Phenotype-based algorithm aims to untangle NSAID hypersensitivity and spare safe drugs

Non-steroidal anti-inflammatory drugs are among the most widely used medicines in the world, yet for a subset of patients they trigger urticaria, angioedema and, in severe cases, anaphylaxis. A new review published in Drugs & Therapy Perspectives argues that these reactions are routinely misdiagnosed and overgeneralized in clinical practice, leading to blanket avoidance of every NSAID even when safe alternatives exist. The authors, Ram Abou Mjahed and Ekatherine Prokopiou, synthesize current evidence on mechanisms, phenotypes, diagnosis and management, and distill it into a pragmatic, phenotype-based algorithm intended to guide classification, testing and therapy.

The review identifies two principal mechanisms behind NSAID-induced urticaria and angioedema, each with distinct clinical consequences. The first is a selective, immunologic allergy: the patient develops drug-specific IgE antibodies to a single NSAID or a chemically related group, producing classic type I hypersensitivity responses such as hives, swelling, bronchospasm or anaphylaxis on exposure to that drug while structurally unrelated NSAIDs are tolerated. The second, and more prevalent, mechanism is non-immunologic. It stems from pharmacologic inhibition of the cyclooxygenase-1 enzyme, which disrupts arachidonic acid metabolism, reduces protective prostaglandin E2 and shunts substrates down the 5-lipoxygenase pathway. The resulting overproduction of cysteinyl leukotrienes and other pro-inflammatory mediators directly activates mast cells without any IgE involvement. Because this pseudoallergic process depends on the pharmacologic effect rather than on a specific drug, susceptible individuals react to multiple COX-1-inhibiting NSAIDs, even ones they have never taken before.

The two pathways are not mutually exclusive in the population. Recent studies indicate that about 7 to 8 percent of patients with immediate reactions to two or more different NSAIDs actually tolerate aspirin, a potent COX-1 inhibitor, a pattern indicating separate IgE allergies rather than true cross-intolerance. The review also highlights a dose-threshold concept: aspirin inhibits COX-1 in a dose-dependent manner, with low doses of 75 to 100 milligrams exerting predominant platelet inhibition and higher doses of 500 to 1000 milligrams producing more complete systemic inhibition. A patient who tolerates low-dose aspirin but reacts to analgesic doses could therefore represent COX-1 intolerance rather than multiple separate allergies, a distinction the authors say should be considered before labeling such patients. Genetic factors add further complexity, with polymorphisms in leukotriene pathway enzymes and certain HLA alleles associated with differential risk, and patients with baseline chronic spontaneous urticaria showing an intrinsically lower threshold for NSAID-induced histamine release.

To bring order to this variability, the review applies the European Academy of Allergy and Clinical Immunology and European Network on Drug Allergy classification, which delineates five major phenotypes. NSAID-exacerbated respiratory disease affects patients with underlying asthma or chronic rhinosinusitis with nasal polyps, who develop bronchospasm and nasal symptoms within 30 to 180 minutes of ingestion; it affects roughly 10 to 20 percent of adults with asthma and up to 30 to 40 percent of those with both asthma and nasal polyposis. NSAID-exacerbated cutaneous disease occurs when NSAIDs worsen active chronic spontaneous urticaria, an intolerance estimated in approximately 20 to 30 percent of patients with active disease. NSAID-induced urticaria and angioedema describes cross-reactive reactions in otherwise healthy individuals and is the most common presentation, accounting for roughly one-third of cases in large series, with symptoms typically appearing within minutes to one to three hours and resolving within hours after the drug is stopped. The remaining phenotypes are single-NSAID-induced urticaria, angioedema or anaphylaxis, the IgE-mediated selective form, and single-NSAID-induced delayed reactions, which are T-cell-mediated, usually appear more than 24 hours after intake and can include severe syndromes such as Stevens-Johnson syndrome and toxic epidermal necrolysis.

Diagnosis is the hardest part, and the review is blunt about the limits of laboratory medicine here. No single laboratory marker definitively confirms most cases. Skin prick and intradermal testing are best established for pyrazolone NSAIDs such as metamizole, where in one series of metamizole allergy skin tests were positive in about 62 percent of cases and basophil activation tests added diagnoses in roughly another 28 percent. For most other NSAIDs, skin testing is not standardized, and serum-specific IgE assays are available only for a few drugs and often lack sensitivity. In one series of IgE-mediated metamizole urticaria, a cellular allergen stimulation test was negative in all six patients despite clear clinical allergy. Excluding pyrazolone allergies, the authors conclude there are no standardized skin tests or reliable in vitro assays for most NSAID hypersensitivity, making careful history taking combined with graded oral challenge testing, where safe, the cornerstone of accurate diagnosis.

Drug challenge testing remains the gold standard for immediate phenotypes. For suspected cross-reactive cases, a controlled oral challenge with aspirin or another strong COX-1 inhibitor verifies the diagnosis: a positive result confirms cross-reactive intolerance, while a negative result in a patient with multiple NSAID reactions suggests those reactions may have been selective allergies or coincidental rather than a class effect. Challenges are performed in a stepwise, graded-dose fashion under medical supervision with resuscitation equipment available, and are contraindicated in uncontrolled asthma or after life-threatening reactions unless absolutely necessary. They are also avoided with the culprit drug itself when a single-drug anaphylaxis history is convincing; instead, challenge with an unrelated NSAID can demonstrate normal tolerance. Notably, the review flags pediatric differences: in children, NSAID-induced urticaria without underlying chronic disease frequently occurs during acute viral infections when ibuprofen is taken as an antipyretic, and such reactions may be transient rather than true persistent intolerance, whereas NERD is rarely seen before adolescence.

