The role of cocaine in heroin-related deaths. Hypothesis on the interaction between heroin and cocaine.

In recent years, there has been an increase in cocaine-related deaths at the Department of Legal Medicine and Public Health of Pavia, probably reflecting the rising trend in cocaine use in Western Europe. Deaths from cocaine alone have increased from 6 cases in 1979-1991 (1.5% of drug-of-abuse deaths) to 13 in 1992-2002 (3.2%) and comparing the same periods, heroin-related deaths (HRDs) involving cocaine more than doubled from 8 (1.9%) to 22 (5.4%). In an attempt to investigate the role of cocaine in HRDs, acute narcotic death cases testing positive for cocaine use (blood cocaine or metabolite concentration >0.01 mg/l, COC+) were examined. Only cases from 1997 to 2001 were considered as in this period all data were obtained using the same analytical procedures (free morphine and total morphine by DPC Coat-A-Count radioimmunoassay before and after enzymatic hydrolysis, cocaine and metabolites in blood by SPE, TMS derivatization and gas chromatography-mass spectrometry (GC-MS)). The median, minimum and maximum concentrations of free morphine in blood (FM) and total morphine in blood (TM), urine (UM) and bile (BM) in the COC+ group (n = 9) were compared with those calculated in the group of "pure" HRDs (no other drugs detected in blood, COC-, n = 30). Differences among the medians in the two groups were statistically evaluated using the two-tailed Mann-Whitney U-Test. Statistical analysis was also carried out including in both groups cases with a blood alcohol concentration (BAC) > 20 mg/L (COC+, n = 19; COC-, n = 76). For the COC+ group, median TM was lower (0.32 mg/l versus 0.90 mg/l, P = 0.0214), median FM was lower, but not statistically significant (0.08 mg/l versus 0.28 mg/l, P = 0.1064), FM/TM ratio was similar (0.33 and 0.35), UM was also similar (21.0 mg/l and 18.0 mg/l), and BM was higher (90.0 mg/l versus 49.0 mg/l, P = 0.0268). Similar comparison results were obtained by repeating statistical analyses after including in the two groups cases with positive BAC. The picture observed for HRD cases involving cocaine is very different from what was previously observed for HRD cases involving ethanol [A. Polettini, A. Groppi, M. Montagna, The role of alcohol abuse in the etiology of heroin related deaths: evidence for pharmacokinetic interactions between heroin and alcohol, J. Anal. Toxicol. 23 (1999) 570-576], and updated with more recent data; in the high-ethanol (HE, BAC > 1000 mg/l) group, TM was lower than in the low-ethanol (LE, BAC <or= 1000 mg/l) group (0.59 mg/l versus 0.90 mg/l, P = 0.0180), FM/TM ratio was higher (0.66 versus 0.43, P = 0.0038), FM was equal, UM was lower (0.21 mg/l versus 13.0 mg/l, P = 0.0001), as was BM (10.0 mg/l versus 26.5 mg/l, P = 0.0156). The observed lower TM in the COC+ group supports the hypothesis of an interaction of cocaine with heroin in the cause of death, as previously observed for the heroin-ethanol interaction. However, ethanol results suggested that a pharmacokinetic interaction is prevalent (inhibition of heroin metabolism, as suggested by the increased FM/TM ratio, resulting in reduced urinary and biliary excretion). In the case of cocaine, a pharmacodynamic interaction seems to prevail, as the FM/TM ratio remains unchanged and UM and BM are not lower in the COC+ group. The hypothesis of a pharmacodynamic interaction of cocaine with heroin could not be confirmed owing to the paucity of data and the many uncontrolled variables involved.