Adverse Drug Reactions — Viva Defence

Long Case Viva — DRESS to Allopurinol in a Complex Elderly Man

Examiner's opening question

"You have 90 seconds to present this patient. A 66-year-old man with chronic kidney disease (eGFR 28), gout, hypertension and ischaemic heart disease was started on allopurinol 300 mg daily four weeks ago for recurrent gout flares. He now presents with fever, a morbilliform rash with facial oedema, generalised lymphadenopathy, eosinophils 1.6 times ten to the ninth per litre, ALT 280 U per litre, bilirubin 36 micromol per litre, and creatinine 195 micromol per litre. His medications are amlodipine, atorvastatin, aspirin, frusemide and allopurinol. Present." [1]

Candidate's opening statement (SASPOP format)

"A 66-year-old man with a five-day history of fever, rash and dark urine (symptoms), aged 66 (age), male (sex), presenting acutely unwell to the emergency department (presentation), who is a retired engineer (occupation), with the following problems: (1) Drug Reaction with Eosinophilia and Systemic Symptoms — DRESS — to allopurinol, a severe Type B adverse drug reaction with hepatitis and acute kidney injury; (2) chronic kidney disease stage 4 with an eGFR of 28; (3) gout; (4) ischaemic heart disease on secondary prevention; (5) hypertension; and (6) a prescribing error — allopurinol started at 300 mg in significant renal impairment without dose adjustment, which increased the risk of this reaction. The immediate threats to life are the hepatitis, which may progress to fulminant hepatic failure, and the acute kidney injury. My immediate plan is to stop the allopurinol permanently, admit for observation, and start systemic corticosteroids for the visceral involvement." [1]

Examiner probe 1 — diagnosis and classification

"Defend your diagnosis of DRESS. What is DRESS, and how does it differ from Stevens-Johnson syndrome?" [1]

"DRESS — Drug Reaction with Eosinophilia and Systemic Symptoms — is a delayed, immune-mediated, multi-organ hypersensitivity reaction that typically occurs 2 to 8 weeks after first exposure to a culprit drug. The culprits are the aromatic anticonvulsants (phenytoin, carbamazepine, lamotrigine, phenobarbital), allopurinol, minocycline, the sulfonamide antibiotics, dapsone and abacavir. The clinical picture is a triad of fever, a morbilliform rash with facial oedema, and internal organ involvement — most often the liver but also the kidney, lung and heart — accompanied by a marked eosinophilia and generalised lymphadenopathy. This patient has the full picture: the latency of four weeks, the fever, the morbilliform rash with facial oedema, the lymphadenopathy, the eosinophilia of 1.6 times ten to the ninth, and the hepatitis and acute kidney injury. On the RegiSCAR scoring system this would be probable or definite DRESS. [1]

The distinction from Stevens-Johnson syndrome is the pattern of the skin and mucosa. SJS and TEN present with painful blistering, target lesions, flaccid bullae, sheet-like epidermal detachment and mucosal involvement of at least two sites — the patient is in pain, the skin sloughs, and the Nikolsky sign is positive. DRESS presents with a morbilliform rash and facial oedema, with no blistering, no detachment and a negative Nikolsky sign, but with eosinophilia and visceral involvement. The two share some triggers — allopurinol can cause either — but the clinical signatures are distinct. The classification matters because the management and the prognosis differ: SJS/TEN has a mortality of 10 to 30 per cent and is managed in a burns unit, while DRESS has a mortality of 5 to 10 per cent and is managed with corticosteroids." [1]

Examiner probe 2 — causality and the prescribing error

"How confident are you that allopurinol caused this? And why was the dose an error?" [1]

"I would apply the Naranjo Adverse Drug Reaction Probability Scale. There are previous conclusive reports of allopurinol-induced DRESS (plus 1); the reaction appeared after the drug was started (plus 1); the reaction will improve when the drug is stopped (plus 1); rechallenge is ethically barred in a severe Type B reaction so cannot be tested; there is no convincing alternative cause — the eosinophilia, hepatitis and timing are all consistent with the drug (plus 1 for no alternative cause); the reaction was more severe because the dose was too high for the renal function (plus 1); and there is objective evidence in the eosinophilia and transaminitis (plus 1). The total is in the probable range. In practice, causality in severe Type B reactions rests on the temporal relationship, dechallenge, and exclusion of alternatives, because rechallenge is not done. [1]

The dose was an error because allopurinol's active metabolite oxypurinol is renally cleared, and the dose must be reduced in renal impairment. In a patient with an eGFR of 28, the starting dose should have been 50 to 100 mg daily, titrated upward against the serum urate — not 300 mg. Starting at 300 mg produced a higher oxypurinol exposure and increased the probability of the Type B hypersensitivity reaction. This is an example of a Type A prescribing error (a dose-related mistake) that increased the likelihood of a Type B reaction. In addition, the HLA-B*5801 allele, prevalent in Han Chinese, Thai and Korean populations, strongly predicts allopurinol-induced SJS/TEN and DRESS, and screening is increasingly standard of care in those populations in ANZ practice." [1]

