Angioedema without urticaria after recent initiation of celecoxib

Abstract

Angioedema is potentially life-threating swelling of integument and mucosa that has multiple potential aetiologies with varying mechanisms. Drug-induced angioedema is often easily correlated with the offending agent and can be prevented with discontinuation of the medication. Many medications have now been implicated in drug-induced angioedema but the two most common are ACE inhibitors and non-steroidal anti-inflammatory drugs (NSAIDs). This case highlights severe angioedema secondary to celecoxib and reviews varying aetiologies of angioedema and NSAID hypersensitivity reactions.

Background

Angioedema is defined as localised swelling involving the subcutaneous, submucosal and/or deep dermal tissues that can involve any area of the body. It has the potential to be life threatening with upper airway mucosal involvement that can lead to airway obstruction.^{1 2} Many medications have been identified as causative agents of angioedema. The most often implicated pharmacological categories include ACE inhibitors (ACE-i) and non-steroidal anti-inflammatory drugs (NSAIDs).³ NSAIDs are one of the most widely available medications worldwide and are a mainstay of treatment for pain and inflammation. NSAIDs provide these benefits by inhibiting cyclooxygenase (COX) 1 and 2 and reducing proinflammatory cytokines, mediated by the conversion of arachidonic acid to prostaglandin H2. Selective COX-2 inhibitors were developed to reduce side effects and intolerance of COX-1 inhibition.⁴ Hypersensitivity reactions to NSAIDs, including angioedema, are known, but the pathophysiology is not well understood and very few cases describe hypersensitivity reactions with COX-2 inhibitor use. The prevalence of NSAID hypersensitivity, including angioedema and urticaria, in the general population is 0.1%–0.3%.⁵ This case highlights severe angioedema without urticaria after initiation of celecoxib, a selective COX-2 inhibitor.

Case presentation

An African American woman in her 70s with a medical history of type 2 diabetes, hypertension, and obesity presented with tongue swelling, difficulty swallowing, and speaking. These symptoms started the morning of presentation. Due to concern for impending airway compromise, she underwent fiberoptic intubation in the emergency department. She had no symptoms of urticaria or anaphylaxis, and no history of angioedema. She was recently started on celecoxib for right shoulder bursitis that was started the day prior to admission. Lisinopril was also on the patient’s medication list, but medication refill history revealed that it was last dispensed 7 months prior to admission for a 90-day supply. She had also been on lisinopril for nearly 5 years. Aspirin was also a long-term medication that was continued on admission. Medication reconciliation confirmed recent use of celecoxib without recent use of ACE-i. Both celecoxib and lisinopril were discontinued on admission. The patient was admitted to the intensive care unit (ICU) and started on solumedrol, diphenhydramine, and famotidine. Over the course of a week, the patient was slowly weaned to lower dose steroids and extubated. She transferred out of the ICU 2 days after extubation with improvement in facial and oral swelling. On the third day post extubation, mild stridor was noted prompting otolaryngology consultation. Bedside laryngoscopy showed poor visualisation of vocal cords, swelling in the right inferior oropharynx/hypopharynx, oedema in the right arytenoid and aryepiglottic fold, and a right neck mass suspected to be of thyroid origin. She was subsequently transferred to the ICU and steroids were re-escalated. Thyroid ultrasound and CT neck showed multinodular goitre with mild narrowing of the trachea. On chart review, this multinodular goitre had been present for at least 8 years prior to admission, with most recent imaging of the thyroid 2 years prior. No imaging on admission included the neck. Acute airway compromise was caused by angioedema as evidenced by laryngoscopy. Oedema at the level of the right arytenoid and aryepiglottic fold would be superior to the level of the thyroid. The patient likely had chronic encroachment of the airway from her longstanding goitre resulting in increased susceptibility to airway compromise from even mild oedema explaining the persistence of symptoms. Repeat laryngoscopy showed improvement in swelling and the patient was ultimately able to be discharged with a steroid taper and plan for outpatient follow-up for her goitre.

