- Top of page
- Patients and Methods
- AUTHOR CONTRIBUTIONS
Summaries of the clinical history, physical examination, and results of laboratory investigations for the 8 patients are shown in Tables 1, 2, and 3, respectively.
Table 1. Clinical history and disease course*
| ||Patient 1||Patient 2||Patient 3||Patient 4||Patient 5||Patient 6||Patient 7||Patient 8|
|History|| || || || || || || || |
| Age, years (sex)||33 (F)||46 (F)||44 (F)||48 (F)||45 (F)||32 (F)||39 (F)||43 (M)|
| Route of cocaine use (duration, years)||N/A||Past IV, snort (25)||Snort (10)||Snort (32)||N/A||Smoke (2.5)||Smoke (4)||N/A|
| Medical history||Hepatitis C||SLE, hepatitis C||SLE||Sweet's syndrome|| ||Hepatitis B, C|| || |
|Disease course|| || || || || || || || |
| Treatments||Prednisone, antibiotics||Prednisone, antibiotics||Antibiotics||Prednisone, antibiotics||Prednisone, antibiotics||Antibiotics||Antibiotics||Antibiotics|
| Followup, months||11||8||4||16||11||10||10||3|
| No. of admissions||1||5||5||2||1||5||2||1|
| Current disease (cocaine use)||Resolving (stopped)||Diseased (using)||Active (using)||Resolving (stopped)||Diseased (using)||Active (using)||Improving (stopped)||N/A|
Table 2. Physical examination findings*
| ||Patient 1||Patient 2†||Patient 3†||Patient 4‡||Patient 5||Patient 6||Patient 7||Patient 8|
|Skin|| || || || || || || || |
| Mouth||Ulcer||Papule, ulcer|| || ||Ulcer|| ||Aphthae||Ulcer, bulla|
| Ear lobe||Erythema||Purpura, necrotic|| || || ||Purpura|| || |
| Face|| ||Cheek: purpura|| || || ||Nose: purpura||Papule|| |
| Extremities||Purpura, pustule, livedo reticularis, necrotic ulcer||Purpura, livedo reticularis||Hemorrhagic bullae, necrotic ulcer||Livedo reticularis, purpura, ulcers, hemorrhagic bullae||Purpura, hemorrhagic bullae||Bullae, petechia, papule||Pap, pustule||Macule–papule, necrotic ulcers, petechia, hemorrhagic bullae|
| Fingers||Tender papule (dorsal)||Tender papule (dorsal)|| || || ||Tender papule (dorsal)||Tender nodule (pads)|| |
|Other|| || || || || || || || |
| General||Fatigue, weight loss||Fatigue, weight loss||Fatigue||Fatigue, low-grade fevers||Fatigue||Fatigue||Fatigue, weight loss|| |
| Arthritis (duration)||Ankles, feet, wrists, knees (7 mos)|| ||Most joints (3 yrs)|| ||Hands, wrists, knees (N/A)||Hands, ankles (6 mos)||Wrists, knees, ankles (1 mo)||Ankles, feet, knees, hands (4 mos)|
| Other||UAO||Pneumonia, UAO||Lymphadenopathy (reactive)|| ||Pneumonia||Pneumonia, pericardial and pleural effusions|| || |
Table 3. Laboratory investigations*
| ||Patient 1||Patient 2†||Patient 3†||Patient 4||Patient 5||Patient 6||Patient 7||Patient 8||Normal value|
|ANCA|| || || || || || || || || |
| cANCA||+||+||+||+||–||+|| |
| pANCA||+||+||+||+||+||+||+||+|| |
| Anti-PR3‡||–||>100 KU/liter||300 AU/ml||400 AU/ml||–||8.9 KU/liter||8 AU/ml||100 AU/ml||<3.5 KU/liter, <99 AU/ml|
| Anti-MPO‡||22 KU/liter||>100 KU/liter||>100 AU/ml||400 AU/ml||–||>100 KU/liter||18 AU/ml||300 AU/ml||<9 KU/liter, <99 AU/ml|
|aPL|| || || || || || || || || |
| LAC||+§||+||§||+||§|| |
| aCL IgG, units/ml||–||–||–||–||–||–||–||+ (19)§||<15|
| aCL IgM, units/ml||+ (64)||+ (23)§||–||+ (15)||+ (34)§||+ (15)§||–||+ (45)§||<12|
| β2GPI IgG, units/ml||–||N/A||N/A||–||–||N/A||N/A||–||<10|
| β2GPI IgM, units/ml||13‡||N/A||N/A||–||–||N/A||N/A||17§||<10|
|Other investigations|| || || || || || || || || |
| ANAs||1:320 (homogenous)||1:160 (homogenous)||1:160 (speckled)||–||1:320 (homogenous)||1:160 (homogenous)||1:160 (homogenous)||–|| |
| Anti-dsDNA, IU/ml||–|| (30)||+ (107)||N/A||+ (41)||+ (41)||–||+ (53)||0–30|
