Author: Xiaorong Lin

Specialty: Medical Oncology
Institution: Breast Tumor Center, Sun Yat-Sen Memorial Hospital, Sun Yat-Sen University
Address: Guangzhou, Guangdong, 510060, China

Author: Shubin Hong

Specialty: Endocrinology
Institution: Department of Endocrinology, The First Affiliated Hospital, Sun Yat-Sen University
Address: Guangzhou, Guangdong, China

Author: Jiefeng Huang

Specialty: Medical OncologyObstetrics and Gynecology
Institution: Diagnosis and Treatment Center of Breast Diseases, Shantou Affiliated Hospital, Sun Yat-Sen University
Address: Guangzhou, Guangdong, China

Author: Yi Chen

Specialty: Medical OncologyObstetrics and Gynecology
Institution: Breast Tumor Center, Sun Yat-Sen Memorial Hospital, Sun Yat-Sen University
Address: Guangzhou, Guangdong, 510060, China

Author: Yufeng Chen

Specialty: OncologySurgery
Institution: Department of Colorectal Surgery, The Sixth Affiliated Hospital, Sun Yat-Sen University
Address: Guangzhou, Guangdong, China

Author: Zhiyong Wu

Specialty: Medical OncologyObstetrics and Gynecology
Institution: Diagnosis and Treatment Center of Breast Diseases, Shantou Affiliated Hospital, Sun Yat-Sen University
Address: Guangzhou, Guangdong, China

Abstract: Strong evidence exists indicating that the risk of breast cancer (BC) occurrence is influenced by complex internal environmental factors, including blood lipid and lipoprotein components. However, the roles of these components in BC development and progression remain controversial. This study examined whether serial serum lipid and lipoprotein measurements were associated with breast cancer risk and whether lipoproteins had BC prognostic properties. We compared the plasma-related parameter levels, including lipid and lipoprotein levels between 299 patients with invasive ductal breast cancer, also known as invasive ductal carcinoma (IDC), and 200 healthy donors. We performed univariate and multivariate logistic regression analyses to assess overall survival (OS) and disease-free survival (DFS). We found that the serum glucose, triacylglycerol, and low-density lipoprotein levels were significantly higher in patients with IDC than in healthy donors. However, high-density lipoprotein and apolipoprotein A1 (apoA1) levels were lower in patients with IDC than in healthy donors. Multivariate regression analysis demonstrated that elevated apoA1 levels were associated with a reduced risk of IDC, and univariate analysis showed that patients with IDC with lower apoA1 levels at diagnosis had larger tumors than patients with high apoA1 levels. Moreover, patients with IDC with lower apoA1 levels were more likely to have positive axillary lymph nodes, and were diagnosed at more advanced disease stages than patients with high apoA1 levels. We used a Cox regression model to assess the relationships between the above parameters and DFS and OS, after adjusting for tumor T and N stages, which were determined using the TNM classification system, and immunohistochemical subtypes. We found that lower apoA1 levels at diagnosis were associated with poor DFS and OS. At 60 months of follow-up, the DFS rate is 74.5% in the apoA1 L1 group, 89.9% in apoA1 L2 group, and 93.1% in apoA1 L3 group (p=0.0002). Similarly, the OS rate is 78.2% in apoA1 L1 group, 91.3% in apoA1 L2 group, and 93.7% in apoA1 L3 group (p=0.0012). In conclusion, our data indicate that low apoA1 levels are an independent predictor of the poor clinical outcomes in IDC patients.

Introduction

Breast cancer (BC) is the most commonly diagnosed cancer and the second most common cause of cancer-related death among females (Ferlay et al., 2015). The incidence of BC in Asia is lower than in America and Europe, but is rising dramatically. Although the association between deregulated lipid and lipoprotein metabolism and the risk of cardiovascular disease is well established (Huxley et al., 2002), the role of lipid and lipoprotein metabolism biomarkers in BC remains controversial.

