Mastocytosis presenting with mast cell‐mediator release‐associated symptoms elicited by cyclo oxygenase inhibitors: prevalence, clinical, and laboratory features

Abstract Background Nonsteroidal anti‐inflammatory drugs (NSAIDs) are frequently avoided in mastocytosis, because of a potential increased risk for drug hypersensitivity reactions (DHRs) due to inhibition of cyclo‐oxygenase (COX), subsequent depletion of prostaglandin E 2 and release of leukotrienes. Objectives Here, we aimed at determining the prevalence of mast cell (MC) mediator release symptoms triggered by NSAIDs in mastocytosis patients and the associated clinical and laboratory features of the disease. Methods Medical records from 418 adults to 223 pediatric mastocytosis patients were retrospectively reviewed. Patients were classified according to tolerance patterns to NSAIDs and other COX inhibitors (COXi) and compared for epidemiological, clinical and laboratory findings. Results Overall, 87% of adults and 91% of pediatric patients tolerated NSAIDs and other COXi. Among adult and pediatric patients presenting DHRs, 5% and 0% reacted to multiple NSAIDs, 4% and 0.7% were single reactors, and 3% and 8% were single reactors with known tolerance to paracetamol but unknown tolerance to other COXi, respectively. Among adults, hypersensitivity to ≥2 drugs was more frequent among females (p = 0.009), patients with prior history of anaphylaxis to triggers other than NSAIDs or other COXi and Hymenoptera venom (p = 0.009), presence of baseline flushing (p = 0.02), baseline serum tryptase ≥48 ng/ml (p = 0.005) and multilineage KIT mutation (p = 0.02). In contrast, tolerance to NSAIDs and other COXi was more frequent among males (p = 0.02), in patients with anaphylaxis caused by Hymenoptera venom (p = 0.02), among individuals who had skin lesions due to mastocytosis (p = 0.01), and in cases that had no baseline pruritus (p = 0.006). Based on these parameters, a score model was designed to stratify mastocytosis patients who have never received NSAIDs or other COXi apart from paracetamol, according to their risk of DHR. Conclusions Our results suggest that despite the frequency of MC mediator related symptoms elicited by NSAIDs and other COXi apart from paracetamol is increased among mastocytosis patients versus the general population, it is lower than previously estimated and associated with unique disease features. Patients that tolerated NSAIDs and other COXi following disease onset should keep using them. In turn, adults with unknown tolerance to such drugs and a positive score should be challenged with a preferential/selective COX‐2 inhibitor, while the remaining may be challenged with ibuprofen.


| INTRODUCTION
Mastocytosis encompasses a heterogeneous group of rare diseases characterized by the accumulation of clonal and phenotypically aberrant mast cells (MC) in different tissues and organs, such as the skin, bone marrow (BM) and the gastrointestinal (GI) tract. 1,2 Mastocytosis might affect children and adults, independently of gender, 1,2 with an estimated prevalence of ≈9 cases per 100,000 individuals. 3,4 Currently, the World Health Organization (WHO) defines seven diagnostic subtypes of mastocytosis that include cutaneous masto- Moreover, a provisional diagnostic subtype of ISM-that is bone marrow mastocytosis (BMM)-is also defined by WHO, 5 and another variant of well differentiated SM has been recently characterized. 6,7 Most mastocytosis patients present with mild to severe symptoms caused by the release of MC mediators, infiltration of tissues by MC, or both, 1,2 in association with an overall higher prevalence of hereditary alpha tryptasemia. 8 Activated MC release multiple vasoactive, proinflammatory, chemotactic and immunomodulatory mediators which are both prestored in granules and produced de novo. 9 A broad variety of triggers for MC activation have been described so far, which, among others, include cyclooxygenase inhibitors (COXi). 10,11 According to the Anatomical Therapeutical Chemical (ATC)/WHO classification, COXi include traditional nonsteroidal anti-inflammatory drugs (NSAIDs)butylpyrazolidines, acetic acid derivatives and related substances, oxicams, propionic acid derivatives, fenamates, and other not otherwise classified drugs (e.g., nimesulide, nabumetone, clonixin lysine) -COX-2 selective NSAIDs (i.e. coxibs), and analgesics and antipyretics (henceforth other COXi) such as salycilates (e.g. acetylsalicylic acid, ASA), pyrazolones and paracetamol. 12 NSAIDs and ASA inhibit COX, which results in pain control and both anti-inflammatory and antipyretic effects. In contrast, paracetamol exerts a limited inhibitory effect on COX-1, lacks anti-inflammatory effects and exerts its analgesic and antipyretic effects through central nervous system COX-2 inhibition. 13 In Spain, the prevalence of hypersensitivity to NSAID ranges between 1% and 3% in the (adult) general population, 14 while it is estimated to be 1.2% in children. 15 In contrast to the general population, previous studies in mastocytosis showed a greater prevalence of NSAID hypersensitivity of up to 14% in adults and 2% in pediatric patients. 16 In addition, NSAIDs have been reported to cause anaphylaxis in between 2% and 11% of adult mastocytosis patients. [17][18][19][20][21] Because of this, strict avoidance of NSAIDs in mastocytosis is often recommended in routine clinical practice, due to safety concerns. 22 From a pathogenic point of view, patients might display NSAID or other COXi hypersensitivity to specific drugs or structurally related groups of drugs-single reactors-and to different structurally unrelated drugs-multiple reactors. 23 Pyrazolones frequently cause IgE-mediated hypersensitivity reactions, 24 while NSAIDs and other COXi most commonly induce IgE-independent reactions through a COX-1-related mechanism. 25 COX-1 inhibition depletes protective PGE 2 , 25,26 resulting in increased production of cysteinylleukotrienes 27 that leads to symptoms ranging from urticaria to anaphylaxis. 25 NSAID hypersensitive patients might react to COX-1 inhibitors while tolerating weak COX-1 inhibitors (e.g. paracetamol) and COX-2 (preferential or selective) inhibitors (i.e. meloxicam or coxibs, respectively), that induce a lower decrease in PGE 2 with lower production of leukotrienes. 28 Whether the mechanisms involved in NSAID-associated release of MC mediators in mastocytosis is similar or not to that described for the general population, currently remains unclear.
Here, we retrospectively analyzed the prevalence of MCmediator release-associated symptoms triggered by NSAIDs and other COXi in a large series of mastocytosis patients, and compared the clinical and laboratory features of these patients with those of other mastocytosis patients, in order to search for a potentially unique clinical and laboratory profile associated with hypersensitivity to these drugs.

