A potential error in modifications of the [14C]2-deoxyglucose technique.

Although the technique developed by Sokoloff et al. employing [14C]2-deoxyglucose as tracer, provides measurements of local cerebral glucose utilization, various modifications of the original technique are presently in widespread use. We have examined the validity of a crucial simplifying assumption which is an integral feature of many of these modifications; namely, that all the radioisotope in the CNS at the moment of sacrifice is present as 2-deoxyglucose-6-phosphate. By employing the kinetic model developed by Sokoloff et al. and arterial plasma [14C]2-deoxyglucose and glucose histories, the proportion of total radioactivity present in the CNS as unphosphorylated 2-deoxyglucose was calculated and found to vary from 5 to 50% of the total amount of isotope present in the CNS under different experimental conditions. It can be shown to be dependent upon (a) the rate of cerebral glucose utilization, (b) the time which elapses from administration of the radioisotope until the sacrifice, (c) the route by which the isotope is administered, and (d) arterial plasma glucose levels. Thus, modifications of the original 2-deoxyglucose technique have introduced a potential source of error which is not a feature of the approach of Sokoloff and his associates.