During the last few years, breast cancer clinicians have seen an increase in novel treatment options for their patients. Disease types with major expansions have been those that are estrogen receptor-positive or amplified by human epidermal growth factor receptor 2-positive (HER2+).
In contrast, clinicians treating triple-negative breast cancer (TNBC) have observed devastating consequences from early relapse; a lack of specific, targeted therapies; premature development of drug resistance; and short overall survival (OS).
Treating patients with TNBC remains a challenge because this heterogeneous disease is not defined by a specific biology. Moreover, compared with other breast cancer subtypes, TNBC is associated with a median survival of only 12 months, visceral metastasis, and recurrence within 3 years of diagnosis. To improve outcomes, clinicians must rely on novel investigational therapies and the identification of subgroups of patients most likely to obtain benefit.
Biologic Characteristics: TNBC Subtypes
TNBC, which constitutes 10-20 percent of all breast tumor cases, is characterized by a lack of expression of estrogen receptor, progesterone receptor, and HER2/neu (Clin Cancer Res 2007;13(15 Pt 1):4429-4434). Although not synonymous with the basal-like (BL) subtype defined by gene expression profiling, approximately 70 percent of TNBCs have BL characteristics.
TNBC is more common in young women, those of African descent, and those with BRCA1 germline mutations (JAMA 2006;295(21):2492-2502). TNBCs are generally high-grade, aggressive tumors with a high rate of distant metastasis. Poor outcomes occur even though standard chemotherapy regimens have activity against these tumors. In fact, the TNBC subgroup is typically more responsive than others to preoperative chemotherapy and associated with a high rate of pathological complete response (pCR); however, it also has higher overall rates of relapse, which is likely attributable to residual disease persisting after neoadjuvant therapy.
Increasingly, TNBC is being divided into several molecular subtypes thanks to a variety of algorithms for classifying this disease. Researchers analyzed gene expression profiles from over 500 TNBC cases and identified six unique subtypes: BL (subtypes BL1 and BL2), immunomodulatory, mesenchymal, mesenchymal stem-like, and luminal androgen receptor (LAR) (Breast Cancer Res 2014;16(4):406).
In a retrospective analysis of TNBC patients who received neoadjuvant chemotherapy at a single institution, the Lehmann and Bauer's algorithm paired with Affymetrix microarrays demonstrated that classifying TNBC by seven subtypes (six subgroups and one unstable subtype) predicts high versus low pCR rate (Clin Cancer Res 2013;19(19):5533-5540). For instance, the BL1 subtype possesses a pCR of 52 percent; meanwhile, the LAR and BL2 subtypes have the lowest rates of pCR, 10 percent and 0 percent, respectively.
Similarly, Burstein and collaborators, using DNA and RNA sequencing on 200 patients with TNBC identified four subtypes: LAR, mesenchymal, BL immunosuppressed (BLIS), and BL immune activated (BLIA) (Clin Cancer Res 2015;21(7):1688-1698). In this classification, the BLIS subtype conferred the poorest prognosis and BLIA the best prognosis for disease-free and disease-specific survival.
These discoveries have generated a move toward precision medicine and characterizing TNBC molecularly in a way that can guide cancer therapy. Prospective clinical studies are necessary to validate these findings.
The genomic revolution has considerably improved clinicians' understanding of breast cancer biology. With the advent of next-generation sequencing, molecular dissection of TNBC is revealing novel targets, several of them under investigation. The Cancer Genome Atlas is a large-scale analysis of primary human breast cancers with several high-throughput technologies (Nature 2012;490(7418):61-70).
Cytotoxic chemotherapy has been the mainstay in treating TNBC. For more than 2 decades, taxanes and anthracyclines have been administered in sequential or concurrent fashion, and they represent the current standard of care. More recently, considerable interest has risen in using platinum salts because dysfunction of homologous recombination DNA repair in TNBC sensitizes tumors to these agents.
