Breast Cancer Prediction Model Research Articles

Percent Mammographic Density and Dense Area as Risk Factors for Breast Cancer.

Purpose: Mammographic characteristics are known to be correlated to breast cancer risk. Percent mammographic density (PMD), as assessed by computer-assisted methods, is an established risk factor for breast cancer. Along with this assessment the absolute dense area (DA) of the breast is reported as well. Aim of this study was to assess the predictive value of DA concerning breast cancer risk in addition to other risk factors and in addition to PMD. Methods: We conducted a case control study with hospital-based patients with a diagnosis of invasive breast cancer and healthy women as controls. A total of 561 patients and 376 controls with available mammographic density were included into this study. We describe the differences concerning the common risk factors BMI, parital status, use of hormone replacement therapy (HRT) and menopause between cases and controls and estimate the odds ratios for PMD and DA, adjusted for the mentioned risk factors. Furthermore we compare the prediction models with each other to find out whether the addition of DA improves the model. Results: Mammographic density and DA were highly correlated with each other. Both variables were as well correlated to the commonly known risk factors with an expected direction and strength, however PMD (ρ = -0.56) was stronger correlated to BMI than DA (ρ = -0.11). The group of women within the highest quartil of PMD had an OR of 2.12 (95 % CI: 1.25-3.62). This could not be seen for the fourth quartile concerning DA. However the assessment of breast cancer risk could be improved by including DA in a prediction model in addition to common risk factors and PMD. Conclusions: The inclusion of the parameter DA into a prediction model for breast cancer in addition to established risk factors and PMD could improve the breast cancer risk assessment. As DA is measured together with PMD in the process of computer-assisted assessment of PMD it might be considered to include it as one additional breast cancer risk factor that is obtained from breast imaging.

Incidence of invasive breast cancer in the presence of competing mortality: the Canadian National Breast Screening Study

Mortality due to causes other than breast cancer is a potential competing risk which may alter the incidence probability of breast cancer and as such should be taken into account in predictive modelling. We used data from the Canadian National Breast Screening Study (CNBSS), which consist of two randomized controlled trials designed to evaluate the efficacy of mammography among women aged 40-59. The participants in the CNBSS were followed up for incidence of breast cancer and mortality due to breast cancer and other causes; this allowed us to construct a breast cancer risk prediction model while taking into account mortality for the same study population. In this study, we use 1980-1989 as the study period. We exclude the prevalent cancers from the CNBSS to estimate the probability of developing breast cancer, given the fact that women were cancer-free at the beginning of the follow-up. By the end of 1989, from 89,434 women, 944 (1.1 %) were diagnosed with invasive breast cancer, 922 (1.0 %) died from causes other than breast cancer, and 87,568 (97.9 %) were alive and not diagnosed with invasive breast cancer. We constructed a risk prediction model for invasive breast cancer based on 39 risk factors collected at the time of enrolment or the initial physical examination of the breasts. Age at entry (HR 1.07, 95 % CI 1.05-1.10), lumps ever found in left or right breast (HR 1.92, 95 % CI 1.19-3.10), abnormality in the left breast (HR 1.26, 95 % CI 1.07-1.48), history of other breast disease, family history of breast cancer score (HR 1.01, 95 % CI 1.00-1.01), years menstruating (HR 1.02, 95 % CI 1.01-1.03) and nulliparity (HR 1.70, 95 % CI 1.23-2.36) are the model's predictors. We investigated the effects of time-dependent factors. The model is well calibrated with a moderate discriminatory power (c-index 0.61, 95 % CI 0.59-0.63); we use it to predict the 9-year risk of developing breast cancer for women of different age groups. As an example, we estimated the probability of invasive cancer at 5 years after enrolment to be 0.00448, 0.00556, 0.00691, 0.00863, and 0.01034, respectively, for women aged 40, 45, 50, 55, and 59, all of whom had never noted lumps in their breasts, had 32 years of menstruating, 1-2 live births, no other types of breast disease and no abnormality found in their left breasts. The results of this study can be used by clinicians to identify women at high risk of breast cancer for screening intervention and to recommend a personalized intervention plan. The model can be also utilized by a woman as a breast cancer risk prediction tool.

