Drug Reactions — Exanthems and DRESS Syndrome

• T-cell mediated delayed hypersensitivity (Type IV) as primary mechanism: DRESS syndrome results from abnormal CD8+ and CD4+ T-cell responses to drug hapten-protein complexes presented via HLA molecules on antigen-presenting cells. The inciting drug or its metabolites covalently bind to endogenous proteins, creating neo-antigens that breach central tolerance. Initially, circulating lymphocytes recognize these altered self-antigens, leading to robust T-cell proliferation and activation. This contrasts with IgE-mediated reactions (Type I hypersensitivity) that produce urticaria within minutes; DRESS syndrome's delayed onset (typically 2-8 weeks after drug initiation) reflects the time required for naive T-cell priming, clonal expansion, and trafficking to cutaneous and systemic sites. Genetic factors including specific HLA allotypes (HLA-A*31:01 for carbamazepine, HLA-B*5801 for allopurinol in certain populations) significantly predispose to severe reactions, suggesting that MHC-peptide binding geometry critically influences whether drug metabolites generate immunogenic epitopes.

• Defective regulatory T cell (Treg) function and loss of immune tolerance: A pathogenic hallmark of DRESS syndrome involves quantitative and qualitative Treg dysfunction. Affected patients demonstrate reduced CD4+CD25+Foxp3+ Treg frequencies and impaired suppressive capacity, allowing unchecked effector T-cell proliferation. This defect may represent either intrinsic Treg dysfunction or preferential Treg apoptosis induced by the offending drug or its reactive metabolites. Additionally, patients with DRESS syndrome exhibit reduced expression of inhibitory molecules (PD-1, CTLA-4) on T-cell surfaces, further reducing negative feedback signaling. The loss of immune homeostasis permits vigorous expansion of self-reactive T-cell clones that would normally be constrained by Treg-mediated suppression, perpetuating systemic inflammation even after drug discontinuation. This explains the clinical observation that DRESS syndrome may progress for weeks despite drug cessation, necessitating corticosteroid-mediated Treg restoration.

• Viral reactivation as amplifier of immune dysregulation (HHV-6 and EBV involvement): Emerging evidence demonstrates that human herpesvirus-6 (HHV-6), Epstein-Barr virus (EBV), and cytomegalovirus (CMV) reactivation occurs in 40-80% of DRESS cases, with HHV-6 being most frequently detected. This reactivation occurs as a consequence rather than primary cause of immune dysregulation—the defective Treg responses and abnormal T-cell control permit latent viral replication. Viral reactivation then amplifies systemic inflammation through multiple mechanisms: (1) viral antigens provide additional T-cell stimulation and bystander activation of non-specific T-cells; (2) viral superantigens bypass HLA-restricted recognition, triggering massive non-specific T-cell expansion; (3) cross-reactive epitopes between viral and drug antigens sustain T-cell activation even after drug clearance. This viral component explains why some DRESS cases demonstrate clinical deterioration despite drug discontinuation and why PCR detection of HHV-6 in serum or PBMCs predicts more severe disease requiring higher corticosteroid doses.

• Hepatic involvement through direct drug metabolite toxicity and immune-mediated injury: The liver represents a primary target organ in DRESS syndrome through two complementary mechanisms. First, phase I cytochrome P450 metabolism (particularly CYP3A4 with aromatic antiepileptic drugs) generates reactive metabolites that form protein-DNA adducts within hepatocytes, creating neoantigens recognized by drug-reactive CD8+ T-cells. Second, the hepatic microenvironment, rich in tissue-resident memory T-cells and antigen-presenting cells, facilitates focal accumulation of drug-specific lymphocytes. Portal tract inflammation with predominantly CD8+ T-cell infiltration leads to hepatocellular apoptosis and transaminitis (typically 3-10x upper limit of normal). Some cases progress to drug-induced liver injury (DILI) with synthetic dysfunction and coagulopathy, potentially progressing to acute liver failure. The degree of hepatic involvement correlates with DRESS severity and predicts need for intensive immunosuppression.

