Background Most cancers are diseases of aging. But the lack of an unbiased inquiry into the cellular basis for the link between aging and tumorigenesis has left the study of cancer incomplete. Filling this gap in knowledge could help characterize the early stages of tumorigenesis and aid the design of cancer prevention strategies. Non-small cell lung cancer (NSCLC) is strongly associated with aging. As it is the most common cause of cancer-related mortality, we sought to decipher the effect of aging on the host response to NSCLC pathogenesis.

Methods We combined heterochronic bone marrow transplant studies with an orthotopic model of primary lung adenocarcinoma and single-cell RNA sequencing to generate the first atlas of aging and cancer. We also performed multi-omic profiling of both primary tissue specimens to establish that mechanisms uncovered by our preclinical models are conserved in human cancer.

Results We show that aging of the immune system, regardless of the age of the stroma, drives lung cancer progression. Hematopoietic aging exacerbates emergency myelopoiesis and results in the extramedullary accumulation of myeloid progenitors in lung tumors. These cells are a primary source of IL-1α, signal to IL-1R1⁺ hematopoietic progenitors in bone marrow in an age-specific manner, and sustain the age-enhanced myelopoietic response in old mice. Disrupting this feedback loop using either an anti-IL-1α antibody or the IL-1R1 antagonist anakinra early during tumor initiation not only slowed tumor growth but also normalized myelopoiesis. We discovered that the increased proclivity of aged myeloid cells to produce more IL-1α is due to an age-associated decline of DNA methyltransferase 3A (DNMT3A). In patients, we show that emergency hematopoiesis is enhanced with age and cancer stage; from older patients, circulating HSPCs exhibit a decrease in DNMT3A, and immature monocytes produce more IL-1α. In human lung lesions, we identified an enrichment for monocyte-derived macrophages that act as a major source of IL-1α and associated with aging, poorer survival, and recurrence of cancer.

Conclusions We established that the relationship between aging and the host response to lung cancer is an immunologically-driven one. Our study not only highlights the importance of immunotherapy in improving aging-dependent antitumor control but more specifically rationalizes an IL-1α-based intervention to deter pathogenic myelopoiesis that is worsened with age. We define a therapeutic window for such an approach. Our findings have direct relevance to the design of immunoprevention trials and to our understanding of aging, DNMT3A mutation-driven clonal hematopoiesis, and lung cancer.