| Complement||Normal|| C3, C4|| C3, C4||Normal||Normal||Normal||Normal||Normal|| |
| Cold agglutinins||4+||4+||N/A||4+||N/A||4+||4+||2+|| |
| ESR, mm/hour|| (17)|| (80)|| (60)|| (43)|| (82)|| (39)||Normal||Normal|| |
| CRP level, mg/liter|| (18)|| (123)|| (>190)|| (21.1)|| (111)|| (30)|| (59.3)|| (107)|| |
| CBC (nadir values, × 109/liter)|| lymphocytes 0.3|| WBCs 1.2|| lymphocytes 0.9|| WBCs 2|| lymphocytes 0.8|| PMNs 1.9|| WBCs 2.4|| lymphocytes 0.88|| |
| || || PMNs 0.3|| || PMNs 0.65|| || lymphocytes 0.7|| PMNs 0.67|| || |
| || || lymphocytes 0.6|| || lymphocytes 0.74|| || || lymphocytes 0.8|| || |
| || || platelets 108|| || || || || || || |
| Biopsy||Thrombotic microangiopathy||Leukocytoclastic vasculitis||N/A||Thrombotic microangiopathy¶||Thrombotic microangiopathy||Thrombotic microangiopathy||Neutrophilic infiltrate||Neutrophilic infiltrate|| |
Relevant medical history included a previous questionable diagnosis of systemic lupus erythematosus (SLE) in 2 patients. This was based on positive serology and arthralgias in both patients, in addition to idiopathic thrombocytopenic purpura in patient 2 and a rash (the details of which are not available) in patient 3. In addition, patient 4 was previously diagnosed with Sweet's syndrome, patients 1 and 2 had chronic hepatitis C, and patient 6 had chronic hepatitis B and C.
In most patients, other symptoms preceded the appearance of the skin lesions. These included nonspecific constitutional symptoms in most patients and prominent weight loss in 3 patients (patients 1, 2, and 7). Six patients had overt arthritis. The ankles, wrists, hands, and knees were most commonly affected. One patient (patient 3) had diffuse lymphadenopathy that was found to be “reactive” on biopsy. Three patients (patients 2, 5, and 6) were treated for pneumonia at the time of presentation, 1 of whom (patient 6) also had pleural and pericardial effusions. In addition, 2 patients had episodes of acute upper airway obstruction, with 1 requiring emergent placement of a tracheostomy tube. The cause for this was a vocal cord fold with a nonspecific pathology in 1 patient. No structural abnormality was found in the other patient.
Most patients were unwilling to discuss the details of their cocaine use, so this information was only available for 5 patients. The duration of cocaine use was between 2.5 years (patient 6) and 32 years (patient 4). All of the patients either smoked or snorted the cocaine, and 1 patient (patient 2) also admitted to previous intravenous cocaine use.
On physical examination, all 8 patients had a mixture of different types of skin lesions (Figures 1 and 2). Five patients (patients 1, 2, 5, 7, and 8) had oral ulcers of variable morphology. Three patients (patients 1, 2, and 6) had lesions on their ears, and 3 patients (patients 2, 6, and 7) had involvement of their face. All of the patients had involvement of the extremities and/or torso: 3 (patients 1, 2, and 4) had livedo reticularis, 4 (patients 1, 2, 4, and 5) had purpuric lesions, 4 (patients 1, 2, 3, and 4) had ulcers of various depth, 5 (patients 3, 4, 5, 6, and 8) had bullous lesions, 4 (patients 1, 6, 7, and 8) had papular and/or pustular lesions, and 4 (patients 1, 2, 3, and 8) had areas of skin necrosis. In addition, 3 patients (patients 1, 2, and 6) had characteristic tender erythematous papules on the extensor surface of their fingers (Figure 2), and 1 patient (patient 7) had tender erythematous nodules on her finger pads. Six patients had active synovitis in at least 1 joint.
