NEW YORK – Interim results from a small study led by researchers at Baylor College of Medicine indicated that treatment with engineered natural killer T (NKT) cells could be useful for treating neuroblastoma.
Detailed in a paper published this week in Nature Medicine, the study is following three pediatric patients with relapsed or refractory neuroblastoma enrolled in a Phase I dose-escalation trial of autologous NKT cells engineered to co-express a GD2-specific chimeric antigen receptor (CAR) and the cytokine interleukin-15.
The interim results suggest that CAR-NKT treatment can be delivered safely and that the engineered cells localized to the tumors. In the case of one patient, the treatment led to the shrinking of bone metastases. The CAR-NKT platform used in the study has been licensed by Baylor to Kuur Therapeutics for further clinical development.
The study also identified nine different CAR-NKT subpopulations present in the treated patients using single-cell RNA-seq data generated and analyzed by immune profiling firm Immunai. Two Immunai researchers were authors on the paper.
NKT cells are known to be particularly effective at localizing to tumor sites, and the presence of NKT cells at tumor sites has been shown to be correlated with a better prognosis in certain types of cancer," said Leonid Metelitsa, an oncologist at Baylor College of Medicine and senior author on the Nature Medicine study.
NKT cells appear particularly effective at attacking tumor-associated macrophages, which play key roles in tumor proliferation, driving angiogenesis and metastases, Metelitsa said.
This has led some researchers to speculate that these cells could be effective as part of CAR treatments, particularly CAR treatments for solid tumors, where traditional CAR T-cell therapy has had limited success.
Use of NKT cells in CAR therapy has been challenging, however, due to the small number of these cells in the body, Metelitsa said, noting that they are present at levels "many-hundred-fold" lower than T cells.
This has made it challenging to expand NKT cell populations for use in CAR-NKT treatments, he said.
In recent years though, Metelitsa and his colleagues identified a subset of NKT cells with high proliferative potential and developed a process for expanding these populations, allowing them to generate the large numbers of NKT cells needed for CAR-NKT therapy.
The NKT cells used in the study were engineered to express a receptor to GD2, a surface protein commonly expressed in neuroblastoma cells along with IL-15, which has been shown to support NKT survival and promote tumor infiltration and anti-tumor activity. The final CAR-NKT products introduced into the three patients had NKT cell purity of 96.5 percent, 97.2 percent, and 90.4 percent, which according to the authors represented a substantial improvement over the most successful previous clinical CAR-NKT effort where average purity was 54.4 percent across nine patients.
At dose level one (3×106 GD2-CAR-IL-15 NKT cells per m2), the researchers observed no dose-limiting toxicities over four weeks of treatment and saw one patient maintain stable disease status. The second patient with stable disease experienced a Curie score decline from 7 to 5, while the third experienced a Curie score decline from 2 to 1 and saw a significant reduction in bone metastasis. Curie scores measure the amount of metaiodobenzyl-guanidine (MIBG) neuroblastomas accumulate on diagnostic imaging. Cancers with high uptake of MIBG are thought to have poorer outcomes than those that do not accumulate MIBG.
The researchers also found that the nine different NKT subpopulations as defined by the Immunai analysis were present in the different patients at different frequencies.
Particularly notable, Metelitsa said, was that one of the nine subpopulations, dubbed cell cluster three, was enriched for CAR transgene expression. Although he and his colleagues are just at the beginning of exploring the implications of the different clusters, he noted that the initial single-cell characterizations suggest that the clusters will be linked to NKT cell functionality, including their ability to persist in vivo and to resist exhaustion. These features, he said, could be key to further improving such therapies.
"If you can understand how NKT cells can resist exhaustion that would be a major targetable discovery that would lead us to new steps for improving NKT cell therapy," he said.
New York City-based Immunai launched in 2018 and recently closed a $20 million seed funding round. The company's platform profiles immune cells via a combination of single-cell RNA-seq, cellular indexing of transcriptomes and epitopes by sequencing (CITE-seq), and VDJ-seq. The platform uses machine learning to analyze this data, mapping it to different immune cell types and generating profiles of the populations of immune cells present in a sample, and then correlates to outcomes like disease state or response to treatment.
Antonino Montalbano, lead, genomics technologies at Immunai and an author on the Nature Medicine study, said that the Baylor collaboration was one of the company's first projects.
"The idea was to try to understand if there were differences [between the NKT cell populations used to treat the different patients], and then if those differences were reflected on the clinical side," he said. "Is there something we could find before the infusion that would correlate or associate with [patient] response?"
Montalbano said that he and his colleagues were surprised that they were able to identify nine distinct clusters of NKT cells, noting that they had expected the population to be more homogeneous. Like Metelitsa, he said the relationship between the proportion of the different cell clusters and patient response could be a promising area for future inquiries.
He added that the researchers planned in the future to try to explore the relationship between the profiles of the NKT cell populations before they are infused and what they look like after they have been active in different patients.
Metelitsa said the researchers are also working with Immunai to include additional data in their analysis, including CITE-seq, which will provide single-cell protein expression data.