| Study design
Randomly selected medical records from a total of 641 patients-418 adults and 223 children and adolescents <18 years old-diagnosed with mastocytosis as per the WHO criteria at the Spanish Network on Mastocytosis (REMA) and that had been followed for a minimum period of 1 year, were retrospectively reviewed. Eighty-two pediatric patients with insufficient or inconsistent clinical data were subsequently excluded from the analysis. Ninety patients (36 adult and 54 pediatric patients) who had never received NSAIDs or other COXi following the onset of mastocytosis, but that had tolerated paracetamol, were also further excluded from the study. Data recorded on the remaining 469 patients ( Figure S1)-382 adults and 87 children and adolescents-included: diagnostic subtypes of mastocytosis, basal MC mediator-related symptoms (pruritus, flushing and GI symptoms) and prior history of anaphylaxis and its respective triggers, among other clinical and laboratory features of the disease, which are described in more detail in Table 1.
This study was approved by the Ethics Committee of the Complejo Hospitalario de Toledo (Toledo, Spain) and every procedure was in accordance with the Declaration of Helsinki. Written informed consent was given by each patient and/or patient legal guardian for collection of clinical data and, in a subgroup of patients also for undergoing drug challenge testing.

| Definitions, diagnostic procedures and laboratory tests
Onset of mastocytosis was defined either as the date of first appearance of cutaneous lesions, or in cases who presented in the absence of the typical skin lesions of mastocytosis as the first episode of anaphylaxis, or detection of B and/or C findings. 29 Diagnosis of mastocytosis was retrospectively revised based on well-established morphological, 30 histopathological, immunohistochemical, 31 immunophenotypic 31 and molecular criteria, 32 according to the WHO classification criteria 5 and more recent criteria for WDSM. 6,7 Multilineage KITD816V mutation was defined as involvement of fluorescence activated cell sorting (FACS)-purified non-MC myeloid and/or lymphoid cell populations by this KIT mutation as assessed by a previously described 32 peptide nucleic acid (PNA) polymerase chain reaction (PCR) clamping (PNA-PCR) technique. In turn, patients with the KITD816 V mutation restricted to BM MC were categorized as carrying a MC-restricted KITD816V mutation in BM. 32 Diagnosis of anaphylaxis followed the 2011 World Allergy Organization Guidelines. 33 Patients over 18 years old (adults) with cutaneous involvement in the absence of a BM study were categorized as mastocytosis in the skin (MIS), since SM could not be confirmed or ruled out. 34 Blood tests performed at diagnosis and at follow-up included: complete blood cell count and differential, routine biochemistry, serum baseline tryptase (sBT; ImmunoCAP Tryptase, Phadia/Thermo Fisher Scientific Inc, Uppsala, Sweden) and both total and specific serum IgE levels (ImmunoCAP total IgE, Phadia/Thermo Fisher Scientific Inc.). Specific IgE levels were measured whenever appropriate (ImmunoCAP allergen components, Phadia/Thermo Fisher Scientific Inc.). In addition, skin tests (e.g. skin prick and intradermal tests) were performed with specific triggers (e.g. Hymenoptera venom, aeroallergens, foods and drugs). Allergic sensitization was defined based on positive specific IgE antibodies or a positive skin test. 31

