The Safety of Celecoxib as an Acute Treatment for Migraine: A Narrative Review

Abstract

Introduction

Nonsteroidal anti-inflammatory drugs (NSAIDs) have been the first-line choice for the acute treatment of migraine attacks for decades; however, the safety of a particular NSAID is related to its treatment dose, duration, and mechanism of action. Although adverse event (AE) risks differ substantially among individual migraine treatments, increased or prolonged exposure to any NSAID elevates risks and severity of AEs.

Methods

For this narrative review, we conducted a literature search of PubMed until July 2022, focusing on the history, mechanism of action, and treatment guidelines informing the safety and efficacy of celecoxib oral solution for the acute treatment of migraine attacks.

Results

Here we discuss the mechanisms of action of nonselective NSAIDs vs. cyclooxygenase-2 (COX-2) inhibitors, and how these mechanisms underlie the AEs associated with these treatments. We review the clinical trials that influenced the regulatory history of NSAIDs, specifically COX-2 inhibitors, the role of traditional and new formulations of NSAIDs including celecoxib oral solution, and special considerations in the acute treatment of migraine attacks.

Conclusions

Low-dose formulations of NSAIDs, such as celecoxib oral solution, provide acute migraine analgesia with similar or fewer associated cardiovascular and gastrointestinal events than previous formulations.

Key Summary Points

The treatment dose, duration of therapy, and mechanism of action contribute to the safety of a nonsteroidal anti-inflammatory drugs (NSAID).

Inhibition of the cyclooxygenase (COX)-1 enzyme leads to increased gastric bleeding, whereas COX-2 inhibition may increase the frequency of cardiovascular events.

In 2020, the US Food and Drug Administration (FDA) approved celecoxib oral solution, a COX-2 selective inhibitor, for the acute treatment of migraine in adults with or without aura.

As the FDA recommends using the shortest duration and lowest effective dose of NSAIDs, celecoxib oral solution is particularly suited for intermittent treatment of acute migraine pain, providing its effective analgesia with relatively low overall drug exposure and AE risks.

Introduction

For over a century, nonsteroidal anti-inflammatory drugs (NSAIDs) have been among the most widely used treatments worldwide [1]. Nonprescription and prescription NSAID formulations are available to treat a variety of inflammatory and pain conditions, including migraine [2]. Compared with other treatments used for migraine, NSAIDs function in both the central and peripheral nervous systems and can provide analgesia at any time during a migraine attack [2–5]. Up to half of all persons with migraine use only nonprescription NSAIDs to treat their attacks, often taking suboptimal dosages [6].

The treatment dose, duration of therapy, and mechanism of action contribute to the safety of a particular NSAID. NSAIDs reduce prostaglandin synthesis by either nonselective cyclooxygenase [COX] inhibition (nsNSAIDs) or selective COX-2 inhibition [7]. The risk of adverse events (AEs) differs substantially among migraine drug treatments, but increased or prolonged exposure to any NSAID heightens the risks and worsens the severity of potential AEs [7–10]. Well-established side effects associated with NSAID usage include an increased risk of gastrointestinal (GI) and cardiovascular (CV) AEs, such as GI bleeding, myocardial infarction (MI), and stroke [3, 4]. Approximately 25% of the US population who have taken NSAIDs required medical attention due to side effects [11]. Serious, yet rare AEs associated with NSAIDs include acute kidney damage, chronic kidney disease, hepatotoxicity, acute and chronic hepatitis, Stevens–Johnson syndrome, skin photosensitivity, and toxic epidermal necrolysis [12, 13].

NSAIDs are a first-line therapy for the acute treatment of migraine attacks [14]. Despite the associated risks, NSAIDs provide established safety and efficacy and remain integral in acute migraine treatment. Current NSAIDs approved by the US Food and Drug Administration (FDA) for the acute treatment of migraine attacks in adults include aspirin, ibuprofen (Advil® migraine), and diclofenac potassium powder for oral solution (Cambia®) [2]. In 2020, the FDA also approved the selective COX-2 inhibitor celecoxib oral solution (Elyxyb™) due to its proven efficacy in the acute treatment of migraine attacks [15, 16]. Often used off-label as a monotherapy, naproxen is FDA-approved for acute migraine treatment but only when used in combination with sumatriptan (Treximet®) [17, 18]. NSAIDs prescribed off-label to treat migraine with limited efficacy data are flurbiprofen, ketoprofen, and ketorolac [2].