Management follows the mechanism. In selective IgE-mediated allergy, the mantra is avoid and substitute: a patient allergic to ibuprofen must avoid other propionic acid derivatives such as naproxen, ketoprofen and flurbiprofen, but can generally use aspirin, diclofenac, oxicams or selective COX-2 inhibitors. In cross-reactive COX-1 intolerance, all strong COX-1 inhibitors, including aspirin and non-selective NSAIDs such as ibuprofen, naproxen, diclofenac, indomethacin and ketorolac, must be avoided. Here selective COX-2 inhibitors are the key alternatives: because they minimally inhibit COX-1, they typically do not trigger cross-reactive reactions, and studies show approximately 98 percent of patients with COX-1-mediated intolerance tolerate celecoxib, though tolerance rates may be slightly lower with meloxicam due to its partial COX-1 activity at higher doses. The consensus is that COX-2 inhibitors are first-line alternatives, but the first dose should be given under medical observation. Acetaminophen is not an NSAID and is generally safe, although at doses above 1000 milligrams it has mild COX-1 inhibition and can rarely precipitate symptoms in highly sensitive individuals, particularly those with respiratory disease.

For patients who genuinely need an NSAID, desensitization offers another route. It is most often applied in NSAID-exacerbated respiratory disease, where expert groups recommend considering aspirin desensitization followed by daily aspirin therapy, for example to allow cardiac protection or to treat sinonasal inflammation. The process involves gradually increasing aspirin doses over a day or two in a controlled hospital setting until a full dose is tolerated; the tolerant state is maintained only by daily dosing, and if doses are missed for more than 48 to 72 hours the hypersensitivity returns and the procedure must be repeated. Protocols start at extremely low doses, around 20 to 40 milligrams, doubling every one to two hours under monitoring. Desensitization deliberately triggers reactions and should be performed only by experienced allergists in settings with intensive care support. Adjuncts play a supporting role: leukotriene receptor antagonists such as montelukast do not reliably prevent cutaneous reactions but can attenuate respiratory reactions, and biologic agents including dupilumab, omalizumab, benralizumab and mepolizumab have shown benefit in severe respiratory disease, while omalizumab can indirectly reduce NSAID sensitivity in cutaneous disease by controlling the underlying chronic urticaria, with case reports of patients regaining aspirin tolerance after such therapy.

The review’s algorithm proceeds through defined steps: confirm the reaction is true hypersensitivity rather than a predictable dose-related side effect; determine whether timing was immediate or delayed; classify the phenotype from history, which alone can phenotype approximately 80 percent of cases; select targeted tests, reserving skin testing for suspected selective allergy and relying on graded oral challenge for cross-reactive forms; evaluate additional work-ups such as pulmonary function testing in respiratory disease; implement a mechanism-tailored management plan; and confirm the diagnosis with comprehensive patient education, including written documentation of agents to avoid and tolerate, label-reading guidance and an epinephrine auto-injector when reactions have been life-threatening. The authors note that a meaningful subset of patients, in the 10 to 20 percent range, present with blended or atypical features that do not fit standard categories, prompting recent algorithm updates to accommodate such variability.

The broader message is aimed squarely at primary care: resist blanket NSAID allergy labels. Only about 20 percent of all adverse reactions to NSAIDs are true hypersensitivity, and most of those are COX-1-mediated, for which skin and in vitro tests are largely unhelpful. Misclassification can cause either unnecessary restriction of all NSAIDs or preventable re-exposures to high-risk agents. A structured, phenotype-led history with referral for confirmation, the authors argue, prevents over-avoidance, preserves analgesic and anti-inflammatory options, and aligns practice with international guidance from EAACI, AAAAI and the World Allergy Organization. Current gaps, including the lack of validated biomarkers, classification gray zones and variable access to challenge testing, remain priorities for research, but consistent application of phenotype-based reasoning, the review concludes, enables safer and more nuanced prescribing for the many patients affected by these reactions.

Subject of Research: Phenotype-based diagnosis and management of NSAID hypersensitivity reactions

Article Title: Navigating NSAID hypersensitivity: a phenotype-based algorithm for diagnosis and safe therapy

Article References: Mjahed, R. A., & Prokopiou, E. (2026). Navigating NSAID hypersensitivity: a phenotype-based algorithm for diagnosis and safe therapy. Drugs & Therapy Perspectives. https://doi.org/10.1007/s40267-026-01274-5

Image Credits: AI Generated

DOI: 10.1007/s40267-026-01274-5

Keywords: NSAID hypersensitivity, urticaria, angioedema, COX-1 inhibition, IgE-mediated allergy, drug challenge testing, COX-2 inhibitors, aspirin desensitization, chronic spontaneous urticaria, anaphylaxis, EAACI classification, drug allergy