Breast Cancer Tissue Markers, Genomic Profiling, and Other Prognostic Factors: A Primer for Radiologists

Age

Generally, the course of breast cancer is considered to be unfavorable in very young and very old patients. Very young patients with breast cancer (younger than 35 years) usually present at a later stage with estrogen receptor (ER)–negative status, and aggressive tumor biology, which leads to shorter time to local-regional recurrence, a shorter distant disease–free interval, and poorer overall survival (Fig 1) (7,8). Patients older than 65 years have lower relative survival rates because their disease is diagnosed at a later stage, and they are more likely to have comorbidities and treatment discrepancies (9). Women aged 45–49 years have the best prognosis (9). The effect of age also depends on breast cancer subtypes. Age may be of greater prognostic importance in luminal cancers than in other types of breast cancer. Younger patients with ER-positive tumors have poorer disease-free survival than do patients with ER-negative tumors, but in older patients, the disease-free survival rate is similar regardless of ER status (10). In patients with human growth hormone receptor 2 (HER2)–positive breast cancer, age is not associated with early relapse or survival, with or without treatment with trastuzumab (11).

Figure 1a.

Download as PowerPoint

Figure 1b.

Download as PowerPoint

Figure 1c.

Download as PowerPoint

Figure 1d.

Download as PowerPoint

Figure 1e.

Download as PowerPoint

Figure 1f.

Download as PowerPoint

Race and Ethnicity

Racial disparities in breast cancer outcomes have multiple causes (12). African American women of lower socioeconomic status present with fewer screening-detected cancers and have delays between diagnosis and initiation of treatment (12). Despite the lower incidence of breast cancer in African American women compared with that in white women, the mortality rate for African American women is higher (29.2 vs 20.6 deaths per 100 000 women, respectively) (1,12). This is related in part to the predisposition to biologically aggressive triple-negative tumors with basal-like subtypes, greater intratumoral genetic heterogeneity, and low pathologic complete response for triple-negative and HER2-positive tumors after African American women receive neoadjuvant chemotherapy (12,13).

Lifestyle Factors

Smoking increases the risk for development of breast cancer, with risk dictated by the number of pack-years of smoking history and age of initiation (14). Patients who smoke before and after a diagnosis of breast cancer also have higher mortality rates. Smoking cessation has been shown to improve breast cancer outcomes (14), and patients should be advised to quit smoking.

Obesity is another important consideration. Excess weight (body mass index >30 kg/m²) has a negative prognostic effect in patients with breast cancer (15). Obesity is associated with older age at diagnosis, advanced stage, and triple-negative status (15–17). However, unlike smoking cessation, weight loss does not lead to increased survival (15–17). Therefore, additional research is necessary (16).

Screening-detected Breast Cancer

Screening mammography has been shown in randomized controlled trials (18,19) to reduce breast cancer mortality by 30%–50%. This benefit is largely attributed to earlier-stage detection of breast cancers compared with those detected at clinical examinations (20–22). The prognosis is particularly excellent if the diagnosis is ductal carcinoma in situ (DCIS). Up to 90% of clinically occult DCIS manifests as microcalcifications seen on mammograms (23). The estimated 20-year mortality rate for DCIS is 3.3%, while the estimated 10-year mortality rate is 23.2% for clinically detected breast cancer and 20.9% for interval breast cancer, or cancer detected between routine screening mammographic examinations (24,25).

Breast Density

Breast density is a known independent risk factor for breast cancer, with a four- to five-fold increased relative risk for breast cancer in women with dense breasts (26). The overlapping dense parenchyma may obscure small masses or lower the sensitivity of mammography, leading to a delay in diagnosis (27). Breast cancers arising in women with dense breasts tend to be larger, have a higher histologic tumor grade, and more often manifest with positive lymph node status and lymphovascular invasion (28–30). However, it is ambiguous if having dense breasts is an independent mortality risk factor once the diagnosis of breast cancer is made. Results of the Swedish mammographic screening trial (31) showed that women with dense breasts have a 1.91 relative risk for breast cancer mortality after controlling for other factors such as age, body mass index, tumor size, histologic tumor grade, and nodal status. However, additional studies (26,32) with a comparable number of patients showed that high breast density does not lead to increased mortality after controlling for confounding factors.

