Determining the appropriate adjuvant systemic therapy regimen for an individual patient requires an estimate of her underlying risk of disease recurrence and a determination of the therapy to which she is most likely to respond. The former is based primarily on tumor-related prognostic factors that reflect tumor biology irrespective of the treatment administered. This contrasts with predictive factors, which are indicative of responsiveness to specific therapies and are discussed later in association with general treatment recommendations. It is important to note, however, that individual factors may be prognostic, predictive, or both (6).

Prognostic markers are patient- or tumor-related factors that allow for a determination of clinical outcome in the absence of systemic therapy. The use of prognostic factors provides the clinician with a means to estimate the benefits of adjuvant therapy and thus forms the basis for systemic treatment recommendations. Prognostic factors tend to reflect the underlying biology of the tumor, such as its ability to proliferate, invade, and induce angiogenesis. Tumor size, hormone receptor status, human epidermal growth factor receptor (HER)-2/neu (i.e., c-erbB2) status, histopathologic features, including tumor type and grade, and the number of involved lymph nodes are all markers that have documented prognostic value. Other indicators, such as proliferative rate, the presence of circulating tumor cells (CTCs), have been evolving in different studies. Gene expression profiles generated by high-throughput, DNA or RNA-based methodologies simultaneously measure the expression of thousands of genes and have led to the identification of biology-based prognostic profiles, several of which have been validated and are in clinical use.

While clinicopathologic parameters aid in decision on adjuvant treatment for breast cancer, substantial tumor heterogeneity occur. Gene expression profiling is a molecular tool that more precisely defines the intrinsic characteristics of each individual tumor.

In 2000, Perou (24) and Sørlie (25) demonstrated that ER-positive and ER-negative breast cancers are completely distinct diseases on a molecular level. In addition, hierarchical cluster analyses of genes that vary more between tumors than between repeated samples of the same tumor, the so-called intrinsic genes, revealed at least four molecular subtypes of breast cancer, namely, luminal, HER-2 enriched, basal like, and normal breast like (24).

Luminal A tumors, which probably make up about 40% of all breast cancers, usually have high expression of ER-related genes (ER alpha, GATA-binding protein 3, X-box binding protein 1), low expression of the HER-2 cluster of genes, and low expression of proliferation-related genes (26,27). Luminal A tumors are the most common subtype and in general, carry the best prognosis of all breast cancer subtypes (28). The less common (about 20%) luminal B tumors have relatively lower (although still present) expression of ER-related genes, variable expression of the HER-2 cluster, and higher expression of the proliferation cluster. Luminal B tumors carry a worse prognosis than luminal A tumors (29).

The HER-2–enriched subtype makes up about 10% to 15% of breast cancers and is characterized by high expression of the HER-2 and proliferation gene clusters (ERBB2 amplicon at 17q22.24, including ERBB2 and GRB7) and low expression of the luminal and basal clusters. For this reason, these tumors are most often negative for ER and PR, and positive for HER-2. However, the HER-2–enriched subtype is not synonymous with clinically HER-2–positive breast cancer. While half of clinically HER-2–positive breast cancers are HER-2 enriched, the other half can include any molecular subtype but is mostly made up of HER-2–positive luminal subtypes (30).

The ER-negative genomic profile includes multiple subtypes, such as basal like, claudin low, and interferon rich, among others. Most of these fall under the category of triple-negative breast cancers (TNBCs) because they are also PR negative and HER-2 negative.

Multiple independent datasets have shown the prognostic effect of intrinsic subtyping of breast cancer with luminal A cases to be at low risk of early recurrence. Apart from its prognostic implication, the intrinsic molecular
subclassification seems to predict responsiveness to chemotherapy. Studies evaluating the association between molecular subtype and pathologic complete response (pCR) rate to preoperative (neoadjuvant) chemotherapy have reported the highest rate of pCR in basal-like (30% to 45%) and HER-2–enriched (33% to 55%) subtypes, while pCR rates for luminal B (1% to 15%) and luminal A disease (0% to 7%) were substantially lower (31).

Standard clinicopathologic features have suboptimal prognostic and predictive utilities in predicting chemotherapy benefit (32). Multigene expression profiles have significantly increased our ability to estimate distant recurrence in ER-positive HER-2–negative breast cancer (33). In recent years, several novel gene expression prognostic tests have been developed. The commonly used multigene profiles are listed in Table 14-1.