Figure 1. Photographs of mucocutaneous findings in patient 2. A, necrotic lesion of the ear, B, punched-out mouth ulcer, C, purpuric lesion on the right arm, D, livedo reticularis on the left arm.
Download figure to PowerPoint
Figure 2. Photograph of the dorsal aspect of a hand in patient 1. Note erythematous papules (black arrows) and scabs at the place of old lesions (white arrows).
Download figure to PowerPoint
These patients had very characteristic laboratory abnormalities. All had elevated inflammatory markers. All displayed a perinuclear immunofluorescence pattern of antineutrophil cytoplasmic antibodies (pANCAs), and 7 of 8 had antimyeloperoxidase (MPO) antibodies. Cytoplasmic ANCA (cANCA) was positive in 5, equivocal in 2, and negative in 1 patient. Anti–proteinase 3 (anti-PR3) antibodies were found in 6 of 8 patients. Most (6 of 8) had positive antinuclear antibodies (ANAs; homogenous or speckled pattern), with titers between 1:160 and 1:320. Five patients had positive anti–double-stranded DNA antibodies, only 1 of which was at a high titer (patient 3). Levels of C3 and C4 complements were low in 2 of these patients (patients 2 and 3), both of whom had a previous diagnosis of SLE. Most patients had at least 1 antiphospholipid antibody (aPL), although in many instances the test was performed only once. A lupus anticoagulant was positive in 3 patients and equivocal in the remaining 5. One patient had anticardiolipin IgG at a low titer, 6 had anticardiolipin IgM (low to medium titer), and 2 patients had a low-titer β2-glycoprotein I IgM. Cold agglutinins were measured in 6 patients, and were strongly positive in all of them. Six patients were anemic (patients 1–6). At some point over the period of observation, 3 patients had leukopenia, all 8 patients had lymphopenia, and 4 were neutropenic. Biopsies of the skin lesions were performed in 7 patients (Figure 3). This showed a leukocytoclastic vasculitis in 1 patient, thrombotic microangiopathy in 3, and neutrophilic dermatosis in 2. One patient (patient 4) had a first biopsy sample showing neutrophilic dermatosis (which was diagnosed as Sweet's syndrome), followed by a repeat biopsy 9 months later that showed thrombotic microangiopathy.
Figure 3. Photographs of different types of pathology specimens. A, Sample from patient 1 showing thrombotic microangiopathy. Note the thrombosed small blood vessels (open arrows) and areas of hemorrhage (double arrows). B, Sample from patient 7 showing neutrophilic infiltrate. Note areas of necrosis (open arrows) and numerous neutrophils (solid arrows), C, Sample from patient 2 showing leukocytoclastic vasculitis. Note the appearance of a normal blood vessel (solid thick arrow) and a vessel involved with leukocytoclastic vasculitis (open arrow), with numerous inflammatory cells (thin arrows) and an occluding thrombus (*).
Download figure to PowerPoint
All 8 patients required at least 1 hospital admission due to the severity of their skin disease. Two patients required 2 admissions, and 3 patients were admitted 5 times. Many of these readmissions were related to patients leaving the hospital against medical advice. At some point in their disease, 4 patients were treated with oral prednisone and all patients received antibiotics.
Seven of the 8 patients were seen on more than one occasion. Two patients (patients 3 and 6) continue to use cocaine and have ongoing skin disease. Three patients (patients 1, 4, and 7) appear to have stopped cocaine use and their skin disease is resolving. Two patients died from cardiac arrest, in both cases thought to be related to substance overdose. Both of these patients had progressive skin disease related to ongoing cocaine use prior to their death.
- Top of page
- Patients and Methods
- AUTHOR CONTRIBUTIONS
We describe a series of 8 patients with cocaine-related cutaneous vasculopathy. All of these patients have characteristic clinical features and laboratory investigations, suggesting the presence of the same clinical syndrome in all of these patients.
