5 à 10% des cancers du sein sont héréditaires mais parmi ceux-ci seulement la moitié est expliquée par une altération constitutionnelle d’un gène de prédisposition connu tels que les gènes BRCA1 et BRCA2. L’importante hétérogénéité génétique qui caractérise les famillesBRCAx rend difficile la réalisation d’études familiales groupées et ne permet pas l’identification de nouveaux gènes de prédisposition au cancer du sein selon les méthodes classiques de liaison génétique ou d’association. Les techniques de séquençage de nouvelle génération (NGS) à l’échelle de l’exome ou du génome entier, autorisent en revanche l’étude de familles individuelles à la recherche de mutations constitutionnelles privées mais le nombre considérable de variants génétiques identifiés impose leur tri sur des critères de pathogénicité ou de récurrence. Un autre critère de tri peut être représenté par l’identification de régions candidates définies en fonction de réarrangements génomiques tumoraux communs à plusieurs tumeurs au sein d’une même famille. Le génotypage tumoral par puces SNP (pour single nucleotide polymorphism) permet en effet la détection d’haplotypes conservés dans des régions récurrentes de LOH (pour loss of heterozygosity) communes à plusieurs tumeurs familiales et donc l’identification de régions candidates suspectes d’abriter des mutations germinales dans des gènes de prédisposition au cancer. La combinaison de ces deux approches, génotypage tumoral puis NGS, a été appliquée à une série de 17 familles avec agrégation de cancers du sein pour lesquelles au moins deux échantillons tumoraux étaient disponibles. Aucun nouveau gène de prédisposition au cancer du sein n’a été identifié mais une mutation délétère constitutionnelle du gène ATM a ainsi été retrouvée, associée à une perte de l’allèle sauvage dans les 2 tumeurs d’une famille BRCAx. L’analyse de 17 tumeurs du sein supplémentaires provenant de 10 familles avec agrégation de cancers du sein et mutation constitutionnelle du gène ATM identifiée chez le cas index, a révélé que l’allèle sauvage d’ATM était fréquemment perdu dans ces tumeurs (>80% contre 20% attendu en situation sporadique ; p<0.001). Ce résultat plaide fortement en faveur de l’implication d’ATM dans la carcinogénèse de ces cancers du sein tel un gène suppresseur de tumeur et suggère que les mutations constitutionnelles d’ATM sont impliquées dans des formes familiales de cancer du sein. / Hereditary breast cancers (BCs) account for 5-10% of all diagnosed BCs, yet only 50% of such tumors arise in the context of a germline mutation in known tumor suppressor genes such as BRCA1 or BRCA2. The vast genetic heterogeneity which characterizes BRCAx families makes grouped studies impossible to perform. Next generation sequencing (NGS) techniques, however, allow individual families to be studied in order to identify private mutations. Single nucleotide polymorphism (SNP) arrays allow the detection of conserved haplotypes within recurrent regions of loss of heterozygosity, common to several familial tumors, therefore identifying genomic loci likely to harbor a germline mutation in cancer predisposition genes. The combination of both exome sequencing and SNP arrays for a series of 17 familial BC did not allow the identification of a novel BC predisposition gene, but revealed a germline ATM mutation associated with a loss of the wild-type allele in a BRCAx family. The analysis of 17 additional breast tumors from ten BC families in which a germline ATM mutation had been identified revealed a high frequency of wild-type allele loss in these tumors (>80% compared to the 20% expected in sporadic BC; p <0.001). This result argues strongly in favor of the involvement of ATM in the carcinogenesis of these tumors as a tumor suppressor gene and suggests that germline ATM mutations are involved in a subset of familial BC.

