A 65-year-old man with longstanding seropositive rheumatoid arthritis and lower extremity ulceration



Chief symptom

A 65-year-old man with longstanding seropositive erosive rheumatoid arthritis (RA) who had recently started taking adalimumab was admitted for evaluation of right foot ulceration.

History of the present illness

The patient reported developing right foot pain and blue discoloration of the toes 2 months prior to the present admission. At that time, he was in a rehabilitation facility following a myocardial infarction that was treated with endovascular stenting of a coronary artery bypass graft. The foot pain was localized to the plantar surface of the right foot and was constant and stabbing in nature. He denied numbness, tingling, or burning of the lower extremities.

He was evaluated by a podiatrist and underwent steroid injection for presumed plantar fasciitis. However, following the steroid injection he developed progressive ulceration and gangrene of this region. He denied fevers, chills, shortness of breath, sicca symptoms, skin rash, or weight loss. His RA was stable, with approximately 1 hour of morning stiffness in the hands and mild swelling in the metacarpophalangeal (MCP) joints. His exercise tolerance was limited due to the foot pain.

Past medical history

The patient had erosive RA dating back to the 1970s, with a positive rheumatoid factor and anti–cyclic citrullinated peptide. Previous records also revealed a positive antinuclear antibody at a titer of 1:1,280 with a homogenous pattern. Prior treatments had included hydroxychloroquine and methotrexate. Due to refractory disease, adalimumab was added 4 months prior to the myocardial infarction, but following the myocardial infarction this was discontinued.

His medical history also included coronary artery disease with bypass surgery in 1995, peripheral vascular disease, and dyslipidemia. He did not have diabetes mellitus.

Family history and social history

Prior to the recent myocardial infarction, the patient had a 40 pack-year smoking history. He did not drink alcohol. He was a retired printer and resided with his wife. He had a family history of diabetes mellitus and cerebrovascular disease.


The medications on admission included aspirin 81 mg, atorvastatin 20 mg, clopidogrel 75 mg, esomeprazole 40 mg, metoprolol 100 mg twice daily, felodipine 5 mg daily, valsartan 160 mg daily, digoxin 125 μg daily, hydroxychloroquine 200 mg twice daily, and tamsulosin 0.4 mg daily.

Physical examination

Physical examination revealed an afebrile, cachectic man with a blood pressure of 106/60 mm Hg and a regular pulse at 90 beats per minute. He had no malar or discoid rashes. The cardiovascular examination was normal without murmurs, the pulmonary examination revealed bibasilar crackles, and the abdominal examination was benign.

There was mild synovitis of the bilateral wrists and the second and third MCP joints, along with radial drift at the wrist and ulnar deviation at the MCP joints. There were no rheumatoid nodules. Radiograph findings confirmed erosive disease.

There was a gangrenous ulcer on the heel of the right foot (Figure 1), but the dorsalis pedis and posterior tibialis pulses were biphasic. On the left foot there was a submetatarsal blister, the dorsalis pedis pulse was monophasic, and the posterior tibialis pulse was absent. There were no other skin rashes, livedo reticularis, splinter hemorrhages, or Bywater's lesions.

Figure 1.

Gangrenous ulcer on the heel of the right foot.

The neurologic examination revealed normal tone, power, and reflexes throughout all muscle groups and there was no evidence of foot drop. Proprioception, pinprick sensation, as well as sensation to light touch were intact.

Initial laboratory evaluation

On admission, the patient had a normochromic, normocytic anemia of chronic disease. His white blood cell count was 12.7 cells/mm3, the erythrocyte sedimentation rate was 60 mm/hour, and the C-reactive protein (CRP) level was 11.3 mg/dl, which is consistent with his gangrene. The creatinine level on admission was elevated at 2.4 mg/dl, but returned to within normal limits with intravenous hydration.

The patient had a positive rheumatoid factor and anti–cyclic citrullinated peptide. The antinuclear antibodies, anti-Sm and anti-RNP antibodies, Sjögren's antibodies, and antineutrophil cytoplasmic antibodies were negative. Hypercoagulable evaluation revealed an elevated homocysteine level of 13.5 μmoles/liter (normal range 4.3–11.4), along with heterozygous plasminogen activator inhibitor 1 gene mutation (Table 1).

