Breast Tumors Shed Their HER2 Target in More Than Half of Cases After Drug Therapy

One of the most celebrated targets in modern oncology may be far more slippery than clinicians have assumed. A new study from The University of Texas MD Anderson Cancer Center reports that when women with HER2-positive breast cancer receive HER2-directed drugs before surgery, more than half of the tumors that survive the treatment no longer overexpress the very protein the drugs were designed to attack. The finding, published in Breast Cancer Research and Treatment, raises uncomfortable questions about how adjuvant therapy should be chosen for the substantial minority of patients whose disease does not fully melt away.

HER2, a growth factor receptor encoded by the ERBB2 gene, is amplified or overexpressed in roughly 15 to 20 percent of breast cancers. That molecular signature once heralded an aggressive disease course, but the arrival of trastuzumab, pertuzumab, and a growing arsenal of HER2-targeted agents transformed the prognosis. Today, giving these drugs before surgery—neoadjuvant therapy—is standard practice for stage I to III HER2-positive disease, because it can shrink tumors, sometimes eliminating them entirely, and allows oncologists to gauge response in real time. Yet a meaningful fraction of patients still reach the operating table with residual invasive carcinoma in the breast or lymph nodes.

For those patients, the landmark KATHERINE trial established trastuzumab emtansine (T-DM1) as the standard of care, improving invasive disease-free survival and overall survival compared with trastuzumab alone. More recently, the DESTINY-Breast05 trial showed that trastuzumab deruxtecan (T-DXd) outperformed T-DM1 in patients with residual HER2-positive early disease. Critically, however, both frameworks assumed that a tumor classified as HER2-positive before treatment remained HER2-positive afterward. The MD Anderson study directly tested that assumption—and found it wanting.

The research team, led by pathologist Payu Raval and senior author Aysegul Sahin, retrospectively examined 161 patients treated with HER2-targeted neoadjuvant therapy between 2016 and 2025. Sixty-five of them had enough residual invasive tumor for reliable biomarker reassessment. Using the same clinically validated immunohistochemistry protocol applied in routine care, two study pathologists independently scored HER2 expression in the surgical specimens under the 2018 ASCO/CAP guidelines, with a third pathologist resolving any discordant readings. Every equivocal case—those scored IHC 2+—underwent reflex fluorescence in situ hybridization to determine whether the ERBB2 gene remained amplified.

The results were striking. Thirty-seven of the 65 tumors, or 57 percent, showed a change in HER2 status, shifting from overexpression or amplification before treatment to IHC 0, IHC 1+, or IHC 2+ without gene amplification afterward. Only 28 cases, 43 percent, retained clear HER2 positivity. The pattern depended on where tumors started: among the 33 tumors that were strongly HER2 IHC 3+ at baseline, just under half lost HER2, whereas among the 32 tumors classified as IHC 2+/FISH-amplified, more than two-thirds—68.8 percent—did so. Several of the converted tumors even fell into the emerging HER2-ultralow category, expressing barely detectable levels of the protein.

Perhaps the most sobering aspect of the study is what the researchers could not find. When the team compared the HER2-retained and HER2-changed groups across a battery of clinicopathologic variables—age, clinical stage, histologic subtype, Nottingham grade, hormone receptor status, residual cancer burden, and type of surgery—the two groups were essentially indistinguishable. Hormone receptor positivity was common in both, present in 82.1 percent of the retained group and 89.2 percent of the changed group, and the distribution of residual cancer burden categories was nearly identical. In other words, there is no easy clinical flag that tells oncologists in advance which tumors will shed their target.

