Todd Fehniger looks into a microscope in his laboratory

Oncologist Todd Fehniger researches natural-killer-cell therapy for use in cancer treatment.Credit: David Russler-Germain

As part of a study published in January, 37 people who had been through various failed treatments for blood cancer received infusions of genetically enhanced immune cells that researchers hoped would clear their disease1. The concept was not new — therapies based on T cells have been approved since 2017. But the source of the cells was.

Researchers at the MD Anderson Cancer Center in Houston, Texas, used frozen donor umbilical-cord blood to grow and prepare a lesser-known component of the immune system: natural killer (NK) cells. In the phase I/II trial, 68% of people with B-cell lymphoma who received the cells survived for at least one year — suggesting that the NK-based therapy could work as well as currently available T-cell therapies2. And NK cells seemed to trigger fewer side effects, which have limited the availability of therapies that use chimeric antigen receptor (CAR) T cells, says Katy Rezvani, an oncologist and immunologist at the MD Anderson Cancer Center who led the recent NK-cell study.

CAR-T therapy requires the extraction of an individual’s own T cells and a two-to-four week wait for the cells to be engineered. After the cells are injected back into individuals, at least one-third of recipients develop cytokine release syndrome, in which a rush of inflammatory molecules causes fever, low blood pressure and other symptoms requiring hospitalization. About one-quarter also develop neurotoxicity, which can be life-threatening3. All of this makes the price of therapy high — around US$400,000 — and limits it to people who are able to withstand the side effects.

So in 2013, at a time when all eyes were on T cells, Rezvani shifted her research focus to NK cells. NK cells have a better safety record and unique features that make them good candidates for off-the-shelf therapies, in which cells from an unrelated donor are prepared ahead of time and frozen, ready to use in anyone. This would save time and money and be a game changer for people who are too unwell to have their own cells extracted. The advantages of NK cells first piqued cancer researchers’ interest two decades ago, but issues such as the cells’ inability to last long in the body hindered development. In the past few years, strategies to take on this challenge have delivered encouraging results in clinical trials.

There are now at least 50 companies testing a wide variety of tactics to give NK cells an edge as a cancer therapy. Several have started phase I trials to test CAR-NK cells like Rezvani’s, grown from a variety of sources including cord blood, induced pluripotent stem cells and cell lines. Other groups, meanwhile, are developing drugs that restore the cancer-killing abilities of NK cells that are already in the body.

“The number of NK companies has just exploded,” says Todd Fehniger, an oncologist at Washington University School of Medicine in St. Louis, Missouri. As researchers test their ideas in early clinical trials, the field is anxious to see which approach, if any, will give people long-lasting results. “We’re still at this inflection point in the field where we need that one clinical success, that one regulatory approval that will spark this torrent of interest,” Fehniger says.

Perked-up persistence

NK cells are well-positioned to steal the immunotherapy spotlight away from T cells. NK cells are equipped with cancer-detecting receptors (see ‘Killer at work’). They got their name in the 1970s for their ability to spot infected and cancerous cells, release molecules that poke holes in cell membranes and trigger their targets’ own self-destruction. NK cells also have a reputation for keeping cancer at bay: one study linked high NK-cell activity to a lower likelihood of developing cancer over the subsequent decade4.

Killer at work: infographic showing how natural killer cells can be used to combat cancer

Source: A. Merino et al. Blood Rev. 60, 101073 (2023).

What really drew Rezvani in, however, was the safety of these cells. NK cells can’t kill other cells until they integrate a combination of activating and inhibitory signals from dozens of receptors; this constraint prevents them from harming healthy cells. Once activated, NK cells don’t replicate in huge numbers like T cells do, making them much less likely to kick off cytokine release syndrome or graft-versus-host disease, a sometimes-fatal condition in which donor cells attack a transplant recipient’s tissues.

But many of the features that contribute to the safety of using NK cells have also hindered their success as a cancer treatment. In studies done in the 2000s, NK cells from donors lasted only one or two weeks before participants’ immune systems cleared them out5. “There was scepticism that NK cells were not going to be as effective as T cells and I think those scepticisms were well justified. The big problem was their limited persistence,” Rezvani says.

