Can Clinicians Now Assure their Breast Cancer Patients that Soyfoods are Safe?

“…we propose that recently published research … warrants a re-evaluation of the default prohibition against soyfoods.”

Considerable confusion exists among both physicians and their breast cancer and high-risk patients regarding the health consequences of soyfoods. Identifying the soybean components (isoflavones) as phytoestrogens, many oncologists routinely discourage and often prohibit soyfoods from inclusion in their breast cancer patients' diets. Although no consensus exists among the research or clinical communities regarding the use of soy, many patients have internalized their physician's concern, despite also hearing opposing messages regarding the beneficial effects of soy (e.g., the ability of soy protein to lower cholesterol levels) [1]. However, we propose that recently published research – especially the results from a large prospective epidemiologic study, which found that postdiagnosis soyfood intake improved the prognosis of breast cancer patients – warrants a re-evaluation of the default prohibition against soyfoods [2].

“…isoflavones have … come to be viewed as selective estrogen receptor modulators…”

This reassessment is especially warranted because this study has generated considerable media coverage, largely extolling the potential benefits of soyfoods for breast cancer patients, and because soyfoods have markedly increased in popularity in non-Asian countries over the past 20 years; for example, in 2009, US survey data indicated that a third of Americans used soy products at least once per month [3]. Although still not a part of traditional western diets, for the many breast cancer patients who transit towards a more plant-based diet for general health reasons, having to avoid soyfoods limits a large group of convenient sources of high-quality protein. Furthermore, there are intriguing data suggesting that, independent of nutrient content, soyfoods offer health benefits for menopausal women, including reducing the risk of heart disease [4] and possibly osteoporosis [5,6], and alleviating menopausal symptoms [7]. Thus, if the data now indicate that soyfoods do not put breast cancer patients at risk, the oncology community needs to recognize this and adjust their recommendations accordingly.

Isoflavones are a subclass of flavonoids but, by comparison, have a much more limited distribution in nature, such that among commonly consumed foods, these diphenolic compounds are only found in physiologically relevant amounts in foods made from soybeans [8]. As a general rule of thumb, in traditional Asian soyfoods such as tofu, soymilk and miso, each gram of soy protein is associated with approximately 3.5 mg of isoflavones [9]. Isoflavones have traditionally been classified as phytoestrogens [10,11] but in recent years, with the knowledge that they preferentially bind to [12] and transactivate estrogen receptor (ER)-β in comparison with ER-α [13], isoflavones have also come to be viewed as selective ER modulators (mixed estrogen agonists/antagonists) [14]. However, it was actually the hormone-independent effects of isoflavones that first attracted the attention of cancer researchers [15,16]. The primary soybean isoflavone, genistein, inhibits a wide range of hormone-dependent and -independent cancer cell lines in vitro (for reviews, see [16,17]).

Nevertheless, it is, in fact, the in vitro growth-promoting effects of genistein that provided the initial impetus for exploring safety concerns [18]. Genistein exerts a biphasic effect on MCF-7 (an ER-positive human breast cancer cell line) cell growth. At physiologically-relevant concentrations, the estrogenic actions of genistein result in MCF-7 cell growth (in estrogen-depleted media), whereas only at much higher concentrations do the ER-independent effects of genistein on signal transduction result in growth inhibition [18]. However, it is the preclinical data that provide the strongest evidence of harm. Genistein stimulates the growth of mammary tumors in ovariectomized athymic mice implanted with MCF-7 cells [19,20].

In the basic model employed, in which stimulation has been observed, an estrogen pellet is implanted into mice in order to stimulate tumor growth. After tumors reach a cross-sectional area of approximately 30–40 mm², the estrogen pellet is removed, which results in tumor regression in mice consuming a soy/isoflavone-free diet [21]. By contrast, after an initial period of regression, tumor regrowth occurs in mice fed this same diet but to which a variety of genistein/genistin (the glycoside form of genistein)-containing products have been added [22].

“This population-based cohort study found that … those in the fourth soy protein intake quartile experienced a reduced risk of tumor recurrence … and lower total mortality.”

Research from the discussed model also shows that genistein inhibits the efficacy of tamoxifen [23] and aromatase inhibitors [24], and that despite containing similar amounts of genistein, tumor growth is positively related to the degree to which the isoflavone-containing product has been processed [25]. In fact, the whole soyfood, soy flour, does not stimulate tumor growth [25]. Thus, the animal data suggest a possible risk primarily from isolated isoflavones and soybean extracts, but do not illuminate a risk for whole soyfoods (note that preliminary evidence indicates the proposed explanation for this processing effect – that circulating levels of unconjugated genistein, which is the biologically active form, increase in proportion to processing, is not applicable to humans [26]). Furthermore, not only are there significant limitations to transferring conclusions from a rodent model to humans, but not all rodent models show that even isolated isoflavones stimulate the growth of existing mammary tumors [27]. Given the conflicting results and the inherent limitations in animal models, the human data should carry the most weight when making clinical decisions.

