Our patient had no history of chronic repetitive trauma to the hands from any occupations or hobbies. Although a hypercoagulable state is unlikely to cause vasculitis, the patient had normal levels of homocystine and antithrombin III. During one of his previous admissions, lupus anticoagulant and factor V Leiden were not detected and both protein C and protein S levels were low because of warfarin therapy.

The purpose of our report was to highlight the role of cocaine in causing arterial diseases. Although vasculitis is an attractive explanation, the exact causal mechanism is unknown. The literature is sparse and consists mainly of cases of cerebral vasculitis proven by either angiography or biopsy (1, 2). We agree that it is extremely difficult to implicate cocaine as the sole cause of any vascular disease since most affected patients have polysubstance abuse or other predisposing conditions, such as diabetes mellitus and hypertension. Cocaine can also precipitate or exacerbate vascular abnormalities in such patients. We used the term cocaine-related rather than cocaine-induced to make this distinction.

Our patient did not meet the strict and well-defined criteria for Buerger disease, such as a significant smoking history, disease onset before age 50 years, infrapopliteal arterial occlusions, upper-limb involvement or phlebitis migrans, and absence of atherosclerotic risk factors other than smoking (3). Apart from the lack of these clinical criteria, the patient's history of relatively light smoking (15 pack-years over 35 years) and almost daily cocaine use would favor cocaine as the causal agent of vasculitis.

Of interest, a recent report suggested a possible pathogenetic association between cocaine and Buerger disease (4). Some cases of vasculitis related to cocaine might have been classified as idiopathic, and the recurrent venous thrombosis was thought to be related to cocaine abuse (5).