Researchers can extend that persistence by giving people IL-2, an NK-activating cytokine. But IL-2 can be given only at low doses because it can be toxic. IL-2 also stimulates T regulatory (Treg) cells, which hamper anti-tumour immunity and NK-cell activity. In the past decade, another cytokine, IL-15, has gained attention as a more specific way to activate NK cells. Rezvani’s group found that engineering CAR-NK cells to coexpress IL-15 improved their metabolic fitness, which increased their numbers and longevity inside the body6. In the recent trial1 conducted by Rezvani and her team, the engineered CAR-NK cells could still be found in some participants’ blood a year after the treatment.

The study also revealed important variables about the cell source that affected NK-cell activity. People who experienced the best outcomes received NK cells grown from umbilical-cord blood that was frozen within 24 hours of collection and lacked nucleated red blood cells, an indicator of physical stress. For those who received NK cells from such optimal cord-blood units, the one-year survival rate was 69%. When the cord blood did not meet these standards, it was just 5%. “The minute we came up with this, we went and changed all of our protocols,” Rezvani says — now she uses only optimal cord blood.

Rezvani says that her team is constantly incorporating ideas from NK-cell-biology research into its efforts to use the cells in treatment. This includes methods of processing NK cells that don’t involve any genetic engineering and that improve the cells’ function in the body. About 15 years ago, for example, researchers at Washington University School of Medicine found that a particular blend of cytokines could reprogram mouse NK cells to become more easily activated when they are stimulated for a second time. The cells retained this memory-like quality weeks after being injected into different mice7.

Working in Fehniger’s lab, immunologist Rizwan Romee, who is now a medical oncologist at the Dana-Farber Cancer Institute in Boston, Massachusetts, and his colleagues found that it was possible to nudge human NK cells into a similar memory-like state in which the cells live longer, divide more, make more pro-inflammatory cytokines and resist inhibitory signals. The team first tested the cells in nine people with acute myeloid leukaemia (AML), and found that the cells cleared the cancer in five people — one of whom had been through four previous failed therapies8. This was encouraging, Romee says, because AML is notoriously difficult to treat and T-cell therapies have made little headway against it. NK cells, by contrast, naturally recognize AML cells well. “This is where NK cells should shine, based on their biology,” Romee says.

Fehniger and his colleagues are testing memory-like NK cells as an add-on therapy to stem-cell transplantation in adults and children with AML. In experiments in which the NK cells came from the same donor, the cells multiplied 1,000-fold and lasted for at least 2 months9. His team is also detailing the characteristics that distinguish memory-like NK cells using tools such as mass cytometry, which can measure dozens of proteins or genes in millions of individual cells at a time. In 2021, the biotechnology company Wugen, which is based in St Louis and was co-founded by Fehniger, obtained the licence for the memory-like NK-cell protocol. Wugen is now testing its own cryopreserved off-the-shelf NK-cell product in people with AML and some solid tumours.

Cell image of lymph note biopsy showing NK cells

A post-treatment tumour biopsy shows active natural killer cells (white).Credit: Ref. 10

Romee, meanwhile, has started an NK-cell programme at Dana-Farber. He and his colleagues are testing memory-like NK cells against AML and other cancers, and engineering the cells with cancer-targeting CARs.

Researchers are continuing to explore pre-treatment ‘recipes’ that could boost NK-cell function once the cells have been transferred into the body. In the past year, the biotechnology company Gamida Cell, which is based in both Boston and Kiryat Gat, Israel, have identified multiple benefits of growing NK cells with IL-15 and nicotinamide — a water-soluble form of vitamin B10. The combination improves the metabolism of NK cells, protects them from oxidative stress, enhances their cancer-killing activity and turns on genes for surface proteins that improve their ability to home in on tumour sites.

Veronika Bachanova, an oncologist at the University of Minnesota Medical School in Minneapolis, partnered with Gamida Cell to examine the effects of nicotinamide on NK cells and test them in a phase I trial in combination with rituximab, a monoclonal antibody that targets B-cell cancers. In a study published last year, the researchers showed that 13 out of 19 people with advanced non-Hodgkin’s lymphoma responded to the combination10. A lymph-node biopsy from one person showed that the treatment encouraged T cells to infiltrate the cancerous tissue. So, even though the NK cells might be around for only a couple of weeks, Bachanova says, that might be enough to stimulate a powerful anti-cancer immune response. Gamida Cell is continuing this work in a phase II trial.

