Breast Cancer Risk Prediction with Heterogeneous Risk Profiles According to Breast Cancer Tumor Markers

American journal of epidemiology (Impact Factor: 5.23). 05/2013; 178(2). DOI: 10.1093/aje/kws457
Source: PubMed


Relationships between some risk factors and breast cancer incidence are known to vary by tumor subtype. However, breast tumors can be classified according to a number of markers, which may be correlated, making it difficult to identify heterogeneity of risk factors with specific tumor markers when using standard competing-risk survival analysis. In this paper, we propose a constrained competing-risk survival model that allows for assessment of heterogeneity of risk factor associations according to specific tumor markers while controlling for other markers. These methods are applied to Nurses' Health Study data from 1980-2006, during which 3,398 incident invasive breast cancers occurred over 1.4 million person-years of follow-up. Results suggested that when estrogen receptor (ER) and progesterone receptor (PR) status are mutually considered, some risk factors thought to be characteristic of "estrogen-positive tumors," such as high body mass index during postmenopause and increased height, are actually significantly associated with PR-positive tumors but not ER-positive tumors, while other risk factors thought to be characteristic of "estrogen-negative tumors," such as late age at first birth, are actually significantly associated with PR-negative rather than ER-negative breast cancer. This approach provides a strategy for evaluating heterogeneity of risk factor associations by tumor marker levels while controlling for additional tumor markers.

