How Immune Checkpoint Inhibitors Work and Why Response Is Uneven

Immune checkpoint inhibitors work by removing a brake, not by adding an accelerator. Healthy tissue relies on molecular "off" switches, called checkpoints, so that activated T cells stop short of attacking the body's own cells. Many tumors survive by pulling those same switches. According to the National Cancer Institute, drugs that block the PD-1, PD-L1, or CTLA-4 interactions prevent the "off" signal from reaching the T cell, which frees the T cell to recognize and kill cancer cells. That single mechanism explains both the striking, sometimes durable responses these drugs can produce and why a large share of patients see no benefit at all: a released brake only helps if a capable immune response was there to begin with.

The checkpoints and what they normally do

Two pathways account for most approved checkpoint inhibitors, and they act at different stages of the immune response.

CTLA-4 operates early, during T cell priming in the lymph node. When a T cell is first being activated by an antigen-presenting cell, CTLA-4 competes for the B7 molecules on that presenting cell and delivers an inhibitory signal that dampens activation. Blocking CTLA-4 lowers the threshold for T cells to become active in the first place.

PD-1 and its partner PD-L1 act later, in the tumor itself. PD-1 sits on the surface of T cells; PD-L1 is the ligand that engages it. As the NCI describes, some tumors produce large amounts of PD-L1, which turns down the T cell response locally. The review by Alsaab and colleagues in the peer-reviewed literature frames this as pharmacologically preventing the PD-1/PD-L1 interaction to restore anti-tumor immunity. Because CTLA-4 and PD-1 blockade act at different points, combining them can be synergistic, though that also compounds the risk of immune-related side effects.

A useful way to hold these apart: CTLA-4 blockade widens the pool of T cells that get activated, while PD-1/PD-L1 blockade keeps already-activated T cells from being switched off inside the tumor.

Why response is uneven

If the biology were as simple as "block the brake, win the fight," response would be near-universal. It is not. The variability traces to a chain of requirements, and a tumor can fail at any link.

The tumor has to be visible to T cells

T cells recognize cancer by the abnormal proteins that arise from mutations. Tumors carrying many mutations tend to display more of these neoantigens, giving T cells more to target. This is the logic behind tumor mutational burden as a biomarker. In the KEYNOTE-158 analysis published by Marabelle and colleagues, which supported the 2020 approval of pembrolizumab for tumor-mutational-burden-high solid tumors (defined as at least 10 mutations per megabase), the objective response rate in that biomarker-selected group was about 29 percent, with a majority of responses lasting a year or more. A closely related marker is mismatch-repair deficiency, or high microsatellite instability, which floods a tumor with mutations. The FDA approval summary by Marcus and colleagues describes the 2017 accelerated approval of pembrolizumab for MSI-high or mismatch-repair-deficient tumors as the agency's first tissue-agnostic indication, granted on a 39.6 percent objective response rate across 15 tumor types. Both figures are notable in two directions: they are far higher than one would expect from unselected tumors, yet even in these favorable groups most patients did not respond.

T cells have to be present and able to enter

A tumor that expresses the right targets still fails treatment if T cells cannot get in. The review literature describes "non-inflamed" tumor microenvironments with poor CD8-positive T cell infiltration, sometimes shorthanded as "cold" tumors versus "hot," inflamed ones. Tumor heterogeneity compounds this: the density of T cells can vary from the edge to the core of the same tumor, shaped by hypoxia and by which mutations are present where. Releasing a brake in a region no T cell has reached accomplishes nothing.

PD-L1 status is informative but not decisive

It is tempting to assume PD-L1 expression should predict who benefits, but the evidence is messier. The review is explicit that not all PD-L1-expressing tumors respond to PD-1/PD-L1 inhibitors, and, conversely, some PD-L1-negative tumors do respond. Part of the problem is measurement: PD-L1 staining varies with tumor location and sampling, so a single biopsy may misrepresent the whole tumor. This is why PD-L1 testing guides some decisions but is not treated as a clean yes-or-no switch.

Tumors actively resist

Even a promising tumor can defeat the drug through resistance. The literature distinguishes two patterns. In innate (constitutive) resistance, signaling pathways such as PI3K-Akt and STAT3 drive PD-L1 expression independent of any immune pressure. In adaptive resistance, the tumor upregulates PD-L1 specifically in response to interferon-gamma released by attacking immune cells, so the immune attack itself provokes the countermeasure. Resistance can also emerge after an initial response, as tumors evolve to escape a working immune attack.

What this means for interpreting the evidence

The honest reading is that checkpoint inhibition is a conditional therapy. Its benefit depends on a tumor being mutated enough to be seen, infiltrated enough to be reached, and not already armored with resistance machinery. Biomarkers like tumor mutational burden, microsatellite instability, and PD-L1 expression each capture part of that picture, which is why they enrich for responders without guaranteeing response, and why no single test yet identifies everyone who will benefit. Reading a checkpoint inhibitor result well means asking which of these conditions a given tumor met, rather than treating the drug class as uniformly effective or uniformly futile. This article is educational and is not medical advice.