Variability in endometrial carcinoma pathology practice: opportunities for improvement with molecular classification.

We assessed the landscape of diagnostic pathology practice and how molecular classification could potentially impact management of patients with endometrial cancer by collecting patient samples, clinicopathologic data, and patient outcomes from EC patients diagnosed in 2016 at 10 Canadian tertiary cancer centers and 19 community hospitals. ProMisE molecular subtype (POLEmut, MMRd, p53abn, No Specific Molecular Profile (NSMP)) was assigned retrospectively. 1357 patients were fully evaluable including 85 POLEmut (6.3%), 380 MMRd (28.0%), 643 NSMP (47.4%), and 249 p53abn ECs (18.3%). Immunohistochemistry (IHC) for MMR proteins was undertaken at the time of primary diagnosis in 2016 in only 42% of the cohort (570/1357; range 3.5-95.4%/center). p53 IHC had only been performed in 21.1% of the cohort (286/1357; range 10.1-41.9%/center). Thus, based on the retrospective molecular subtype assignment, 54.7% (208/380) of MMRd EC had not been tested with MMR IHC (or MSI) and 48.2% (120/249) of p53abn ECs were not tested with p53 IHC in 2016. Molecular subtype diversity within histotypes was profound; most serous carcinomas were p53abn (91.4%), but only 129/249 (51.8%) p53abn EC were serous. Low-grade (Gr1-2) endometrioid carcinomas were mostly NSMP (589/954, 61.7%) but included all molecular subtypes, including p53abn (19/954, 2.0%). Molecular subtype was significantly associated with clinical outcomes (p < 0.001) even in patients with stage I disease (OS p = 0.006, DSS p < 0.001, PFS p < 0.001). Assessment of national pathologic practice in 2016 shows highly variable use of MMR and p53 IHC and demonstrates significant opportunities to improve and standardize biomarker reporting. Inconsistent, non-reflexive IHC resulted in missed opportunities for Hereditary Cancer Program referral and Lynch Syndrome diagnosis, and missed potential therapeutic implications (e.g., chemotherapy in p53abn EC, immune blockade for MMRd EC). Routine integration of molecular subtyping into practice can improve the consistency of EC pathology assessment and classification.

Association between alcohol consumption and DNA methylation in blood: a systematic review of observational studies.

Aim: We systematically reviewed and evaluated current literature on alcohol consumption and DNA methylation (DNAm) at the genome-wide and probe-wise level in blood of adults. Materials & methods: Five databases (PubMed, Embase, Web of Science, CINAHL and PsycInfo) were searched until 20 December 2020. Studies assessing the effect of alcohol dependence on DNAm were not eligible. Results: 11 cross-sectional studies were included with 88 to 9643 participants. Overall, all studies had a risk of bias criteria unclear or unmet. Epigenome-wide association studies identified between 0 and 5458 differentially methylated positions, and 15 were observed in at least four studies. Conclusion: Potential methylation markers for alcohol consumption have been identified, but further validation in large cohorts is needed.

Endometrial carcinoma molecular subtype correlates with the presence of lymph node metastases.