1 Follower
5 Reads
    • "In this paper, we consider the estimation of regression models when the population under study is naturally divided into strata, or sub-groups, defined by a categorical variable Z such as gender, dosage or type of treatment, geographical area, etc, or a combination thereof (Rosner et al., 2013; Gertheiss and Tutz, 2012; Viallon et al., 2015). More formally, our aim is to study the relationship 30 between a response variable y ∈ IR and a set of p ≥ 1 predictors x ∈ IR p over these strata, and in particular to determine whether this relationship varies across these strata, that is whether the categorical covariate Z modifies the effect of x on y (Gertheiss and Tutz, 2012). "
    [Show abstract] [Hide abstract]
    ABSTRACT: We consider regression models to be estimated on $K>1$ pre-defined strata of a sample. Denote by $\beta^*_{k,j}$ the theoretical parameter associated to the $j$-th covariate in the $k$-th stratum. It is common practice to first arbitrarily chose a reference stratum, say $\ell$, and perform inference based on the decomposition $\beta^*_{k,j} = \beta^*_{\ell,j} + \gamma^*_{k,j}$. In particular, $\ell_1$-penalized regression models can be constructed to recover non-zero parameters among the $\beta^*_{\ell,j}$'s and the $ \gamma^*_{k,j}$'s. In this paper, we present a simple though efficient method that bypasses the arbitrary choice of the reference stratum at no cost. Its implementation can be done with available packages under a variety of models and, in the linear regression model, we show it is sparsistent under conditions similar to those ensuring sparsistency for an oracular version of the reference stratum strategy. Our empirical study further shows that our proposal performs at least as well as its competitors under the considered settings. As a final illustration, an analysis of road safety data is provided.
  • Source
    • "It is generally recognized that the risk factors affect the ER, PR and HER2 expression: Post-menopause state, high body mass index and increased height, were associated with positive PR tumors and breastfeeding period equal to or longer than seven months was negatively associated with TN breast cancer [10,11]. The increased frequency of positive ER breast tumors in African American, Hispanic and white women, was associated with changing incidence of obesity and parity [12]. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Background The frequencies of molecular breast cancer subtypes vary among different human populations. The Northeastern region of Brazil has a mixed population of African, Indigenous and European ancestry. This retrospective study investigated breast cancer subtypes and applied therapies in a public hospital of Northeastern Brazil. Methods Data of 633 patients with invasive breast cancer from 2005 to 2011 were obtained from medical records. Status of hormone receptor (HR), HER2 and Ki67 expression index of 269 out of 633 patients were used to define subtypes of Luminal A and B, HER2 and triple negative (TN) breast cancer. Expression index of Ki67 ≥ 14% was applied to distinguish Luminal A from Luminal B subtypes. Results Overall, 185 (68.77%) and 132 (49.07%) patients showed positive hormone receptor (HR+) and positive HER2 (HER2+) tumors. The mean age ranged from 53.33 to 58.25 years for patients with tumors of Luminal B and Luminal A subtypes, respectively (p = 0.0182). In general, 67.39% of patients with TN tumors aged over 50 and 19.57% aged between 31 and 40 years (p = 0.0046). The rate of small tumors (T1: ≤ 2.0 cm) varied from 22.73% to 52.46% for TN and Luminal A subtypes (p = 0.0088). The rate of high graded (G3) tumors was increased for HER2 and TN subtypes (35.29% and 34.28%) compared to Luminal A and Luminal B subtypes (3.92% and 12.62%), respectively (p < 0.0001). The five-year survival rate ranged from 92.86% to 75.00%, for Luminal A, HER2 and TN subtypes, respectively (HR: 0.260 to 1.015; 95% CI: 0.043 to 3.594; p = 0.2589). Patients with HER2 positive (HER2+) breast tumors did not receive immunotherapy and chemotherapy application varied from 54.84% to 86.49% for Luminal A and HER2 subtypes, respectively (p = 0.0131). Conclusions The results of this study revealed a high percentage of HER2+ breast tumors and an increased rate of patients with TN tumors aged over 50 years. This emphasizes the need for establishing immunotherapy as an additional therapeutic option to improve clinical outcomes for patients with HER2+ tumors and to investigate the risk factors of TN breast cancer.
    BMC Women's Health 09/2014; 14(1):110. DOI:10.1186/1472-6874-14-110 · 1.50 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Breast cancer remains a significant scientific, clinical and societal challenge. This gap analysis has reviewed and critically assessed enduring issues and new challenges emerging from recent research, and proposes strategies for translating solutions into practice. More than 100 internationally recognised specialist breast cancer scientists, clinicians and healthcare professionals collaborated to address nine thematic areas: genetics, epigenetics and epidemiology; molecular pathology and cell biology; hormonal influences and endocrine therapy; imaging, detection and screening; current/novel therapies and biomarkers; drug resistance; metastasis, angiogenesis, circulating tumour cells, cancer 'stem' cells; risk and prevention; living with and managing breast cancer and its treatment. The groups developed summary papers through an iterative process which, following further appraisal from experts and patients, were melded into this summary account. The 10 major gaps identified were: (1) understanding the functions and contextual interactions of genetic and epigenetic changes in normal breast development and during malignant transformation; (2) how to implement sustainable lifestyle changes (diet, exercise and weight) and chemopreventive strategies; (3) the need for tailored screening approaches including clinically actionable tests; (4) enhancing knowledge of molecular drivers behind breast cancer subtypes, progression and metastasis; (5) understanding the molecular mechanisms of tumour heterogeneity, dormancy, de novo or acquired resistance and how to target key nodes in these dynamic processes; (6) developing validated markers for chemosensitivity and radiosensitivity; (7) understanding the optimal duration, sequencing and rational combinations of treatment for improved personalised therapy; (8) validating multimodality imaging biomarkers for minimally invasive diagnosis and monitoring of responses in primary and metastatic disease; (9) developing interventions and support to improve the survivorship experience; (10) a continuing need for clinical material for translational research derived from normal breast, blood, primary, relapsed, metastatic and drug-resistant cancers with expert bioinformatics support to maximise its utility. The proposed infrastructural enablers include enhanced resources to support clinically relevant in vitro and in vivo tumour models; improved access to appropriate, fully annotated clinical samples; extended biomarker discovery, validation and standardisation; and facilitated cross-discipline working. With resources to conduct further high-quality targeted research focusing on the gaps identified, increased knowledge translating into improved clinical care should be achievable within five years.
    Breast cancer research: BCR 10/2013; 15(5):R92. DOI:10.1186/bcr3493 · 5.49 Impact Factor
Show more

Similar Publications