It is important to note that the cutaneous biopsy specimen from patient 2 showed mild leukocytoclastic vasculitis along with vessel thrombosis and necrosis. Mild leukocytoclasis has been described in a patient diagnosed with cocaine-associated vasculopathy previously (2). Although the term cocaine-related cutaneous vasculitis may be more appropriate in this setting, we have included this patient in this series because the similarity of clinical features and laboratory findings to other patients described here suggests a common pathologic process.
Case reports describe cocaine use in association with a wide range of rheumatologic complications (2). Due to the difficulties inherent in studying the population of cocaine users, precise epidemiologic data such as prevalence or incidence of the abovementioned rheumatologic conditions in cocaine users are not known. This makes it difficult to determine whether the frequency of occurrence of these conditions differs in cocaine users from the general population.
Several case reports have previously described cutaneous vasculopathy in association with cocaine use. Friedman and Wolfsthal (6) described a patient with papular and pustular lesions and ulcers on the face, legs, arms, and back. Laboratory investigations included elevated inflammatory markers, positive ANA (1:640), and positive pANCA associated with antibodies to PR3. Her skin biopsy sample showed a neutrophilic infiltrate. The patient died 2 years after presentation from narcotic intoxication. Bhinder and Majithia (2) described a chronic cocaine user who presented with palpable purpura and blisters distributed in a reticular pattern on the upper and lower extremities associated with fatigue, arthritis, and a dry cough. In contrast to our patients, this patient's laboratory investigations were normal. Her biopsy sample showed the presence of fibrin thrombi and mild leukocytoclasis. She was treated with oral prednisone with clinical improvement.
It is not clear what causes the abovementioned complications in cocaine users. Some may be the consequence of chronic cocaine use itself, through its vascular effects and the effects of the associated aPL (7). Other complications could be caused by one or more contaminants of cocaine. Street market cocaine is frequently adulterated or “cut” with various powdery fillers, such as baking soda, various sugars, or other stimulants, to increase its weight (8). Recently, various complications of cocaine have been linked to levamisole, an antihelminthic drug belonging to the class of imidazothiazole derivatives (9). Levamisole was previously used as an immunomodulator for certain malignancies and inflammatory conditions. In a series of children treated for nephrotic syndrome, levamisole was associated with the development of vasculitic skin lesions particularly involving the ear lobes and the face, as well as positive ANCA, ANA, and aPL (10). In addition, a case of cutaneous necrotizing vasculitis was described in a woman receiving levamisole for breast cancer (11). It is estimated that up to 70% of street cocaine in parts of the US and Canada is contaminated with levamisole (12). Two cocaine users presenting with purpuric and necrotic lesions of the ear lobes and cheeks (in addition to the lower extremities in 1 of the patients) have recently been described (12). Also, both of these patients had neutropenia, positive anticardiolipin IgM antibodies, and positive pANCA. Levamisole was implicated as the causative agent, although the presence of levamisole in the patients' urine was not confirmed (12). A case series presented in abstract form at the 2010 American College of Rheumatology Annual Scientific Meeting describes 5 cocaine users with purpuric skin lesions affecting the ears and extremities. In 2 of these patients the presence of levamisole was confirmed biochemically. The authors of this series believe that levamisole is responsible for the clinical presentation of their patients (7). Levamisole has a short half-life of approximately 5 hours, and it may become undetectable in the urine in as little as 48 hours after exposure (7). Levamisole has also been associated with bone marrow suppression. A recent report describes neutropenia in 2 patients who used cocaine contaminated with levamisole (9).
Most of the laboratory abnormalities seen in our 8 patients have been previously described in patients who use cocaine. Such patients often have elevated inflammatory markers and positive ANA (6, 7). Cocaine users are more likely to produce aPL (13). It is generally believed that patients with the combination of aPL and history of cocaine use are at a particularly high risk of thrombotic complications (13, 14). In addition, it was suggested that IgM anticardiolipin antibodies in particular predispose cocaine users to medical complications (13). In the case series presented by Ullrich et al (7), 4 of 5 patients had elevated levels of IgM anticardiolipin antibodies. The mechanism of aPL production in cocaine users is not clear. More recently, an association between cocaine use and ANCA, antibodies usually found in primary ANCA-associated vasculitis (AAV), is being increasingly recognized. In the reported literature to date, cocaine users are most often described to express pANCA, usually with specificity for the anti-MPO antigen (3, 6, 12, 15). However, with ANCA related to cocaine use, inconsistency between the ANCA immunofluorescence pattern and the target antigen measured by enzyme-linked immunosorbent assay is frequently observed, most commonly as a positive pANCA immunofluorescence with reactivity against the PR3 antigen (4, 6). It has also been suggested that drug-related ANCAs may be directed against more than 1 antigen (15), and in the case series by Ullrich et al, 4 of 5 patients had both MPO and PR3 antibodies (7). In contrast, the co-occurrence of cANCA and pANCA immunofluorescence patterns has not been described in association with cocaine use. In our report, 5 of 8 patients displayed both pANCA and cANCA immunofluorescence patterns as well as reactivity to both MPO and PR3 antibodies.
