Erythema Multiforme, Stevens-Johnson, TEN

USMLE Step 1 trap: Incorrectly places EM on the same disease spectrum as SJS/TEN rather than recognizing them as distinct entities with different triggers. Erythema multiforme is a distinct entity most commonly triggered by HSV and is generally self-limited; SJS and TEN are drug-induced reactions that form their own severity spectrum based on BSA involvement.

Erythema multiforme (EM), Stevens-Johnson syndrome (SJS), and toxic epidermal necrolysis (TEN) are three distinct but frequently confused hypersensitivity reactions involving the skin and mucous membranes. The biggest conceptual trap on USMLE Step 1 is treating all three as a single severity spectrum — they're not. EM is its own disease, triggered predominantly by HSV and mycoplasma, characterized by classic target lesions, and usually self-limited. SJS and TEN are drug-induced reactions that form their own spectrum, separated by how much skin is sloughing off.

The exam tests this at multiple levels. Straightforward questions will show you a target lesion and ask about the trigger (HSV) or show a patient on carbamazepine with skin blistering and mucosal erosions and ask what's happening (SJS). Harder questions will give you a vignette with a patient of Han Chinese descent on a new anticonvulsant and ask which HLA allele predicts risk — that's HLA-B*1502 and carbamazepine. Application questions may give you a patient with 15% BSA epidermal detachment and ask how to classify it (SJS-TEN overlap, 10–30% range). USMLE Step 1 also tests management, and the steroid question is a classic trap.

What makes this tricky is that students memorize 'immune-mediated = steroids' as a reflex. For TEN, that reflex will get you the wrong answer. The exam also exploits the assumption that mucosal involvement is what separates SJS from TEN — it doesn't. Both involve mucosa. The separator is BSA of epidermal detachment. Get those two facts locked in, and most of the hard questions become straightforward.

From real student decks

51%have cards covering this topic

30%have mature cards

Common misconceptions

Common mistake

Wrong: Erythema multiforme, Stevens-Johnson syndrome, and TEN are a continuous severity spectrum of the same disease.

Right: Erythema multiforme is a distinct entity most commonly triggered by HSV and is generally self-limited; SJS and TEN are drug-induced reactions that form their own severity spectrum based on BSA involvement.

EM and SJS look superficially similar because both cause skin lesions and can involve mucosa, but they have different triggers, pathophysiology, and prognoses. EM is predominantly triggered by infections — especially HSV — and is generally self-limited; SJS and TEN are almost always drug-induced and carry significant mortality. Placing EM on the SJS/TEN spectrum leads to wrong answers about triggers and management, so treat them as separate diseases that happen to both involve skin.

Common mistake

Wrong: SJS and TEN are distinguished by mucosal involvement rather than percentage of BSA with epidermal detachment.

Right: SJS involves <10% BSA epidermal detachment, SJS-TEN overlap is 10–30%, and TEN is >30% BSA; both SJS and TEN involve mucosal surfaces.

Both SJS and TEN cause mucosal erosions — oral, ocular, and genital involvement are expected in both conditions — so mucosal involvement is not the distinguishing feature. The classification is entirely based on percentage of body surface area with full-thickness epidermal detachment: <10% is SJS, 10–30% is the overlap syndrome, and >30% is TEN. When a vignette gives you a percentage, that's your signal to use the BSA cutoffs, not the mucosal findings.

Common mistake

Gap: Misses the high-yield drug triggers and HLA association for SJS/TEN

The most common drug triggers for SJS/TEN are sulfonamides, allopurinol, anticonvulsants (lamotrigine, carbamazepine, phenytoin), and NSAIDs; carbamazepine-induced SJS has a strong HLA-B*1502 association in Han Chinese populations.

The highest-yield drug triggers for SJS/TEN are sulfonamide antibiotics, allopurinol, anticonvulsants (especially lamotrigine and carbamazepine), and NSAIDs — these appear repeatedly on USMLE Step 1 vignettes. The HLA-B*1502 association with carbamazepine is particularly testable because it illustrates pharmacogenomics and explains why screening is recommended in patients of Han Chinese or other Southeast Asian ancestry before starting carbamazepine. Allopurinol-induced SJS/TEN has a separate HLA association (HLA-B*5801) that is lower yield but worth recognizing.

Common mistake

Wrong: Systemic corticosteroids are the standard of care for TEN because it is an immune-mediated disease.

Right: Systemic corticosteroids are controversial in TEN and may worsen outcomes by increasing infection risk; supportive care in a burn unit, IVIg, and cyclosporine are preferred management strategies.

The instinct to use corticosteroids in TEN because it's immune-mediated is logical but wrong in practice. Corticosteroids suppress the immune response but also impair skin barrier recovery and dramatically increase the risk of sepsis — the leading cause of death in TEN. The preferred approach is aggressive supportive care (fluids, wound care, nutrition) in a burn unit, with IVIg and cyclosporine as adjuncts; steroids are at best controversial and at worst harmful in established TEN.

Free Deck audit

See if your Anki deck covers this topic.

Upload your deck →

Guided session

Stuck on this? An AI tutor that probes your understanding.

Start a session →

What the exam tests

Identify the classic target lesion morphology of EM (three concentric zones: central dusky/necrotic, surrounding pale ring, outer erythematous halo) and recognize the most common triggers — HSV infection and Mycoplasma pneumoniae.
Know which drug classes most commonly cause SJS/TEN — sulfonamides, allopurinol, anticonvulsants (lamotrigine, carbamazepine, phenytoin), and NSAIDs — and know that carbamazepine-induced SJS has a strong pharmacogenomic association with HLA-B*1502, particularly in Han Chinese populations.
Distinguish SJS from TEN using BSA epidermal detachment cutoffs: SJS is <10%, SJS-TEN overlap is 10–30%, and TEN is >30%; recognize that mucosal involvement occurs in both SJS and TEN and cannot be used to differentiate them.
Apply the correct management approach for SJS/TEN: immediate drug withdrawal, supportive care in a burn unit setting, and consideration of IVIg or cyclosporine; recognize that systemic corticosteroids are controversial and potentially harmful in TEN due to increased infection risk.

Can you avoid these mistakes?

A 24-year-old man presents with three-zoned 'target' skin lesions on his palms and forearms that appeared one week after a cold sore outbreak. He has no mucosal involvement and no new medications. What is the diagnosis, what is the causative trigger, and what is the expected clinical course?

A 45-year-old Han Chinese woman with newly diagnosed trigeminal neuralgia is being started on carbamazepine. Before prescribing, what genetic test should you consider, what does it predict, and what alternative drug might be used if she tests positive?

A 60-year-old man on allopurinol for gout presents with fever, painful erosions of the oral mucosa and conjunctiva, and blistering skin lesions covering approximately 12% of his body surface area with epidermal detachment. How do you classify this condition, what is the first step in management, and why are systemic corticosteroids not the answer?

You are given three patients: one with <10% BSA epidermal detachment and mucosal erosions, one with 25% BSA detachment and mucosal erosions, and one with 40% BSA detachment and mucosal erosions — all drug-induced. Classify each and explain why mucosal involvement alone cannot be used to distinguish between these diagnoses.

Erythema Multiforme, Stevens-Johnson, TEN

Common misconceptions

What the exam tests

Can you avoid these mistakes?

Related topics