In an OncLive Peer Exchange® discussion, experts in breast malignancies reviewed some exciting new developments in TNBC.
Adam M. Brufsky, MD, PhD,
Adam M. Brufsky, MD, PhD
Approximately 1 million breast cancer cases are diagnosed annually worldwide, with about 17% (n = 170,000) being triple-negative breast cancer (TNBC).1 Most TNBCs, about 75%, have been considered “basal-like,” meaning their expression signature has been shown to be comparable to that of the basal/myoepithelial cells of the breast, with transcriptomic characteristics similar to tumors arising in BRCA1 germline mutation carriers;2 however, more recently, genomic analysis has shown TNBCs to have substantial transcriptional, mutational, and copy number heterogeneity, indicating these tumors are significantly more heterogeneous than previously thought.3,4 Such findings have helped spur new developments in treating patients with TNBC, a cancer field that has seen relatively few meaningful clinical advancements since the introduction of adjuvant taxane therapy approximately 20 years ago.4
TNBC Heterogeneity and BRCA Testing
In an OncLive Peer Exchange® discussion led by Adam M. Brufsky, MD, PhD, experts in breast malignancies reviewed some exciting new developments in TNBC. The discussion focused on BRCA testing and other predictive biomarkers and on the promise of 2 novel treatment approaches currently under investigation for TNBC: immunotherapy and anti-body—drug conjugates.The complexity of TNBCs is increasingly being understood as gene expression profiling is revealing more subtypes. “We’re learning over the last couple of years that these really are likely a collection of multiple different diseases,” said Brufsky. The most recent classification, which revised a widely accepted earlier classification by the same investigators, breaks them into 4 tumor-specific subtypes: 2 basal-like (BL1 and BL2), a mesenchymal (M), and a luminal androgen receptor (LAR) subtype (Table 1).5 When using these subtypes to assess response to neoadjuvant chemotherapy, significant differences between patient groups have been observed, with a pathological complete response (pCR) found in 41% of BL1 patients, 29% of LAR patients, and 18% of BL2 patients.5 The relevance of these subtypes has yet to be elucidated in prospective studies, and these subtypes have the potential to guide patient selection with regard to some of the newer treatments that are being investigated in clinical trials, including immunotherapy and antibody—drug conjugates.4
BL indicates basal-like; LAR, luminal androgen receptor; M, mesenchymal.
Currently, BRCA mutation status, which has been associated with an increased response to platinum agents and poly (ADP-ribose) polymerase (PARP) inhibitors, is the only biomarker that is being used in routine clinical practice to guide treatment. The NCCN Guidelines recommend BRCA mutation screening in any patient who develops TNBC at ≤60 years, regardless of family history or other risk factors.6 This is partially because BRCA status has been strongly correlated with TNBC, with 80% of BRCA1 mutation carriers and 50% of BRCA2 mutation carriers developing TNBC. However, the likelihood of BRCA mutations decreases with advancing age (present in 43.8% of those <40 years vs 16.6% of those >70 years).4,7 Although some panelists found this recommendation “pretty reasonable,” it doesn’t fully clarify the testing of patients over 60 years old with TNBC, which was a point of discussion during the Peer Exchange®.
“In an Ashkenazi woman, you could potentially go up even older,” said Mark E. Robson, MD, who said he was liberal about testing but acknowledged it might not be possible to test older non-Jewish women who are on Medicare. The increased leeway with testing Ashkenazi Jewish women is rooted in their significantly increased risk of TNBC and BRCA mutations compared with other populations. In one study, 50% of Ashkenazi Jewish women were found to have BRCA1 mutations.7 In contrast, BRCA mutation prevalence in African-American women, another TNBC high-risk group, was less than half that at 20.4%.7 Panelist Aditya Bardia, MBBS, MPH, appeared to be less liberal with BRCA testing. “If there’s a clinical indication to do BRCA testing or an indication in terms of a clinical trial, we would consider that, but we don’t do it for everyone with TNBC,” he said. Kimberly L. Blackwell, MD, said she would test regardless of age in the first line to ensure that opportunity to participate in a clinical trial is not missed.
Immunotherapy for TNBC
In clinical practice, the results of BRCA testing are used to guide treatment, often directing which patients go on to receive immediate platinum-based chemotherapy, but Blackwell said she felt this common practice ultimately limits patients’ options. “What I’m seeing in my referral patterns is that for people who have TNBC, in particular BRCA mutations, there’s a rush to put them on a platinum-based chemotherapy and then send them to a place for trial, but I’m really trying to encourage referring physicians to think twice about the BRCA mutation carriers so that they get these patients enrolled in a study before they use a platinum-based agent. For the most part, I’d say that over half of the BRCA-mutation randomized studies require patients to be platinum naïve,” she said. She prefers to reserve platinum-based agents for the second- and third-line settings whenever possible.Patients with the M subtype, which has a high prevalence of immune-infiltrated lymphocytes, have been thought to be especially primed to benefit from programmed death-ligand 1 (PD-L1) inhibition, and patients with TNBC are quickly being enrolled in checkpoint inhibitor trials, regardless of subtype, because predictive biomarkers have yet to be identified in this setting. “I’m encouraging my patients to participate in both the KEYNOTE studies and some of the single-arm phase I/II studies where the checkpoint inhibitors are being compared,” said Blackwell, who noted many checkpoint inhibitor trials are available and enrolling patients (Table 2).
aTrial is ongoing but not recruiting participants. ER+ indicates estrogen receptor positive; PD-L1, programmed death-ligand 1; TNBC, triple-negative breast cancer.
