Edited by: Pascal Demoly
Response to a selective COX-2 inhibitor in patients with urticaria/angioedema induced by nonsteroidal anti-inflammatory drugs
Version of Record online: 11 AUG 2011
© 2011 John Wiley & Sons A/S
Volume 66, Issue 11, pages 1428–1433, November 2011
How to Cite
Doña, I., Blanca-López, N., Jagemann, L. R., Torres, M. J., Rondón, C., Campo, P., Gómez, A. I., Fernández, J., Laguna, J. J., Rosado, A., Blanca, M. and Canto, G. (2011), Response to a selective COX-2 inhibitor in patients with urticaria/angioedema induced by nonsteroidal anti-inflammatory drugs. Allergy, 66: 1428–1433. doi: 10.1111/j.1398-9995.2011.02684.x
- Issue online: 7 OCT 2011
- Version of Record online: 11 AUG 2011
- Accepted for publication 13 July 2011
- nonsteroidal anti-inflammatory drugs;
To cite this article: Doña I, Blanca-López N, Jagemann LR, Torres MJ, Rondón C, Campo P, Gómez AI, Fernández J, Laguna JJ, Rosado A, Blanca M, Canto G. Response to a selective COX-2 inhibitor in patients with urticaria/angioedema induced by nonsteroidal anti-inflammatory drugs. Allergy 2011; 66: 1428–1433.
Background: In subjects with hypersensitivity reactions with cross-intolerance to nonsteroidal anti-inflammatory drugs (NSAIDs), tolerance to selective COX-2 inhibitors has not been evaluated in large series of well-phenotyped cases.
Methods: We evaluated 252 patients with urticaria and/or angioedema caused by hypersensitivity owing to cross-intolerance to NSAIDs. In addition to the clinical history, diagnosis was confirmed by provocation to an alternative NSAID. Two groups were considered: (A) patients with cross-intolerance to NSAIDs and intolerance to paracetamol and (B) patients with cross-intolerance to NSAIDs and good tolerance to paracetamol. Etoricoxib was administered to Group A patients and to a representative sample of Group B patients. In the event of a positive response, serum tryptase levels were determined and skin biopsy was performed in five patients in each group.
Results: Ibuprofen was the most commonly implicated drug, followed by acetylsalicylic acid (ASA). Urticaria was the most common manifestation, followed by angioedema. Most of the patients developed symptoms within 1 h. Twenty-five percent in Group A (n = 47) and 6% in Group B (n = 50) were intolerant to etoricoxib. Skin biopsy showed mast cell activation with the release of tryptase to the extracellular space but without the increase in serum tryptase levels.
Conclusion: Selective COX-2 inhibitors may be unsafe in subjects with urticaria and/or angioedema caused by hypersensitivity reactions to NSAIDs with cross-intolerance if they are intolerant to paracetamol. A quarter of patients who were intolerant to this drug were also intolerant to etoricoxib. In subjects with hypersensitivity to NSAIDs and intolerance to paracetamol, selective COX-2 inhibitors should be administered as a controlled, incremental dose provocation test to assess tolerance.
nonsteroidal anti-inflammatory drugs
Nonsteroidal anti-inflammatory drugs (NSAIDs) are one of the most commonly prescribed compounds (1). Among the side-effects reported, those belonging to the type B classifications of adverse drug reactions are of wide concern (2). Although different classifications of adverse drug reactions are of concern, the term ‘hypersensitivity’ has been proposed for those cases where histamine and other inflammatory mediators are released, independent of the mechanism (3). This occurs with NSAIDs in 0.6–2.5% of the population (4), although these figures increase considerably in patients with bronchial asthma, particularly those with concomitant nasal polyposis, reaching values of 30–40% (5). From 21% to 30% of patients with chronic urticaria also develop symptoms after the intake of NSAIDs (6).
One of the mechanisms proposed involves the inhibition of cyclooxygenase (COX) by blocking prostaglandin E2 formation and deviating toward the 5-lipooxygenase pathway, with an excess production of cysteinyl-leukotrienes LTC4, LTD4, and LTE4 (7). Two isoforms of COX exist: COX-1, which is expressed constitutively and is involved in the mechanisms of homeostasis, and COX-2, which is inducible and mediates inflammation (5). More recently, three splice variants of COX-1 have been described and termed as COX-3, COX-1b, and COX-1v. These isoforms seem to be selectively inhibited by paracetamol (8). The therapeutic effects of NSAIDs are primarily related to their ability to inhibit COX-2, whereas some of their most frequent adverse effects may be caused by COX-1 inhibition (5). The pathogenesis of hypersensitivity reactions owing to cross-intolerance seems to be related to COX-1 inhibition (5).
