Belly Fat Weakens Immune Fitness

Ovarian cancer is genetically heterogeneous and has several important histologic types. High-grade serous adenocarcinoma is the most common and the type to which we refer here by “ovarian cancer.” There is generally uniform agreement that ovarian cancer is immunogenic, originally suggested by the finding that tumor CD8⁺ T-cell infiltration predicted improved survival and reduced recurrence (1), followed by a demonstration that in this same patient cohort, a higher number of detrimental regulatory T cells predicted reduced survival (2). Recent collaborative studies identified a distinctive signature for T-cell clonotypes that exert sustained and active pressure against malignant progression (3). Nonetheless, ovarian cancer has not enjoyed the same success in cancer immunotherapy trials as other carcinomas despite much effort from highly qualified research teams. Resistance to cancer immunotherapy is multifactorial. Many groups have tackled known immunosuppression mediators in preclinical studies and early-stage clinical trials, but success remains elusive.

Recent studies demonstrate that many types of tumors, including ovarian cancers, can thwart type I IFN (e.g., IFNα/β) production and signaling as a major strategy to evade immune control and mount resistance to multiple forms of therapy. However, the mechanisms mediating these immunosuppressive processes remain incompletely understood. Whereas much attention has been paid to secreted proteins (e.g., transforming growth factor-β, vascular endothelial growth factor), surface mole­cules (e.g., PD-L1), and cells (e.g., regulatory T cells, myeloid-derived suppressor cells) as mediators of immunosuppression in the ovarian cancer (or any cancer, for that matter) tumor microenvironment, much less attention has been paid to lipid mediators of immunosuppression.

In this issue of Cancer Discovery, the Cubillos-Ruiz team showed that lysophosphatidic acid, a bioactive lipid messenger enriched in the tumor microenvironment (4), functions as a dominant regulator of otherwise potentially potent anticancer type I IFN responses (ref. 5; Fig. 1). Lysophosphatidic acid is generated by the extracellular enzyme autotaxin that is recognized by multiple G protein–coupled receptors expressed by tumor and immune cells. Signaling in tumor cells promotes their tumorigenic attributes, while signaling through the LPA5 receptor has been shown to inhibit the effector function of CD8⁺ T cells (6). Emerging studies now indicate that lysophosphatidic acid also directly suppresses T-cell activity by altering the cytoskeleton, which prevents the formation of a proper immune synapse (7). However, its effects on innate immune cells or effects in ovarian cancer are little or not described.

Chae and colleagues now offer a new and important layer of mechanistic understanding of the immunosuppressive effects of lysophosphatidic acid. First, they found high levels of secreted autotaxin, a potent lysophosphatidic acid inducer, in metastatic ovarian tumors. Autotaxin-engendered lysophosphatidic acid suppressed the antitumor immune responses driven by type I IFNs induced through distinct approaches, including PARP inhibition and Toll-like receptor 3 agonist [poly(I:C)] treatment. The specific mechanism of action appeared to be lysophosphatidic acid–mediated inhibition of type I IFNs in distinct dendritic cell populations (Fig. 1) seen in mouse models and corroborated with human tissues. Dendritic cells are innate immune cells critical to generating de novo immune responses (immune priming) and providing specific instructions to T cells, including guiding their differentiation into protumor or antitumor immune cells. Specifically, lysophosphatidic acid induced prostaglandin E₂ production from dendritic (and other) cells, which suppressed autocrine type I IFN production from distinct dendritic cell subsets, thereby blunting antitumor immunity. Lysophosphatidic acid also reduced tumor infiltration of innate natural killer cells that can mediate antitumor immunity (Fig. 1). Thus, in addition to direct effects on tumor-specific, antitumor CD8⁺ T cells, lysophosphatidic acid can indirectly suppress antitumor immunity by inhibiting dendritic cell functions and by suppressing tumor ingress of innate natural killer cells.

The research team analyzed 44 ovarian cancer patients in the TOPACIO/KEYNOTE-162 clinical trial testing a combination of niraparib (a small-molecule PARP inhibitor) plus pembrolizumab, an anti–PD-1 antibody as immune-checkpoint blockade, to define a signature of seven lysophosphatidic acid–controlled genes associated with poor responses to this treatment combination. In their analysis, they found that defective homologous recombination DNA damage repair in the tumor predicted prolonged progression-free survival. However, 23% of nonresponding tumors did not express this signature, suggesting that other factors were also contributing to the clinical efficacy resistance of this particular combination treatment approach. The research team hypothesizes that lysophosphatidic acid–driven immunosuppression could also contribute to efficacy here based on its ability to suppress the type I IFN response, which could otherwise be engendered from defective homologous recombination DNA repair alone, or from PARP inhibition addition.

Taken together, these data suggest an important role for lysophosphatidic acid immunopathogenesis in ovarian cancer and that suppressing the autotaxin/lysophosphatidic acid axis could improve the efficacy of other approaches such as combinations with PARP inhibitors or immune-checkpoint blockade. Understanding and targeting the immunobiology of lipids in ovarian cancer, which grows in an environment rich in adipocytes, could open new therapeutic avenues that synergize with immunotherapy. For instance, seminal studies by Herber and colleagues found that dendritic cells with high lipid content were not able to stimulate T cells effectively (8). This current work adds important mechanistic insight into that initial observation, now over a decade old. X-box binding protein 1 activation is augmented in ovarian cancer. This group previously showed that lipid peroxidation byproducts can induce X-box binding protein 1 activation in dendritic cells at tumor beds, inhibiting their capacity to activate antitumor T cells (9). These studies have since been extended to other myeloid cells, where oxidized lipids accumulate and are transferred between cells (10). Lipids also reprogram neutrophils to acquire immunosuppressive activity (11). Targeting the appropriate lipid receptors therefore emerges as a promising intervention to improve the efficiency of cancer immunotherapy, but much mechanistic work remains to be done to capitalize on these initial discoveries.

Clinical testing of inhibiting the autotaxin/lysophosphatidic acid axis is merited based on these data and, if found to be effective, would be a welcome addition to the immunotherapy armamentarium in ovarian cancer. There is no particular reason to expect that these insights pertain exclusively to ovarian cancer. As lysophosphatidic acid is present in many types of cancers, such approaches could also extend to other tumor types that thus far stubbornly resist response to available immunotherapies, including pancreatic, breast, and prostate cancers. The gene signature this team developed and other signal effects could be useful to develop much-needed, informative, and actionable treatment response biomarkers in any of these cancers.

Various small-molecule autotaxin inhibitors are currently being tested in human clinical trials, not only in cancer but also in other diseases, particularly pulmonary fibrosis. These autotaxin inhibitors also merit consideration in cancer trials to test if they can unleash the potency of type I IFN responses through suppressing detrimental lysophosphatidic acid effects. Discoveries in the role of bioactive lipids as novel cancer immunotherapy targets such as that just reported by Chae and colleagues should move us along the road to understanding ovarian cancer immunopathogenesis and the requirements for developing effective immunotherapies for it and for other cancers as well.

J.R. Conejo-Garcia reports personal fees from Alloy Therapeutics and grants and personal fees from Anixa Biosciences outside the submitted work, as well as a patent for methods and compositions for treating cancer (Anixa Biosciences) issued to The Wistar Institute and a patent for antigen binding agents that bind CD277 and uses thereof (Compass Therapeutics) pending to Compass Therapeutics. No disclosures were reported by the other author.