Abstract
Background: Fatty acid composition of adipose tissue is a most reliable biomarker of long-term dietary fatty acid intake. Few studies have implemented biomarkers of fatty acid intake in relation to breast cancer. In this study the relation between adipose tissue composition and breast cancer was investigated. Patients and Methods: Fatty acid composition in breast and buttock adipose tissue from 94 Greek women with breast cancer and 57 with benign breast tumors was determined. Multivariate analysis was performed to determine the association between fatty acid groups and breast cancer risk. Results: In pre-menopausal women, elevated total polyunsaturated fatty acids (PUFA) in breast adipose tissue and N-3 PUFA in buttock adipose tissue were associated with reduced odds of breast cancer (odds ratio, OR=0.19; 95% confidence interval, CI=0.05-0.76, p<0.02 and OR=0.02; 95% CI=0.0009-0.36, p<0.009). Conclusion: Adipose total PUFA and N-3 PUFA were inversely-related to breast cancer risk in Greek pre-menopausal women. These results may have specific impact on habitual fat intake recommendations.
Although the exact etiology of breast cancer is not yet known, there are reports that dietary factors, including dietary fat, may be implicated in this disease. Meta-analyses of epidemiological studies in which dietary fat intake was assessed with questionnaires or diet history have suggested that increased total fat intake is associated with increased breast cancer risk, especially in post-menopausal women (1, 2). A number of epidemiological case control and prospective cohort studies have shown that the type of fat consumed is related to breast cancer, albeit results have been considerably inconsistent (1, 2).
The inconsistency in epidemiological findings on the role of dietary fat in breast cancer may be due to the fact that it may be a heterogeneous disease, as well as due to methodological issues relating to study design, statistical analyses and small samples used, heterogeneity in fat intake among the different study samples, and dietary assessment tools and associated measurement errors in assessing fat intake (3, 4). Dietary questionnaires, used by the vast majority of these studies, may be flawed with bias due to inaccurate recall and reporting, changes in the day-to-day or seasonal eating patterns, limitations in the use of the food consumption tables, and over- or under-reporting of foods regarded as more or less socially acceptable, respectively (5, 6). There are also considerable amounts of hidden fat in foods, which are in general not known by the studied subjects. By contrast, studies on biomarkers of dietary fat intake, such as the fatty acid composition of adipose tissue, erythrocyte membrane, serum and plasma, are free of such bias. Owing to its slow turnover and insensitivity to acute disease (7, 8), adipose tissue is probably an ideal medium for the study of long-term fatty acid intake. The use of the adipose tissue as a biomarker of long-term fatty acid intake may be especially suited to the study of chronic disease or diseases that can have a relatively long course and duration, such as cancer. Although such biomarkers and, in particular, adipose tissue composition, seem to represent most accurately long-term fatty acid intake, they have been used to determine the association of habitual fat intake and breast cancer risk in relatively few studies (9-14).
In our previous study (13), adipose tissue composition of Greek patients with breast cancer was compared with that of healthy control women. In a multivariate analysis, the relative amount of monounsaturated fatty acids (MUFAs) in general, and of oleic acid in particular, were higher in healthy women, while the percentage of saturated fatty acids such as myristic acid was elevated in those with breast cancer. The purpose of the present study was to investigate whether adipose tissue composition relates to the type of breast tumor, comparing adipose tissue analysis of Greek women with malignant breast disease with those with benign breast tumors.
Patients and Methods
Patients. Patients included in this study were Greek females presenting with breast tumor at the Breast Unit of the Department of Surgical Oncology at the University Hospital of Heraklion, Crete, Greece. Final diagnosis was determined by histological examination of the surgical specimen.
