deCODE BreastCancer™

The first test measuring genetic risk of the common forms of breast cancer. Complements standard risk modeling tools. More complete assessment of risk and more personalized and effective screening, prevention and treatment.

Better screening & prevention

How can deCODE BreastCancer™ help?

Comprehensive risk assessment

deCODE BreastCancer™ detects genetic risk that is independent of and complementary to the risk factors used in standard clinical risk assessment tools. The results are presented as a numerical value of individual risk relative to the population average, by which the scores of traditional risk modeling tools can be multiplied.

Informing prophylactic therapy

Approximately 14% of women of European descent aged 55 and older who take the deCODE BreastCancer™ test will receive a score placing them at or above a 1.7% absolute risk of developing breast cancer within the following five years. This is the threshold used by the American Society of Clinical Oncology for consideration of the use of tamoxifen (or similar drugs) as preventive therapy. deCODE BreastCancer™ also provides information on whether a woman is more likely to develop estrogen-positive or estrogen-negative breast cancer.

Modification of BRCA-associated risk

The risk conferred by four of the SNPs detected by the deCODE BreastCancer™ test interacts in a multiplicative manner with that conferred by the highly penetrant mutations in the BRCA1 and BRCA2 genes. The results can thereby indicate increased lifetime risk in women positive for the BRCA1 or BRCA2 gene mutations.

How deCODE BreastCancer™ can help to optimize risk assessment, screening and therapy in clinical practice

The deCODE BreastCancer™ test makes a significant contribution to modifying basic or modeled risks since it is, to a great extent, independent of the individual factors currently used to estimate future breast cancer risk. deCODE and others have shown that the markers used are independent of immediate family history, breast density, menarche, age of first pregnancy, and nulliparity ^{(2,4,14,44,45)}. The deCODE BreastCancer™ test results represent an independent, validated set of risk factors that is an important addition to the currently recognized risk factors for breast cancer.

Current recommendations for breast cancer screening and diagnosis are based on clinically validated studies of the effectiveness of screening interventions among women of predefined risk categories. These risk categories have been defined by conventional risk factors and their relative contribution, the strongest risk factors being age and close family history of breast cancer (2-fold increase in risk for one first degree relative with breast cancer ⁽²⁸⁾). The deCODE BreastCancer™ test provides new information on the genetic risk contribution of 16 highly significant DNA variants, risk information that seems to be an addition to the currently used risk assessment tools, but not a replacement. It seems to be particularly important for the group of women who will be moved up or down in risk category by multiplication of their basic or modeled risk by the deCODE BreastCancer™ results.

Identifying patients who would benefit from MRI screening

Large-scale studies on the genetic risk factors included in deCODE BreastCancer™ demonstrate that approximately 6% of women of European descent (0.1% of women of East Asian descent) will receive a score putting them at ≥ 1.65-times the general population risk: that is, 65% above the general population risk. This corresponding to a 20% or greater lifetime risk, the threshold at which use of tamoxifen (or similar drugs) as preventive therapy. deCODE BreastCancer™ also provides information on whether a woman is more likely to develop estrogen-positive or estrogen-negative breast cancer.

Modification of BRCA-associated risk

The risk conferred by four of the SNPs detected by the deCODE BreastCancer™ test interacts in a multiplicative manner with that conferred by the highly penetrant mutations in the BRCA1 and BRCA2 genes ^(12-25). The results can thereby indicate increased lifetime risk in women positive for the BRCA1 or BRCA2 gene mutations. This can be relevant for clinical decision making regarding preventive surgery. The test is also useful in parallel with BRCA1 and BRCA2 testing since it may help to underscore the fact that women who test negative for BRCA1 and BRCA2 mutations may still be at increased risk for the later-onset, common forms of breast cancer.

