Drug Allergy

Table of Contents

  1. Overview
  2. Epidemiology
  3. Classification: Gell and Coombs Types I–IV
  4. Non-Immune Adverse Drug Reactions
  5. Penicillin Allergy Overdiagnosis and Delabeling
  6. Cross-Reactivity: Penicillins and Cephalosporins
  7. SCAR — Severe Cutaneous Adverse Reactions
  8. Drug Rechallenge and Desensitization
  9. Allergy Testing and Evaluation
  10. Key Drug Classes and Allergies
  11. Management Algorithm and Prevention
  12. Research Papers
  13. Connections
  14. Featured Videos

Back to Table of Contents

1. Overview

Drug allergy is an immune-mediated adverse drug reaction (ADR) that accounts for 5–10% of all adverse drug reactions. Unlike predictable pharmacologic side effects, drug allergies involve an inappropriate immune response to a drug or its metabolites, which can range from mild skin rash to life-threatening anaphylaxis or severe cutaneous reactions.

Understanding drug allergy has enormous practical consequences because of a pervasive public health problem: penicillin allergy overdiagnosis. Approximately 10% of the US population — roughly 33 million people — carries a documented "penicillin allergy" label. However, when formally evaluated, more than 90% of these patients can actually tolerate penicillin-class antibiotics. This widespread mislabeling forces clinicians to prescribe broader-spectrum alternatives (vancomycin, fluoroquinolones, clindamycin) that drive antibiotic resistance, increase Clostridioides difficile rates, worsen surgical outcomes, and add billions of dollars to US healthcare costs annually.

The 2022 Joint Task Force on Practice Parameters (JTFPP) and the 2019 NIAID Drug Allergy Guidelines provide current frameworks for the evaluation and management of drug allergy across all drug classes. A key theme in both is the move toward active delabeling — systematically removing inaccurate drug allergy labels through structured clinical evaluation.

Back to Table of Contents

2. Epidemiology

Drug allergy is among the most common and clinically consequential problems in medicine. Key epidemiologic facts:

Back to Table of Contents

3. Classification: Gell and Coombs Types I–IV

The Gell and Coombs classification (1963, updated) organizes immune hypersensitivity reactions into four mechanistic types. Most drugs that cause immune reactions act as haptens — they are too small to trigger immune responses alone and must bind to carrier proteins (usually serum proteins or cell surface molecules) to become immunogenic.

Type I — IgE-Mediated (Immediate)

Drug antigen binds IgE antibodies pre-loaded onto mast cells and basophils → cross-linking triggers immediate degranulation → release of histamine, tryptase, prostaglandins, leukotrienes. Onset: seconds to 1 hour. Clinical manifestations: urticaria, angioedema, bronchospasm, hypotension, anaphylaxis. Examples: penicillin anaphylaxis, latex allergy. Type I reactions require prior sensitization (no reaction on first exposure); the second or subsequent exposure triggers the response.

Type II — Cytotoxic (Antibody-Dependent Cell Cytotoxicity)

Drug coats cell surfaces → IgG or IgM antibodies bind the drug-coated cells → complement activation and NK cell-mediated destruction of the coated cells. Onset: days to weeks. Examples: drug-induced hemolytic anemia (penicillin at high doses, cephalosporins, methyldopa), drug-induced thrombocytopenia (quinine, heparin — note that heparin-induced thrombocytopenia is primarily driven by platelet factor 4/heparin complex antibodies with a distinct mechanism), drug-induced agranulocytosis (clozapine, propylthiouracil).

Type III — Immune Complex (Serum Sickness)

Drug-antibody immune complexes form in excess → deposit in vascular walls and tissues → complement activation → inflammation and tissue damage. Onset: 1–3 weeks after drug initiation. Clinical: serum sickness-like reaction (fever, urticaria, arthralgias, lymphadenopathy), drug-induced vasculitis, drug-induced lupus. Examples: serum sickness-like reaction to cefaclor and amoxicillin in children, hydralazine-induced lupus, minocycline-induced lupus, TNF inhibitor-induced anti-drug antibodies.