Table 1. Laboratory results*
 Normal valueOn admissionDay 4Day 27
  • *

    AI = antibody index; pANCA = perinuclear antineutrophil cytoplasmic antibody; cANCA = cytoplasmic antineutrophil cytoplasmic antibody; ESR = erythrocyte sedimentation rate; CRP = C-reactive protein; MTHFR = methylenetetrahydrofolate reductase; PAI-1 = plasminogen activator inhibitor 1.

White blood cell count, cells/mm34.8–10.812.76.8 
Hemoglobin, gm/dl14–1812.110.6 
Hematocrit, %42–5236.332 
Platelet count, cells/mm3130–400229  
Blood urea nitrogen, mg/dl6–1962  
Creatinine, mg/dl0.5– 
Glucose, mg/dl70–10597  
Sodium, mmoles/liter135–145132  
Potassium, mmoles/liter3.5–55.2  
Protein, gm/dl6–8.36.8  
Albumin, gm/dl3.5–52.9  
Aspartate aminotransferase, IU/liter10–3736  
Alanine aminotransferase, IU/liter10–4021  
Alkaline phosphatase, IU/liter39–117103  
International normalized ratio 1.15  
Partial thromboplastin time, seconds23–3534.9  
Prothrombin time, seconds11.8–14.514.5  
Urinalysis Clean  
Antinuclear antibodiesNegative  Negative
SSA/Ro, AI< 0.1  < 0.1
SSB/La, AI< 0.1  < 0.1
Anti-Sm, AI< 0.1  < 0.1
Anti-RNP, AI< 0.1  < 0.1
pANCA, units/ml< 6  < 6
cANCA, units/ml< 6  < 6
C3, mg/dl86–184137 130
C4, mg/dl20–5947.2 35.6
ESR, mm/hour0–2060  
CRP level, mg/dl0.0–0.811.3  
Anticardiolipin antibody IgA, IgG, IgM, units/ml< 10  < 10
Protein S activity, %65–140  65
Protein C activity, %70–130  105
Antithrombin III activity, %87–123  87
Prothrombin gene mutationNegative  Negative
Factor V Leiden gene mutationNegative  Negative
MTHFR gene mutationNegative  Negative
PAI-1 gene mutationNegative  Heterozygous 4G/5G
Homocysteine, μmoles/liter4.3–11.4  13.5
CryoglobulinNegative  Negative
Cholesterol, mg/dl140–200  103
Triglycerides, mg/dl35–160  64
High-density lipoprotein, mg/dl40–80  37
Low-density lipoprotein, mg/dl< 130  53

Initial hospital course

The patient was taken to the operating room for a right partial calcanectomy. However, due to the presence of poor intraoperative blood flow, a below-knee amputation was performed.

On postoperative day 10, he developed a pulmonary embolism, and was started on enoxaparin 60 mg subcutaneously twice daily. A lower extremity Doppler ultrasound examination was negative for deep vein thrombosis. The following day, he developed mottled discoloration of the skin and toes of the left foot, and an arteriogram revealed a moderate-length near-total occlusion of the superficial femoral artery along with bilateral renal artery stenosis and atherosclerosis in the infrarenal aorta and left common iliac and femoral arteries. There was poor blood flow distal to the stenosis and in the peroneal and posterior tibial arteries. Endovascular recanalization of the superficial femoral artery was unsuccessful, but left femoropopliteal bypass reestablished flow to the left lower extremity.

Within 24 hours of the endovascular procedure, the right lower extremity amputation flap became dusky with the progressive development of gangrene despite patent vasculature and no evidence of infection. The rheumatology team was asked to evaluate for vasculitic processes.


A 65-year-old white man with longstanding seropositive erosive RA, coronary artery disease, and peripheral vascular disease presented with progressive lower extremity ulceration despite patent vasculature to the lower extremities.


The differential diagnosis in this case included rheumatoid vasculitis, peripheral arterial disease, drug-induced vasculitis or lupus-like syndromes related to adalimumab, coexistent prothrombotic state with small-vessel thrombosis, and cholesterol embolization related to the numerous endovascular procedures that were performed.

Rheumatoid vasculitis

Rheumatoid vasculitis was initially suspected in this patient. Rheumatoid vasculitis typically develops in patients with longstanding erosive disease (1) and it most commonly involves small and medium-sized vessels of the skin, digits, peripheral nerves, eyes, and heart. There may be associated systemic signs including fatigue, weight loss, and fever, but often the synovitis has quiesced by the time the vasculitis develops (2, 3).