• Hematologic consequences of excessive cytokine production: Elevated levels of Th1 cytokines (IFN-γ, TNF-α, IL-2) and Th2 cytokines (IL-4, IL-5, IL-13) characterize DRESS syndrome, with relative Th1 predominance in early phases and Th2 shift in later stages. IL-5 overproduction drives marked peripheral eosinophilia (commonly >1500/μL) through STAT5 signaling in bone marrow hematopoietic cells, promoting eosinophil differentiation and survival. TNF-α and IL-6 trigger splenic and lymph node hyperplasia with reactive proliferation of lymphoblasts, creating atypical lymphocytosis mimicking acute leukemia on blood smear. In severe cases, macrophage hyperactivation driven by excessive IFN-γ and TNF-α can progress to hemophagocytic lymphohistiocytosis (HLH), characterized by macrophages engulfing RBCs, WBCs, and platelets within lymphoid tissues. This progression—occurring in 2-5% of DRESS cases—represents a medical emergency with rapid clinical deterioration and markedly elevated ferritin (often >5000 ng/mL), elevated soluble IL-2 receptor, and depressed NK cell function.

• Cutaneous manifestations driven by lymphocyte trafficking and skin-homing receptors: The characteristic DRESS exanthem results from CD8+ T-cell infiltration of superficial and mid-dermal regions, driven by upregulation of cutaneous lymphocyte-associated antigen (CLA) on drug-reactive T-cells and corresponding E-selectin expression on skin endothelial cells. The infiltrating lymphocytes release cytotoxic granules (perforin, granzyme B) targeting keratinocytes, leading to individual cell apoptosis visible as dyskeratotic cells on histopathology. This explains why DRESS exanthems are predominantly lichenoid or interface dermatitis patterns histologically, rather than urticarial (which shows mast cell degranulation without lymphocytic infiltration). The timing of rash appearance (typically 3-6 weeks) reflects the duration required for drug-specific T-cell priming in regional lymph nodes, homing to skin, and accumulation to sufficient density to produce visible inflammation.

• Aromatic antiepileptic drugs (carbamazepine, phenytoin, phenobarbital) as prototype offenders: These agents represent the most common DRESS triggers in developed nations, with carbamazepine implicated in approximately 20-25% of DRESS cases. The mechanism involves CYP2C9 and CYP3A4-mediated formation of reactive arene oxide metabolites that form protein adducts. Carbamazepine carries particularly high DRESS risk in HLA-A*31:01-positive individuals (primarily East/Southeast Asian descent), with relative risk exceeding 1000-fold compared to HLA-A*31:01-negative individuals. Phenytoin can produce a similar reaction termed "phenytoin hypersensitivity syndrome," with overlapping features. These drugs demonstrate a dose-dependent risk, with DRESS emerging within 2-8 weeks of initiation or dose escalation. Cross-reactivity occurs among aromatic antiepileptics (carbamazepine with phenytoin, phenobarbital), so switching between agents is contraindicated in DRESS cases.

• Allopurinol as second most common culprit with severe systemic toxicity: Allopurinol-induced DRESS accounts for 10-15% of cases in tertiary referral centers, with particular virulence and high mortality (approximately 20-30% if severe). The xanthine oxidase inhibitor undergoes hepatic metabolism to the active metabolite oxypurinol, which accumulates and generates immune reactivity. HLA-B*5801 positivity strongly predisposes to severe allopurinol reactions (relative risk 80-1000 fold in Han Chinese, Thai, and Korean populations), prompting current guidelines recommending HLA-B*5801 screening prior to allopurinol initiation in these populations. Allopurinol-DRESS frequently presents with particularly severe hepatic involvement and renal dysfunction, likely reflecting both direct cytotoxicity and immune-mediated injury. Dose and renal function interact; patients with pre-existing renal impairment have reduced oxypurinol clearance, leading to accumulation and heightened immune responses.

• Sulfonamide antibiotics (trimethoprim-sulfamethoxazole, sulfasalazine) and non-aromatic drug triggers: Sulfonamides cause DRESS in approximately 10% of cases, with particular frequency in HIV-positive patients receiving trimethoprim-sulfamethoxazole (TMP-SMX) prophylaxis. The mechanism involves CYP2C9 metabolism to reactive N-hydroxylamine and nitroso metabolites. Interestingly, non-aromatic drugs also trigger DRESS syndrome: NSAIDs (particularly naproxen, ibuprofen), beta-blockers (atenolol, labetalol), ACE inhibitors, and antiretrovirals (particularly abacavir, nevirapine) have all been implicated. This broader drug repertoire emphasizes that DRESS represents a generic immune response pathway, not exclusive to aromatic compounds. The latency period varies by drug class (aromatic drugs: 2-8 weeks; sulfonamides: 1-2 weeks; NSAIDs: variable).

