Checkpoint Occupancy Versus Functional T-Cell Rescue

Target Engagement Is Not the Same as Functional Rescue

A checkpoint antibody can bind most of its target receptors without producing a large increase in T-cell activity. This is not necessarily a contradiction: binding is a measure of target engagement, whereas cytokine production, proliferation, and tumor-cell killing are downstream functions shaped by multiple steps. The useful question is therefore not simply “Was occupancy high?” but “What does this occupancy measurement establish, and what functional change would the model predict under the stated conditions?”

In the one-site binding model below, occupancy depends on free antibody concentration and affinity. Relating occupancy to function requires additional assumptions: the bound antibody must actually inhibit the checkpoint signal, and the measured T cells must have functional capacity that can be released by that inhibition. Even then, occupancy and function need not change proportionally.

A Simple Binding Model for Occupancy

For a single reversible antibody–receptor interaction at equilibrium, let \(R\) be unbound receptor, \(D\) be free antibody, and \(RD\) be antibody-bound receptor. The dissociation constant is \(K_D=[R][D]/[RD]\). If the free antibody concentration is \(C\), total receptor is \(R_{\mathrm{tot}}=[R]+[RD]\), and \(O=[RD]/R_{\mathrm{tot}}\), then:

This expression assumes one class of independent binding sites, reversible binding at equilibrium, and a free antibody concentration that is approximately constant during binding. It uses free concentration at the target, not automatically the total concentration measured in plasma. It also assumes that the assay measures the relevant receptor population without materially changing binding. In practice, heterogeneous receptor states, antibody distribution, receptor turnover, and non-equilibrium sampling can make this expression only an approximation.

The equation describes binding, not the fraction of checkpoint signaling inhibited. To represent that distinction explicitly, let \(q\) denote the fraction of antibody-bound receptors whose checkpoint signaling is functionally blocked, with \(0 \leq q \leq 1\). The effective blocked fraction is then \(qO\) in this simplified representation. If binding does not prevent signaling, \(q\) may be low even when \(O\) is high. If the antibody fully blocks signaling at every bound receptor, the idealized assumption is \(q=1\).

Why Occupancy and Function Can Separate

Even when an antibody is a functional antagonist, the relationship between the fraction of receptors occupied and a measured T-cell response may be nonlinear. A simple empirical model for the increase in a functional readout is:

Here \(\Delta F(O)\) is the change from the matched comparator in the chosen functional readout; \(E_{\max}\) is the maximum attainable increase under the assay conditions; and \(EC_{50}\) is the effective blocked fraction \(qO\) that produces half of \(E_{\max}\). Both \(O\) and \(q\) are dimensionless fractions. The units of \(\Delta F\) and \(E_{\max}\) depend on the readout, such as cytokine concentration or percentage points of target-cell lysis. This saturating model is a useful description, not a universal mechanistic law. It assumes a monotone response to effective blockade and fixed cellular conditions.

The derivative with respect to occupancy is:

For positive \(E_{\max}\), \(q\), and \(EC_{50}\), this derivative is positive, but it decreases as \(O\) increases. Thus, the model predicts diminishing functional gains as occupancy approaches saturation. The model also shows why high occupancy alone cannot ensure a large response: if \(E_{\max}\) is small because the cells have little recoverable functional capacity, the predicted change remains small at every occupancy.

Interpreting an Occupancy Assay Alongside Function

Occupancy can be estimated with assays that distinguish antibody-bound from unbound receptor, or indirectly by measuring the remaining antibody-accessible receptor. An indirect calculation such as one minus the fraction of accessible receptor signal is valid only when the signal is proportional to accessible receptor, background is handled correctly, and treatment does not alter detection independently of occupancy. Assay design matters: an antibody that interferes with the detection reagent can make accessible receptor appear reduced even if the intended binding interpretation is not justified.

A functional comparison must specify the readout, the comparator, and the observation time. For example, a cytokine increase after stimulation is not interchangeable with an increase in tumor-cell killing. Readouts may have different response ceilings and different sensitivities to changes in T-cell state. Paired functional assays can therefore help determine whether high occupancy accompanies a change in the particular function being measured, but they do not make occupancy a proxy for every T-cell activity.

Diagnostics, Alternatives, and Limitations

When occupancy is high but measured T-cell activity changes little, check the chain of inference rather than treating the result as a paradox. First verify that occupancy is estimated against an appropriate reference, with background correction and a detection method that distinguishes bound from accessible receptor. Then check whether the occupancy and function measurements refer to comparable samples and time points. Finally, establish what the functional assay measures and whether the cells had a measurable capacity to respond under its conditions.

A second diagnostic is to ask whether the binding model is plausible. The equilibrium expression \(O=C/(K_D+C)\) may mislead if concentration at the target is unknown, binding has not equilibrated, receptor abundance changes, or binding sites are heterogeneous. In those cases, a direct occupancy assay may be more informative than calculating occupancy from a nominal concentration. Conversely, an indirect assay should not be treated as definitive merely because its output is called “occupancy.”

These approaches answer different questions and are most informative when interpreted together. A concentration–affinity calculation is not a substitute for measured target engagement when its inputs are uncertain. An occupancy assay is not a substitute for a functional assay. A functional response alone does not prove that checkpoint binding caused the change. These are related measurements, not interchangeable endpoints.

Common Pitfall: Treating Occupancy as a Surrogate for Activity

Wrong reasoning: “The antibody occupied 95% of the checkpoint receptors, so T-cell activity must have increased by about 95%.” This confuses a fraction of bound receptors with a fraction of functional capacity restored. The binding equation contains no functional-response term, and even a functional-response model depends on its own parameters and assumptions.

Correction: Report occupancy as target engagement, report the functional readout and its comparator separately, and interpret any relationship in light of the assay conditions and available functional capacity. High occupancy with little measured rescue can be consistent with a response plateau, weak functional blockade per bound receptor, another limiting process, or measurement limitations. The observed discordance is a prompt to evaluate those possibilities, not a diagnosis on its own.

A practical summary is to keep three quantities distinct: the fraction of receptors bound (\(O\)); the extent to which binding blocks checkpoint signaling (\(q\), if it can be estimated); and the change in the chosen functional output (\(\Delta F\)). A high value for the first does not determine the other two without additional evidence.