BACKGROUND: The role of lymph node assessment/dissection (LND) in endometrial cancer (EC) has been debated for decades, with significant practice variation between centers. Molecular classification of EC provides prognostic information and can be accurately performed on preoperative endometrial biopsies. We assessed the association between molecular subtype and lymph node metastases (LNM) in order to determine if this tool could be used to stratify surgical decision making. METHODS: All EC patients undergoing primary staging surgery with planned complete pelvic +/- para-aortic LND from a single institution in the 2015 calendar year were identified, with clinicopathological and outcome data assessed in the context of retrospectively assigned molecular classification. RESULTS: 172 patients were included. Molecular classification of the total cohort showed 21 POLEmut (12.2%), 47 MMRd (27.3%), 74 NSMP (43.1%), and 30 p53abn (17.4%) ECs. Complete pelvic +/- para-aortic LND was performed in 171 of 172 patients, and LNM were found in 31/171 (18.1%). This included macrometastases (19/31), micrometastases (5/31), and isolated tumour cells (ITCs) (7/31). LNM were pelvic only in 83.9%, and pelvic plus para-aortic in 16.1%. There were no isolated para-aortic LNM. Molecular subtype was significantly associated with LNM (p = 0.004). There was a strong association between the presence of LNM and p53abn EC (nodal involvement in 44.8% of cases), with LNM detected in 14.2% of POLEmut, 14.9% of MMRd, and 10.8% of NSMP EC. On multivariate analysis, molecular subtype and preoperative CA 125 > 25 were significantly associated with LNM (p = 0.021 and p = 0.022 respectively) but preoperative grade and histotype were not (p = 0.24). CONCLUSION: EC molecular subtype is significantly associated with the presence of LNM. As molecular classification can be obtained on preoperative diagnostic specimens, this information can be used to guide surgical treatment planning and may reduce the cost and morbidity of unnecessary lymph node staging in EC care.

Association of Breast Tumour Expression of Cannabinoid Receptors CBR1 and CBR2 with Prognostic Factors and Survival in Breast Cancer Patients.

Cannabinoid receptors (CBR) are potential therapeutic targets for breast cancer. However, the role of CBR in breast cancer survival remains poorly understood. Data from a prospective cohort of 522 women diagnosed with invasive breast cancer between 2010 and 2012 were analysed. Clinical and pathological features were retrieved from electronic medical records. CBR expression was measured by immunohistochemistry. Adjusted partial Spearman correlations and multivariate Cox models were used to estimate associations with breast cancer prognostic factors and survival, respectively. The median follow-up was 92.0 months (range 7.0-114.0). CBR expression was heterogenous in tumours. Cytoplasmic expression of CBR1 was positively correlated with lymph node invasion (rs = 0.110; p = 0.0155) and positive status of the human epidermal growth factor receptor 2 (HER2) (rs = 0.168; p = 0.0002), while nuclear CBR2 was negatively correlated with grade (rs = -0.171; p = 0.0002) and positively correlated with oestrogen receptor and progesterone receptor-positive status (rs = 0.173; p = 0.0002 and rs = 0.121; p = 0.0084, respectively). High cytoplasmic expression of CBR2 was associated, with 13% higher locoregional and distant recurrences (HR = 1.13 [0.97-1.33]), though this association did not reach statistical significance. Although the few events occurring during follow-up may have limited the detection of significant associations, these results indicate that CBR expression in breast cancer deserves further investigation.

TBC1D9: An Important Modulator of Tumorigenesis in Breast Cancer.

Triple-negative breast cancer (TNBC) is a major concern among the different subtypes of breast cancer (BC) due to the lack of effective treatment. In a previous study by our group aimed at understanding the difference between TNBC and non-TNBC tumors, we identified the gene TBC1 domain family member 9 (TBC1D9), the expression of which was lower in TNBC as compared to non-TNBC tumors. In the present study, analysis of TBC1D9 expression in TNBC (n = 58) and non-TNBC (n = 25) patient tumor samples validated that TBC1D9 expression can differentiate TNBC (low) from non-TNBC (high) samples and that expression of TBC1D9 was inversely correlated with grade and proliferative index. Moreover, we found that downregulation of the TBC1D9 gene decreases the proliferation marginally in non-TNBC and was associated with increased migratory and tumorigenic potential in both TNBC and luminal BC cell lines. This increase was mediated by the upregulation of ARL8A, ARL8B, PLK1, HIF1α, STAT3, and SPP1 expression in TBC1D9 knockdown cells. Our results suggest that TBC1D9 expression might limit tumor aggressiveness and that it has a differential expression in TNBC vs. non-TNBC tumors.

Is Carboxypeptidase B1 a Prognostic Marker for Ductal Carcinoma In Situ?