A breast cancer prediction model incorporating familial and personal risk factors

Risk factors for breast cancer can be allocated to one of four groups: 1 Family History/genetic 2 Reproductive/hormonal 3 Proliferative benign breast disease 4 Mammographic density These four factors have now been thoroughly studied and accurate quantitative estimates for the risk are now available for many of them. The most useful summary comes from the Oxford collaboration, which has now produced a series of papers estimating risk for individual factors. Less is known about the possible interaction between these factors and virtually nothing is known about how different factors influence the risk of different types of breast cancer e.g. oestrogen receptor positive versus negative tumours. Risk factors appear to be largely independent and this facilitates building a model to predict risk for individuals. Previous models have focused on either non-genetic factors [5] where important factors relating to genetic risk are not considered, or strictly familial factors in which the modifying effect of other factors is not included [4]. Mammographic density has not been included in any of these models, although it is currently the one risk factor with the largest population attributable risk [2]. Other authors have looked at combined models however [6]. Here we briefly review the main risk factors for breast cancer and describe our own model [7] and a computer programme for synthesizing the factors into an individual risk profile. The model has novel features in terms of combining family history data using segregation analysis with phenotypic factors using the proportional hazards model. Some comparisons between models will be made [1].

Abstract IA21: Risk prediction models for lung cancer

Abstract IA21: Risk prediction models for lung cancer

P4-10-11: Genetic and Environmental Predictors, Endogenous Hormones and Growth Factors and Risk of Estrogen Receptor-Positive Breast Cancer in Japanese Women.

Abstract The incidence of breast cancer in Japanese women has doubled in all age groups over the past two decades. We have recently demonstrated that this marked increase is mostly due to an increase in the estrogen receptor (ER)-positive subtype. It is necessary to establish risk factors capable of predicting the risk of ER-positive breast cancer which will enable the efficient selection of candidates for preventive chemotherapy. We analyzed genetic factors, including 14 single nucleotide polymorphisms (SNPs), environmental risk factors (body-mass index (BMI), age at menarche, pregnancy, age at first birth, breastfeeding, family history of breast cancer, age at menopause, use of hormone replacement therapy, alcohol intake and smoking), serum hormones and growth factors (estradiol, testosterone, prolactin, insulin-like growth factor 1 (IGF1) and IGF binding protein 3 (IGFBP3)) and mammographic density in 913 women with breast cancer and 278 disease-free controls. To identify important risk factors, risk prediction models for ER-positive breast cancer in both pre- and postmenopausal women were created by logistic regression analysis. In premenopausal women, 1 SNP (CYP19A1-rs10046), age, pregnancy, breastfeeding, alcohol intake, serum levels of prolactin, testosterone and IGFBP3 were considered to be risk predictors. In postmenopausal women, 1 SNP (TP53-rs1042522), age, BMI, age at menopause, serum levels of testosterone and IGF1 were identified as risk predictors. Risk factors may differ between women of different menopausal status, and inclusion of common genetic variants and serum hormone measurements as well as environmental factors might improve risk assessment models. Further validation studies will clarify appropriate risk groups for preventive chemotherapy. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P4-10-11.

A systematic review of breast cancer incidence risk prediction models with meta-analysis of their performance

A risk prediction model is a statistical tool for estimating the probability that a currently healthy individual with specific risk factors will develop a condition in the future such as breast cancer. Reliably accurate prediction models can inform future disease burdens, health policies and individual decisions. Breast cancer prediction models containing modifiable risk factors, such as alcohol consumption, BMI or weight, condom use, exogenous hormone use and physical activity, are of particular interest to women who might be considering how to reduce their risk of breast cancer and clinicians developing health policies to reduce population incidence rates. We performed a systematic review to identify and evaluate the performance of prediction models for breast cancer that contain modifiable factors. A protocol was developed and a sensitive search in databases including MEDLINE and EMBASE was conducted in June 2010. Extensive use was made of reference lists. Included were any articles proposing or validating a breast cancer prediction model in a general female population, with no language restrictions. Duplicate data extraction and quality assessment were conducted. Results were summarised qualitatively, and where possible meta-analysis of model performance statistics was undertaken. The systematic review found 17 breast cancer models, each containing a different but often overlapping set of modifiable and other risk factors, combined with an estimated baseline risk that was also often different. Quality of reporting was generally poor, with characteristics of included participants and fitted model results often missing. Only four models received independent validation in external data, most notably the 'Gail 2' model with 12 validations. None of the models demonstrated consistently outstanding ability to accurately discriminate between those who did and those who did not develop breast cancer. For example, random-effects meta-analyses of the performance of the 'Gail 2' model showed the average C statistic was 0.63 (95% CI 0.59-0.67), and the expected/observed ratio of events varied considerably across studies (95% prediction interval for E/O ratio when the model was applied in practice was 0.75-1.19). There is a need for models with better predictive performance but, given the large amount of work already conducted, further improvement of existing models based on conventional risk factors is perhaps unlikely. Research to identify new risk factors with large additionally predictive ability is therefore needed, alongside clearer reporting and continual validation of new models as they develop.