The results of many extensive epidemiologic and experimental investigations have shown that the dysmetabolism of blood lipids or lipoproteins which are used for energy storage and membrane production (Currie et al., 2013), may influence carcinogenesis through insulin resistance, inflammation, and oxidative stress pathways (Ackerman et al., 2014). Cholesterol is a structural component of the cell membrane, and is associated with proteins involved in key cellular signaling pathways (Lingwood et al., 2010). Furthermore, cholesterol is also a steroid hormone precursor and the vast majority of BCs are known to be hormone responsive (Yager et al., 2006). A number of studies have investigated the relationship between total cholesterol (TC) and the BC risk (Eliassen et al., 2005; Fagherazzi et al., 2010; Gaard et al., 1994; Hiatt et al., 1986; His et al., 2014; Hoyer et al., 1992; Iso et al., 2009; Kitahara et al., 2011; Knekt et al., 1988; Melvin et al., 2013; Steenland et al., 1995; Strohmaier et al., 2013; Tornberg et al., 1988; Tulinius et al., 1997; Vatten et al., 1990). However, the results of these studies are inconsistent, as some have shown that high plasma TC levels increase the risk of BC (Kaye et al., 2002; Kitahara et al., 2011), while others have revealed no association between TC levels and the risk of BC (Eliassen et al., 2005; Ha et al., 2009; Iso et al., 2009) or have even demonstrated the existence of an inverse relationship between TC levels and the risk of BC (Touvier et al., 2015; Tulinius et al., 1997; Vatten et al., 1990).

High-density lipoprotein cholesterol (HDL-C) is the lipoprotein responsible for the cholesterol transportation. Evidence regarding the possible mechanisms by which HDL-C can influence carcinogenesis has been provided by many experimental studies (Silvente-Poirot et al., 2014; Soran et al., 2012; von Eckardstein et al., 2005). The majority of HDL-C particles contain either a single copy or multiple copies of apoA1, which plays a role in promoting cholesterol release from cells, possesses anti-inflammatory, antioxidant, anti-apoptotic properties, and also influences innate immunity (Mineo et al., 2012). However, the relationship between apoA1 levels and BC risk is not clear, as the results of published studies are contradictory (Borgquist et al., 2016; Huang et al., 2006). The objective of this study was to investigate the prospective associations between TC, HDL-C, LDL-C, apoA1, apoB, and triglycerides (TG) levels and breast cancer risk during the indicated follow-up period (5 years), and to determine whether some lipoproteins acted as prognostic factors in BC.

Methods

Patients

We retrospectively collected the clinicopathological data for BC patients who were first diagnosed with and then underwent modified radical mastectomy for BC between January 2007 and May 2011 at Shantou Affiliated Hospital of Sun Yat-Sen University. The following patients were included in the study: patients with confirmed IDC by pathology, and patients who had not received any anti-tumor treatments before blood sampling for biochemical data collection. The following patients were excluded from the study: (1) patients with coexisting cancers; (2) patients lacking of original blood biochemical test samples acquired before they began treatment; (3) patients with diseases characterized by abnormal lipid metabolism diseases; (4) patients who had taken lipid-lowering drugs, anti-diabetic drugs (statins, fibrates, oral anti-diabetics, insulin) or corticosteroids within the previous year; (5) patients lacking of follow-up data; and (6) patients lacking of other necessary information. Treatments were determined by the clinicopathological stages and patient characteristics, according to the institutional protocols (in accordance with the NCCN BC guidelines), and were not modified specifically for this study. We also collected blood samples from 200 healthy donors, who served as a control group for biomarkers assay.

Follow-up and study endpoint

Post-surgery follow-up visits were scheduled every 3 months during first 2 years after treatment, every six months during the second 2 years after treatment and then annually thereafter to determine if patients had relapsed or died. The last follow-up date on which the final conditions of all available patients were confirmed was May 2016 and the median follow-up time was 6.5 years.

One of the primary endpoints of the study was OS which was defined as the period of time extending from the date of definite pathological diagnosis to the date of death or the date of the last follow-up. DFS was the other primary endpoint and was defined as the period of time extending from the date of definite pathological diagnosis to the date of local recurrence or distant metastasis, death, or new neoplasm diagnosis.

Clinical data collection

Each patient’s medical history, age, BMI, menopause status, pathology parameters (such as tumor size, lymph node status, hormone status, HER-2 status, and histological grade), and laboratory data were collected. The clinical stages of the diseases used TNM staging system according to the AJCC (American Joint Committee on Cancer Classification, 7th edition, http://www.cancerstaging.org). The Immunohistochemical subtypes were determined according to 2013 St. Gallen Consensus Conference.

Serum lipid and lipoprotein assay

Fasting lipid profiles were measured at the time of diagnosis. Blood samples were collected into EDTA-coated tubes and the serum levels of glucose, TC, LDL-C, HDL-C, TG, apoA1, and apoB levels were measured automatically by electrophoresis (Hitachi Automatic Analyzer 7600-020; Hitachi, Tokyo, Japan).