The 21-gene Recurrence Score (RS) (Oncotype DX; Genomic Health, Inc., Redwood City, CA) uses a real-time reverse transcriptase-polymerase chain reaction–based method and requires formalin-fixed, paraffin-embedded tumor tissue. The assay includes genes related to cellular proliferation, ER signaling, the HER-2/neu pathway, and tumor invasive potential in addition to five reference genes that serve as an internal control to generate scores of 0 to 100, where higher scores indicate a higher probability of distant recurrence at 10 years in patients treated with adjuvant tamoxifen therapy alone (34). The RS result represents an individualized estimate of risk of distant recurrence and/or breast cancer–specific mortality (prognosis) and predicts the likelihood of adjuvant chemotherapy benefit (35). The Oncotype DX assay achieved level IB evidence and has been incorporated into current National Comprehensive Cancer Network and American Society of Clinical Oncology guidelines, to aid in treatment decisions for patients with HR-positive, node-negative, early-stage breast cancer (36). Results of the TAILORx study suggest little-to-no benefit of adjuvant chemotherapy for patients with HR+, HER-2–negative, node-negative breast cancer and RS ≤25 (35,37). Endocrine treatment alone was noninferior to chemoendocrine treatment in women ≤ 50 years of age with RS of 11 to 15 and in women > 50 years of age with RS of 11 to 25. When adjusted for age, chemotherapy benefit was supported in women ≤ 50 years of age with RS of 16 to 25. An ongoing study, Treatment for Positive Node, Endocrine Responsive Breast Cancer (RxPONDER; SWOG S1007), should provide further insight into the RS cutoff at which a chemotherapy benefit can be detected in patients with N+ breast cancer (38).

The 70-gene profile (MammoPrint; Agendia, Huntingdon Beach, CA), originally designed for unfixed tissue, has recently been adapted to formalin-fixed tissue. The 70-gene MammaPrint assay stratifies patients into low or high risk for distant metastases at 5 years. Large-scale retrospective validation of this profile was performed on frozen breast tumor specimens collected from 326 node-negative patients through a collaborative effort of the TRANSBIG consortium (39). The Microarray in Node-Negative Disease May Avoid Chemotherapy (MINDACT) trial reported that women with a high–clinical risk and low-risk MammaPrint result had a risk of distant metastases only 1.5% higher at 5 years without chemotherapy, suggesting that these patients can safely forgo chemotherapy (40).

In 2009, Parker et al. (41) developed an efficient 50-gene classifier, called Prediction Analysis of Microarray (PAM50) that reanalyzed the previous subgroups defining the four major intrinsic subtypes currently known: luminal A, luminal B, HER-2 enriched, and basal like. The PAM50 gene signature has been developed into a clinical test, the Prosigna gene signature assay (Prosigna; NanoString Technologies, Inc., Seattle, WA) validated to estimate the prognosis for postmenopausal patients with ER-positive early-stage breast cancer. Along with the identification of subtypes, PAM50 has been shown to be an independent predictor of survival in breast cancer (42). PAM50 generates a numerical score ROR that along with clinical features estimates the risk of relapse at 10 years in postmenopausal women with stage I/II node-negative or stage II node-positive (one to three positive lymph nodes) and HR-positive breast cancer (43).

The Breast Cancer Index (BCI; Biotheranostics, Inc., San Diego, CA) combines two biomarkers (HOXB13:IL17BR ratio and molecular grade index) to derive the BCI score. The BCI score has been validated to provide prognostic information in the NCIC MA.14 study (44). The BCI linear (BCI-L) model was also found to be strongly prognostic of both early and late recurrences, and its strength may be the ability to identify patients at risk of late recurrence as candidates for extended endocrine therapy (45). For node-positive patients treated in the aTTom trial BCI predicted benefit of extending tamoxifen therapy from 5 to 10 years (46).

EndoPredict (Myriad Genetics, Salt Lake City, UT) is a 12-gene RT-polymerase chain reaction–based assay. The EP classifies tumors into high- and low-risk categories and can also be combined with tumor size and nodal status to derive the EPclin. EPclin can potentially reduce the use of adjuvant chemotherapy by identifying patients with a good prognosis that would have been classified as intermediate or high risk based on clinicopathologic criteria and spares the use of chemotherapy in this group (47). The EPclin has been shown to be able to accurately prognosticate recurrence risk up to 10 years in LN0 and LN+ patients (48). There is no prospective data relating to chemotherapy benefit for EndoPredict or EPclin.