The TNT clinical trial (NCT00532727) randomized 376 untreated TNBC patients to a total of six cycles with one initiated every 3 weeks of either carboplatin at the full area-under-the-curve dose times six (AUC6) or standard-of-care first-line therapy with 100 mg/m2 of docetaxel. No difference in progression-free survival (PFS) or OS was detected between the treatments. A prespecified subset analysis showed that patients with germline BRCA1 or BRCA2 mutations had a higher overall response rate (ORR) than those without the mutations (68% vs. 33%; p=.03) and longer PFS (6.8 months vs. 4.8 months) when treated with carboplatin rather than with docetaxel. However, patients with homologous recombination deficiency experienced similar outcomes.
The Triple-Negative Albumin-Bound Paclitaxel Combination International Treatment Study (tnAcity) (NCT01881230) randomized 191 untreated patients with metastatic TNBC to receive nab-paclitaxel in combination with carboplatin, or nab-paclitaxel with gemcitabine, or gemcitabine plus carboplatin. In this study, patients on the nab-paclitaxel with carboplatin regimen demonstrated a PFS rate that was statistically significantly longer, and they enjoyed a better risk-benefit profile than did those taking nab-paclitaxel plus gemcitabine or gemcitabine plus carboplatin.
Beyond Cytotoxic Chemotherapy
The current standard of care in the treatment of TNBC offers limited therapeutic options aside from traditional cytotoxic chemotherapy agents. On the horizon, however, are several potential signal pathways to target with novel drugs.
About 20 percent of TNBC tumors express PD-L1, which suggests they may represent a viable target for immunotherapy, such as checkpoint blockade, which targets CTLA-4 and the PD-1/PD-L1 interaction to take the brakes off the tumor immune response, allowing antitumor T cells to proliferate.
The KEYNOTE-012 study (NCT01848834) was a multicenter, non-randomized phase Ib trial of single agent pembrolizumab to patients with advanced PD-L1-positive advanced cancer (J Clin Oncol 2016;34(21):2460-2467). In the TNBC cohort, 32 women were enrolled, and the ORR was 18.5 percent (one complete response [CR] and four partial responses [PRs]). Twenty-six percent (7/32 patients) had stable disease. The median duration of response was not reached, and three responders remained on study for at least 1 year.
These encouraging results led to the initiation of KEYNOTE-086, an international phase II study evaluating pembrolizumab monotherapy (200 mg IV every 3 weeks) in metastatic TNBC. The primary endpoint was ORR. Cohort A included 170 patients with metastatic TNBC who had been previously treated, and responses occurred in eight patients, including one CR and seven PRs. Twenty-seven percent of patients had a decrease in target lesion size from baseline (J Clin Oncol 2017; doi:10.1200/JCO.2017.35.15_suppl.1008). PD-L1 status was not associated with response (PD-L1-positive and PD-L1-negative cases had an ORR of 4.8% and 4.7%, respectively).
Preliminary results from cohort B (previously untreated TNBC patients), which included 52 patients with PD-L1-positive disease, showed an ORR of 23.1 percent with 4 percent of patients experiencing CRs and 50 percent having targeted lesions that decreased in size (J Clin Oncol 2017; doi: 10.1200/JCO.2017.35.15_suppl.1088). All-grade immune-mediated adverse events occurred in 18.8 percent of patients. Only 1.2 percent of these were grade 3/4 and none led to death.
An update of biomarker analysis of this trial was presented at the 2017 European Society for Medical Oncology Annual Meeting. The results demonstrated that pembrolizumab response was significantly related to stromal tumor-infiltrating lymphocyte (TIL) levels in both cohorts (Ann of Oncol 2017;28(Supp 5): Abstr LBA13).