Throughout her admission, she received no additional NSAIDs and no new NSAIDs were prescribed on discharge. She did not receive any sulfa-containing agents during admission and had been prescribed courses of sulfamethoxazole-trimethoprim within 5 years of admission without reported side effects. Therefore, sulfa allergy was not further investigated. Other potential causative agents on her medication list that rarely report angioedema included tramadol and tizanidine. These medications were longstanding, with no recent change in dosage or administration prior to admission. The temporal association between celecoxib and onset of angioedema increased the likelihood of celecoxib being the causative agent compared with other medications. Lisinopril and celecoxib were listed as allergies in the patient’s chart and discontinued on discharge.

The patient was readmitted to the ICU 1 week after discharge with worsening hoarseness, odynophagia, and shortness of breath without any noticeable lip or mucosal swelling. She was started on positive pressure ventilation and intravenous dexamethasone without improvement. Otolaryngology re-evaluated the patient and symptoms were thought to be due primarily to thyromegaly. The patient underwent successful right hemithyroidectomy without complication and was discharged home with a steroid taper. It is suspected that the patient had a functionally reduced airway calibre due to the goitre. In the absence of haemorrhage, it is unlikely that the goitre caused airway compromise. Additionally, her readmission was unlikely related to angioedema in the absence of mucosal swelling, lack of improvement with intravenous steroids, and cessation of the suspected offending agent. Her readmission may be attributed to tracheomalacia, which was later formally diagnosed as an outpatient. Tracheomalacia may have developed due to recent intubation and compression from her chronic goitre. Therefore, hemithyroidectomy would improve dyspnoea by relieving compression but not entirely resolve tracheomalacia.

Outcome and follow-up

Two months after discharge, patient was able to return to daily activities. However, she reported continued hoarseness and mild dyspnoea. Outpatient bronchoscopy was performed which ruled out tracheal stenosis but showed tracheomalacia, erythematous vocal cords, and oedematous larynx. She continues to follow up with her primary care physician, pulmonologist, and otolaryngologist.

Discussion

There are multiple types of NSAID hypersensitivity reactions as described by Kowalski et al⁶ that are widely recognised. The five types are: NSAIDs-exacerbated respiratory disease (NERD), NSAIDs-exacerbated cutaneous disease (NECD), NSAIDs-induced urticaria/angioedema (NIUA), single-NSAID-induced urticaria/angioedema or anaphylaxis (SNIUAA), and single-NSAID-induced delayed hypersensitivity reactions (SNIRD). NERD, NECD, and NIUA are all non-immunological hypersensitivity reactions, whereas SNIUAA and SNIRD are immunologically mediated. NECD describes patients with chronic urticaria/angioedema, that is worsened by NSAID use. SNIRD describes skin and/or organ-specific hypersensitivity reactions 24–48 hours after exposure. Our patient had acute onset angioedema and no history of chronic urticaria/angioedema; therefore, we will focus on NIUA and SNIUAA.

Angioedema has multiple aetiologies, including hereditary, allergic, non-allergic, and idiopathic; however, pathophysiology is not consistent among the varying aetiologies.^{2 7} NSAID-induced angioedema may be due to an acute allergic, IgE-mediated, reaction seen in SNIUAA or a ‘pseudoallergic’ reaction in NIUA.^{2 8 9} Neither are seen in patients with chronic urticaria/angioedema. In SNIUAA, other allergic findings such as urticaria and/or anaphylaxis would be expected in addition to angioedema. It has been suggested that the most common presentation of SNIUAA is anaphylaxis.¹⁰ The pathogenesis of angioedema without urticaria is poorly understood; according to one 10-year retrospective study, most cases are idiopathic, and no causative agent was found.¹¹ Currently, there is no validated, widely available method to test IgE, specific to the causative agent, in vitro.2 5 6 12 Skin testing may be beneficial for diagnosing SNIUAA but is of little value if non-immunological hypersensitivity reactions are suspected.⁶

