What is trained immunity, and how does it work against cancer?
Trained immunity is a newly recognized form of memory in the innate immune system—the body's first line of defense. Unlike adaptive immunity (which relies on T cells and B cells that recognize specific antigens), trained immunity results from epigenetic and metabolic reprogramming of innate cells like macrophages, monocytes, and natural killer cells after an initial stimulus, making them respond more vigorously to later, unrelated threats [2][3]. This means a vaccine or microbial product can 'train' these cells to attack cancer cells even though the cancer wasn't the original target.
The mechanism involves lasting changes in gene expression without altering the DNA sequence. For instance, after exposure to BCG (the tuberculosis vaccine) or β-glucan (a yeast-derived compound), innate immune cells undergo histone modifications and shifts in metabolism that prime them for stronger cytokine production and phagocytosis [2][3][9]. In the context of cancer, this trained state can help overcome the immunosuppressive tumor microenvironment, which normally disables immune attacks [6][9].
What the best evidence shows: clinical and preclinical successes
The most compelling clinical evidence comes from bladder cancer treatment with BCG. Intravesical BCG has been the standard therapy for high-risk non-muscle-invasive bladder cancer for decades, and a 2023 study of 11 patients showed that after four instillations, their innate immune cells produced significantly higher levels of trained immunity cytokines (IL-1β, IL-6, TNFα) when exposed to SARS-CoV-2, even though they had no adaptive immunity to the virus [8]. This proves BCG induces a broad, non-specific trained response. Another study is now prospectively measuring this trained phenotype in bladder cancer patients before BCG therapy to predict who will respond best [4].
In pancreatic cancer—one of the deadliest and most immunotherapy-resistant tumors—a 2023 study combined β-glucan-induced trained immunity with irreversible electroporation (IRE), a non-thermal tumor ablation technique. In a mouse model, this combination reduced both local and distant tumor burden and doubled survival compared to either treatment alone. Remarkably, the effect persisted even in mice lacking adaptive immune cells (Rag-/- mice), proving the trained innate system alone drove the anti-tumor response [5]. The same study also showed that oral β-glucan altered the innate cell landscape in the blood of five patients with stage III pancreatic cancer who had undergone IRE, suggesting the approach is feasible in humans [5].
The catch: where trained immunity falls short or backfires
Despite these successes, trained immunity is not a guaranteed win. A 2025 study found that training macrophages with β-glucan actually decreased their ability to clear dead cancer cells (efferocytosis)—a process that normally helps prevent tumor spread. Trained macrophages showed lower levels of active caspase-1 and produced less interleukin-1β when encountering apoptotic tumor cells [10]. This means the same training that boosts some anti-tumor functions can impair others, potentially allowing cancer cells to persist.
Another major limitation is the plasticity of tumor-associated macrophages (TAMs). These cells can switch between a tumor-fighting (M1) and a tumor-promoting (M2) phenotype. Current therapies struggle to selectively target the harmful M2 type without affecting healthy tissue macrophages, leading to off-target toxicity and limited efficacy [1]. Even the most advanced approach—chimeric antigen receptor macrophages (CAR-M)—faces hurdles: first-generation CAR-M cells cannot expand in the body, requiring large infusion doses, and the targeted antigens are often present on healthy cells, risking side effects [1]. While second- and third-generation CAR-M designs aim to solve these issues (e.g., using iPSCs to generate unlimited cells or delivering CAR genes via nanoparticles), none have yet been approved for clinical use [1].
Finally, trained immunity can be a double-edged sword. Inappropriate activation by endogenous stimuli (e.g., from chronic inflammation or metabolic disease) can contribute to inflammatory and neurodegenerative disorders [2][7]. So while harnessing trained immunity for cancer is promising, it must be carefully controlled to avoid fueling other diseases.
About These Sources
This answer is built on 10 peer-reviewed studies — published from 2022 to 2025, 5 from 2024 or later, 8 in Q1 journals, collectively cited 412 times — selected as the most relevant from 15 studies that passed quality screening, drawn from 53 papers retrieved from a database of over 500 million.
Sources used in this answer
CAR-macrophage: An extensive immune enhancer to fight cancer
Reviews CAR-macrophage therapy, noting that first-generation CAR-M cells cannot expand in vivo and face off-target toxicity, but second- and third-generation designs (e.g., using iPSCs or in vivo gene delivery) aim to overcome these limitations.
Trained innate immunity: Concept, nomenclature, and future perspectives
Defines trained immunity as epigenetic and metabolic reprogramming of innate immune cells, and states that activation of trained immunity can lead to novel strategies for cancer treatment.
Trained immunity: induction of an inflammatory memory in disease
Reviews the role of trained immunity in various diseases, including cancer, and notes that it can counteract immunosuppression in the tumor microenvironment.
Defining innate immune training potential as a predictor of Bacillus Calmette-Guérin immunotherapy response in nonmuscle-invasive bladder cancer.
Presents proof-of-principle data that primary monocytes from bladder cancer patients can be trained in vitro to produce more TNFα, and this phenotype will be correlated with future BCG response in a prospective cohort.
Irreversible electroporation augments β-glucan induced trained innate immunity for the treatment of pancreatic ductal adenocarcinoma
In a murine orthotopic pancreatic cancer model, β-glucan combined with IRE ablation reduced tumor burden and doubled survival, even in mice lacking adaptive immunity; oral β-glucan also altered innate immune cells in five stage III pancreatic cancer patients.
Exploiting innate immunity for cancer immunotherapy
Reviews strategies exploiting innate immunity for cancer, including CAR-macrophage and CAR-NK cell therapies, and notes that innate cells are the cornerstone of anti-tumor immune response.
Innate Immunity and MASLD
Reviews the role of innate immunity in metabolic dysfunction-associated steatotic liver disease (MASLD) and its progression to hepatocellular carcinoma, noting that targeting innate pathways may be therapeutic.
Intravesical BCG in bladder cancer induces innate immune responses against SARS-CoV-2
In 11 unvaccinated, SARS-CoV-2-naïve bladder cancer patients, intravesical BCG did not induce adaptive immunity to SARS-CoV-2 but significantly increased mRNA and protein expression of trained immunity cytokines (IL-1β, IL-6, TNFα) after four instillations.
Trained Immunity in Cancer: Pathways, Strategies and Emerging Therapies
Reviews molecular mechanisms of trained immunity (histone modifications, DNA methylation, metabolic shifts) and highlights novel nanoparticle-based approaches to deliver therapeutics that stimulate trained immunity within the tumor microenvironment.
Trained innate immunity attenuates macrophage efferocytosis of cancer cells
In mice, β-glucan-induced trained immunity decreased macrophage efferocytosis of apoptotic tumor cells, reduced active caspase-1, and lowered IL-1β production, revealing a potential downside of trained immunity in cancer.