Serum sickness and serum-sickness-like reactions — branching viva

Viva format

The examiner takes the candidate through four escalating branches, each anchored to a patient. At each stage the candidate must state the reasoning before the action, name the mechanism, and respect the true-serum-sickness versus SSLR split. The examiner probes the why and the corners. [5]

Branch 1 — Cefaclor SSLR in a young child

Examiner: A four-year-old boy presents with a moving, blotchy rash, low-grade fever and stiff, swollen knees. He finished a cefaclor course for an ear infection eight days ago. He looks miserable but not toxic. How do you manage this? [8]

The model answer confirms the serum-sickness-like reaction from the latency and the triad — six to twenty-one days after a drug, with cefaclor the paradigmatic paediatric cause. The first step is to exclude the dangerous mimics (anaphylaxis, SJS/TEN, DRESS) by morphology and tempo. The complement is expected to be normal. Management is to stop and document the cefaclor, give a weight-based non-sedating antihistamine, and reserve a short course of oral corticosteroid for severe arthralgia or rash. The condition is self-limiting over days to weeks. [8]

Examiner probe: "What does the complement do in this condition, and why?" [5]

The candidate explains that the SSLR is not an immune-complex reaction — the proposed mechanism is a reactive metabolite or T-cell/histamine-mediated mimicry — so complement is not consumed and C3, C4 and CH50 remain normal. This is the laboratory feature that distinguishes the SSLR from true immune-complex serum sickness, where the complement falls. [5] [8]

Branch 2 — The viral confounder and the avoidance decision

Examiner: His mother insists he is now allergic to penicillin for life and wants a MedicAlert. What do you say? [1]

The model answer resists a blanket lifelong label. The modern paediatric evidence — the systematic review, the avoidance paper and the graded oral challenge data — supports an individualised approach: a substantial proportion of children labelled with a beta-lactam SSLR tolerate the drug on graded challenge, confirming that many labels reflect a coincidental viral illness rather than true drug hypersensitivity. The candidate documents the reaction, avoids cefaclor and closely related drugs for now, and arranges a specialist drug-allergy review with a graded oral challenge if a beta-lactam is clinically important in the future. The family is also told that this was serum sickness, not anaphylaxis or SJS, so the danger is not acute re-exposure. [1] [3]

Examiner probe: "How does a graded oral challenge actually separate an SSLR from a viral exanthem?" [3]

The candidate explains that a graded challenge gives incremental supervised doses of the suspected beta-lactam; tolerance confirms that the original episode was not a true drug reaction, removing the label, while reproducible symptoms confirm a true SSLR and inform a tailored avoidance. The challenge is specialist-led and reserved for cases in which future use of the drug class matters. [3] [13]

Branch 3 — Excluding the dangerous mimics

Examiner: Now picture a different child: a fever and rash two weeks into a new anticonvulsant, with target lesions and painful skin. Is this serum sickness? [12]

The model answer says no. Target lesions, skin pain and mucosal involvement point to Stevens-Johnson syndrome or toxic epidermal necrolysis, not serum sickness. DRESS is the other mimic to exclude — eosinophilia, transaminitis, onset two to eight weeks after the drug, and systemic illness. Anaphylaxis is excluded by the latency (it arrives within the hour) and the morphology. The candidate stops all suspect drugs, admits, and involves dermatology or burns; these conditions are never re-challenged. The point is that the serum-sickness label must not be applied until the dangerous drug reactions are excluded. [12] [13]

Examiner probe: "What single feature most reliably separates serum sickness from SJS/TEN?" [12]

The candidate names mucosal involvement and skin pain with epidermal detachment as the SJS/TEN signature; serum sickness has a polymorphic rash without mucosal blistering or skin necrosis, and it is self-limiting rather than life-threatening. The latency overlaps, so morphology — not timing — is the separator here. [12]

Branch 4 — True serum sickness after rituximab

Examiner: Finally, a fourteen-year-old on rituximab for an autoimmune disease presents ten days after her second infusion with fever, an annular rash, swollen knees and tender lymph nodes. What is going on, and what is the single most important test? [11]

The model answer recognises true immune-complex serum sickness. The trigger is a foreign-protein biologic, the autoimmune indication raises the risk relative to haematological malignancy, and the latency fits. The single most important test is the complement panel — C3, C4 and CH50 are expected to be low because the immune complexes consume complement. The candidate also checks urinalysis and renal function for nephritis, and looks for proteinuria, eosinophilia and serositis. [11] [5]

Examiner probe: "Do you re-challenge rituximab, and why is the answer different from the cefaclor case?" [11]

The candidate states that rituximab is not re-challenged, because recurrence of immune-complex serum sickness can be faster and more severe on repeat exposure. The answer differs from the cefaclor SSLR because the mechanism and the risk differ: true serum sickness is a complement-consuming immune-complex hypersensitivity to a foreign protein, whereas the SSLR is a normal-complement, non-immune-complex mimic of a small-molecule drug that allows an individualised, challenge-informed avoidance. The candidate coordinates re-treatment of the autoimmune disease with an alternative agent and documents the non-re-challenge status. [11] [5]

Closing synthesis

A passing candidate keeps four ideas in tension throughout: the latency of six to twenty-one days is the diagnostic signature and rules out immediate allergy; the complement level is the single most useful discriminator between true serum sickness (low) and the SSLR (normal); the dangerous mimics (anaphylaxis, SJS/TEN, DRESS) must be excluded by morphology before the serum-sickness label is applied; and the avoidance decision is mechanism-dependent — a true biologic serum sickness is not re-challenged, whereas a drug-induced SSLR supports an individualised, challenge-informed approach rather than a blanket lifelong label. [1] [5] [11]