Immunology is the specialty that learned to work with the immune system, not against it.

In January 2009, the New England Journal of Medicine published a study of gene therapy for children born without a working immune system. The disease is a form of severe combined immunodeficiency. It comes from a missing enzyme called ADA, and it is fatal. The usual cure was a bone marrow transplant from a matched brother or sister, but these children had none. The researchers took each child’s own bone marrow stem cells and added a working copy of the ADA gene. Then they gave the cells back. Ten children took part. All ten were alive after a median of four years. Nine of them rebuilt a working immune system. Eight no longer needed enzyme replacement. Five stopped their antibody infusions and made their own antibodies after vaccines. Some children had serious side effects, such as infections from their intravenous lines. This was a small study with no comparison group. A missing immune system could now be rebuilt from a child’s own cells.

Almost ten years later, in November 2018, the same journal published a trial about peanut allergy. Peanut allergy had no approved treatment, and a small amount of peanut could cause a dangerous reaction. Patients could only avoid it. The trial asked whether the immune system could be trained to tolerate peanut instead. It gave allergic children and adults either a peanut protein powder, in doses that rose slowly over months, or a placebo. The trial measured its main result in 496 children aged 4 to 17. The researchers then tested how much peanut each child could eat without a serious reaction. Two peanuts’ worth of protein was safe for 67.2% of the treated children. It was safe for only 4.0% of the children on placebo. Severe side effects during treatment affected 4.3% of the treated children. They affected 0.8% of the children on placebo. The treatment did not work in adults. A daily dose of peanut had taught allergic children to tolerate it.

Two years later, in December 2020, the same journal published the trial of an mRNA vaccine against covid-19. The vaccine carried genetic instructions for the virus’s spike protein, so the body could learn to recognise it. The trial gave 43,448 people two doses of the vaccine or a placebo, three weeks apart. Covid-19 struck 8 people who had the vaccine, counted from a week after the second dose. It struck 162 people who had the placebo. The vaccine was 95% effective. Nine of the 10 severe cases were in the placebo group. Side effects were mostly short-lived pain, tiredness and headache. A new kind of vaccine had taught the immune system to prevent a new disease.

These are three of the thousands of studies that shaped modern immunology. They taught immunology that a child’s own cells could rebuild a missing immune system, that slowly rising doses of peanut could teach allergic children to tolerate it, and that an mRNA vaccine could teach the body to prevent a new disease, and each time the answer was in a paper years before it was in a textbook.

It is still happening. Recently, in March 2026, Nature Medicine published a study about resetting the immune system in autoimmune disease. Some patients with lupus, systemic sclerosis or inflammatory muscle disease do not respond to standard treatment. The study gave 24 of these patients a single infusion of their own immune cells, engineered to destroy the B cells that drive these diseases. Every patient first stopped their immune-suppressing drugs. Twenty-two of the 24 patients reached the study’s goals within 24 weeks. Nine of the 10 patients with lupus went into remission. All of them stayed off steroids and other immune-suppressing drugs throughout. No patient had a severe reaction to the infusion. This was a small study with no comparison group. A single infusion had calmed the disease without daily drugs to hold it down. Practice follows the literature as it goes. It is a good culture. It has exactly one cost.