Sweet syndrome

Meet the patient

A 52-year-old woman arrives with a two-day crop of tender, angry red-violet plaques on her arms, face, and neck, a fever of 39 degrees, and a neutrophilia of 16. The plaques look almost vesicular but feel solid on palpation. She had a sore throat a week ago.[1][6]

The two questions that decide her next 48 hours are the two that decide every Sweet case: is this a neutrophilic dermatosis, not a vasculitis or cellulitis? (biopsy answers it) and what lit the fuse? (infection, drug, IBD, or — the one you must not miss — leukaemia). Steroids will melt her plaques within two days; the malignancy screen decides her future.[4][8]

Why Sweet earns its name — and its trap

Sweet syndrome is a hypersensitivity reaction, not an infection and not a vasculitis. Cultures are sterile; antibiotics alone do nothing. The neutrophils pour into the dermis but leave the vessel walls standing — and that single histological fact (vessel walls intact) is the most-tested line in the whole topic.[4][6]

First described by the British dermatologist Robert Douglas Sweet in 1964, the disease is the prototype of the neutrophilic dermatoses — a family sharing a final common pathway of sterile neutrophilic skin infiltration, alongside pyoderma gangrenosum, subcorneal pustular dermatosis, and erythema elevatum diutinum.[5]

The classic trap: every junior reaches for antibiotics, seeing fever, redness, and a neutrophilia. Sweet is culture-negative and worsens without corticosteroids. A biopsy before the first steroid dose excludes cellulitis, vasculitis, and leukaemia cutis in one cut.[4]

Three subtypes — and the malignancy screen that never stops

Recognising the subtype is not academic — it dictates the workup and the management. Treat the trigger, not just the skin.[1]

[1]

The number rule for Sweet: GAMPS — G-CSF is the number one drug, AML the number one malignancy, MDS the number two (and the hallmark of histiocytoid Sweet), Prednisolone the first-line drug, Streptococcus the classical infection trigger.[1]

The cytokine cascade — and why G-CSF is the number-one drug

Sweet is a runaway innate immune response. A distant trigger activates keratinocytes, dendritic cells, macrophages, and marrow myeloid precursors to release IL-1, IL-6, IL-8 (CXCL8), G-CSF, GM-CSF, and TNF-alpha. G-CSF and GM-CSF expand the neutrophil pool; IL-8 — the most potent neutrophil chemoattractant known — drags them into the skin.[4][3]

Why G-CSF tops the drug list: granulocyte colony-stimulating factor does exactly what Sweet syndrome does — it expands and activates the neutrophil pool. Exogenous G-CSF (filgrastim, pegfilgrastim) for chemotherapy neutropenia or stem-cell mobilisation can therefore directly precipitate the disease. Endogenous G-CSF is also elevated in classical disease, which is why the biology is so consistent.[3]

Why IL-1 is the master switch: IL-1 sits upstream of the whole cascade, driving IL-6 and IL-8 release. That is the mechanistic basis for IL-1 blockade (anakinra, canakinumab) in refractory disease and in VEXAS — and it is the viva answer to "why does an IL-1 blocker work in a neutrophilic disease?"[2][3]

The paraneoplastic engine: in malignancy-associated disease the tumour itself secretes the cytokines — AML and MDS blasts release G-CSF and IL-1. Successful chemotherapy removes the source and often clears the skin without any steroid at all.[8]

Pathergy closes the loop: trauma — venepuncture, biopsy, sunburn, BCG vaccination — can seed a Sweet lesion at that site, mirroring pyoderma gangrenosum and Behçet disease. A pathergic reaction at the BCG site is a classic exam stem.[4]