The study sample consisted of a total of 151 patients, 94 women with breast cancer and 57 with benign breast tumors, including fibroadenoma (n=48), fibrocystic changes (n=6) and papilloma (n=3). The patients with breast cancer were between 23 and 84 years of age (mean=55.9 years) and the same as included in our previous study (13), while the patients with a benign breast tumor were between 17 and 77 years of age (mean=41.2 years). Overall, there were 75 pre- and 76 post-menopausal patients. Among patients with a benign breast tumor, 43 (75.4%) were pre- and 14 (24.6%) post-menopausal, while in the breast cancer group, 32 (34.0%) patients were pre- and 62 (66.0%) post-menopausal.
Additionally, educational level, smoking, alcohol consumption, age of onset of menarche, breastfeeding, miscarriages, hormone replacement therapy, contraceptive use, waist to hip ratio (WHR) and body mass index (BMI) were recorded for each patient. All patients were informed about the nature and the purpose of this study and signed a consent form. The Ethical Committee at the Medical School of Crete University Hospital had previously approved the research protocol (approval #801).
Adipose tissue measures. Breast and buttock adipose tissue was collected from 39 and 32 patients with benign breast tumors and from 71 and 60 patients with breast cancer, respectively. A fat tissue sample of the breast, approximately 0.5 cm in diameter, was harvested at least 1 cm from the tumor and stored in glass tubes. Sub-cutaneous buttock tissue samples were collected by aspiration, using the method described by Beynen and Katan (15). This particular method has been reported to be rapid and safe, and to cause no more discomfort than a routine venipuncture. Buttock adipose tissue samples can be safely stored for up to 1.5 years without changes in the component fatty acids (15). Samples were taken from the upper outer quadrant of the gluteal area, through the use of a 10-ml vaccutaneous tube. Aspiration sites were sprayed with local anesthetic (ethyl chloride) prior to sampling. Adipose tissue samples were stored at −80°C.
Prior to analysis, samples were thawed and the fat was transferred to 10 ml screw-capped tubes with the aid of Pasteur pipettes and several drops (~0.5 ml) of chloroform:methanol (2:1, v/v). Methyl esters of the fat component fatty acids were prepared in the screw-capped vials according to the method described by Metcalfe et al. (16). Briefly, 20-30 mg of fat sample were saponified with 1.0 ml NaOH in methanol and the fatty acid methyl esters (FAME) were prepared with 14% boron trifluoride in methanol following extraction with hexane after washing with saturated NaCl. The hexane (upper layer) containing the FAME was transferred to gas chromatography vials and stored at −20°C until analysis. The FAME were separated on a 100×0.25 mm Id.SP-2560 fused silica capillary column, coated with a 0.25 μm of cyanopropyl silicone provided by Supelco (Bellefonte, PA, USA), using a Shimadzu GC-17A/FID gas chromatograph equipped with an AOC-20I auto injector (Kyoto, Japan). The Class-VP Chemstation software (Shimadzu Co., Kyoto, Japan) was used for quantitation and identification of peaks. Baseline separation of over 50 FAME peaks was accomplished by means of mixed FAME standards (Sigma, Poole, UK). The analytical conditions employed were as follows: volume injected 1 μl, carrier gas helium (1.1 ml/min), injector temperature 250°C, FID 260°C, split ratio 1:4 to 1:20 (depending on the sample quantity), and oven temperature from 140°C to 245°C with stepped temperature program: within a total run time of 54 min. The fatty acids are expressed as percentage of the total fatty acids determined in the chromatogram.
Statistical analysis. Data were analyzed through the use of the SPSS statistical package (Chicago, IL, USA). The statistical methods used were one-way ANOVA and multivariate conditional logistic regression analysis. Multivariate conditional logistic regression analyses were carried out to obtain odds ratio (OR) estimates and associated 95% confidence interval (CI), while controlling for potential confounders. Disease status was the dependent measure (breast cancer=1, benign breast tumor=0), while groups of adipose tissue fatty acids of the breast and buttock, i.e. saturated, trans, MUFA, total polyunsaturated fatty acids (PUFA), N-3 (also called omega-3) and N-6 (also called omega-6) PUFA, were the independent measures. Groups of adipose fatty acids were coded 1 if their levels exceeded the median and 0 otherwise. Potential confounders included age, BMI, WHR, educational level, smoking, alcohol consumption, age of onset of menarche, breastfeeding, miscarriages, hormone replacement therapy and contraceptive use. Analysis was performed for the entire groups of patients, as well as for pre- and postmenopausal patients separately.