Tumor Size

Tumor size was one of the earliest recognized independent prognostic factors. The size of the tumor can be determined with a combination of radiologic, gross pathologic, and microscopic findings. The largest measurement of an invasive component without an adjacent in-situ component is used to determine T staging. If there is multifocal disease (additional cancer seen in the same quadrant) or multicentric disease (additional cancer seen in different quadrants), only the largest diameter of the largest mass is used to determine T staging. Carter et al (33) showed that larger tumors have a higher likelihood of nodal involvement and worse outcomes, with 91%, 80%, and 63% 5-year overall survival for T1, T2, and T3 tumors, respectively. Given similar lymph node status, a 1-mm increase in tumor size resulted in a 1% decrease in the 15-year survival rate (34). T4 staging is reserved for tumors of any size with direct extension to the chest wall (T4a), to the skin (T4b), or both (T4c).

Multifocal and Multicentric Disease

Studies (35–37) show that patients with multifocal or multicentric disease have worse disease-free survival rates and higher rates of locally recurrent and metastatic disease. Lynch et al (36) suggested that this may be secondary to larger tumor size, higher histopathologic grade, lymphovascular invasion, and lymph node involvement in patients with multifocal and/or multicentric disease compared with patients with unifocal disease. However, whether a multifocal and/or multicentric tumor is an independent prognostic factor is controversial and currently not considered a part of the TNM staging system (Fig 2).

Figure 2a.

Download as PowerPoint

Figure 2b.

Download as PowerPoint

Extensive Intraductal Component

Invasive carcinoma is considered to have an extensive intraductal component if intraductal foci or DCIS constitute more than 25% of the adjacent breast tissue separate from the main tumor mass (38) (Fig 3). Although the final diagnosis of an extensive intraductal component is determined on the basis of pathologic examination, imaging findings such as a mass with associated segmental microcalcifications on a mammogram (Fig 3a, 3b) or clumped nonmass enhancement on breast MR images (Fig 3d) can help in surgical planning to ensure negative margins (39). Achieving negative resection margins decreases the risk of local recurrence in the surgical bed, while incomplete resection of an extensive intraductal component results in higher risk of local recurrence but does not increase the risk for a new primary cancer in the ipsilateral or contralateral breast or of distant metastatic disease (38,40).

Figure 3a.

Download as PowerPoint

Figure 3b.

Download as PowerPoint

Figure 3c.

Download as PowerPoint

Figure 3d.

Download as PowerPoint

Figure 3e.

Download as PowerPoint

Figure 3f.

Download as PowerPoint

Skin and Chest Wall Involvement

Local extension to the skin or the chest wall is classified as T4 or at least stage IIIB regardless of tumor size (6). For T4a, the tumor must invade beyond the pectoralis muscles into the chest wall (defined as the ribs, intercostal muscles, and serratus anterior muscle) (6). For T4b, skin involvement must extend beyond the dermis but not meet criteria for inflammatory breast cancer (Figs 2, 4) (6). However, direct skin invasion may not imply a worse outcome. Authors of several studies (41–44) demonstrated that skin invasion has no substantial effect on prognosis when the researchers controlled for tumor size and nodal status. Silverman et al (42) showed that the adjusted 5-year disease-specific survival for stage I-II (based on tumor size and lymph node status only) are similar with or without skin invasion (42).

Figure 4a.

Download as PowerPoint

Figure 4b.

Download as PowerPoint

Inflammatory Breast Cancer

For diagnosis of inflammatory breast cancer (stage T4d), in addition to pathologic diagnosis of invasive breast carcinoma, the patient must present with rapid onset or less than 6 months of skin erythema or peau d’orange that occupies at least one-third of the breast (Fig 5) (6). Although survival has improved with neoadjuvant chemotherapy, surgery, and radiation therapy, the prognosis remains poor, with a 5-year overall survival rate of 30%–70% (45–49). Triple-negative receptor status (ER, PR, and HER2), lymph node and/or chest wall involvement, age older than 50 years, and lack of response to neoadjuvant chemotherapy are poor prognostic factors for inflammatory breast cancer (48–51). Results of some studies (52) suggest that an accelerated or high radiation dose may improve local-regional control in patients with close or positive margins or suboptimal response to neoadjuvant chemotherapy treatment.