Ductal carcinoma in situ (DCIS) is considered a non-obligatory precursor for invasive ductal carcinoma (IDC). Around 70% of women with atypical ductal hyperplasia (ADH) undergo unnecessary surgery due to the difficulty in differentiating ADH from low-grade DCIS. If untreated, 14-60% of DCIS progress to IDC, highlighting the importance of identifying a DCIS gene signature. Human transcriptome data of breast tissue samples representing each step of BC progression were analyzed and high expression of carboxypeptidase B1 (CPB1) expression strongly correlated with DCIS. This was confirmed by quantitative PCR in breast tissue samples and cell lines model. High CPB1 expression correlated with better survival outcome, and mRNA level was highest in DCIS than DCIS adjacent to IDC and IDC. Moreover, loss of CPB1 in a DCIS cell line led to invasive properties associated with activation of HIF1α, FN1, STAT3 and SPP1 and downregulation of SFRP1 and OS9. The expression of CPB1 could predict 90.1% of DCIS in a cohort consisting of DCIS and IDC. We identified CPB1, a biomarker that helps differentiate DCIS from ADH or IDC and in predicting if a DCIS is likely to progress to IDC, thereby helping clinicians in their decisions.

Associations between markers of mammary adipose tissue dysfunction and breast cancer prognostic factors.

BACKGROUND: Obesity fosters worse clinical outcomes in both premenopausal and postmenopausal women with breast cancer. Emerging evidence suggests that an android body fat distribution in particular is deleterious for breast cancer prognosis. The extent of adipose tissue dysfunction, especially how it relates to breast cancer prognostic factors and anthropometric measurements, has not been fully investigated. OBJECTIVE: Our objective was to examine if markers of adipose tissue dysfunction, such as hypertrophy and macrophage accumulation, are relevant for the pathophysiology of breast cancer and its associated prognostic factors in a well-characterised cohort of women with breast cancer who did not receive treatment before surgery. METHODS: A consecutive series of 164 women with breast cancer provided breast adipose tissue sample. Multivariate generalised linear models were used to test associations of anthropometric indices and prognostic factors with markers of adipose tissue dysfunction. RESULTS: We found associations of breast adipocyte size and macrophage infiltration (number of CD68+ cells/100 adipocytes) with adiposity, particularly a strong association between breast adipocyte size and central obesity, independent of total adiposity, age and menopausal status (ßadj = 0.87; p = 0.0001). We also identified relationships of adipocyte hypertrophy and macrophage infiltration with prognostic factors, such as cancer stage and tumour grade (p < 0.05). RNA expression of pro-inflammatory cytokines (IL6, TNF) and leptin was also increased as a function of adipocyte size and CD86+/CD11c+ macrophage number/100 adipocytes (p < 0.05). CONCLUSIONS: Our findings support the model of dysfunctional adipose tissue in obesity-associated breast cancer.

Epigenome-wide DNA methylation and risk of breast cancer: a systematic review.