VEXAS — the unifying diagnosis in the older man

Since 2020, VEXAS syndrome (vacuoles, E1 enzyme, X-linked, autoinflammatory, somatic) is the diagnosis to send for in any older man with a refractory Sweet-like neutrophilic dermatosis, relapsing polychondritis, cytopenias (macrocytic anaemia, thrombocytopenia), fever, and pulmonary infiltrates. The cause is a somatic mutation in UBA1, the E1 ubiquitin-activating enzyme; loss of cytoplasmic ubiquitination triggers constitutive innate immune activation. Send UBA1 sequencing and treat with IL-1 blockade (anakinra, canakinumab) or JAK inhibitors (ruxolitinib) — corticosteroids alone are usually inadequate.[2][9]

What you see at the bedside

The signature is sudden, painful, erythematous-to-violaceous lesions in a febrile patient, evolving over hours to days. The lesions are tender papules, plaques, and nodules with a smooth, often pseudovesicular surface — they look vesicular but feel solid (the look comes from marked papillary dermal oedema). The colour runs bright red to deep violaceous; lesions coalesce into polycyclic plaques.[1]

Distribution is classical: upper extremities (arms, dorsa of hands, fingers), face, neck, and upper trunk. The lower legs host the subcutaneous variant (neutrophilic panniculitis). Mucosa is usually spared; severe mucosal involvement should push you toward Stevens-Johnson syndrome or Behçet disease.[1]

Systemic features: fever precedes or accompanies the rash in about 80 per cent; malaise, myalgia, arthralgia, and headache are common; ocular involvement (conjunctivitis, episcleritis, scleritis, iridocyclitis) occurs in about a third. Extracutaneous neutrophilic infiltration can reach bone (sterile osteomyelitis), lung (neutrophilic pneumonitis), kidney, liver, and rarely CNS.[4][6]

Morphology in malignancy-associated disease is often bullous, ulcerative, or atypical, in unusual sites, or recurs despite corticosteroid — each should escalate the malignancy workup.[8]

The face-off — Sweet against its closest cousins

The differential includes true emergencies (SJS, leukaemia cutis, necrotising fasciitis) and conditions whose management is the opposite of Sweet's — cellulitis treated with antibiotics, vasculitis treated with immunosuppression after infection is excluded. This is the highest-yield part of the viva.[1]

[4]

The one-line discriminator: Sweet has vessel walls intact; vasculitis has fibrinoid necrosis. Both share triggers — infection, drug, malignancy, autoimmunity — so the biopsy, not the history, draws the line.[4]

Su and Liu — the criteria that make the diagnosis

Diagnosis requires both major criteria and at least two of the four minor criteria (von den Driesch modification).[1][3]

[1]

Confirm on biopsy — and exclude the killers

The punch biopsy is the single most important investigation. Classical Sweet shows a dense diffuse dermal neutrophilic infiltrate (upper and mid-dermis), marked papillary dermal oedema (the pseudovesicular look), karyorrhexis that reflects neutrophil turnover rather than vessel death, and — the cardinal feature — vessel walls intact. Order special stains (PAS, Grocott, Ziehl-Neelsen, Fite) to exclude infection.[4][6]

Histiocytoid Sweet shows immature myeloid cells with histiocyte-like morphology but a myeloid immunophenotype (MPO-positive, CD15-positive, CD68-positive). It is the variant most easily confused with leukaemia cutis — immunohistochemistry (CD34, CD117/KIT) and bone marrow biopsy settle it. Treating leukaemia cutis with corticosteroids delays oncology referral, so do not skip the stains.[7][8]

Blood tests: FBC with differential shows neutrophilia (a minor criterion); in malignancy-associated disease the count may be normal, low, or show cytopenias or blasts. ESR and CRP are raised; U&E and LFTs screen for organ involvement. Skin and blood cultures are sterile — expected, and reassuring.[1]

The malignancy workup is mandatory in adults: FBC with differential and peripheral film first-line; bone marrow aspirate and trephine if cytopenias, blasts, or dysplasia; LDH and beta-2 microglobulin; age-appropriate cancer screening; and UBA1 sequencing for the VEXAS phenotype.[2][8][9]

