Keywords

6.1 Introduction

Breast cancer is one of the most common cancers among women in industrialized countries and is one of the deadliest. In 2020, 2.26 million new cases of breast cancer have been reported worldwide and 685,000 deaths [1]. In France, there has been a steady increase of +1.1% each year between 1990 and 2018 in the number of new cases diagnosed. However, in contrast to the number of new cases per year, there has been a progressive decrease of −1.6% per year between 2010 and 2018 [2] in the number of deaths.

The epidemiology of cancers in women has continuously changed over the last 30 years, including a progressive increase in the frequency of advanced breast cancer in young women [3].

This change in the profile of patients with breast cancer has motivated the exploration of numerous avenues to identify new and previously unknown environmental risk factors.

Numerous studies have examined the potential carcinogenic effect of endocrine disruptors: molecules present in food, water, ambient air, industrial products, cosmetics, and many everyday objects.

This carcinogenic effect in humans has been proven for certain organs such as the prostate, liver, and blood [4, 5].

Diethylstilbestrol is one of the endocrine disruptors whose carcinogenic effect in women has been widely demonstrated for several years, notably for breast cancer in mothers who used it and vaginal cancer and cervical cancer in daughters exposed in utero [6]. Considering the large number of studies and reviews on the subject, we have not retained this substance in this research.

The main objective of this review is to identify the scientific studies concerning other endocrine disruptors and their potential impacts on the risk of developing breast cancer.

The secondary objective is to study the potential increase in the risk of developing endometrial cancer or ovarian cancer in the presence of these disruptors.

6.2 The Importance of Studying Endocrine Disruptors When Studying the Genesis of Breast Cancer

The human endocrine system is a complex communication system involving many organs producing different hormones.

Hormones are chemical mediators that circulate in the blood to their target organs to exert a specific function. They are secreted by different glands in the human body such as the thyroid, thymus, liver, pancreas, pituitary, hypothalamus, stomach, adrenals, ovaries, kidneys, and testes.

These hormones circulate systemically until they bind to their specific receptors present on the cells of the target tissue. After binding, the hormone–receptor complex is internalized within the cell nucleus and then binds to a specific region of the hormone-dependent gene promoter, triggering gene expression.

From this definition of a hormone comes the definition of an endocrine disruptor (ED). According to the World Health Organization (WHO), endocrine disruptors are “chemical substances of natural or synthetic origin, foreign to the organism and likely to interfere with the functioning of the endocrine system, i.e. the cells and organs involved in the production of hormones and their effect on target cells via receptors” [7].

These endocrine disruptors are thus at risk of inducing harmful effects on the organism or on its descendants.

Indeed, EDs can have an impact on physiological hormonal functioning in different ways. They can bind to hormone receptors naturally present in the target organs (the direct effect of EDs) or interfere with the mechanisms of production or regulation of hormones or their receptors (the indirect effect of EDs).

EDs can act directly via membrane or nuclear hormone receptors, resulting in either an agonist effect by mimicking the effects of hormones or an antagonist effect by blocking these effects.

They can also act indirectly through different mechanisms:

  • By degrading the molecular structure of natural hormones (enzymatic interference, via cytochrome P450 in particular)

  • By invading hormone receptors, which reduces the number of receptors available to bind natural hormones

  • By short-circuiting the transport of natural hormones (interference with the internalization or deoxyribonucleic acid (DNA) binding of the hormone–receptor complex)

  • By maintaining high levels of natural hormones (interference with elimination by altering plasma clearance)

  • By modifying gene expression, that is, by causing epigenetic modification without changing the nucleotide sequence

The endocrine disruptors currently identified are very numerous, and their list is constantly growing. They can be of natural or synthetic origin (Table 6.1). Synthetic endocrine disruptors are found in products from the pharmaceutical industry (ethinylestradiol and diethylstilbestrol); in products used in everyday life such as food packaging, plastics, and cosmetics (bisphenol A, phthalates, and parabens); in food (dioxins); in products from the construction industry such as paints, carpets, solvents, and flame retardants (organochlorines and polychlorinated biphenyls); in air pollution (polycyclic aromatic hydrocarbons), etc.