Results
ANOVA indicated that compared to patients with benign breast tumors, patients with breast cancer had significantly higher age (F=39.4, p<0.0005), BMI (F=9.03, p<0.003), waist circumference (F=8.7, p<0.004), hip circumference (F=5.5, p<0.03), WHR (F=5.1, p<0.03), total MUFA in buttock adipose tissue (F=6.1, p<0.02) and breast adipose tissue total MUFA (F=10.04, p<0.002, Table I). The following parameters were lower in breast cancer patients: height (F=7.1, p<0.009), total trans-fatty acids (F=4.2, p<0.05), total PUFA (F=13.09, p<0.0005), N-3 PUFA (F=5.8, P<0.02), N-6 PUFA (F=14.2, p<0.0005) and N-6/N-3 PUFA ratio (F=17.8, p<0.0005) in buttock adipose tissue and total PUFA (F=10.7, p<0.001), N-6 PUFA (F=12.36, p<0.001), and N-6/N-3 PUFA ratio (F=7.1, p<0.006) in breast adipose tissue (Table I).
Age, anthropometric and adipose tissue fatty acids.
Multivariate conditional logistic regression analysis with groups of breast adipose tissue fatty acids as categorical independent variables and age, BMI, WHR, educational level, smoking, alcohol consumption, age of onset of menarche, breastfeeding, miscarriages, hormone replacement therapy and contraceptive use as potential confounders, indicated that higher total PUFA in breast adipose tissue was associated with a reduction in breast cancer risk in premenopausal women, comparing patients with breast cancer versus those with benign breast tumors (OR=0.19; 95% CI=0.05-0.76, p<0.02) (Table II). A multivariate conditional logistic regression model using the same predictors as previously was subsequently performed for analysis of buttock adipose tissue. The particular multivariate model indicated that higher N-3 PUFA in buttock adipose tissue was associated with a reduction in breast cancer risk in pre-menopausal women (OR=0.02; 95% CI=0.0009-0.36, p<0.009) (Table II). Among groups of all (pre- and post-menopausal) and post-menopausal women, multivariate analysis revealed no statistically significant associations between disease status and any of the variables.
Adipose tissue fatty acids and significant breast cancer risk. Multivariate conditional logistic regression analysis of the association of adipose fat with breast cancer risk, comparing data from patients with breast cancer with those with benign breast tumors.
Among patients with breast cancer, there were no statistically significant relationships between tumor grade, tumor size, nodular status or disease stage and any group of adipose tissue fatty acids.
Discussion
In the present study, after controlling for potential confounders, multivariate conditional logistic regression analysis of fatty acid composition in breast adipose tissue demonstrated that only for total PUFA was there a significant difference found between malignant and benign breast tumors in premenopausal women. While there was no difference in total PUFA among groups of all women (pre- and post-menopausal), and post-menopausal women only, total PUFA was inversely-related to breast cancer risk in pre-menopausal women. While in one study total PUFA was lower in erythrocyte membrane of post-menopausal breast cancer patients (17), most biomarker studies did not identify a difference of total PUFA between patients with breast cancer and healthy women, regardless of their menopausal status (18-20). Similarly, in our previous study in which fatty acid composition of adipose tissue in Greek breast cancer patients was compared with that of healthy women, regardless of their menopausal status, we were unable to detect a difference of the total PUFA amount among the two groups (13). The combination of the results of both the former and the present study in a similar population may indicate that PUFA intake itself does not contribute to the induction of breast tumors (proliferation), but promotes the development of their benign biological behavior (differentiation) in pre-menopausal women.