Type IV — T Cell-Mediated (Delayed)

T lymphocytes recognize drug-MHC complexes and mount a cytotoxic or inflammatory response. Onset: 48–72 hours to days or weeks. This is the most heterogeneous category, with four distinct subtypes that explain the diverse clinical presentations of delayed drug reactions:

The pharmacological interaction with immune receptors (p-i) concept (Pichler, 2002) extends Gell-Coombs: some drugs bind directly and non-covalently to MHC molecules or T cell receptors without hapten formation, triggering immediate T cell activation. This explains why some individuals react on first drug exposure without prior sensitization — particularly relevant for abacavir (HLA-B*57:01) and carbamazepine SCARs.

Back to Table of Contents

4. Non-Immune Adverse Drug Reactions (Mimickers of Allergy)

A critical clinical skill is distinguishing true immune-mediated drug allergy from the far more common non-immune adverse drug reactions that mimic allergy. Accurate distinction matters because non-immune reactions do not require avoidance of the drug class — the patient is not truly allergic.

Pharmacologic Side Effects

Predictable, dose-dependent effects of a drug's mechanism of action. Examples: NSAIDs causing GI bleeding (COX-1 inhibition in gastric mucosa — not allergy), opioids causing nausea and pruritus (direct mu-receptor effects), metformin causing GI upset (not allergy), ACE inhibitors causing cough (see below). These should never be labeled as drug allergy in the medical record.

Pseudoallergic / Non-Immune Mast Cell Activation

Drugs trigger mast cell degranulation directly (without IgE) → identical clinical picture to Type I allergy but without immune sensitization. This means: (1) can occur on first exposure, (2) not dose-dependent on prior sensitization, (3) reactions may improve with premedication or dose adjustment.

ACE Inhibitor Cough and Angioedema

ACE inhibitors block the degradation of bradykinin → bradykinin accumulates → kinin-mediated effects. Cough occurs in 10–20% of patients (higher in East Asian populations). Angioedema occurs in 0.5–1% of patients, with potentially life-threatening laryngeal involvement in 0.1–0.3%. Because the mechanism is bradykinin-mediated (not IgE), antihistamines and epinephrine have limited efficacy; icatibant (bradykinin B2 receptor antagonist) is the preferred acute treatment. ARBs are generally safe alternatives (cross-reactivity for angioedema is rare, approximately 5–10%).

Vancomycin "Red Man Syndrome"

Rapid vancomycin infusion directly triggers mast cell degranulation via MRGPRX2 → flushing, erythema, and pruritus of the face, neck, and upper trunk ("red man" distribution). This is rate-dependent and not IgE-mediated. Management: slow the infusion rate (infuse over at least 60 minutes for standard doses), pretreat with diphenhydramine. True IgE-mediated vancomycin anaphylaxis is distinct and rare.

Back to Table of Contents

5. Penicillin Allergy Overdiagnosis and Delabeling

The penicillin allergy "epidemic" is a major public health problem. The vast majority of penicillin allergy labels were assigned during childhood for rashes that were most likely viral exanthems (especially amoxicillin rashes during EBV mononucleosis, which occur in ~80% of cases) or non-immune drug reactions. These labels persist in electronic health records for decades without re-evaluation.

Consequences of Inaccurate Penicillin Allergy Labels

Penicillin Skin Testing (PST)

The gold standard for evaluating IgE-mediated penicillin allergy. Testing uses:

Skin prick testing is performed first; if negative, intradermal testing follows. A negative PST (both major and minor determinants) has a 97–99% negative predictive value for IgE-mediated reactions. After negative PST, a graded oral amoxicillin challenge confirms tolerance: 10% of the therapeutic dose followed 30 minutes later by the remaining 90%. Over 90% of labeled "penicillin-allergic" patients pass this evaluation.

A positive PST confirms current IgE-mediated sensitization and is an indication for allergy referral. For patients who require penicillin despite a positive PST (e.g., neurosyphilis in pregnancy), desensitization protocols are available.

Oral Amoxicillin Challenge Without Prior PST

For patients with very low-risk penicillin allergy histories (remote childhood rash of unknown type, GI symptoms labeled as allergy, family history of penicillin allergy only, or reactions to an unknown drug decades ago), direct oral amoxicillin challenge without prior skin testing is increasingly practiced. This approach is safe, cost-effective, and endorsed by the JTFPP 2022 guidelines for low-risk patients.