Table 6.1 Examples of families of molecules with endocrine-disrupting effects and their potential sources of diffusion in the environment (INSERM) [8]
Full size table

These disrupting agents can be absorbed by humans via the respiratory tract, the cutaneous or mucous membrane tract, or the digestive tract. They can be absorbed in low doses in a chronic and repeated manner (accumulation in the body) or in an acute manner, which occurs mainly during workplace accidents, namely in industrial settings.

During chronic and repeated exposure to low doses, integrated endocrine disruptors can be stored in the body in a variable manner depending on its eliminated half-life, ranging from a few days to a few years.

There are four key points about the mode of action of these endocrine disruptors on human health:

  • The most critical period of exposure seems to be during embryonic life, but effects may not manifest until adulthood; this is the mechanism of delayed programmed toxicity.

  • The effects are mainly manifested in the next generation and not in the exposed subjects; this is the transgenerational effect.

  • As the quantity of hormones necessary for the normal functioning of the endocrine system is extremely small, disruption can result from a very small quantity of disruptive substances, this is, a nonmonotonic dose–response relationship with a toxicity threshold that is difficult to define.

  • There are interactions between different endocrine disruptors acting via various mechanisms (synergistic and antagonistic); therefore, possible potentiated effects may be suspected.

The large number of women affected by hormone-sensitive breast cancer has motivated many years of research into its risk factors, genesis, and therapeutic management.

Classically, oncogenesis is divided into three key stages:

  • Initiation, which represents a rapid, irreversible, and transmissible lesion of the DNA, induced by a carcinogenic factor (physical, chemical, viral, etc.)

  • Promotion, which corresponds to a prolonged, repeated, or continuous exposure to a substance that maintains and stabilizes the initiated lesion (mitogenic stimuli such as cytokines and growth factors). This leads to clonal expansion of pretumor cells.

  • Progression, characterized by the acquisition of proliferation capacities.

This oncogenesis develops through different mechanisms within healthy cells such as the release of growth factors, escape from tumor suppressor genes, facilitation of cell movement capacity, induction of neoangiogenesis, and the ability to resist apoptosis mechanisms.

Once modified, tumor cells are dedifferentiated, have highly developed motility capabilities, and respond very sensitively to chemoreceptors. In female breast cancer, these cells are mostly hormone-dependent, with an overexpression of estrogen receptor (ER) and progesterone receptor (PR), resulting in a high sensitivity to the latter.

Several risk factors have been studied and are known to increase its occurrence, such as age, parity, alcohol consumption, body mass index, physical activity, the use of menopausal hormone therapy, personal history of atypical breast hyperplasia, lobular/ductal carcinoma in situ or thoracic irradiation, and the presence of genetic mutations such as breast cancer gene 1 (BRCA1), breast cancer gene 2 (BRCA2), and partner and localizer of BRCA2 (PALB2). By looking at the pathophysiology and oncogenesis of the breast, it is inferred that both endogenous and exogenous estrogens may play a major role in tumor proliferation.

Other factors remain poorly studied in terms of their involvement in the development of breast cancer, and endocrine disruptors are among them.

The search for a cause-and-effect relationship between a woman’s exposure to endocrine disruptors during her life and the development of breast cancer seems coherent when we know their potential impacts on the hormonal system.

Indeed, certain molecules of the endocrine disruptor family could be responsible for a promoter effect on mammary hormone receptors and thus favor the clonal expansion of tumor cells previously modified by DNA lesions.

6.3 Endocrine Disruptors and Breast Cancer

6.3.1 Organochlorines (DDT, dichlorodiphenyldichloroethylene [DDE], and PCB)

Organochlorines are a large family of molecules with an endocrine-disrupting effect. They are neurotropic toxins that alter the functioning of the sodium channels essential for the transmission of nerve impulses. They are synthetic organic compounds in which one or more hydrogen atoms are substituted by one or more chlorine atoms.