CIMDL is a distinct entity that frequently mimics GPA, the most common AAV. Patients with CIMDL are usually positive for pANCA that reacts with human neutrophil elastase (HNE) or, less frequently, with PR3, but none of these patients seem to react with MPO (3–5). Such specificity of anti-MPO to cutaneous vasculopathy and not CIMDL suggests (although it does not prove) a different pathogenesis of these 2 conditions, despite both of them being triggered by cocaine use. On the other hand, whereas normal controls and patients with GPA do not react with HNE (3, 5), 2 of 2 patients with cutaneous vasculopathy but no CIMDL reacted to HNE in a case series by Ullrich et al (7). None of our patients had CIMDL, and we did not test our patients for the presence of antibodies to HNE.
Of note, the association between cocaine use and cold agglutinins, found in 100% of tested patients in our series, was not described previously. The clinical significance of this finding, however, is not known.
The optimal treatment for cocaine-related vasculopathy is not known. All 8 patients in this series were treated with antibiotics, and 4 patients were given oral prednisone at some point in the course of their disease. In most instances, steroids were given either due to concern for active previously diagnosed SLE (patient 2) or due to concern for the presence of a systemic vasculitis. It is generally thought that abstinence from cocaine is crucial for resolution of clinical symptoms of cocaine-related vasculopathy (7, 12). Immunosuppression, most commonly oral steroids, may be necessary for severe disease (12).
Notably, 2 of our patients had a previous diagnosis of SLE, and 1 was diagnosed with Sweet's syndrome 1 year prior to the diagnosis of cocaine-related vasculopathy. All 3 of these patients were using cocaine at the time these diagnoses were made. Although it is difficult to make definitive conclusions retrospectively, it is possible that the positive lupus serologies, arthralgias, and the rash in patients diagnosed with lupus, as well as skin pathology demonstrating neutrophilic infiltrate consistent with Sweet's syndrome, were in fact early manifestations of cocaine-related vasculopathy.
Importantly, 2 of 8 patients described in this report died as a result of their cocaine use within 1 year of diagnosis of their skin condition. In addition, one of the previous case reports described a similar death of their patient 2 years after presentation (6). It is possible that cocaine-related vasculopathy is associated with increased mortality. However, it is more likely that the prolonged heavy cocaine use puts these patients at risk for other complications of cocaine use, which could be more life threatening than skin vasculopathy.
In this case report we describe a series of 8 patients who used cocaine, all presenting with a similar clinical syndrome. These patients also had a characteristic set of laboratory abnormalities, including elevated inflammatory markers, strong positivity for cold agglutinins, and the presence of pANCA in all patients. In most of these patients, the pANCA was specific to the anti-MPO antigen, and often coexisted with cANCA with specificity to anti-PR3. The majority of these patients also had ANA, aPL (particularly anticardiolipin IgM), and at least 1 cytopenia. The precise mechanism of such a characteristic clinical picture could be related to the pharmacokinetics of cocaine itself, or of one of the cocaine adulterants such as levamisole. It may be difficult to distinguish cocaine-related vasculopathy from primary rheumatologic conditions, particularly systemic vasculitides and SLE. One needs to be particularly cautious of the fact that these conditions may coexist, and a positive history of cocaine use should not preclude the clinician from searching for potentially life-threatening rheumatologic conditions. However, in the setting of the high prevalence of cocaine use in certain geographic areas, it is important to recognize this constellation of clinical features and laboratory abnormalities as being potentially caused by cocaine use. This may eliminate inappropriate treatment with immunosuppressive medications in place of the appropriate, and potentially life-saving, intervention—abstinence from cocaine use.