Optimizing Treatment of "Triple-Negative" Breast Cancer

Introduction

The use of modern genomic techniques has significantly enhanced our understanding of breast cancer biology. Five distinct breast cancer tumor subsets have been recognized, including hormone receptor (HR)-positive luminal A and B, human epidermal growth receptor 2 (HER2)-positive, "normal"-like, and basal-like. This final group is frequently identified by conventional immunohistochemical techniques as "triple negative" because it lacks staining for estrogen receptor (ER), progesterone receptor (PR), and HER2. Triple-negative tumors, which often overexpress the epidermal growth factor receptor (EGFR) 1 and are positive for CK 5 and/or 6, are typically high grade and have a high risk of relapse within the first several years after initial diagnosis. Long-term follow-up^[1] of triple-negative cohorts has demonstrated a worse prognosis for the triple-negative subgroups than for those that are HR-positive. Given the differential outcomes for patients with this group of cancers, there is significant interest in better understanding the natural history of and best treatment for triple-negative disease; multiple presentations at the 2007 San Antonio Breast Cancer Symposium (SABCS) addressed this topic.

Classification and Prognosis

Several posters explored classification and prognosis of triple-negative disease. Conforti and colleagues^[2] attempted to characterize the use of immunohistochemical analysis in the identification of basal-like tumors. In their analysis of over 800 specimens from a single institution, triple-negative staining had positive and negative predictive values of 67% and 99%. This finding suggests that while not all triple-negative tumors are basal-like, non-triple-negative tumors are almost never basal-like.

Two posters examined outcomes in patients with small stage 1 triple-negative tumors. Kaplan and colleagues^[3] evaluated their institutional registry of over 800 patients with stage 1 disease. In general, those with triple-negative disease experienced more than 3 times the risk of recurrence compared with non-triple-negative patients, even after adjusting for tumor size and use of adjuvant chemotherapy. Similarly, a retrospective analysis^[4] of a small cohort of ≤ 1 cm triple-negative cancers demonstrated a significantly higher rate of recurrence (12.5%) compared with HR-positive or HER2-positive groups (1.3% and 7.4%, respectively, P = .007), despite increased use of adjuvant chemotherapy in the triple-negative subgroup.

Data on outcomes after metastatic recurrence were also presented. Dent and colleagues^[5] suggested triple-negative cancers not only were more likely to lead to distant recurrence and death than non-triple-negative cancers, but they also more often had visceral metastases as the first site of metastatic spread.

In general, these data supported previous observations while highlighting the need to optimize adjuvant therapies for this group of high-risk patients given the risks for recurrence and for visceral metastatic disease.

Treatment

Chemotherapy

Triple-negative tumors do not respond to endocrine agents or trastuzumab and can only be treated with chemotherapy. Fortunately, increasing evidence suggests that the triple-negative subgroup derives substantial and preferential benefit from chemotherapy. The Cancer and Leukemia Group B (CALGB)^[6] recently reviewed data from 3 randomized trials: 8541, investigating escalating dose intensity; 9344, exploring the addition of paclitaxel to the doxorubicin and cyclophosphamide (AC) backbone; and 9741, studying the impact of dose density. A combined analysis of over 6600 patients treated on these studies, stratified by HR status, demonstrated greater reductions in risk of recurrence for the HR-negative subset, translating into larger absolute benefits in both disease-free and overall survival.

The neoadjuvant setting also provides an opportunity to determine in vivo tumor responses to chemotherapy. The National Breast and Bowel Project (NSABP) B27 trial^[7] of neoadjuvant AC with neoadjuvant vs adjuvant taxane demonstrated a higher rate of pathologic complete response (pCR) in ER-negative rather than ER-positive tumors. Prospective analysis of outcomes after neoadjuvant chemotherapy stratified by molecular subtype has demonstrated that the "basal-like" group has a greater frequency of pCR compared with HR-positive/HER2-negative subgroups.^[8,9]

At the 2007 SABCS, several abstracts supported these findings. Four studies^[10-13] presented in poster format reported outcomes after exposure to neoadjuvant anthracycline ± taxane-containing regimens. In all 4 studies, pCR rates were significantly higher in the triple-negative subgroup compared with the HR-positive subgroups, with similarly high pCR rates in women with HER2-positive tumors.^[10-12] Among women with triple-negative tumors, the pCR rates were 23% to 36%. For trials with longer-term follow-up, inferior outcomes were observed among patients in the triple-negative group, especially in those with residual disease at surgery.^[13] Taken together, these outcomes demonstrate that not only is chemotherapy the primary choice of systemic therapy for triple-negative tumors, but it may also be a more efficacious choice.

With respect to the standard chemotherapy repertoire, some agents may be more effective than others in the treatment of triple-negative tumors. The newest chemotherapeutic agent available for treatment of metastatic breast cancer is ixabepilone, an epothilone analog. Epothilones bind tubulin, leading to stabilization of microtubules, cell cycle arrest, and subsequent apoptotic cell death; preclinical study has suggested activity in both taxane-sensitive and taxane-refractory tumors.^[14] In the primary phase 3 study leading to US Food and Drug Administration (FDA) approval,^[15] ixabepilone 40 mg/m² every 3 weeks plus capecitabine 2000 mg/m² were superior to capecitabine 2500 mg/m² monotherapy in 752 patients with anthracycline- and taxane-pretreated disease. A subgroup analysis was performed in the 25% of patients with triple-negative disease from the phase 3 study. In a group of 187 patients with triple-negative disease,^[16] response rate (RR) increased from 9% to 27% with the addition of ixabepilone to the capecitabine therapy, and progression-free survival (PFS) improved from 2.1 to 4.1 months (hazard ratio 0.68, 95% confidence interval 0.50-0.93). These relative improvements in RR and PFS were comparable to those seen in the study cohort as a whole, although the low RR of 9% with capecitabine monotherapy highlights the difficulty in treating this patient population. In the neoadjuvant setting, monotherapy with ixabepilone in patients with triple-negative tumors led to a pCR rate of 26%, comparable to that seen with other chemotherapeutics.^[17] Overall, ixabepilone appears to have reasonable activity in the triple-negative population.