Figure 5a.

Download as PowerPoint

Figure 5b.

Download as PowerPoint

Figure 5c.

Download as PowerPoint

Figure 5d.

Download as PowerPoint

Nodal Involvement

The N stage is determined according to the size of nodal metastatic deposits, and the number and location of involved lymph nodes (6).

The ipsilateral axillary lymph node is the most common site of involvement, and in isolation, axillary lymph node status is the most important prognostic factor (53,54).

Figure 6a.

Download as PowerPoint

Figure 6b.

Download as PowerPoint

Figure 7a.

Download as PowerPoint

Figure 7b.

Download as PowerPoint

Peritumoral Lymphovascular Invasion

Peritumoral lymphovascular invasion currently is not considered part of the TNM staging system but is a negative prognostic factor (53). In particular, lymphovascular invasion is associated with larger tumor size, higher histologic grade, lymph node involvement, and ER-negative status (Fig 1e). Although lymphovascular invasion is confirmed at pathologic examination, imaging findings such as increased peritumoral edema, increased intratumoral T2 signal intensity, increased ipsilateral whole-breast vascularity, and a higher peritumor to tumor apparent diffusion coefficient ratio can suggest lymphovascular invasion (61,62).

Tumoral Necrosis

Tumoral necrosis is a recognized adverse pathologic feature and can be seen as cystic or nonenhancing areas on images (Fig 8). It is often associated with other poor prognostic findings in patients with invasive breast cancer such as younger patient age, larger tumor size, and lymphovascular invasion (63). DCIS with central necrosis or comedonecrosis is associated with earlier disease recurrence (Fig 3f) (64). Furthermore, a subset of breast cancers with central necrosis and fibrosis are related to the basal-like molecular subtype and have a higher rate of lung and brain metastasis and lower rate of 5-year disease-free survival (70% vs 84% of patients with and without necrotic and fibrotic tissue, respectively) (65). One exception is medullary carcinoma, which has a superior prognosis compared with other triple-negative subtypes (66).

Figure 8a.

Download as PowerPoint

Figure 8b.

Download as PowerPoint

Additional Imaging Workup

According to the American College of Radiology (67) appropriateness criteria, additional imaging workup such as fluorodeoxyglucose PET/CT, technetium 99m bone scanning, and CT of the chest, abdomen, and pelvis usually are not appropriate for evaluation of early breast cancer. The National Comprehensive Cancer Network (NCCN) guideline recommends staging studies in patients diagnosed with stage IIIA and higher cancers, greater than four positive axillary nodes, and recurrent or inflammatory breast cancer (68). However, breast MRI may be appropriate in selected patients with early breast cancer such as those who have mammographic or US findings suggestive of multifocal and/or multicentric disease and those suspected of having chest wall involvement or a high-risk subtype of cancer (67,69).