Genetic and environmental predictors, endogenous hormones and growth factors, and risk of estrogen receptor‐positive breast cancer in Japanese women

The incidence of breast cancer in Japanese women has doubled in all age groups over the past two decades. We have recently shown that this marked increase is mostly due to an increase in the estrogen receptor (ER)-positive subtype. It is necessary to establish risk factors capable of predicting the risk of ER-positive breast cancer that will enable the efficient selection of candidates for preventive therapy. We analyzed genetic factors, including 14 single nucleotide polymorphisms (SNPs), environmental risk factors (body mass index, age at menarche, pregnancy, age at first birth, breastfeeding, family history of breast cancer, age at menopause, use of hormone replacement therapy, alcohol intake, and smoking), serum hormones and growth factors (estradiol, testosterone, prolactin, insulin-like growth factor 1 [IGF1] and IGF binding protein 3 [IGFBP3]), and mammographic density in 913 women with breast cancer and 278 disease-free controls. To identify important risk factors, risk prediction models for ER-positive breast cancer in both pre- and postmenopausal women were created by logistic regression analysis. In premenopausal women, one SNP (CYP19A1-rs10046), age, pregnancy, breastfeeding, alcohol intake, serum levels of prolactin, testosterone, and IGFBP3 were considered to be risk predictors. In postmenopausal women, one SNP (TP53-rs1042522), age, body mass index, age at menopause, serum levels of testosterone, and IGF1 were identified as risk predictors. Risk factors may differ between women of different menopausal status, and inclusion of common genetic variants and serum hormone measurements as well as environmental factors might improve risk assessment models. Further validation studies will clarify appropriate risk groups for preventive therapy.

The Numbers Game: The Risky Business of Projecting Risk

JNCI Vol. 103, Issue 13 | July 6, 2011 In this issue of the Journal, Petracci et al. ( 1 ) present a model to estimate the degree of benefit in terms of breast cancer risk reduction that may be achieved with risk factor modification. By how much would breast cancer risk change over a 10-year period for a 45-yearold Italian woman making changes in modifiable risk factors such as exercise, weight , and/or alcohol use? According to this well-designed and conducted study, all of the following answers are correct: 0.4% , 0.6%, and 20.5%. These figures refer to individual absolute risk reduction, population-level absolute risk reduction, and population fractional risk reduction (akin to the population attributable risk), respectively, and are associated with change in one (individual absolute risk reduction) or all modifiable risk factors (population-level risk reduction estimates). The compatible, yet disparate numbers, for the absolute risk reduction and fractional risk reduction risk estimates is a difficult concept to grasp — even more challenging is when and how to apply the estimates. The results of this study provide a welcome spotlight on the complexity of predicting, interpreting, and communicating risk on both an individual and population level. By sound statistical methodology and data from a welldesigned , large , Italian , multicenter case – control study of invasive breast cancer, Petracci et al. ( 1 ) extend previous work on breast cancer prediction models by incorporating modifi able risk factors and assessing the potential reductions in risk that might result if the modifi able risk factor profi le could be optimized. The calibration and discriminatory ability of the model when evaluated in an independent cohort was similar to other prediction models ( 1 , 2 ). The Numbers Game: The Risky Business of Projecting Risk

Variance computations for functionals of absolute risk estimates

We present a simple influence function based approach for computing the variances of estimates of absolute risk and functions of absolute risk. We apply this approach to criteria that assess the impact of changes in the risk factor distribution on absolute risk for an individual and at the population level. As an illustration we use an absolute risk prediction model for breast cancer that includes modifiable risk factors in addition to standard breast cancer risk factors. Influence function based variance estimates for absolute risk and the criteria are compared to bootstrap variance estimates.