Although adipose tissue composition, in general, reflects long-term dietary intakes (7, 8) there are indications that adipose tissue analysis of the buttock may be a better index of long-term or habitual fat intake than that of the breast (21). In the present case control study, only N-3 PUFA of the buttock adipose tissue was related (inversely) to breast cancer risk in pre-menopausal women after controlling for potential confounders. Considering all dietary fatty acid biomarker studies, only in two case control studies was a relationship between N-3 PUFA level and breast cancer risk found, which also appeared to be inverse (11, 12, 14, 22). Such a protective effect was confirmed by a meta-analysis of three cohort studies, but not in a meta-analysis of seven case-control studies (11). Only four other studies have examined the fatty acid composition of the buttock adipose tissue in relation to breast cancer (13, 14, 18, 23). In our previous study, comparing buttock adipose tissue of patients with breast cancer with that of Greek women without breast disease regardless of their menopausal status, only MUFA and saturated fatty acids appear to be independent breast cancer risk factors (13). Unlike the latter and the present study, however, the two case control studies recruited only post-menopausal women. One of these studies revealed only a borderline significant inverse relation between breast cancer risk and N-3/N-6 PUFA ratio in buttock adipose tissue (23), while the other, as well as the present study, failed to identify any fatty acid as being a risk factor in post-menopausal women (18). The fourth, a Danish cohort study, investigating only the role of N-3 PUFA in women aged 50-64 years, failed to demonstrate any association with breast cancer risk (14). Notably, analysis was not performed per menopausal status. Hence, the present study, to our knowledge, showed for the first time a significant inverse relation between N-3 PUFA in buttock adipose tissue, a superior index of habitual fat intake, and breast cancer risk in pre-menopausal women with breast tumors. The combination of the results of both the previous and the present study in the same Greek population may indicate that N-3 PUFA intake itself does not contribute to breast cell proliferation and the development of breast tumors, but to benign differentiation of proliferating breast cells in pre-menopausal women.
Some examples of the various possible molecular mechanisms to explain the potentially protective effect of N-3 PUFA on breast cells are: growth inhibition and differentiation of mammary tissue by binding to mammary-derived growth inhibitor-related gene (MRG) (10), down-regulation of v-erb-b2 erythroblastic leukemia viral oncogene homolog 2 (ErbB2 or Her-2/neu) expression (24), apoptosis of human breast cancer cells and mammary tissue by up-regulation of syndecan-1 expression (25), and inhibition of fatty acid synthase or oncogenic antigen-519 (26, 27). N-3 PUFA docosahaexaenoic acid has been reported to down-regulate expression of the antiapoptotic proto-oncogene B-cell lymphoma 2 (Bcl-2), consequently inducing apoptosis (28), and to reduce phosphorylated mitogen-activated protein kinase (MAPK) and cyclin-D1 levels in MCF-7 human breast cancer cells (29). However, as gene expression was not measured in the present study, it remains a tentative hypothesis that the decrease in breast cancer risk with increasing adipose N-3 PUFA in our pre-menopausal women may be mediated by effects of these fatty acids on the above mentioned genes.
Animal studies suggest that N-6 PUFAs may be precursors of intermediates involved in the development of mammary tumors, whereas long-chain N-3 PUFAs, found for example in fish oil, can inhibit these effects (9). Data of biomarker studies regarding the impact of N-6 PUFA leves on breast cancer risk have been inconsistent (9, 11). Some investigators have postulated that N-3 PUFAs inhibit breast cancer and that the degree of inhibition depends on background levels of N-6 PUFAs, suggesting that the ratio between those fatty acid families may be of importance (9, 23). In both studies, an inverse relation of borderline significance between breast cancer risk and the N-3/N-6 PUFA ratio in adipose tissue had been found. In the present, as well as in our previous study (13), no significant relation for N-6 PUFA or the balance between N-3 and N-6 PUFAs with breast cancer risk was demonstrated in multivariate analysis.