Apropos to the use and concerns of NSAIDs for migraine therapy, here we review the history, mechanism of action, and treatment guidelines informing the safety of celecoxib oral solution for the acute treatment of migraine attacks [19]. This manuscript is based on previously performed and published studies and does not contain any new human participant or animal studies performed by any of the authors.

History of Safety Concerns of COX-2 Inhibitors and Nonselective NSAIDs

Inhibition of both COX-1 and COX-2 activity and subsequent reduction of prostanoid production drives the function of nonselective NSAIDs (nsNSAIDs) and increases the risk of AEs (Fig. 1). Platelets are a major source of the platelet aggregating and vasoconstricting hormone thromboxane (TxA₂), which is produced from arachidonic acid by COX-1. NSAIDs (e.g., low-dose aspirin treatment) provide cardioprotection by inhibiting COX-1, thereby reducing TxA₂ [20–22]. The gastric mucosa also produces TxA₂, which functions locally to provide gastric protection by increasing blood flow, mucus production, and bicarbonate secretion [23]. Inhibition of gastric TxA₂ leads to the gastric irritation and bleeding associated with NSAID use [24]. Prostaglandin I₂ (PGI₂, prostacyclin) is produced from arachidonic acid by COX-2 in cardiac, renal, and endothelial tissue [4, 20]. PGI₂ normally functions to increase nitric oxide generation, vasodilation, and pain signaling [4, 20]. Further, PGI₂ suppresses platelet aggregation, thereby counteracting TxA₂ and maintaining platelet activity in a COX-dependent manner. COX-2 inhibition produces anti-inflammatory and analgesic effects by reducing the production of PGI₂ [4, 20]. However, the reduction of PGI₂ increases platelet aggregation and may be the mechanism underlying the increased thrombotic CV AEs (e.g., stroke, MI) associated with COX-2 inhibitors, albeit without conclusive proof [22, 25, 26].

Fig. 1.

Effects of COX inhibition on prostaglandin signaling in the GI, cardiovascular, and renal systems and the associated AEs related to COX inhibition. Created with BioRender.com. AEs adverse events, COX cyclooxygenase, GI gastrointestinal, MI myocardial infarction, nsNSAID nonselective nonsteroidal anti-inflammatory drug, PGE2 prostaglandin E2, PGI2 prostaglandin I2 (prostacyclin)

Due to its selective COX-2 inhibition, increased CV events were hypothesized before the original FDA approval of celecoxib for the treatment of arthritis pain in 1998 [4, 27]. Celecoxib reduces PGI₂, which may lead to the three potential mechanisms by which selective COX-2 inhibitors cause CV AEs: elevated blood pressure (BP), accelerated atherogenesis/atherosclerosis, and exaggerated thrombotic response to atherosclerotic plaque rupture [4, 25, 27]. Deletion of the COX-2 gene increased BP in preclinical mouse models [28]; however, later clinical trials reported heterogenous effects on BP by COX-2 selective inhibitors, as rofecoxib, but not celecoxib, increased BP [29–31]. Studies in mice also showed that PGI₂ protects from atherogenesis and atherosclerosis, but a reduction of PGI₂ by COX-2 inhibition may diminish this protection [4]. Additionally, COX-2 inhibitors may accelerate atherogenesis by suppressing production of PGE₂, a prostaglandin formed by both COX-1 and COX-2 [4]. Finally, PGI₂ is a potent inhibitor of platelet function, and COX-2 inhibition accelerated thrombotic processes in several animal models [4]. Together, these three mechanisms may contribute to the CV AEs sometimes observed with NSAIDs, particularly selective COX-2 inhibitors.

