Self & Other · Cancer and the Immune System
What Checkpoint Inhibitors Look Like in a Person
Allison and Honjo shared the 2018 Nobel Prize for Medicine. The Nobel committee was unusually direct in its citation: they had, the committee said, established an entirely new principle for cancer therapy. The phrase the field had begun using, a little informally, was “taking the brakes off.” The drugs were called checkpoint inhibitors, after the regulatory checkpoints they blocked. They had, by 2018, started saving the lives of people whose cancers would otherwise have killed them.
What Checkpoint Inhibitors Look Like in a Person
Pembrolizumab — sold under the brand name Keytruda — has become, in the years since its approval, one of the most-prescribed cancer drugs in the world. It is given as an infusion, typically every three or six weeks. It does not, in itself, kill cancer cells. It releases a brake on the patient’s own T cells and lets the immune system do the work it had been prevented from doing.
The most famous patient ever treated with pembrolizumab is Jimmy Carter. In August 2015, at age ninety, the former U.S. president was diagnosed with metastatic melanoma that had spread to his liver and his brain. He received surgery, radiation, and pembrolizumab. By the end of that year his cancer was undetectable. He lived nearly a decade more, entering hospice in early 2023 and dying peacefully at the end of 2024 at the age of one hundred. Most of the years he had after his cancer treatment were the result of a drug that did not exist when he turned eighty.
The Carter case is famous because of who he was, but it is not unrepresentative of what checkpoint inhibitors can do for some patients with some cancers. People with metastatic melanoma, certain advanced lung cancers, certain bladder cancers, certain head and neck cancers, certain mismatch-repair-deficient colon cancers, and a growing list of others can sometimes have prolonged complete responses on these drugs. The responses can last years, sometimes for the rest of the patient’s life.
The cost is real. Because the drugs work by removing a brake on T cells, they can sometimes allow T cells to attack the body’s own tissues. Patients on checkpoint inhibitors can develop, in effect, drug-induced autoimmune disease — inflammation of the colon, of the lungs, of the thyroid, of the pituitary gland, of the heart in rare cases. Most of these side effects can be managed with steroids and careful monitoring. Some are severe. A small number are fatal. The risk-benefit calculation, for any individual patient, depends on the cancer they have, the alternatives available, and how their body handles the treatment.
The drugs do not work for everyone. For many patients with many cancers, response rates remain stubbornly low. Some cancer types — most notably pancreatic and prostate cancers, but others as well — have so far responded poorly to checkpoint inhibition. Researchers are still trying to understand why. The leading hypothesis is that some tumors are “cold” — they have few mutations for the immune system to recognize, or they have walled themselves off from immune cell access — while “hot” tumors, the ones with many mutations and active immune infiltration, are the ones most likely to respond. Turning a cold tumor hot is one of the major projects of current cancer research.
CAR-T
The other dramatic development in cancer immunology over the last decade is a different kind of treatment, with a different kind of logic. Where checkpoint inhibitors take the brakes off T cells the patient already has, CAR-T therapy gives the patient new T cells — engineered, in the lab, to recognize their cancer specifically.
The acronym is awkward but worth knowing: CAR stands for chimeric antigen receptor. A T cell with a CAR has had its DNA modified to express a custom-built receptor on its surface — a receptor that is part antibody and part T cell receptor, and that recognizes a specific protein on the surface of cancer cells. The engineered T cell, given back to the patient, finds and kills any cell wearing that protein.
The first CAR-T cells were developed in the 1990s and 2000s by several research groups, but the breakthrough that turned them into a clinical reality came primarily from Carl June’s lab at the University of Pennsylvania. June’s team designed a CAR that recognized a protein called CD19, which is found on most B cells — normal and cancerous. The idea was to use the engineered T cells against cancers of B cells, particularly leukemias and lymphomas, by simply killing all of the patient’s B cells. The patient would lose their normal B cells too, but the cancer would be gone, and the lost normal B cells could be replaced by infusions of antibodies, the same IVIG treatment we discussed in the last chapter.
In April 2012, a six-year-old girl named Emily Whitehead became the first child treated with CAR-T cells. She had acute lymphoblastic leukemia that had relapsed twice and that no further chemotherapy could control. She was, by every standard medical measure, dying. Her parents agreed to enroll her in a clinical trial of June’s CAR-T therapy because there was nothing else left.
The treatment nearly killed her before it cured her. Within days of the infusion, the engineered T cells had multiplied enormously and were destroying her cancerous B cells in such numbers that she developed exactly the kind of cytokine storm we described in Chapter 8 — a runaway inflammatory response that put her in the ICU on ventilators and a long list of other supports. The team, in a moment that has since become medical legend, identified the dominant cytokine driving the storm and treated her with tocilizumab, an antibody normally used for rheumatoid arthritis that blocked the relevant signal. Within hours, her temperature dropped. Within days, she was awake.