BACKGROUND: DNA methylation is a potential biomarker for early detection of breast cancer. However, robust evidence of a prospective relationship between DNA methylation patterns and breast cancer risk is still lacking. The objective of this study is to provide a systematic analysis of the findings of epigenome-wide DNA methylation studies on breast cancer risk, in light of their methodological strengths and weaknesses. METHODS: We searched major databases (MEDLINE, EMBASE, Web of Science, CENTRAL) from inception up to 30th June 2019, for observational or intervention studies investigating the association between epigenome-wide DNA methylation (using the HM450k or EPIC BeadChip), measured in any type of human sample, and breast cancer risk. A pre-established protocol was drawn up following the Cochrane Reviews rigorous methodology. Study selection, data abstraction, and risk of bias assessment were performed by at least two investigators. A qualitative synthesis and systematic comparison of the strengths and weaknesses of studies was performed. RESULTS: Overall, 20 studies using the HM450k BeadChip were included, 17 of which had measured blood-derived DNA methylation. There was a consistent trend toward an association of global blood-derived DNA hypomethylation and higher epigenetic age with higher risk of breast cancer. The strength of associations was modest for global hypomethylation and relatively weak for most of epigenetic age algorithms. Differences in length of follow-up periods may have influenced the ability to detect associations, as studies reporting follow-up periods shorter than 10 years were more likely to observe an association with global DNA methylation. Probe-wise differential methylation analyses identified between one and 806 differentially methylated CpGs positions in 10 studies. None of the identified differentially methylated sites overlapped between studies. Three studies used breast tissue DNA and suffered major methodological issues that precludes any conclusion. Overall risk of bias was critical mainly because of incomplete control of confounding. Important issues relative to data preprocessing could have limited the consistency of results. CONCLUSIONS: Global DNA methylation may be a short-term predictor of breast cancer risk. Further studies with rigorous methodology are needed to determine spatial distribution of DNA hypomethylation and identify differentially methylated sites associated with risk of breast cancer. PROSPERO REGISTRATION NUMBER: CRD42020147244.

DNA Methylation and Breast Cancer Risk: An Epigenome-Wide Study of Normal Breast Tissue and Blood.

Differential DNA methylation is a potential marker of breast cancer risk. Few studies have investigated DNA methylation changes in normal breast tissue and were largely confounded by cancer field effects. To detect methylation changes in normal breast epithelium that are causally associated with breast cancer occurrence, we used a nested case-control study design based on a prospective cohort of patients diagnosed with a primary invasive hormone receptor-positive breast cancer. Twenty patients diagnosed with a contralateral breast cancer (CBC) were matched (1:1) with 20 patients who did not develop a CBC on relevant risk factors. Differentially methylated Cytosine-phosphate-Guanines (CpGs) and regions in normal breast epithelium were identified using an epigenome-wide DNA methylation assay and robust linear regressions. Analyses were replicated in two independent sets of normal breast tissue and blood. We identified 7315 CpGs (FDR < 0.05), 52 passing strict Bonferroni correction (p < 1.22 × 10-7) and 43 mapping to known genes involved in metabolic diseases with significant enrichment (p < 0.01) of pathways involving fatty acids metabolic processes. Four differentially methylated genes were detected in both site-specific and regions analyses (LHX2, TFAP2B, JAKMIP1, SEPT9), and three genes overlapped all three datasets (POM121L2, KCNQ1, CLEC4C). Once validated, the seven differentially methylated genes distinguishing women who developed and who did not develop a sporadic breast cancer could be used to enhance breast cancer risk-stratification, and allow implementation of targeted screening and preventive strategies that would ultimately improve breast cancer prognosis.

Secreted Frizzled-Related Protein 1 as a Biomarker against Incomplete Age-Related Lobular Involution and Microcalcifications' Development.

As a downregulator of the Wnt signaling pathway, SFRP1 is involved in several components of the age-related lobular involution process such as inflammation, apoptosis, and adipogenesis. Because microcalcifications are associated with inflammation, we aimed to demystify the cross talk between SFRP1, inflammatory markers, and microcalcifications by assessing SFRP1 expression (immunohistochemistry) in a cohort of 162 women with different degrees of lobular involution. SFRP1 expression was inversely associated with the degree of lobular involution (OR = 0.84; p-value < 0.01). SFRP1 expression, age at mastectomy, and waist circumference taken together predicted the degree of lobular involution (AUC = 78.1). This predictive model was best in patients with microcalcifications (AUC = 81.1) and in parous women (AUC = 87.8). SFRP1 expression was correlated with leptin (rho = 0.32), TNF-α (rho = 0.21), and IL-6 (rho = 0.21) expression by epithelial cells (all p-values <0.001). SFRP1 expression was lower in nulliparous women with involuted breast tissue compared with parous women with involuted breast tissue (Δmean = -2.31; p-value < 0.01) and was higher in nulliparous women with microcalcifications compared with nulliparous women without microcalcifications (Δmean = 2.4; p-value < 0.05). In this study, we highlighted two SFRP1-based predictive models for incomplete lobular involution and the development of microcalcifications and identified two distinct inflammatory profiles associated with age-related lobular involution in parous and nulliparous women.

