In 69 Women With Metastatic Uterine Cancer, a Breast Cancer Drug Finds New Purpose

A genomic atlas of HER2-positive metastatic endometrial cancer maps who carries the target, who gets the therapy, and whether it changes survival.

Every year, roughly 66,000 American women are diagnosed with endometrial cancer — cancer of the uterine lining. Most cases are caught early and cured with surgery. But a fraction are not. They arrive late, or they recur aggressively, seeding metastases in the lungs, liver, and peritoneum. For these women, treatment options narrow fast.

A molecular detail offers one route forward. About 20 to 30 percent of serous endometrial cancers — the most lethal subtype — overexpress a protein called HER2, the same target that transformed breast cancer treatment two decades ago. The question is straightforward: can trastuzumab, the antibody engineered to block HER2 in breast tumors, do the same in the uterus? Memorial Sloan Kettering profiled 69 women with metastatic endometrial cancer whose tumors were selected for HER2/ERBB2 status, sequenced them with MSK-IMPACT targeted panels, and tracked their outcomes. The resulting atlas — part genomic portrait, part clinical ledger — maps the terrain of HER2-positive metastatic endometrial cancer with unusual granularity.

Who Is in the Room

The cohort is small and deliberately constructed. All 69 patients are women with metastatic endometrial carcinoma whose tumors were flagged for HER2/ERBB2 — either by genomic amplification of the ERBB2 gene or by protein overexpression on immunohistochemistry (IHC). The first question is how many actually carry the genomic amplification that defines the target. In a cohort selected by HER2 criteria, you would expect the amplified bar to dominate. But some patients may have entered via equivocal IHC (a score of 2+) or discordant testing, meaning their genomic profiles tell a more complicated story.

The Biology Points to Serous Histology

HER2 amplification is not distributed evenly across endometrial cancer subtypes. It concentrates in serous carcinoma — the high-grade, copy-number-driven subtype that accounts for a disproportionate share of endometrial cancer deaths. That pattern holds here. The histological breakdown shows serous tumors as the dominant group, with smaller contributions from endometrioid, carcinosarcoma, and other rare subtypes.

This matters because HER2-positive endometrioid tumors may represent a different molecular pathway — one that overlaps with the copy-number-high, TP53-abnormal (CN-H/TP53abn) molecular subtype even though endometrioid tumors are more commonly associated with lower-grade, hormone-driven biology. Since all patients carry metastatic disease, the stage distribution skews toward stage III and IV at initial diagnosis, though a subset originally presented with early-stage disease and later progressed.

Not Everyone Gets the Drug

Having the target does not guarantee receiving the therapy. Among the 69 women profiled, only a fraction received HER2-directed treatment — predominantly trastuzumab, the monoclonal antibody that has been the backbone of HER2-targeted therapy since its approval for breast cancer in 1998. The treated-versus-untreated split is critical, because it determines whether the survival comparisons that follow have enough statistical power to detect a real difference.

The reasons some patients did not receive HER2 therapy are not captured in the genomic data, but they likely include the timing of diagnosis (before HER2-directed therapy was considered in endometrial cancer), contraindications, or clinical judgment that other treatments took priority. The specific agents used skew heavily toward trastuzumab, with limited representation of newer drugs like trastuzumab deruxtecan (T-DXd) or pertuzumab.

Having the target does not guarantee receiving the therapy. Among the 69 women profiled, only a fraction received HER2-directed treatment.

The Mutational Landscape: Low Burden, High Copy Numbers

HER2-amplified endometrial cancers are, at their core, copy-number-driven tumors. They accumulate large-scale chromosomal gains and losses rather than piling up point mutations. The tumor mutational burden (TMB) distribution across these 69 specimens confirms that pattern: most tumors cluster in the low-to-moderate range, reflecting the CN-high molecular subtype.

But there is a tail. A small number of specimens show markedly elevated TMB, and those outliers overlap almost perfectly with tumors classified as microsatellite instability–high (MSI-H). MSI-H tumors have defective DNA mismatch repair, which causes mutations to accumulate at microsatellite repeats throughout the genome. These tumors sit at an unusual intersection: they carry HER2 amplification and the high mutational burden that makes them candidates for immune checkpoint inhibitors like pembrolizumab. The TMB-versus-MSI scatter plot makes this separation vivid — a dense cluster of low-TMB, microsatellite-stable tumors, and a few specimens escaping toward the upper-right corner.

The Oncoprint: TP53 and ERBB2 as Twin Pillars

The gene-by-patient oncoprint — a matrix showing which genes are altered in each patient and how — reveals the architecture of these tumors. Two features dominate. First, TP53 mutations are near-universal, consistent with the CN-high/TP53-abnormal molecular subtype that defines most serous endometrial cancers. Second, ERBB2 amplification runs as a dark band across the cohort, confirming the selection criterion.

Beyond those two pillars, PIK3CA mutations appear in a meaningful fraction of patients. This is clinically relevant: the PI3K pathway sits downstream of HER2 signaling, and co-occurring PIK3CA mutations have been implicated in resistance to trastuzumab in breast cancer. Whether the same resistance mechanism operates in endometrial cancer is an open question, but the co-occurrence pattern here suggests it warrants testing. Conversely, mutations in KRAS and CTNNB1 — hallmarks of the endometrioid, hormone-driven molecular subtype — are largely absent, reinforcing the boundary between copy-number-driven and mutation-driven endometrial cancers.

Does the IHC Match the Genome?

One of the most practically important questions in HER2-driven oncology is whether the two standard ways of measuring HER2 status — genomic amplification of the ERBB2 gene and protein overexpression by immunohistochemistry — agree. Discordance between the two tests creates clinical uncertainty: a patient whose tumor is amplified but scores 0 or 1+ on IHC might not be offered trastuzumab, even though her genomic profile suggests she could benefit.

The concordance panel cross-tabulates the two measures. High agreement — amplified tumors scoring 3+ on IHC, non-amplified tumors scoring 0 or 1+ — validates both assays and supports using either for treatment decisions. Discordant cases, however, deserve scrutiny. They may reflect tumor heterogeneity (where one part of the tumor overexpresses the protein and another does not), fixation artifacts that degrade IHC staining, or genuinely borderline biology.

Sixty-nine patients is not enough to rewrite treatment guidelines. But it is enough to sketch a map. The terrain of HER2-positive metastatic endometrial cancer turns out to be more structured than it first appears — dominated by serous histology, TP53 loss, and copy-number-driven biology, with a small but distinct MSI-high enclave that bridges two therapeutic worlds. The concordance between genomic amplification and protein expression is reassuringly high but not perfect, and imperfect concordance is exactly where clinical decisions get hard. The survival data hint at a benefit from trastuzumab, but observational data can only hint. What this atlas does best is identify the patients who sit at the edges — the ones whose tumors are MSI-high and HER2-amplified, the ones whose IHC and genomics disagree, the ones who carry PIK3CA co-mutations that might blunt a drug's effect. Those edge cases are where the next clinical trials should focus, and where 69 carefully profiled patients become worth far more than their number suggests.