In clinical trials, elevated CV risks appear to be class-specific to selective COX-2 inhibitors with increased COX-2 specificity, higher dose, and longer duration increasing the AE risk exposure [21, 29–32]. Many NSAIDS (e.g., ibuprofen, diclofenac, naproxen, and celecoxib) increase the risk of heart failure by approximately two-fold, linked in part to the renal COX-2-mediated effects on BP and sodium retention [22, 32]. In 2005, the Adenomatous Polyp Prevention on Vioxx (APPROVe) trial found elevated CV risks (i.e., thrombotic events, heart failure, or pulmonary edema) in participants taking rofecoxib compared with placebo approximately 18 months after treatment initiation [33]. The Adenoma Prevention With Celecoxib (APC) trial also reported elevated risk of death from CV causes, MI, stroke, or heart failure beginning approximately 15 months after initiation of celecoxib compared with placebo [34]. The Celecoxib for the Prevention of Colorectal Sporadic Adenomatous Polyps (PreSAP) trial did not replicate these findings of increased CV risk [35, 36]. By 2008, evidence was published that the risk of MI increased with higher doses and longer duration of treatment with NSAIDs; low-dose celecoxib (< 200 mg/d) only slightly elevated MI risk (relative risk [RR]: 1.23), whereas higher doses exhibited a more pronounced increase in MI risk (RR: 1.57) [7]. These data support avoidance of COX-2 inhibitor use in patients with, or at high risk for, CV disease [20].

All NSAIDs carry some degree of increased risk of CV events [3, 37]. Due to its high RR of MI and CV events, valdecoxib (Bextra®) was withdrawn from the market, and for the same reason, the FDA never approved lumiracoxib [4, 38]. Diclofenac, which is a partially selective COX-2 inhibitor, also slightly increases RR for MI. Although generally considered safe, treatment with ibuprofen or naproxen elevates the risk of CV events, with ibuprofen also increasing the risk of GI and renal AEs as well as death from any cause [7, 32, 37–39]. The similar AE rates reported for ibuprofen and naproxen in clinical trials argue against a single “safe” NSAID [20]. Thus, ongoing use of NSAIDs for the acute treatment of migraine attacks requires medical supervision [40].

A recent meta-analysis and two epidemiological studies did not corroborate the elevated CV risks observed in previous clinical trials and showed no increased CV risk with celecoxib, regardless of dosage [26, 41, 42]. Real-world prescribing patterns of COX-2 inhibitors revealed that most patients received lower dosages (≤ 200 mg/d) and shorter durations (≤ 90 days) than those used in clinical trials (200–800 mg/d for up to several years), which may explain these findings [43].

Regulatory History of Nonselective NSAIDs and COX-2 Selective NSAIDs

Nonselective NSAID use may account for upward of 30% of GI bleeding cases in the US, resulting in 3200–16,500 deaths annually in the 1990s [11, 13, 44–47]. For this reason, the prescribing information for all NSAIDs contains a black box warning that “NSAIDs cause an increased risk of serious GI AEs including bleeding, ulceration, and perforation of the stomach or intestines, which can be fatal” [48]. Selective COX-2 inhibitors were originally designed to reduce GI complications associated with nsNSAIDs. Celecoxib capsule formulation (Celebrex®), a selective COX-2 inhibitor, was FDA-approved in December 1998 for the acute treatment of pain associated with osteoarthritis and rheumatoid arthritis in adults, followed shortly thereafter by rofecoxib [9, 49]. Compared with nsNSAIDs, treatment with celecoxib was associated with reduced risk of upper GI ulcerations by two- to eight-fold, which contributed to its approval [50–52].

Within 2 years of FDA approval, clinical trial results revealed an increase in CV events in participants taking COX-2 inhibitors, particularly rofecoxib. The Vioxx Gastrointestinal Outcomes Research (VIGOR) study results showed an increased incidence of thrombotic CV events; those taking supramaximal doses of rofecoxib experienced increased RRs of up to twofold for hypertension, 2.5-fold for CV AEs, and fourfold for MI [53–55]. By 2005, an external review board for the APPROVe trial terminated the study because participants receiving rofecoxib had a fourfold increase in the risk of serious thromboembolic events, mainly acute MI and ischemic stroke, compared with placebo [33, 56]. Although no deaths resulted from CV events in the APPROVe or VIGOR trials, due to the nature of the CV events and availability of other analgesics, the manufacturer voluntarily withdrew rofecoxib from the market in September 2004 [54, 55, 57].

In 2005, the FDA published that there may be a class effect of increased CV risk for all NSAIDs and required the addition of a black box warning that all prescription NSAIDs, including COX-2 inhibitors, carry risk of serious CV thrombotic events (i.e., MI and ischemic stroke) [35, 58–60]. The FDA also required all sponsors of prescription NSAIDs to conduct a comprehensive review and analysis evaluating the risk of serious CV AEs [35].

