages were computed for all medications for which there were 2 or more randomized trials with the medication used as a single agent (that is, was not combined with any other medications). The following lists these average percentages of pain relief from greatest to least (total number of patients represented is in parentheses): droperidol 82% (229), sumatriptan 78% (351), prochlorperazine 77% (312), tramadol 76% (37), metamizole (not available in the USA) 75% (164), metoclopramide IV 70% (184), DHE 67% (188), chlorpromazine 65% (158), ketorolac 30 mg IV 60% (77), meperidine 58% (79), metoclopramide IM 45% (128), magnesium 43% (169), ketorolac 60 mg IM 37% (64), and valproate 32% (39) (Fig. 1). Of note, promethazine is not listed here in this analysis because it was only studied in combination with meperidine.

Figure 1.—.

Weighted averages of the percentages of pain relief for all medications for which there were ≥2 randomized trials with the medication used as a single agent. DHE = dihydroergotamine; IM = intramuscular; IV = intravenous.

The weighted averages for percentages of patients who were pain free were also computed for all medications for which there were 2 or more randomized trials where the medication was used as a single agent. These are presented from greatest to least as follows (total number of patients represented is in parentheses): 53% for both prochlorperazine (90) and chlorpromazine (115), droperidol 40% (214), magnesium 36% (91), sumatriptan 35% (166), tramadol 32% (37), and DHE 21% (34) (Fig. 2). Magnesium fared much better in this analysis than in the analysis of headache relief.

Figure 2.—.

Weighted averages of the percentages of pain free for all medications for which there were ≥2 randomized trials with the medication used as a single agent. DHE = dihydroergotamine; IM = intramuscular; IV = intravenous.

The choice of headache treatment should be based on considerations of efficacy, side effects, and cost. The calculated expense of a treatment should include not only immediate ED treatment, but also the hidden costs of continuing headache, early headache recurrence, functional disability, return trips to the ED, and the need for follow-up care in an outpatient clinic. Headache recurs in more than 50% of patients after ED discharge. Ducharme et al found that those who are pain free at discharge are significantly less likely to have recurrence, but Friedman et al could not independently support this finding.1 When they looked at their ED records for administered medications, Sheftell et al also reported a link between low recurrence and pain-free response with naratriptan PO.46 Overall, it seems reasonable that a concerted effort should be made to discharge patients from the ED pain free.

Parenterally administered dopamine antagonists are not only effective anti-emetics but also can reduce or terminate migraine headache. As a class, however, they frequently cause side effects (including sedation, akathisia, and dystonia) that can outlast the symptoms of the migraine itself and thereby prolong patients' functional disability. There is a need to pre-dose patients receiving phenothiazines (especially chlorpromazine) with IV fluid to prevent postural hypotension, as well as with a drug possessing anticholinergic properties to reduce the likelihood of extrapyramidal side effects. Droperidol and haloperidol are not recommended as first-line therapy because of potential QTc prolongation and the consequent need for electrocardiogram monitoring.

For a variety of reasons (discussed in some detail earlier in this paper), opiates/opioids generally are not recommended as first-line treatment for migraine. One argument used in support of using opioids for first-line treatment is that they are quick to administer and act rapidly, such that patients can be discharged in a timely fashion. It remains unclear, however, whether the use of opioids does indeed save time. Coleman et al assessed migraine treatment patterns in 5 linked Canadian EDs.47 Opiates/opioids were used as first-line treatment for 59.6% of the migraine patients. The odds of receiving an opioid as first-line therapy was increased in patients who took medications prior to ED admission and was decreased in patients who had a longer-lasting headache or a more urgent triage score. Those who received opioids first line spent less total time in the ED (177 minutes vs 237 minutes, P < .001). This contrasts with the findings of Tornabene et al, who found that patients who received opioids spent more time in their ED than patients who did not (160 minutes vs 125 minutes, P = .015), regardless of whether they often sought treatment for headache in the ED.48

Sumatriptan SQ, a relatively migraine-specific medication, is as effective as droperidol and prochlorperazine in providing pain relief. When limited to patients with no contraindications, it is very well tolerated. Adverse events are generally limited to transient chest tightness, shoulder pain, and neck pain, all of which rarely outlast the migraine pain itself and require no treatment to resolve.

