T-DM1 Shows Strongest Activity in Salivary Gland Cancer in HER2-Amplified Basket Trial

Why This Study Matters

HER2 (ERBB2) amplification is best known as a targetable driver in breast and gastric cancer, but it also occurs in a smaller subset of other solid tumors — including salivary gland, lung, colorectal, and endometrial cancers. Ado-trastuzumab emtansine (T-DM1), an antibody-drug conjugate, is FDA-approved for HER2-positive breast cancer, but it carries no approved indication in these other HER2-amplified tumor types. That leaves oncologists treating HER2-amplified disease outside of breast and gastric cancer without a standard, HER2-directed option.

This single-center, Memorial Sloan Kettering-led basket trial tested T-DM1 across five disease-specific cohorts of HER2-amplified cancers, asking a simple but clinically important question: does a HER2-amplified biomarker predict benefit from a HER2-targeted drug regardless of the tumor’s tissue of origin — or does response depend heavily on histology?

Study Design

This was a single-center, non-randomized, open-label Phase 2 basket trial conducted entirely at Memorial Sloan Kettering Cancer Center. There was no comparator arm — all patients received intravenous T-DM1 at 3.6 mg/kg every 21 days. HER2 amplification was confirmed by next-generation sequencing or in-situ hybridization prior to enrollment. The primary endpoint was overall response rate (ORR) by investigator assessment; overall survival (OS), progression-free survival (PFS), duration of response (DOR), and safety were secondary endpoints. The randomization scheme was not applicable, since this was a non-randomized, single-arm basket trial, and the number of study sites or countries beyond the single MSK site was not specified in the source.

Patient Population

All 95 patients had advanced or metastatic HER2-amplified solid tumors and were enrolled into one of five disease-specific cohorts: salivary gland cancer (n=19), lung cancer (n=23), colorectal cancer (n=7), endometrial cancer (n=23), and an "other" histology cohort (n=23), based on the denominators reported for each cohort’s response rate. Detailed eligibility criteria, such as prior lines of therapy or ECOG performance status thresholds, were not reported in the abstract and are not specified in the source.

Primary Endpoint: Overall Response Rate

Across all 95 treated patients, 22 achieved a confirmed response by investigator assessment, for an overall response rate of 23% (95% CI, 16–33%). Median progression-free survival across the full population was 3.6 months (95% CI, 2.6–5.4), and median duration of response was 13.1 months (95% CI, 7.3–30.5).

Subgroup (Cohort) Analyses

Response rates varied substantially by tumor type. All figures below come directly from the published abstract — no subgroup statistic in this section has been estimated, interpolated, or sourced from anywhere other than the abstract itself.

Across cohorts, median PFS ranged from 1.4 to 10.2 months, median OS from 7.8 to 29.2 months, and median DOR from 6.4 to 18.4 months, with the best/worst cohort not separately itemized per metric beyond these ranges. Salivary gland cancer had the longest median PFS (10.2 months) and longest median OS (29.2 months) of any cohort; the "other" cohort had the shortest median PFS (1.4 months) and shortest median OS (7.8 months). The lung cohort had the shortest median DOR (6.4 months) and the "other" cohort the longest (18.4 months).

Safety Profile

Safety was a secondary endpoint of this trial, but the published abstract does not report specific adverse event types, grades, or rates by cohort or overall — this run relied on the abstract only, since no PMC full text was available for this "online ahead of print" record at the time of writing. Readers seeking safety data should consult the full-text article once available or the trial’s ClinicalTrials.gov results record.

Interpretation and Broader Context

The central finding of this basket trial is heterogeneity: a shared molecular feature (HER2 amplification) did not translate into a uniform treatment effect across tumor types. Salivary gland cancer — a rare, historically undertreated malignancy with no FDA-approved HER2-targeted therapy — showed by far the strongest signal, with a 58% response rate and a median overall survival approaching two and a half years. By contrast, the colorectal cohort saw no confirmed responses at all.

T-DM1 demonstrated heterogenous efficacy among HER2-amplified tumor types, with promising response and outcomes among patients with salivary gland cancer.
— Study authors, Conclusions section, Ross et al., Clinical Cancer Research (2026)

For clinicians, this argues against treating "HER2-amplified" as a single actionable category that automatically predicts benefit from a HER2-targeted antibody-drug conjugate. Instead, it supports a disease-specific approach to development — and it strengthens the case for salivary gland cancer as a priority indication for further T-DM1 study, given the absence of any FDA-approved HER2-directed option in that disease today.

Limitations

This was a single-center, non-randomized, open-label basket trial with no comparator arm, so treatment effects cannot be separated from patient selection or disease course at a single referral center. Several cohorts are small — the colorectal cohort in particular enrolled only 7 patients, giving a 95% confidence interval spanning 0% to 35% around its 0% observed response rate, too wide to conclude T-DM1 has no activity there. The published abstract did not report detailed safety data by cohort or overall, and full-text peer-reviewed results were not yet available in PubMed Central at the time of writing, since the article was listed as "online ahead of print." Detailed eligibility criteria and randomization details beyond what is stated above were also not specified in the source.