• Host genetic and immunologic risk factors beyond HLA associations: While HLA genotyping has greatest predictive value (HLA-A*31:01, HLA-B*5801), additional genetic polymorphisms contribute to DRESS susceptibility. Genetic variants affecting drug metabolism (CYP3A4 and CYP2C9 polymorphisms), glutathione S-transferase variants affecting reactive metabolite conjugation, and functional TNF-α and IL-10 promoter polymorphisms influence individual disease risk. Functional Treg deficiencies (reduced IL-2 signaling, FOXP3 polymorphisms) may predispose to loss of immune tolerance. Additionally, prior exposure to structurally similar drugs increases risk through molecular mimicry and pre-existing T-cell sensitization.

• Clinical risk factors and comorbidities: Older age (>50 years) carries increased risk, possibly reflecting altered drug metabolism and reduced immune regulation. Patients with HIV infection, particularly those with CD4+ counts <200 cells/μL, demonstrate dramatically elevated DRESS incidence with sulfonamides. Concurrent polypharmacy increases risk by both exposing patients to multiple potential triggers and complicating causality assessment. Underlying autoimmune diseases and prior severe drug reactions suggest immune dysregulation predisposing to future reactions.

• Fever as cardinal opening symptom with variable patterns: Fever typically precedes rash by days to weeks, presenting as the initial manifestation in approximately 80% of DRESS cases. The fever pattern is frequently spiking and quotidian (daily), reflecting underlying systemic immune activation and elevated IL-6 production. Temperatures typically range from 38.5-40°C (101-104°F), often accompanied by malaise, myalgias, and arthralgias suggesting systemic infection. The fever may be refractory to antipyretics and persists despite negative blood cultures, distinguishing drug reaction from infectious etiologies. In some cases, fever begins weeks after drug initiation, coinciding with viral reactivation detected by PCR.

• Polymorphic exanthem with variable morphology complicating diagnosis: The DRESS rash classically presents as a maculopapular eruption involving the face (particularly periorbital edema), neck, chest, and proximal upper extremities, with variable extension to trunk and extremities. Notably, the morphology is highly variable: lesions may remain pure maculopapular, evolve to lichenoid papules, develop vesiculation or purpura, or demonstrate "atypical lymphocytosis-like" appearance with target lesions mimicking erythema multiforme. This polymorphism reflects underlying lymphocytic infiltration and keratinocyte apoptosis, but lacks the geometric precision of true erythema multiforme. Pruritus is nearly universal and often severe, driven by T-cell lymphokine production and skin inflammation. The rash typically appears 2-6 weeks after drug initiation and may continue evolving for weeks despite drug discontinuation, occasionally persisting for months in untreated cases.

• Lymphadenopathy completing diagnostic triad with cytologic implications: Generalized lymphadenopathy occurs in approximately 75% of DRESS cases, most commonly involving cervical, axillary, and inguinal nodes. Nodes are typically moderately enlarged (1-3 cm), firm, and non-tender, representing reactive lymphoid proliferation rather than malignant infiltration. Lymph node biopsy reveals reactive follicular hyperplasia with variable paracortical expansion, contrasting with the architectural preservation seen in lymphoma. The degree of lymphadenopathy correlates with peripheral eosinophilia and systemic inflammation severity. This presentation frequently triggers concern for lymphoproliferative malignancy, leading to unnecessary biopsies and delayed DRESS diagnosis.

• Facial edema and periorbital swelling as distinctive clinical sign: Facial puffiness and periorbial edema occur in 50-70% of DRESS cases, contributing to the distinctive "puffy face" appearance sometimes described. This edema reflects both lymphoid hyperplasia in face-draining lymph nodes (submandibular, cervical) and T-cell-mediated inflammation in facial skin and subcutaneous tissues. The edema is typically non-pitting, firm, and symmetric, worse upon waking. This finding, combined with facial erythema and fever, creates a characteristic clinical gestalt that experienced clinicians recognize as potentially DRESS.

• Hematologic abnormalities reflecting systemic immune activation: Peripheral blood evaluation reveals marked lymphocytosis (commonly 4,000-30,000/μL) with atypical lymphocytes on blood smear, created by activated CD8+ and CD4+ T-cell blasts. These atypical lymphocytes are morphologically similar to those seen in infectious mononucleosis or acute leukemia, necessitating careful flow cytometry to exclude malignancy. Eosinophilia is characteristic and often marked (1,500-15,000/μL, occasionally >25,000/μL), driven by IL-5 and IL-13 production and significantly more pronounced than in simple drug exanthems. Monocytosis frequently accompanies lymphocytosis. Mild thrombocytopenia (100,000-150,000/μL) occurs in 20-25% of cases, reflecting immune-mediated platelet consumption or bone marrow suppression. Severe leukopenia occurs rarely but indicates progression toward HLH or bone marrow failure.