I-SPY 2 (NCT01042379) is an ongoing neoadjuvant, phase II, adaptively randomized, multicenter study in newly diagnosed, locally advanced breast cancer. The addition of pembrolizumab to standard chemotherapy increased the rate of pCR response approximately threefold (J Clin Oncol 2017; doi: 10.1200/JCO.2017.35.15_suppl.506). Reviewing the results, investigators reported that pembrolizumab “graduated” (met a prespecified threshold) in all HER2-negative signatures, giving a high Bayesian predictive probability of success (>85%) for the drug in a confirmative phase III trial of 300.
The PD-L1 antagonist atezolizumab (MPDL3280A) also is being evaluated in TNBC. A phase I study of atezolizumab has an expansion cohort in heavily pretreated patients with PD-L1-positive and PD-L1-negative TNBC. Initial data show the ORR was 19 percent, including 9.5 percent CRs and 9.5 percent PRs (Cancer Res 2017;77(13 Suppl):2986). The phase III IMpassion 130 trial (NCT02425891) is investigating the combination of atezolizumab plus nab-paclitaxel as a first-line therapy for TNBC.
The Southwest Oncology Group's S1418 study (NCT02954874) is an ongoing randomized trial in which pembrolizumab is being compared with observation in patients who have TNBC with >1 cm residual cancer or any lymph-node involvement after neoadjuvant chemotherapy. The primary endpoint is disease-free survival.
Breast tumors have TILs, and substantial evidence indicates that patients with high TIL concentrations had higher rates of pCR when patients received neoadjuvant chemotherapy than did patients whose tumors had no lymphocytic infiltrate. Evidence suggests that PD-L1 expression in infiltrating tumor cells may be a predictive biomarker. Furthermore, mutational load or the presence of neoantigens might represent a predictive biomarker for the use of immune checkpoints.
What we know today is that in most cases immune checkpoint inhibitors work well against hypermutated cancers. Despite the expanding landscape of immunotherapy use in lung and melanoma patients, a difficult biology puzzle remains to be solved. For instance, why do immune checkpoint inhibitors work so well in some cancers and not at all in others? Among the same types of cancer, why do some respond while others do not?
Antibody-drug conjugates (ADCs) represent an emerging therapeutic modality in which antibodies are used as vehicles for delivering a cytotoxic drug selectively to a target such as a tumor-associated antigen. Glembatumumab vedotin (CDX-011) is a fully human IgG2 monoclonal antibody against the tumor-associated antigen glycoprotein NMB (gpNMB) and the potent microtubule inhibitor monomethyl auristatin E. Glembatumumab is designated to bind to gpNMB and after internalization release monomethyl auristatin E, resulting in tumor cell death by microtubule inhibition.
The EMERGE clinical trial (NCT01156753) was a randomized phase II study in heavily pretreated TNBC patients in whom expression of gpNMB occurred in more than 5 percent with treatment of glembatumumab compared with a physician's choice of therapy (PCT) (J Clin Oncol 2015;33(14):1609-1619). A total of 124 patients were enrolled, and the ORR was 6 percent versus 7 percent for patients treated with glembatumumab versus PCT, respectively. Interestingly, in patients with gpNMB overexpression greater than 25 percent, the ORR ascended to 30 percent.
This study suggests that patients with TNBC and tumor gpNMB overexpression may potentially derive the greatest benefit from glembatumumab. Another study with glembatumumab (NCT01997333) has been initiated to confirm these findings. In this study, 300 patients with metastatic gpNMB-overexpressing TNBC will be randomized to receive glembatumumab or capecitabine in a 2:1 ratio. This trial is active but not accruing patients, according to clinicaltrials.gov.
Sacituzumab govitecan (IMMU-132) is an anti-Trop-2 ADC. Trop-2 is a 46-kD glycoprotein initially identified in a trophoblast cancer cell line and is overexpressed in several solid tumors, including TNBC. Sacituzumab comprises a toxic payload, SN-38, which is a topoisomerase I inhibitor that causes double-stranded DNA breaks that lead to apoptosis. Irinotecan, the prodrug of SN-38, has activity in several solid cancers.