Association Between BMI and DNA Methylation in Blood or Normal Adult Breast Tissue: A Systematic Review.

BACKGROUND/AIM: Several studies have investigated the influence of obesity on DNA methylation (DNAm) to find biomarkers associated with the detection of chronic diseases, including breast cancer. The aim of the study was to systematically review studies examining the association of body mass index (BMI) and DNAm in blood or normal breast tissue. MATERIALS AND METHODS: Three scientific literature databases (PubMed, Embase and Web of Science) were screened until May 2018. RESULTS: Twenty-four studies were included along with ours in which we investigated this relation in the normal breast tissue of 40 breast cancer patients. CONCLUSION: BMI-associated CpG sites were highly variable with few identified in less than half of the studies. Nevertheless, a few genes potentially associated with BMI were highlighted in blood (CPT1A, ABCG1, SREBF1 and LGALS3BP) and in normal breast tissue (PTPRN2 and ABLIM2). The variability of the results could be explained by the tissue and cell-specificity of methylation and differences in methodology.

Progesterone receptor status modifies the association between body mass index and prognosis in women diagnosed with estrogen receptor positive breast cancer.

The role of progesterone receptor (PR) status on the association between obesity and prognosis of estrogen receptor positive (ER+) breast cancer (BC) remains poorly understood. We aim to examine whether this association varies according to the tumor PR status. Data for 3,747 women diagnosed with nonmetastatic ER+ invasive BC between 1995 and 2010 were analyzed. Women were classified according to their body mass index (BMI) (<18.5, 18.5-24.9, 25.0-29.9 or ≥30.0 kg/m2 ). Tumor PR status (PR-, PR+) was evaluated by immunohistochemistry. Hazard ratios (HR) for survival outcomes were estimated using multivariable Cox regression models. Effect modification was assessed on both additive and multiplicative scales using relative excess risk due to interaction and ratio of HRs, respectively. After a median follow-up of 5.9 years (range: 3.4-9.2), women with PR- tumors and underweight (HR = 2.76, 95% CI: 1.40-4.91), overweight (HR = 2.02, 95% CI: 1.43-2.81) or obese (HR = 2.51, 95% CI: 1.67-3.65) had increased risk of all-cause mortality, when compared to normal weight women with PR+ tumors. A similar pattern of associations was observed for BC-specific mortality. In contrast, women with PR+ tumors had similar risks for both mortality outcomes, regardless of BMI. On the additive scale, all-cause mortality was modified by PR status for overweight and obese women, whereas for BC-specific mortality, it was only modified for underweight women. The same observations were found on the multiplicative scale. These results suggest that poorer survival associated with low and high BMI among women diagnosed with ER+ BC may depend on the tumor PR status.

Persistent Organic Pollutants and Breast Cancer: A Systematic Review and Critical Appraisal of the Literature.

Persistent organic pollutants (POPs) bioaccumulate in the food chain and have been detected in human blood and adipose tissue. Experimental studies demonstrated that POPs can cause and promote growth of breast cancer. However, inconsistent results from epidemiological studies do not support a causal relationship between POPs and breast cancer in women. To identify individual POPs that are repeatedly found to be associated with both breast cancer incidence and progression, and to demystify the observed inconsistencies between epidemiological studies, we conducted a systematic review of 95 studies retrieved from three main electronic databases. While no clear pattern of associations between blood POPs and breast cancer incidence could be drawn, POPs measured in breast adipose tissue were more clearly associated with higher breast cancer incidence. POPs were more consistently associated with worse breast cancer prognosis whether measured in blood or breast adipose tissue. In contrast, POPs measured in adipose tissue other than breast were inversely associated with both breast cancer incidence and prognosis. Differences in biological tissues used for POPs measurement and methodological biases explain the discrepancies between studies results. Some individual compounds associated with both breast cancer incidence and progression, deserve further investigation.