• Hepatic dysfunction with transaminitis and synthetic dysfunction in severe cases: Liver involvement occurs in approximately 50-80% of DRESS cases, with variable severity ranging from asymptomatic transaminitis to fulminant hepatic failure. Early DRESS typically presents with ALT/AST elevation to 2-10x upper limit of normal, with ALT>AST (contrasting with alcoholic hepatitis). Bilirubin elevation lags transaminitis, indicating hepatocellular rather than cholestatic pattern. Alkaline phosphatase and GGT elevation occurs but is less pronounced. Coagulopathy (elevated PT/INR) indicates synthetic dysfunction and represents a severe prognostic marker necessitating hospital admission. Hepatic involvement typically peaks 2-4 weeks into systemic disease, though some cases demonstrate delayed progression. Fulminant hepatic failure occurs in

Initial laboratory panel (obtain in every suspected case)

• CBC with manual differential: eosinophilia and atypical lymphocytes are the hematologic signature; RegiSCAR uses an eosinophil threshold in the range of ≥700/µL (or ≥10% of leukocytes).
• Hepatic panel and INR: hepatocellular transaminitis is the commonest visceral finding; a rising INR or bilirubin signals synthetic failure.
• Creatinine, urinalysis with microscopy: sterile pyuria, WBC casts, or eosinophiluria suggest acute interstitial nephritis.
• Troponin and ECG, with echocardiography if abnormal: screens for eosinophilic myocarditis, which is often clinically silent early.
• Ferritin, triglycerides, fibrinogen, soluble IL-2 receptor: obtain when cytopenias or unremitting fever raise concern for HLH.
• Exclusion studies: blood cultures, hepatitis A/B/C serologies, HIV, EBV/CMV, Mycoplasma, and ANA — RegiSCAR awards points only when these are negative.

Named criteria

• RegiSCAR scoring system (the one examiners use): points for fever >38.5°C, lymphadenopathy at ≥2 sites, eosinophilia, atypical lymphocytes, rash >50% body surface area with suggestive morphology, biopsy compatible with DRESS, organ involvement, resolution taking >15 days, and negative alternative workup. Totals stratify to no case, possible, probable, and definite.
• Japanese (J-SCAR) criteria additionally require HHV-6 reactivation for "typical" DRESS; HHV-6/CMV PCR is optional in US practice but supports the diagnosis and predicts severity.

Immediate steps

• Withdraw the culprit drug and all non-essential medications — the single most important intervention; delay in withdrawal correlates with mortality. Substitute a structurally unrelated agent (never cross-substitute among aromatic antiepileptics, which cross-react).
• Stabilize and admit: fluids, electrolytes, thermoregulation, and skin barrier care. Hypotension, coagulopathy, hypoxemia, or troponin elevation mandates ICU-level care.

First-line therapy

• Topical high-potency corticosteroids (e.g., clobetasol) plus emollients and antihistamines for pruritus — sufficient for skin-limited disease with normal organ function.
• Systemic corticosteroids (e.g., prednisone ~0.5–1 mg/kg/day) for visceral involvement: transaminitis, renal injury, pneumonitis, or myocarditis. Because DRESS relapses when steroids are stopped abruptly, the taper is deliberately slow — over roughly 8–12 weeks.

Escalation and second-line

• Pulse IV methylprednisolone for fulminant hepatitis, myocarditis, or HLH.
• Steroid-sparing immunosuppressants: calcineurin inhibitors (cyclosporine) or antimetabolites (mycophenolate); IVIG is reserved for refractory or steroid-contraindicated cases and should not be used alone.
• Antivirals (ganciclovir/valganciclovir) only for documented, symptomatic CMV reactivation with end-organ disease.
• Biologics targeting the IL-5 axis and JAK inhibitors are reported anecdotally, not standard.

Contraindicated / avoid

• Rechallenge with the culprit or a cross-reactive congener — absolute; document the reaction as a hard allergy and counsel first-degree relatives (HLA-linked risk).
• Empiric antibiotics added "for fever" — they muddy causality and add a second potential trigger.
• Prevention: the ACR gout guideline conditionally recommends **HLA-B*58:01 testing before allopurinol in Southeast Asian and African American patients; FDA labeling** requires HLA-B*15:02 screening in patients of Asian ancestry before carbamazepine (SJS/TEN risk), with **HLA-A*31:01** conferring DRESS risk.