A phase I dose-finding and phase II dose-expansion study in advanced solid tumors including TNBC demonstrated an important therapeutic activity (NCT01631552). A total of 69 patients with TNBC who received a median of five prior therapies received a 10 mg/kg sacituzumab starting dose on days 1 and 8 of 21-day repeated cycles. The confirmed ORR was 30 percent, clinical benefit ratio (CBR) was 46 percent, median PFS was 6 months, and median OS was 16.5 months. In terms of toxicity, adverse events over grade 3 occurred in 41 percent of patients, with the most common being neutropenia (39%), leukopenia (16%), anemia (14%), diarrhea (13%), and vomiting (10%). On the basis of these promising results, in February 2016, sacituzumab obtained the designation of breakthrough therapy by the FDA.
Therapies Targeting Defective DNA Repair
Approximately 5 percent of unselected patients with breast cancer carry a germline BRCA1 or BRCA2 mutation. These mutations are tumor-suppressor genes that encode proteins involved in the repair of DNA double-strand breaks by a homologous recombination repair pathway. Inhibition of PARP blocks the alternative mechanism of DNA repair, leading to multiple double-strand breaks and cell death.
OlympiAD (NCT02000622) was a randomized phase III study that assessed efficacy and safety of olaparib versus PCT in patients with HER2-negative and germline BRCA-mutated breast cancer (N Engl J Med 2017;377(6):523-533). A total of 301 patients were randomized and the ORR was 60 percent and 28 percent in olaparib and PCT arm, respectively. At 77 percent of data maturity, PFS was significantly longer in patients treated with olaparib instead of the PCT (7 months vs. 4.2 months; HR, 0.58; 95% CI, 0.43, 0.80; p=.0009). The risk of disease progression or death was 42 percent lower and the median PFS survival was 2.8 months longer with olaparib than with standard therapy.
Olaparib was well-tolerated, with less than 2 percent of patients discontinuing treatment because of toxicity. The most important toxicities were nausea, anemia, and fatigue. In October 2017, the FDA granted a priority review to a supplemental new drug application for olaparib in patients with previously treated metastatic HER2-negative, germline BRCA-positive breast cancer.
EMBRACA (NCT01945775) is a randomized, phase III trial to compare the efficacy and safety of talazoparib versus PCT. Talazoparib, a dual mechanism PARP inhibitor that inhibits the PARP enzyme and also traps PARP on DNA, prevents DNA damage repair, leading to death in BRCA1/2-mutated cells. The median PFS was 8.6 months for patients in the talazoparib arm compared to 5.6 months for those in the PCT arm (HR, 0.54; p<.0001) (SABCS 2017; Abstract GS6-07). Global health status and/or quality of life showed overall improvement from baseline and a delay in the time to clinical deterioration in patients receiving talazoparib (HR, 0.38; p<.0001). An interim analysis of OS was conducted; a positive trend favoring talazoparib was observed with a 24 percent reduction in the risk of death.
Therapies Targeting Androgen Receptor
The androgen receptor (AR) is present in approximately 80 percent of invasive breast cancer patients and up to 30 percent of patients with TNBC (Endocr Relat Cancer 2016;23(4):323-334). A meta-analysis published in 2013 demonstrated that high AR expression predicted a favorable outcome in the total population of patients (J Natl Cancer Inst 2014;106(1):djt319).
In contrast, in the subgroup of patients with BL disease, of whom approximately 80 percent had TNBC, the presence of AR was associated with poor outcome (Int J Cancer 2012;131(11):2471-2477). The TBCRC 011 study (NCT00468715) was the first phase II trial using bicalutamide for AR-positive, estrogen receptor- and progesterone receptor-negative metastatic breast cancer (Clin Cancer Res 2013;19(19):5505-5512). The primary endpoint was CBR. Overall, 26 patients were enrolled, and no CRs or PRs were reported. Nineteen percent had stable disease for at least 6 months. Besides the small sample and lack of objective responses, this trial served as proof of concept for anti-AR treatment in TNBC.