An argument can be made for the use of fixed drug combinations to treat migraine. These combinations can target multiple symptoms and have synergistic effects, and there is also the possibility that 1 or more medications can counteract side effects of another medication when they are combined. One such combination is DHE combined with metoclopramide intravenously, with the latter treating the nausea induced by the former. Both of these medications were individually as effective as chlorpromazine in providing pain relief, although the combination did not appear to have additional pain-relieving effects.

Magnesium can be an effective agent in migraine treatment, either alone or as adjuvant treatment, especially in those patients who have aura. It provided as much freedom from migraine pain (with and without aura) as did sumatriptan or ketorolac IV, although average pain relief was relatively low, only surpassing ketorolac IM and valproate.

There is some support for using corticosteroids to prevent headache recurrence, especially if the presenting migraine has lasted longer than 72 hours. The optimal regimen likely involves IV doses in the ED followed by oral dosing for several days post-discharge.

It is recommended that all future studies be randomized and double blinded (including double dummy). For those rescue treatments for which there have been insufficient studies employing a placebo arm, it is recommended that this be done to ascertain effectiveness more accurately. When drug combinations are compared, the same second agent should be used in both or all arms. An ideal design for studying drug combinations would have 4 arms: drug A/placebo B, drug B/ placebo A, drug A/ drug B, and placebo A/ placebo B. Recording of headache status should be done at 2 hours (in addition to any other outcome measures), even if this involves the patient reporting their status post-discharge, in order to determine the percent pain free at 2 hours. All studies should include 24-72 hour follow-up evaluations.

In summary, the ideal acute migraine rescue therapy administered in the urgent care or ED setting would provide complete headache relief, possess no side effects, and prevent early headache recurrence. Because no such therapy currently exists, treatment must be tailored to the needs of the individual patient.

Triptans and DHE are most effective in injectable formulations and should be avoided in those at risk for vascular complications. DHE is particularly effective when given IV but can cause increased nausea.

NSAIDs such as ketorolac have the advantage of being appropriate for patients with vascular risk factors, and they do not cause sedation. They can be combined with most other treatments for increased efficacy.

Dopamine antagonists can be given alone or with other agents. As this review indicates, they are surprisingly effective for migraine, even late in the attack, and they are inexpensive. When used alone, prochlorperazine, droperidol, and metoclopramide were superior to placebo. Droperidol and prochlorperazine were superior or equal in efficacy to all other treatments, but they commonly cause side effects that are especially unpleasant for patients (notably dystonia and akathisia). Metoclopramide was equivalent to prochlorperazine, and, when combined with diphenhydramine, was superior in efficacy to triptans and NSAIDs. Meperidine was arguably equivalent when compared with ketorolac and DHE but was inferior to chlorpromazine and equivalent to the other neuroleptics. Sumatriptan was inferior or equivalent to the neuroleptics and equivalent to DHE when considering only paired comparisons. The overall percentage of patients with pain relief after taking sumatriptan was equivalent to droperidol and prochlorperazine.

Once again, opiate/opioid rescue sometimes can be effective, but such therapy also may lead to early headache recurrence, central sensitization, sedation, nausea and dizziness, as well as raise concerns for overuse and abuse. While commonly administered for treatment of acute migraine, ideally, these medications should be a last resort.

Magnesium can be an effective treatment for migraineurs with aura and can reduce the photophobia and phonophobia of all migraineurs. It can be added on to any of the said medications to boost effectiveness without sedation. Magnesium also can be very useful as a therapy for pregnancy-associated migraine.

STATEMENT OF AUTHORSHIP

Category 1

  • (a)Conception and Design
    • Nancy E. Kelley, Deborah E. Tepper

  • (b)Acquisition of Data
    • Nancy E. Kelley, Deborah E. Tepper

  • (c)Analysis and Interpretation of Data
    • Nancy E. Kelley, Deborah E. Tepper

Category 2

  • (a)Drafting the Article
    • Nancy E. Kelley, Deborah E. Tepper

  • (b)Revising It for Intellectual Content
    • Nancy E. Kelley, Deborah E. Tepper

Category 3

  • (a)Final Approval of the Completed Article
    • Nancy E. Kelley, Deborah E. Tepper

Ancillary