Efficacy of intrathecal delivery of peptide-pulsed type 1 conventional dendritic cell vaccine for the treatment of breast cancer-associated leptomeningeal disease in preclinical models.

2030 Background: Approximately 5% of advanced stage breast cancer (BC) will develop leptomeningeal disease, a rare form of metastasis in the meninges, a membrane surrounding the brain and spinal cord. The prognosis for BC-LMD is dismal and currently available treatments are ineffective. For example, intrathecal (IT) therapy such as trastuzumab, and proton cranial spinal radiotherapy have only showed modest success. Dendritic cells are a master regulator of the immune system and play a critical role in bridging the innate and adaptive immune response. Here, we used a novel preclinical model to administer MHC class II peptide-pulsed conventional Type 1 dendritic cells (cDC1s) intrathecally in mice that harbor either HER2+ or triple negative breast cancer (TBNC) LMD. Our murine BC-LMD murine data lead us to open a Phase 1 clinical trial of cDC1s in HER2+ and TNBC LMD (NCT05809752; in progress). Methods: Our lab developed a “murine Ommaya” which mimics the clinically used Ommaya reservoir and allows for repeated intrathecal (IT) injections of novel drugs at 3 – 7µl volume directly into the CSF. We next developed a pipeline for screening immunogenic MHC class II peptides from tumor-associated oncodrivers and using these targets to generate tumor-targeting cDC1s. We tested the resulting vaccines in a preclinical model that IT delivers peptide-pulsed cDC1 cell therapy directly into the BC-LMD microenvironment. Results: We tested the efficacy of treatment in HER2+ LMD and TNBC-LMD models and found effective responses; LMD mice that received IT cDC1 therapy exhibited reduced tumor burden and prolonged survival that was significantly better than systemic therapy. Approximately 70% of mice from HER2+ LMD and 30% from TNBC LMD demonstrated complete responses. This was CD4+ T cell dependent. Notably, cured mice also showed signs of resistance against LMD recurrence upon rechallenge. Subsequent single cell RNA-seq (scRNA-seq) analyses of the CSF showed a shift in the innate immune landscape found in untreated LMD to an adaptive immune landscape. In addition, we noted the elevated secretion of Th1 proinflammatory cytokines such as IFN-γ in the CSF. This was recapitulated in the CSF samples of patients enrolled in an on-going phase 1 trial, where early data showed IT cDC1 triggered a high concentration of several proinflammatory cytokines. Conclusions: Our preclinical data suggest IT cDC1 vaccine is effective against BC-LMD and prevents LMD recurrence. We are currently investigating the mechanism(s) by which IT cDC1-initiates the CD4 Th1 adaptive immune response against LMD. By employing an scRNA-seq approach, we hope to further determine the roles of different immune cells and their subtypes in the CSF to IT cDC1s in mice and in patients. A future approach includes testing IT cDC1 platform in LMD from other cancers (e.g. melanoma etc.).