Statistical analysis

Continuous variables are presented as means (standard deviations) for normally distributed data or medians (interquartile ranges) for non-normally distributed data. Categorical variables are presented as absolute values and frequencies. Spearman rank correlations coefficients were calculated to examine the correlations between continuous variables.

X-tile software (version 3.6.1, Yale University School of Medicine, New Haven, CT) was utilized to optimize the cutoff points for plasma apoA1 levels, according to breast cancer patients’ outcomes (Cai et al., 2011; Camp et al., 2004). Statistical significance was assessed by using the cutoff values derived from a training set to parse a separate validation set with a standard log-rank method. Values were obtained from an index table. The X-tile plots allowed the determination of optimal cutoff values while correcting for the use of minimal statistics using the Miller-Siegmund minimal p value correction (Raeside, 1976).

Univariate analysis of relationships between lipid profiles, BC risk factors (menstruation status), traditional prognostic factors (tumor size, lymph nodes positivity), estrogen receptor (ER) status, progesterone receptor (PR) status, and Her2-neu receptor (HER2) status was performed using parametric tests for variables with a normal distribution and non-parametric tests for variables without a normal distribution.

Univariate logistic regression analysis of the relationships between the above parameters and the risk of IDC incidence and the relationships between the above parameters and tumor size and axillary lymph node stage was performed. Then, multivariate logistic regression analysis was performed by inducing the parameters that were significant difference in the univariate analysis, in which stepwise conditional forward analysis method was used (entry 0.05; removal 0.10).

Kaplan-Meier curves were used to determine OS and DFS rates using log-rank tests. Cox proportional hazards models were used to estimate hazard ratios with 95% confidence intervals (CIs), that related apoA1 levels to DFS and OS. The multivariate Cox model was adjusted for tumor T stage, N stage, and subtype.

Table 1.

Table 1. Clinical characteristics of the breast cancer patients in this study.

Likelihood ratio p values are reported for whole variables in the model. All p values are two-tailed. Statistical analysis was performed using IBM SPSS Statistics for Windows, Version 19.0 (IBM Corp., Armonk, NY), and GraphPad Prism5.

Results

Characteristics of the subjects

A total of 702 women were recruited for the study. A total of 211 patients were excluded from the study because they did not satisfy the inclusion criteria. 15 patients were excluded from the study because they suffered from coexisting cancers and 57 patients had no original blood biochemical tests before treatment. There were 68 patients suffering from metabolic diseases, or taking anti-dysmetabolic medicines, or having BMI≥25. There were also 31 patients lacking follow-up data and 21 patients without the required information. Ultimately, 299 patients were included in this study.

Table 2.

Table 2. Tumor-related characteristics of the IDC patients.

Table 1 showed selected characteristics of patients and control subjects included in the study. Both groups had similar average ages, the proportion of pre- or postmenopausal individuals, weight, height, and BMI. Significant differences in lipid and lipoprotein components were noted between control subjects and patients in Table 1. The clinical characteristics of the breast cancer patients were shown in Table 2.

Univariate and multivariate logistic regression for the risk of IDC incidence

Univariate logistic regression was performed to calculate the relative risk of IDC according to each potential BC risk factor. No statistically significant relationships were observed between TC levels, TG levels, apoB levels, the apoA/apoB ratio, the HDL-C/apoA ratio, BMI and menstruation status and IDC risk (Table 3).

Multivariate logistic regression was performed to calculate the risk of IDC according to patient HDL-C, LDL-C, and apoA1 levels and age, all of which were found to be significantly associated with IDC incidence in univariate regression analysis. A strong negative association between apoA1 and IDC risk was demonstrated in this model [OR: 0.092, 95%CI (0.028, 0.305), p<0.001], whereas the age and LDL-C increased the risk of IDC prominently [age: OR 1.064, 95%CI (1.030, 1.099), p<0.001; LDL-C: OR 3.750, 95%CI (2.213, 6.355), p<0.001] (Table 3).

Table 3.

Table 3. Univariate and multivariate logistic regression for the risk of IDC incidence.