Histologic Tumor Grade

Histologic tumor grade is an important prognostic factor, with higher tumor grade generally portending a worse prognosis (6,70). Pathologists assess the grade of breast cancers by using the Nottingham Histologic Score System (the Elston-Ellis [71] modification of the Scarff-Bloom-Richardson grading system), which allows objective scoring of tumors on the basis of three characteristics: tubule formation (or glandular differentiation), nuclear features, and mitotic activity (Fig 9). Tumors are scored on a three-point scale in each of the three categories, and the scores are added for a total of up to nine points. Tumors with a score of 1 for tubule formation have glands in greater than 75% of the tissue and exhibit better differentiation, those with a score of 2 have glands in 10%–75% of the tissue, and those with a score of 3 have glands in less than 10% of the tissue (6,71). Scoring nuclear pleomorphism is arguably the most subjective aspect of the Nottingham system, incorporating features such as nuclear size and prominence of nucleoli. Tumors that are given a score of 1 generally have small regular nuclei with rare nucleoli. Score 2 tumors have larger and mildly pleomorphic nuclei with inconspicuous nucleoli, and score 3 tumors have markedly enlarged and often bizarre nuclear features with prominent nucleoli (6,53). Mitotic activity is scored on the basis of a count of mitotic figures in 10 high-powered fields; the exact scoring is dependent on microscopic field diameter but follows the principle that more mitoses are found in higher-grade tumors. Total scores of 3–5 indicate a low-grade tumor (grade 1), while a score of 8–9 corresponds to a high-grade tumor (grade 3) with aggressive features likely (53). The eighth edition of the AJCC TNM staging manual incorporates histologic grades into a clinical prognostic stage and pathologic prognostic stage.

Figure 9a.

Download as PowerPoint

Figure 9b.

Download as PowerPoint

Figure 9c.

Download as PowerPoint

Figure 9d.

Download as PowerPoint

Figure 9e.

Download as PowerPoint

Figure 9f.

Download as PowerPoint

Figure 9g.

Download as PowerPoint

Figure 9h.

Download as PowerPoint

Figure 9i.

Download as PowerPoint

Histologic Subtype

To some extent, histologic type also has a role in determining breast cancer prognosis. Because of their tendency to occur multifocally with a single-cell infiltrative pattern, invasive lobular carcinomas tend to manifest at a higher overall stage with frequent lymph node metastases (53). Although there are conflicting reports of the short-term prognostic differences between invasive ductal carcinomas and invasive lobular carcinomas, invasive lobular carcinomas appear to show higher rates of relapse and bone metastasis in the long term (66). This may be, in part, because invasive lobular carcinomas tend to be ER positive, a type that has a lower recurrence rate in the initial year compared with that of the ER-negative type, but the trend is reversed in later years (66). Various subtypes of invasive ductal carcinoma also have prognostic implications. Usually well-differentiated breast cancers such as tubular and mucinous carcinomas have a favorable prognosis, while less-differentiated breast cancers such as metaplastic and micropapillary carcinomas tend to have worse outcomes.

Estrogen and Progesterone Receptors

ER is an estrogen-activated nuclear transcription factor that regulates the development and proliferation of breast tissue, both benign and neoplastic (72). Similarly, PR is also a nuclear transcription factor and is a component of the ER pathway (53). The majority of tumors that express ER also express PR, and most ER-negative tumors also lack PR expression (53). There is debate as to whether ER-positive PR-negative tumors are a true entity, although there is some evidence that these tumors may represent a more aggressive subtype of ER-positive tumors (53).

Generally, ER and PR status are considered together as hormone receptor status.

Approximately 75% of invasive breast carcinomas are hormone receptor positive (53). These tumors tend to be lower grade and are associated with more indolent behavior, although hormone receptor status is not known to correlate with tumor stage (53,72). PR and particularly ER status are known to be strong positive predictive factors since the advent of targeted hormone therapy. Adjuvant endocrine therapy is associated with a favorable treatment response in up to one-third of patients with ER-positive tumors and also has been shown to decrease the rate of recurrence by up to 50% in these patients (49,53,73).

Although molecular assays for ER and PR are currently available, hormone receptor status is most commonly determined with immunohistochemical staining on formalin-fixed paraffin-embedded tissue in the pathology laboratory. They are generally scored as a percentage or range of percentages of nuclear positivity, with greater than or equal to 1% as the cutoff value for positivity (74). The average intensity of nuclear ER and PR staining is generally reported as weak, moderate, or strong (72).

Human Epidermal Growth Factor Receptor 2

The ERB-B2 gene encodes HER2, which is a growth factor receptor found in both normal and neoplastic breast cells (53). The gene is amplified with resulting HER2 protein overexpression in approximately 20% of invasive breast cancers, leading to increased tumor growth (72). Breast cancers that overexpress HER2 are generally more aggressive, being associated with high rates of metastasis and decreased overall survival in the absence of treatment (72). It is important to assess invasive breast carcinoma for HER2 status, because treatment with targeted therapies (ie, trastuzumab) has been shown to improve overall and disease-free survival in some patients with HER2-positive tumors (75). HER2 overexpression also has been associated with resistance to hormonal modulation (72).