Why does this happen? The leading explanation is intratumoral heterogeneity: a single tumor can harbor mixed populations of cells with differing levels of HER2 expression and ERBB2 amplification. Potent HER2 blockade wipes out the dependent cells, while subclones with little or no HER2 survive and repopulate the residual disease—a classic case of clonal selection under therapeutic pressure. Supporting this model, a phase II trial of T-DM1 plus pertuzumab found zero pathologic complete responses in tumors with HER2 heterogeneity, compared with 55 percent in homogeneous tumors. Other work has shown that HER2 heterogeneity is far more prevalent in tumors with low-level amplification, which may explain why the IHC 2+/FISH-amplified group in the new study was so vulnerable to status change. Resistant subclones, genomic studies suggest, often exist before treatment even begins.

The clinical stakes are considerable. Current national and international guidelines recommend adjuvant HER2-targeted therapy for residual disease regardless of post-treatment HER2 status, largely because trials like KATHERINE enrolled patients based on their original diagnosis and never required repeat testing. Whether T-DM1 or T-DXd actually benefits tumors that have lost HER2 remains unknown. There are reasons for cautious optimism: antibody-drug conjugates deliver cytotoxic payloads via the HER2 antibody, so even low levels of the protein may suffice for activity, and the DESTINY-Breast04 trial demonstrated survival benefits of T-DXd in HER2-low metastatic disease. Real-world data from China reported remarkably low recurrence rates among patients who continued HER2-directed therapy despite conversion. But preclinical studies complicate the picture, indicating that HER2-low cells may resist ADCs while remaining susceptible to HER2 kinase inhibitors—hinting that combination strategies may ultimately be needed.

Prognostic evidence is equally unsettled. Several meta-analyses have linked HER2 loss after neoadjuvant therapy to worse recurrence-free and overall survival, with hazard ratios approaching 2, and one registry analysis of more than 21,000 Japanese patients found that 21.4 percent of initially HER2-positive tumors became HER2-negative after treatment. Yet other cohorts, including a surgical series from 2021, found no oncologic penalty for losing HER2 positivity, and some even reported better outcomes with reduced expression. These contradictions likely reflect differences in definitions, therapy intensity, and sampling, but they underscore that complete loss and partial reduction of HER2 may carry different prognostic meanings. Intriguingly, one study found that shorter intervals between therapy and tissue sampling were associated with greater HER2 loss, raising the possibility that some of the change is a reversible pharmacodynamic effect rather than permanent clonal elimination.

The MD Anderson authors are careful about their limitations: the study was retrospective, single-institution, and modest in size, excluded residual tumors smaller than 5 millimeters, and lacked outcome data and molecular profiling. Still, their conclusion is pointed. HER2 should be viewed not as a fixed binary label but as a dynamic continuum reshaped by treatment, and routine reassessment of residual disease deserves serious consideration. Prospective randomized phase II and III trials, they argue, are now needed to determine whether patients whose tumors lose HER2 actually benefit from continuing HER2-targeted therapy—or whether, in the era of antibody-drug conjugates, the biopsy taken after neoadjuvant treatment should become the new starting point for every adjuvant decision.

Subject of Research: HER2 biomarker status change in residual breast carcinoma after neoadjuvant HER2-targeted therapy

Article Title: HER2 status change in residual breast carcinoma after neoadjuvant HER2-targeted therapy in patients with HER2-positive breast cancer

Article References: Raval, P., Ding, Q., Singh, P., Alrohaibani, A., Sun, H., Ai, D., Wanis, K. N., Zhao, M., Chen, H., Valero, V., Chavez-MacGregor, M., & Sahin, A. (2026). HER2 status change in residual breast carcinoma after neoadjuvant HER2-targeted therapy in patients with HER2-positive breast cancer. Breast Cancer Research and Treatment, 219(3), Article 25. https://doi.org/10.1007/s10549-026-08088-z

Image Credits: AI Generated

DOI: 10.1007/s10549-026-08088-z

Keywords: HER2-positive breast cancer, neoadjuvant therapy, HER2 status change, biomarker discordance, immunohistochemistry, FISH, trastuzumab, pertuzumab, antibody-drug conjugates, intratumoral heterogeneity, residual disease, adjuvant therapy