Organochlorines are used as solvents, pesticides, insecticides, fungicides, refrigerants, and intermediary molecules in the chemistry and in the pharmaceutical industry.

These molecules are therefore used in agriculture, where they are administered to animals or to plants as growth regulators. They are also used as defoliants (herbicides), desiccants (water removal), and fruit thinners or used to prevent premature fall of fruit from the trees.

Among these compounds, the most studied are dichlorodiphenyltrichloroethane (DDT), dichlorodiphenyldichloroethylene (DDE), dieldrin, hexachlorobenzene (HCB), polychlorinated biphenyls (PCBs) derived from dioxin, chlordecone, and hexachlorocyclohexane (HCH).

A large number of in vitro and in vivo studies have explored the potential impacts of organochlorines on cancer cell proliferation, some of which are listed below.

One of the hypotheses of action of endocrine disruptors on breast cancer cell proliferation is the interaction of these compounds with membrane or intracellular proteins of breast cells, potentially responsible for tumor proliferation.

6.3.1.1 In Vitro Studies and Organochlorines

The study by Montes-Crajales et al. in 2016 explored this avenue by analyzing in vitro the affinity between certain endocrine-disrupting compounds and widely studied cellular proteins such as estrogen receptors (ESR1), progesterone receptors (PGR), human epidermal growth factor 2 receptors called HER 2 (ERBB2), BRCA susceptibility proteins type 1 (BRCA1), BRCA susceptibility proteins type 2 (BRCA 2), and sex hormone–binding globulin (SHBG).

This study targeted the affinity between these proteins and different forms of dioxin (belonging to the organochlorine family) and bisphenol A via a high-throughput virtual screening technique, followed by an experimental validation in silico by spectroscopy of the protein/ligand affinity suspected during the screening.

This work highlighted the potential for several endocrine disruptors, including some dioxin and bisphenol A derivatives, to bind to breast cancer–associated proteins, not just hormone receptor proteins [9].

Other in vitro study techniques have been implemented, including the analysis of human breast cell lines grown in culture.

A study by MA Garcia et al. in 2010 evaluated the in vitro effects of different doses of an organochlorine pesticide called hexachlorobenzene (HCB) on human cell cultures MCF-7 and MDA-MB-231.

MCF-7 is one of the estrogen receptor (ER)-positive breast tumor cell lines and is the most widely used line in breast cancer research laboratories. MDA-MB-231 represents an ER-negative tumor cell line.

This study showed an impact of HCB on MCF-7 cell proliferation, but not on the MDA-MB-231 line. It was also shown that exposure to certain doses of HCB induces cytochrome P450 gene expression and stimulates the insulin-like growth factor 1 (IGF-1) signaling pathway, but only on ER-positive cells (MCF-7).

This study raises the potential impact of this pesticide on the proliferation of ER-positive tumor cells [10].

The work of J. Payne et al. in 2001 also investigated the effect of several types of organochlorines on cell lines, such as o,p’-DDT, p,p’-DDE, β-HCH, and p,p’-DDT, which are persistent compounds found in human tissues. The objective of this study is to analyze the impact of these endocrine disruptors on human MCF-7 cell cultures after a standardized exposure to one or a mixture of several of these compounds, thus avoiding any exposure bias found in the general population due to the multiple possible uncontrollable sources of exposure.

The regression analysis showed combined effects even when each component of the mixture was present at or below its individual no-effect concentration. Assessments of the proliferation induced by the individual components of the mixture revealed that the effects of the combination of several components were stronger than the effects of the most potent component of the combination. These combined effects of organochlorines can therefore be described as synergistic on human cells.

In addition, comparisons with the expected effects as predicted by the summation of concentrations and independent action showed a strong agreement between prediction and observation. The effects of organochlorines can therefore be described as additive [11].