Current risk assessment tools and screening guidelines

Current screening guidelines aimed at early detection of breast cancer are based upon distinguishing between women who are at normal risk (at or near the average population risk) and those who are at significantly increased risk. The goal is to focus screening and resources on those at higher risk, and thereby catch more cancer early for more effective treatment.

In broad terms, the currently used risk assessment models aim at stratifying the population into three general risk categories. The first is the common or average risk group, which also includes those at below average risk. The second comprises those women who as a result of various risk factors or risk modelling are at higher risk, defined as having either a greater than 1.7% absolute risk of developing breast cancer within the next five years or a 20% or greater lifetime risk. Age and family history, including a history of early-onset disease or of related cancers such as ovarian cancer, are the factors that bring the largest number of women into this second category. Other factors taken into account by ACS and the “National Comprehensive Cancer Network (NCCN):http://www.nccn.org/professionals/physician_gls/f_guidelines.asp algorithms include radiation therapy to the chest; lobular carcinoma in situ (LCIS)/atypical hyperplasia; ductal carcinoma in situ (DCIS); and previous diagnosis with breast cancer. (For further reference, see www.cancer.org). The third category is comprised of those women who are likely to test positive for known, rare genetic mutations that can confer very high risk.

The deCODE BreastCancer™ test can impact the risk assessment of women in all three categories in clinically relevant ways. The genetic risk factors for the common forms of the disease detected by this test appear to be largely independent of the parameters used in the most common risk models, most importantly age and family history. Additionally, three of the deCODE BreastCancer™ markers have been shown to modify the risk of women who are known carriers of disease-associated mutations in the BRCA1 and/or BRCA2 genes^(12,25).

Application of the deCODE BreastCancer™ results

deCODE BreastCancer™ and the common-risk group

Average lifetime risk of developing breast cancer is estimated to be 12.3% in the United States (see SEER Cancer Statistics). About 5% of women of European descent taking the deCODE BreastCancer™ test will receive a result putting them at or above 1.65-times average risk, corresponding to a lifetime risk of 20% or greater. This is the ACS guideline threshold for high risk. deCODE BreastCancer™ places a further 10% of women of European descent and 3% of women of East Asian descent at between 1.25- and 1.64-times average risk. This corresponds to a lifetime risk of 15-20%, defined by ACS guidelines as moderate risk. For women within this range ACS recommends consideration and discussion of MRIs in addition to annual mammograms.

It is important to remember that as deCODE BreastCancer™ results represent only the genetic risk for the common forms of breast cancer, they complement and should be used in tandem with risk information on other factors commonly measured in clinical practice. The results of the deCODE BreastCancer™ test can be multiplied by the risk scores from these other factors to yield a comprehensive risk assessment.

deCODE BreastCancer™ and modelled risk

Family history is at the heart of and, in general, of the greatest significance in the most commonly used models for calculating risk of breast cancer. With good reason: having just one first-degree relative with breast cancer doubles a woman’s lifetime risk. But family history, which is designed to capture risk of rarer, high-risk genetic mutations often associated with early-onset disease, by definition does not capture the genetic risk for the common forms of breast cancer that comprise the vast majority of cases. So, family history can be defined narrowly or broadly: to include only very close relatives with diagnosed breast cancer; to include histories of second- and third-degree relatives, if known; or to include diagnosis of related cancers, such as ovarian, as well as prostate. But none of the available risk modeling tools incorporate in their calculations the genetic risk factors for the common forms of the disease measured by deCODE BreastCancer™. This is part of the reason that most of these models tend to underestimate risk, and explains how deCODE BreastCancer™ can provide complementary information missing from these assessments.