HER2 testing in the laboratory typically begins with immunohistochemical staining of formalin-fixed paraffin-embedded tissue. It is given a score ranging from 0 to 3+ on the basis of the amount and intensity of membranous staining (Fig 11). A score of 3+ is considered HER2 positive at immunohistochemical analysis, while tumors with scores of 1+ and 0 are considered to be HER2 negative. A score of 2+ is considered to be equivocal and warrants further testing with in situ hybridization (76).

Figure 11.

Download as PowerPoint

Although immunohistochemical staining targets the HER2 protein on the cell membrane, in situ hybridization is a technique whereby a probe is used to locate, in the case of HER2, a target DNA sequence corresponding to the ERB-B2 gene. It is useful when immunohistochemical results are equivocal, because in situ hybridization is reported as either a copy number of the ERB-B2 gene present in each cancer cell or a ratio of the ERB-B2 gene copies to CEP17 (HER2/CEP17 ratio) to determine if there truly is amplification of HER2 (Fig 12) (76). Fluorescence in situ hybridization is the criterion standard for in situ hybridization, but it requires special fluorescence microscopes and software for interpretation (77). Techniques such as chromogenic in situ hybridization and silver-enhanced in situ hybridization have emerged; they rely on standard light microscopy for interpretation, combining the techniques of immunohistochemical staining and in situ hybridization (77). These nonfluorescence in situ hybridization methods correlate well with immunohistochemical and fluorescence in situ hybridization results and are useful when fluorescence in situ hybridization is not available (77).

Figure 12a.

Download as PowerPoint

Figure 12b.

Download as PowerPoint

Ki-67

Ki-67 is a nuclear protein involved in cell proliferation. Because uncontrolled proliferation is a hallmark of malignancy, Ki-67 expression commonly is tested in patients with breast cancer. Its prognostic and predictive importance is somewhat uncertain, although a higher proliferative rate generally is considered to be unfavorable (78,79). For stratification of breast cancers according to their intrinsic molecular subtypes, Ki-67 immunohistochemical staining can be used to differentiate luminal A from luminal B tumors, because luminal B tumors have a higher proliferation rate (Fig 13) (78). The 2015 St Gallen Consensus (80,81) proposed that a laboratory’s median Ki-67 value serve as the cutoff value for high versus low proliferation in breast cancers; this value usually coincides with the previously suggested cutoff value of 14%. The methods of Ki-67 evaluation in the pathology laboratory are not as standardized as they are for ER, PR, and HER2. Immunohistochemical staining is performed in formalin-fixed paraffin-embedded tissue; the most commonly used antibody is the MIB-1 clone (82). It is reported only as a percentage of positive nuclei, with care to evaluate nuclear positivity only in tumor cells. This analysis can be done visually or with a digital analysis algorithm as for ER and PR.

Figure 13.

Download as PowerPoint

Molecular Subtype

There are currently four defined breast cancer subtypes based on hormone receptor status and gene expression patterns: luminal A, luminal B, HER2-enriched, and triple-negative (Fig 13) (83).

Luminal Subtype

The luminal epithelial cells that line the milk ducts express low–molecular weight cytokeratins (such as CK8 and CK18) and genes associated with activation of estrogen receptors; therefore, the luminal subtype overlaps with ER-positive cancers and makes up the most common subtype (approximately 75%–80%) (78,83). These are further divided into luminal A and luminal B tumors. Luminal A tumors are the most common among the luminal types and have the best prognosis (84). These are characterized by high expression of ER-related genes and low expression of HER2 and proliferation-related genes (eg, Ki-67) (83,84). Luminal B tumors are less common and have a slightly worse prognosis than do luminal A tumors. These are characterized by lower expression of ER-related genes, variable expression of HER2 gene clusters, and higher expression of proliferation-related genes (83,84). Luminal B tumors are further stratified into two groups on the basis of HER2 expression, highlighting the importance of using the combination of Ki-67 expression with hormone receptor status and HER2 expression in clinical subtyping (85,86). When assessed with Oncotype DX (Genomic Health, Redwood City, Calif) or MammaPrint (Agendia, Amsterdam, the Netherlands), luminal B tumors have higher recurrence scores and higher clinical risk.