The analysis of these various in vitro studies supports the hypothesis that organochlorines are carcinogenic via their endocrine-disrupting effect, but the data on cell cultures is not sufficient to conclude that there is a probable relationship between these compounds and the development of breast cancer in women.

The proof of a possible adverse effect requires human studies, but the ethical issue strictly prohibits any randomized interventional study.

The best way to explore the subject is therefore epidemiological studies with a retrospective analysis of cohorts of women with breast cancer or a prospective analysis of women exposed to endocrine disruptors in their environment.

6.3.1.2 In Vivo Studies and Serum Organochlorine Levels

Two Danish (Hoyer et al.) and Norwegian (Ward et al.) teams explored, respectively, in 1998 (Danish team [12]), in 2000 (Norwegian team [13]), and in 2001 (Danish team [14]) via case–control studies the serum levels of certain organochlorine compounds in women with breast cancer and those in women from a control group.

These three studies showed discordant results regarding the correlation between high serum levels of compounds and the presence of breast cancer, which are presented in Table 6.2.

Table 6.2 Comparison of three studies
Full size table

The first study by Hoyer et al. found a significant increase in the number of women with elevated serum levels of dieldrin in the study group compared with women in the control group, irrespective of the immunohistochemical profile of the breast cancers in the cases. No other significant correlation was found for the other organochlorine compounds [12].

The second study by Ward et al. found no significant increase in serum organochlorine levels in the study group compared with the control patients, despite stratification by age, the presence of ER and PR on immunohistochemical analysis of breast tumors, and time between serum level measurement and the diagnosis of breast cancer [13].

Finally, in their second study in 2001, the team of Hoyer et al. explored the relationship between serum organochlorine levels and the presence of breast cancer, according to the amount of serum levels and according to the RE+ or RE− status of the breast tumors in the cases. No significant association was found except for a higher number of women with ER− breast cancer at high dieldrin exposure [14].

In the 1990s to 2000, several other US studies showed nonsignificant results for the association between organochlorines and breast cancer. However, it appeared that in several of these studies associations tended to be significant in certain subgroups, but the lack of power in each study limited interpretation. This is why the team of Laden et al., in 2001, published a meta-analysis of five of these American studies in order to increase their power [15].

This meta-analysis examined retrospectively collected data from 2042 patients regarding DDE and PCB concentrations in adipose tissue in women with breast cancer compared with controls.

No significant association was observed between elevated adipose levels of DDE (OR 0.99, 95% CI 0.77–1.27, p = NS) or PCB (OR 0.94, 95% CI 0.73–1.21, p = NS) and the presence of breast cancer.

Although some of the individual studies suggested an increased risk of PCB-associated breast cancer in certain stratified subgroups, these findings were reversed in this meta-analysis.

The teams of Millikan et al. (the USA) in 2000 [16] and Charlier et al. (Belgium) in 2004 [17] carried out two case–control studies with large cohorts and found contradictory results.

The Millikan et al. study included 748 cases (292 African American and 456 White women) and 659 controls (270 African American and 389 White women). No statistically significant association was found between elevated serum DDE or PCB levels and the presence of breast cancer in this study, even when subgrouped by ethnicity with adjustment for age, BMI, parity, history of breastfeeding, menopausal status, the use of hormone replacement therapy, and annual income.

The only slight increases in OR found were for elevated serum PCB and DDE levels in African American women in some subgroups (including BMI) and without p-value calculations, which further limits interpretation [16].

In contrast, the study by Charlier et al. retrospectively compared the presence of DDE and HCB in 231 cases and 290 controls. The variables were reported both continuously and in a binary manner (the presence or absence of DDE and HCB) according to a serum level below or above the limit of quantification set by the team.

In this study, there was a significant increase in the number of patients with DDE (OR 2.21, 95% CI 1.41–3.48, p = 0.0006) and HCB (OR 4.99, 95% CI 2.95–8.43, p < 0.0001) in the study group compared with the control group. This difference is also significant when analyzing DDE and HCB as continuous variables.

