Probabilistic approach to the establishment of maximal content limits of impurities in drug formulations: The case of parenteral diphenylhydantoic acid
D. Manca et al., Probabilistic approach to the establishment of maximal content limits of impurities in drug formulations: The case of parenteral diphenylhydantoic acid, REGUL TOX P, 29(1), 1999, pp. 1-14
Diphenylhydantoic acid (DPHA) is a degradation product in parenteral formul
ations of the anticonvulsant phenytoin and the prodrug fosphenytoin. DPHA h
as also been reported to be a minor metabolite of phenytoin. Levels found i
n the urine of various species, including humans, after oral or intravenous
(iv) phenytoin ranged from undetected to a few percent of administered dos
e. In the present analysis, the toxicologic profile of DPHA was integrated
with exposure data in order to characterize its safety under recommended cl
inical regimens of fosphenytoin administration. In preclinical safety studi
es, DPHA was without effect in the Ames assay and at concentrations up to 3
000 mu g/plate in the presence or absence of metabolic activation, and in t
he in vitro micronucleus test with acute and 2-week repeated dose studies i
n Wistar rats at iv doses up to 15 mg/kg. In 4-week studies conducted in ra
ts and dogs receiving fosphenytoin containing DPHA levels up to 1.1%, and i
n an in vitro structural chromosome aberration test with DPHA levels up to
2.0%, all findings were consistent with known effects of phenytoin (such as
CNS signs and increased liver weight), and none were attributed to DPHA. R
eports in the literature indicate that in murine in vivo and in vitro model
s, DPHA has much lower potential for reproductive toxicity than phenytoin.
A no-observed-effect level (NOEL) of 15 mg/kg established from the S-week s
tudy in rats was used with probabilistic techniques to estimate tolerable d
aily doses (TDDs) of DPHA. In this approach, interspecies correction was pe
rformed by allometrically scaling the NOEL based on a distributional power
of body weight while intraindividual variability was accounted for by selec
ting the lower percentiles of the population-based distribution of TDDs. Th
e results indicate that a DPHA content limit of 3.0% in an administered dos
e of fosphenytoin is unlikely to cause adverse effects in patients, (C) 199
8 Academic Press.