Spearman correlations for the relationships between lipid profiles and tumor size or axillary lymph node positivity

Analysis of the correlations between age, BMI and lipid profiles and primary tumor size or axillary lymph node positivity, showed that systemic LDL-C and apoA1 levels and the apoA1/apoB ratio were negatively correlated with tumor size (Spearman r=-0.198, p=0.007; Spearman r=-0.219, p=0.003; Spearman r=-0.215, p<0.001; respectively). However, the numbers of positive lymph nodes and the HDL-C/apoA1 ratio were positively correlated with tumor size (Spearman r=0.279, p=0.0001; Spearman r=0.233, p=0.001, respectively). Moreover, plasma apoA1 levels, the apoA1/apoB ratio, and the HDL-C/apoA1 ratio were negatively correlated with the number of positive lymph nodes (Spearman r=-0.180, p=0.002; Spearman r=-0.145, p=0.012; Spearman r=0.204, p<0.001; respectively). There were no statistical correlations between the other lipid profile parameters and tumor size or lymph node positivity, irrespective of disease stages (Additional Table 1, Additional Table 2).

Figure 1.

Figure 1. Tumor characteristics in apoA1 tertile groups. A. Tumor size increases across apoA1 tertile groups. Each bar represents the mean value of tumor size in each apoA1 tertile. B. Frequency of tumor characteristics in apoA1 tertile groups. Kruskal-Wallis test. The frequency of apoA1 L3 was markedly higher than L1 and L2 in ER positive subtypes (Luminal A; Luminal B Her-2(-); Luminal B Her-2(+)); the frequency of apoA1 L3 was lower than L1 and L2 in TNBC.

Cutoff values for pretreatment ApoA1 levels and univariate associations between ApoA1 levels and clinicopathological characteristics

The study population was stratified according to the following apoA1 level cutoff values: apoA1 L1: apoA1≤ 1.11 g/L; apoA1 L2: 1.29 g/L ≥apoA1 >1.11 g/L, and apoA1 L3: apoA1>1.29 g/L. Patients in the first apoA1 level had larger tumors (p<0.001) (Figure 1A) and more metastasis lymph nodes (p=0.05) (Figure 1B), and were diagnosed at more advanced stages than patients in the other levels. Patients with hormone receptors-positive BC presented with higher plasma apoA1 levels than patients with hormone receptor-negative BC (Figure 1B). However, patients with the more invasive breast cancer subtype TNBC were more likely to be within the first apoA1 level (Figure 1B).

Univariate and multivariate logistic regression for the predictors of tumor size or axillary lymph node positivity

Multivariate logistic regression for the predictors of tumor T stage was modeled. All the variables that were significantly associated tumor T stage in the univariate analysis (Table 4) were introduced in the model. We found that HDL-C levels, apoA1 levels, the apoA1/apoB ratio, and the HDL-C/apoA1 ratio were predictive factors for tumor sizes >T1 at diagnosis (Table 4). HDL-C levels and the HDL-C/apoA1 ratio increased the risk of a large tumor size with the OR 2.563 and 6.437 respectively. However, apoA1 reduced much more risk of big tumor size by 80%, and the apoA1/apoB ratio reduced the risk of a large tumor size by 50%.

Another multivariate logistic regression model was used to determine which of the above-mentioned variables were associated with lymph node metastasis in the univariate analysis (Table 5) were introduced in this model. The analysis showed the apoA1 levels and the ratio of apoA1/apoB could be negative predictors of metastasis when more than 3 lymph nodes (N1 stage) were present, as such a scenario reduced the risk of lymph node metastasis by approximately 50%.

Table 4.

Table 4. Univariate and multivariate logistic regression for the risk of tumor sizes >T1.

Survival and Cox regression model

At 60 months of follow-up the DFS rates in apoA1 L1, apoA1 L2, and apoA1 L3 groups were 74.5%, 89.9%, and 93.1%, respectively (log-rank test 0.0002) (Figure 2A) and the OS rates in apoA1 L1, apoA1 L2, and apoA1 L3 groups were 78.2%, 91.3%, and 93.7%, respectively (log-rank test 0.012) (Figure 2B) after 60 months of following-up. We used a Cox regression model to assess the relationships between various parameters and DFS and OS after adjusting for tumor T and N stages, and BC immunohistochemical subtypes. The regression model showed that apoA1 levels lower than 1.12 g/L at diagnosis were associated with poor DFS and OS [DFS: HR=2.628, 95%CI (1.339, 4.936), p=0.03; OS: HR=2.717, 95%CI (1.273, 5.802), p=0.01] (Tables 6 and 7).

Table 5.

Table 5. Univariate and multivariate logistic regression for the predictors of metastasis LNs >N1.

