September 17, 2025 | Ovarian cancer relapse is a remarkably diverse and dynamic immunological event that when mapped in conjunction with the genomic characteristic of tumors can provide actionable insights about how to enhance the effectiveness of chemotherapy in extending remission and prolonging survival. That’s the promising possibility suggested by researchers in the field who have newly defined four immunologic subtypes of recurrent ovarian cancers.
Among their recent findings is that the immunogenic landscape of original cancers persists in recurrent tumors in predictable and targetable ways (Cancer Cell, DOI: 10.1016/j.ccell.2025.07.005). “Combining immune profiling and genomic features like BRCA status we can identify the patients who may benefit from immunotherapy and tailor new strategies for this immunotherapy-resistant tumor microenvironment,” according to study lead Denarda Dangaj Laniti, Ph.D., who heads the Tumor Microenvironment and Biomarker Discovery lab at the Department of Oncology of the Lausanne University Hospital and the Lausanne Branch of the Ludwig Institute for Cancer Research (Ludwig Lausanne, Switzerland).
Specifically, the research team was able to successfully treat mice with BRCA1-mutated tumors with a therapeutic combination of poly (ADP-ribose) polymerases (PARP) and cyclooxygenase (COX) inhibitors and an immune checkpoint blockade, she reports. They also demonstrated how recurrent homologous recombination proficient tumors having few if any CD8 + T lymphocytes, the primary killers of cancer cells, can be made more susceptible to immunotherapy by targeting a population of the tumor microenvironment called tumor associated macrophages.
When ovarian tumors recur, response to subsequent therapy tends to be dismal because differences between cancer patients are not fully understood, says Dangaj Laniti. But with additional vetting in clinical studies, the simple new immune classification tool combined with genomic biomarkers could enable a more personalized approach to improve survival outcomes for women with ovarian cancer.
In developing the classification system, researchers used digital pathology to analyze 697 ovarian cancer samples from 595 patients in five independent clinical cohorts which included patient-matched treatment-naïve and recurrent tumors. Researchers built an algorithm to capture the heterogeneity of CD8+ T cell densities and their spatial distribution in whole tissue slides as well as in tissue microarrays.
This immune classifier assigns tumors to one of four immunologic buckets. Tumors were categorized as “purely inflamed” if their core and the stromal area of their microenvironment were extensively and homogeneously infiltrated with the T cells, Dangaj Laniti says. In the “mixed-inflamed” category were tumors with hot spots of high inflammation but overall had a highly heterogenous assortment of CD8 + T cells. Tumors were deemed “excluded” if they had T cells only on their border, and “desert” in the small percentage of cases where CD8 + T cells were “low or completely absent.”
To better understand the mechanistic reasons for these immune patterns of cancer recurrence, researchers also used DNA sequencing to identify specific mutations. These notably included BRCA1. In addition to demonstrating that BRCA1-mutated tumors maintain their inflamed phenotype at recurrence, the study showed that they also displayed a network comprised of myeloid and T cells that sustain antigen presentation.
Patients having purely inflamed and mixed-inflamed tumors tended to survive significantly longer than those with excluded and desert tumors, Dangaj Laniti says. Better overall survival following chemotherapy was most strongly associated with tumors whose DNA repair machinery was disabled (loss of homologous recombination proficiency)—by a BRCA1 mutation, for example—since they were also more likely to be inflamed due to the genomic instability. The survival outlook was less rosy for patients having recurrent tumors with mixed, excluded or desert immune subtypes, as well as those that were proficient in DNA repair.
As Dangaj Laniti and her team have shown in previous studies on BRCA1-mutated tumors, cancer cells also carry a lot of damage in their DNA (Cell Reports, DOI: 10.1016/j.celrep.2021.109412). “This can be inflammatory itself and could partly justify why, when these tumors reassemble themselves and their genomic instability patterns, their immune landscape is also rendered to be inflamed,” she says.
The main driver of the study was to better understand how immune inflammation and the immune microenvironment evolves from the primary stage to the recurrent stage in patients with returning ovarian cancer, says Dangaj Laniti. “In between those stages are the standard-of-care treatments, which are chemotherapy and surgery.”
Despite decades-long knowledge that ovarian cancer has endogenous immunity to control tumor growth, the workarounds it develops for evading those natural defenses remain understudied, she continues. “Most patients with ovarian cancer do not consistently respond to immunotherapy, at least immunotherapy in the form of checkpoint blockade.”
Even for scientists, it is a “big paradox” why immune cells can extensively infiltrate ovarian tumors, making them immunologically hot, yet not eliminate the cancer entirely, says Dangaj Laniti. Getting at answers meant looking beyond just the tumor-infiltrating lymphocytes. So, in their latest study, she and her team turned their attention to the role of the immune system’s myeloid cells, including dendritic cells and macrophages, in determining how resistant the tumor microenvironment is to anti-tumor immune responses.
Among the discoveries was that tumors proficient in DNA repair (desert variety) produce proteins drawing a significant number of macrophages into their microenvironment where they suppress anti-tumor immunity, Dangaj Laniti explains. Macrophages have the potential to promote as well as inhibit tumor growth, she says, and in their unhelpful role here they express the ApoE and Trem2 proteins involved in lipid metabolism.
Dendritic cells, meanwhile, work cooperatively with tumor-infiltrating lymphocytes in purely inflamed tumors with DNA repair deficiency to enable effective anti-tumor immunity and immunotherapy in both ovarian cancer and melanoma, as previously demonstrated by Ludwig Lausanne investigators. They now additionally know that these niches are recapitulated upon disease recurrence and draw in anti-tumor macrophages to assist in the fight, says Dangaj Laniti.
In cases where the BRCA1 gene is normal and healthy, and the DNA repair machinery is functioning properly, the study authors report that tumors were more likely to transition into the excluded or desert immune phenotype at recurrence. They also uncovered tumor-associated macrophages as additional components of the tumor microenvironment that help turn away or tune-off T cells. So, they tested an antibody inhibitor of Trem2 in mouse models to show that it could synergize with chemotherapy for tumor types that tend to be excluded at recurrence.
BRCA1-mutated tumors are inflamed and have more T cells and dendritic cells at cancer recurrence. However, in response to standard-of-care chemotherapy and PARP inhibition, they increase the production of prostaglandin E2, a lipid mediator produced by COX enzymes that functionally disables and induces the “suicide” of T cells in the tumor microenvironment via a specialized form of death called ferroptosis, Dangaj Laniti says.
Researchers exploited this vulnerability and were able to show that COX pathway can be blocked with an existing COX inhibitor called Celecoxib that acts in a similar way as common aspirin, along with the standard-of-care PARP inhibitor drug olaparid and chemotherapy, to improve the survival of mice with inflamed, DNA-repair-deficient tumors, she adds. The addition of checkpoint blockade immunotherapy doubled that survival time by activating anti-tumor lymphocytes, including the CD8 + variety.
The current reality is that ovarian cancer does not have a good response rate to immunotherapy, but there has also been no biomarker available to guide upfront patient selection, Dangaj Laniti notes.
The latest study provides a paradigm for “biomarker identification” of these tumor types and for discovery of their vulnerabilities and susceptibility to different forms of immunotherapy.
Dangaj Laniti and her colleagues hope to implement a biomarker-driven approach for clinical practice as well as identify new combinations of therapy based on this patient stratification schema. This will require collaborations with clinical translation partners from academic consortia and cancer centers as well as industry collaborators to aid in the design of prospective studies using the immune classifier and substantial funding, she says.