Percutaneous ethanol injection therapy under sonographic guidance suff
ers from the occasional adverse spread of ethanol. We tried to optimiz
e the technique to obtain the maximal echogenicity of injections. Eigh
t dead pig livers were injected with ethanol and the consequent echoge
nicity was correlated with the time lapse after the injection, the spe
ed of injection, the injected dose and the concentration of ethanol. F
or the sake of comparison, ethanol was also injected into processed so
ur milk. The injected lesions became smaller (P < 0.001), less sharply
demarcated and less echogenic (P < 0.001) with time. Speeding up the
injection (P < 0.05) and concentrating the ethanol (P < 0.05) increase
d echogenicity, whereas enlarging the dosage did not. Injections of et
hanol into processed sour milk induced hyperechoic zones similar to th
e liver injections. It is concluded that quick injections of concentra
ted ethanol would be best detected. Injections should be monitored whi
le they are still taking place, because the echogenicity will soon dec
rease. This immediate echogenicity seems to be caused by flow and the
inherent properties of ethanol.