Secular trends in breast cancer incidence in younger women suggest environmental factors, like exposure to environmental chemicals, may play a role in rising incidence. One of the strongest risk factors for developing breast cancer, next to family history, is high mammographic breast density, which is defined as the proportion of fibroglandular breast tissue relative to fat as seen on a mammogram. Phthalates, a ubiquitous endocrine disrupting chemical, have the potential to interfere with endogenous hormones like estrogen and androgens. There is growing evidence from animal and epidemiologic studies indicating distinct periods of heightened susceptibility to endocrine disrupting chemicals throughout the life course, particularly during critical windows of breast development. Exposure to hormonally active environmental chemicals like phthalates may be a modifiable risk factor for breast cancer, therefore reducing or eliminating exposure could have substantial public health benefits. The overarching goal of this dissertation was to assess the relationship between exposure to phthalates during two critical windows of susceptibility, the prenatal and pregnancy periods, and its effect on breast tissue composition in adolescence and adulthood. First, a comprehensive review of epidemiologic studies summarized the body of evidence for the association between phthalate exposure and intermediate markers known to be in the causal pathway of breast cancer risk (age at breast development, menarche, and breast tissue composition). This systematic review of the literature aimed to identify potential patterns of evidence by outcome and timing of exposure. Evidence from this review suggested that phthalate exposure during the prenatal and childhood periods may play a role in altering menarche. Findings for phthalate exposure and age at breast development were inconclusive. There was a considerable lack of epidemiologic data on phthalate exposure and breast tissue composition throughout the life course. Based on one study, there is a potential association between phthalate exposure during pre-puberty and altered breast tissue density in adolescent girls. No study assessed the relationship between phthalate exposure during the prenatal or pregnancy period and subsequent breast tissue composition. Second, an examination for the association between prenatal phthalate exposure and breast tissue composition measured in adolescence (Chapter 3) and the association between phthalate exposure during pregnancy and breast tissue composition measured during or after the postpartum transient period (Chapter 4) aimed to address this major gap identified from the comprehensive review. The empirical chapters of this dissertation used data from an ongoing longitudinal birth cohort study of mothers and their children conducted by the New York City Columbia Center for Children's Environmental Health and the Breast Cancer and the Environment Research Project (CCCEH-BCERP). The CCCEH-BCERP study cohort has prospective data on nine phthalate metabolite concentrations measured during the third trimester of pregnancy and breast tissue composition measured in a subsample of mother-daughter dyads. Notably, we used novel non-invasive methods (optical breast spectroscopy) in this younger cohort of mothers and daughters to objectively measure specific components of the bulk breast composition before mammography screening age. There was significant evidence of altered breast tissue composition in both mothers and daughters. For daughters (n=127, mean age 15.2 ± 1.9 years), prenatal exposures to select low molecular weight (LMW) and high molecular weight (HMW) phthalate metabolites altered overall breast density in opposing directions, which appears to be driven by significant altered percent breast water. There was a significant association between higher prenatal levels of a LMW phthalate metabolite (monobutyl phthalate) and lower levels of overall breast density (adjusted β = -0.32; 95% CI: -0.51, -0.13) and significant association between sum of di(2-ethylhexyl) phthalate (∑DEHP), a HMW phthalate metabolite, and higher levels of overall breast density in girls (adjusted β = 0.20; 95% CI: 0.05, 0.34). For mothers (n=133, mean age 41 ± 5.3 years at follow-up), there was a significant association between two LMW phthalate metabolites and lower levels of percent breast collagen. Additionally, there was a significant inverse relationship between levels of mono-(3-carboxypropyl), a HMW phthalate metabolite, and percent total hemoglobin of the breast (adjusted β =-0.03; 95% CI: -0.06, 0.00, p=0.05). Overall, this dissertation increased our understanding of the impact that exposure to phthalates during critical windows of susceptibility may have on specific components of the breast. Reducing exposure to both HMW and LMW phthalates may have an impact in reducing breast cancer risk, particularly for girls prenatally exposed, as there was stronger evidence of higher overall breast density and percent water from exposure to select HMW phthalates. Future prospective studies should confirm these results as findings might provide an opportunity for modifying potential breast cancer risk.

Breast cancer is the most common form of cancer among women, with around 9000 new diagnoses in Sweden yearly. Detecting and localizing risk of breast cancer could give the opportunity for individualized examination programs and preventative measures if necessary, and potentially be lifesaving. In this study, two deep learning methods have been designed, trained and evaluated on mammograms from healthy patients whom were later diagnosed with breast cancer, to examine how well deep learning models can localize suspicious areas in mammograms. The first proposed model is a ResNet-18 regression model which predicts the pixel coordinates of the annotated target pixel in the prior mammograms. The regression model produces predictions with an average of 44.25mm between the predictions and targets on the test set, which for average sized breasts correspond to a general area of the breast, and not a specific location. The regression network is hence not able to accurately localize suspicious areas in mammograms. The second model is a U-net segmentation model that segments out a risk area in the mammograms. The segmentation model had a 25% IoU, meaning that there is on average a 25% overlap between the target area and the prediction area. 57% of the predictions of the segmentation network had some overlap with the target mask, and predictions that did not overlap with the target often marked high density areas that are traditionally associated with high risk. Overall, the segmentation model did better than the regression model, but needs further improvement before it can be considered adequate to merge with a risk value model and used in practice. However, it is evident that there is sufficient information present in many of the mammogram images to localize the risk, and the research area holds potential for future improvements. / Bröstcancer är den vanligaste cancerformen bland kvinnor, med cirka 9000 nya diagnoser i Sverige årligen. Att upptäcka och lokalisera risken för bröstcancer kan möjliggöra individualiserade undersökningsprogram och förebyggande åtgärder vid behov och kan vara livräddande. I denna studie har två djupinlärningsmodeller designats, tränats och utvärderats på mammogram från friska patienter som senare diagnostiserades med bröstcancer, för att undersöka hur väl djupinlärningsmodeller kan lokalisera misstänkta områden i mammogram. Den första föreslagna modellen är en ResNet-baserad regressionsmodell som förutsäger pixelkoordinaterna för den utmarkerade målpixeln i de friska mammogrammen. Regressionsmodellen producerar förutsägelser med ett genomsnitt på 44,25 mm mellan förutsägelserna och målpunkterna för testbilderna, vilket för medelstora bröst motsvarar ett allmänt bröstområde och inte en specifik plats i bröstet. Regressionsnätverket kan därför inte med precision lokalisera misstänkta områden i mammogram. Den andra modellen är en U-net segmenteringsmodell som segmenterar ut ett riskområde ur mammogrammen. Segmenteringsmodellen hade ett IoU på 25%, vilket innebär att det i genomsnitt fanns en 25-procentig överlappning mellan målområdet och förutsägelsen. 57% av förutsägelserna från segmenteringsnätverket hade viss överlappning med målområdet, och förutsägelser som inte överlappade med målet markerade ofta områden med hög täthet som traditionellt är förknippade med hög risk. Sammantaget presterade segmenteringsmodellen bättre än regressionsmodellen, men behöver ytterligare förbättring innan den kan anses vara adekvat nog att sammanfogas med en riskvärdesmodell och användas i praktiken. Det är dock uppenbart att det finns tillräcklig information i många av mammogrambilderna för att lokalisera risken, och att forskningsområdet har potential för framtida förbättringar.

