Symmetrical peripheral gangrene in a 77-year-old woman with ulcerative colitis

Key points

• Symmetrical peripheral gangrene is ischemic necrosis of 2 or more extremities without large-vessel occlusion and should be considered a potential consequence in any patient with bilateral, distal, symmetric, and acral ischemia and preserved or palpable pulses.

• Symmetrical peripheral gangrene is a diagnosis of exclusion that is usually driven by disseminated intravascular coagulation and microvascular thrombosis, diagnosed only after arterial occlusion, venous gangrene (phlegmasia), vasculitis, antiphospholipid syndrome, cryoglobulinemia, and embolic causes have been considered and excluded.

• A severe inflammatory bowel disease flare can generate the prothrombotic, consumptive state that precipitates symmetrical peripheral gangrene, even without classically described shock or sepsis.

• Management is directed toward treating the underlying inflammatory driver, with anticoagulation balanced against bleeding (favouring reversible agents) and necrosis allowed to demarcate before débridement or amputation.

A 77-year-old woman presented to our emergency department with a 10-day history of bloody diarrhea, abdominal pain, tenesmus, and nocturnal stools, as well as 2 to 3 days of progressive, painful swelling and dusky discoloration of both feet. Four months earlier, she had undergone a colonoscopy performed for a positive fecal immunochemical test, which had shown mild colitis, with biopsy-confirmed chronic active colitis in the rectum and the left and transverse colon, consistent with ulcerative colitis. She had not attended follow-up. This episode represented a severe flare of newly diagnosed, untreated ulcerative colitis. The patient had hypertension, was a nonsmoker, and had no prior venous thromboembolism, bleeding disorder, Raynaud phenomenon, or vascular disease. She was not on anticoagulant or antiplatelet therapy.

We admitted the patient to our internal medicine teaching unit on the day of presentation. She was hemodynamically stable and afebrile. Both feet were cool, with acral violaceous mottling of the toes and forefeet, nonpitting edema, and pulses difficult to palpate. She did not have cellulitis, wet gangrene, skin change elsewhere, or upper-limb involvement.

Initial investigations showed normocytic anemia (hemoglobin 101 [reference 120 to 160] g/L), a normal leukocyte level (10.0 [reference 4.0 to 11.0] × 10⁹/L), a normal platelet level (193 [reference 140 to 400] × 10⁹/L), an elevated C-reactive protein (CRP) level (52.5 [reference < 8.0] mg/L), an elevated fecal calprotectin level (4310 [reference < 50] μg/g; > 250 μg/g suggests active inflammation) and hypoalbuminemia (22 [reference 30 to 45] g/L). Over the next 1 to 2 days, before any anticoagulation, the patient’s platelet count fell to 101 × 10⁹/L, her international normalized ratio (INR) was 1.3 (reference 0.8 to 1.2), her activated partial thromboplastin time (aPTT) was 107 (reference 25 to 37) seconds, her fibrinogen level was 0.6 (reference 2.0 to 4.0) g/L, and her D-dimer level was 98.87 (reference < 0.50) mg/L fibrinogen equivalent units. These findings were concerning for a consumptive coagulopathy without hepatic or renal failure.

Venous ultrasonography on admission showed a nonocclusive thrombus in the left common femoral vein extending into the femoral vein (approximately 3.5 cm) and extensive occlusive thrombi in the calf veins of both legs, without right femoral or great saphenous involvement. Computed tomography (CT) pulmonary angiography was negative for pulmonary embolism. A vascular surgeon assessed the patient the day after admission and considered that the bilateral, distal findings with preserved pulses were not in keeping with phlegmasia cerulea dolens, meaning no arterial intervention was indicated. We did not administer pharmacologic prophylaxis for venous thromboembolism on admission; it was withheld for active hematochezia (1 unit of packed red blood cells were transfused) and pending endoscopy.

