STUDY OBJECTIVE: To evaluate BAL cells obtained from habitual users of alkaloidal ("crack") cocaine alone or in combination with tobacco, for evidence of cocaine-associated alveolar injury. DESIGN: Prospective cohort study. PATIENTS: A total of 36 healthy men and women (mean age [SD], 37.5 [7.5] years), including 10 cocaine-only smokers (CS), 6 cocaine-plus-tobacco smokers (CTS), 10 tobacco smokers (TS), and 10 nonsmokers (NS), underwent fiberoptic bronchoscopy and BAL. METHODS: Cytospins were prepared from BAL cells and stained with Wright-Giemsa for cell differentials and Gomori's stain for detection of hemosiderin. Endothelin (ET)-1 levels were determined from lavage fluid by enzyme-linked immunosorbent assay. RESULTS: None of the cocaine users reported episodes of hemoptysis or respiratory distress, and routine spirometry findings were within normal limits in all subjects. While there was little effect on total cell numbers or differential counts, the percentages of hemosiderin-positive alveolar macrophages (AMs) were markedly increased in CS (33.8 +/- 8.7% [SEM]) compared to TS and NS (< 2%; p < 0.05). The percentages of hemosiderin-laden AMs were also numerically increased in CTS (11.8 +/- 7.8%), but this value was not statistically significant from that of TS or NS. ET-1 levels were significantly increased in the fluid recovered from CS (6.2 +/- 0.8 pg/mL) when compared to NS (1.2 +/- 0.4 pg/mL) and TS (1.3 +/- 0.2 pg/mL) [p < 0.05], while ET-1 levels were elevated to a lesser extent in CTS (2.5 +/- 0.6 pg/mL). ET-1 levels correlated with the percentage of hemosiderin-positive AMs when CS were analyzed in conjunction with CTS (r = 0.64; p = 0.0004). CONCLUSION: Clinically inapparent alveolar hemorrhage occurs frequently in otherwise healthy crack cocaine smokers and is associated with elevated levels of ET-1, indicative of cocaine-induced pulmonary microvascular injury.

Although cocaine-induced constriction of cerebral vessels may play an important negative role in the pathogenesis of cocaine-related stroke, the mechanism underlying the vasospasm remains unclear. This study investigated the role of endothelin-1 in mediating the spasm. Intracisternal cocaine infusion (10 µl/h via osmotic pump) into the cisterna magna of rabbits induced time- and concentration-dependent spasm. Maximal spasm was achieved with 100 µM cocaine infusate, and was observed as early as 0.5 days and reached a maximum at 2 days. Coinfusion of 100 µM cocaine with the endothelin receptor antagonist PD145065 (100 µM) prevented the spasm. Cerebral spinal fluid levels of cocaine and benzoylecgonine, a major cocaine metabolite, were below the limit of assay detection. This study demonstrates the novel finding that endothelin-1 mediates cocaine-induced cerebral vasospasm.