A 53-year-old woman was treated with intravenous unfractionated heparin (UFH) for an acute pulmonary embolism. She was started on 10 mg of warfarin on her fourth day of heparin therapy. Her baseline platelet count was normal, but decreased to 86 × 10⁹/L by day 7 of treatment. Her international normalized ratio (INR) was 3.2 at that time and heparin was discontinued. A platelet factor 4 (PF 4)–heparin immunoassay was found to be positive, consistent with a diagnosis of heparin-induced thrombocytopenia (HIT). Three days later she developed a violaceous discoloration of both breasts that progressed to full-thickness skin necrosis and warfarin-induced skin necrosis was diagnosed (Figure 62-1). Warfarin was discontinued and the direct thrombin inhibitor (lepirudin) instituted as the platelet count dropped further to 22 × 10⁹/L. She ultimately required extensive surgical debridement of both the breasts and reconstructive surgery. Warfarin was later resumed cautiously at 1 mg/d after full platelet recovery (overlapped with lepirudin) until the INR was therapeutic (>2) for two consecutive days.

EPIDEMIOLOGY

• 
  

  Severe anticoagulant-induced skin reactions that can develop following subcutaneous injection of heparin (UFH) or any of the low–molecular weight heparin (LMWH) preparations.^{1,2,3,4, and 5} It can also be seen following intravenous UFH administration.^3,4,7

  
  
• 
  

  More commonly seen in patients receiving UFH than LMWH. It has been reported in most of the LMWH preparations including dalteparin, enoxaparin, tinzaparin, tedelparin, certoparin, and nadroparin.⁴

  
  
• 
  

  Incidence likely underestimated due to either under-recognition and/or under-reporting.

  
  
• 
  

  An uncommon manifestation of HIT.

  
  
• 
  

  Most patients (50%-75%) will not develop thrombocytopenia despite the presence of heparin-dependent antibodies. However, if antibodies are present they are a marker for HIT.^2,4,5

  
  
• 
  

  Only 10% to 20% of patients who develop HIT antibodies during subcutaneous administration of UFH or LMWH will develop skin lesions.²

ETIOLOGY AND PATHOPHYSIOLOGY

• 
  

  Generally develops in conjunction with HIT, an immunologic reaction triggered by the binding of heparin to PF 4 resulting in the formation of heparin-dependent antibodies that can cause thrombocytopenia and venous and/or arterial thrombosis.^1,2,4,5

  
  
• 
  

  Patients often have other comorbid conditions such as hypertension, diabetes mellitus and obesity, underlying malignancy, or a connective tissue disease.¹

DIAGNOSIS

Clinical Features

• 
  

  Consider in any patient who develops a skin reaction while receiving subcutaneous UFH or LMWH or intravenous UFH (Figure 62-2).

  
  
• 
  

  Lesions appear suddenly and vary from tender, local erythematous plaques to painful sharply delineated necrotic lesions with a cuff of erythema or violaceous discoloration surrounding the necrosis (Figure 62-3).^2,5,6

  
  
• 
  

  Lesions may be multiple.

  
  
• 
  

  Skin manifestations have also been reported with other unusual presentations of HIT such as adrenal hemorrhagic infarction and severe acute anaphylactoid reaction.^2,5

Typical Distribution

• 
  

  Most commonly found at the injection site, although they may be found at distal sites, especially regions with extensive subcutaneous fat tissue such as the buttocks, thighs, breasts, or abdomen.⁴

LABORATORY STUDIES

• 
  

  Thrombocytopenia, defined as a platelet count less than 150 × 10⁹/L or 50% or greater drop from preadministration levels in patients receiving UFH or LMWH who are considered at risk for HIT.

  
  
• 
  

  Obtain heparin antibodies, an immunologic study (PF 4–heparin immunoassay), and/or a functional test (serotonin release assay).

  
  
• 
  

  Majority of patients with heparin-induced skin necrosis (HISN) will have serologic confirmation of HIT.

  
  
• 
  

  Skin biopsy demonstrates extensive focal epidermal necrosis with marked neutrophil infiltration and extensive fibrin deposition within the capillaries and venules of the dermis without any evidence of vasculitis.¹