Flexible sigmoidoscopy on day 3 of admission showed severe rectosigmoid inflammatory disease (Mayo endoscopic subscore 3) with ulceration. Biopsies confirmed chronic active colitis; stool studies, including testing for Clostridioides difficile, were negative. After sigmoidoscopy, we started the patient on reversible intravenous unfractionated heparin, intravenous methylprednisolone, infliximab, and rectal 5-aminosalicylate.

Despite anticoagulation, acral changes progressed to hemorrhagic bullae and necrosis (Figure 1). On hospital day 7, CT angiography showed patent arteries of both lower legs, with normal 3-vessel runoff and no arterial thrombus, stenosis, or occlusion. Inflammatory markers had decreased (calprotectin 152 μg/g and CRP 38.9 mg/L). A dermatologist performed a skin biopsy of the left anterior shin; histopathology showed ischemic, vasculopathic changes consistent with microvascular occlusion, without vasculitis, immune deposits (negative direct immunofluorescence test), or infection. Blood cultures and echocardiography excluded septic emboli. Serology showed cytoplasmic anti-neutrophil cytoplasmic antibody (ANCA) with anti-proteinase-3 positivity, negative anti-myeloperoxidase antibody, and positive speckled antinuclear antibody (1:80). Inpatient testing did not support antiphospholipid syndrome, as lupus anticoagulant was not detected, and the patient tested negative for anticardiolipin antibodies, anti-β-2 glycoprotein antibodies, and cryoglobulins. A reactive thrombocytosis later developed (peak 730 × 10⁹/L on hospital day 14). With these alternative diagnoses excluded, we diagnosed symmetrical peripheral gangrene.

Figure 1:

Progression of bilateral distal foot ischemia from discoloration to dry gangrene during a flare of ulcerative colitis. (1A, 1B) On day 1, (A) dorsal and (B) plantar views show acral violaceous discoloration and mottling consistent with early ischemia, with preserved pulses. On day 13, (C) dorsal and (D) plantar views show demarcating necrosis of the distal toes and forefeet, without edema, purulence or other features of wet gangrene or secondary infection. Two months after presentation, (E) dorsal and (F) plantar views show dry gangrene with partial autoamputation and healing, residual digital loss, and preserved proximal perfusion. The concurrent C-reactive protein (reference < 8.0 mg/L) levels and platelet counts (reference 140 to 400 × 10⁹/L), respectively, were 52.5 mg/L and 193 × 10⁹/L on day 1, 5.2 mg/L and 597 × 10⁹/L on day 13, and 2.2 mg/L and 464 × 10⁹/L at 2 months.

As the patient’s symptoms of colitis and bleeding settled, we transitioned her anticoagulation from intravenous unfractionated heparin infusion to enoxaparin (1 mg/kg subcutaneously every 12 h) on hospital day 13 and then apixaban (5 mg twice daily) on hospital day 20; we added sildenafil (25 mg twice daily) on hospital day 15 for its vasodilatory effects. We discharged her on day 21 with demarcating dry gangrene. By 2 months, the toes had partially autoamputated without further progression.

Discussion

Symmetrical peripheral gangrene is defined as symmetric ischemic necrosis of 2 or more extremities that is not explained by large-vessel arterial or venous occlusion. ¹ It should be suspected whenever acral discoloration suggesting ischemia is bilateral, distal, and symmetric, and peripheral pulses are preserved or palpable. This combination can be misleading because intact pulses and patent large arteries may falsely reassure providers and lead them to exclude acute arterial occlusion; however, among patients with symmetrical peripheral gangrene, mortality exceeds 40%, and digital or limb loss occurs in as many as 70% of survivors. ² The occlusion is at the microvascular level, thus normal large-vessel flow on duplex and CT angiography does not exclude severe, full-thickness tissue loss either. ² In our patient, bilateral distal necrosis with preserved pulses, patent arteries on CT angiography, consumptive coagulopathy, and a microvascular thrombotic skin biopsy collectively pointed to symmetrical peripheral gangrene.

