Susan G Komen  
I've Been Diagnosed With Breast Cancer Someone I Know Was Diagnosed Share Your Story Join Us And Stay Informed Donate To End Breast Cancer
    Home > Research & Grants > Grants Program > Research Grants > Research Grants Awarded > Abstract

    Research Grants Awarded

    Development of biological models to differentiate aggressive from indolent DCIS

    Study Section:
    Detection, Diagnosis and Prognosis

    Scientific Abstract:
    Background: Despite improvements in detection of breast disease, management of ductal carcinoma in situ (DCIS) remains difficult. Treatments, which vary widely from excisional biopsy, to radiation and/or chemopreventive treatment, to mastectomy, are based largely on pathological grade, with high-grade DCIS often treated more aggressively than low-grade. This system of classification is inaccurate as low-grade DCIS treated less aggressively may recur while high-grade DCIS treated aggressively may be indolent in nature. Objective: We have developed genetic models that classify DCIS by pathological grade; additional genotyping data from DCIS samples with outcome information will allow us to develop biological models to identify aggressive DCIS and the underlying genetic lesions contributing to tumor behavior. Specific Aims: 1) to collect DCIS specimens and outcome information from the CBCP (n=40) and pathology archives (n=200), 2) to generate genotype data representing 26 chromosomal regions frequently altered in breast tumors, 3) to perform statistical analysis and biological modeling using the genotyping data to develop prognostic models, and 4) to utilize molecular technologies to identify genes associated with aggressive phenotype. Study Design: Paraffin-embedded tumor specimens and referent samples will be collected and DNA isolated from pure DCIS cell populations after laser microdissection. Genotype data representing 26 chromosomal regions associated with breast tumor etiology will be generated. Preliminary models will be evaluated in the context of outcome data and new models developed to ensure the most accurate models of classification are developed. Chromosomal regions involved in predicting tumor behavior will be subjected to fine-mapping to delimit the critical regions and identify the underlying genetic lesions. Benefits of the Research: By using allelic imbalance to develop models that discriminate aggressive from indolent DCIS, these models can be converted into clinical tests, requiring a single 5-um section from a paraffin-embedded specimen. Identification of underlying genetic alterations within prognostic chromosomal regions will allow development of routine clinical tests, allowing customized treatment regimens to be developed so that each woman is given optimal medical care to prevent recurrence with minimal physical or psychosocial side effects.

    Lay Abstract:
    Improvements in mammography have greatly increased the ability to detect early breast disease. Despite these improvements, treatment of ductal carcinoma in situ (DCIS), a preinvasive disease, is difficult - options range from lumpectomy with or without radiation or chemopreventive therapy to double mastectomy. Characterization of DCIS by a pathologist can place the disease type into three grades associated with different levels of risk for recurrence; however, this grading system cannot accurately predict which DCIS will progress to invasive breast cancer versus which will not. Hence, women with lower grade DCIS may not be treated aggressively but progress to invasive disease, while women with high-grade DCIS may be managed with unnecessary treatments, some of which have serious side effects. We have developed biological models, based on patterns of chromosomal alterations called allelic imbalance (AI) that can accurately separate DCIS into pathological groups. In these models, three chromosomal regions could correctly classify DCIS by grade with 94% accuracy. In this study, we will apply these models to 200 DCIS specimens from women diagnosed at least 10 years ago. Use of archival specimens will allow us to develop models that can be associated with recurrence, progression, and length of disease-free status. Using a panel of 52 microsatellite markers representing 26 chromosomal regions commonly altered in breast cancer, data will be used to determine how well the preliminary, pathological-based models can predict the underlying behavior, aggressive or indolent, of these archived DCIS specimens. New models will also be generated using artificial intelligence-based software to ensure that these new predictive models are as accurate as possible. Once predictive models have been developed, those chromosomal regions that can discriminate aggressive from indolent DCIS will be studied so that the causative genes and the associated genetic alterations can be identified. This work will improve the diagnosis of DCIS so that pathological examination of tumor specimens can be accompanied by molecular assays that will accurately predict tumor behavior, and DCIS effectively managed with a minimum of unwanted side effects. With the knowledge gained from identifying genes that contribute to aggressive tumor behavior, new treatments can be designed so that DCIS is effectively treated without the associated risk of recurrence.