Examiner probe 3 — immediate management

"Take me through your immediate management in detail." [1]

"First, I stop the allopurinol immediately and permanently, and document the reaction in the medication record and the adverse-reaction field, naming the drug, the reaction, the date and the severity. The patient must never take allopurinol again. [1]

Second, I admit him for observation and supportive care, with daily liver function tests and renal function for the first week, because the hepatitis can progress to fulminant hepatic failure and the acute kidney injury may worsen. I monitor for a falling ALT with a rising INR and encephalopathy, which would signal decompensation and trigger a liver transplantation referral per the King's College criteria. [1]

Third, because he has visceral involvement — hepatitis and acute kidney injury — I start systemic corticosteroids. I would use oral prednisolone 0.5 to 1 mg per kg per day, approximately 50 mg per day, and plan a slow taper over weeks to months to prevent the relapses that characterise DRESS. If the hepatitis were fulminant or he were haemodynamically unstable, I would use intravenous methylprednisolone. [1]

Fourth, I exclude alternative causes — viral hepatitis A, B, C and E serology, EBV and CMV, HIV, an autoimmune hepatitis screen (ANA, smooth muscle antibody, immunoglobulins), and an abdominal ultrasound to exclude biliary obstruction. I hold the atorvastatin temporarily because it can contribute to transaminitis and rarely to DILI, and I resume it only when the DRESS attribution to allopurinol is secure. [1]

Fifth, I will report this reaction to the Therapeutic Goods Administration via the Blue Card, because it is a severe reaction to a marketed drug, and I will tell the patient I have done so." [1]

Examiner probe 4 — follow-up and the long view

"What are the long-term issues, and how will you manage his gout going forward?" [1]

"DRESS can relapse for weeks after the drug is stopped, partly via human herpesvirus 6 reactivation, so the corticosteroid taper must be slow and supervised. Long-term autoimmune sequelae are described — autoimmune hepatitis, type 1 diabetes and autoimmune thyroiditis — and I would arrange follow-up with clinical immunology or dermatology for the taper and for surveillance for these sequelae over the following months. [1]

For his gout, allopurinol is now contraindicated permanently. The alternatives are colchicine for acute flares (with dose adjustment for the renal function), a uricosuric such as probenecid (which is less effective in significant renal impairment), or pegloticase infusions for severe tophaceous gout. I would involve a rheumatologist. The broader lesson is a systems-level one: the prescribing error that started allopurinol at 300 mg in renal impairment should be reviewed with the team, and I would reinforce the practice of dose adjustment for renal function and the consideration of HLA-B*5801 screening in at-risk populations." [1]

Examiner probe 5 — the ABCDEF classification

"Give me the ABCDEF classification with one example each." [1]

"Augmented — Type A — an exaggerated normal pharmacology, dose-dependent and common; example, beta-blocker bradycardia. Bizarre — Type B — idiosyncratic, dose-independent, rare, high mortality; example, DRESS, SJS/TEN, anaphylaxis. Chronic — Type C — cumulative dose over time; example, bisphosphonate osteonecrosis of the jaw. Delayed — Type D — time-related, including teratogenesis and carcinogenesis; example, thalidomide phocomelia, diethylstilboestrol and clear-cell vaginal carcinoma. End-of-use — Type E — a withdrawal syndrome; example, adrenal suppression after long-term steroids. Failure — Type F — unexpected therapeutic failure; example, oral contraceptive failure with rifampicin." [1]

Examiner probe 6 — a high-yield discriminating question

"A patient on long-term prednisolone for polymyalgia rheumatica is admitted hypotensive, hyponatraemic and hyperkalaemic after a gastrointestinal illness. What is the diagnosis and what do you do?" [1]

"This is adrenal suppression — a Type E withdrawal reaction — unmasked by the stress of the gastrointestinal illness. The long-term exogenous steroid has suppressed the hypothalamic-pituitary-adrenal axis, and the patient cannot mount a cortisol response to the illness. I would give empirical intravenous hydrocortisone 100 mg and intravenous fluid resuscitation immediately, without waiting for a cortisol level — adrenal crisis is fatal if untreated and the empirical steroid is safe. I would send a 9 am cortisol and a short synacthen test before the first dose if it can be obtained immediately, but treatment must not wait. The HPA axis may take months to a year to recover, and the patient will need a slow supervised steroid taper." [1]

Short Case Viva — Skin Examination of a Patient with a Drug Eruption

Examiner's instruction

"Examine this patient's skin. The instruction is: examine the skin and discuss your findings." [1]

