Isoflavone Research Initiative

Skip to content

Information

Subscribe to Isoflavone News

Did Petrarkis et al. (1996) observe an increased risk of breast cancer?

Thursday, 01 October 2009 08:15

The hypothesis of an increased risk of breast cancer derived from experimental studies is, according to the German BfR and some reviews (Duffy et al. 2007; Wuttke et al. 2007), apparently conformed by clinical observations. In fact, there is only one regularly cited study, which seems suggestive for such a risk (Petrakis et al. 1996).

Petrarkis et al. performed an uncontrolled pilot study in an inhomogeneous female population. At least one of the participants used an oral contraceptive, several received hormonal replacement therapy. No inclusion criteria were defined. It is therefore impossible to know whether even more women took hormones next to isoflavones. Only 15 women regularly terminated the study (the abstract mentions 24 women, but this figure does not fit to the data presented in the publication). Seven of these 15 women had epithelial hyperplasia – a phenomenon which was reversible on discontinuation of the intake of isoflavones (38 mg/day over 6 months), although some of the women still had hormonal replacement therapy (Petrakis et al. 1996). If isoflavones had in fact caused this effect, their estrogenic potential would be higher than that of estrogen itself, which is clearly not the case.

Based on these inconsistent results, Petrarkis at al. called for a systematic trial. Shortly thereafter, such a trial was presented by McMichael-Phillips and coworkers. The first impression was that isoflavones in fact caused an increase in breast tissue density (McMichael-Phillips et al. 1998) – and this was referenced to by the German BfR and by Wuttke. However, both do not mention that the same working group corrected their results shortly thereafter. The first analysis had been made in a subgroup of the examined women. When the full set of data was analyzed and published, no more proliferation-enhancing effect was detected (Hargreaves et al. 1999). Consequently, the authors had to withdraw their initial assessment. Since then, no more clinical hints to potential proliferation-enhancing effects of isoflavones at the breast have been published.

In contrast, a lack of an effect of isoflavones in tissues sensitive to estrogen-effects (breast, uterus, vagina) has regularly been confirmed in clinical trials. The safety of application of isoflavones at the breast and endometrium has explicitly been tested under long-term conditions. Data has been published for the endometrial safety of the daily supplementation of 54 mg of genistein (measured by biopsies), and the safety at the breast (by mammography) for a duration of intake of one year (Crisafulli et al. 2004; D'Anna et al. 2007), two years (Marini et al. 2007), and three years (Marini et al. 2008). Even after three years there was no significant change in breast tissue density or the thickness of the endometrium. These effects of isolated genistein can be transferred to soy extracts, as the same observation of safety was also found after one year of supplementation of soy extract with 70 mg of isoflavones daily. Again, no impact of isoflavones on the thickness of the endometrium was found (Palacios et al. 2007). A recently published placebo-controlled study excluded effects of 80-120 mg of isoflavones daily on breast tissue density after two years of supplementation (Maskarinec et al. 2009), thus confirming the experience from other long term studies over one year (Sammartino et al. 2003; Verheus et al. 2008) and two years (Burke et al. 2003; Caserta et al. 2005). Long-term experience with the safety of red clover extract covers a minimum of three years (Powles et al. 2008).

In view of these results from well-designed controlled trials, the study results of the open, uncontrolled pilot study of Petrarkis et al. (1996) in a highly heterogeneous study population should no longer be used as an argument for assumed safety problems with isoflavones.