Subjects with cross-intolerance are intolerant to acetylsalicylic acid (ASA) and other strong COX-1 inhibitors and, in some cases, to weak inhibitors such as meloxicam (9), nabumetone (9), and paracetamol (10). Reports have shown that up to 20% of subjects with cross-intolerance may also be intolerant to paracetamol (10). The need to choose a safe, alternative NSAID for these patients is a frequent problem in clinical practice. Recently, much attention has been given to the new class of selective COX-2 inhibitors. Their mechanism of action makes them good candidates as alternative NSAIDs for patients with a history of cross-intolerance in NSAID-induced reactions (11, 12).
Cross-intolerance may affect the skin and/or the respiratory airways (2). Although most studies have been carried out on the respiratory airways (5, 7, 10), the skin is the most common organ affected (13). Two types of entities exist: reaggravation of urticaria (14) or acute urticaria in the absence of any history indicative of chronic urticaria (15). Recent evidence suggests that acute urticaria induced by NSAIDs is the most common manifestation within the group of hypersensitivity reactions (13).
Selective COX-2 inhibitors have been found to be well tolerated in patients with ASA-exacerbated respiratory disease (16–18), and only a few cases of asthma attack precipitated by selective COX-2 inhibitors have been reported (19–21). On the other hand, evidence also indicates that COX-2 inhibitors may cause urticaria in patients with cross-intolerance to NSAIDs (22–24). However, these studies may have limitations because of the small sample size and/or their design.
The purpose of this study was to assess the tolerance to etoricoxib, a widely used selective COX-2 inhibitor, in a large group of patients with a confirmed history of urticaria and/or angioedema caused by hypersensitivity owing to cross-intolerance in the absence of past or recent history of chronic urticaria.
We evaluated patients referred to the allergy clinics of five hospitals in Spain (Carlos Haya Hospital, Malaga; Infanta Leonor Hospital, Madrid; Elche Hospital, Elche, Alicante; Cruz Roja General Hospital, Madrid; and Alcorcon Foundation-Hospital, Madrid) with skin symptoms attributed to the intake of different NSAIDs. The patients were evaluated between 2006 and 2009, and all were aged over 14 years.
Confirmation of cross-intolerance
Subjects had to have experienced at least two episodes of cutaneous symptoms after the intake of two or more different NSAIDs to be considered as potential candidates. It has been reported that in certain instances, subjects with two episodes may not develop a positive reaction after subsequent exposure (25). Thus in all cases, cross-intolerance was confirmed by single-blind, controlled, challenged administration of one drug.
Following are the exclusion criteria: suspicion or evidence of a delayed-type hypersensitivity reaction to NSAID compounds, such as fixed drug eruption, erythema multiforme, Stevens–Johnson syndrome, acute generalized exanthematous pustulosis, or drug hypersensitivity syndrome; patients with chronic urticaria or renal, cardiac or liver diseases; subjects with a history of ASA-induced asthma and/or nasal polyposis; subjects with a history of anaphylaxis/anaphylactoid reactions after the administration of NSAID compounds; patients taking beta-blockers or ACE inhibitors or with contraindications for epinephrine administration; pregnant or breast-feeding women; and subjects with psychosomatic disorders.
Drug provocation test to confirm NSAID cross-intolerance
A drug provocation test was performed in a single-blind procedure. Either a placebo (lactose) or an NSAID capsule was administered orally, in divided doses at intervals of 90 min up to a cumulative single therapeutic dose, depending on the drug (see Table 1). The same procedure as for the drug was employed for the placebo. All substances were given in opaque capsules prepared by the hospital pharmacy service. The patients received increasing doses of the drug (see Table 1) and stopped receiving as soon as any symptoms appeared. If no symptoms were observed, the maximum therapeutic dose of the specific drug was given. After an observation period of 24 h, a therapeutic course was given over 2 days. Each challenge was carried out separately, with a minimum of 7 days between the administrations of different drugs. Drug administration was made in an ascending order according to the capacity of the drug for COX inhibition, starting with paracetamol as described (see Fig. 1). If a positive response occurred, no further drugs were administered. If the response was negative, we then challenged with the next drug until there was a positive response. If the patient reported one or several episodes to the same drug, the procedure was continued with the next noninvolved drug to confirm cross-intolerance.
