The amount of drug remaining after previous doses, or drug accumulation, is closely related to drug efficacy and safety. An
accurate calculation of the accumulation index or ratio (Rac) is crucial for dose finding. However, in drug accumulation studies little consensus
exists with regard to experimental design or data analysis. We conducted a systematic review of the literature to produce a detailed
profile of drug accumulation studies of the last 30 years (1980-2011). Ninety-six articles comprising 122 studies were analyzed. A typical
drug accumulation study enrolled 10 to 20 subjects randomly assigned into treatment groups of 1 or 2 dose levels to observe pharmacokinetic
behaviors. The median washout period between single and multiple dosing was 7 days, and the dose interval was 1-2 elimination
half-lives in non- or one-compartmental models. Generally, the number of repeated times of administration for multiple dosing was 7-14,
and the median number of sampling time points was 11. Eight different methods were used to calculate Rac. The most frequently used
method, in 72.9% of the studies, was to set Rac equal to the ratio of the area under a plasma concentration-time curve (AUC) during a
dosage interval at steady state to the AUC of a dosage interval after the first dose, i.e., Rac = AUC0-ι,ss / AUC0-ι,1. The values of Rac in the
included studies ranged from 0.85 to 18.8, and 68.03% were <2. We suggest that sample size estimation for an accumulation study should
be similar to that of a bioequivalence study, and in most studies, 18-24 subjects will be needed.
Appropriate calculation methods for Rac should be selected based on the experimental design and data characteristics. The crucial values
for non-, weak, moderate, and strong accumulation can be set at Rac < 1.2, 1.2 ≤ Rac < 2, 2 ≤ Rac < 5, and Rac ≥ 5, respectively. Accumulations
studies should also give more regard to drug metabolism and increased accumulation in kidney or liver damaged patients.