Steroids first — then treat the trigger

The treatment ladder reflects the disease's dramatic corticosteroid responsiveness and the primacy of treating the underlying cause.[1][3]

[1] [1]

Oral prednisolone 0.5 to 1 mg/kg/day is first-line for classical Sweet. The response is dramatic and characteristic — fever settles within 24 to 72 hours, lesions flatten and fade, no new lesions appear. Because that response is itself a minor criterion, observing it both confirms the diagnosis and treats the disease. Taper over 4 to 6 weeks; tapering too quickly is the commonest cause of the roughly 30 per cent recurrence. For severe disease, intravenous methylprednisolone 0.5 to 1 g daily for 3 to 5 days is an option.[3]

Steroid-sparers and alternatives — chosen regionally (EADV favours colchicine and potassium iodide; AAD literature favours early IL-1 blockade). Useful when steroids are contraindicated, on relapse, or in recurrent disease:[3]

[1]

Management by subtype — classical: prednisolone, taper over 4 to 6 weeks, treat the trigger. Malignancy-associated: treat the malignancy; successful chemotherapy often clears the skin without steroid, and recurrent Sweet may herald tumour relapse. Drug-induced: stop the drug (G-CSF first); a short steroid course accelerates healing. Histiocytoid: same ladder plus a mandatory haematology workup. VEXAS: anakinra or ruxolitinib — steroids alone are inadequate.[2][8]

The scenarios that change the plan

Histiocytoid Sweet carries the same steroid ladder but a mandatory bone marrow workup for myelodysplasia; IL-1 blockade is increasingly reported as effective. The pitfall is misreading it as leukaemia cutis — immunophenotyping (CD34, CD117/KIT) settles it.[7][8]

Subcutaneous Sweet (neutrophilic panniculitis) presents as tender nodules on thighs, arms, and lower legs, strongly linked to IBD — distinguish from erythema nodosum (septal panniculitis with a lymphohistiocytic infiltrate, predominantly shins).[4]

Sweet in pregnancy is rare, typically second or third trimester, resolving postpartum. First-line is super-potent topical corticosteroid; the safest systemic agent is IVIG (2 g/kg). Avoid colchicine, dapsone, and high-dose first-trimester steroids.[6]

Sweet in children is rare but well described, usually classical and post-infectious; histology and criteria are identical and malignancy must still be excluded (paediatric AML can present with Sweet). Use weight-based prednisolone dosing.[6]

Complications and the pitfalls that change management

Recurrence (around 30 per cent) is the commonest complication, driven by rapid taper or an untreated underlying cause. Corticosteroid morbidity accumulates over repeated courses. Cutaneous complications include secondary infection and postinflammatory hyperpigmentation; scarring is uncommon in classical disease but can occur in bullous or malignancy-associated variants.[1]

Prognosis and disposition

Classical Sweet has an excellent prognosis — corticosteroids produce a dramatic response within 24 to 72 hours and most cases resolve over 4 to 6 weeks without scarring; recurrence in about 30 per cent should always prompt reassessment for a cause. Malignancy-associated disease tracks the underlying malignancy — Sweet resolves with successful cancer treatment and may relapse with tumour relapse, a useful clinical marker. Drug-induced disease has the best prognosis, resolving on withdrawal. VEXAS is more guarded: a chronic relapsing autoinflammatory disease with significant morbidity and early 1-year mortality around 10 to 20 per cent, improved by early IL-1 or JAK blockade.[2][8]

Stable Sweet is managed in the outpatient dermatology clinic with multidisciplinary input from haematology, gastroenterology, rheumatology, and ophthalmology. Admit for extensive or rapidly progressive disease, diagnostic uncertainty, intravenous therapy, or acute systemic involvement.[1]

Ward-round test

A 60-year-old man with tender red-violet plaques, fever, neutrophilia, and a dense dermal neutrophilic infiltrate on biopsy — but the lesions recur every time you taper prednisolone. Name the single most important next test.