|Drug (total dose, mg)||Doses administered at 90-min intervals (mg)|
|Etoricoxib (60)||15, 15, 30|
|Paracetamol (1000)||250, 250, 500|
|Piroxicam (20)||5, 5, 10|
|Ibuprofen (600)||150, 150, 300|
|Diclofenac (50)||12.5, 12.5, 25|
|ASA (1000)||50, 100, 250, 500|
The blood pressure, heart rate, and peak expiratory flow rate were measured at baseline and prior to each dose escalation during the drug provocation test. After the last dose, the patients were kept under observation for 3 h; after discharge, they could return or telephone if they developed reactions after a longer interval from the last challenge dose. The procedure was stopped if cutaneous or respiratory symptoms developed or alterations were noted in vital signs (rhythm alterations, decrease in peak expiratory flow rate, or hypotension). At the time of the challenge, none of the patients had urticaria and/or angioedema. Drugs that could interfere with the result of the challenge, such as H1-receptor antagonists and leukotriene receptor antagonists, were stopped at least 1 week before the challenge. All participants provided written informed consent. The study protocol was approved by the ethics committee of each participating hospital.
When a positive event occurred during drug provocation, total serum tryptase levels were quantified in patients at the time of symptoms onset and again 2 and 24 h after the onset of symptoms. Tryptase levels were determined by immunoassay (UniCap; Phadia, Uppsala, Sweden). Levels above 15 μg/l were considered as increased. A skin biopsy was also obtained at the time of the reaction. The biopsies were processed using hematoxylin–eosin and immunohistochemical staining for tryptase (26).
Atopy status assessment
The atopy status of both paracetamol-tolerant and paracetamol-intolerant patients was assessed from a clinical questionnaire that included questions on nasal and bronchial symptoms such as sneezing, itching, watery nose, nasal blockade, difficulty breathing, cough and wheezing, as well as symptomatology attributed to food, hymenoptera stings, or exposure to latex. A skin prick test was performed with a battery of common inhalant allergens that included pollens, house dust mites, molds, animal danders, and foods such as milk, egg, cod, shrimp, peach, apple, kiwi, peanut, walnut, wheat, and soybean (ALK, Madrid, Spain); 10 mg/ml of histamine hydrochloride and phenolated glycerol saline were used as positive and negative controls, respectively. A positive skin prick test response was defined as a wheal diameter of 3 mm or larger to at least one of these allergens. Total IgE serum was determined by ImmunoCAP (Phadia, Upssala, Sweden), in accordance with the manufacturer’s instructions. A cutoff value of 130 IU/ml was used to consider values as low or high. Although subjects were considered atopic if the skin test was positive or the total IgE was higher than 130 IU/ml, we also compared the clinical symptoms reported in both groups to determine any possible differences in atopic status.
Data analysis was performed using the statistical package spss 15.0 for Windows (SPSS Inc., Chicago, IL, USA). Chi-square analysis was used to test the differences for nominal variables, and t-tests were used for interval variables with two groups. A P-value of <0.05 was considered statistically significant.
The study included 252 patients with a confirmed skin reaction after taking NSAIDs. All the reactions were confirmed by challenge to at least one alternative NSAID. Of the 252 patients, 151 (60%) were women and 101 (40%) men, and the mean age was 39.71 ± 15.54 years (14–80 years). No age differences were found.
Patients reported an average of 2.56 ± 2.03 (1–15) episodes prior to challenge. The drugs involved in the putative reported reactions are shown in Table 2. Propionic acid derivatives, especially ibuprofen, were the drugs most frequently reported (39%), followed by ASA (23.8%) and pyrazolones (16.95%). According to the clinical description, 134 (53.2%) patients had experienced urticaria, 59 (23.4%) angioedema, 58 (23%) urticaria plus angioedema, and 1 (0.4%) nonurticarial rash.
|Culprit drug from history||n (%)||Drug used in challenge||Positive challenge (intolerant)||Negative challenge (tolerant)|
|Propionic acids||228 (39.04)||Paracetamol||47||205|
|Arylacetic acids||47 (8.04)||ASA||21||0|
|Indolacetic acids||3 (0.51)|
The time interval between the drug intake and the reaction reported by the patient was <1 h in 204 (80.95%) patients and between 1 and 6 h in the remaining. Patients with longer time interval between the drug intake and the appearance of symptoms were not included.
