CD47 Expression Linked to Durable Responses With Zanidatamab/Evorpacept in Pretreated HER2+ Metastatic Breast Cancer

Exploratory biomarker data from a phase 1b/2 trial (NCT05027139) indicated that CD47 expression and circulating tumor DNA (ctDNA) dynamics emerged as potential biomarkers of response to zanidatamab-hrii (Ziihera) plus evorpacept (ALX148) in patients with heavily pretreated HER2-positive metastatic breast cancer, according to findings presented at the 2026 ESMO Breast Cancer Congress.¹

Among 24 evaluable patients, all of whom had received prior fam-trastuzumab deruxtecan-nxki (T-DXd; Enhertu) and a median of 5 prior HER2-targeted therapies (range, 3-7), the confirmed objective response rate (cORR) was 33.3%, and the median progression-free survival (PFS) was 3.6 months (95% CI, 1.7-11.0). In the subset of patients with centrally confirmed HER2-positive disease (n = 10), the cORR increased to 60.0%, and the median PFS was 8.3 months (95% CI, 0.6-not estimable [NE]). Responses were notably durable in the HER2-positive patient cohort, with a median duration of response (DOR) of 20.2 months (95% CI, 5.6-NE).

CD47 expression emerged as a key determinant of response in patients with HER2-positive metastatic breast cancer. Of 17 evaluable patients with CD47 immunohistochemistry data, all 5 patients with HER2-positive disease whose tumors expressed CD47 on at least 20% of tumor cells achieved a complete response (CR) or a partial response (PR), with a median DOR of 20.2 months (95% CI, 5.6-NE). In contrast, only 1 of 4 patients with HER2-positive disease with CD47 expression levels lower than 20% achieved a response, with a median DOR that was NE (95% CI, NE-NE).

The median PFS was 22.1 months (95% CI, 7.4-NE) in patients with HER2-positive disease and CD47 expression levels of at least 20%, compared with 3.4 months (95% CI, 1.5-NE) in those with CD47 expression levels below 20%. CD47 expression levels of at least 20% were observed in tissue samples from 5 of 9 (55.6%) patients with HER2-positive disease and in 1 of 8 (12.5%) patients without HER2-positive disease.

“Zanidatamab plus evorpacept showed promising antitumor activity in patients with heavily pretreated HER2-positive metastatic breast cancer, all of whom had received prior T-DXd; greater antitumor activity was observed in patients with centrally confirmed HER2-positive disease,” lead study author Funda Meric-Bernstam, MD, and coauthors wrote in a poster of the data. “Durable responses in centrally confirmed HER2-positive disease were largely observed in patients with higher CD47 expression, supporting a CD47-dependent HER2-driven biology that resulted in prolonged PFS.”

Meric-Bernstam is the Nellie B. Connally Chair in Breast Cancer and department chair of the Department of Investigational Cancer Therapeutics, as well as medical director of the Department of Khalifa Institute for Personalized Cancer Therapy, a professor in the Department of Breast Surgical Oncology, and a regular member of the Department of Cancer Biology at the University of Texas MD Anderson Cancer Center in Houston.

What was the rationale for evaluating zanidatamab plus evorpacept in patients with HER2-positive metastatic breast cancer?

Zanidatamab is a humanized, IgG1-like, HER2-directed bispecific antibody that promotes HER2 clustering and internalization, inhibits downstream HER2 signaling, and elicits immune-mediated effects, including complement-dependent cytotoxicity, antibody-dependent cellular cytotoxicity, and antibody-dependent cellular phagocytosis (ADCP). Evorpacept is a high-affinity CD47 blocker with an inactive Fc domain designed to enhance ADCP by blocking the CD47-SIRPα interaction and avoiding on-target toxicities observed with conventional CD47 inhibitors. Investigators hypothesized that these 2 agents could act synergistically by combining direct HER2 targeting with phagocytic immune cell activation.

What was the design of the phase 1/2 trial investigating zanidatamab plus evorpacept in HER2-positive breast cancer?

This trial enrolled patients with unresectable, locally advanced and/or metastatic HER2-expressing cancer and an ECOG performance status of 0 or 1. Patients with treated, stable brain metastases were permitted to enroll. Patients with HER2-positive breast cancer had to have received at least 3 prior regimens, including trastuzumab (Herceptin); pertuzumab (Perjeta); and ado-trastuzumab emtansine (Kadcyla), tucatinib (Tukysa), or T-DXd.