Candidate's examination routine

"I introduce myself, confirm the patient's identity and consent, position the patient comfortably with adequate lighting and privacy, and ask whether there is any pain. I examine the skin systematically from head to toe, including the mucous membranes, the nails, the palms and soles, and the dependent areas. I am looking for: the distribution and morphology of any rash; the presence of target lesions, vesicles, bullae or epidermal detachment; mucosal involvement (oral, conjunctival, urogenital); lymphadenopathy; and the features of systemic involvement such as jaundice or fever. I test the Nikolsky sign by applying gentle lateral pressure to apparently normal skin adjacent to a blister. I estimate the percentage of body surface area involved using the rule of nines. I examine the relevant systems for organ involvement — the chest for pneumonitis, the abdomen for hepatomegaly, and the vital signs for haemodynamic stability." [1]

Key physical signs and presentation

"Sir, this patient has a confluent morbilliform rash over the trunk and face with marked facial oedema, generalised lymphadenopathy, and no blistering, no mucosal involvement, and a negative Nikolsky sign. There is no epidermal detachment. The patient is febrile. The rash spares the palms and soles. These findings, in a patient started on a new drug 2 to 8 weeks ago, with eosinophilia and hepatitis on the blood tests, are those of Drug Reaction with Eosinophilia and Systemic Symptoms (DRESS). I would distinguish this from Stevens-Johnson syndrome, which presents with painful blistering, target lesions, flaccid bullae, sheet-like epidermal detachment with a positive Nikolsky sign, and mucosal involvement of at least two sites. The two reactions share some triggers — allopurinol can cause either — but the clinical signatures are distinct." [1]

Discussion questions

Examiner: How would you classify this reaction and what does that mean for management? [1]

"This is a Type B — bizarre — adverse drug reaction in the Rawlins and Thompson ABCDEF classification. It is idiosyncratic, dose-independent, rare, and carries a high mortality when severe. The management implication is that the offending drug must be stopped permanently and never re-challenged. This contrasts with a Type A reaction, which is an exaggerated normal pharmacology and is managed by dose reduction or temporary cessation." [1]

Examiner: How do you assess causality? [1]

"I would use the Naranjo Adverse Drug Reaction Probability Scale, a 10-item questionnaire that scores the temporal relationship, dechallenge, rechallenge, the presence of an alternative cause, and objective evidence. A score above 9 is definite, 5 to 8 is probable, 1 to 4 is possible, and 0 or less is doubtful. Naranjo assesses causality, not severity — severity is a separate axis of mild, moderate or severe. In severe Type B reactions rechallenge is ethically barred, so causality usually rests on the temporal relationship, dechallenge, and exclusion of alternatives." [1]

Examiner: When would you calculate a SCORTEN, and what are the components? [1]

"I would calculate SCORTEN in any patient with suspected Stevens-Johnson syndrome or toxic epidermal necrolysis, at admission and again at 48 hours. The seven components, each scoring one point, are: age above 40 years, the presence of malignancy, heart rate above 120 beats per minute, initial detached body surface area above 10 per cent, serum urea above 10 mmol per litre, serum glucose above 14 mmol per litre, and serum bicarbonate below 20 mmol per litre. A SCORTEN of 0 to 1 predicts a mortality of about 3 per cent; 2 about 12 per cent; 3 about 35 per cent; 4 about 58 per cent; and 5 or more above 90 per cent. The score guides the level of care — a high score mandates transfer to an ICU or specialist burns unit." [1]

Examiner: What is Hy's Law, and why does it matter? [1]

"Hy's Law predicts the lethality of a drug-induced hepatocellular injury. It is present when ALT or AST is more than three times the upper limit of normal, serum bilirubin is more than two times the upper limit of normal, there is no significant cholestasis (ALP less than two times the upper limit of normal), and there is no alternative cause. Its presence carries a case fatality of around 10 per cent from acute liver failure, and it is used by the FDA and EMA as a stopping rule in drug development and post-marketing surveillance. Any drug that produces a Hy's Law case must be presumed to be capable of causing fatal hepatotoxicity until proven otherwise." [1]

Examiner: How would you report this reaction? [1]

"I would report every severe or previously unrecognised adverse drug reaction to the Therapeutic Goods Administration via the Blue Card in Australia, or to the MHRA Yellow Card scheme in the United Kingdom. I would report all suspected reactions to new (black triangle) drugs and all serious reactions regardless of how long the drug has been marketed. A reaction is serious if it is fatal, life-threatening, requires or prolongs hospitalisation, causes persistent or significant disability, or is a congenital anomaly. Reporting is voluntary — a suspicion is enough; I do not need certainty of causality. Pharmacovigilance depends on clinician reporting." [1]