PENFAST Clinical Decision Score

The 5-question PENFAST score (Penicillin allergy — 5 questions; validated 2020) stratifies patients by predicted risk of IgE-mediated allergy based on: (1) time since reaction (≥5 years scores 0, <5 years scores 2), (2) reaction type (urticaria/angioedema = 2, anaphylaxis/severe = 3), (3) age at reaction (<30 = 0, ≥30 = 1), (4) treatment required (no = 0, antihistamine = 0, epinephrine = 2), (5) prior tolerated penicillin (yes = −2). Scores 0–2 = low risk (direct oral challenge reasonable); 3–4 = moderate risk (PST first); ≥5 = high risk (full allergy evaluation required).

Back to Table of Contents

6. Cross-Reactivity: Penicillins and Cephalosporins

The widely taught "10% cross-reactivity between penicillins and cephalosporins" is incorrect. This figure derived from flawed 1970s-era studies using first-generation cephalosporins that were contaminated with residual penicillin in the manufacturing process — the reactions observed reflected penicillin allergy, not true cross-reactivity. Multiple subsequent well-controlled studies have established the true picture:

Clinical implication: Cephalosporins — especially cefazolin, the most widely used surgical prophylaxis antibiotic — should not be universally avoided in patients with reported penicillin allergy. The 2022 guidelines support the use of cefazolin in most penicillin-allergic patients because cefazolin's R1 side chain has no structural analog in the penicillin family. Confirming the specific prior penicillin reaction and matching the R1 side chain is the appropriate clinical approach, not blanket cephalosporin avoidance.

Carbapenems: Cross-reactivity with penicillins is less than 1% based on side-chain dissimilarity; the 2022 JTFPP guidelines support carbapenem use in most IgE-mediated penicillin-allergic patients with appropriate monitoring. Aztreonam: The monobactam aztreonam shares an R1 side chain with ceftazidime — caution in patients allergic to ceftazidime specifically.

Back to Table of Contents

7. SCAR — Severe Cutaneous Adverse Reactions

Severe cutaneous adverse reactions (SCARs) are rare but potentially fatal T cell-mediated (Type IV) drug reactions that require immediate drug discontinuation and often intensive inpatient care. All SCARs are contraindications to re-exposure to the causative drug.

Stevens-Johnson Syndrome (SJS) and Toxic Epidermal Necrolysis (TEN)

SJS and TEN represent a disease spectrum defined by the extent of epidermal detachment: SJS = <10% body surface area (BSA) detachment; TEN = >30% BSA; SJS/TEN overlap = 10–30% BSA. Both are medical emergencies.

Mechanism: Drug-specific CD8+ cytotoxic T cells recognize drug-MHC-I complexes on keratinocytes and trigger apoptosis via the Fas-FasL pathway, perforin/granzyme B, and — most importantly — granulysin secretion, which is the principal cytotoxin causing the widespread epidermal necrosis of TEN.

HLA pharmacogenomics: Specific HLA alleles strongly predict SJS/TEN risk for certain drugs in specific ethnic populations:

Clinical presentation: Prodromal flu-like illness (fever, sore throat, conjunctival burning) 1–3 days before skin involvement → painful erythematous macules progressing to targetoid lesions → epidermal detachment → denuded skin resembling burns + erosive mucosal involvement (oral mucosa, conjunctiva, genital mucosa essential for diagnosis) → positive Nikolsky sign (gentle lateral pressure causes skin to slip). Systemic toxicity, fluid/electrolyte losses, sepsis risk.

Mortality: SJS 5–10%; TEN 25–35%. The SCORTEN score (Score of TEN; validated at 48h) incorporates age, HR, BUN, BSA involvement, bicarbonate, blood glucose, and presence of malignancy to predict mortality (score ≥5 = >90% mortality).

Treatment: Burn unit or ICU transfer; immediate withdrawal of causative drug (each day of delay worsens mortality); meticulous wound care and supportive therapy. Cyclosporine (3–5 mg/kg/day) has the most consistent evidence for improving outcomes. Etanercept (TNF-α inhibitor) showed benefit in a Phase III RCT (Taiwanese, 2018). IVIG at high doses remains controversial (conflicting evidence). Systemic corticosteroids are NOT recommended (may increase infection risk).