“There’s definitely a signal in a population of patients, but I don’t think we’ve figured out who those patients are, so, unfortunately, we’ve gone back to old-school clinical trial design, which enrolls a lot of people to standard of care versus the immuno-oncology agent,” said Blackwell, noting she remains excited but somewhat guarded about the promise of checkpoint inhibition in TNBC.
Another factor that has been thought to make TNBCs more suitable for immunotherapy when compared with other breast cancer subtypes is the high mutational load of these tumors. In neoadjuvant studies, immune-cell infiltrates have been correlated with responses and long-term outcomes, lending evidence that TNBCs may be more immunogenic.8,9 But the panelists noted that the mutational picture of these tumors is not fully in line with the characteristics that have previously been tied to improved response. “If you look at the types of mutations that are in the diseases that are most associated with response, they tend to be point mutations and small [insertions or deletions], and so they’re generating abnormal proteins, which can be recognized, whereas in the homologous recombination deficient tumors [including BRCA defects], you’ve got big gains and losses [of large chromosomal regions],” said Robson.
Antibody—Drug Conjugates for TNBC
Bardia concurred with the other panelists’ cautious optimism. “There’s definitely excitement about immunotherapy, and it works for a subset of patients, but the challenge is we haven’t found that subset,” he said. “The first step is obviously to look at PD-L1, but that might not be the answer. It could be mutation burden of the tumor, and it could be other antigens that are expressed by the tumor, so what the field needs is predictive biomarkers that can show who is going to bene t from immunotherapy.”Like immunotherapy, antibody—drug conjugates represent another class of medications that have the potential to change the TNBC treatment land- scape. Development of these agents for TNBC was inspired by the success of trastuzumab conjugated to emtansine (T-DM1) in HER2-positive breast cancer. “The idea is that you have an antibody that binds to the cancer cells and then releases chemo only to the cancer cells while sparing the normal cells, which has a better risk pro le by enabling very high doses of chemotherapy to be delivered only to cancer cells,” Bardia said.
An antibody—drug conjugate being developed for TNBC is sacituzumab govitecan (IMMU-132), which binds to tumor-associated calcium signal transducer 2 (Trop-2), a cell-surface glycoprotein that is overexpressed in approximately 85% of TNBCs. “We observed an impressive objective response rate of about 30% and a median progression-free survival [PFS] of about 6 months in patients who had been heavily pretreated,” Bardia said. The patients in the study had received a median of 5 prior lines of therapy. The median overall survival was 16.6 months; 2 patients had complete responses, and tumor shrinkage was observed in 1 patient who had disease progression while on checkpoint inhibitors.10 “In that heavily pretreated population, it was quite encouraging to see the signal,” Bardia said. Based on the results, the FDA granted sacituzumab govitecan a breakthrough therapy designation. Other trials are planned but are not yet recruiting: sacituzumab govitecan with or without carboplatin (NCT02161679) and sacituzumab govitecan for refractory/relapsed TNBC (NCT02574455).
Other antibody—drug conjugates in development for TNBC that were not discussed during the Peer Exchange® include SGN-LIV1A, an LIV-1-targeted monoclonal antibody linked to the cytotoxin monomethyl auristatin E (MMAE) by a protease-cleavable linker, and glembatumumab vedotin (CDX-011), which targets transmembrane glycoprotein NMB and is also conjugated to MMAE.11,12 During the recent 2016 San Antonio Breast Cancer Symposium, study investigators reported interim data from an ongoing phase I clinical trial of SGN-LIV1A, which revealed a partial response in 37% of patients (n = 11), a disease control rate of 67%, and a clinical bene t rate of 47%.11 The median PFS was 12 weeks, but 7 patients remained on treatment.11 All patients were heavily pretreated. A safety study of SGN-LIV1A is currently recruiting (NCT01969643).
The results for glembatumumab vedotin appear promising in individuals with tumors that have a high level of glycoprotein NMB expression. In the randomized phase II EMERGE study, the overall objective response rate was only 6% versus 7% for investigator-selected chemotherapy; however, response rates reached 30% in patients with glycoprotein NMB expression in ≥25% of tumor cells.12 Response rates were also significantly improved in patients with TNBC, reaching 18% in this population overall, and 40% when considering those with glycoprotein NMB-overexpressing TNBC. Glembatumumab vedotin is being assessed for metastatic TNBC in the pivotal METRIC study (NCT01997333).The treatment landscape of TNBC is finally evolving as these tumors and their complex genomic profiles are becoming better understood, Bardia concluded. Clinical trials are essential for these patients, including in the metastatic setting, and should not be overlooked, Blackwell said. “All of us who have busy practices dread the rigmarole that’s involved in putting people on trial, but it’s so worth it to give these options to our patients facing meta- static disease,” she said.
In addition to finding important new treatments and giving hope to patients with advanced disease, clinical trials are helping to find predictive biomarkers that can improve treatment selection for future patients. As a clearer picture of TNBC emerges, treatment sequencing will become an increasingly important piece of the puzzle. “We need to start thinking about the deepest possible characterization of patients, both from a genomic level and potentially from a germline level, so that we can figure out which is the best way of putting these things together for each individual patient,” Robson concluded.