Answer:
Ovarian tumors have been studied most extensively because of initial alarming reports based upon case reports. Case reports are subject to significant biases and larger, better-designed studies have been generally reassuring. There may be slight increase in the frequency of "borderline ovarian tumors" which are not frankly malignant but are also not entirely benign.
The greatest difficulty in designing such studies lies in finding appropriate controls, namely infertile women who do not use fertility drugs. Infertility, particularly disorders of ovulation, is itself associated with increased risk of hormonally-dependent cancers, in particular endometrial cancer. Thus comparing infertile women using fertility drugs with fertile women is not a good strategy; one needs to compare infertile women using fertility drugs with infertile women who do not. As infertility treatments have gained wider acceptance, the latter group becomes ever harder to find.