Drug eruptions including DRESS and AGEP — branching viva

Examiner opening (Examiner)

You are the general paediatric registrar on the ward. A seven-year-old boy who started phenytoin for new-onset epilepsy six weeks ago presents with a fever of 39 degrees, a widespread red rash, facial swelling, enlarged cervical lymph nodes, and blood tests showing eosinophilia and a hepatitis with raised transaminases. Talk me through your assessment and your working diagnosis. [2]

Exemplar opening (Candidate)

This is DRESS — drug reaction with eosinophilia and systemic symptoms, also called drug-induced hypersensitivity syndrome — until proven otherwise, caused by the phenytoin he started six weeks ago. The combination of the long latency of two to eight weeks, the fever, the facial oedema, the lymphadenopathy, the eosinophilia and the hepatitis is the classic presentation of a severe cutaneous adverse reaction to an aromatic anticonvulsant. My immediate priorities are to stop the phenytoin, to assess his airway breathing and circulation, to exclude sepsis with cultures, and to confirm the pattern of organ involvement, because this is not a simple drug allergy — it carries up to ten per cent mortality and can relapse. [2]

Branch 1 — why this is not a simple morbilliform rash (Examiner)

What features distinguish this from a common morbilliform drug eruption, and why does the distinction matter? [1]

Exemplar (Candidate)

A morbilliform eruption is common and benign: it appears five to fourteen days after a drug, produces a fine blanching maculopapular rash from trunk to limbs, and has no significant systemic upset. This boy's reaction differs in three ways that each mark severity. First, the latency is six weeks, well beyond the five to fourteen day window of a benign exanthem and squarely in the two to eight week window of DRESS. Second, he has striking systemic features — fever, facial oedema and lymphadenopathy — that a morbilliform eruption lacks. Third, the eosinophilia and hepatitis signal multi-organ involvement, which is the hallmark of DRESS and the source of its mortality. The distinction matters because a morbilliform eruption is managed with drug cessation and an antihistamine, while DRESS demands drug cessation plus systemic corticosteroid, organ monitoring and specialist input. [1]

Branch 2 — the mechanism (Examiner)

Explain the pathophysiology of DRESS, and tell me what is special about the aromatic anticonvulsants. [2]

Exemplar (Candidate)

DRESS is a T-cell mediated type IV hypersensitivity reaction amplified by herpes-virus reactivation. The aromatic anticonvulsants — phenytoin, carbamazepine and phenobarbital — are oxidised by the cytochrome P450 system to a reactive arene-oxide metabolite that is normally detoxified by the enzyme epoxide hydrolase. A child with an inherited or acquired defect in this pathway cannot clear the metabolite, which accumulates and triggers a polyclonal drug-specific T-cell activation. In parallel, a latent human herpesvirus, most often HHV-6 but also cytomegalovirus and Epstein-Barr virus, reactivates and drives a self-perpetuating inflammatory cascade: eosinophils are recruited and activated, atypical lymphocytes appear, and the inflammation spills from the skin into the liver, the kidneys and other organs. The viral reactivation explains the long latency, the eosinophilia, the relapsing course and the high mortality. [2]

Branch 3 — investigations and diagnostic criteria (Examiner)

What investigations would you arrange, and how do you confirm the diagnosis? [3]

Exemplar (Candidate)

I would arrange a full blood count for eosinophilia and atypical lymphocytes, liver function tests and coagulation for the hepatitis, renal function, inflammatory markers, and blood cultures to exclude sepsis. I would send herpes-virus serology for HHV-6, CMV and EBV, because a rising HHV-6 viral load over the first weeks supports the diagnosis and predicts the relapsing course. I would apply the RegiSCAR validation score, which standardises the diagnosis by weighting fever, lymphadenopathy, eosinophilia, atypical lymphocytes, the extent and morphology of the skin rash, biopsy findings, organ involvement and the time to resolution, and stratifies each case as possible, probable or definite DRESS. No single acute test confirms DRESS — the pattern, the latency, the culprit drug and the eosinophilic hepatitis make the diagnosis. [3]

Branch 4 — management (Examiner)

How do you manage him, and what is your corticosteroid strategy? [3]

Exemplar (Candidate)

The first step is to stop the phenytoin immediately and to review all his medications. I assess his airway breathing and circulation, exclude infection with cultures taken before I start corticosteroid where possible, and support the failing organs with close monitoring of his liver function, coagulation and renal function. I treat him with systemic corticosteroid — oral prednisolone at one milligram per kilogram per day — to suppress the inflammation, and I continue it until the fever, the rash and the organ dysfunction settle. I then taper the corticosteroid slowly over weeks to months, because the herpes-virus-driven relapse is common, dangerous and more severe than the original presentation if the steroid is stopped abruptly. I involve dermatology, allergy and, if he deteriorates, intensive care. [3]

Branch 5 — future anticonvulsant choice (Examiner)

He needs ongoing epilepsy treatment. What anticonvulsant would you choose, and why? [2]

Exemplar (Candidate)

I would avoid all aromatic anticonvulsants — phenytoin, carbamazepine and phenobarbital, and the related oxcarbazepine and lamotrigine — because they share the arene-oxide metabolite pathway and a child who has reacted to one may react to the others. I would choose a non-aromatic alternative such as sodium valproate, levetiracetam or topiramate, in partnership with his neurologist, and I would document the phenytoin DRESS clearly in his record and on an allergy card. I would counsel the family that the reaction is serious and lifelong, that he must avoid the aromatic anticonvulsants in future, and that they should inform every clinician of the allergy. [2]

Branch 6 — a different scenario (Examiner)

A four-year-old girl started amoxicillin yesterday and today has hundreds of tiny pinhead spots over a red swollen skin with fever and no mucosal involvement. What is this and how does it differ from your first patient's DRESS? [4]

Exemplar (Candidate)

This is acute generalised exanthematous pustulosis, or AGEP, triggered by the amoxicillin. The rapid onset within a day, the fever, and the hundreds of sterile non-follicular pinhead pustules on an erythematous oedematous skin make the diagnosis. It differs from DRESS in every key dimension: the latency is hours to a day rather than two to eight weeks, the morphology is pustular rather than morbilliform and exfoliative, the cell is the neutrophil driven by interleukin-8 rather than the eosinophil driven by herpes-virus reactivation, the systemic involvement is mild rather than multi-organ, and the prognosis is usually good with self-resolution in one to two weeks. I manage it supportively — stop the drug, antipyretics, fluids, a topical corticosteroid for comfort — and I exclude SJS-TEN by confirming the absence of target lesions, mucosal erosions and a positive Nikolsky sign. [4]

Examiner wrap-up (Examiner)

Thank you. Summarise the three points you most want the examiner to remember. [2]

Exemplar (Candidate)

First, timing names the reaction — morbilliform at five to fourteen days, AGEP within hours to a day, DRESS at two to eight weeks — and the latency is the single strongest clue. Second, DRESS is the delayed, eosinophilic, multi-organ syndrome with hepatitis and HHV-6 reactivation, carrying up to ten per cent mortality, and it is managed by stopping the drug and giving systemic corticosteroid tapered slowly over weeks to months, with avoidance of cross-reactive aromatic anticonvulsants in future. Third, the first action in every drug eruption is to stop the drug, and the features of skin pain, mucosal involvement, blistering, facial oedema and eosinophilia with organ dysfunction are the red flags that separate a severe cutaneous adverse reaction from a benign exanthem. [2] [4]