Pyoderma gangrenosum

Meet the patient

A 44-year-old with ulcerative colitis arrives with a lower-leg ulcer that appeared at a venepuncture site four days ago and has trebled in size. The border is purple, overhanging, and exquisitely painful; the surrounding skin is dusky. The surgical registrar has booked it for debridement tomorrow.[1][2]

That debridement is the single most harmful act possible — it will enlarge the ulcer and seed new lesions along the incision. Cancel it. This is pyoderma gangrenosum, where pathergy means the wound you make becomes the next ulcer. Hold that sentence and every exam answer on PG falls into place.[3]

The mantra — and the name that lies

The one line every candidate must reproduce: do not debride a suspected pyoderma gangrenosum ulcer — pathergy worsens the disease. The wound that deteriorated dramatically after debridement is the classic PG exam stem.[1]

The name is a 1908 coinage of Louis Brocq, and it lies: pyoderma implies pus-forming, but the lesion is sterile — cultures are negative and antibiotics alone do not heal it. PG is driven by aberrant innate immune activation with sterile neutrophilic infiltration of the skin, the same family as Sweet syndrome, subcorneal pustular dermatosis, and erythema elevatum diutinum.[3][4]

Etymology for viva gold: gangrenosum scared a century of clinicians into debriding these ulcers like wet gangrene — the very act pathergy forbids. The word outlived the misunderstanding, but the lesson is permanent.[1]

Two features dominate every question

Two clinical features separate PG from every ordinary leg ulcer, and they anchor the whole topic.[1]

[1]

The classic trap: post-surgical PG declares itself four to seven days after the operation — the wound fails to heal, becomes exquisitely painful and purple rather than pink, and enlarges dramatically when "debrided" as a presumed infection. The covering surgeon reaches for the scalpel; the dermatologist cancels the list.[3]

Five variants — and the association each one forces

Distinguish the variant at first presentation, because it directs the workup. A bullous lesion on the hands mandates a haematological malignancy screen; a peristomal ulcer points to ulcerative colitis after colectomy.[1]

[1] [1]

The variant-association links to memorise: bullous points to AML or MDS (send FBC, film, bone marrow); pustular points to an IBD flare (stool calprotectin, colonoscopy); peristomal points to UC after colectomy; vegetative usually needs only topical clobetasol.[3]

IMAH — the four associations you screen for every time

About half of PG is linked to systemic disease, and the workup is the same every time. The mnemonic IMAH — Inflammatory bowel disease, Monoclonal gammopathy, Arthritis, Haematological malignancy — covers it; the rest is idiopathic, still a diagnosis of careful exclusion.[1]

[1]

Everyone forgets: PG can precede the IBD by years, so a negative bowel history does not exclude it — a stool calprotectin is cheap and catches the silent colitis. And a familial form exists, linked to PSTPIP1 mutations and the autoinflammatory syndromes PAPA, PASH, PAPASH, PsAPASH, where IL-1 blockade (anakinra, canakinumab) can be highly effective.[9]

The neutrophil, the cytokines, and why pathergy happens

PG is a skin-limited autoinflammatory disease. The neutrophil, not the lymphocyte, is the effector cell. A trigger — skin trauma, an IBD flare, malignancy-derived cytokines — fires a runaway cascade of IL-1 beta, IL-17, IL-23, TNF-alpha, IL-8, and G-CSF that recruits neutrophils into otherwise normal skin.[3]

The cytokine map that explains the drugs: IL-23 to IL-17 sustains the infiltrate (hence ustekinumab and the IL-17 biologics); TNF-alpha is the central amplifier (hence infliximab and adalimumab); IL-8 is the dominant neutrophil chemoattractant; G-CSF both drives neutrophil release and is itself a recognised drug trigger of PG. The loop closes on itself through pathergy — trauma recruits more neutrophils, which make more ulcer.[1]

Pathergy is the defining clinical-pathological link. Trauma provokes a localised neutrophilic infiltrate that in a healthy person resolves in hours; in PG it escalates into a destructive ulcer. That is why PG appears at surgical wounds, cannulae, and tattoos, why debridement enlarges it, and why a skin graft fails when the donor site itself becomes PG.[3]

