Heavier cancer patients often respond better to immunotherapy than thin ones, a result oncologists call the obesity paradox and have mostly shrugged at. A new study pins part of it on a single fat-derived hormone that does opposite things in different tumors, and on one enzyme that decides which way it goes.
Working with human tumor samples and lab-grown immune cells, a team led by Pawel Kalinski at Roswell Park Comprehensive Cancer Center in Buffalo reported July 10 in the Journal of Immunology ↗ that intratumoral leptin predicted better survival in breast and liver cancers but worse survival in colon and esophageal cancers. Leptin is the hormone fat tissue releases roughly in proportion to how much fat a person carries. Same hormone, opposite sign, depending on the organ.
The dividing line was an enzyme called IDO1. In the cancers where leptin tracked with worse survival, tumor-associated myeloid cells (immune cells that are supposed to help) carried high levels of IDO1, which starves and suppresses the T cells that attack tumors. Leptin turned out to be double-edged on those cells directly. It pushed human macrophages to make T-cell-activating signals, but through the NF-kB and STAT3 pathways it also switched on IDO1. IDO1 working alongside the inflammatory signal interferon-gamma then flipped the same cells into T-cell suppressors, routing them down the kynurenine pathway that quiets the immune attack.
The useful part is that the switch is druggable. Blocking IDO1 reprogrammed the leptin-exposed macrophages back from suppression to stimulation. IDO1 inhibitors already exist, and they largely failed when tested broadly. The class leader, epacadostat, missed its main goal in a large 2018 melanoma trial that paired it with the checkpoint drug pembrolizumab (Keytruda) in unselected patients, and enthusiasm for the target collapsed. This work maps out a subgroup where an IDO1 blocker might actually earn its place: obese patients whose tumors are leptin-high and IDO1-high in the cancer types where that combination reads as bad news. That is the kind of selection the melanoma trial never made.
That reframes the obesity paradox from a statistical curiosity into a targetable mechanism. Body fat is not simply good or bad for the immune fight against a tumor. It sends a hormone whose effect depends on what the local immune cells do with it, and a single enzyme makes the call. If the finding holds, the clinical question stops being whether a patient's weight helps and becomes whether a given tumor is routing its leptin through IDO1, which is something a pathologist can measure.
The work is mechanistic and correlational. It rests on patient-sample associations plus macrophages exposed to leptin in a dish, not on a trial of anyone actually receiving an IDO1 inhibitor. The cancer-type split is a pattern that a prospective study still has to earn. But it is a specific, testable claim about who benefits, which is more than the paradox offered before.
Leptin is a peptide hormone, not a drug, so peptidemodel hosts no card for it. The lever the paper points at sits in the platform's anticancer ↗ and immune ↗ collections, where the peptides being designed to steer tumor immunity live. The fat signal is upstream of all of them.