What medications can cause Stevens-Johnson syndrome (SJS)?

Medications That Cause Stevens-Johnson Syndrome

The highest-risk medications for Stevens-Johnson syndrome include allopurinol, carbamazepine, lamotrigine, phenytoin, phenobarbital, sulfonamide antibiotics, sulfasalazine, nevirapine, and oxicam NSAIDs, all of which should be discontinued immediately at the first sign of rash. 1

Highest-Risk Medications (Immediate Discontinuation Required)

Anticonvulsants

• Carbamazepine carries a strong association with SJS/TEN, particularly in patients with HLA-B*1502 allele (>90% of cases occur within first few months of treatment) 2
• Phenytoin can cause SJS/TEN and should be discontinued if any rash appears; the risk is increased in black patients and those with HLA-B*1502 allele 3
• Lamotrigine is among the most common causes requiring immediate withdrawal 1
• Phenobarbital shows a crude relative risk of 45 during the first two months of treatment 4
• Valproic acid demonstrates a relative risk of 25 in the initial treatment period 4

Anti-Infective Agents

• Sulfonamide antibiotics (particularly trimethoprim-sulfamethoxazole) have the highest relative risk at 172, making them the most dangerous class 4
• Sulfasalazine is listed among the most common SJS causes 1
• Aminopenicillins carry a multivariate relative risk of 6.7 4
• Quinolones show a multivariate relative risk of 10 4
• Cephalosporins demonstrate a multivariate relative risk of 14 4

Antiretroviral Agents

• Nevirapine is among the most common causes and can progress to severe hepatotoxicity; it shows early-onset rash that can rapidly evolve to SJS 5, 1
• Efavirenz is associated with early-onset rash that can rarely progress to SJS, making differentiation from acute seroconversion difficult 5

NSAIDs and Other Agents

• Oxicam NSAIDs (meloxicam, piroxicam) are most commonly associated with SJS/TEN among NSAIDs, with a crude relative risk of 72 during initial treatment 5, 4
• Allopurinol shows a relative risk of 52 in the first two months, requiring immediate discontinuation at first sign of rash 1, 4
• Chlormezanone demonstrates a crude relative risk of 62 4

Important Clinical Caveats

Timing Considerations

• For drugs used chronically (anticonvulsants, allopurinol, NSAIDs), the increased risk is confined largely to the first two months of treatment 4
• Over 90% of carbamazepine-related SJS/TEN cases occur within the first few months 2

Pediatric Populations

• Anticonvulsants and antibiotics are the primary implicated medications in children 1
• Ibuprofen has been associated with higher complication risk in pediatric SJS cases 1
• Important pitfall: NSAIDs (particularly ibuprofen) and acetaminophen may be falsely implicated because they are often started for prodromal fever and mucosal symptoms (protopathic effect) 5

Genetic Risk Factors

• HLA-B*1502 screening should be performed before initiating carbamazepine in patients of Asian ancestry (>15% prevalence in Hong Kong, Thailand, Malaysia, Philippines; ~10% in Taiwan; ~4% in North China) 2
• HLA-A*3101 is moderately associated with hypersensitivity reactions including SJS in European, Korean, and Japanese patients taking carbamazepine (carried by >15% of Japanese, Native American, Southern Indian populations) 2
• Limited evidence suggests HLA-B*1502 may predict SJS/TEN risk with phenytoin in Asian populations 3

Management Algorithm

Immediate Actions

• Document all medications taken over the 2 months prior to symptom onset 1
• Withdraw suspected medication immediately to decrease mortality risk 1
• Never rechallenge with the suspected drug or drugs within the same chemical class for severe reactions 5

Documentation Requirements

• Permanently document the suspected medication in medical records 1
• Report to pharmacovigilance authorities 1
• Ensure the medication is flagged as an allergy in all medical systems

When Causative Drug Cannot Be Identified

• Consider infectious etiologies, particularly Mycoplasma pneumoniae in children (up to 50% of pediatric cases are infection-related) 1
• Herpes simplex virus is another well-documented non-drug cause 6

Cross-Reactivity Warnings

• Within NSAID chemical classes: Cross-reactivity is well described but not universal (e.g., lack of cross-reactivity between ibuprofen and naproxen has been reported for fixed drug eruption) 5
• For severe reactions: Avoidance of the entire chemical class without rechallenge is recommended due to unpredictable recurrence risk 5
• Carbamazepine and phenytoin: Limited evidence suggests HLA-B*1502 may confer cross-risk between these agents in Asian populations 2