Patients in the initial safety cohort received zanidatamab at 1200 mg (if weighing less than 70 kg) or 1600 mg (if weighing at least 70 kg) plus evorpacept at 20 mg/kg (part 1A); or zanidatamab at the same doses plus evorpacept at 30 mg/kg (part 1B) every 2 weeks in 28-day cycles.

This exploratory biomarker analysis includes data from patients in parts 1A and 1B, as well as data from patients in part 2 who received zanidatamab at the same doses above plus evorpacept at 30mg/kg ever 2 weeks in 28-day cycles.

Safety served as the primary end point of part 1. In part 2, the primary end point was cORR, and key secondary end points included PFS, disease control rate (DCR), DOR, and safety. Exploratory end points included biomarkers of response and resistance.

What efficacy findings have been previously reported with zanidatamab plus evorpacept in HER2-positive breast cancer?

Data previously presented at the 2024 San Antonio Breast Cancer Symposium showed that among evaluable patients in part 2 (n = 21), irrespective of HER2 status, the cORR was 33.3% (95% CI, 14.6%-57.0%); all responses were partial responses.² Additionally, the DCR was 71.4% (95% CI, 47.8%-88.7%), the median DOR was NE (range, 3.6-25.9), and the median PFS was 3.6 months (95% CI, 1.8-11.0).

Among patients with HER2-positive disease per central assessment (n = 9), the cORR was 55.6% (95% CI, 21.2%-86.3%). The DCR was 77.8% (95% CI, 40.0%-97.2%), the median DOR was NE (range, 5.6-25.9), and the median PFS was 7.4 months (95% CI, 0.6-NE).

What did ctDNA and ERBB2 copy number dynamics reveal about the efficacy of zanidatamab plus evorpacept in HER2-positive metastatic breast cancer?

In the current biomarker analysis, matched baseline and on-treatment ctDNA samples were available for 14 of 24 patients, of whom 10 had baseline ERBB2 amplification by ctDNA next-generation sequencing (NGS).¹ Among patients with CR or PR, early on-treatment reductions in plasma ERBB2 copy number exceeded 90%, and molecular response (MR) scores, a measure of overall ctDNA change from baseline, also decreased by more than 90%. Patients with stable disease showed variable MR score changes ranging from 5% to 94% decreases, whereas those with progressive disease had MR scores ranging from a 40% decrease to a 267% increase.

Patients who achieved CR or PR had lower alteration burden across RTK/RAS/PI3K signaling pathways at baseline, along with ERBB2 amplification, and minimal co-alterations were observed in EGFR, MET, and ERBB3. Patients with stable or progressive disease had higher alteration burdens, including ERBB2 amplification, as well as co-alterations in EGFR, MET, and ERBB3. Patients in this population were also shown to have PIK3CA mutations (E545K and E542D), NF1 loss, KRAS/MAP3K alterations, and activating ERBB2 L755S kinase domain mutations, a pattern consistent with bypass pathway activation and intrinsic resistance.

Concordance between central tissue fluorescence in situ hybridization (FISH) and plasma ctDNA NGS results for ERBB2 amplification status was 86% in FISH-positive patients (n = 6/7) and 54% in FISH-negative patients (n = 7/13). Discordance in FISH-negative cases may reflect tumor heterogeneity, methodological differences, and sampling timing, with ctDNA providing amplification status immediately prior to treatment initiation.

“Most patients with centrally confirmed HER2-positive disease remained [with] ERBB2-amplified disease after prior T-DXd, supporting continued use of HER2-targeted therapies, including zanidatamab,” the authors concluded. “A biomarker-driven approach incorporating CD47 may optimize patient selection for this combination regimen and warrants further study.”

References

1. Meric-Bernstam F, Wisinski KB, Fang B, et al. Exploratory biomarker analysis from a phase 1b/2 trial of zanidatamab + evorpacept in patients with HER2-positive metastatic breast cancer. Presented at: 2026 ESMO Breast Cancer Congress. May 6-8, 2026; Berlin, Germany. Abstract 72P.
2. Montero AJ, Wisinski KB, Fang B, et al. Zanidatamab in combination with evorpacept in HER2-positive and HER2-low metastatic breast cancer: results from a phase 1b/2 study. Presented at: 2024 San Antonio Breast Cancer Symposium. December 10-13, 2024; San Antonio, Texas. Abstract PS8-09.