Many medical devices have been shifting to personal platforms such as smartphones due to its ubiquitous availability, variety of included sensors, robust communication, and user-friendliness. By utilizing smartphones as a medical sensing device should improve the early detection of abnormalities and the long-term monitoring of health conditions. Tissue abnormalities will be detected by touch sensation due to mechanical property changes within the tissue. However, touch sensation is unquantifiable and subjective. We integrate the smartphone with a tactile sensor to build a portable and personalized tissue assessment device based on changes in mechanical properties. The Smartphone Tactile Imaging Probe (STIP) is developed to quantify the mechanical properties of the tissue. The proposed system has a dual-sensing mode: compression-based sensing (STIP-C) and indentation-based sensing (STIP-I). STIP–C is designed to detect and measure the size and hardness of the inclusion. It assesses mechanical property changes caused by the tumor inside the tissue. STIP–I is designed to measure the pitting parameters and viscoelastic properties of the tissue. This system will assess the viscoelasticity changes caused by fluid retention within the tissue. STIP estimates mechanical and viscoelastic behavior changes in the tissue and provides the risk evaluation of an underlying health problem. Breast cancer risk assessment and edema severity level classification are the main applications of STIP. We estimate the breast cancer risk by incorporating the patient’s personal risk value into the STIP-C data associated with the tumor mechanical properties to improve the risk assessment accuracy. To classify the edema severity level, the STIP-I measures the pitting parameters and viscoelastic properties of the tissue. From these parameters, we build a Viscoelastic Pitting Recovery (VPR) model. The model illustrates the changes in tissue viscoelastic behavior associated with the edema severity level. Using the VPR model, we use the thresholding method to classify the edema cases. We also developed customized phantoms representing the different amounts of fluid retention in the tissue. The experimental result found a relationship between the amounts of pitted depth from STIP-I and the fluid amount of a phantom. In this dissertation, we developed and tested a portable tissue mechanical property estimation system. The interchangeable dual-mode STIP sensing probe and risk assessment methods were developed for the breast tumor malignancy and edema severity applications. / Electrical and Computer Engineering

No description available.

Breast cancer poses a serious public health concern. This case-control study describes the relationship of night-shift working on the development of breast cancer in 57 women diagnosed with breast cancer compared to 49 women with other types of cancer in the Vaal Triangle area (selected through non-probability purposive sampling from CANSA). The study revealed that women who work night-shift developed breast cancer 1.24 times more often than women who do not work nightshift (OR=1.24 [95% CI 0.52 to 2.89]). The odds ratio was further increased in women who worked rotating-shift (OR=1.44 [95% CI 0.58 to 3.59]). Night-shift work exposure was not statistically related to the development of breast cancer. It is recommended that the relationship between night-shift exposure and breast cancer risk be further explored through cross-sectional and cohort studies, and other breast cancer pathways. / Health Studies / M.A. (Public Health)

Breast cancer poses a serious public health concern. This case-control study describes the relationship of night-shift working on the development of breast cancer in 57 women diagnosed with breast cancer compared to 49 women with other types of cancer in the Vaal Triangle area (selected through non-probability purposive sampling from CANSA). The study revealed that women who work night-shift developed breast cancer 1.24 times more often than women who do not work nightshift (OR=1.24 [95% CI 0.52 to 2.89]). The odds ratio was further increased in women who worked rotating-shift (OR=1.44 [95% CI 0.58 to 3.59]). Night-shift work exposure was not statistically related to the development of breast cancer. It is recommended that the relationship between night-shift exposure and breast cancer risk be further explored through cross-sectional and cohort studies, and other breast cancer pathways. / Health Studies / M. A. (Public Health)