Should we routinely measure free plasma phenytoin concentration?

The plasma protein binding of phenytoin was investigated in 56 epileptic patients attending the outpatient clinic. The free phenytoin fraction was measured by equilibrium dialysis at 37 degrees C and the total concentration by a homogenous enzyme immunoassay technique. The free fraction ranged from 0.123 to 0.177 (median 0.144, mean +/- s.d. = 0.145 +/- 0.12). Distribution was consistent with normality. Four of the patients were also taking sodium valproate. The median free fraction of phenytoin in these patients was 0.174, 21% higher than that of the total group (P less than 0.05). The total concentration of phenytoin varied from 0.3 to 29.4 micrograms/ml (median 12 micrograms/ml, mean +/- s.d. = 13.31 +/- 6.13 micrograms/ml) and the free fraction was not related to the total drug concentration. There was a highly significant relationship between free phenytoin concentration and total phenytoin concentration (r = 0.986, P less than 0.001). There appears to be very little variability in protein binding of phenytoin in epileptic patients and thus total plasma phenytoin concentration closely reflects the free (unbound) drug concentration. Routine estimation of free plasma phenytoin concentration is therefore unnecessary and should be reserved for those patients where alteration in binding is likely, e.g. renal or hepatic disease or where adverse effects occur at unexpectedly low total phenytoin concentrations.