DRESS (Drug Reaction with Eosinophilia and Systemic Symptoms)

DRESS has a characteristically long latency of 2–8 weeks after drug initiation — longer than other drug reactions and often causing the drug to be overlooked as a cause. Also called DIHS (Drug-Induced Hypersensitivity Syndrome) in the Japanese literature.

Diagnostic features (RegiSCAR criteria): Morbilliform rash (often with facial edema, purpuric or bullous features), fever (>38.5°C), lymphadenopathy (≥2 sites), visceral organ involvement (liver most common — transaminitis, hepatitis; also kidneys, lungs [pneumonitis], heart [myocarditis], thyroid), eosinophilia (>1,500/μL), atypical lymphocytosis, reactive lymphadenopathy. HHV-6 and HHV-7 reactivation occurs during DRESS and is thought to amplify organ damage — serial HHV-6 PCR titers can confirm reactivation and track disease activity.

Common causative drugs: Allopurinol (most common cause globally), aromatic anticonvulsants (carbamazepine, phenobarbital, phenytoin — cross-reactivity within this group is ~50%), sulfonamides, dapsone, minocycline, vancomycin, lamotrigine.

Treatment: Systemic corticosteroids (prednisone 1 mg/kg/day, tapering over 6–12 weeks to prevent rebound) are standard of care. IVIG is used for severe cases with significant organ involvement. Critically: thyroid involvement occurs in 15–30% of patients, often months after resolution of the acute rash — thyroid function must be monitored for 6–12 months after DRESS. Mortality: 5–10% in hospitalized cases.

AGEP (Acute Generalized Exanthematous Pustulosis)

AGEP is distinctive for its rapid onset (within 1–2 days of drug exposure) — the fastest onset of any SCAR. Clinical: an eruption of hundreds of sterile, non-follicular, 1–2 mm pustules on a diffuse erythematous base, beginning in intertriginous areas and major flexures (axillae, groin, neck), then spreading; fever; peripheral neutrophilia (>7,000/μL). Mucous membrane involvement is absent or minimal (distinguishes from TEN/DRESS).

Common culprits: Beta-lactam antibiotics (most common), macrolides, diltiazem, terbinafine, hydroxychloroquine. Most cases are drug-induced; a small proportion are viral triggers.

Prognosis: AGEP is generally self-limiting — the pustular eruption resolves within 1–2 weeks of drug discontinuation, followed by superficial desquamation. Mortality is low (<5%), primarily in elderly patients with comorbidities and secondary infections. Distinction from generalized pustular psoriasis (GPP) is important: GPP is typically recurrent, lacks a clear drug trigger, and may have psoriasis history.

Back to Table of Contents

8. Drug Rechallenge and Desensitization

Drug Rechallenge

After evaluation establishes that a prior reaction was non-immune (pseudoallergic, pharmacologic side effect, or unlikely drug allergy), a graded oral drug challenge under observation can confirm tolerability. Rechallenge is appropriate for: non-immune urticaria/flushing (contrast media, opioids, vancomycin RMS), low-risk delayed maculopapular rash with negative skin testing, and probable viral exanthem (amoxicillin during EBV mononucleosis). Rechallenge is absolutely contraindicated after confirmed SJS/TEN or DRESS — re-exposure to the causative drug can trigger even more severe and rapid recurrence, and is potentially fatal.

Drug Desensitization

Desensitization is a procedure to induce temporary clinical tolerance to a drug in a truly allergic patient when no medically acceptable alternative exists. It is not a cure — it temporarily renders mast cells and basophils refractory to drug-induced degranulation through sustained receptor downregulation.

Mechanism: Stepwise administration of increasing drug doses (typically starting at 1/10,000 to 1/1,000 of the therapeutic dose and doubling at 15–30 minute intervals) progressively occupies and downregulates IgE-receptor complexes without triggering full degranulation. The desensitized state requires continuous maintenance dosing — any interruption leads to resensitization, and the protocol must be completely repeated if the drug is stopped.

