"H3K27M-mutated diffuse midline gliomas are universally lethal central nervous system tumors in children and young adults," observed Sneha Ramakrishna, MD, in Pediatric Hematology-Oncology at Stanford University. In these patients, the lysine 27 residue of histone 3 is replaced by a methionine residue. Gliomas with this mutation are characterized as being particularly aggressive with typically poor clinical outcomes.
At the 2023 AACR Annual Meeting, Ramakrishna presented a poster detailing her research for the treatment of patients with this aggressive malignancy (Abstract 959). "At Stanford, we are treating these diffuse midline glioma patients with GD2-targeting chimeric antigen receptor (CAR)-T cells, with clear evidence of both radiographic and clinical benefit for many of our (10 of 12) patients. The goal of this work is to identify immune populations and signatures that may be affecting durability of response in our patients. These findings will allow us to iterate and improve this therapy with the goal of providing patients with the hope of cure."
Prior Research
In a 2022 Nature article, researchers from Stanford, including Ramakrishna, discussed the initial results obtained for the first-in-human Phase I clinical trial (NCT04196413) evaluating the use of GD2-directed CAR-T cells in four patients with H3K27M-mutated diffuse intrinsic pontine glioma or spinal cord-based diffuse midline glioma (doi: 10.1038/s41586-022-04489-4). In their findings, the researchers noted "clinical and radiographic improvement" in three of the four patients included in the trial.
Interestingly, the plasma and cerebrospinal fluid levels of pro-inflammatory cytokines were elevated. Transcriptomic analyses were performed on approximately 66,000 individual cells isolated from CAR-T products and cerebrospinal fluid. These analyses displayed heterogeneity in routes of administration, as well as inter-participant responses. The researchers felt these preliminary results showed significant promise for this therapeutic strategy in patients with H3K27M-mutant diffuse intrinsic pontine glioma or diffuse midline glioma.
"In our original manuscript," "we also identified CAR T-cell activity and myeloid signatures in cytokines and single-cell RNA sequencing analysis of cells from cerebrospinal fluid," Ramakrishna noted. "The work presented in our poster builds on that finding and provides more insight into the fluctuations of these various immune populations in the context of CAR T-cell infusions. As we develop a better understanding of the immune biology surrounding CAR T-cell activity in our patients, we can take steps to optimize these treatments for our patients."
Study Details
In this study, blood draws and cerebrospinal fluid samples were submitted for analyses to assess CAR T-cell expansion and "cytokine signatures." Additionally, immune cell profile assessments were performed for pre-manufactured CD4/CD8 T-cell-enriched samples and GD2-targeting CAR-T cells using single-cell RNA sequencing. The data obtained were then analyzed with respect to the patients' clinical progress.
"The most important findings of this study include that we are able to capture the cellular compartment serially over time in patients who are giving repeated intracerebroventricular CAR T-cell infusions and that, by evaluating these cells, we identified both Treg (regulatory T cells) and myeloid suppressive populations associated with response," Ramakrishna explained. "Perhaps the most exciting piece of this was that the myeloid immune suppressive populations seemed to emerge in the setting of infusions where patients stopped responding. This suggests an immune switch at the time of progression that may be targetable and modifiable to improve outcomes."
Regarding the clinical relevance for patients with H3K27M-mutated diffuse midline gliomas, Ramakrishna stated, "While these findings are specifically in patients who have been given CAR-T cells, we have identified similar suppressive myeloid populations at baseline of patients. If we can understand the contribution of these immune suppressive populations, we may be able to alter the tumor microenvironment to allow treatments to be more effective for these patients," she elaborated.
Commenting on the current state of their research, Ramakrishna stated, "We have now collected 169 samples (the data represented at the AACR was for 91 of those samples) and are embarking on integrating and analyzing these data in the context of patient response. Future directions include analysis of completed single-cell RNA sequencing dataset for dose level 1 and 2 patients, including T-cell clonal tracing and myeloid trajectory analysis over time and across treatment infusions.
"Key findings from these data will allow for iterative improvement in CAR-T therapies for patients with H3K27M-positive diffuse intrinsic pontine glioma or diffuse midline glioma, providing hope to shift the paradigm of this fatal disease," Ramakrishna concluded.
Richard Simoneaux is a contributing writer.