Large-scale, peer-reviewed studies demonstrate that the genetic risk for the common forms of breast cancer measured by deCODE BreastCancer™ can be considered to be essentially independent of risk measured by the Gail model, the most commonly used tool for clinical risk assessment. This is the finding of deCODE’s own published studies on the markers included in the test, and is supported by the work of independent researchers. A recent study by Wacholder et al (2010) ⁽³²⁾, for example, found that not only does the genetic risk seem to be independent of the risk derived by Gail modeling but also to be of similar effect and size. Mealiffe et al (2010) ⁽³³⁾ found risk scores derived by measuring the SNPs in the test were consistently related to breast cancer odds across Gail risk strata, very largely independent of the Gail risk with the combined predictor being more predictive than either Gail risk or SNP risk alone.

The Gail model is known to underestimate risk for many women because in terms of family history it takes into account only the number of first-degree relatives, thereby excluding the risk of the 50% of families with cancer in the paternal lineage. Nor does it take into account the age of onset ^(34,35). But again, these shortcomings are relevant only to the family history component of risk, which is independent of the risk factors measured by deCODE BreastCancer™. The results of deCODE BreastCancer™ can therefore be used to directly complement Gail scores, and the deCODEhealth system provides a tool that enables doctors to do this on their desktops.

deCODE BreastCancer™ can help to complement the risk captured by other models as well, but the test results should not be indiscriminately multiplied together with the risk scores from other models without considering the components of these models. The most commonly used other models are the Claus model ⁽³⁶⁾, the BRACAPRO model ⁽³⁷⁾ and the Cuzick–Tyrer model ⁽³⁸⁾.

The Claus model is exclusively focused on familial risk and does not include any nonhereditary risk factors.

The BRACAPRO model provides estimates for the likelihood of finding either a BRCA1 mutation or a BRCA2 mutation based on family history of breast cancer and ovarian cancer, but allows for neither non-hereditary factors nor other “genetic” elements to be included, thereby underestimating the risk in breast-cancer-only families.

The Cuzick–Tyrer model integrates family history, surrogate measures of endogenous estrogen exposure, and benign breast disease in a comprehensive fashion, in addition to producing readout of BRCA1/2. It is particularly important to note that the Cuzick–Tyrer model does take into account the possible presence and effect of multiple genes of differing penetrance, an attempt to implicitly predict the impact of genetic factors of the sort detected by deCODE BreastCancer™ but none of which had been discovered at the time the model was developed. Therefore, the modeled risk derived from this tool should not be multiplied by the deCODE BreastCancer™ results since the effect of the risk variants may already be accounted for to an unknown, albeit imperfect degree. This also applies to the emerging BOADICEA model ⁽³⁹⁾, though this model has the potential to be updated to accommodate the deCODE BreastCancer™ risk alleles.

Because of the limitations of the available computerized risk models, many physicians find manual risk estimation to be one of the best ways to assess risk by considering the strongest risk factor ⁽⁴⁰⁾. In many clinics, this is family history. Manual breast cancer risk assessment is largely based on the published Claus risk tables ⁽⁴¹⁾ and use of data in clear-cut BRCA1/2 families from penetrance data for breast cancer ⁽⁴²⁾. If first-line risk can be assessed on this basis, then adjustments can be made for other factors ^(40,43), including the results of the deCODE BreastCancer™ test. A good manual assessment will alert the assessor to any spurious readout from a computer model.

deCODE BreastCancer™ and hereditary breast cancer

All of the above mentioned models seem to accurately predict risk in women with multiple relatives affected by breast cancer (that is, 2 first-degree relatives and 1 first-degree relative plus 2 other relatives). The deCODE BreastCancer™ test does not assess risk from rare, high-penetrance mutations in genes such as BRCA1, BRCA2, TP53, and PTEN, which confer high risk of familial, early-onset breast cancer. However, some of the variants in the deCODE BreastCancer™ test modify the risk of breast cancer in subjects who carry mutations in the BRCA1 and/or BRCA2 genes. The deCODE BreastCancer™ test reports a factor based on the relevant SNP genotypes by which the lifetime risk of women known to be BRCA mutation carriers should be multiplied in order to determine overall lifetime risk.

This content was last reviewed on April 07, 2011.