| Diagnosis and classification of hypersensitivity reactions to NSAID or other COXi
Data on MC activation-associated symptoms induced by NSAIDs and other COXi was retrospectively recorded for each individual patient, through specific anamnesis and review of medical records and it included: the age at disease onset, the type and severity of symptoms and (later) tolerance to each drug used by individual patients, and diagnostic procedures, as follows. Skin tests were performed whenever an underlying IgE-mediated hypersensitivity (i.e., single reaction to metamizole performed and positive in two adult patients) was suspected. Drug challenge tests were performed using the suspicious drug or a preferential/selective COX-2 inhibitor, following risk/benefit assessment. Specifically, the suspicious drug (for children) was used when true reactions were unlikely and when the suspicious reaction was mild. Meloxicam or coxibs (for adults) were used when true reactions were likely, if the reaction had been moderate/severe, and when NSAIDs or other COXi were not tolerated following the index episode, as previously recommended for the general population. 35 Patients who had no MC mediator release-related symptoms caused by the administration of NSAIDs and other COXi were classified as tolerant. Multiple reactors were defined by the presence of MC activation (MCA)-associated symptoms caused by ≥2 structurally unrelated NSAIDs or COXi. Those patients who presented reactions elicited by a single NSAID or other COXi were subclassified either as single reactors (those who avoided only that particular drug or group of structurally related drugs while tolerating other NSAIDs after that episode) or as single reactors with known tolerance to paracetamol but unknown tolerance to NSAID and other COXi (those who avoided all NSAIDs and other COXi apart from paracetamol from that moment on).

| Pediatric patient series
From 87 pediatric patients analyzed, 38 (44%) were girls, with a median age at the moment of entering the study of 10 years (range: 2-17 years) and at disease onset of 4 months (range: birth-10 years). Further details on patient demographics and clinical and laboratory characteristics are shown in Table 1 and causes for anaphylaxis according to the diagnostic subtype of mastocytosis are displayed in Table S1.

| Pediatric cohort
A total of 85 (98%) pediatric patients received ibuprofen (among them, 9 and 4 also received metamizole and ASA, respectively) and 2  and laboratory characteristics, no statistically significant differences were found among pediatric patients that were tolerant versus reactive to NSAIDS and other COXi (Table S2).
Anaphylaxis caused by NSAIDs or other COXi occurred in 2/8 (25%) patients who had a DHR, and it was caused by ibuprofen and metamizole, respectively. Urticaria was the most frequent symptom presented during DHRs in 4/8 (50%) patients (Table S2).

| DISCUSSION
Prescription of NSAIDs and COXi other than paracetamol, is often empirically avoided in mastocytosis patients, because of a general concern for a greater frequency of MCA symptoms triggered by these drugs, particularly anaphylaxis. 11 However, data about the true prevalence of MCA induced by NSAIDs and other COXi, largely varies in the literature [19][20][21] and is usually based on limited series of patients with mastocytosis. In this study, based on a large retrospective series of mastocytosis patients we confirmed that a significant fraction (13%) of adults diagnosed with mastocytosis who had previously used these drugs did actually experienced DHRs after using them, less than half of them being reactors to multiple NSAIDs or other COXi. In pediatric patients, the data collected here are less informative, because the sample size and NSAID use were relatively limited. Overall, these results confirm the greater frequency of hypersensitivity reactions to NSAIDs and other COXi in adult mastocytosis compared to the general population in Spain, although the rates might vary depending on the studied patient cohort, as well as potential regional differences. [19][20][21] Interestingly, the frequency of DHRs to NSAIDs and other COXI found in our adult mastocytosis patients is slightly higher than that previously reported for asthma,  19,20 show that sBT among single reactors is similar to that found in Similarly to the general population, propionic acid derivatives were the most frequent cause for NSAID hypersensitivity, which might be due to the greater use of these versus other NSAIDs, for all age groups. 40 In our series however, these were amongst the safest NSAIDs, as only a small fraction (≤8%) of mastocytosis patients that had used ibuprofen, naproxen and dexketoprofen referred reactions to these drugs. In contrast, a significantly higher rate of hypersensitivity to ASA and metamizole was observed among patients who had received these drugs. Aceclofenac and clonixin were also associated with frequent reactions, but the number of patients who received them is rather limited to draw any definitive conclusions.
Previous studies suggested that mastocytosis patients who have hypersensitivity to ASA, would be tolerant to NSAIDs, 40  no MCA-associated symptoms other than anaphylaxis, low sBT and BMMC burden, in the absence of multilineage KIT mutation. 37 Based on these results, a score model was built with a high sensitivity (associated with optimal negative predictive value), but a still limited positive predictive value, that would allow identification of one multiple reactor among each seven patients identified to be at risk.
Based on these results, it might be wise that adult mastocytosis patients who require NSAIDs and who had never used them are submitted to a drug challenge with ibuprofen when the score here proposed is negative, or with either a coxib or meloxicam when the score is positive. In our experience, ibuprofen is safe in most children, but the first administration should be performed under medical surveillance whenever previous tolerance to the drug is not known (Figure 2). Patients who have tolerated specific NSAIDs following the onset of the disease do not require further testing and may be instructed to use the previously tolerated drug(s) 11

CLINICAL IMPLICATIONS
Hypersensitivity to NSAIDs and other COXi in mastocytosis patients is less frequent than previously estimated and it is associated with unique disease features.