Indications:

Setting: Desensitization must be performed in a monitored inpatient or supervised outpatient setting with immediate access to epinephrine, IV antihistamines, bronchodilators, and resuscitation equipment. Standardized protocols (12-step, 16-step) are available and routinely used in allergy centers.

Back to Table of Contents

9. Allergy Testing and Evaluation

Skin Testing (Prick and Intradermal)

Skin testing is the most clinically validated method for diagnosing IgE-mediated drug allergy. The procedure: prick testing first (lower sensitivity, safer), then intradermal testing at increasing drug concentrations (higher sensitivity) if prick is negative. Standardized reagents are available only for penicillin (Pre-Pen major determinant + minor determinant mixture) and insulin. For cephalosporins, carbapenems, sulfonamides, and most other drugs, skin testing reagents lack standardized validation and negative skin tests do not reliably exclude IgE-mediated allergy.

Serum Specific IgE (ImmunoCAP)

Commercial specific IgE assays exist for penicilloyl-G, penicilloyl-V, amoxicilloyl, ampicilloyl, and cefaclor. Sensitivity is lower than skin testing (approximately 50–70% for penicillin vs. 80–90% for PST). A negative ImmunoCAP does not rule out IgE-mediated allergy. These assays are most useful when skin testing is not available or in patients with extensive skin disease (eczema, dermatographism) that makes skin test interpretation unreliable.

Drug Patch Testing

Patch testing evaluates delayed-type (Type IV) drug hypersensitivity. Applied under occlusion for 48 hours, read at 48 and 96 hours. Positive results are informative for maculopapular exanthems, contact dermatitis, and can identify culprit drugs in DRESS and AGEP. Sensitivity varies significantly by drug and reaction type; a negative patch test does not exclude T cell-mediated allergy for most drugs.

Basophil Activation Test (BAT)

BAT measures CD63 or CD203c upregulation on basophils after drug stimulation by flow cytometry. Currently a research tool with no standardized clinical protocols for drug allergy. May offer value for drugs lacking validated skin test reagents. Not routinely available in clinical practice.

Graded Drug Challenge

The definitive test to confirm drug tolerance. After negative skin testing (or for low-risk history without skin testing), a supervised graded dose challenge administers the drug in 2–5 incremental doses (e.g., 1/100, 1/10, full dose at 30-minute intervals) under observation. Considered the gold standard for "ruling out" drug allergy for a specific patient. Should be performed in a setting capable of treating anaphylaxis.

Back to Table of Contents

10. Key Drug Classes and Allergies

Beta-Lactam Antibiotics

The largest and most clinically significant group. All beta-lactams share a four-membered beta-lactam ring but differ in side chains and fused ring structure (thiazolidine in penicillins, dihydrothiazine in cephalosporins, no fused ring in monobactams/aztreonam, oxazolidine ring in carbapenems). Cross-reactivity is governed by R1 side-chain similarity, not ring structure (see Section 6). Management hierarchy: (1) confirm true allergy, (2) if confirmed IgE-mediated, use a cephalosporin with a dissimilar R1 side chain (e.g., cefazolin, ceftriaxone) — or desensitize if first-line penicillin is necessary.

Sulfonamides (TMP-SMX, Sulfadiazine)

Sulfonamide antibiotics (TMP-SMX, sulfadiazine) cause both IgE-mediated urticaria and Type IV delayed reactions including SJS and DRESS. Reaction rates are dramatically higher in HIV-positive patients (40–60% vs. ~3% in HIV-negative) due to impaired detoxification of hydroxylamine sulfonamide metabolites. A common clinical question: does sulfonamide antibiotic allergy predict cross-reactivity to "sulfa-containing" non-antibiotic drugs (furosemide, hydrochlorothiazide, celecoxib, sulfonylureas)? Answer: No clinically meaningful cross-reactivity. Non-antibiotic sulfonamides lack the N1-substituent and p-amino group that are the immunogenic determinants of sulfonamide antibiotic reactions. The clinical decision to avoid non-antibiotic sulfonamides in patients with sulfonamide antibiotic allergy is not evidence-based.