More than 90% of cases of rheumatoid vasculitis involve the skin (4). Patients most commonly present with malleolar ulcers due to a medium-vessel arteritis. However, digital ischemia and gangrene may also occur.

Risk factors for the development of rheumatoid vasculitis include male sex, high-titer rheumatoid factor, joint erosions, pleuritis, subcutaneous nodules, and the presence of nailfold lesions (5). Patients will often have low complement levels. Our patient did not have rheumatoid nodules, splinter hemorrhages, or serositis, and his complement levels were normal.

Cardiac arteritis has been reported in rheumatoid vasculitis, and in this case the recent myocardial infarct prompted concern for cardiac involvement. However, myocardial infarctions secondary to rheumatoid vasculitis are rare (6) and the stenosed vessel in this patient was one of the prior bypass graft vessels.

Peripheral arterial disease

The pathogenesis of atherosclerosis involves abnormal retention of cholesterol in the endothelial wall, often with inflammation secondary to this (7). Atherosclerotic peripheral artery disease is more common in patients with RA than in healthy controls independent of other cardiovascular risk factors (8). Inflammatory markers, glucocorticoid use, and extraarticular features of RA are independent risk factors in this population (9).

In patients with connective tissue diseases, the etiology of lower extremity ulcers is often multifactorial, with peripheral arterial and venous disease, endothelial damage from chronic inflammation, and superimposed prothrombotic states all playing a role in the development of these lesions (10).

The patient described in this case report did have peripheral artery disease in the left leg (Figure 2); additionally, he was a smoker with coronary artery disease and poorly controlled RA, placing him at risk for endothelial damage. However, the presence of biphasic pulses in the right foot at presentation and intact blood flow to the amputation flap all suggested that another factor must be playing a role in the stepwise progression of the right extremity gangrene.

Figure 2.

Arteriogram of the left lower extremity showing occlusion of the superficial femoral artery.

Vasculitis related to tumor necrosis factor α inhibitor use

Adalimumab is a fully humanized monoclonal antibody against tumor necrosis factor α (TNFα). Three different TNFα inhibitors (etanercept, infliximab, and adalimumab) have been approved by the US Food and Drug Administration for the treatment of RA (11). Several reports suggest that TNFα inhibitors induce vasculitis (12–15), whereas others have reported a successful response of rheumatoid vasculitis to these drugs (16–18). The initiation of treatment with adalimumab just 4 months prior to presentation in this patient raised concern that it could have been involved in the illness.

Ramos-Casals et al reviewed 379 cases of documented autoimmune diseases associated with TNFα inhibitors, including 118 who developed vasculitis (19). Most vasculitis cases (86%) involved skin lesions, including purpura, ulcerative lesions, nodules, digital vasculitis, maculopapular rash, and chilblain lesions. Vasculitis appeared after a mean of 38 weeks of therapy and in most cases resolved with discontinuation of therapy. Adalimumab was implicated in only 4% of the cases in this study, and there were no reports of cardiac involvement. However, it is possible that the lower incidence of vasculitis with the use of adalimumab may simply reflect lower use of adalimumab relative to the other TNFα inhibitors.

Our patient had started treatment with adalimumab only 12 weeks prior to the myocardial infarct and subsequently discontinued it; therefore, it is unlikely that the medication played a causative role in the lower extremity ulceration.

TNFα-induced lupus syndrome

The TNFα inhibitors have also been associated with the development of a lupus-like syndrome (19), and this syndrome is reported more frequently with the use of infliximab and etanercept than adalimumab, although as with vasculitis, this may be attributable to differences in prescribing. Although lower extremity leukocytoclastic vasculitis may occur in association with drug-induced lupus, other clinical features of lupus such as fever, myalgias, rash, arthralgias, arthritis, and serositis were absent in this patient. Antihistone antibodies were not tested and repeat antinuclear antibody screens were negative.

Prothrombotic states

The role of prothrombotic states in leg ulceration in patients with connective tissue disease (10) is currently the focus of a longitudinal study. Our patient was screened for superimposed prothrombotic states found to have heterozygous plasminogen activator inhibitor 1 mutation and elevated homocysteine levels. Although these 2 risk factors may have contributed to the development of the pulmonary embolus, they are not typically associated with arterial thrombi.