What you see at the bedside

The classic ulcer evolves in two phases. First a small tender papule, pustule, or nodule — mistaken for an insect bite or folliculitis — that becomes haemorrhagic and breaks down within 24 to 72 hours. Then the ulcerative phase: a deep, painful ulcer with the signature undermined, overhanging, violaceous border, a ragged necrotic base, and surrounding dusky erythema, enlarging concentrically — often 1 to 2 cm per day untreated.[1]

Distribution favours the lower leg (shin, ankle, dorsum of foot), then trunk and peristomal areas; the bullous variant favours the upper limbs and dorsal hands. Mucous membranes are typically spared — a point that helps separate PG from Behçet disease.[1]

Pathergy at presentation is almost always in the history: the ulcer appeared at a recent venepuncture, cannula, insect bite, scratch, or surgical wound. Many patients feel systemically well, but fever, malaise, and arthralgia may herald an underlying IBD or malignancy flare.[3]

The face-off — PG against its ulcer mimics

PG is a diagnosis of exclusion, and the price of error runs both ways. Treating an infected ulcer with high-dose steroids is catastrophic; treating PG as infection with serial debridement is equally so. Exclude the four mimics before any immunosuppression.[3][7]

[1]

The four mimics to exclude before steroid: infection (wound swab and deep tissue culture, special stains), vasculitis (palpable purpura, biopsy with immunofluorescence, ANCA), malignancy (Marjolin or SCC in a chronic wound — biopsy the edge), and vascular insufficiency (ABPI, arterial and venous Doppler). Necrotising fasciitis is the one you must never miss — severe pain out of proportion, systemic sepsis, crepitus, a surgical emergency.[1]

Diagnosis of exclusion — the criteria and the score

There is no pathognomonic test for PG. The biopsy shows a non-specific sterile dermal neutrophilic infiltrate with central necrosis and a mixed infiltrate — a pattern shared with Sweet syndrome — so its job is to exclude vasculitis, infection, and malignancy, not to confirm PG.[3][4]

Two validated instruments standardise the diagnosis once mimics are excluded. The Cheng 2018 Delphi consensus requires two major (pain; rapid progression with cribriform scarring) and two of four minor (pathergy; undermined violaceous border; a systemic association; compatible histology excluding infection). The PARACELSUS score of Dissemond 2019 uses ten diagnostic categories plus four exclusionary items; a score of 6 or more supports PG.[7][8]

[1]

Investigations — exclude mimics, find the trigger

Every investigation serves one of two purposes: exclude a mimic, or detect an associated systemic disease.[1]

Histopathology of the active advancing edge shows the non-specific sterile neutrophilic infiltrate — and, crucially, excludes leucocytoclastic vasculitis, lymphoma, and infection (order Ziehl-Neelsen, Fite, PAS, and Grocott stains). Direct immunofluorescence excludes IgA vasculitis and bullous disorders.[3][4]

Microbiology: wound swab AND deep tissue biopsy for bacterial, mycobacterial, and fungal culture, plus tissue PCR for Leishmania and atypical mycobacteria if the context fits. A positive superficial swab may be colonisation — deep tissue culture is more reliable.[1]

The associated-disease blood screen is the IMAH workup: FBC with differential and film (AML, MDS), ESR and CRP, SPEP and urine Bence-Jones (IgA monoclonal gammopathy), rheumatoid factor and ANA, HLA-B27, stool calprotectin (IBD), ANCA and complements (vasculitis), and the pre-biologic screen — hepatitis B and C, HIV, and QuantiFERON-TB.[1]

Imaging and endoscopy: chest X-ray and CT-PET for occult malignancy; Doppler and ABPI for venous and arterial disease; bone marrow aspirate when FBC or SPEP suggest haematological malignancy; colonoscopy with biopsies when stool calprotectin is raised or bowel symptoms exist — ulcerative colitis and Crohn are the strongest associations.[2]

Resuscitation — the rule that saves the limb

PG is not a "fast" emergency like necrotising fasciitis, but the wrong first action is disastrous. The resuscitative phase has one absolute rule and four steps around it.[1]