In 223 patients (88.5%), although the clinical questionnaire showed that patients had at least two episodes with two or more different NSAIDs, the diagnosis was confirmed by challenge. In the remaining 29 patients (11.5%), the diagnosis was established by challenge.
Thus, a total of 511 challenges were performed in 252 patients, with a mean of 1.40 ± 0.69 (1–5) challenges per patient. The drugs used are shown in Table 2. Forty-seven patients were intolerant to paracetamol (Group A) and 205 were tolerant (Group B). In all instances, tolerance to paracetamol was established by challenge. Considering the presence of skin prick test positivity to allergens, the difference in the frequency of response to at least one allergen was not significant (75% atopic patients in Group A and 62.8% in Group B). The same occurred concerning total IgE value (33.8%vs 30.2%). Considering the clinical entities, the most frequently reported was rhinitis, followed by asthma and food allergy. These were all similar to those reported previously, with no differences between them (13).
Tolerance to etoricoxib was assessed by challenge in all cases in Group A and in a representative sample of 50 subjects randomly selected from Group B who were matched to Group A in age, sex, clinical entity, and NSAIDs involved. In Group A, 12 (25.53%) patients showed positive response to etoricoxib, and in Group B, 3 (6%) patients showed positive response to this drug (P < 0.03). In all cases, the symptoms caused by etoricoxib consisted of pruritus and wheals localized on different parts of the body. No patient had any respiratory symptoms. The symptoms were mild, disappearing within 1–2 h of taking oral antihistamine.
In the group of 12 patients intolerant to both paracetamol and etoricoxib, nine were atopic, and among the three patients tolerant to paracetamol and intolerant to etoricoxib, two were atopic. No statistical comparison could be made. Total tryptase levels in serum were determined and a skin biopsy was also taken in ten cases (five from each group). Sequential tryptase levels showed no significant variation during an evaluation period of 24 h after the episode. The skin biopsy showed the release of tryptase to the extracellular space. A biopsy taken at an asymptomatic time from the area where the lesions appeared showed tryptase within cytoplasmic granules of mast cells (Fig. 2). Similar observations were noted in the Group A and the Group B patients.
NSAIDs are widely used for the treatment of different diseases, and when a patient develops a hypersensitivity reaction to one of these drugs, it is relevant to assess whether other NSAIDs are tolerated (27). Excluding those cases where a specific immunological mechanism occurs, in the remaining cases with the involvement of the skin and/or respiratory airways, it is necessary to assess tolerance to other different NSAIDs.
Clinical evidence indicates that subjects who develop reactions to ASA are also usually intolerant to potent COX inhibitors and very occasionally to weak COX inhibitors, although the possibility exists of being a selective responder (28). Among selective NSAID responders, a specific IgE mechanism has been implicated (28), although it has only been well established with pyrazolones (29). Nonetheless, because of the lack of validated tests, in selective responders, a controlled administration of an alternative NSAID is required in most instances.