However, there is a potential selection bias due to the significant decrease in the number of postmenopausal patients in the control group. There is also a significant decrease in the number of patients receiving hormone replacement therapy in the case group, which reinforces the suspected role of the organochlorines studied [17].

Among the most recent studies on DDE and DDT are those conducted by Cohn et al. in 2007, 2015, and 2019. Dr. Cohn looked at the potential impact of organochlorines on the development of breast cancer depending on the age at which the exposure to these compounds began. She assumed that the risks are higher when exposure occurs when the mammary gland is still developing. For this research, she used data from the Child Health and Development Studies (CHDS), among others.

The 2007 nested case–control study measured serum levels of two forms of DDT and DDE in young women, dividing them into four age-groups at the time of peak exposure to these endocrine disruptors in the United States.

Statistical analyses revealed a significant increase in the risk of breast cancer when exposed to p,p’-DDT at any age, especially at an age below 14 years (OR 5.4, 95% CI 1.7–17.1, p < 0.01). However, there was no significant difference in each age-group (< 4 years, 4–7 years, 8–13 years) [18].

The 2015 nested case–control study is very interesting because it is one of the first to study the impact of in utero exposure to certain organochlorines, notably DDT.

Indeed, 9300 girls had been prospectively followed since their in utero growth started in the 1960s. Their mothers had blood samples taken for future analysis of serum levels of endocrine disruptors in the perinatal period. Of these, 103 cases who had developed invasive and/or noninvasive breast cancer by the age of 52 were selected, and 354 controls were matched. Perinatal maternal serum levels of DDE and two forms of DDT were measured and classified into four groups according to four quartiles.

The statistical analysis showed a significant increase in the risk of developing cancer in the highest quartile of o,p’-DDT compared with the lowest quartile in girls exposed in utero (OR 3.7, 95% CI 1.5–9, p = 0.04), irrespective of the mothers’ cancer status [19].

Finally, the 2019 study looked at the risk of breast cancer after DDT exposure according to exposure before or after puberty and according to the onset of cancer before or after menopause [20].

This study found a significant increase in breast cancer risk after menopause for postpubertal exposure and a significant increase before menopause for prepubertal exposure.

Two prospective cohort studies have analyzed the potential link between dioxin exposure and the development of breast cancer. However, these two studies contain cohorts of different sizes, which limits their comparison.

The Italian team of Warner et al. in 2011 analyzed data from a cohort of 833 women and showed an increased risk of breast cancer with serum dioxin levels, but not significantly (hazard ratio [HR] = 1.44, 95% CI 0.89–2.33) [21].

The French team of Danjou et al. in 2015 looked at the potential effect of dioxin through a prospective cohort of 63,830 women and found no significant increase in breast cancer risk associated with dioxin exposure (HR = 1.0, 95% CI 0.96–1.05). It should be noted that the degree of exposure is measured via a food questionnaire [22].

Finally, there are a few case–control studies with small cohorts giving discordant results for DDE, DDT, or PCB (Table 6.3).

Table 6.3 Comparison of nine studies
Full size table

6.3.1.3 In Vivo Studies and Fat Levels of Organochlorines

The analysis of the levels of endocrine disruptors in fat tissue is interesting because they remain stored for a longer period without being eliminated. The measurement of this level therefore makes it possible to study more reliably the cumulative exposure to certain compounds and their potential harmful effects long after the first exposure.

Furthermore, as the breast is predominantly composed of fat, the mammary gland is exposed locally to endocrine disruptors accumulated over the years in the fat cells.

Other studies published between 2000 and 2005 investigated the possible association between organochlorine fat levels and breast cancer through case–control studies in the USA, Canada, and Denmark and also presented conflicting results.

Of these four studies, the study by Stellman et al. (in the USA) in 2000 concluded that there was no significant association between high organochlorine body fat levels and the presence of breast cancer. The same is true for the individual analysis of each of the two organochlorine families studied (DDE and PCBs). The only significant association found concerned a specific subgroup of PCBs called PCB-183 (adjusted OR 2, 95% CI 1.2–3.4, p = not reported (NR)) [32].