HER2-enriched Tumors

The HER2-enriched tumor subtype accounts for approximately 5%–10% of breast cancers (85,87). It is characterized by high expression of HER2 and low expression of ER and PR (83). This is not synonymous with clinically HER2-positive breast cancers, which may be tumors of a luminal type (which have higher expression of ER when compared with a HER2-enriched tumor). This subtype has a worse prognosis than the luminal types, although treatment directed against HER2 in addition to the standard chemotherapy regimen has shown substantial additional survival benefits in this group (88).

Triple-Negative Tumors

ER-negative, PR-negative, and HER2-negative tumors, collectively called triple-negative tumors, account for approximately 15%–20% of breast cancers. They have the worst prognosis among all subtypes, with higher proliferation rates, and are predominantly high-grade tumors (89,90). They are divided into two subtypes, the basal and nonbasal tumors.

The basal subtype, as its name implies, has a gene expression profile similar to that of the basal epithelial cells of normal breast tissue and expresses several basal markers, the most widely accepted of which are high-molecular-weight cytokeratins (such as CK5 and CK6) and epidermal growth factor receptor (85). The nonbasal subtype represents 7%–14% of breast cancers (85,87) and includes interferon-rich, claudin-low, and metaplastic subtypes.

The interferon-rich subtype is characterized by overexpression of immune response genes, offers a comparable prognosis to that of luminal B tumors, and has the best prognosis among triple-negative breast cancers (85). Claudin-low tumors and metaplastic breast cancer tumors are characterized by epithelial to mesenchymal transition cancer stem cell–like features with higher frequency of metaplastic and medullary differentiation (85,89). Claudin-low tumors are distinguished from metaplastic breast cancer tumors by their low gene expression of tight junction claudin proteins. Metaplastic breast cancer accounts for 1% of breast cancers and carries mutations such as PIK3CA, AKT, or KRAS, which also differentiate it from claudin-low tumors (85). Metaplastic breast cancer tumors have a poor prognosis similar to that of claudin-low tumors and have higher local recurrence and metastatic rates (91).

In a study by Grimm et al (69), use of preoperative MRI allowed detection of multifocal and/or multicentric disease and nodal involvement more times in the higher-risk subtypes including luminal B and HER2-enriched cancers. An ongoing clinical trial (Alliance A011104/ACRIN 6694) (92) has the aim of comparing the role of MRI in rates of recurrence in women with triple- negative or HER2-enriched breast cancer.

Gene Expression Profiles

With the advent of the human genome project, a multitude of genes involved in cell proliferation, cell differentiation, and cell death, particularly in breast tumors, were identified and categorized. These breast cancer genes were measured and profiled to determine the correlation between gene expression and the risk of cancer recurrence. In the continued search for personalized therapy, several multigene genomic assays were developed and validated in clinical trials to guide decision making regarding appropriate treatment with adjuvant chemotherapy and/or endocrine therapy, or in the case of DCIS, treatment with radiation therapy in patients with early stage breast cancer.

Adjuvant systemic therapy is responsible for much of the reduction in mortality from breast cancer (93). However, not every early-stage breast cancer may benefit from adjuvant chemotherapy. This is where genomic assays become practical tools in decision making. Two well-known trials are the Trial Assigning Individualized Options for Treatment (Rx) (TAILORx), a trial to validate the use of Oncotype DX that was performed in the United States, and the ongoing Microarray in Node-Negative Disease May Avoid Chemotherapy (MINDACT) trial, which is a prospective trial in Europe to validate MammaPrint (94,95).

