Target expression versus ADC response

In one sentence

For antibody-drug conjugates, the amount of target protein measured in a tumor sample predicts benefit only weakly, because delivery, payload potency, spread to neighboring cells and the limits of the test all sit between expression and response.

The intuition

A mail carrier needs a mailbox to deliver a parcel, but one small mailbox is enough if the parcel holds something powerful. Counting mailboxes on a street tells you whether delivery is possible, not whether the parcel will do its job once inside. If parcels can also leak into the house next door, even houses without a mailbox are affected. The analogy breaks because tumors are not streets of identical houses: a biopsy counts mailboxes on one block, at one time.

How it works

An antibody-drug conjugate (ADC) response depends on a chain of steps. Target expression measures only the first:

Immunohistochemistry sees the first box; everything after it can raise or lower response independently.

Five reasons the link is loose:

1. A little target can be enough. Payloads are potent and some ADCs carry about eight per antibody, so modest uptake can deliver a lethal dose.
2. Payload can spread. Membrane-permeable payloads kill nearby cells that lack the target (the bystander effect), so target-negative patches are not necessarily spared.
3. The test is built for a different job. HER2 immunohistochemistry was designed to find very high HER2 (score 3+). At the low end it is unreliable: when 18 pathologists scored the same 170 breast biopsies, they agreed on 0 versus 1+ only 26% of the time, versus 58% for 2+ versus 3+ [1].
4. Tumors vary in space and time. One core, often from an earlier biopsy, may not represent every metastasis now; see tumor heterogeneity.
5. Resistance can sit after delivery. A tumor full of target can still pump out or tolerate the payload; see cross-resistance between ADCs.

What the trials show

Two lessons follow. There is a gradient: more target goes with more response on average (DAISY, ASCENT). There is no clean floor: benefit or response appeared in the lowest groups tested, and DAISY found responses even in HER2-zero tumors, where it also saw that HER2 staining and drug uptake went together in a few biopsies [4]. The labels reflect this. Sacituzumab govitecan and datopotamab deruxtecan are used in triple-negative disease without a TROP2 test, while trastuzumab deruxtecan requires an authorized HER2 test to place a tumor in the low or ultralow category [6].

Why it matters in cancer

• "Low" stopped meaning "negative." HER2-low and HER2-ultralow are treatment-eligibility categories created by these trials. They are defined by a test score; whether they form a distinct biological group is debated. See HER2.
• Retesting can change eligibility. Because 0 versus 1+ is hard to read, a second biopsy or a second reading can move a tumor across the line.
• New ADCs need their own evidence. A gradient seen for one drug does not set a cutoff for another; that requires a validated companion diagnostic.

Common confusions

• Weakly predictive is not irrelevant. On average, more target goes with more response; the measurement just cannot sort individuals well.
• No test required does not mean the target does not matter. It means trials did not find a group in which the drug failed to help.
• Prognostic is not predictive. A marker can track outcome with any treatment without telling you who benefits from this one; see prognostic versus predictive.
• A response rate in a single-arm cohort is not a comparison. DAISY shows activity across HER2 levels but has no control arm.

How it is measured

HER2 is scored by immunohistochemistry on a 0 to 3+ scale, with in situ hybridization for gene amplification at 2+. HER2-ultralow means IHC 0 with some membrane staining, as defined in DESTINY-Breast06 [3]. TROP2 has been scored in research by H-score, which combines staining intensity and the percentage of stained cells. Computer-based scoring of membrane staining is a research method for some ADCs, not a validated selection test for breast-cancer ADCs.

In Diana's case

How her tumor's HER2 and TROP2 results bear on ADC options is owned by the ADC therapies page.

Related concepts

• HER2
• Immunohistochemistry
• Companion diagnostics
• ADC conjugation chemistry
• Prognostic versus predictive

Sources and scope

Source check: November 7, 2026. Randomized and single-arm trial results limited to their populations; subgroup and biomarker analyses are exploratory. Expert and learner review remain pending.

References

1. Fernandez et al., 2022 — concordance of HER2 immunohistochemistry scoring in the low range.
2. Modi et al., 2022: DESTINY-Breast04 — trastuzumab deruxtecan in previously treated HER2-low metastatic breast cancer.
3. Bardia et al., 2024: DESTINY-Breast06 — trastuzumab deruxtecan in HER2-low and HER2-ultralow hormone-receptor-positive metastatic breast cancer.
4. Mosele et al., 2023: DAISY — trastuzumab deruxtecan across HER2 expression levels.
5. Bardia et al., 2021: ASCENT biomarker analyses — sacituzumab govitecan outcomes by TROP2 expression.
6. US prescribing information (DailyMed), section 1: trastuzumab deruxtecan, sacituzumab govitecan, datopotamab deruxtecan.

Used in

• Compare related ideas
• Match the payload to the failure
• Aiming a drug at a surface target