CTCs represent the liquid component of solid tumors and indicate the presence of residual disease. CellSearch (Menarini Silicon Biosystems, Inc; Bologna, Italy) is the only FDA-cleared assay for CTC analysis that uses a positive selection method with antibody-coated magnetic beads for detection and enumeration of EpCAM-positive CTCs. Ten percent to 30% of patients with localized breast cancer have detectable CTCs (≥1 CTC per 7.5 mL blood) at the time of diagnosis, before or after neoadjuvant or adjuvant chemotherapy, which is prognostic for higher recurrence and breast cancer mortality rates (49,50). In the phase III SUCCESS-A trial at 2 years, a positive CTC assay was associated with a 3.9-fold higher risk of death and a 2.3-fold higher ROR in multivariate models that included clinicopathologic features and CTC status at baseline; sensitivity analysis showed this effect only in HER-2–negative disease (49). CTCs were reported in 5% of patients with localized HR-positive breast cancer 5 years after diagnosis and were associated with higher ROR (51). In a meta-analysis that included individual patient data from 21 studies, CTCs were detected in 25.2% of patients before NACT, and a higher CTC number had a detrimental impact on OS (P <.001), distant DFS (P <.001), and locoregional relapse–free interval (P <.001) but not on pCR (52). CTCs have not been predictive of benefit from any specific systemic therapy or change in therapy. A recent study from a large national database cohort and the SUCCESS trial demonstrated that CTC status was a predictive factor of the benefit of radiotherapy in early-stage breast cancer (53).

Neoadjuvant therapy originally was developed as a way of reducing tumor size, to facilitate surgical resection and increase breast-conserving surgery (BCS) rates. More recently, neoadjuvant treatment has also been used for operable early breast tumors to eliminate axillary node metastases and avoid axillary node dissection (54).

Neoadjuvant systemic therapy is at least as effective as in an adjuvant setting but has several additional benefits. It allows for in vivo determination of tumor chemosensitivity. pCR after NACT has been identified as an independent prognostic factor defined by ypT0/Tis (55). Lastly, the neoadjuvant setting is a powerful model for the development of new drugs by using pCR as a surrogate outcome for event-free and overall survival and enables seeking prognostic and predictive biomarkers (56).

Selection criteria for neoadjuvant therapy include axillary node involvement, tumor size >2 cm, triple-negative or HER-2–positive subset, high proliferative index, unresectable tumors, and inflammatory carcinoma (57). The best management approach involves a multidisciplinary tumor board discussion at the time of diagnosis. The current standard treatment for triple-negative, and HER-2–positive and HR-positive subtypes is reviewed.

The main aim of systemic treatment in addition to local treatment is to eradicate distant micrometastases to increase OS. The selection of patients for neoadjuvant therapy is based on tumor characteristics, stage of breast cancer, and patient performance status and comorbidities. Although TNBC is generally associated with the poorest prognosis among breast cancer types, the rate of pCR to NACT in TNBC is relatively high. Up to 45% of triple-negative tumors show a pCR after NACT. Achieving pCR is associated with longer PFS and OS (58). In contrast, triple-negative tumors with residual disease have a high probability of recurrence (59).

The chemotherapy regimens employed in the neoadjuvant setting are the same as those used in the adjuvant setting. Standard anthracycline–taxane (A/T)-based
NACT combinations yield pCR rates between 25% and 40%. A meta-analysis of the EBCTCG in 44,000 patients showed that the addition of a taxane to a fixed anthracycline-based regimen improves BCSS, with a hazard ratio (HR) of 0.86 (standard error [SE] 0.04, p = 0.0005) (60). Taxanes are equally effective if administered concurrently or sequentially with anthracyclines, although concurrent regimens such as docetaxel/doxorubicin/cyclophosphamide (TAC) show increased toxicity and require prophylactic administration of granulocyte colony–stimulating factor (Sparano-3) (61). Within the sequential regimens, weekly paclitaxel improves DFS and OS compared with 3-weekly paclitaxel. Three-weekly docetaxel also improves DFS compared with 3-weekly paclitaxel, but not OS (Sparano-3).

More frequent administration of cytotoxic therapy (dose dense) is a more effective way of minimizing residual tumor burden than dose escalation. In a meta-analysis of 10 randomized controlled trials (RCTs), dose-dense–administered chemotherapy improved OS by 16% (HR 0.84, 95% CI 0.72 to 0.98, p = 0.03) and DFS by 17% (HR 0.83, 95% CI 0.73 to 0.94, p = 0.005) (62). In the few trials that were designed to analyze the pure effect of dose-dense compared with standard-dose chemotherapy, the benefit on both OS and DFS was largest for HR-negative tumors (62,63). A recent pooled analysis of two Italian trials showed a larger benefit of the dose-dense regimen for premenopausal women (64).