Recently, the American Society of Clinical Oncology (ASCO) updated their clinical practice guidelines for women with node-negative ER-positive early-stage invasive breast cancer to include the use of biomarker tests including Oncotype DX, MammaPrint, Endopredict (Myriad Genetics, Zurich, Switzerland), and Prosigna PAM50 risk of recurrence score (nanoString, Seattle, Wash) to help predict whether patients will benefit from adjuvant chemotherapy (95).

Table 1: Multigene Prognostic Genomic Assays to Evaluate Risk of Recurrence Treatment in Early Breast Cancers

*American Society of Clinical Oncology (ASCO) guidelines and NCCN guidelines currently endorse use in ER positive, early stage, node-negative cancers only.

^†Patient must be categorized as at high clinical risk per MINDACT categorization (Adjuvant!Online).

^‡Includes eight breast cancer genes and three reference genes.

Table 2: Scoring for Multigene Prognostic Genomic Assays

Note.—NA = not applicable, ROR = risk of recurrence.

*Data are percentages.

^†Data in parentheses are percentages.

^‡The EPclin score is derived from a combination of the EndoPredict score, the tumor size, and the nodal status.

Oncotype DX and Oncotype DX DCIS are 21-gene prognostic markers developed to quantify the risk of distant breast cancer recurrence. This test allows analysis of tumor cells and quantification of expression of any of the 21 selected genes. The test was validated in patients with early-stage ER-positive HER2-negative node-negative breast cancer (96) with 0–3 positive nodes (97), and recently, in DCIS (98). The results for invasive cancers are reported by using a numerical recurrence score with a scale of 1–100 to categorize tumors as having a low risk (recurrence score < 11), intermediate risk (recurrence score of 11–25), and high risk (recurrence score > 25) of recurrence. In the case of DCIS, the numbers differ for low risk (recurrence score < 39), intermediate risk (recurrence score 39–54), and high risk (recurrence score > 54) tumors. The test allows patients with invasive cancers who would benefit from adjuvant chemotherapy in addition to tamoxifen to be identified (94) and those with DCIS who would benefit from adjuvant radiation therapy to be identified. For example, a patient with stage 1 hormone receptor–positive node-negative breast cancer and a recurrence score of 30 likely would benefit from adjuvant chemotherapy.

MammaPrint is a 70-gene expression profile dichotomous risk classifier. Researchers chose the top 70 genes from a list of 231 genes implicated in distant metastasis within 5 years (99). A patient with hormone receptor–positive HER2-negative node-negative breast cancer is categorized as being at low or high clinical risk of developing distant metastasis within 5 years with the use of an online risk assessment tool such as Adjuvant!Online (www.adjuvantonline.com). The MammaPrint assay can be used in patients at high clinical risk to guide decision making regarding adjuvant treatment (strong recommendation according to ASCO guidelines [95]). If the same patient has 1–3 positive lymph nodes and high clinical risk, the strength of the recommendation according to ASCO is only moderate. MammaPrint allows accurate discrimination of low and high risk of recurrence in early-stage node-negative breast cancer (100) and provides binary results, unlike Oncotype DX. This test is not to be used in patients at low clinical risk according to results of the Adjuvant!Online tool (101,102).

Endopredict is a gene expression profile that includes eight breast cancer genes and three reference genes used in hormone receptor–positive HER2-negative node-negative endocrine-treated breast cancers to estimate residual risk of early (0–5 years) and late (5–10 years) distant recurrence and is reported with a score on a scale of 0–15 (a score of ≤5 corresponds to low risk, indicating a less than 10% probability of distant recurrence) (103). Epclin incorporates nodal status and tumor size into the EndoPredict score to provide more prognostic information (104,105).

The Prosigna PAM50 risk of recurrence score, BCI (Breast Cancer Index, bioTheranostics, https://www.breastcancerindex.com), or uPA (urokinase plasminogen activator) and PAI-1U (plasminogen activator inhibitor type 1) are other genomic assays available that are also recommended for hormone receptor–positive HER2-negative node-positive or node-negative breast cancers to guide decision making for adjuvant chemotherapy. The strength of the recommendations varies according to the test.