Exclude infection first — wound swab and deep tissue biopsy before any steroid or biologic; empirical broad-spectrum antibiotics only for clinical sepsis. Wound care aims for a moist, low-bioburden, non-traumatic environment (silicone foams, hydrofibres, cadexomer iodine or silver for high bioburden); compression only if venous insufficiency is documented and ABPI is adequate. Analgesia is stepwise — paracetamol, then a short-acting opioid, then a neuropathic agent for the burning component opioids under-treat. Decide disposition — admit if the ulcer grows over 5 cm in 24 hours, the variant is bullous, pain is uncontrolled orally, or there is suspected sepsis.[1]

The treatment ladder — localised to refractory

Definitive therapy follows severity and variant. The ladder runs from a topical steroid for a solitary vegetative ulcer to infliximab for refractory or IBD-linked disease.[1]

[1]

Mild or localised disease — super-potent topical clobetasol 0.05 per cent BD under occlusion, topical tacrolimus 0.1 per cent (steroid-sparing, useful on face, folds, and peristomal skin), or intralesional triamcinolone 10 to 40 mg per mL at the active border.[1]

Moderate disease — add oral prednisolone 0.5 to 1 mg/kg/day (max about 80 mg) as a single morning dose, taper after response; consider concurrent ciclosporin 3 to 5 mg/kg/day for a steroid-sparing effect.[1]

Severe, rapidly progressive, or refractory disease — IV methylprednisolone pulse 0.5 to 1 g/day for 3 to 5 days, then oral prednisolone; ciclosporin 3 to 5 mg/kg/day added early; and infliximab 5 mg/kg IV at weeks 0, 2, 6, then 8-weekly — the biologic of choice for refractory PG, especially with IBD, because it treats both diseases.[1][6]

Steroid-sparing and adjunct agents — azathioprine (check TPMT first), mycophenolate, methotrexate, dapsone (check G6PD first), colchicine, and IVIG; ustekinumab, the IL-17 biologics, and canakinumab (IL-1 beta) for the autoinflammatory syndromes; JAK inhibitors with emerging evidence.[1]

The red-flag drug monitoring that examiners reward: ciclosporin needs BP and creatinine every two weeks for three months; azathioprine needs TPMT before starting and FBC weekly for four weeks; dapsone needs G6PD before starting and FBC plus methaemoglobin; infliximab needs IGRA and hepatitis serology before starting to avoid reactivating latent TB or hepatitis B.[1]

The scenarios that change the plan

Bullous PG. Rapidly spreading superficial haemorrhagic bullae on the dorsal hands and forearms — the differential that matters is AML, MDS, and the myeloproliferative neoplasms, not the morphological mimic Sweet syndrome. Send FBC, film, and bone marrow; priority is to treat the malignancy, and PG often resolves with chemotherapy.[3][4]

Peristomal PG. Around an ileostomy, colostomy, or urostomy, classically in UC after colectomy — pathogenesis is appliance trauma plus pathergy. Refit the appliance with a stoma nurse, use topical tacrolimus 0.1 per cent (no atrophy with chronic use), inject intralesional triamcinolone at the border, and reserve systemic therapy or stoma relocation for refractory disease.[2][5]

Post-surgical PG. Four to seven days after abdominopelvic, breast, or cardiac surgery — intense pain out of proportion, violaceous rather than pink margins, dramatic deterioration with debridement. Stop debridement, confirm the diagnosis, exclude infection, start systemic corticosteroid plus ciclosporin.[3]

PG in pregnancy. Use topical corticosteroids and oral prednisolone; avoid ciclosporin in the first trimester, methotrexate, and mycophenolate; continue a biologic only with multidisciplinary input if PG or IBD is severe.[1]

Complications and the pitfalls that cost limbs

Cribriform ("meshed") scarring — a pathognomonic pattern of small punched-out scars from healing pustules — remains after the ulcer heals. Secondary infection (Staphylococcus aureus, Pseudomonas, Streptococcus pyogenes) can progress to sepsis, the leading cause of PG-related death. PG is also a prothrombotic state — venous thromboembolism is recognised, particularly in the bullous variant with malignancy. Chronic pain drives depression and opioid dependence.[1]

Prognosis and disposition

PG is a chronic relapsing-remitting disease. Mean time to complete healing with appropriate therapy is 3 to 9 months, with substantial pain relief within 1 to 2 weeks of starting systemic corticosteroid; recurrence is 20 to 40 per cent over 5 years, higher with active IBD or persistent haematological malignancy. Overall mortality attributable to PG itself is low but runs 3 to 30 per cent in series, driven by underlying malignancy, secondary sepsis, and thromboembolism. Once healing is complete, continue immunosuppression for 1 to 3 months then taper slowly to avoid rebound flare.[3][5]

