Background/Purpose: Exposure to illicit cocaine and its frequent adulterant, levamisole, is associated with the development of serum anti-neutrophil cytoplasmic autoantibodies (ANCA) and a variety of hematologic, renal and skin abnormalities (e.g., thrombotic purpura). The mechanisms of cocaine/levamisole-associated autoimmunity (CLAA) are unknown but may involve drug-induced neutrophil activation. Most reports of CLAA have examined small cohorts (< 20 subjects) or have focused on select clinical abnormalities. In order to gain a fuller picture of CLAA and generate new hypotheses regarding pathogenesis, we analyzed a cohort of 51 CLAA subjects identified in a large, academic healthcare system between 2000 and 2015.

Methods: We queried an electronic database and rheumatology referral records for individuals with positive ANCA testing results (ANCA+) and evidence of cocaine exposure by urine toxicology (cocaine+). Individuals were excluded if they had biopsy-proven ANCA-associated vasculitis. Clinical and immunologic parameters were extracted from the electronic medical record. We compared rates of certain parameters with rates within the cocaine+ (n = 6,233) or ANCA+ (n = 310) populations in a de-identified clinical data repository (DCDR) using univariate (Chi square, Fisher’s exact) and multivariate (logistic regression) analyses. 

Results: Hematologic abnormalities were almost universal (94.1%), with anemia (84.3%), leukopenia (64.7%) and neutropenia (47.1%) the most frequently reported. Renal (74.5%), skin (58.8%) and musculoskeletal (37.3%) abnormalities were also common. When compared to the cocaine+ population, the CLAA group showed similar estimated age at first positive cocaine result (46.6 vs. 47.7 years), but it was more female (56.9% vs. 38.8%, p = 0.0016) and less African-American (19.6% vs. 37.5%, p = 0.0130). Among those tested, the CLAA group showed higher rates of anti-phospholipid Abs (ACLAs) (69% vs. 29.1%, p < 0.0001) and lupus inhibitor (LI) (80% vs. 34.9%, p = 0.0002) but not ANA (22.4% vs. 19.4%) or active HCV infection (46% vs. 52.7%). ANCA in CLAA subjects were generally high titer (often > 1:2048) and of P-ANCA pattern (60.8%). Compared to the ANCA+ population, the CLAA population had significantly higher rates of MPO/PR3 double-positivity (31.4% vs. 2.6%, p < 0.0001). Finally, multivariate logistic regression analysis identified a significant association in the CLAA population between anti-phospholipid Abs and skin abnormalities (OR 7.1, 95%CI 1.3 – 49.2). 

Conclusion: CLAA subjects showed a variety of previously reported hematologic, skin and MSK abnormalities. However, results from our demographic analysis suggest that, in cocaine users, the development of ANCA and/or clinical abnormalities prompting ANCA testing are influenced by sex and race. Moreover, exposure to cocaine in this population may also lead to development of LI and APLAs, which could contribute to skin abnormalities such as thrombotic purpura. Finally, high-titer P-ANCA with double-positivity may help clinicians differentiate CLAA from other forms of ANCA-associated disease.

« Back to 2016 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/immunologic-clinical-and-demographic-correlates-in-51-cocaine-users-with-serum-anti-neutrophil-cytoplasmic-antibodies/