Send UBA1 sequencing for VEXAS syndrome. In an older man with refractory or recurrent Sweet-like neutrophilic dermatosis, especially with cytopenias (macrocytic anaemia, thrombocytopenia), relapsing polychondritis, or pulmonary infiltrates, a somatic UBA1 mutation is the unifying diagnosis. Corticosteroids alone are usually inadequate; treat with IL-1 blockade (anakinra, canakinumab) or a JAK inhibitor (ruxolitinib). Also re-screen the FBC and film for AML and MDS, which are the other drivers of recurrent Sweet.[2][9]

Biopsy of a Sweet-like plaque shows immature myeloid cells with MPO, CD15, and CD68 positivity. What is the diagnosis, and what must you exclude?

This is histiocytoid Sweet syndrome — immature neutrophilic granulocytes with a histiocyte-like morphology but a myeloid immunophenotype. It is strongly associated with myelodysplastic syndrome and other myeloid neoplasms, and it is the variant most easily confused with leukaemia cutis. Exclude leukaemia cutis with CD34 and CD117/KIT immunohistochemistry and a bone marrow biopsy — treating leukaemia cutis with corticosteroids delays oncology referral.[7][8]

A febrile patient with tender red plaques is started on antibiotics for cellulitis and deteriorates over 48 hours. Biopsy shows dense dermal neutrophils with intact vessel walls. What is the diagnosis and the first treatment step?

Sweet syndrome — misdiagnosed as cellulitis. The biopsy is diagnostic: a dense dermal neutrophilic infiltrate with vessel walls intact (no fibrinoid necrosis). Sweet is culture-negative and worsens without corticosteroids. Stop expecting antibiotics to work and start oral prednisolone 0.5 to 1 mg/kg/day — the response within 24 to 72 hours is itself a diagnostic minor criterion. Then hunt the trigger: infection, G-CSF, IBD, and screen the FBC and film for AML and MDS.[1][4]

References

1. [1]Villarreal-Villarreal CD, Ocampo-Candiani J, Villarreal-Martínez A. Sweet Syndrome: A Review and Update Actas Dermosifiliogr, 2016.PMID 26826881
2. [2]Loeza-Uribe MP, Hinojosa-Azaola A, Sánchez-Hernández BE, et al. VEXAS syndrome: Clinical manifestations, diagnosis, and treatment Reumatol Clin (Engl Ed), 2024.PMID 38160120
3. [3]Calabrese L, Satoh TK, Aoki R, et al. Sweet syndrome: an update on clinical aspects, pathophysiology, and treatment Ital J Dermatol Venerol, 2024.PMID 39560338
4. [4]Nelson CA, Stephen S, Ashchyan HJ, et al. Neutrophilic dermatoses: Pathogenesis, Sweet syndrome, neutrophilic eccrine hidradenitis, and Behçet disease J Am Acad Dermatol, 2018.PMID 29653210
5. [5]Delaleu J, Lepelletier C, Calugareanu A, et al. Neutrophilic dermatoses Rev Med Interne, 2022.PMID 35870984
6. [6]Cohen PR. Sweet's syndrome--a comprehensive review of an acute febrile neutrophilic dermatosis Orphanet J Rare Dis, 2007.PMID 17655751
7. [7]Requena L, Kutzner H, Palmedo G. Histiocytoid Sweet syndrome: a dermal infiltration of immature neutrophilic granulocytes Arch Dermatol, 2005.PMID 16027297
8. [8]Maronese CA, Derlino F, Moltrasio C, et al. Neutrophilic and eosinophilic dermatoses associated with hematological malignancy Front Med (Lausanne), 2023.PMID 38249974
9. [9]Tan IJ, Ferrada MA, Ahmad S, et al. Skin Manifestations of VEXAS Syndrome and Associated Genotypes JAMA Dermatol, 2024.PMID 38865133