Genetics and Breast Cancer Risk Prediction—Are We There Yet?

Whether information on genetic variants can be used for better prediction of disease risk is of current clinical interest. Previous studies have shown that mutations in the BRCA1 and BRCA2 genes are associated with high risk of breast cancer (1). Theoretical considerations suggest that further work will lead to the discovery of more genes that are involved in breast carcinogenesis (2). However, it requires more than proof of association to assess improvement in risk prediction by including genetic information. In this issue of the Journal, Mealiffe et al. (3) use data from the Women’s Health Initiative Clinical Trial to provide additional support for including genetic information in risk prediction models for breast cancer. This study finds that adding a small number of single-nucleotide polymorphisms to the well-known and established Gail model can improve the prediction of breast cancer. The authors (3) indicate that previous studies that added single-nucleotide polymorphisms to the Gail model for breast cancer risk prediction suffer from methodological limitations or are merely theoretical in nature. What statistical methods are appropriate to use in comparing risk prediction models has been subject to much recent debate. Standard analyses comparing receiver operating characteristic curves have been criticized as being insensitive to meaningful changes in predicted risk and not well suited to prospective data (4,5). Mealiffe et al. (3) use newer methods based on reclassification, or changes in risk strata, which have become relatively widespread in the clinical literature. A reclassification table sorts the predicted probabilities from each risk prediction model into clinically relevant risk categories and tabulates one against the other. The extent and accuracy of any reclassification, or change in risk category, indicate whether a model can generate improvements in risk stratification, and ultimately, decisions regarding therapies. The authors compute the net reclassification improvement (NRI), which compares the proportion of case patients appropriately moving up vs down a risk category with

Abstract LB-432: Validation studies for Korean breast cancer prediction model

Abstract Background: A model for predicting breast cancer risk has been developed in Korea. Associated risk factors included in the model were 1st degree relatives with breast cancer, age at menarche, menopausal status, number of children, age at first fullterm pregnancy, mammography, and exercise for patients less than 50 years of age and 1st degree relatives with breast cancer, age at menarche, age at menopause, number of children, BMI, mammography, and exercise for patients greater than 50 years of age. The area under the curve among over 50 year olds was 0.655 and 0.611 among under 50 year old. Objectives: The objective of the study was validation of an absolute risk prediction model for breast cancer in Korea by using a large, population cohort and patient registry. Method: The Korean Multi-center Cancer Cohort (KMCC) and registry data of breast cancer patients constructed by the Korean Breast Cancer Society were used to validate the model. 12221 women more than 30 years of age at baseline answered questions about demographic characteristics, reproductive factors, lifestyle, family history since 1993 and we identified 11 and 5 incident breast cancer patient more than 50 and less than 50 years of age, respectively. Age, resident area, and included year were matched to select controls; women aged less than 50 were matched 1:20 while women aged greater than 50 were matched 1:10. In the breast cancer registry, clinical, demographic characteristics and reproductive factors of breast cancer patients were collected since 1980. We selected 2727 patients under 50 years old and 2000 patients over 50 years old matched by age and diagnostic year and calculated their 10, 20, 30, and 40 year absolute risk. Result: In the KMCC data, 10 year absolute risk was 0.39 (95% CI 0.36-0.42) among cases and 0.36 (95% CI 0.35-0.37) among controls under 50 years old and 0.16 (95% CI 0.10-0.22) among cases and 0.08 (0.07-0.09) among controls over 50 years old and it was significantly different in the median test (P-value: 0.02 in both groups). Among breast cancer patient group over 50 years old, 10, 20, 30 and 40 year absolute risk were 0.22 (95% CI 0.21-0.23), 0.35 (95% CI 0.34-0.36), 0.43 (95% CI 0.42-0.45), 0.46 (95% CI 0.44-0.48) respectively and 0.56 (95% CI 0.55-0.57), 1.18 (95% CI 1.15-1.21), 1.53 (95% CI 1.50-1.57), 1.74 (95% CI 1.70-1.78). Conclusion: Due to the short follow up time and small number of cancer incidence in the KMCC cohort, although absolute risk was small, a significant difference was confirmed. Future validation study with a large number and enough follow up cohort time is needed. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr LB-432.