Information on cross-intolerance has come from studies in patients with asthma and/or nasal polyposis and ASA intolerance (5, 30). Although paracetamol has been recommended as a relatively safe drug, a certain number of patients will be intolerant, especially when doses of 1 g or higher are given. This proportion varies from 2% to 34% (10). In addition to affecting the respiratory airway, these drugs can also induce other types of reactions, with the skin being the most common organ involved (2, 13). Classically, in patients with chronic urticaria plus angioedema, it has been shown that skin symptoms may appear after NSAID intake, even if patients are asymptomatic (14). The related mechanisms have also been attributed to COX-1 inhibition with an increase in the release of leukotrienes (31). Selective COX-2 inhibitors can usually be safely administered to patients with cutaneous symptoms with cross-intolerance to NSAIDs (11, 12, 22–24). However, rofecoxib and valdecoxib have been withdrawn from the market because of an increased incidence of cardiovascular complications (32). To assess in detail the tolerance to COX-2 inhibitors in subjects with hypersensitivity owing to cross-intolerance, we studied two groups: (A) a well-defined group of patients who were intolerant to paracetamol and (B) a group of patients who were tolerant to paracetamol. In all these patients, tolerance to paracetamol and etoricoxib was verified by challenge. Twenty-five percent of patients in Group A and 6% in Group B had etoricoxib intolerance (P < 0.03). Our data contrast with previous studies on tolerance to COX-2 inhibitors. In patients with airway involvement, limited evidence indicates that a very low proportion of patients are intolerant to COX-2 inhibitors (16–18). Concerning skin involvement, it has been shown that 0.2–3% of patients are intolerant to COX-2-selective inhibitors (22–24). Our figures are higher, with one of four patients intolerant to both etoricoxib and paracetamol. One reason for this may be a better patient selection and stratification/classification. Several reports showed a fewer number of cases, and patients were not classified into two groups (33). According to our study, intolerance to paracetamol seems to be a strong predictor for response to COX-2 inhibitors as has been previously suggested by other authors (33, 34). To determine whether these differences could be attributed to the presence of atopy in any of the groups, we analyzed, in detail, the atopic status by performing a skin test and determining total IgE level in serum, as well as by a clinical questionnaire recording all symptoms reported by the patients. Although a high prevalence was observed in both groups, the results were similar to previously reported data (13). In addition, no differences were observed between the two groups. In the group of 12 patients intolerant to both paracetamol and etoricoxib, nine were atopic, giving similar figures of atopy than in the total group A (75%). Concerning the group of patients tolerant to paracetamol, three showed positive response to etoricoxib. Of these three patients, two were atopic and one nonatopic, which gives similar values of atopy than the total group B (66.6%). Owing to the small size of the samples, no statistical comparison could be made.
Immunohistochemical studies showed the presence of tryptase in the extracellular space, indicating mast cell activation. However, these levels were not detected in peripheral blood, probably due to the lower intensity of the stimulus induced by etoricoxib compared with the drug responsible for the reaction. These data contrast with cases of severe anaphylaxis with typical increase in tryptase level in peripheral blood (35).
Cases of intolerance to NSAIDs, tolerance to paracetamol and yet intolerance to selective COX-2 inhibitors during drug provocation tests have been reported (36). Whether these patients developed a cell-mediated response was not verified. In our study, three patients with good tolerance to paracetamol were also intolerant to etoricoxib, experiencing in challenge pruritus and wheals localized on the back within 1 h of administering etoricoxib. Skin biopsy in these three patients also showed mast cell activation with the release of tryptase to the extracellular space.
The reasons why etoricoxib induces a positive response in subjects with good tolerance to paracetamol are at present not known. Paracetamol has a weak inhibitor action for COX-1 enzyme, but in some instances, subjects may react (8). Etoricoxib is 100-fold more selective for COX-2 than for COX-1 enzyme and only appears to have minimal interaction with COX-1 (37). However, the capacity for inhibiting COX-1 is lower than paracetamol (8, 37). Further studies are in progress for monitoring the local production of histamine and leukotriene in these patients.
In summary, we conclude that in patients with urticaria and/or angioedema with hypersensitivity owing to NSAIDs’ cross-intolerance, similar intolerance may be elicited with the selective COX-2 inhibitor etoricoxib.
We thank Ian Johnstone for help with the final English language version of the manuscript. This study was supported by grants from the Spanish Health Ministry Fund for Health in Spain (FIS) network RIRAAF (RD07/0064), PI071220, and PD09/02419.
Conflict of interest
None of the authors has any conflict of interest. None has received any money for this study. Research is part of their daily activities. All authors had full access to all the data (including statistical reports) and take responsibility for the integrity of the data and the accuracy of the data analysis. This study was supported by grants from the Spanish Health Ministry Fund for Health in Spain (FIS) network RIRAAF (RD07/0064), PI071220, and PD09/02419.
- 1Analgesic-antipyretic and antiinflammatory agents and drugs employed in the treatment of gout. In: Hardman JG, Limbird LL, Goodman Gilman A, editors. Goodman & Gilman’s. The Pharmacological Basis of Therapeutics, 10th edn. New York: McGraw-Hill, 2001: pp. 687–731., .
- 34A novel phenotype of nonsteroidal anti-inflammatory drug hypersensitivity. The high-risk patient. World Allergy Organiz J 2009;2:17–19..