The addition of carboplatin to standard anthracycline–taxane-containing NACT in unselected TNBCs improved pCR rates from 37% to 53% in the GeparSixto trial (65), and from 31% to 58% in the BrighTNess trial (66), albeit with higher toxicity. Longer follow-up is needed to detect change in event-free survival or OS. However, despite an early signal in the metastatic disease setting that platinum agents may preferentially benefit TNBC patients with BRCA mutations, this observation has not been replicated in the early-stage setting (66,67).

Patients unfit for anthracyclines (e.g., due to cardiac dysfunction) or having node-negative early-stage triple-negative cancers may benefit from four cycles of docetaxel/cyclophosphamide (TC) every 3 weeks. This regimen improved OS compared with four cycles of Adriamycin/cyclophosphamide (AC) after a median follow-up of 7 years (HR 0.69, 95% CI 0.50 to 0.97, p = 0.032) (68). An anthracycline-free docetaxel and carboplatin combination × 6 cycles yielded a 55% pCR rate with 3-year OS of 87% (69).

Adding the programmed death-1 (PD-1) inhibitor pembrolizumab to NACT may improve outcomes in early-stage TNBC. In KEYNOTE-522, researchers assigned 1,174 patients with newly diagnosed, early-stage TNBC in a 2:1 ratio to receive neoadjuvant pembrolizumab plus chemotherapy or a placebo plus chemotherapy. In an analysis of the first 602 patients, performed after a median follow-up of 15.5 months, 64.8% of those in the pembrolizumab arm experienced a pCR compared with 51.2% of those in the chemotherapy-only group—a statistically significant difference that was seen regardless of nodal stage, tumor size, patient age, chemotherapy regimen, or PD-L1 expression (70).

Poly ADP-ribose polymerase (PARP) inhibitors have been evaluated in this scenario, targeting the potential homologous recombination deficiency in TNBC. The PARTNER study (NCT03150576) will test olaparib in addition to platinum-based NACT for TNBC and/or BRCA-mutant patients.

Approximately 15% to 20% of breast cancers are HER-2 positive. Without HER-2–directed treatment, HER-2–positive breast cancer is characterized by an aggressive course of disease and a poor prognosis. Anti–HER-2 treatment has changed the natural biology of this disease. Several studies have demonstrated a high pCR rate (up to 60%) with neoadjuvant therapy that includes anti–HER-2 regimen. The response rates are higher in ER-negative than in ER-positive cases. Targeted agents currently in use are summarized in Table 14-2.

Endocrine therapy is the mainstay of treatment for ER-positive (ER+) breast cancer. At approximately 75% of all breast cancers, ER-positive constitutes the most common subtype of the disease. Although medical therapy for localized breast cancer is primarily used in the adjuvant setting, it can also be effectively used in the neoadjuvant (preoperative) setting.

The rationale for neoadjuvant therapy in HR+ breast cancer is similar to other clinical subtypes of breast cancer: (1) the ability to downstage breast cancers in postmenopausal women with ER-positive tumors, thus making breast-conserving therapy possible; (2) to understand prognosis based on progression-free survival (PFS) or OS; and (3) for biomarker evaluation and targeted therapy development. pCR to a given therapy is often used as a surrogate of long-term survival and cure from breast cancer.

Neoadjuvant endocrine therapy (NET) historically has been reserved to treat elderly patients with ER-positive breast cancer, who were not considered good candidates for systemic chemotherapy or surgery (82). More recently, the ability to identify early endocrine responsiveness and the development of highly effective aromatase inhibitors (AIs) has resulted in a broader use of NET. NET, however, has been less frequently incorporated into practice due to the slow tumor response requiring prolonged therapy as well as the less defined prognostic information that is obtained after treatment (84). A systemic review and meta-analysis
of 20 randomized clinical trials with a total sample size of 3,490 women, NET, even as monotherapy, was associated with response rates like those of neoadjuvant combination chemotherapy but with lower toxicity (83). NET can also be used in selected premenopausal women and is often combined with ovarian suppression (85).