By 2005, another COX-2 inhibitor, valdecoxib, was shown to increase the risks up to fourfold for CV AEs (MI, cardiac arrest, stroke, pulmonary embolism) and Stevens–Johnson syndrome compared with placebo in individuals receiving treatment for postoperative pain after a coronary-artery bypass graft (CABG) [61–63]. In April 2005, the FDA decided the overall risk–benefit profile was unfavorable for valdecoxib and requested its withdrawal from the market [35, 41, 57, 58].

While multiple clinical trials provided unfavorable safety results for rofecoxib and valdecoxib, celecoxib has exhibited a reassuring safety profile in comparison with other COX-2 inhibitors and some nsNSAIDs [35, 52]. In 2000, the Prevention of Colorectal Sporadic Adenomatous Polyps Study (CLASS) found that celecoxib (400 mg twice daily) was as safe as diclofenac or ibuprofen in approximately 8000 individuals with osteoarthritis or rheumatoid arthritis; no differences were observed for serious CV AEs among the celecoxib, diclofenac, and ibuprofen treatment arms after 6 months of treatment [52]. In 2005, the US National Cancer Institute suspended the APC trial after participants taking supramaximal doses of celecoxib daily displayed a 2.5-fold (for 400 mg) or 3.4-fold (for 800 mg) increase in major CV event risk compared with placebo [34, 56]. The PreSAP study also reported an RR for serious CV AEs of 1.30 for participants taking celecoxib with no additional risk of death (RR: 1.06) [36]. Conversely, the placebo-controlled Alzheimer’s Disease Anti-inflammatory Prevention Trial (ADAPT) showed no increase in CV events in more than 2500 elderly persons taking 200 mg of celecoxib daily for up to 46 months [64]. In 2005, an FDA panel declared that the “benefits of celecoxib outweighed the potential risks in properly selected and informed patients” and requested that the manufacturer conduct additional long-term controlled clinical trials of celecoxib to better evaluate the potential increased risk of serious CV AEs [35].

The FDA approved diclofenac potassium powder for oral solution for the acute treatment of migraine attacks in 2009, which reinforced the idea that NSAIDs are safe and effective treatments [65, 66]. Although diclofenac is somewhat selective for COX-2 inhibition, at therapeutic IC₈₀ doses, it inhibits COX-1 by approximately 70%, effectively functioning as an nsNSAID [65, 67].

Evidence from several large clinical trials supports the elevated risk of CV events across the entire class of NSAIDs [38, 39, 68–70]. The manufacturer of celecoxib capsules initiated two studies enrolling participants with arthritis: the Prospective Randomized Evaluation of Celecoxib Integrated Safety versus Ibuprofen or Naproxen (PRECISION) trial for the FDA and the Standard Care vs. Celecoxib Outcome (SCOT) trial for the European Medicines Agency (EMA) [39, 70]. Treatment with celecoxib was noninferior to ibuprofen, naproxen, and diclofenac with regard to the frequency of CV events [39, 70]. Meta-analysis provided evidence that the risk of CV events is slightly elevated across many NSAIDs, but the PRECISION study showed that these risks do not differ significantly between celecoxib and other commonly used NSAIDs (Table 1) [38, 39]. Additionally, treatment with celecoxib was associated with significantly fewer GI AEs than ibuprofen or diclofenac and fewer renal AEs than ibuprofen [39]. These results may have contributed to a subsequent increase in the usage of COX-2 inhibitors for the treatment of pain. With the small number of published trials concerning the use of celecoxib in single migraine attacks, the preceding results relating to the chronic use of celecoxib provide valuable insights into the single-dose situation.

Table 1.