Symmetrical peripheral gangrene is classically described in critically ill patients with septic shock, low-output states, or other causes of severe disseminated intravascular coagulation (DIC). ¹ Some scholars have proposed that symmetrical peripheral gangrene arises when DIC is combined with disturbed natural anticoagulant (protein C) pathways and a precipitating acute illness. ¹ Our patient was neither septic nor in shock. However, the intense systemic inflammation and consumptive, DIC-like coagulopathy of a severe ulcerative colitis flare can serve as a precipitant even without hemodynamic shock. ² Protein C and antithrombin levels were not measured in our patient. Through elevated clotting factors, thrombocytosis, endothelial activation and reduced natural anticoagulant levels, active inflammatory bowel disease is itself strongly prothrombotic. ³ It can drive macro- and microvascular thrombosis. ³ However, our inference of symmetrical peripheral gangrene was based on the laboratory pattern (e.g., thrombocytopenia, hypofibrinogenemia, prolonged coagulation times, markedly elevated D-dimer level) that was present before anticoagulation and resolved as the flare was treated, rather than on a single confirmatory test.

A structured work-up is needed to diagnose symmetrical peripheral gangrene and exclude its mimics (Table 1). Venous duplex ultrasonography defines the thrombotic burden. Arterial imaging, usually CT angiography, confirms patent large vessels and excludes acute arterial occlusion. A coagulation panel — which includes platelet count, INR, aPTT, fibrinogen, and D-dimer — characterizes the consumptive coagulopathy. Targeted testing should include lupus anticoagulant, anticardiolipin antibodies, and anti-β-2 glycoprotein antibodies for antiphospholipid syndrome and its catastrophic variant, and cryoglobulins for cryoglobulinemia. When the timing fits, testing for anti–platelet factor 4 (PF4) antibodies can identify a heparin-induced or autoimmune anti-PF4 thrombotic disorder. ¹ Blood cultures and echocardiography evaluate septic emboli and endocarditis. A skin biopsy distinguishes microvascular thrombotic vasculopathy from vasculitis. Colonic histology distinguishes inflammatory bowel disease from ischemic or infectious causes of colitis.

Table 1:

Differential diagnosis of bilateral acral (foot) ischemia

Condition	Typical features	In our patient
Acute arterial limb ischemia or large-vessel occlusion	Asymmetric, proximal Absent pulses Arterial occlusion on imaging	Preserved pulses CT angiography showing normal 3-vessel runoff, no arterial occlusion
Phlegmasia cerulea dolens (venous gangrene)	Massive proximal DVT Whole-limb cyanosis and edema	Distal, symmetric, bilateral pulses Preserved pulses No whole-limb cyanosis
Disseminated intravascular coagulation or consumptive coagulopathy	Thrombocytopenia, low fibrinogen level, prolonged coagulation times, high D-dimer level	Present at admission before anticoagulation, favoured mechanism
Antiphospholipid syndrome	Detectable lupus anticoagulant, anticardiolipin, and anti-β-2 glycoprotein	Lupus anticoagulant not detected Negative for anticardiolipin and anti-β-2 glycoprotein antibodies
Cryoglobulinemia	Positive for cryoglobulins Often caused by hepatitis C or lymphoproliferative disease	Negative for cryoglobulins
Anti-PF4 or heparin-induced thrombotic disorder	Thrombosis with falling platelets after PF4 exposure	Unlikely as coagulopathy preceded heparin, platelet counts increased
ANCA-associated vasculitis	Systemic features (ENT, lung, kidney) Vasculitis on biopsy	No systemic features No vasculitis on biopsy Anti-PR3 nonspecific in IBD
IgA vasculitis or Henoch–Schönlein purpura	Palpable purpura IgA deposits on biopsy	No palpable purpura Negative direct immunofluorescence
Cholesterol embolization	Recent arterial instrumentation Cholesterol clefts on biopsy	No instrumentation No cholesterol clefts on biopsy
Ischemic or ulcerative colitis	Histology distinguishes ischemic injury from chronic IBD	Chronic active colitis typical of IBD, not ischemic colitis

Note: ANCA = anti-neutrophil cytoplasmic antibody, CT = computed tomography, DVT = deep vein thrombosis, ENT = ear, nose, and throat, IBD = inflammatory bowel disease, IgA = immunoglobulin A, PF4 = platelet factor 4, PR3 = proteinase-3.