Complete pathologic responses are infrequent with NET; however, residual disease does not imply poor prognosis. AIs are more effective than tamoxifen in downstaging ER-positive tumors. From the available data in studies of postmenopausal women, 4 months to 6 months of an AI seems optimal with modest persistent benefits thereafter. NET has been shown to cause a significant reduction in Ki67, with the degrees of suppression related to the level of ER expression. Ki67 levels after NET have been shown to be prognostic. The IMPACT trial demonstrated that high Ki67 expression levels after 2 weeks of NET was associated with a poorer recurrence-free survival (RFS) (86). The 21-gene RS has also been shown to have predictive value. Ueno et al. evaluated pretreatment and posttreatment tumor tissue from patients with estrogen-positive tumors treated with neoadjuvant exemestane. The clinical response rate was 59% in patients with a low RS compared with 20% in patients with a high RS (87). The TransNEOS study validated the 21-gene test as a predictor of clinical response to neoadjuvant hormonal therapy. Among patients with large tumors (≥ 2 cm), 54% of those with RS <18 achieved complete response or partial response (CR or PR) with neoadjuvant letrozole, and 79% were BCS recipients, including many who were BCS noncandidates before neoadjuvant treatment. In contrast, patients with RS ≥ 31 had a higher rate of PD with neoadjuvant letrozole. Multivariate analyses showed that the RS result significantly predicted clinical response to neoadjuvant hormonal therapy, even after adjustment for clinical covariates (age, tumor size, and tumor grade). RS group was significantly associated with rate of BCS after neoadjuvant treatment (RS < 18 vs. RS ≥ 31, p = 0.010) (88).

Multiple additional studies are presently underway to better define the impact of NET on early-stage breast cancer. The New Primary Endocrine-therapy Origination Study (NEOS), a multicenter phase III randomized trial, will assess the need for adjuvant chemotherapy in postmenopausal women with stage T1c-T2N0M0, HR-positive tumors who responded to neoadjuvant letrozole (89). The Alternate Approaches for Clinical Stage II or III Estrogen Receptor–Positive Breast Cancer Neoadjuvant Treatment (ALTERNATE) in Postmenopausal Women: A Phase III Study, is an ongoing trial that seeks to define the rate of endocrine-resistant disease, RFS, and rates of pCR in this population after treatment with fulvestrant, anastrozole, or the combination of the two agents (90). The ability to identify good and poor responses to NET early in treatment provides a strategy to triage poor responders into clinical trials of targeted agents to address endocrine resistance. Multiple trials are ongoing that combine molecularly targeted agents with endocrine therapy in the neoadjuvant setting. NeoPalAna is enrolling the endocrine-resistant population based on high Ki67 on endocrine therapy to receive the combination of the cyclin dependent kinase (CDK)4/6 inhibitor palbociclib and anastrozole to overcome endocrine resistance.

The Capecitabine for Residual Cancer as Adjuvant Therapy (CREATE-X) trial was conceived with the aim of improving the prognosis of patients who did not achieve a pCR after NAT. This multicenter, open-label, randomized, phase III trial enrolled patients with stage I to IIIB HER-2–negative early breast cancer who had residual invasive disease and/or tumor-positive lymph nodes after standard NAT containing anthracycline, taxane, or both. After surgery, the capecitabine group received oral capecitabine (1,250 mg/m², twice per day, on days 1 to 14) every 3 weeks for 6 to 8 cycles. The trial was terminated early after 887 patients for meeting its primary end point: at the final analysis, the 5-year DFS improved from 67.6% to 74.1% (HR 0.70, 95% CI 0.53 to 0.92, p = .01) and OS improved from 83.6% to 89.2% (HR 0.59, 95% CI 0.39 to 0.90, p = .01). The most prominent HR was seen in the subgroup of 286 patients (32.2%) with TNBC (DFS: HR 0.58, 95% CI 0.39 to 0.87; OS: HR 0.52, 95% CI 0.30 to 0.90) (91).

Two randomized phase III trials are currently investigating the role of immune checkpoint inhibitors in this setting: the A-Brave trial (NCT02926196), with the anti–PD-L1 avelumab tested as adjuvant or post-NAT treatment for high-risk TNBC patients, and the SWOG S1418 trial (NCT02954874), in which TNBC patients with residual disease (RD) after NAT are randomized to adjuvant pembrolizumab or observation.

No evidence is available on the efficacy of post-NAT platinum-based chemotherapy for TNBC, but the ongoing ECOG-ACRIN EA1131 study (NCT02445391) will examine this issue further, with patients who did not achieve pCR randomized to adjuvant carboplatin, cisplatin, or observation. Based on the results of the CREATE-X
trial, the observation arm was replaced by adjuvant capecitabine as comparator.