An isoform of AIF1 involved in breast cancer.

BACKGROUND: Inflammation is a major player in breast cancer (BC) progression. Allograft-inflammatory factor-1 (AIF1) is a crucial mediator in the inflammatory response. AIF1 reportedly plays a role in BC, but the mechanism remains to be elucidated. We identified two AIF1 isoforms, AIF1v1 and AIF1v3, which were differentially expressed between affected and unaffected sisters from families with high risk of BC with no deleterious BRCA1/BRCA2 mutations (BRCAX). We investigated potential functions of AIFv1/v3 in BC of varying severity and breast adipose tissue by evaluating their expression, and association with metabolic and clinical parameters of BC patients. METHODS: AIF1v1/v3 expression was determined in BC tissues and cell lines using quantitative real-time PCR. Potential roles and mechanisms were examined in the microenvironment (fibroblasts, adipose tissue, monocytes and macrophages), inflammatory response (cell reaction in BC subgroups), and metabolism [treatment with docosahexaenoic acid (DHA)]. Association of AIF1 transcript expression with clinical factors was determined by Spearman's rank correlation. Bioinformatics analyses were performed to characterize transcripts. RESULTS: AIF1v1/v3 were mostly expressed in the less severe BC samples, and their expression appeared to originate from the tumor microenvironment. AIF1 isoforms had different expression rates and sources in breast adipose tissue; lymphocytes mostly expressed AIF1v1 while activated macrophages mainly expressed AIF1v3. Bioinformatics analysis revealed major structural differences suggesting distinct functions in BC progression. Lymphocytes were the most infiltrating cells in breast tumors and their number correlated with AIF1v1 adipose expression. Furthermore, DHA supplementation significantly lowered the expression of AIF1 isoforms in BRCAX cell lines. Finally, the expression of AIF1 isoforms in BC and breast adipose tissue correlated with clinical parameters of BC patients. CONCLUSIONS: Results strongly suggest that AIF1v1 as much as AIF1v3 play a major role in the crosstalk between BC and infiltrating immune cells mediating tumor progression, implying their high potential as target molecules for BC diagnostic, prognostication and treatment.

Telomere Length and Breast Cancer Prognosis: A Systematic Review.

Telomeres ensure genome integrity during replication. Loss of telomeric function leads to cell immortalization and accumulation of genetic alterations. The association of telomere length (TL) with breast cancer prognosis is examined through a systematic review. Electronic databases (MEDLINE, EMBASE, CENTRAL), from inception to December 2015, and relevant reviews were searched. Studies that evaluated TL (blood and/or tumor) in association with breast cancer survival or prognostic factor were included. Thirty-six studies met inclusion criteria. Overall risk of bias was critical. Eight studies reported survival outcomes. Overall, there was a trend toward an association of longer telomeres with better outcomes (tumor, not blood). Of the 33 studies reporting associations with prognostic factors, nine adjusted for potential confounders. Among the latter, shorter telomeres were associated with older age (blood, not tumor), higher local recurrence rates (normal tissue), higher tumor grade (tumor), and lower physical activity (blood), which were reported in one study each. TL was not associated with molecular subtype (blood, one study), family history (tumor, one study), chemotherapy (blood, three of four studies), and stress reduction interventions (blood, two of two studies). Although major methodologic differences preclude from drawing conclusive results, TL could be a valuable breast cancer prognostic marker. Cancer Epidemiol Biomarkers Prev; 26(1); 3-10. ©2016 AACR.

Association of Telomere Length with Breast Cancer Prognostic Factors.