Discussion

BC development and progression comprise multiple processes that are influenced by environmental factors, lipid and lipoprotein metabolism, and the tumor microenvironment (Giovannucci et al., 1995; Hagymasi et al., 2007). The role of deregulated metabolism of lipid and lipoprotein metabolism in breast cancer remains under investigation. This retrospective study has shown that significant differences in plasma lipid and lipoprotein components exist between healthy donors and IDC patients (Table 1); however, only age, apoA1, and LDL-C levels were correlated with the risk of IDC incidence (Table 3). The results of this study also revealed that apoA1 can be a protective factor for IDC, as it was associated with a reduced risk of IDC incidence (Table 3), and an independent diagnostic factor for IDC, as low apoA1 levels predicted poorer DFS and OS in IDC (Figure 2; Tables 6 and 7).

Figure 2.

Figure 2. Overall and disease-free survival in apoA1 groups. Kaplan-Meier Curves. A. At 60 months, disease-free survival is 74.5% in apoA1 L1, 89.9% in apoA1 L2, and 93.1% in apoA1 L3 (Log rank test p=0.0002). B. At 60 months, overall survival is 78.2% in apoA1 L11, 91.3% in apoA1 L2, and 93.7% in apoA1 L3 (Log rank test p=0.0012).

The results of previous studies on the relationship between apoA1 and breast cancer were mixed. One of small case-control studies showed that elevated plasma apoA1 levels contributed to a high risk of breast cancer in Chinese women after adjustment for BMI (Han et al., 2005). In contrast, another study revealed the existence of an inverse correlation between apoA1 levels and breast cancer risk in Taiwanese women, in which BMI was not adjusted for (Chang et al., 2007). A large nested case-control study showed high apoA1 levels increased the risk of BC. That study did not adjust for BMI because its authors found that BMI was no association with lipoprotein profile parameters (Martin et al., 2015). Our results regarding apoA1 levels supported the hypothesis that high apoA1 levels reduced the risk of IDC, one of the most important pathological types of BC, when all the lipid component markers and BMI were adjusted for the study.

ApoA1 acts as a protective factor for the IDC development and progression. ApoA1 is the major protein constituent of HDL-C and plays an important role in reverse cholesterol transport by extracting cholesterol and phospholipids from peripheral cells and transferring them to the liver for excretion. The role of apoA1 in cancer is under investigation. Decreased apoA1 levels have been reported in the serum of patients with pancreatic cancergastric cancer, and ovarian cancer (Ehmann et al., 2007; Kozak et al., 2005). Studies using in vivo models have shown that mice lacking apoA1 develop tumors at an extremely rapid pace and are at a significant survival disadvantage, whereas mice expressing apoA1 exhibit dramatically restrained tumor progression and improved survival in a dose-dependent manner (Zamanian-Daryoush et al., 2013). ApoA1 represses tumor development and progression via two prominent mechanisms. ApoA1 first inhibits the tumor-associated angiogenesis, and then reduces protein expression of MMP9 which is a critical matrix-degrading enzyme needed for metastasis (Bauvois, 2012). The inhibitory role of apoA1 with respect to tumor growth and metastasis was supported by our clinical data, which showed that apoA1 expression levels were negatively associated with not only the incidence of IDC, but also tumor sizes and lymph node stages (Figure 1; Tables 3, 4, and 5). ApoA1 was an independent factor for the IDC prognosis, and lower apoA1 levels were strongly associated with a higher risk of IDC incidence, and poorer DFS and OS (Tables 6 and 7; Figure 2). Moreover, the levels of apoA1 found in breast tumors were positively correlated with chemotherapyresistance in malignant tumors (Morale et al., 2013). Breast tumors sensitive to chemotherapy secreted more apoA1 than those not sensitive to chemotherapy, whereas normal tissues secreted the largest amount of apoA1 (Lee et al., 2008).

Table 6.

Table 6. Cox multivariate regression model for disease-free survival.

