To approve a drug means to release it for use in a specific patient population which should be characterized as clearly as possible in the labelled indications in the Summary of Product Characteristics (eg patients with major depression according to DSM III-R).
Ideally, the patient samples in various clinical trials should be representative of this specific patient population, and the crucial question is whether the study data allow an extrapolation to the wider patient population which we intend to treat.
Basically, what is valid for preclinical data and animal exposure, should also be valid for humans: more or less all regulatory authorities agree that they should accept valid data and information from all regions and all clinical studies available in order to minimize unnecessary human exposures and needless duplication. This is especially important in therapeutic fields where there is an urgent need for treatment due to the lack of suitable alternatives.