INTRODUCTION: Telomere length, a marker of cell aging, seems to be affected by the same factors thought to be associated with breast cancer prognosis. OBJECTIVE: To examine associations of peripheral blood cell-measured telomere length with traditional and potential prognostic factors in breast cancer patients. METHODS: We conducted a cross-sectional analysis of data collected before surgery from 162 breast cancer patients recruited consecutively between 01/2011 and 05/2012, at a breast cancer reference center. Data on the main lifestyle factors (smoking, alcohol consumption, physical activity) were collected using standardized questionnaires. Anthropometric factors were measured. Tumor biological characteristics were extracted from pathology reports. Telomere length was measured using a highly reproducible quantitative PCR method in peripheral white blood cells. Spearman partial rank-order correlations and multivariate general linear models were used to evaluate relationships between telomere length and prognostic factors. RESULTS: Telomere length was positively associated with total physical activity (rs = 0.17, P = 0.033; Ptrend = 0.069), occupational physical activity (rs = 0.15, P = 0.054; Ptrend = 0.054) and transportation-related physical activity (rs = 0.19, P = 0.019; P = 0.005). Among post-menopausal women, telomere length remained positively associated with total physical activity (rs = 0.27, P = 0.016; Ptrend = 0.054) and occupational physical activity (rs = 0.26, P = 0.021; Ptrend = 0.056) and was only associated with transportation-related physical activity among pre-menopausal women (rs = 0.27, P = 0.015; P = 0.004). No association was observed between telomere length and recreational or household activities, other lifestyle factors or traditional prognostic factors. CONCLUSIONS: Telomeres are longer in more active breast cancer patients. Since white blood cells are involved in anticancer immune responses, these findings suggest that even regular low-intensity physical activity, such as that related to transportation or occupation, could be recommended to breast cancer patients.

Effect of physical activity on sex hormones in women: a systematic review and meta-analysis of randomized controlled trials.

INTRODUCTION: Exposure to high levels of endogenous estrogens is a main risk factor for breast cancer in women, and in observational studies was found to be inversely associated with physical activity. The objective of the present study is to determine the effect of physical activity interventions on sex hormone levels in healthy women. METHODS: Electronic databases (MEDLINE, EMBASE, CENTRAL), from inception to December 2014, and reference lists of relevant reviews and clinical trials were searched, with no language restrictions applied. Randomized controlled trials (RCTs) were included if they compared any type of exercise intervention to no intervention or other interventions, and assessed the effects on estrogens, androgens or the sex hormone binding globulin (SHBG) in cancer-free women. Following the method described in the Cochrane Handbook for Systematic Reviews of Interventions, data on populations, interventions, and outcomes were extracted, and combined using the inverse-variance method and a random-effects model. A pre-established protocol was drawn up, in which the primary outcome was the difference in circulating estradiol concentrations between the physical activity (experimental) and the control groups after intervention. Pre-specified subgroup analyses and sensitivity analysis according to the risk of bias were conducted. RESULTS: Data suitable for quantitative synthesis were available from 18 RCTs (1994 participants) for total estradiol and from 5 RCTs (1245 participants) for free estradiol. The overall effect of physical activity was a statistically significant decrease of both total estradiol (standardized mean difference [SMD] -0.12; 95 % confidence interval [CI] -0.20 to -0.03; P = 0.01; I (2) = 0 %) and free estradiol (SMD -0.20; 95 % CI -0.31 to -0.09; P = 0.0005; I (2) = 0 %). Subgroup analyses suggest that this effect is independent of menopausal status and is more noticeable for non-obese women and for high intensity exercise. Meta-analysis for secondary outcomes found that physical activity induces a statistically significant decline of free testosterone, androstenedione, dehydroepiandrosterone-sulfate and adiposity markers, while a significant increase of SHBG was observed. CONCLUSIONS: Although the effect is relatively modest, physical activity induces a decrease in circulating sex hormones and this effect is not entirely explained by weight loss. The findings emphasize the benefits of physical activity for women.