ApoA1 contributes more to hormone receptor positive IDC than to hormone receptor-negative IDC by inducing the estrogen receptor (ER) expression. ERa, which is encoded by the estrogen receptor 1 (ESR1) gene has been demonstrated to influence the expression of proteins involved in regulating plasma lipid metabolism. The results regarding the association between ESR1 polymorphisms and lipid profile parameter levels, such as TC, TG, HDL-C, LDL-C, and apoA1 levels were mixed (Kikuchi et al., 2000; Sosa et al., 2004). Plasma apoA1 binds to ATP-binding cassette (ABC) lipid transporters, which regulates breast cancer ER/PR expression in vitro and in vivo by inducing Cdc42 and PAK-1 signaling (Bourguignon et al., 2005; Holm et al., 2006). In this study, we found that most patients with estrogen receptor/progesterone positive IDC had high plasma apoA1 levels. In contrast, more TNBC patients were in the first apoA1 level (<1.11 g/L) than in the third level (>1.29 g/L) (Figure 1B). This result revealed that the tumor with high plasma apoA1 levels was likely to be luminal-type breast cancer than other types of breast cancer. ApoA1 levels may be useful as a predictive marker for prognosis and sensitivity to the hormone therapy.

Table 7.

Table 7. Cox multivariate regression model for overall survival.

There were limitations to this study. First, this was a retrospective study based on a single-institutional database. Thus, the power of the study was reduced because of inevitable selection bias. Second, indicators of invasiveness, such as Ki67 and lymphovascular invasion (LVI) were not included in the whole model. Moreover, the detailed mechanisms underlying the effects of apoA1 and the biological significance of apoA1 in other pathological types of breast cancer should be investigated in future studies. In conclusion, to our knowledge, this retrospective study was the first report to evaluate the prognostic value of the pretreatment serum apoA1 levels in IDC. The findings of our study demonstrate that the apoA1 is a protective factor for IDC development and progression, and is an independent predictive biomarker for DFS and OS in IDC. Additional clinical trials, including prospective trails and multiple center studies, are needed to better define the role of apoA1 in BC.

Disclosure

The authors report no conflicts of interest.

Corresponding Author

Zhiyong Wu, Ph.D., Diagnosis and Treatment Center of Breast Diseases, Shantou Affiliated Hospital of Sun Yat-Sen University, 114 Waima Rd., Shantou, Guangdong, China.

References

Ackerman D, Simon MC. Hypoxia, lipids, and cancer: surviving the harsh tumor microenvironment. Trends Cell Biol 24(8):472-478, 2014.

Bauvois B. New facets of matrix metalloproteinases MMP-2 and MMP-9 as cell surface transducers: outside-in signaling and relationship to tumor progression. Biochim Biophys Acta 1825(1):29-36, 2012.

Borgquist S, Butt T, Almgren P, Shiffman D, Stocks T, Orho-Melander M et al. Apolipoproteins, lipids and risk of cancer. Int J Cancer 138(11):2648-2656, 2016.

Bourguignon LY, Gilad E, Rothman K, Peyrollier K. Hyaluronan-CD44 interaction with IQGAP1 promotes Cdc42 and ERK signaling, leading to actin binding, Elk-1/estrogen receptor transcriptional activation, and ovarian cancer progression. J Biol Chem 280(12):11961-11972, 2005.

Cai MY, Tong ZT, Zhu W, Wen ZZ, Rao HL, Kong LL et al. H3K27me3 protein is a promising predictive biomarker of patients’ survival and chemoradioresistance in human nasopharyngeal carcinomaMol Med 17(11-12):1137-1145, 2011.

Camp RL, Dolled-Filhart M, Rimm DL. X-tile: a new bio-informatics tool for biomarker assessment and outcome-based cut-point optimization. Clin Cancer Res 10(21):7252-7259, 2004.

Chang SJ, Hou MF, Tsai SM, Wu SH, Hou LA, Ma H, et al. The association between lipid profiles and breast cancer among Taiwanese women. Clin Chem Lab Med 45(9):1219-1223, 2007.

Currie E, Schulze A, Zechner R, Walther TC, Farese RJ. Cellular fatty acid metabolism and cancer. Cell Metab 18(2):153-161, 2013.

Ehmann M, Felix K, Hartmann D, Schnolzer M, Nees M, Vorderwulbecke S et al. Identification of potential markers for the detection of pancreatic cancer through comparative serum protein expression profiling. Pancreas 34(2):205-214, 2007.

Eliassen AH, Colditz GA, Rosner B, Willett WC, Hankinson SE. Serum lipids, lipid-lowering drugs, and the risk of breast cancer. Arch Intern Med 165(19):2264-2271, 2005.

Eliassen AH, Colditz GA, Rosner B, Willett WC, Hankinson SE. Serum lipids, lipid-lowering drugs, and the risk of breast cancer. Arch Intern Med 165(19):2264-2271, 2005.