Cholesterol crystal embolism

Cholesterol crystal embolization (blue toe syndrome) is a syndrome in which portions of atherosclerotic plaque embolize distally, occluding small arteries and causing tissue ischemia. Although this can occur spontaneously (20), it is often associated with invasive vascular procedures such as arteriograms and vascular surgery, and the inciting event may precede the onset of the clinical syndrome by days or weeks.

Clinical presentation of cholesterol emboli can be nonspecific. Constitutional symptoms including fevers, myalgias, headache, and weight loss occur in 21% of patients (21). Skin findings are reported in 34%, with the mottled purple pattern of livedo reticularis seen in 16%, gangrene in 12%, cyanosis in 10%, cutaneous ulceration in 6%, and blue toes occurring in 5% of patients. Most patients have patent proximal vasculature with detectable pulses. Embolism to renal, mesenteric, and iliac arteries can lead to ischemia of these organs, but coronary events and stroke are relatively rare.

Transient elevation of inflammatory markers along with hypocomplementemia and eosinophilia are sometimes seen. However, a definitive diagnosis requires a biopsy sample demonstrating cholesterol clefts in the small or medium-sized arteries or arterioles. These crescentic or elongated ovoid spaces are the result of dissolution of the cholesterol crystal during tissue fixation and are pathognomic of cholesterol embolization syndrome.


A diagnostic procedure was performed. The debrided tissue from the amputation site was reviewed in detail, and cholesterol clefts were identified in a small arteriole confirming the diagnosis of cholesterol crystal embolization syndrome (Figure 3).

Figure 3.

Skin biopsy sample showing cholesterol cleft in a small arteriole (hematoxylin and eosin stained).


Cholesterol crystal embolism is a well-recognized mimicker of vasculitis (22), which should be considered in all patients presenting with tissue ischemia following endovascular procedures, even if they have known autoimmune disease. Major risk factors include advanced age (>60 years), vascular procedures, and peripheral vascular disease (23). Based on a prospective observational study of 1,786 cardiac catheterizations, the rate of cholesterol embolization syndrome is estimated to be approximately 1.4%. However, this study was limited because it relied only on clinical manifestations (cutaneous findings or renal impairment) for diagnosis of cholesterol crystal embolism rather than requiring histologic confirmation. The authors also noted a 4.6-fold increase in the risk of cholesterol crystal embolism in association with elevated CRP levels, consistent with recent data implicating CRP as a biomarker of “inflamed, unstable” plaque (24). The prognosis of cholesterol crystal embolism is poor, with a 72% fatality rate due to concomitant visceral ischemia.

Some studies have implicated anticoagulation as a precipitant for cholesterol crystal embolism (25). However, in a study of 519 patients with severe aortic plaque, the rate of cholesterol emboli in those who were receiving warfarin was similar to those who were not (1%) (26).

The patient reported here developed livedoid skin changes and ischemia of the right foot 24 hours after the cardiac catheterization and endovascular manipulation. Progressive gangrene culminated in amputation. Subsequently, he developed a pulmonary embolus without detectable deep venous thrombosis. Cholesterol crystal embolism has been reported in the pulmonary vasculature, possibly via passage through the systemic capillary bed into the venous system and lungs (27). In our patient, anticoagulation was introduced secondary to the pulmonary embolus, and this may have predisposed to further plaque rupture during the second endovascular procedure and bypass surgery, resulting in acute ischemia of the previously viable amputation flap. The presence of cholesterol crystals in the necrotic tissue confirmed the diagnosis.

To our knowledge, there are no clinical trials evaluating interventions for cholesterol crystal embolism and treatment remains supportive. Statins, which stabilize and may cause regression of atherosclerotic plaques, improve renal and overall outcome (28, 29). Additionally, steroids (30, 31), iloprost (32), and low-density lipoprotein (LDL) apheresis (33, 34) have been beneficial in small numbers of patients.


The patient received 20 mg of prednisone, which was tapered over the subsequent weeks. LDL apheresis was considered, but was deferred due to clinical stabilization. Since the patient had prothrombotic risk factors and a recent pulmonary embolus, anticoagulation therapy was continued. The patient had no further ischemic events and was discharged to a rehabilitation facility.


Cholesterol crystal embolism.


All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Ms Shanmugam had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Charabaty, Shanmugam.

Acquisition of data. Charabaty.

Analysis and interpretation of data. Charabaty, Shanmugam.


We appreciate the assistance of Dr. Christopher Attinger and the Georgetown Center for Wound Healing in the management of this patient.