Stable PG is managed in the outpatient dermatology clinic with shared care from gastroenterology, haematology, rheumatology, and stoma therapy; a documented flare action plan — early presentation, photography, swab, escalation pathway — reduces avoidable admissions.[5]

Ward-round test

A 50-year-old develops a rapidly enlarging, exquisitely painful purple-bordered ulcer at a laparotomy wound six days after surgery. The surgical team plans debridement. What do you do, and why?

Cancel the debridement. This is post-surgical pyoderma gangrenosum — onset four to seven days after surgery, intense pain out of proportion, violaceous rather than pink margins, and the signature undermined border. The single most important rule in PG is do not debride: pathergy means the surgical wound enlarges and new lesions seed along the incision. Confirm with biopsy (non-specific neutrophilic infiltrate, excluding infection and vasculitis), send wound swab and deep tissue culture, then start systemic corticosteroid plus ciclosporin. Apply the Cheng Delphi criteria — pain and rapid progression with cribriform scarring are major; pathergy and the undermined violaceous border are minor.[1][3]

A 65-year-old has rapidly spreading haemorrhagic bullae on the dorsal hands. Biopsy shows a neutrophilic infiltrate. What is the single most important next investigation, and why?

Send FBC with differential and peripheral blood film, and prepare for bone marrow biopsy. This is bullous (atypical) pyoderma gangrenosum, the variant most tightly linked to haematological malignancy — AML, MDS, and the myeloproliferative neoplasms. The priority is to find and treat the malignancy; PG often resolves with chemotherapy. Use systemic corticosteroid cautiously, because immunosuppression in a marrow-failure patient risks sepsis.[3][4]

Name the four mimics you must exclude before starting corticosteroid for suspected PG, and the single discriminator for each.

Infection (wound swab and deep tissue culture, special stains on biopsy), vasculitis (palpable purpura, biopsy with immunofluorescence, ANCA), malignancy (Marjolin or SCC in a chronic wound — biopsy the edge; cutaneous lymphoma), and vascular insufficiency (ABPI, arterial and venous Doppler). PG is a diagnosis of exclusion — the biopsy excludes mimics rather than confirming PG, because the neutrophilic infiltrate is non-specific.[1][7]

References

1. [1]Maronese CA, Pimentel MA, Li MM, et al. Pyoderma Gangrenosum: An Updated Literature Review on Established and Emerging Pharmacological Treatments Am J Clin Dermatol, 2022.PMID 35606650
2. [2]Rogler G, Singh A, Kavanaugh A, et al. Extraintestinal Manifestations of Inflammatory Bowel Disease: Current Concepts, Treatment, and Implications for Disease Management Gastroenterology, 2021.PMID 34358489
3. [3]Maverakis E, Marzano AV, Le ST, et al. Pyoderma gangrenosum Nat Rev Dis Primers, 2020.PMID 33033263
4. [4]Delaleu J, Lepelletier C, Calugareanu A, et al. Neutrophilic dermatoses Rev Med Interne, 2022.PMID 35870984
5. [5]Dissemond J, Marzano AV, Hampton PJ, et al. Pyoderma Gangrenosum: Treatment Options Drugs, 2023.PMID 37610614
6. [6]Tan MG, Tolkachjov SN. Treatment of Pyoderma Gangrenosum Dermatol Clin, 2024.PMID 38423680
7. [7]Maverakis E, Ma C, Shinkai K, et al. Diagnostic Criteria of Ulcerative Pyoderma Gangrenosum: A Delphi Consensus of International Experts JAMA Dermatol, 2018.PMID 29450466
8. [8]Jockenhöfer F, Wollina U, Salva KA, et al. The PARACELSUS score: a novel diagnostic tool for pyoderma gangrenosum Br J Dermatol, 2019.PMID 29388188
9. [9]DeFilippis EM, Feldman SR, Huang WW. The genetics of pyoderma gangrenosum and implications for treatment: a systematic review Br J Dermatol, 2015.PMID 25350484