In contrast, the Canadian study by Aronson et al. in 2000 showed a significant association between high levels of certain groups of PCBs and the presence of breast cancer.

Indeed, high concentrations of PCB-105 (OR 3.17, 95% CI 1.15–6.68, p < 0.01) and PCB-118 (OR 2.31, 95% CI 1.11–3.18, p < 0.01) are significantly more frequent in patients in the study group than in the control group. This association was also found in the nonmenopausal subgroup, but not in the menopausal subgroup. The analysis of data for other organochlorines (DDE, DDT, HCB, β-HCH, trans-nonachlor, cis-nonachlor, oxychlordane, and mirex) did not show a significant association [33].

Raashou-Nielsen et al. (Denmark) in 2005 built its study on the same spectrum studied by Aronson et al. and showed very different results.

The team investigated the possible association between given fatty levels of organochlorines and the presence of breast cancer in exclusively postmenopausal women, using a study of 409 cases and 409 controls. The relative risk was calculated according to the quartiles of organochlorine fat levels for HR+ breast cancer, HR− breast cancer, and all breast cancers combined. The results of this study are unexpected, as they show the absence of a significant association between high levels and the presence of breast cancer, but also a decrease in the relative risk of HR− breast cancer for high levels of certain organochlorines, notably DDE, β-HCH, oxychlordane, trans-nonachlor, and HCB. This decrease in RR was also found in the all-cancer group for β-HCH, oxychlordane, and HCB [34].

At the same time, no significant association was found when analyzing the data for the different molecules of the PCB family.

However, these results remain difficult to interpret, given the nonexhaustive analysis of endocrine-disrupting compounds and the absence of data concerning the period of exposure of each patient.

The 2003 study by Muscat et al. (USA) looked at the levels of certain organochlorines found in the fat of surgical specimens after lumpectomy or mastectomy and the potential correlation with disease recurrence. There was an increased relative risk of breast cancer recurrence in patients with high levels of PCBs in fat in general (RR 2.9, 95% CI 1.02–8.2), but none of the results from this study are specified in the article. In addition, some biases were raised by the project team, including the small cohort size and the fact that patients with stage 3 or 4 breast cancer are more likely to have a recurrence, whereas no correlation was found between organochlorine fat levels and disease stage [35].

The recent study by Huang et al. (China) in 2019 also found discordant results, with a nonsignificant decrease in the odds ratio risk of breast cancer in women with average fat levels of certain organochlorines (PCB-28, PCB-52, PCB-101, and DDT by adjusted odds ratio only), but it also showed a significant increase in risk in women with high body fat concentrations of PCB-188, PCB-138, PCB-153, and PCB-180, when calculated with the synthesis of all PCBs, as well as with high body fat levels of DDE. This significant association was found with both the unadjusted and the adjusted ORs.

However, it should be noted that in this study there were significantly fewer postmenopausal women in the control group than in the study group, which represents a potential bias [36].

Finally, we can mention the meta-analysis carried out by Lopez-Cervantes et al. and published in 2001, which brings together 22 studies, some of which are mentioned above, and many others were published before the 2000s and therefore not included in this review. The statistical results of this meta-analysis did not show a significant increase in breast cancer risk associated with serum or adipose levels of DDT or DDE [37].

6.3.2 Bisphenol A and Phthalates

Bisphenol A is a synthetic organic compound found primarily in the manufacture of plastics and resins. It is a molecule that has long been used to create materials used in everyday life, such as cables and adhesives, in the world of childcare and in plastic packaging in contact with foodstuffs. It is also used as a flame retardant or developer in thermal papers.

Bisphenol A is one of the compounds recognized as toxic for reproduction and endocrine disruptors, and its use has been banned in France in the composition of food containers since 2015 [38].

Phthalates are a family of chemical compounds derived from phthalic acid, used in the manufacture of floor coverings, cables, pipes, plastic films, shower curtains, and certain medical devices and childcare equipment. Finally, they are found in certain cosmetics such as nail polish and lacquer for their fixing properties.