Relative risk of AEs compared with placebo or nonuse for each NSAID commonly used to treat migraine attacks or other headache disorders from selected publications

NSAID	CV events, RR, % (95% CI)	MI, RR, % (95% CI)	Ischemic stroke, RR, % (95% CI)	GI AEs, RR, % (95% CI)
Aspirin	0.89 (0.84–0.94) [85]	0.85 (0.73–0.99) [85]	0.81 (0.76–0.87) [85]	1.56 (1.38–1.78) [85]
	0.91 (0.87–0.96) [86]	0.87 (0.77–0.97) [86]	0.88 (0.80–0.96) [86]	1.57 (1.38–1.78) [86]
Celecoxib	1.10 (0.67–1.79) [64]	0.69 (0.41–1.14) [8]	0.96 (0.13–6.92) [26]	1.45 (1.17–1.81) [76]
	1.04 (0.81–1.33) [70]	0.91 (0.38–2.19) [64]	1.04 (0.87–1.23) [88]	< 2 [3]
	1.30 (0.65–2.62) [36]	1.24 (0.91–1.82) [37]	1.20 (1.00–1.44) [89]
	2.3 (0.9–5.5) [34]	1.36 (0.95–1.96) [87]	1.47 (0.52–4.20) [64]
		1.87 (0.39–8.90) [26]
Diclofenac	1.40 (1.27–1.55) [90]	1.70 (1.19–2.41) [32]	0.94 (0.59–1.49) [88]	1.89 (1.16–3.09) [32]
	1.41 (1.12–1.78) [32]	1.50 (1.06–2.04) [37]	1.18 (0.79–1.78) [32]	3.34 (2.79–3.99) [76]
	1.5 (1.4–1.7) [74]	1.7 (1.4–2.0) [74]	1.55 (1.45–1.66) [89]	4.4 (3.5–5.5) [74]
		1.80 (1.49–2.18) [7]	1.7 (1.4–2.0) [74]
Ibuprofen	1.18 (1.11–1.25) [90]	1.48 (1.00–2.26) [37]	0.88 (0.73–1.06) [88]	1.84 (1.54–2.20) [76]
	1.44 (0.89–2.33) [32]	1.56 (0.90–2.71) [7] 2.22	0.97 (0.42–2.24) [32]	< 2 [3]
		(1.10–4.48) [32]	1.45 (1.31–1.61) [89]	3.97 (2.22–7.10) [22, 32]
Indomethacin1	1.30 (1.19–1.41) [90]	1.22 (0.82–1.83) [87]	1.20 (0.85–1.69) [88]	4.14 (2.91–5.90) [76]
		1.36 (0.67–2.79) [7]	1.24 (1.00–1.54) [89]
Ketorolac2	–	2.02 (1.00–4.09) [87]	1.90 (1.39–2.60) [89]	> 5 [3]
				11.5 (5.56–23.78) [76]
Naproxen3	0.93 (0.69–1.27) [32]	0.84 (0.52–1.35) [32]	0.94 (0.80–1.11) [88]	4.10 (3.22–5.23) [76]
	1.09 (1.02–1.16) [90]	1.12 (0.74–1.69) [7]	0.97 (0.59–1.60) [32]	4.22 (2.71–6.56) [22, 32]
	1.63 (1.04–2.55) [64]	1.49 (0.69–3.22) [64]	1.46 (1.22–1.74) [89]
Rofecoxib4	1.45 (1.33–1.59) [90]	1.58 (1.07–2.17) [37]	1.28 (1.06–1.53) [88]	2.32 (1.89–2.86) [76]
	2.32 (0.89–6.74) [33]	1.15 (0.74–1.79) [8]	2.13 (0.81–5.60) [64]
		6.50 (0.70–60.33) [7]

AEs adverse events, CI confidence interval, CV cardiovascular, GI gastrointestinal, MI myocardial infarction, RR relative risk

¹Indicated for headache, not migraine

²Available for intravenous or intramuscular administration

³Indicated for migraine in combination with sumatriptan

⁴Withdrawn from the market

Notably, the increased risks of CV and renal AEs are associated with longer-term usage of NSAIDs; thus, low-dose and/or acute administration of NSAIDs should confer much lower risk of AEs while still effectively reducing pain. Recent results from two pivotal trials support that celecoxib oral solution demonstrates significant efficacy in the acute treatment of migraine attacks [15, 16]. Participants reported significant improvements in freedom from the most bothersome symptom and freedom from headache pain at 2 h postdose compared with placebo (see Fig. 2) [15, 16]. These two clinical trials reported no serious treatment-related or unexpected AEs and few mild, manageable AEs—only dysgeusia and nausea increased in frequency [15, 16]. In light of these safety and efficacy results, in 2020 the FDA approved celecoxib oral solution for the acute treatment of migraine in adults with or without aura.