Fagherazzi G, Fabre A, Boutron-Ruault MC, Clavel-Chapelon F. Serum cholesterol level, use of a cholesterol-lowering drug, and breast cancer: results from the prospective E3N cohort. Eur J Cancer Prev 19(2):120-125, 2010.

Ferlay J, Soerjomataram I, Dikshit R, Eser S, Mathers C, Rebelo M et al. Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer136(5):E359-E386, 2015.

Gaard M, Tretli S, Urdal P. Risk of breast cancer in relation to blood lipids: a prospective study of 31,209 Norwegian women. Cancer Causes Control 5(6):501-509, 1994.

Giovannucci E, Ascherio A, Rimm EB, Colditz GA, Stampfer MJ, Willett WC. Physical activityobesity, and risk for colon cancer and adenoma in men. Ann Intern Med 122(5):327-334, 1995.

Ha M, Sung J, Song YM. Serum total cholesterol and the risk of breast cancer in postmenopausal Korean women. Cancer Causes Control 20(7):1055-1060, 2009.

Hagymasi K, Tulassay Z. [Role of obesity in colorectal carcinogenesis]. Orv Hetil 148(51):2411-2416, 2007.

Han C, Zhang HT, Du L, Liu X, Jing J, Zhao X et al. Serum levels of leptin, insulin, and lipids in relation to breast cancer in China. Endocrine 26(1):19-24, 2005.

Hiatt RA, Fireman BH. Serum cholesterol and the incidence of cancer in a large cohort. J Chronic Dis39(11):861-870, 1986.

His M, Zelek L, Deschasaux M, Pouchieu C, Kesse-Guyot E, Hercberg S et al. Prospective associations between serum biomarkers of lipid metabolism and overall, breast and prostate cancerrisk. Eur J Epidemiol 29(2):119-132, 2014.

Holm C, Rayala S, Jirstrom K, Stal O, Kumar R, Landberg G. Association between Pak1 expression and subcellular localization and tamoxifen resistance in breast cancer patients. J Natl Cancer Inst98(10):671-680, 2006.

Hoyer AP, Engholm G. Serum lipids and breast cancer risk: a cohort study of 5,207 Danish women. Cancer Causes Control 3(5):403-408, 1992.

Huang HL, Stasyk T, Morandell S, Dieplinger H, Falkensammer G, Griesmacher A et al. Biomarker discovery in breast cancer serum using 2-D differential gel electrophoresis/MALDI-TOF/TOF and data validation by routine clinical assays. Electrophoresis 27(8):1641-1650, 2006.

Huxley R, Lewington S, Clarke R. Cholesterol, coronary heart disease and stroke: a review of published evidence from observational studies and randomized controlled trials. Semin Vasc Med2(3):315-323, 2002.

Iso H, Ikeda A, Inoue M, Sato S, Tsugane S. Serum cholesterol levels in relation to the incidence of cancer: the JPHC study cohorts. Int J Cancer 125(11):2679-2686, 2009.

Iso H, Ikeda A, Inoue M, Sato S, Tsugane S. Serum cholesterol levels in relation to the incidence of cancer: the JPHC study cohorts. Int J Cancer 125(11):2679-2686, 2009.

Kaye JA, Meier CR, Walker AM, Jick H. Statin use, hyperlipidaemia, and the risk of breast cancer. Br J Cancer 86(9):1436-1439, 2002.

Kikuchi T, Hashimoto N, Kawasaki T, Uchiyama M. Association of serum low-density lipoprotein metabolism with oestrogen receptor gene polymorphisms in healthy children. Acta Paediatrica89(1):42-45, 2000.

Kitahara CM, Berrington DGA, Freedman ND, Huxley R, Mok Y, Jee SH et al. Total cholesterol and cancer risk in a large prospective study in Korea. J Clin Oncol 29(12):1592-1598, 2011.

Knekt P, Reunanen A, Aromaa A, Heliovaara M, Hakulinen T, Hakama M. Serum cholesterol and risk of cancer in a cohort of 39,000 men and women. J Clin Epidemiol 41(6):519-530, 1988.

Kozak KR, Su F, Whitelegge JP, Faull K, Reddy S, Farias-Eisner R. Characterization of serum biomarkers for detection of early stage ovarian cancer. Proteomics 5(17):4589-4596, 2005.