136 New recurrence prediction model for breast cancer by data mining

136 New recurrence prediction model for breast cancer by data mining

Putting Risk Prediction in Perspective: Relative Utility Curves

Risk prediction models based on medical history or results of tests are increasingly common in the cancer literature. An important use of these models is to make treatment decisions on the basis of estimated risk. The relative utility curve is a simple method for evaluating risk prediction in a medical decision-making framework. Relative utility curves have three attractive features for the evaluation of risk prediction models. First, they put risk prediction into perspective because relative utility is the fraction of the expected utility of perfect prediction obtained by the risk prediction model at the optimal cut point. Second, they do not require precise specification of harms and benefits because relative utility is plotted against a summary measure of harms and benefits (ie, the risk threshold). Third, they are easy to compute from standard tables of data found in many articles on risk prediction. An important use of relative utility curves is to evaluate the addition of a risk factor to the risk prediction model. To illustrate an application of relative utility curves, an analysis was performed on previously published data involving the addition of breast density to a risk prediction model for invasive breast cancer.

Breast cancer epidemiology according to recognized breast cancer risk factors in the Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial Cohort

BackgroundMultidisciplinary attempts to understand the etiology of breast cancer are expanding to increasingly include new potential markers of disease risk. Those efforts may have maximal scientific and practical influence if new findings are placed in context of the well-understood lifestyle and reproductive risk factors or existing risk prediction models for breast cancer. We therefore evaluated known risk factors for breast cancer in a cancer screening trial that does not have breast cancer as a study endpoint but is large enough to provide numerous analytic opportunities for breast cancer.MethodsWe evaluated risk factors for breast cancer (N = 2085) among 70,575 women who were randomized in the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial. Using Poisson regression, we calculated adjusted relative risks [RRs, with 95% confidence intervals (CIs)] for lifestyle and reproductive factors during an average of 5 years of follow-up from date of randomization.ResultsAs expected, increasing age, nulliparity, positive family history of breast cancer, and use of menopausal hormone therapy were positively associated with breast cancer. Later age at menarche (16 years or older vs. < 12: RR = 0.81, 95% CI, 0.65–1.02) or menopause (55 years or older vs. < 45: RR = 1.29, 95% CI, 1.03–1.62) were less strongly associated with breast cancer than was expected. There were weak positive associations between taller height and heavier weight, and only severe obesity [body mass index (BMI; kg/m2) 35 or more vs. 18.5–24.9: RR = 1.21, 95% CI, 1.02–1.43] was statistically significantly associated with breast cancer.ConclusionThe ongoing PLCO trial offers continued opportunities for new breast cancer investigations, but these analyses suggest that the associations between breast cancer and age at menarche, age at menopause, and obesity might be changing as the underlying demographics of these factors change.Clinical Trials Registrationhttp://www.clinicaltrials.gov, NCT00002540.

0094 Breast cancer prediction models for non sentinel axillary lymph node metastases - assessing validity and correlation to disease recurrence

0094 Breast cancer prediction models for non sentinel axillary lymph node metastases - assessing validity and correlation to disease recurrence

Introducing Molecular Subtyping of Breast Cancer Into the Clinic?

Introducing Molecular Subtyping of Breast Cancer Into the Clinic?

Using Clinical Factors and Mammographic Breast Density to Estimate Breast Cancer Risk: Development and Validation of a New Predictive Model

Current models for assessing breast cancer risk are complex and do not include breast density, a strong risk factor for breast cancer that is routinely reported with mammography. To develop and validate an easy-to-use breast cancer risk prediction model that includes breast density. Empirical model based on Surveillance, Epidemiology, and End Results incidence, and relative hazards from a prospective cohort. Screening mammography sites participating in the Breast Cancer Surveillance Consortium. 1,095,484 women undergoing mammography who had no previous diagnosis of breast cancer. Self-reported age, race or ethnicity, family history of breast cancer, and history of breast biopsy. Community radiologists rated breast density by using 4 Breast Imaging Reporting and Data System categories. During 5.3 years of follow-up, invasive breast cancer was diagnosed in 14,766 women. The breast density model was well calibrated overall (expected-observed ratio, 1.03 [95% CI, 0.99 to 1.06]) and in racial and ethnic subgroups. It had modest discriminatory accuracy (concordance index, 0.66 [CI, 0.65 to 0.67]). Women with low-density mammograms had 5-year risks less than 1.67% unless they had a family history of breast cancer and were older than age 65 years. The model has only modest ability to discriminate between women who will develop breast cancer and those who will not. A breast cancer prediction model that incorporates routinely reported measures of breast density can estimate 5-year risk for invasive breast cancer. Its accuracy needs to be further evaluated in independent populations before it can be recommended for clinical use.

