ADC Bystander Killing

Bystander killing is when the cytotoxic payload released from an ADC diffuses out of the targeted, antigen-positive tumor cell and kills neighboring antigen-negative tumor cells that the antibody never bound. It's a key mechanism for overcoming antigen heterogeneity, a major challenge in solid tumors where not every cancer cell expresses the target antigen.

Mechanism

The canonical pathway involves five steps: the ADC binds its target antigen on an antigen-positive (Ag+) cell, undergoes receptor-mediated endocytosis into an early endosome, is trafficked to the lysosome where acidification (pH 4.5–5.0) triggers linker cleavage or full antibody degradation, and the released payload — if lipophilic and membrane-permeable — crosses the lysosomal and cell membranes to diffuse into the extracellular space and enter adjacent antigen-negative (Ag−) cells, killing them too. More recent studies show a second, non-canonical route: extracellular cleavage, where proteases like cathepsin B or cathepsin L cleave the linker outside the cell (in the tumor microenvironment) before or without full internalization, releasing free payload directly into the extracellular space. This was demonstrated for trastuzumab deruxtecan (T-DXd), where cathepsin L cleaved the linker extracellularly and killed HER2-negative cells in a target-independent manner.

What Determines Bystander Activity

Two factors are decisive: linker cleavability and payload physicochemistry.

• Cleavable linkers (peptide/Val-Cit, hydrazone, disulfide) are required to release free, chemically unmodified payload — this is described as "the major binary determinant of bystander activity."

• Non-cleavable linkers (e.g., T-DM1's thioether/SMCC linker) prevent bystander killing almost entirely, because complete lysosomal antibody degradation leaves a permanently charged amino-acid residue attached to the payload, blocking membrane diffusion.

• Payload properties matter just as much: an effective bystander payload must be neutral in charge, lipophilic, and hydrophobic to cross membranes efficiently — MMAE (brentuximab vedotin) and DXd (T-DXd, datopotamab deruxtecan) are classic examples.

Clinical Trade-off: Efficacy vs. Toxicity

Bystander-capable ADCs show meaningfully higher efficacy in antigen-heterogeneous or even antigen-negative tumors — one meta-analysis of 40 trials and 7,879 patients found cleavable-linker ADCs achieved a 29.7% tumor response rate even without target expression, versus a strictly antigen-dependent response for non-cleavable linkers. However, this comes at a real safety cost: the same analysis found a 47% severe toxicity rate for cleavable-linker ADCs versus 34% for non-cleavable ones, because the same "leaky" premature or extracellular cleavage that enables bystander killing also causes off-target payload exposure in healthy tissue. This is precisely why linker cleavability is considered the central design lever balancing potency against systemic toxicity, and why newer "conditionally cleavable" linkers — engineered to activate only within the tumor microenvironment — are an active area of next-generation ADC research.

Clinical Evidence Across Approved ADCs

Four FDA-approved ADCs are recognized for strong bystander activity, each pairing a cleavable linker with a lipophilic payload:

By contrast, trastuzumab emtansine (T-DM1/Kadcyla), which uses a non-cleavable SMCC linker, shows minimal bystander activity and is strictly antigen-dependent, underscoring the linker's decisive role.

Standard Ways to Measure Bystander Effect Experimentally

Two in vitro assay formats are the accepted gold standard for confirming and quantifying bystander killing:

• Co-culture assay: antigen-positive (Ag+) and fluorescently labeled antigen-negative (Ag−) cells are seeded together (e.g., ratios of 1:1, 1:3, 3:1), treated with the ADC, and Ag− cell viability is compared against an Ag− monoculture treated identically; greater Ag− killing in co-culture confirms bystander activity.

• Conditioned-medium transfer assay: Ag+ cells are first treated with ADC, the medium is then removed and transferred onto separate Ag− cells; if the transferred medium kills Ag− cells, it confirms that free payload diffused out and remains active.

Readouts typically use flow cytometry, fluorescence plate readers, high-content imaging, or real-time impedance systems (e.g., Agilent xCELLigence) to quantify Ag− cell viability over 48–120 hours. A widely cited 2016 study formalized this into a quantitative pharmacodynamic (PD) model using a "Bystander Effect Coefficient" (φBE), showing that bystander killing scales with the fraction of Ag+ cells present and with the antigen expression level on those Ag+ cells, and that the effect exhibits a lag time and can diminish as the Ag+ population is depleted.

Emerging Nuance: It's Not Just About Internalization

Newer research complicates the "internalize-then-cleave-then-diffuse" model. Studies show ADCs like T-DXd can undergo extracellular linker cleavage by proteases such as cathepsin L present in the tumor microenvironment, releasing active payload without requiring antibody internalization at all. This means bystander killing can occur through two distinct routes — intracellular processing followed by diffusion, or direct extracellular cleavage — and the relative contribution of each likely varies by ADC and tumor context. This extracellular mechanism has also been linked to immunogenic cell death, where bystander-killed cells release danger signals that further stimulate anti-tumor immunity, adding another layer to the therapeutic benefit beyond direct cytotoxicity.