By combining NK cells with rituximab, the biotech firm is one of many companies taking advantage of a natural interaction between antibodies and the potent NK-cell activating receptor, CD16. “We are actually using the antibody for the targeting,” says Fred Aslan, chief executive of Artiva Biotherapeutics in San Diego, California, which is also testing this combination. “We have the luxury of not having to engineer our product because it naturally will kill the [cancer] cells once they are tagged by the antibody.” Aslan says that Artiva has enhanced the effect by using only NK cells from umbilical-cord blood donors who have a naturally occurring CD16 variant that binds extra tightly to antibodies.

Initial results from a phase I trial showed that Artiva’s combination shrank or eliminated signs of tumours in four out of six people with non-Hodgkin’s lymphoma — three of whom had previously undergone CAR-T therapy without success11. As far as the issue of persistence is concerned, Aslan says, the team’s remedy is simply to give people high doses of NK cells, which is possible because of their safety. “NK cells are not really designed to be one-and-done like CAR-T cells, in terms of their expansion,” he says. He hopes that updated results from the non-Hodgkin’s lymphoma trial, expected later this year, will help to make the case that, at the right doses, even unedited NK cells can produce lasting remission.

Working with what you’ve got

Because of their multitude of cancer-detecting receptors, NK cells that are already present in the body are becoming a popular target for drugs that aim to activate anti-tumour immunity. In fact, for many types of cancer, an abundance of NK cells in a tumour is linked to a better outcome. The problem is that tumours can disarm NK cells by, for example, triggering inhibitory receptors or producing enzymes that snip off activating receptors such as CD16. NK cells are also sensitive to nutrient competition and metabolic stress, which are common in tumour microenvironments.

“The tumour microenvironment is a challenging environment for NK cells to operate in,” says Nicholas Huntington, a cancer immunologist at Monash University in Melbourne, Australia. “We’re looking at ways of improving their metabolism, fitness and function so we can restore some of that innate natural killing ability.” Huntington is a co-founder of oNKo-innate, a company in Melbourne focused on finding targets for small-molecule drugs that could enhance NK-cell function. One of its drug candidates, for example, enhances NK-cell and T-cell sensitivity to IL-15.

Others are taking advantage of known activating and inhibitory receptors to bolster existing NK cells. Innate Pharma, a biotechnology firm based in Marseille, France, is testing an NK-cell checkpoint inhibitor in a phase III trial as a complement to T-cell checkpoint-inhibitor therapy for non-small-cell lung cancer. The drug blocks an inhibitory receptor found on NK cells and some T cells called NKG2A, unleashing both branches of the immune response.

Innate Pharma has also designed a suite of molecules called engagers, which simultaneously activate immune cells and bring them into close proximity to their targets by binding known tumour antigens. Engagers that bind T-cell receptors are already in clinical use, but they carry a high risk of cytokine release syndrome. NK-cell engagers pose a safer alternative that could still ignite anti-cancer immunity. The company’s tri-specific engager binds to blood-cancer proteins, CD16 and another NK-activating receptor; its tetra-specific engager has an added region that stimulates the IL-2 receptor on NK cells, but not on Treg cells. In mouse studies, the tetra-specific engager caused NK cells to infiltrate solid lung tumours12, a feat that has been difficult in immunotherapy.

Researchers have started pairing engagers with NK cells to get the most from both therapies. Last year, Rezvani and her colleagues reported their findings on a CD16-based engager made by Affimed, a pharmaceutical firm in Mannheim, Germany, in combination with cord-blood-derived memory-like NK cells. The trial tested different doses of cells in people with blood cancer, most of whom had chemotherapy-resistant Hodgkin’s lymphoma and had already been through an average of seven lines of therapy. Of the 36 people who were given the highest dose of NK cells plus the engager, 72% experienced complete remission13.

Aslan says that Artiva is testing Affimed’s engager with its NK cells in a phase I/II trial designed to gather the data needed to apply for FDA approval if the results are good. Fehniger expects the first regulatory approval for NK-cell therapy to come in the next several years. “There are a lot of examples where we’re really close,” he says.

For Bachanova, the work and the wait are well worth it on the basis of the remissions she has seen over the years that avoid the intense side effects common with T-cell therapies. “The patients feel good,” she says. “You witness that and it is powerful.”