NSAIDs and Aspirin

NSAID reactions are predominantly pharmacologic/pseudoallergic via COX-1 inhibition (see Section 4), not IgE-mediated. Two distinct NSAID hypersensitivity syndromes: (1) NSAID-exacerbated respiratory disease (AERD) — cross-reacts with all COX-1 inhibitors; COX-2-selective agents tolerated; aspirin desensitization can benefit with maintained daily aspirin. (2) NSAID-exacerbated cutaneous disease (NECD) — urticaria/angioedema in patients with chronic spontaneous urticaria, triggered by any COX-1 inhibitor. A third, rare pattern is single-NSAID-induced urticaria/angioedema (SNIUA) — truly IgE-mediated to a single NSAID; other NSAIDs are tolerated.

Radiocontrast Media

Radiocontrast reactions are pseudoallergic (non-IgE mast cell activation). Use of low-osmolar non-ionic contrast media (LOCM) reduces reaction rates from ~5% (high-osmolar ionic) to ~0.2–0.5%. Prior reaction to contrast media is a risk factor for subsequent reactions (~35% recurrence without premedication). Premedication protocol (see Section 4) reduces recurrence to ~10%. The historical belief that shellfish allergy predicts contrast allergy (the "iodine allergy" myth) is debunked — shellfish allergy is a food protein allergy unrelated to iodine or contrast media; it is not a contraindication or risk factor for contrast use.

Biologic Agents and Monoclonal Antibodies

Infusion reactions to TNF inhibitors (infliximab), anti-CD20 agents (rituximab), HER2 inhibitors (trastuzumab), and EGFR inhibitors (cetuximab) are increasingly common and can be IgE-mediated (cetuximab — IgE to alpha-gal oligosaccharide in the Fab region, in patients with alpha-gal syndrome) or non-IgE complement-mediated cytokine release syndrome. PEG (polyethylene glycol) hypersensitivity is an emerging entity: IgE to PEG-2000 in lipid nanoparticles (relevant to mRNA vaccines and PEGylated drugs). Standardized desensitization protocols for monoclonal antibodies have been developed and validated by Castells et al. at Brigham and Women's Hospital.

Back to Table of Contents

11. Management Algorithm and Prevention

Structured Evaluation Algorithm for Reported Drug Allergy

  1. Detailed history: What drug? What was the exact reaction (rash description, systemic symptoms)? When relative to drug start (immediate <1h vs. delayed days/weeks)? What treatment was required (antihistamine, epinephrine, hospitalization)? Any subsequent uneventful exposures? Was the patient actually taking the drug when the reaction occurred? (Reaction to a new drug often misattributed to a longstanding drug.)
  2. Risk stratification: Classify as low risk (remote uncharacterized rash, GI intolerance, family history only), moderate risk (urticaria/angioedema, unclear timing), or high risk (anaphylaxis, prior positive skin test, SCAR).
  3. Select evaluation strategy: Low risk → direct oral challenge; moderate risk → penicillin skin testing + oral challenge; high risk → comprehensive allergy evaluation ± desensitization if drug required.
  4. Document clearly: Record the result of the evaluation (tolerated, true IgE-mediated allergy confirmed, label removed) with date and evaluating clinician. Inaccurate allergy documentation propagates through the medical record for decades.
  5. Provide patient education: Written documentation of allergy status, what specific drug is avoided and why, what alternatives are safe, emergency action plan (Medic-Alert bracelet for anaphylaxis history).

Pharmacogenomic Prevention of SCARs

Pre-prescription genetic screening for high-risk HLA alleles is now standard of care for specific drug-population combinations:

These pharmacogenomic screens are cost-effective given the high mortality and hospitalization costs of SCARs, and represent a successful model of precision medicine preventing drug toxicity.

Back to Table of Contents

12. Research Papers

Explore peer-reviewed literature on drug allergy through these PubMed topic searches:

  1. Drug allergy diagnosis
  2. Penicillin allergy delabeling
  3. Penicillin skin testing
  4. Drug hypersensitivity classification
  5. Stevens-Johnson syndrome treatment
  6. DRESS drug reaction eosinophilia
  7. Drug allergy cross-reactivity beta-lactam
  8. Penicillin allergy prevalence
  9. Severe cutaneous adverse reactions
  10. Drug desensitization protocols
  11. HLA pharmacogenomics drug allergy
  12. SCAR toxic epidermal necrolysis

Back to Table of Contents

Connections

Back to Table of Contents