Acute, life-threatening (emergencies)

• Fulminant hepatic failure: massive CD8+ T-cell–mediated hepatocyte apoptosis progressing from transaminitis to synthetic failure. Signaled by rising INR, rising bilirubin with falling transaminases, and encephalopathy — the classic ominous "improving AST with worsening INR." Hepatic failure is the leading cause of DRESS death; contact a transplant center early, consistent with AASLD acute liver failure guidance.
• Hemophagocytic lymphohistiocytosis (HLH): IFN-γ/TNF-α–driven macrophage hyperactivation. Signaled by unremitting fever, new cytopenias in a previously eosinophilic patient, splenomegaly, markedly elevated ferritin, hypertriglyceridemia, and hypofibrinogenemia. Apply HLH-2004 criteria; this is an ICU emergency.
• Acute necrotizing eosinophilic myocarditis: eosinophil degranulation (major basic protein) injures myocytes. Signaled by chest pain, tachycardia out of proportion, hypotension, troponin elevation, or new heart failure — obtain ECG, troponin, and echocardiography. May present weeks after the rash and can be fatal.
• Eosinophilic pneumonitis/ARDS: hypoxemia with diffuse infiltrates.

Subacute organ injury

• Acute interstitial nephritis: rising creatinine, sterile pyuria, WBC casts; allopurinol-associated DRESS is the classic setting.
• Colitis, pancreatitis, and CNS involvement (meningoencephalitis), the latter often tied to HHV-6 reactivation.

Delayed autoimmune sequelae (the tested long-tail)

• Autoimmune thyroid disease (Hashimoto or Graves) appearing months after recovery — check TSH at follow-up.
• Fulminant type 1 diabetes, alopecia areata, vitiligo, and lupus-like disease.

Treatment-related

• Corticosteroid toxicity: hyperglycemia, hypertension, psychosis, adrenal suppression, and — with prolonged high-dose therapy — Pneumocystis pneumonia and Strongyloides hyperinfection in patients from endemic areas (screen before immunosuppression).
• Viral reactivation unmasked by immunosuppression: CMV colitis or retinitis.
• Relapse/flare on rapid taper, the most common iatrogenic misstep.
• Cyclosporine: nephrotoxicity and hypertension; IVIG: thrombosis and aseptic meningitis.

• Latency is the discriminator: DRESS begins 2–8 weeks after drug initiation. A simple morbilliform exanthem appears within days to ~2 weeks; SJS/TEN within ~1–3 weeks; AGEP within 1–3 days. A stem that says "rash three days after starting the drug" is steering you away from DRESS.
• The triad plus two: fever + morbilliform rash + lymphadenopathy, with eosinophilia and atypical lymphocytes on the differential and facial/periorbital edema. Eosinophilia in a drug rash is DRESS until proven otherwise — simple exanthems do not produce marked eosinophilia.
• Single best next step in a suspected case: stop the culprit drug and send CBC with differential, LFTs, creatinine with urinalysis. Do not order a skin biopsy first — it is supportive, never diagnostic.
• The scoring system to name: RegiSCAR. The virus to name: HHV-6 reactivation (also EBV, CMV), which explains paradoxical worsening after the drug has been stopped.
• The pharmacogenetic association examiners love: **HLA-B*58:01 and allopurinol; HLA-A*31:01 and carbamazepine-associated DRESS. Do not confuse these with HLA-B*15:02, which predicts carbamazepine SJS/TEN** in patients of Asian ancestry, not DRESS.
• Mucosa and skin detachment argue against DRESS: prominent erosions at two or more mucosal sites with a positive Nikolsky sign and full-thickness epidermal necrosis is SJS/TEN. DRESS mucosal involvement, when present, is mild.
• Taper slowly: relapse on abrupt steroid withdrawal is the classic management error; plan weeks-to-months, not days.
• Common distractor: the amoxicillin-induced morbilliform rash of EBV mononucleosis — benign, self-limited, and not a true penicillin allergy or DRESS. Also do not label an isolated late-onset maculopapular exanthem with normal labs as DRESS; without fever, eosinophilia, or organ involvement it is a simple exanthem requiring only drug withdrawal and supportive care.
• Follow-up pearl: check TSH months after recovery — delayed autoimmune thyroiditis is a recognized sequela.