Recently there have been some local media reports questioning the safety of Evista^® (raloxifene hydrochloride), Lilly’s new drug for the prevention of postmenopausal osteoporosisThese recent allegations are based on unsubstantiated claims and an uninformed interpretation of findings in rodents (rats and mice). 
The findings in rodents do not translate into human health concerns for several important scientific reasons. First, rodents do not go through menopause. It is also well-established thatmice are especially susceptible to development of ovarian tumors. Because the animals were treated with raloxifene throughout their entire reproductive lives, hormonal changes occurred that are known to cause ovarian tumors in rodents. In contrast, raloxifene does not produce similar hormonal changes in postmenopausal women. Additionally, other marketed products, including one which has been on the market for more than 15 years, have produced similar effects in rodents, but have not been found to cause an increased risk of ovarian cancer in women. 
         The United States Food and Drug Administration (FDA), an independent FDA advisory committee of experts in the field, and Lilly Research Laboratories all rigorously reviewed the animal and human data for raloxifene prior to the product’s approval and came to the same conclusion: Evista is safe and effective for the prevention of osteoporosis in postmenopausal women. Lilly supports the integrity of the FDA’s review and approval process.

Compare with this extract from The Changing Years, Madeline Gray (1967)

Estrogens and cancer 
Now to the most serious question of all: Can the estrogens cause cancer? Because if they can, no matter what else they can or cannot do, they are ruled out.  Let me start by making a strong statement: The. general answer to the question is NO. Estrogen. alone cannot cause cancer in human beings. We don't know what can, but none of the available evidence to date points to estrogen alone. It must be estrogen combined with something else. 
 How you and I got confused about this subject is as follows: Some few years ago, around 1936 or before, a number of experiments were done on cancer with mice. Some young female mice of a special cancer-prone strain were injected with large quantities of estrogen; others were not. The mice receiving the hormones did develop a higher percentage of breast cancer than the mice that got none. Next, other experiments were performed backward. The ovaries were removed from some of the mice that had breast cancer. A certain percentage of the cancer did disappear. 

 Naturally people got tremendously excited. An excess of female sex hormones causes breast cancer, they cried. This proves that taking hormones is bad for mice, and probably for women as well." Well, it didn't prove anything of the kind. For the more cautious people soon remembered several things. First, they remembered that the mice used were a special cancer-prone strain of mice. Second, they remembered that the mice used were all young, or of the age equivalent to menstruating women, not equivalent to menopausal age. Third, the cancers were breast only, not cancers of the uterus or womb. (!!!!!!  so?  Tishy) And fourth, only a certain percentage of the treated mice developed cancer. So other factors unquestionably had to be involved. 
What could these other factors be? Heredity might be one. Some of the mice might have inherited the tendency to breast cancer through their mother's milk. When other experiments were done, sure enough-a hereditary "milk factor" was found to be involved. 
 Soon most scientists had become convinced that several factors had to be at work simultaneously to cause breast cancer in mice. Excess hormone stimulation was only one of them; the mice had also to be of a cancerous strain, and they had to receive the tendency through their mother's milk. Above all, still other research convinced them, animals are so different from humans that relating animal experiments to humans often produces false reports, especially in the field of hormones. What works in animals often does not turn out to work in humans at all.