Fig. 2.

A greater proportion of participants reported headache pain freedom after administration of 120 mg of celecoxib oral solution than with placebo (by postdose timepoint using the last observation carried forward).

Reproduced from Lipton et al. [15] under CC BY-NC 4.0 license terms [https://creativecommons.org/licenses/by-nc-nd/4.0/]

Nonselective NSAIDs and COX-2 Selective NSAIDs in Migraine Treatment Guidelines

The American Headache Society (AHS) published a consensus statement for the acute treatment of migraine attacks in 2021 to include an array of first-line treatments for migraine, including acute therapies such as triptans, ergotamine derivatives, small-molecule calcitonin-gene related peptide (CGRP) receptor antagonists (“gepants”), combination analgesics (acetaminophen/aspirin/caffeine), NSAIDs (aspirin, diclofenac, naproxen, and ibuprofen), and neuromodulation devices [2]. Recently approved first-line therapies with proven efficacy include lasmiditan, rimegepant, ubrogepant, remote electrical neuromodulation, and celecoxib oral solution [2]. The AHS advised that patients using NSAIDs for more than 15 days per month may develop medication-overuse headache and should be candidates for preventive migraine treatments [2]. The latest treatment guidelines for migraine from the European Headache Federation and the European Academy of Neurology list only NSAIDs as first-line therapies; they recommend triptans as second-line treatments and ditans and gepants as third-line treatments [71]. NSAIDs continue to have strong support internationally for the acute treatment of migraine attacks; thus, clinicians and patients are in need of safe and effective NSAID formulations.

The Role of Traditional NSAIDs and New Formulations in Migraine Treatment

Given the COX-2 hypothesis that increased dose, duration, and COX-2 selectivity of an NSAID increases its risk of CV and other AEs, the development of NSAID reformulations allow for lower and safer effective doses [3, 26, 48]. The decision of which NSAID to reformulate is critical as higher relative COX-2 selectivity (see Fig. 3) is associated with increased CV and bleeding AE risks. For example, valdecoxib is more selective for COX-2 than celecoxib in vitro, and valdecoxib is associated with higher CV risk. Thus, the relatively lower COX-2 selectivity of celecoxib, compared with valdecoxib, positions it favorably from a pharmacologic perspective to achieve efficacy outcomes with less CV risk [26]. As the FDA recommends using the shortest duration and lowest effective dose of NSAIDs [35, 48], celecoxib oral solution is particularly suited for diseases that cause acute pain and require intermittent treatment, such as migraine, allowing for effective analgesia with lower overall drug exposure and AE risks.

Fig. 3.

Relative COX-2 selectivity of NSAIDS commonly used in the treatment of migraine attacks or headache disorders with rofecoxib included for reference. COX Clyclooxygenase, IC₈₀ concentration that inhibits 80% of the maximal response

It is important to evaluate the relative safety of NSAIDs in individual patients, whether used on- or off-label, for the acute treatment of migraine attacks (Table 1). Low-dose diclofenac potassium powder for oral solution is available via prescription for the acute treatment of migraine attacks with efficacy as early as 30 min postdose [72, 73]. Common AEs reported with diclofenac treatment include vomiting and upper abdominal pain; further, diclofenac poses an increased CV health risk compared with no usage of NSAIDs or use of paracetamol or ibuprofen [74]. Nasal ketorolac tromethamine (Sprix®) is the only nasal spray NSAID currently marketed and FDA-approved for moderate to severe pain [75]. It is not FDA-approved for the acute treatment of migraine attacks, but it is often prescribed for these attacks. Due to its high COX-1 selectivity, ketorolac has the highest associated risk of serious GI events with more than a fivefold increase compared with no usage of NSAIDs [13, 76].

Ibuprofen and naproxen are common non-prescription medications used for the acute treatment of migraine attacks, although they are not FDA-approved for the indication. They are not exempt from the elevated AE risks associated with all NSAIDs, particularly with longer-term use. Several studies (ADAPT, PRECISION, SCOT, and a recent meta-analysis) provided evidence that celecoxib confers risks of CV events similar to naproxen and ibuprofen (see Table 1) [37, 39, 64, 70]. Both naproxen and ibuprofen are associated with significantly higher risk of GI events than celecoxib, and ibuprofen carries significantly more risk for renal events [39].