Investigations over the last 10 years have led to the classification of certain phthalates as reproductive toxicants and potential endocrine disruptors by the European Chemicals Agency (ECHA) [39].

Unlike organochlorines, there are relatively few studies on the effect of bisphenol A and phthalates on the animal or human population.

The three studies conducted by Vandenberg et al. in 2010 [40], Lozada et al. in 2011 [41], and Acevedo et al. in 2013 [42] have explored the potential mammary carcinogenic risks of bisphenol A in the animal population. All the three studies showed consistent results with a significant increase in breast cancer risk in mice after in utero exposure and during lactation.

A meta-analysis by the team of Liu et al. was published in 2021 and brought together nine case–control studies carried out between 1998 and 2019, allowing the creation of a cohort of 7820 women and the study of certain phthalates and bisphenol A [43].

This meta-analysis found no significant association between urinary or adipose levels of bisphenol A and the development of breast cancer in women of all ages (OR 0.85, 95% CI 0.69–1.05).

The study also looked at urine and fat levels of phthalates, and the statistical analysis of this cohort did not find a significant increase in breast cancer risk. There was even a slight significant decrease in risk for some phthalate subgroups such as mono-benzyl phthalate (OR 0.73, 95% CI 0.60–0.90, p = NC).

A large study by Ahern et al. published in 2019 prospectively collected data from a cohort of 1,122,042 women. The aim was to compare the risk of breast cancer in 161,737 women exposed to phthalates in pharmaceuticals and 960,305 women considered unexposed [44].

The proportion of women with breast cancer among the unexposed was used as a reference with a hazard ratio of 1.0. After a multivariate analysis, only high exposure to dibutyl phthalate appeared to be associated with a significant increase in breast cancer risk (HR 2.0, 95% CI 1.1–3.6, p = NC).

This study also looked for a possible association between high exposure to phthalates in pharmaceuticals and the presence of hormone-dependent breast cancer, but no significant results were found.

Finally, there are four case–control studies that explored serum or urine levels of endocrine disruptors and breast cancer risk (Table 6.4).

Table 6.4 Comparison of four studies
Full size table

6.4 Endocrine Disruptors and Endometrial/Ovarian Cancer

For endometrial cancer and ovarian cancer, very few studies are available in the scientific literature. One example is the large prospective cohort study led by Donat-Vargas et al. in 2016, which included 36,777 women from 1997 [49]. This surveillance recruited 1593 women with breast cancer, 437 with endometrial cancer, and 195 with ovarian cancer. The proportion of women with cancer was compared between those with high PCB exposure (serum levels in the third tertile) and those with low PCB exposure (serum levels in the first tertile).

This large study showed no significant increase in breast (RR = 0.96, 95% CI 0.75–1.24), endometrial (RR = 1.21, 95% CI 0.73–2.01), or ovarian (RR = 0.90, 95% CI 0.45–1.79) cancer in women with higher serum PCB levels.

6.5 Conclusion

The relationship between endocrine disruptors and cancer in women is not clearly demonstrated and remains potential.

The various studies concerning organochlorines and breast cancer are mainly case–control studies with prospective data collection and retrospective analysis that highlight heterogeneous results that do not allow us to draw conclusions on the role played by these compounds.

Studies on phthalates, bisphenol A, and breast cancer remain few, and their results are also heterogeneous, which does not allow them to be incriminated.

Rare prospective cohort studies increase the power of the results, but the methods for measuring exposure remain disparate and the comparisons of these studies are limited.

Studies concerning endometrial cancer and ovarian cancer are rare, contain very small cohorts [50], or show nonsignificant results.

Finally, many studies have focused on old exposure periods, during the peak of the use of the compounds concerned and before the regulations limiting their use and exposure. It would therefore appear interesting to carry out additional studies with a more recent exposure period and large cohorts and by developing reliable in vivo measurements of endocrine disruptors, representative of the previous exposure.