In NIUA, the leading hypothesis is the ‘pseudoallergic’ reaction, which states that NSAID inhibition of COX enzymes results in overproduction of cysteinyl leukotrienes (LT) by shunting arachidonic acid down the lipoxygenase pathway.1 2 8 9 These LT have been documented to play a role in vasoconstriction and increase vascular permeability.^{1 13–15} Additionally, prostaglandin E₂, which is responsible for inhibition of mast cell degradation and LT production is decreased due to COX inhibition.² However, LT overproduction should theoretically be less significant with the use of selective COX-2 inhibitors, such as celecoxib, given that the COX-1 pathway is still available for arachidonic acid metabolism.^{1 4} NIUA is almost exclusively seen with COX-1 inhibitors, and it has been suggested that selective COX-2 inhibitors are less likely to induce angioedema.^{1 4 16} In addition to our case, there have been a few reported cases of selective COX-2 inhibitors causing NSAID hypersensitivity reactions.^17–19 Concomitant use of aspirin and celecoxib may have resulted in excessive inhibition of COX enzymes and increased LT production resulting in NIUA. It has also been documented that NIUA is associated with cross-reactivity among aspirin and other NSAIDs.^{1 2 6} Aspirin has been a long-term medication for our patient. It was continued during admission and after discharge without cross-reactivity, arguing against NIUA. Therefore, her pathophysiology cannot be fully explained by the ‘pseudoallergic’ pathway.

ACE-i-induced angioedema is well known and documented. It has been demonstrated that ACE-i-induced angioedema is most prevalent closer to the time of initiation of therapy, but a continuous risk of angioedema is present years after stabilisation of therapy.^20–22 There are also rare case reports that describe the development of angioedema in patients with long-term ACE-i use with recent NSAID initiation.^{3 23} Concurrent use of ACE-i and NSAIDs may potentially induce or worsen angioedema. In many cases of angioedema, patients are being actively treated with a combination of ACE-i and NSAIDs.^{3 23 24} The cases that described these reactions used diclofenac and naproxen and did not include COX-2 inhibitors. Our patient’s poor adherence to her ACE-i, based on medication reconciliation and dispense history, makes the possibility of angioedema solely from ACE-i or from drug–drug interaction less likely.

Aetiology is vital for preventing future instances of life-threatening angioedema, and further research is needed to determine additional mechanisms and pathophysiology of NSAID hypersensitivity and angioedema. However, in the acute setting, focus should be maintained on treatment and stabilisation.

The initial evaluation of angioedema should investigate the possibility of rapidly developing airway compromise, followed by immediate intubation if deemed necessary. Further evaluation of breathing, circulation, disability, and exposure is essential to prevent further decompensation.^{25 26} The suspected culprit of angioedema should be discontinued promptly and determination of underlying type of angioedema should be conducted simultaneously during initial treatments. One study broke down the aetiology of angioedema into idiopathic (38%), C1 inhibitor deficiency (25%), related to specific factor (food, drug, insect bite, environmental allergen, or physical stimulus (16%)), ACE-i induced (11%) and autoimmune and generalised oedema (7% and 3%, respectively).¹¹ Mainstay of treatment revolves around epinephrine, intravenous antihistamines, and glucocorticoids, although no controlled studies have demonstrated their efficacy.^26–28 In cases of angioedema without a known cause, bradykinin-mediated treatment can be considered if patient’s do not improve with epinephrine, steroids, and antihistamines.²⁶ In angioedema without urticaria, treatment with a histamine H1 receptor blocker has been more responsive.²⁷ In acute management, there are no absolute contraindications for epinephrine administration in the setting of anaphylaxis. Close outpatient follow-up with primary care physician and allergist after resolution of acute symptoms is recommended for secondary prevention and determination of aetiology if unclear.

Learning points.

• Angioedema has many potential aetiologies with varying pathophysiology.

• ACE inhibitors and non-steroidal anti-inflammatory drugs (NSAIDs) are two of the most common culprit medications but angioedema secondary to cyclooxygenase 2 inhibitors is rare.

• There are a wide range of well-described NSAID hypersensitivity reactions that span from exacerbation of existing cutaneous or airway disease to ‘pseudoallergic’ (NSAIDs-induced urticaria/angioedema) and true allergic reactions (single-NSAID-induced urticaria/angioedema or anaphylaxis).

• For all patients that present with angioedema, a thorough medication reconciliation should be performed to identify potential causative agents. Initial management should focus on prevention of airway compromise and administration of medications to reduce mucosal swelling.

Ethics statements

Patient consent for publication

Consent obtained directly from patient(s)