You’ve had food stuck in your teeth at some point during a conversation you thought was going well. Or toilet paper trailing from the back of your shoe on a day you felt particularly put together. Maybe you’ve even forgotten to color in your very blonde eyebrows and spent the rest of the day looking like someone erased the top half of your face. But you had no idea. You’re walking around with the quiet, complete confidence of someone operating on incomplete information.
You weren’t wrong about anything you could see. You just couldn’t see everything, which is a different problem than getting a bad result. A bad result tells you something is wrong, but this doesn’t. This happens in the lab too. The number is clean, the program is advancing, and somewhere in the data, something is happening that your readout has no way to show you.
ADC development has a version of this problem. The cytotoxicity readout is real, reliable, and correct. It’s also an aggregate, and an aggregate compresses everything that happened into a single number. That number can’t tell you which mechanisms produced it, which ones are underperforming, or what to change if the program stops working. It just tells you cells died, or they didn’t. You’re walking around with the quiet, complete confidence of someone operating on incomplete information.
Cytotoxicity: Where Every Program Begins
Cell viability assays were built to answer one question: did the cells die? Increase the dose, more cells die. Decrease it, fewer do. The curve is clean, the data is reliable, and the payload is doing what it was designed to do. For cytotoxicity, it’s the right question to ask.
That question has an established platform with consistent, reproducible data: our CellTiter-Glo® and RealTime-Glo™ Assays.
For most ADC programs, this is where the measurement work starts and stops, but it should only be where it starts. The question they answer well is only one of several your ADC is raising. The number means what it says, but the question it answers has a boundary, and the boundary leaves you with an incomplete picture.
Internalization: The Step That Already Happened
Before the payload can kill anything, it has to get inside the cell. By the time your cytotoxicity assay is measuring results, that step has already shaped them, whether or not it was measured.
A candidate can look unremarkable in a cytotoxicity readout for any number of reasons. Maybe the payload isn’t potent enough. Maybe the linker isn’t releasing efficiently. Or maybe internalization was slow and the payload never reached the intracellular space in sufficient quantity. The cytotoxicity assay can’t distinguish between those scenarios. It just returns a number, and the number doesn’t come with an explanation.
Internalization rate is one of those explanations. It determines how much payload reaches the intracellular space, and the rate can vary significantly across candidates that look equally promising on the surface.
Researchers screening large panels need to know which antibodies are actually crossing the membrane. Our Antibody Internalization Bioassay identifies which candidates are internalizing early in screening. pHAb Reactive Dyes track how internalization unfolds in real time.
Internalization rate was already shaping the cell death outcome before the cytotoxicity assay started. Two candidates that look identical in the readout may have gotten there very differently, leaving you no way to know what’s actually driving the result.
Fc Effector Function: The Job Nobody Checked
Before anyone attached a payload to it, the antibody in your ADC was already doing another job: recruiting immune cells to help kill the tumor.
The antibody class most ADCs are built on includes an Fc region capable of signaling to the immune system. When an antibody binds to a tumor cell, the Fc region recruits immune cells to help destroy it. This function is innate to the antibody, meaning the ADC inherited it. Because it existed before the ADC was engineered, it rarely gets monitored the way payload delivery does. The assumption is that it’s still intact, but too many programs don’t check.
As more payload molecules are attached per antibody, the physical bulk can interfere with the Fc region’s ability to signal immune cells. An ADC can look highly potent in development even as its ability to recruit immune cells has quietly diminished. The standard cytotoxicity readout can’t distinguish between payload activity and the Fc immune response. If one has diminished, the number won’t show it.
Measuring immune engagement requires a different readout entirely. Our ADCC and ADCP Reporter Bioassays use effector cells to measure whether the Fc region can still recruit immune cells to help kill the tumor.
Measuring Fc effector function distinguishes between payload killing and immune engagement. Without it, you don’t know what’s driving the result.
Bystander Killing: The Cells That Were Never In Range
Not every cancer cell in a solid tumor displays the target antigen. Tumors are genetically unstable, and cells within the same tumor can express very different levels of the same antigen. Some have less of it, and some have none at all. Because the ADC is designed to find that specific protein, cells without it are invisible to the treatment. They aren’t rare exceptions; they’re part of the tumor and just as capable of driving disease as the cells it can reach.
Certain ADCs are designed to reach them anyway. When the payload is released inside a target cell, some of it escapes into the surrounding tissue and kills neighboring cells the ADC couldn’t directly target. The cell it found becomes the delivery vehicle for the cells it couldn’t find.
Bystander killing is a newer design consideration, and its clinical impact has only recently become clear. In a recent study, two ADCs targeting the same protein on the same tumor type, one with bystander activity and one without, produced dramatically different outcomes. The ADC with bystander activity (trastuzumab deruxtecan) outperformed the one without it. That difference in mechanism translated directly into a difference in survival.
Measuring bystander killing means distinguishing which cells actually died in a culture that includes both antigen-positive and antigen-negative cells. The HiBiT Target Cell Killing Bioassay is built around that question, generating a luminescent signal specific to target cell death in a mixed culture.
To design for bystander killing, optimize around it, or distinguish it from other killing mechanisms in your data, you need an assay built to see it.
The Questions Behind the Number
An ADC isn’t one mechanism doing one job. It’s several running at once, some you designed, some the molecule inherited before the payload arrived. The cytotoxicity number captures the outcome of all of them together. What it can’t show you is what the others are doing, which ones have quietly changed, and which ones were never part of your program at all.
Without that visibility, programs advance on incomplete information. Candidates get deprioritized for the wrong reason. An ADC can reach later development stages with a mechanism that stopped contributing two steps ago, and nobody knew.
Some days the answer is as simple as looking in a mirror. The assays exist. The picture is available.
References
Hillman K and Gilden J. Unlocking the Multifaceted Mechanisms of Antibody–Drug Conjugates. BEBPA. https://bebpa.org/unlocking-the-multifaceted-mechanisms-of-antibody-drug-conjugates/
Cortés J, et al. Trastuzumab Deruxtecan versus Trastuzumab Emtansine for Breast Cancer. New England Journal of Medicine. 2022. https://www.nejm.org/doi/full/10.1056/NEJMoa2115022
Elise Johnson
Latest posts by Elise Johnson (see all)
- ADC Development: The Questions Running in the Background - July 30, 2026
- Time Capsules in the Language of Science - July 23, 2026
- AI Identifies the Target. Someone Still Has to Validate It. - July 9, 2026