Celecoxib oral solution is the only selective COX-2 inhibitor currently FDA-approved for the treatment of acute attacks in adults with migraine with or without aura [35, 48]. Its self-micro-emulsifying drug delivery system (SMEDDS) allows for rapid absorption and increased bioavailability, compared with the capsule formulation [77]. Many of the AEs associated with celecoxib and other NSAIDs develop over time from accumulated drug exposure due to relatively high or prolonged dosing. When administered at low, intermittent doses, the acute use of celecoxib oral solution appears to have fewer safety concerns than other NSAIDs [15, 16, 26, 48]. In two identical clinical trials of celecoxib oral solution for the acute treatment of migraine, there were no serious treatment-related AEs reported [15, 16]. These trials provide evidence that celecoxib oral solution is safe and effective within 2 h post-dose for the treatment of migraine attacks [15, 16].

Considerations in the Use of COX-2 Selective NSAIDs, including celecoxib oral solution, in Migraine

Certain patients require special considerations prior to the initiation of celecoxib oral solution and other NSAIDs. Those with a history of MI have an increased risk of reinfarction and death with NSAID treatment [78, 79]. However, celecoxib appears to be associated with some of the lowest RR for CV events or bleeding in this clinical scenario—lower than diclofenac, ibuprofen, and naproxen (Table 1) [10].

Many individuals with migraine take daily aspirin, which is cardioprotective, but there is no consistent evidence that concurrent use of aspirin with NSAIDs mitigates the increased risk of serious CV thrombotic events [48]. The AHS warns about GI AEs (bleeding, ulceration, perforation) associated with NSAIDs, particularly in the elderly and those with a history of GI bleeding or peptic ulcer disease [2]. Notably, the decreased frequency of GI AEs is likely mechanistic and class-related for selective COX-2 inhibitors [80].

All NSAIDs, except aspirin, are contraindicated in those with a history of CABG [35]. This is due to an extrapolation of results from a study of patients undergoing CABG who received valdecoxib or parecoxib showed a fourfold increase in CV events [35, 62].

When used too frequently, certain acute treatments for migraine, such as triptans, ergots, and butalbital-containing medications, are associated with an increased risk of progression from episodic to chronic migraine [81, 82]. NSAIDs do not carry this risk in those with fewer than ten monthly migraine days, a group that constitutes more than 80% of persons with migraine [83].

Individuals with gastroparesis, nausea, difficulty swallowing, or a history of bariatric surgery may not be able to quickly administer or absorb oral medications [84]. Rather than tablets, alternative formulations that are absorbed more rapidly with or without food or drink are generally more effective; these include nasal or intramuscular ketorolac, diclofenac potassium powder for oral solution, and celecoxib oral solution [2, 15, 66, 75].

Conclusions

NSAIDs remain an integral part of the acute treatment paradigm for migraine. All NSAIDs, including selective COX-2 inhibitors, confer some elevated risk of CV and GI events, although the level of risk differs between individual NSAIDs. With their high RR of CV events, rofecoxib and valdecoxib were withdrawn from market in 2005, and all NSAIDs, including celecoxib, received a black box warning about increased risk of CV events.

The degree of AE risk appears to increase with dose, duration, and COX-2 selectivity. CV and GI events occur more frequently with long-term usage or high daily dosages; lower exposure to NSAIDs is preferable. Disease states associated with acute pain, like migraine, are more amenable to employing NSAIDs with lower risk of the AEs associated with chronic or high-dose NSAID usage. Thus, the benefits of NSAID treatment may outweigh the risks for many patients. Individualizing treatment according to a patient’s specific needs and risk factors optimizes safe and effective NSAID analgesic use.

New technologies and formulations may help mitigate risks associated with NSAIDs. Celecoxib oral solution is the only selective COX-2 inhibitor designed as a low-dose oral solution for the acute treatment of migraine attacks. Clinical trials demonstrated that it is safe and effective and associated with mild AEs and minimal increased risk of CV events, similar to other NSAIDs. New formulations of NSAIDs may provide alternative low-dose, acute analgesia with fewer associated CV and GI events than previous NSAID formulations.