The second-generation AR inhibitor abiraterone acetate, an inhibitor of 17-a-hydroxylase/17,20-lyase (CYP17), was evaluated in a phase II trial of postmenopausal women with estrogen receptor-positive metastatic breast cancer (Ann Oncol 2016;27(1):106-113). A total of 297 patients were randomized: 89 to receive abiraterone and prednisone, 106 to receive abiraterone, prednisone, and exemestane; and 102 to receive exemestane alone.
The primary endpoint was PFS. The median PFS with exemestane alone was 3.7 months, compared with 4.5 months with abiraterone, prednisone, and exemestane (HR, 0.95; p=.795), and 3.7 months with abiraterone plus prednisone (HR, 1.1; p=.437). Trial data showed that, while ORR was higher among those receiving abiraterone, prednisone, and exemestane (22.7%) compared with those receiving exemestane alone (12.7%), these results were not statistically significant (p=.137)
Enzalutamide is a pure AR-signaling inhibitor. Enzalutamide was evaluated in the phase II MDV3100-11 open-label, Simon two-stage study (NCT01889238) in patients with TNBC who were AR-positive (immunohistochemistry [IHC] staining ≥ 10%). The primary endpoint was CBR at 16 weeks (J Clin Oncol 2015;33(suppl):abstr 1003). An androgen-driven signature for diagnosis was created from gene profiling. The trial met its primary endpoint at 16 weeks with a CBR of 35 percent and a median PFS of 14.7 weeks. In terms of objective response, two CRs and five PRs were reported.
This study also highlights the fact that recognizing that AR detected by IHC may not be the ideal predictor of benefit to anti-androgen therapy. Stratifying by diagnosis status in the intention-to-treat population, patients with diagnosis-positive TNBC experienced more favorable outcomes in all response measures than their biomarker-negative counterparts, including on CBR at 16 weeks (39% vs.11%), CBR at 24 weeks or more (36% vs. 6%), and median PFS (16.1 weeks vs. 8.1 weeks). There were several planned clinical trials using enzalutamide in breast cancer. In May 2017, the pharmaceutical company developing enzalutamide announced the discontinuation of trials for breast cancer patients.
Therapies Targeting Pi3K/AKT/mTOR Pathway
Several studies have shown that the Pi3K/AKT/mTOR pathway is activated in TNBC. Moreover, the luminal subgroup phenotype that expresses AR has a high frequency of Pi3K mutations.
Preclinical study demonstrated that activating PiK3CA mutations are enriched in AR-positive TNBC (Breast Cancer Res 2014;16(4):406). In this study, the growth and viability of AR-positive TNBC cell line models were significantly reduced after treatment with Pi3K inhibitors used in combination with an AR antagonist.
We are seeing explosive activity in basic, clinical, and translational research that includes immune checkpoints, platinum salts, targeted agents, and other therapies.
To develop personalized therapies for TNBC, we need a comprehensive understanding of the molecular basis of the oncogenic pathways and microenvironment changes present in TNBC. We also need to grasp the effects of the immune system and therapies on these pathways.
Research shows us that TNBC is not a single disease. Immunotherapy holds much promise, particularly in combination with other therapies, and the subgroups that define patients most likely to benefit from specific therapies, such as AR inhibition and novel ACDs, are becoming increasingly well-defined.
Options for resistant disease may lie in differences in potential therapeutic options and in patient chemosensitivity. Much of the biology of TNBC is now being defined. There is no single target for TNBC, and several promising targeted options are currently being tested in clinical trials.
To improve outcomes in TNBC, we urgently need one or more clinical laboratory improvement amendment-certified biomarkers for patient selection and novel clinical trials designs using smaller patient populations for go/no-go decision-making.
RICARDO H. ALVAREZ, MD, MSC, is Director of Cancer Research, Cancer Treatment Centers of America (CTCA), and Medical Director of the Breast Cancer Center, CTCA at Southeastern Regional Medical Center in Atlanta.