Across a range of immunotherapies, from cell-based approaches to bispecific antibodies, therapeutic activity depends on how cells interact. In cell therapies such as CAR-T and TCR-T, functional outcome is determined by whether therapeutic cells establish productive immunological synapses with their targets and how those contacts evolve over time. In bispecific antibodies and BiTEs, the therapy itself is essentially a mechanism for creating and shaping contact between cells. In both cases, cell-cell interactions are not a secondary detail. They are the level at which function is executed. Yet most experimental readouts in immunotherapy development still treat interactions as secondary. Interaction is often inferred from what happens downstream, such as cytokine secretion, activation markers, proliferation, or bulk cytotoxicity. These readouts are useful and often robust, but they summarize the outcome of many heterogeneous encounters. They rarely reveal how those encounters unfolded.

The core challenge is that cell-cell interactions are dynamic and context dependent. They are not single events. Cells form contact, exchange signals, disengage, and sometimes re-engage. The duration of contact, the spatial organization of receptors at the interface, and the local balance of activating and inhibitory cues can shift rapidly. Two systems can produce similar downstream profiles while differing substantially in interaction behavior. A second challenge is scale. Many interaction phenomena are easiest to observe at the single-interaction level, but therapeutically relevant conclusions often require population-level inference across large numbers of cells and across conditions. Measuring a small number of interactions in high detail can be scientifically informative, but it does not always translate into a practical assay for development decisions. A third challenge is that interactions are relational by nature. Unlike single-cell profiling, which can treat each cell as an independent unit, interaction measurements must preserve information about pairs or groups of cells. That requirement immediately introduces additional complexity in experimental design, sample handling, and data representation. It also complicates standardization across laboratories and studies.

These technical constraints have shaped the tools that are commonly used. Many assays deliberately disrupt tissue architecture to make samples compatible with high-throughput measurements. Dissociation enables scalable profiling, but it also removes spatial context and breaks many physical interactions. What remains is a snapshot of cellular states after interactions have occurred, rather than a direct readout of interaction dynamics. Even when interactions can be observed directly, they are often difficult to quantify in a way that is comparable across experiments. Microscopy-based approaches can capture rich spatial and temporal information, but they can be limited in throughput and can require complex analysis pipelines. Flow-based approaches offer scale, but historically have struggled to preserve and identify true interacting cell pairs without introducing artifacts.

The consequence is a methodological gap. Biology and therapy functions are frequently interaction-driven, but the dominant experimental workflows have evolved around measurements that are easier to standardize and scale. This is not a failure of the field. It reflects practical constraints and trade-offs that have shaped what is measurable. At the same time, the incentives are changing. As immunotherapies expand into more complex indications, earlier treatment lines and as development decisions become more costly, there is increasing pressure to use readouts that are closer to the mechanism. That is where interaction-centric measurements become attractive. They offer the possibility of connecting therapeutic design choices to the actual cellular events that determine efficacy and toxicity.The most important question is therefore not whether cell-cell interactions matter. It is what would count as a useful interaction readout in practice. An assay must be interpretable, scalable, and robust enough to guide decisions. It must also integrate with existing workflows rather than requiring a complete reinvention of how studies are run. If cell-cell interactions are to become primary readouts, the field needs technologies and analytical frameworks that can capture interacting cell pairs reliably, at scale, and with sufficient biological specificity to be actionable.