Applying this approach with our patient, acute arterial limb ischemia and large-vessel occlusion were unlikely given preserved pulses and normal 3-vessel runoff, and would typically be asymmetric and more proximal than what we observed. ⁴ We considered phlegmasia cerulea dolens because of the extensive bilateral occlusive calf-vein thrombosis, but this was less favoured based on the distal, symmetric pattern, preserved pulses, and absence of whole-limb cyanosis or massive edema. ⁵ Inpatient testing did not support antiphospholipid syndrome, although confirmatory repeat testing at 12 weeks was not performed. ⁶ Cholesterol embolization was unlikely without arterial instrumentation or cholesterol clefts on biopsy. An anti-PF4 or heparin-induced disorder was unlikely because the patient’s coagulopathy and thrombosis preceded heparin and the platelet count later increased rather than dropping. The rheumatologist did not consider ANCA-associated and other small-vessel vasculitides, including IgA vasculitis, to be in keeping with systemic vasculitis given that the patient had no sinus, ear, nose, throat, ocular, oral-ulcer, pulmonary, renal (no hematuria, normal creatinine), neurologic, or purpuric features, and that vasculitis was not apparent on the biopsy. ⁷ Isolated anti-proteinase-3 positivity is a recognized, nonspecific finding in inflammatory bowel disease and rarely changes management. ⁷ Finally, colonic histology showed chronic active colitis typical of inflammatory bowel disease rather than ischemic colitis. Our patient had a markedly elevated fecal calprotectin level on admission during the flare, with markers falling as the disease was treated and the gangrene stabilized.

Management of symmetrical peripheral gangrene prioritizes treating the underlying inflammatory condition and managing thrombosis without precipitating hemorrhage. Treatment of our patient’s colitis with corticosteroids, infliximab, and a 5-aminosalicylate also resulted in resolution of the secondary coagulopathy. ⁸ Anticoagulation was indicated for established venous thromboembolism but had to be balanced against active gastrointestinal bleeding. When bleeding precludes therapeutic anticoagulation in a patient with proximal venous thromboembolism, reduced-intensity anticoagulation should be used, where feasible. For patients for whom pharmacologic anticoagulation is not feasible, a retrievable inferior vena cava filter is preferable to mechanical calf compression, which is potentially harmful in limbs that already have extensive calf-vein thrombosis and threatened perfusion. ⁹

Adjunctive therapies have limited evidence. We used a vasodilator (sildenafil) in our patient to support perfusion, and intravenous prostaglandins such as iloprost are used for digital ischemia in some centres, but neither is proven in symmetrical peripheral gangrene. Protein C concentrate and plasma exchange have been described for severe DIC-associated disease but lack robust support and were not used in our patient. ¹ A limb-preservation approach is preferred; dry gangrene should be allowed to demarcate before débridement or amputation, with urgent surgery reserved for wet gangrene, infection, or another acute indication. In our patient, bedside débridement of bullae and wound care sufficed, and the toes later autoamputated.

Bilateral, symmetric, acral ischemia with preserved pulses should prompt early consideration of symmetrical peripheral gangrene, a structured work-up for a consumptive coagulopathy and its trigger, and exclusion of arterial, venous, embolic, and vasculitic causes, as well as antiphospholipid syndrome. Multidisciplinary care should involve specialists in internal medicine and subspecialists in gastroenterology, hematology, vascular surgery, dermatology, rheumatology, and podiatry as clinically indicated. The priority is prompt recognition and treatment of the underlying driver. Although established necrosis is often irreversible, controlling the inflammatory and coagulopathic process can limit progression and preserve viable tissue.