Multivariate analysis in our previous study demonstrated that total MUFA and, in particular oleic acid, was significantly lower, while saturated fatty acids, especially myristic acid, were significantly higher in patients with breast cancer than in healthy control women (13). This was not observed in the present study. In a meta-analysis of 10 biomarker-based studies (11), saturated fatty acids were not related to breast cancer risk, while total MUFA was related to an increased risk of breast cancer in cohort, but not in case-control studies.
Once again, it has to be stressed-out that adipose fat composition of patients with breast cancer was compared with that of women with benign breast tumors and not with women without any breast disease. In the majority of studies, comparison was made with adipose tissue of healthy women without any known breast disease. In a few studies, women with ‘benign breast disease’ (22, 30), benign proliferative disease (18) and mastomegaly (9) served as controls. In only one study, did women with fibroadenoma and women with benign breast tumors form two control groups (31). A one-way ANOVA of nine adipose fatty acids and their derived ratios did not show any significant differences among the groups.
Most of the studies on dietary fat and breast cancer risk have relied on diet questionnaires and few have used more reliable long-term dietary fat intake biomarkers, such as adipose tissue. Adipose tissue has a very slow turnover (0.11% daily) and a half-life of approximately 600 days (7, 8). The fatty acid composition of adipose tissue has been reported to reliably reflect dietary intake of the preceding two to three year period (32). This is especially true for trans-fatty acids, the odd-chain saturated fatty acids (e.g. C15:0 and C17:0) and the essential PUFAs (both N-3 and N-6), which cannot be synthesized endogenously but are derived exclusively from dietary sources (33, 34). However, a weakness in using adipose tissue measures is that they represent qualitative (proportion of the total fatty acids in the chromatogram), rather than quantitative estimates. Another disadvantage of using adipose tissue composition as biomarker of dietary fatty acid intake is the inability to consider food sources, as well as interactions between fatty acids and antioxidants. It appears that associations between dietary PUFAs and breast cancer risk vary according to food sources and antioxidant intake (35). For example, breast cancer risk of French women has been inversely-associated with alpha-linolenic acid (ALA) intake from fruit and vegetables and positively associated with ALA intake from nut mixes and processed foods as well as with total ALA intake in those with a high antioxidant (vitamin E) intake (35). This may also explain the inconsistent findings regarding the breast cancer risk according to the groups of fatty acids among studies in various geographical areas with different dietary habits. Another significant limitation of our study is the relatively small number of women included.
In conclusion, the present study is the second study on biomarkers of dietary fatty intake and the risk of breast cancer in a Greek population. In the present study, total PUFA in breast adipose tissue and N-3 PUFA in gluteal adipose tissue were higher in pre-menopausal patients with benign versus those with malignant breast tumors. No other significant differences were observed. Translating these findings into habitual fat intake may suggest that the use of fish, krill, sunflower or soybean oil instead of butter fat, as well as consumption of fish and whole-grain bread instead of meat and white bread may reduce the chance of malignant breast cell differentiation in pre-menopausal women.
Acknowledgements
Funding was provided by the Greek Ministry of Health and the Hellenic Anticancer Society. The role of the funding was to provide personnel support and make the technical analysis of the adipose tissue composition possible. We would like to acknowledge the contribution of Dr. Alexandros Zafiropoulos, Manolis Linardakis and Dr. Joanna Moschandreas. In addition, many thanks are expressed to Maria Dafermou, Maria Chrisou and Sotiris Economou (Nursing Department, Technological Institute of Crete, Heraklion) for their invaluable assistance in this project. Furthermore, we would like to thank Mrs. Eugenia Bolbasis for linguistically reviewing this manuscript.
Footnotes
-
Conflicts of Interest
The Authors declare no potential conflict of interests.
- Received January 26, 2013.
- Revision received March 8, 2013.
- Accepted March 12, 2013.
- Copyright© 2013 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