Lee HJ, Paul S, Atalla N, Thomas PE, Lin X, Yang I et al. Gemini vitamin D analogues inhibit estrogen receptor-positive and estrogen receptor-negative mammary tumorigenesis without hypercalcemic toxicity. Cancer Prev Res (Phila) 1(6):476-484, 2008.

Lingwood D, Simons K. Lipid rafts as a membrane-organizing principle. Science 327(5961):46-50, 2010.

Martin LJ, Melnichouk O, Huszti E, Connelly PW, Greenberg CV, Minkin S et al. Serum lipids, lipoproteins, and risk of breast cancer: a nested case-control study using multiple time points. J Natl Cancer Inst 107(5):djv032, 2015.

Melvin JC, Holmberg L, Rohrmann S, Loda M, Van Hemelrijck M. Serum lipid profiles and cancer risk in the context of obesity: four meta-analyses. J Cancer Epidemiol 2013:823849, 2013.

Mineo C, Shaul PW. Novel biological functions of high-density lipoprotein cholesterol. Circ Res111(8):1079-1090, 2012.

Morale W, Patane D, Incardona C, Seminara G, Malfa P, L’Anfusa G et al. [Project work: formation of health-care personnel for self-care of tunnelled central venous catheters in hemodialysis patients of the territory]. G Ital Nefrol 30(4), 2013.

Raeside DE. Monte Carlo principles and applications. Phys Med Biol 21(2):181-197, 1976.

Silvente-Poirot S, Poirot M. Cancer. Cholesterol and cancer, in the balance. Science 343(6178):1445-1446, 2014.

Soran H, Hama S, Yadav R, Durrington PN. HDL functionality. Curr Opin Lipidol 23(4):353-366, 2012.

Sosa M, Jodar E, Arbelo E, Dominguez C, Saavedra P, Torres A et al. Serum lipids and estrogen receptor gene polymorphisms in male-to-female transsexuals: effects of estrogen treatment. Eur J Intern Med 15(4):231-237, 2004.

Steenland K, Nowlin S, Palu S. Cancer incidence in the National Health and Nutrition Survey I. Follow-up data: diabetes, cholesterol, pulse and physical activity. Cancer Epidemiol Biomarkers Prev4(8):807-811, 1995.

Strohmaier S, Edlinger M, Manjer J, Stocks T, Bjorge T, Borena W et al. Total serum cholesterol and cancer incidence in the Metabolic syndrome and Cancer Project (Me-Can). PLoS One 8(1):e54242, 2013.

Tornberg SA, Holm LE, Carstensen JM. Breast cancer risk in relation to serum cholesterol, serum beta-lipoprotein, height, weight, and blood pressure. Acta Oncologica 27(1):31-37, 1988.

Touvier M, Fassier P, His M, Norat T, Chan DS, Blacher J et al. Cholesterol and breast cancer risk: a systematic review and meta-analysis of prospective studies. Br J Nutr 114(3):347-357, 2015.

Tulinius H, Sigfusson N, Sigvaldason H, Bjarnadottir K, Tryggvadottir L. Risk factors for malignant diseases: a cohort study on a population of 22,946 Icelanders. Cancer Epidemiol Biomarkers Prev6(11):863-873, 1997.

Tulinius H, Sigfusson N, Sigvaldason H, Bjarnadottir K, Tryggvadottir L. Risk factors for malignant diseases: a cohort study on a population of 22,946 Icelanders. Cancer Epidemiol Biomarkers Prev6(11):863-873, 1997.

Vatten LJ, Foss OP. Total serum cholesterol and triglycerides and risk of breast cancer: a prospective study of 24,329 Norwegian women. Cancer Res 50(8):2341-2346, 1990.

Vatten LJ, Foss OP. Total serum cholesterol and triglycerides and risk of breast cancer: a prospective study of 24,329 Norwegian women. Cancer Res 50(8):2341-2346, 1990.

von Eckardstein A, Hersberger M, Rohrer L. Current understanding of the metabolism and biological actions of HDL. Curr Opin Clin Nutr Metab Care 8(2):147-152, 2005.

Yager JD, Davidson NE. Estrogen carcinogenesis in breast cancer. N Engl J Med 354(3):270-282, 2006.

Zamanian-Daryoush M, Lindner D, Tallant TC, Wang Z, Buffa J, Klipfell E et al. The cardioprotective protein apolipoprotein A1 promotes potent anti-tumorigenic effects. J Biol Chem 288(29):21237-21252, 2013

Leave a Reply

Your email address will not be published. Required fields are marked *