When i was studying PG at IPGTRA , Jamnagar, In 2003, animal experimental study using Kuberaksha [Caesalpinia bonduc (L.) ], the findings were hyperglycemic in animals and in humans hypoglycemic
there are reports where results of clinical study differs from animal study and vice versa. however, this type of study should be repeated to get and final conclusion taking into naimal model and clinical conditions in to considerations.
Although the animal species may be mammalian, there are differences in physiology that account for contradicting results of experiments in animal species and clinical trials in humans.
This precisely is the reason for the phase I studies after proper authorization from regulatory authority - to safeguard both the investigator and the trial subjects.
What often happens is that drugs seem promising in animal studies and then fail clinical trials. This may be due to many reasons, the most common is probably toxicity. In animal you (almost) only look at systemic toxicity (weight loss) and haematological toxicity. Other toxic side effects may become evident in clinical trials.
I am convinced that careful analysis of animal experiments and inclusion of pharmacodynamic parameters might reduce the number of molecules that proceed from preclinical to clinical trials....just to fail there!