Comparisons of predictive modeling techniques for breast cancer in Korean women

Objective: To develop breast cancer prediction models and to compare their predictive performance by using Bayesian Networks (BN), Nave Bayes (NB), Classification and Regression Trees (CART), and Logistic Regression (LR). Methods: The dataset consisting of 109 breast cancer patients and 100 healthy women was used. Hugin ResearcherTM 6.7 and PoulinHugin 1.5, both of which are NB modeling software, were used. For the LR model and CART, ECMiner was used. Results: The highest area under the receiver operating characteristic curve (AUC) was shown in the Tree augmented NB model as .90. The lowest AUC was CART with .48; that of the LR model was .86. Two BN models with prior knowledge and without prior knowledge did not show any difference at all (.64 vs. .65). The lifts of four models (Simple NB, Tree Augmented NB, Hierarchical NB, LR) were 1.9. The AUCs in both the NB and LR models were higher than that of the previously established models that have been published by using LR methods. Conclusion: NB could be preferred to LR in the development of a predictive model to promote regular screening tests and early detection, which is more or less free from statistical assumptions and limitations. (Journal of Korean Society of Medical Informatics 14-1, 37-44, 2008)

Prospective Breast Cancer Risk Prediction Model for Women Undergoing Screening Mammography

Risk prediction models for breast cancer can be improved by the addition of recently identified risk factors, including breast density and use of hormone therapy. We used prospective risk information to predict a diagnosis of breast cancer in a cohort of 1 million women undergoing screening mammography. There were 2,392,998 eligible screening mammograms from women without previously diagnosed breast cancer who had had a prior mammogram in the preceding 5 years. Within 1 year of the screening mammogram, 11,638 women were diagnosed with breast cancer. Separate logistic regression risk models were constructed for premenopausal and postmenopausal examinations by use of a stringent (P<.0001) criterion for the inclusion of risk factors. Risk models were constructed with 75% of the data and validated with the remaining 25%. Concordance of the predicted with the observed outcomes was assessed by a concordance (c) statistic after logistic regression model fit. All statistical tests were two-sided. Statistically significant risk factors for breast cancer diagnosis among premenopausal women included age, breast density, family history of breast cancer, and a prior breast procedure. For postmenopausal women, the statistically significant factors included age, breast density, race, ethnicity, family history of breast cancer, a prior breast procedure, body mass index, natural menopause, hormone therapy, and a prior false-positive mammogram. The model may identify high-risk women better than the Gail model, although predictive accuracy was only moderate. The c statistics were 0.631 (95% confidence interval [CI] = 0.618 to 0.644) for premenopausal women and 0.624 (95% CI = 0.619 to 0.630) for postmenopausal women. Breast density is a strong additional risk factor for breast cancer, although it is unknown whether reduction in breast density would reduce risk. Our risk model may be able to identify women at high risk for breast cancer for preventive interventions or more intensive surveillance.

Risk Prediction Models For Familial Breast Cancer

A positive family history of breast cancer, reflecting genetic susceptibility, is one of the strongest risk factors for the disease. A number of breast cancer susceptibility genes have been identified to date, with the most important being BRCA1 and BRCA2. Risk prediction models can be used to identify individuals likely to carry BRCA1 and BRCA2 mutations and individuals at high risk of developing the disease. This information can then be used to target genetic testing, screening and interventions more effectively. In this article, the authors review the risk models that have been developed for familial breast cancer and discuss their applicability, strengths and weaknesses, and present examples of classifying women into risk categories according to the predictions by the various models. The review concludes with a discussion of the ways in which risk models could be improved in the immediate- and long-term future.