In this regard, the results from three epidemiologic studies, two of which were conducted in China [2,28] and one in the USA [29], are especially relevant. In the first Chinese study to be published, neither soy nor isoflavone intake was related to the disease-free survival of Chinese breast cancer patients over the 5.2-year follow-up period [28]. In this study, of the 1001 patients for whom data on receptor status were available, approximately 63% were ER-positive. The major limitation of this research is that soy intake was assessed prior to diagnosis. In the US study, 1954 US breast cancer survivors were followed for 6.3 years [29]. During this time, there were 282 breast cancer recurrences. Suggestive trends for a reduced risk of cancer recurrence were observed with increasing quintiles of isoflavone intake compared with no intake among postmenopausal women and among tamoxifen users. Interestingly, the benefit of soyfood intake on survival was more pronounced among women with ER-positive breast cancer.

The suggestive trend towards an improved prognosis noted in the US study is substantially reinforced by the results of the second Chinese study, which was recently published in the Journal of the American Medical Association [2]. This population-based cohort study found that over the mean 3.9-year follow-up period, among the 5033 breast cancer patients, those in the fourth soy protein intake quartile experienced a reduced risk of tumor recurrence (hazard ratio: 0.68; 95% CI: 0.54–0.87) and lower total mortality (hazard ratio: 0.71; 95% CI: 0.54–0.92) [1]. These benefits remained significant even after the findings were adjusted for 17 factors including tumor, node, metastasis stage, ER and progesterone receptor status and the type of treatment received. Soy intake was assessed at 6, 18, 30 and 60 months after diagnosis. During the follow-up, there were 444 deaths and 534 breast tumor recurrences or breast cancer-related deaths. Importantly, intake of soyfoods was associated with improved survival, regardless of tamoxifen use. Furthermore, women who had the highest level of soyfood intake and did not take tamoxifen had a lower risk of mortality and a lower recurrence rate than women who had the lowest level of soyfood intake and used tamoxifen.

Despite these favorable findings, clinicians would understandably prefer to have direct trial data to inform their recommendations. Unfortunately, no randomized controlled trials have evaluated the effects of either soyfoods or isoflavone supplements on breast tumor recurrence and, thus, definitive safety data are lacking and are not likely to be forthcoming. However, those studies that have evaluated the impact of these interventions on intermediary markers of risk are reassuring. Soy and its isoflavones have minor, and probably clinically irrelevant, effects on circulating estrogen levels [30] and, in contrast to conventional hormone therapy [31–33], they have no effects on breast tissue density (n = 4 studies) or breast cell proliferation (n = 5 studies) in healthy postmenopausal women, high-risk women or breast cancer patients (for reviews, see [34,35]). While each of these studies has some important limitations (mainly sample size and to some extent duration) the collective lack of a harmful effect is notable.

“…clinicians can adopt a stance of permitting use in patients who want to begin eating soyfoods or for whom soyfoods already represent a normal part of their diet (mainly vegetarians and patients of Asian ethnicity).”

In summary, the animal data, at least on the basis of one widely used model, are supportive of the safety of minimally, but not highly processed, isoflavone-containing soyfoods and products [25]. The epidemiologic data suggest that soy-food intake may benefit breast cancer patients, although objections concerning extrapolating the results from Chinese patients, who have almost certainly consumed soy throughout their lives, to Caucasian breast cancer patients have been raised. In addition, the follow-up period in the Journal of the American Medical Association study was relatively short [34] and it had limited power to address the subgroups by ER status and tamoxifen use [36]. The clinical data are supportive of safety but not benefit, in that markers of breast cancer risk are neither favorably nor adversely affected [30,34,35].

Faced with the totality of this information, clinicians have several possible options. First, to continue to routinely prohibit soy from the diet of breast cancer patients. We feel that the current state of the science does not support this option for whole soyfoods and it ignores the nutrition and health benefits conferred by including soyfoods in a plant-based diet. Second, clinicians can adopt a stance of permitting use in patients who want to begin eating soyfoods or for whom soyfoods already represent a normal part of their diet (mainly vegetarians and patients of Asian ethnicity). The data would support this option and it is one that is consistent with the position of the American Cancer Society [37]. Third, advice could be given by clinicians, based on the lack of harm noted in the clinical studies and the benefits reported in the epidemiologic studies, for breast cancer patients to begin eating whole soyfoods. However, we do not feel that the existing data are sufficiently strong to justify the use of soyfoods in the treatment of breast cancer patients. This having been said, even in the absence of clinical data demonstrating that soyfoods improve prognosis, should new clinical data confirm that soyfoods have a neutral or favorable effect on breast cancer markers and/or new epidemiologic data show that soyfood intake improves prognosis, then the third option – the more assertive position of actively recommending soyfoods for breast cancer patients – would need to be revisited.