The possibility that postmortem biochemical changes in blood might par
allel drug redistribution and thus serve as markers was explored in a
detailed case study. Eighteen blood and 14 tissue and fluid samples we
re taken at autopsy 16 h after the death of a 34-year-old female from
amitriptyline overdose. Ranges of drug concentrations in blood were am
itriptyline 1.8 to 20.2 mu g/mL, nortriptyline 0.6 to 7.3 mu g/mL, lev
els were lowest in femoral vein and highest in pulmonary vein blood. C
orresponding levels of 17 amino acids showed markedly different patter
ns of site-to-site variability. There was a strong positive correlatio
n between individual amino acid and drug concentrations in pulmonary b
lood samples (n = 5), particularly for glycine, leucine, methionine, s
erine, and valine. In blood samples from the great veins and right hea
rt (n = 10), the correlation was less strong (r = 0.6 to 0.7). Methion
ine showed a strong positive correlation in pulmonary samples (r = 0.9
3), and negative correlation in great veing samples (r = -0.68). Lacti
c acid showed a strong negative correlation in pulmonary samples (r =
-0.93) but a positive correlation in great vein samples (r = 0.71). Al
anine aminotransferase, alkaline phosphatase, aspartate aminotransfera
se, gamma-glutamyl transferase, glucose, and bilirubin had a weak posi
tive correlation with drug levels in great vein samples but not pulmon
ary samples. The results suggest that hepatic enzymes are relatively p
oor markers for postmortem hepatic drug shifts but that amino acids, p
articularly methionine, may be useful markers for pulmonary drug shift
s.