Targeted Therapy

Bevacizumab. There has also been interest in the role of angiogenesis inhibitors in the treatment of triple-negative disease. Histologic examination of "basal-like" triple-negative tumors has demonstrated the presence of glomeruloid microvascular proliferation. These focal endothelial tufts, which portend a worse prognosis in node-positive breast cancer,^[18] may serve as targets for angiogenesis inhibitor therapy. Bevacizumab, a humanized anti-VEGF monoclonal antibody, has a demonstrated role in the treatment of several solid malignancies. In breast cancer, bevacizumab was studied in the first-line metastatic setting in ECOG 2100, a phase 3 randomized study evaluating paclitaxel with or without bevacizumab in essentially HER2-negative patients.^[19] Overall results demonstrated a significant and persistent improvement in disease-free survival with the addition of bevacizumab. Subset analysis has demonstrated activity of the bevacizumab-containing arm in the sizeable triple-negative population, although results were statistically not different from those seen in the HR-positive subgroup. Ongoing studies with angiogenesis inhibitors will attempt to further elucidate the role of these agents in the triple-negative population.

Cetuximab. Triple-negative tumors are known to overexpress EGFR,^[20] and 2 studies presented at SABCS 2007 explored the role of cetuximab, a humanized monoclonal antibody directed against EGFR, in this tumor type. Both of the trials included a platinum agent because preclinical data suggested that triple-negative tumors may have defects in BRCA1-mediated DNA repair and thus may be sensitive to DNA-damaging agents such as platinums.^[21]

Carey and colleagues^[22] presented initial results from TBCRC 001, a phase 2 study that randomized patients with triple-negative stage IV breast cancer to either cetuximab monotherapy at 250 mg/m² or cetuximab combined with weekly carboplatin at AUC 2. This report described the results from an interim analysis of the monotherapy arm only. In the analysis, 21 patents were accrued; out of this group, 1 patient (4%) experienced a prolonged partial response, 4 patients (19%) had stable disease for at least 8 weeks, and the rest developed disease progression, some very quickly. Upon progression, 19 patients were crossed over to the combination therapy arm, wherein 4 (28%) had a partial response and 4 more had stable disease. Due to the low observed response rate with monotherapy, that arm was closed to further accrual; results from the combination arm are awaited.

O'Shaughnessy and colleagues^[23] also presented data on the use of cetuximab in a randomized phase 2 study of weekly irinotecan (90 mg/m²) and carboplatin (AUC 2), with or without cetuximab (250 mg/m²). In the overall intention-to-treat study population, response rate improved marginally from 28% to 33% with the addition of cetuximab. Approximately half of the study population (78 patients) had triple-negative tumors. In that subgroup, cetuximab was associated with a response rate of 49% compared with 30% when chemotherapy was used alone. No significant differences in PFS or overall survival were observed in any subgroup, and significant toxicity led to dose reductions in starting doses of the chemotherapy agents. Based on the available evidence, there is little reason to believe that either single-agent cetuximab or a small molecule tyrosine kinase inhibitor of EGFR will have substantial activity in the triple-negative setting. It remains unknown whether these agents will prove useful in this context when combined with chemotherapy.

Novel Approaches

Some of the newest work with triple-negative breast cancers is based on genomics. Jones and colleagues^[24] presented an evaluation of alphaVbeta6 expression, an integrin typically upregulated in processes such as wound healing, inflammation, and neoplasia, in both triple-negative and HER2-positive subgroups. Using a tissue database, the investigators demonstrated that alphaVbeta6 status proved a stronger predictor of reduced survival than ER-negative status. AlphaVbeta6-negative status was associated with a 10-year survival rate of 70% in the triple-negative group compared with 55% among those in the triple-negative group who are positive for alphaVbeta6. Multivariate analysis demonstrated this marker to be independent of traditional prognostic features such as size, grade, and lymph node status. Similar results were seen in an analysis by Osborne and colleagues^[25] that used immunohistochemistry of triple-negative tumor samples from a neoadjuvant study. Long-term evaluation for recurrence demonstrated that tumors with high levels of cyclin E pretreatment were at significantly higher risk of relapse compared with those with lower cyclin E. These results not only suggest diversity within the triple-negative subgroup but also possibly identify targets for novel therapy. It is expected that further work dissecting the category of triple-negative breast cancer will demonstrate further heterogeneity in this group and offer the opportunity for even more tailoring of therapy.

Ongoing trials are attempting to develop improved regimens for patients with triple-negative tumors. The PACS 08 study will examine the role of sequential adjuvant FEC-100 and ixabepilone specifically for triple-negative disease.^[26] An ongoing study in metastatic triple-negative breast cancer is evaluating the role of sunitinib antiangiogenesis monotherapy compared with best available chemotherapy.^[27] Other emerging targets for treatment incorporate components of cellular proliferative pathways, including the phosphoinositide 3-OH kinase pathway and the mitogen-activated protein kinase pathway, DNA repair, and growth-factor receptors such as EGFR and c-kit. Agents currently in phase 1/2 evaluation for triple-negative disease include dasatinib and PARP1 inhibitors. It is hoped that further advances in targeted treatment and optimization of chemotherapy will provide more effective treatment and improved outcomes for this aggressive subclass of breast cancer.

This activity is supported by independent educational grants from Bristol-Myers Squibb, Roche, and Susan G. Komen for the Cure.

References