Trial Evaluates Antibody-Drug Conjugate in Advanced Triple-Negative Breast Cancer

An ongoing randomized trial in triple-negative breast cancer (TNBC) is evaluating the safety and efficacy of an antibody-drug conjugate that targets glycoprotein NMB (gpNMB), which is associated with reduced metastasis-free and overall survival.

Denise A. Yardley, MD

An ongoing randomized trial in triple-negative breast cancer (TNBC) is evaluating the safety and efficacy of an antibody-drug conjugate that targets glycoprotein NMB (gpNMB), which is associated with reduced metastasis-free and overall survival.

A previous phase II trial showed a significant correlation between gpNMB expression and response to glembatumumab vedotin (GV) but not comparator regimens. The phase III METRIC trial is comparing GV and capecitabine in 300 women who have gpNMB-positive TNBC.

“Patients are selected based on the level of gpNMB expression in tumors obtained in the advanced [locally advanced, recurrent, or metastatic] setting,” according to principal investigator Denise Yardley, MD, a medical oncologist at the Sarah Cannon Cancer Research Institute in Nashville, Tennessee. “Patients with overexpression of gpNMB—greater than or equal to 25% of malignant epithelial cells [which is] determined by central laboratory immunohistochemistry [IHC] methods—are eligible for the trial.”

An accrual goal of 300 patients is expected to continue into 2016, she added. The trial has a primary endpoint of progression-free survival (PFS), and final data analysis will occur after 203 PFS events.

About 40% of TNBC, and 15% to 20% of breast cancers overall, exhibit overexpression of the internalizable transmembrane gpNMB. Overexpression of the protein has been associated with reduced survival in breast cancer, small-cell lung cancer, and glioblastoma.

GV consists of glembatumumab, a monoclonal antibody that binds gpNMB and delivers the cytotoxin monomethyl auristatin E to gpNMB-expressing cells. Three clinical trials of GV were recently completed, two in breast cancer and one in melanoma, and patients with high-level gpNMB expression appeared to derive the greatest benefit from GV.

In the phase II EMERGE trial, GV was evaluated in a treatment refractory/resistant population of patients with advanced breast cancer (J Clin Oncol. 2015;33:1609-1619). Yardley et al randomized 124 patients 2:1 to GV or investigator’s choice, and the trial design allowed crossover to GV at disease progression.

Almost all of the patients had metastatic disease at enrollment, and they had received a median of six prior lines of breast cancer therapy. The primary endpoint was overall response rate.

The results showed no difference in overall response in the analysis of all patients. However, prespecified analyses did show differences by gpNMB expression. In the subgroup of patients with high gpNMB expression, the response rate was 30% (7 of 23) for those with GV versus 9% (1 of 11) for patients randomized to investigator’s choice. Additionally, four of ten patients with TNBC and high gpNMB expression had objective responses with GV versus none of the six patients assigned to the control group.

In the gpNMB-overexpression subgroup, 65% of patients had stable disease or better response with GV versus 27% in the control group, including 90% versus 17% in the triple-negative subgroup. Median PFS in the gpNMB-overexpression favored the GV subgroup (2.8 vs 1.5 months), and the difference achieved statistical significance in patients with TNBC (3.5 vs 1.5 months,P= .0017).

Median overall survival was 10.0 months in the gpNMB-overexpression subgroup and the patients with triple-negative disease compared with 5.7 and 5.5 months with investigator’s choice. The difference in the triple-negative subgroup achieved statistical significance as well (P= .003).

The most common treatment-related toxicities were rash, neutropenia, fatigue, nausea, vomiting, alopecia, decreased appetite, and peripheral neuropathy. Less hematologic toxicity occurred with GV than with investigator’s choice regimens.

The ongoing METRIC trial has a primary endpoint of PFS, determined by independent central review. Eligibility criteria include, TNBC, gpNMB overexpression by centrally reviewed IHC, no more than two prior chemotherapy regimens, prior taxane therapy, and prior anthracycline therapy, unless contraindicated. The trial is being conducted at almost 100 sites in the United States, Canada, and Australia.

Information on the current status of the trial is available at

“Interest in therapeutic application of GV goes beyond breast cancer,” said Yardley. The agent demonstrated promising activity in advanced melanoma in a phase I/II trial, and additional trials have begun. Planned studies include squamous-cell lung cancer, osteosarcoma, uveal melanoma, and pediatric sarcomas, all of which exhibit high expression of gpNMB. Preliminary research has suggested high-level expression of the glycoprotein by pancreatic and head and neck cancers.

Biologic plausibility already exists for using GV in combination with other targeted therapies, according to Yardley.

“In vitro studies have shown that blockade of the mitogen-activated protein kinase pathway using RAF, MEK, or ERK inhibition leads to enhanced gpNMB expression in BRAF- and NRAS-mutated melanoma cells, and pretreatment with a MEK inhibitor sensitizes melanoma cell lines to glembatumumab vedotin,” said Yardley. “These results suggest MAPK pathway inhibition may be synergistic with glembatumumab vedotin, particularly in tumors with low baseline gpNMB expression levels.”

“A rationale may exist for evaluating GV in combination with immune-checkpoint inhibitors,” she continued. Tumor cells that express gpNMB may negatively regulate immune responses, and GV-induced tumor killing may liberate tumor antigens that could augment checkpoint inhibitor-mediated antitumor immune responses. This has been demonstrated with chemotherapy and radiation therapy.

“However, these combinations will need to be approached with caution, given the potential for overlapping cutaneous and GI toxicities,” said Yardley.